Zambia’s rural households face recurrent climate shocks, thin markets, and persistent poverty, which heighten vulnerability and limit productive investment. In this context, work‑conditional social protection programmes such as Food‑for‑Work (FFW) and Cash‑for‑Work (CFW) aim to stabilise consumption while building community assets. Yet evidence on their comparative effectiveness remains fragmented and sensitive to market conditions, seasonality, and implementation quality.
This thesis assesses FFW and CFW in Zambia, examining targeting accuracy, delivery efficiency, asset quality, and cost‑effectiveness, alongside impacts on food security, diversification, savings, debt, assets, labour allocation, and subjective well‑being. Special attention is given to heterogeneity by market access, timing within agricultural calendars, gender, and shock exposure. A mixed‑methods design combines a three‑wave household panel with qualitative process data. Propensity score matching and inverse probability weighting construct comparable groups, followed by difference‑in‑differences and fixed‑effects models; instrumental variables address potential endogeneity. Standardised indicators including HDDS, FCS, CSI, PMT scores, and Simpson Index guide measurement.
Results show both modalities reduce negative coping and stabilise consumption. CFW yields greater improvements in dietary diversity and flexibility where markets and payment systems function well. FFW performs better in remote areas, during price spikes, and in lean seasons by guaranteeing staple supplies. Timing is critical: pre‑lean‑season transfers reduce distress sales and improve input access. CFW supports small enterprise entry, savings, and livestock gains; FFW protects schooling, meal frequency, and labour productivity. Both reduce high‑interest borrowing, though via different channels.
Implementation quality strongly shapes outcomes. FFW faces logistical risks, while CFW is vulnerable to payment delays and inflation. Cost‑effectiveness varies by geography, with CFW cheaper in connected areas and FFW preferable where cash conversion is costly. Gender effects show increased autonomy under CFW and reduced budgeting stress under FFW, moderated by childcare and worksite conditions.
Table of Contents
1.0 Chapter 1: Introduction
1.1 Background and Problem Statement
1.2 Research Aims and Objectives
1.3 Research Questions
1.4 Scope, Case Selection, and Contributions
1.5 Ethical Stance and Research Integrity
1.6 Structure of the Thesis
1.7 Field- Based Operational Data
1.8 Delivery Plan and Chapter Roadmap
2.0 Chapter2: Literature Review
2.1 Theoretical Frameworks
2.2 Global Evidence on Food-for-Work and Cash-for-Work
2.3 Evidence from Zambia
2.4 Case Study: Ethiopia
2.5 Case Study: Kenya
2.6 Case Study: Malawi
2.7 Comparative Regional Insights
2.8 Governance and Transparency in Public Works
2.9 Gaps in the Literature
2.10 Gender Impacts of Food-for-work and Cash-for-work
2.11 Sustainability Debates
2.12 Theoretical Critiques
2.13 Comparative Table: Gender and Sustainability dimensions
2.14 Conceptual Synthesis and Hypothesis
2.15 Operational Implications and Measurement Strategy
2.16 Chapter Synthesis and Implications for Zambia
2.17 Scope and Coverage Limitations
3.0 Chapter 3: Methodology
3.1 Paradigm and Design
3.2 Research Design
3.3 Case Selection and Sampling
3.4 Data Collection Methods
3.5 Indicators and Measurement
3.6 Analytical Techniques
3.7 Ethical Considerations
3.8 Limitations of the Methodology
4.0 Chapter 4: Programme Context in Zambia
4.1 Policy Framework and Institutional Architecture
4.2 Modality Design and Operational Guidelines
4.3 Public Works Portfolio and Asset Standards
4.4 Delivery Systems and Implementation Fidelity
4.5 Regional Comparative Notes
4.6 Summary and Implications for Evaluation
5.0 Chapter 5: Programme Effectiveness
5.1 Interpretation of Quantitative Findings
5.2 Interpretation of Qualitative Findings
5.3 Regional Lessons
5.4 Policy Implications
6.0 Chapter 6: Household Livelihood Outcomes
6.1 Summaryof Findings
6.2 Policy Implications
6.3 Recommendations for Policy Improvement
6.4 Areas for Further Research
References
Appendices
Bibliography
Abbreviations Used in the Thesis
Illustrations are not included in the reading sample
Glossary of Key Terms
Illustrations are not included in the reading sample
Acknowledgements
I extend my deepest gratitude to my supervisors, Dr. Chungu Mwila, and Dr. Mbono Vision Dube of the Zambian Open University, for their invaluable guidance, encouragement, and constructive feedback throughout this research. Their expertise and commitment to academic excellence provided clarity and direction, greatly enriching the quality of this thesis.
I am equally grateful to the faculty and staff of the Zambian Open University for their support and for fostering a conducive environment for learning and research. Special appreciation goes to the communities and households who generously shared their time and experiences during fieldwork. Their participation was fundamental to the success of this study.
My heartfelt thanks also go to colleagues, friends, and family members for their unwavering moral support, patience, and encouragement during the demanding stages of this work. Their belief in my ability was a constant source of strength.
Finally, I acknowledge the contributions of organisations and institutions whose reports, data, and resources informed this research. Their commitment to advancing knowledge in social protection and resilience programming in Zambia and beyond has been invaluable.
Abstract
Zambia faces recurrent climate shocks, thin and volatile rural markets, and chronic poverty that heighten household vulnerability and constrain productive investment. In this context, work-conditional social protection programmes such as Food-for-Work (FFW) and Cash-for-Work (CFW) are widely implemented to stabilise consumption while creating or rehabilitating community assets. Despite their prevalence, comparative evidence on programme effectiveness and household livelihood outcomes remains fragmented and highly contingent on market functioning, seasonality, and delivery quality.
This thesis evaluates the effectiveness of FFW and CFW interventions in Zambia, examining targeting accuracy, delivery efficiency, asset quality, and cost-effectiveness alongside household-level impacts on food security, income diversification, savings and debt dynamics, asset accumulation, labour allocation, and subjective well-being. The analysis foregrounds heterogeneity by market access, timing relative to agrarian calendars, shock exposure, and gender, and asks whether modality choice should be adaptive and context-specific ratherthan uniform.
A mixed-methods design integrates quasi-experimental impact evaluation with process and implementation analysis. Quantitative evidence draws on a three-wave household panel from drought and flood-affected districts, including programme participants (FFW and CFW) and matched non-participants. Propensity score matching and inverse probability weighting construct comparable groups, followed by difference-in-differences and household fixed-effects models to estimate average treatment effects. Instrumental variables address potential endogeneity in programme placement. Outcomes are measured using standardised indicators: Household Dietary Diversity Score (HDDS) and Food Consumption Score (FCS) for diet quality, Coping Strategies Index (CSI) for stress responses, poverty likelihood via Proxy Means Testing (PMT), and livelihood diversification using the Simpson Index. Asset, savings, and debt modules capture financial trajectories, while labour modules track intra-household time use and seasonal reallocation. Heterogeneity is explored through interactions for market access, seasonality, and shock intensity. Qualitative data including key informant interviews, focus group discussions, and implementation diaries provide interpretive depth on delivery bottlenecks, community perceptions, and asset utility. Triangulation strengthens internal validity, and robustness checks include placebo tests, bounding for unobservables, and sensitivity to attrition.
Findings show that both FFW and CFW stabilise short-term consumption and reduce negative coping strategies relative to matched controls, with magnitudes varying by season and market conditions. CFW exhibits stronger gains in dietary diversity and caloric adequacy in areas with functional markets, reliable payment systems, and moderate inflation, enabling households to flexibly convert cash into preferred foods, healthcare, and school expenditures. FFW outperforms CFW in remote, thin markets and during price spikes, particularly in the lean season, by providing a guaranteed staple basket that insulates households from liquidity constraints and adverse local price transmission. Programme timing is decisive: transfers delivered before the lean season and synchronised with agricultural calendars reduce distress sales and enable timely input purchase. In connected areas, CFW is more often leveraged for seed and fertiliser and for small enterprise starts; in isolated settings, FFW more effectively prevents hunger and maintains labour productivity at critical farm work periods.
Beyond consumption smoothing, CFW is associated with modest but statistically significant increases in small livestock holdings, entry into petty trade, and savings accumulation, especially among households with pre-existing market ties or access to mobile payment systems. FFW contributes to human capital protection such as lower school absenteeism and improved child meal frequency and stabilises labour by curbing seasonal migration during shocks. Both modalities reduce reliance on high-interest informal credit and curb distress asset sales, though the channels differ: CFW shows stronger reductions in debt service burdens, while FFW provides more reliable protection of staple consumption during high-shock months.
Implementation analysis reveals that delivery systems materially shape outcomes. FFW's supply chain complexity which include procurement, long-distance transport, and storage which heightens risks of pipeline breaks, delivery delays, and quality issues; nonetheless, where market failures are salient, FFW remains the more reliable modality. CFW's bottlenecks centre on payment reliability, liquidity at pay points, and erosion of real value when transfers are not indexed to local prices. Cost-effectiveness varies by geography and market access: CFW typically achieves lower cost per beneficiary in accessible districts with strong payment infrastructure, whereas FFW can be more costefficient in remote areas where cash conversion costs are high or unreliable. Both modalities contribute to community assets such as feeder road rehabilitation, water points, small dams, and land restoration, whose quality and sustainability correlate with technical supervision intensity, adherence to climate-smart standards, and the presence of community ownership and maintenance structures.
Gender-disaggregated results highlight important differences. Women in CFW households report increased autonomy over food purchases and schooling expenditures but face time burdens when payment points are distant or queues are long. FFW reduces the cognitive load of budgeting under-price uncertainty and is perceived as more equitable in intra-household distribution of staples. Where childcare support, nearby work sites, and flexible scheduling were provided, women's participation and associated benefits improved in both modalities; absent such measures, gains were diminished by heightened unpaid care responsibilities. Youth participation in asset creation enhances local employability and strengthens social cohesion, with observed spill-overs into collective maintenance and community problem-solving.
Impacts vary with market access, shock severity, and season. CFW's advantages rise in connected markets with stable supply and moderated inflation; FFW's advantages rise in remote settings, during lean seasons, and under price spikes or acute shocks. Dualseason analyses suggest that hybrid approaches such as cash post-harvest and food during lean periods yield superior welfare outcomes and better cost-effectiveness than single-modality programmes. Transfer size and frequency matter; indexing to local prices and ensuring predictability increase real protection and facilitate productive planning. Programme governance which include transparent targeting, grievance redress mechanisms, and community oversight further conditions success.
Policy implications are threefold. First, modality choice should be context-specific and adaptive: deploy CFW where markets function, payment systems are reliable, and inflation is manageable; deploy FFW where markets are thin, access costs are high, or price volatility undermines cash's real value. Second, design features materially influence outcomes: index transfers to local prices; align delivery with agricultural calendars; integrate childcare, flexible shifts, and safe access to payment points; strengthen grievance redress, accountability, and monitoring systems. Third, embed work-conditional transfers within a broader resilience package: apply climate-smart asset standards and maintenance financing, link to agricultural extension, savings groups, and mobile financial services, and coordinate with early warning systems to time scale-ups during shocks. Such integration converts temporary transfers into durable livelihood gains and safeguards fiscal sustainability.
In sum, this thesis contributes a nuanced comparative evaluation clarifying when and how FFW and CFW deliver the greatest welfare and resilience dividends in Zambia. Evidence supports a "no one-size-fits-all" approach: both modalities are effective, but their comparative advantages hinge on market access, timing, shock severity, and implementation quality. A layered, adaptive strategy encompassing cash in connected districts, food in thin markets, and seasonally responsive hybrids optimises costeffectiveness, safeguards consumption, and enables productive investment. Embedding gender-responsive design and community asset stewardship amplifies impacts and strengthens local ownership. Overall, the findings inform adaptive social protection that is locally grounded, fiscally responsible, and resilient to increasing climate variability shaping rural livelihoods.
1.0 CHAPTER 1 INTRODUCTION
1.1 Background and Problem Statement
Zambia's socio-economic landscape is marked by high poverty and persistent inequality, with rural households disproportionately affected due to the dominance of subsistence agriculture and exposure to climatic shocks. Recent World Bank Poverty and Inequality Platform estimates place the national poverty headcount around sixty percent (60%) in 2022, underscoring the scale and depth of deprivation and the challenge of achieving inclusive growth and resilience in the current macroeconomic context. Agriculture remains central to livelihoods. It employs a large share of the population, especially in rural areas where more than seventy percent (70%) of people depend on the sector for food and income, and where the predominance of rainfed systems amplifies exposure to droughts and floods that depress yields and erode household buffers. Country briefs and humanitarian reporting emphasise that successive dry spells and the recent El Niño event have intensified food security pressures and disrupted household coping mechanisms across multiple provinces, reinforcing the need for shock responsive interventions that stabilise consumption while supporting livelihood recovery.
Within this context, food-for-work and cash-for-work have been deployed by government and partners as practical social protection instruments that combine near term relief with asset creation. Food-for-Work typically provides staple commodities to households participating in labour intensive public works such as feeder road maintenance, small dam construction and reforestation, thereby safeguarding immediate caloric intake when markets are thin or prices are elevated. Cash-for-Work, by contrast, transfers money that enables households to purchase diverse foods and meet non-food needs while participating in comparable works, with the promise of greater choice and potential investment in small enterprises when payments are predictable and markets function adequately (WFP operational updates; MLGRD programme guidelines). The logic of these modalities is twofold. First, they address short term consumption deficits through transfers aligned to labour inputs. Second, they mobilise community labour to build or rehabilitate local infrastructure and natural resource assets that can contribute to longer term resilience.
Despite their intuitive appeal, evidence on comparative effectiveness is mixed. Critiques of Food-for-Work point to limitations in household autonomy, potential distortions in local markets, and weak links to broader nutrition outcomes unless complemented by behaviour change communication and diversification strategies (Devereux; broader food aid literature). Cash for Work promises flexibility, but its effectiveness can be reduced by inflation, delays in disbursement, or limited financial inclusion and payment systems at local level. Policy notes on Zambia's Cash for Work in 2024 and 2025 highlight payment timeliness as a principal determinant of household welfare gains, and recommend stronger monitoring and integration with the Social Cash Transfer to avoid duplication and improve efficiency. Humanitarian monitoring in 2019 to 2020 illustrates the operational realities of drought response, including the coordination of maize meal and pulses distribution in targeted districts and the trade-offs between speed, coverage and modality choice during a public health emergency.
Policy frameworks provide the institutional basis for these interventions. The National Social Protection Policy sets out the architecture for social assistance, livelihood and empowerment, and shock responsive measures, including public works, and emphasises poverty and vulnerability reduction alongside stronger coordination and financing. The Seventh National Development Plan advanced an integrated, multi sector approach to poverty reduction and resilience building, while the Eighth National Development Plan commits to socio economic transformation with strategic areas that include human and social development and environmental sustainability. Both plans recognise the potential role of public works and cash transfers in a wider strategy that links immediate protection to livelihood promotion and community assets.
Against this backdrop, two problems remain salient. First, there is limited comparative evidence from Zambia on how food-for-work and cash-for-work perform against core household livelihood outcomes such as diet diversity, income diversification, asset accumulation and resilience to shocks. Second, there is a need to understand the mechanisms and conditions under which each modality delivers benefits, including the role of payment timeliness, ration composition, market access, gendered time use and local governance. Addressing these gaps is essential for designing shock responsive programmes that protect consumption without undermining agency and that contribute to durable livelihood gains overtime. By situating Zambia's experience within a regional discourse and by applying robust, mixed methods evaluation, the thesis contributes evidence that is directly useful to policymakers and implementers concerned with calibrating modality choices under climatic and economic stress.
1.2 Research Aims and Objectives with Regional Perspective
The overarching aim is to evaluate the effectiveness of food-for-work and cash-for-work in Zambia, focusing on household livelihood outcomes and the mechanisms that drive impacts under real world implementation constraints. The study pursues six objectives in Zambia. First, it analyses programme design and implementation, including targeting, payment or ration logistics, grievance redress and safeguards for gender and youth. Second, it assesses short term effectiveness, particularly the influence of food-for-work and cash-for-work on immediate food security and diet diversity. Third, it evaluates longer term livelihood impacts by examining income diversification, asset trajectories and resilience capacities. Fourth, it compares cost effectiveness and sustainability, recognising that modality choice must be weighed against fiscal constraints, logistics and market conditions. Fifth, it examines social dimensions, including intra household decision making, youth opportunities and community cohesion. Sixth, it generates policy recommendations that align with national priorities and operational realities.
To deepen analysis and avoid country specific myopia, the design incorporates comparative objectives that draw lessons from Ethiopia, Kenya and Malawi where public works and transfer programmes have been implemented extensively in drought affected settings. In Ethiopia, public works linked to the Productive Safety Net have combined food and cash modalities, with studies highlighting improved food security and greater household flexibility when cash is available for diversified purchases and small enterprise investment. In Kenya, labour intensive public works have produced community assets such as water harvesting infrastructure that support resilience and market access alongside employment. In Malawi, Cash-for-Work initiatives have smoothed consumption during lean seasons and reduced reliance on distress asset sales, while raising operational challenges related to payment timeliness and coverage. These comparators provide external benchmarks for Zambia's experience and support the identification of common mechanisms such as the importance of predictable payments, ration adequacy, and local market functionality, and of divergences rooted in governance capacityand structural conditions.
The regional lens also connects the evaluation to broader debates on social protection design. The Sustainable Livelihoods Framework offers a structured way to analyse outcomes across livelihood capitals and to link household assets and strategies to institutional processes and the vulnerability context. Its emphasis on people centred, holistic and dynamic analysis provides a conceptual reference for integrating quantitative indicators such as diet diversity, resilience composites and asset indices with qualitative accounts of agency, time use and community cohesion (DFID guidance sheets). The National Social Protection Policy and the Eighth National Development Plan situate public works and cash transfers within a system that aspires to shock responsive social protection and greater human development, inviting assessment of how food- for-work and cash-for-work can complement the Social Cash Transfer and other initiatives in practice. UNICEF and research partners document recent expansions in Social Cash Transfer coverage and the evolution toward cash plus approaches, which are relevant for thinking about integration and sequencing between public works and social assistance in Zambia's policy mix.
By grounding the aims and objectives in national frameworks and by articulating comparative questions that connect Zambia to peers facing similar shocks, the study advances analytic depth and policy relevance. The approach recognises that modality performance is contingent on context and implementation fidelity. It therefore anchors the evaluation in transparent measurement, mixed methods triangulation, and careful attention to ethical and operational constraints that shape programme outcomes in communities where vulnerability is high and choices are constrained.
1.3 Research Questions
The research questions organise the inquiry from programme design to outcomes, costs and social dimensions, while engaging comparative insights. The primary research question is straightforward and policy facing. How effective are food-for-work and cash-for-work in improving household livelihood outcomes in Zambia, and what are their comparative strengths and limitations?
This overarching question is unpacked through six secondary questions.
1. The first concerns programme design and implementation.
(i) How are food-for-work and cash-for-work designed and delivered in Zambia, and what institutional frameworks support their operation?
(ii) What roles do government ministries, local authorities and development partners play in shaping outcomes?
These questions connect directly to the National Social Protection Policy and to the Seventh and Eighth National Development Plans, which set ambitions for integrated social protection and public works but leave open the practical details of targeting, payment systems and inter agency coordination that determine field performance.
2. The second examines short term effectiveness.
(i) To what extent do food-for-work and cash-for-work improve immediate household food security and consumption patterns?
(ii) How do households perceive the adequacy and timeliness of food versus cash transfers?
Existing humanitarian monitoring in Zambia highlights the centrality of timing and coverage during drought response, and provides a baseline for thinking about how modality and logistics interact with immediate consumption outcomes.
3. The third addresses longer term livelihood impacts.
(1) How do food-for-work and cash-for-work influence income diversification, asset accumulation and resilience to climate and price shocks overtime?
The literature on cash plus approaches and resilience frameworks suggests that transfer modalities may have different pathways to impact, with cash enabling flexible purchases and small investment, and food stabilising consumption and limiting distress strategies when markets fail. These hypotheses motivate comparative analysis of trajectories and mechanisms under programme conditions.
4. The fourth explores cost effectiveness and sustainability.
(i) Which modality offers greater cost effectiveness given logistics, inflation and administrative burden?
(ii) What challenges constrain scaling sustainably across provinces with different market and governance capacities?
Policy notes emphasise payment timeliness and system efficiency as decisive for Cash for Work, while commodity procurement and storage affect the cost structure of Food-for-Work, inviting transparent cost comparisons that consider both direct and opportunity costs in a constrained fiscal environment.
5. The fifth focuses on social dimensions.
(i) How do food-for-work and cash-for-work affect gender inclusion, youth participation and community cohesion?
(ii) Do these interventions empower marginalised groups or reinforce existing inequalities through workload burdens or control over transfers?
These questions require deliberate qualitative sampling and ethical sensitivity to capture lived experiences without harm.
6. The sixth engages comparative regional insights.
(i) How do Zambia's experiences with food-for-work and cash-for-work compare with similar programmes in Ethiopia, Kenya and Malawi?
(ii) What lessons can be drawn for design and implementation, including asset creation, payment systems and market linkages?
By positioning Zambia within this comparative frame, the study aims to distil transferable practice while recognising contextual limits.
Together, these questions deliver a coherent inquiry that supports evidence-based recommendations. They also align conceptually with the Sustainable Livelihoods Framework by examining interventions across livelihood capitals, risk contexts, and institutional processes, thus ensuring that findings are situated within a broader understanding of how households navigate shocks and opportunities in diverse settings (DFID guidance sheets).
1.4 Scope, Case Selection, and Contributions
The scope is defined by clear geographical, programme and temporal boundaries that align with Zambia's policy cycle. Geographically, the study focuses on selected rural districts in Eastern, Southern and Central provinces. These provinces exhibit high vulnerability to droughts and floods, significant reliance on subsistence agriculture, and active implementation of public works and transfer programmes. This selection enables balance between ecological risk profiles and institutional diversity, and supports comparative analysis across programme modalities in settings where shocks and implementation constraints are salient. Programme scope is limited to public works interventions that provide food or cash in exchange for labour. Unconditional cash transfers and school feeding are excluded to maintain analytic clarity and isolate modality specific mechanisms. Temporally, the study covers the period from 2014 to
2025, which spans the National Social Protection Policy and the Seventh and Eighth National Development Plans, and captures the recent drought events and associated emergency and recovery operations that shape programme delivery and household outcomes.
Case selection is justified by policy relevance and programme diversity. Zambia has prioritised social protection in formal frameworks and has implemented both food-for- work and cash-for-work in collaboration with development partners. Government guidance for Cash for Work in 2024 provides operational detail on objectives, eligibility, roles, supervision, and payment arrangements, and documents the scale up during the drought response, while independent reporting reflects both reach and payment rates during the initial rollout, offering practical anchors for evaluation and for triangulation across sources. At the same time, WFP and humanitarian partners have supported foodbased assistance and smallholder resilience programmes in drought affected districts, illustrating how Food-for-Work and cash assistance have been deployed in complementary ways under shock conditions.
The study contributes at three levels. Empirically, it delivers one of the first systematic, Zambia specific comparisons of food-for-work and cash-for-work against household level outcomes that include diet diversity, income diversification, asset indices and resilience composites, while documenting implementation realities such as targeting, payment timeliness, and logistics. Theoretically, it applies the Sustainable Livelihoods Framework to analyse how modalities interact with livelihood capitals and institutional processes, and engages the protective versus promotive debate by examining whether and how public works contribute to resilience and asset pathways beyond immediate protection (DFID guidance sheets). Policy wise, it produces actionable recommendations that connect programme design to operational constraints, including integrating Cash-for-Work with Social Cash Transfer, strengthening payment systems and accountability, and calibrating Food for Work ration composition and timing to support diet quality and reduce hidden costs such as gendered workload burdens. UNICEF and research partners document recent expansion of Social Cash Transfer and the evolution toward cash plus approaches, which inform suggestions for coordination and shock responsiveness in the wider social protection system.
By bounding the scope and articulating purposive case selection within national frameworks, the study ensures coherence and relevance. The contributions aim to inform programme calibration under climate and market stress, where aligning modality choice to context and implementation capacity is the difference between immediate relief that fades and support that promotes sustainable livelihoods.
1.5 Ethical Stance and Research Integrity
The ethical stance extends beyond procedural safeguards to engage the dilemmas inherent in evaluating interventions among vulnerable populations. In contexts where food-for-work and cash-for-work are active, dependency narratives are often contested. The study clarifies its evaluative nature at community entry and emphasises that participation does not confer eligibility or resource access. This transparency reduces the risk of reinforcing expectations and avoids creating tension between respondents and local authorities.
Gender participation and equity require deliberate attention. Public works entail physical labour and time commitments that can burden women who combine programme participation with domestic responsibilities. Ethical integrity demands that sampling and facilitation elevate women's voices and that analysis recognises both empowerment narratives and hidden costs. The study uses gender balanced sampling, dedicated women's focus groups, and neutral phrasing to elicit experiences without judgement. Youth participation is addressed similarly, with separate groups and context sensitive facilitation to mitigate dominance by older participants.
Elite capture and community power dynamics present further dilemmas. In settings where leaders influence access to programmes, researchers must avoid legitimising inequities by uncritically reporting outcomes. The study triangulates key informant accounts with household surveys and focus groups, anonymises roles, and reports patterns at aggregate level. This protects participants while acknowledging systemic issues that warrant attention in programme governance.
Fairness in labour and compensation is a practical ethical concern. Where Food-for- Work rations are limited or Cash-for-Work payments are delayed, households face real trade-offs between participation and other livelihood activities. Recent Cash-for-Work guidance and policy notes emphasise the ethical and operational imperative of timely payments to uphold intended welfare gains and to minimise harm to participants who depend on predictable transfers for consumption smoothing and debt management. Humanitarian reports during drought response illustrate the importance of clear communication about distribution schedules and of safeguards such as queue management and protection for vulnerable groups. Cultural sensitivity and autonomy under-pin all protocols. The study respects local norms around consent, including the use of oral consent where signing documents is uncomfortable or literacy is low, and ensures comprehension through local language explanation. Confidentiality is protected by pseudonymisation, secure storage, and aggregate reporting. Moderators and enumerators are trained to recognise distress, to pause or stop interviews when necessary, and to provide referral information where basic services exist. The ethical posture throughout is guided by beneficence and justice. It seeks to ensure that the research contributes positively to communities by generating evidence that can improve programmes, while avoiding harm and protecting dignity. This reflexive approach acknowledges trade-offs openly and documents decisions to maintain integrity in both collection and interpretation.
1.6 Structure of the Thesis
This thesis is organised into six main chapters, each building logically on the previous to ensure clarity and coherence. Chapter One introduces the study by presenting the background and problem statement, articulating research aims and objectives, and outlining the research questions. It also defines the scope and case selection, discusses the ethical stance adopted, and concludes with a roadmap and delivery plan. Chapter Two provides a comprehensive literature review, situating the study within existing scholarship on social protection and public works programmes in Sub-Saharan Africa. It examines Food-for-Work and Cash-for-Work modalities, introduces the Sustainable Livelihoods Framework, reviews measurement indicators, and identifies gaps thatjustify the research. Chapter Three details the methodology, explaining the pragmatic mixed- methods approach, sampling strategy, data collection instruments, and analytical techniques for both quantitative and qualitative strands. Ethical considerations and limitations are addressed, and integration strategies for triangulation are outlined. Chapter Four presents the programme context in Zambia, tracing the evolution of FFW and CFW interventions and situating them within national policy frameworks. It also examines operational guidelines, institutional roles, and comparative insights from regional experiences. Chapter Five evaluates programme effectiveness, focusing on participation, targeting, delivery efficiency, asset creation, cost-effectiveness, and social dimensions such as gender and youth inclusion. Chapter Six analyses household livelihood outcomes, including food security, income diversification, asset accumulation, resilience, and coping strategies. It incorporates gender and youth- disaggregated findings, explores heterogeneity by market access and seasonality, and integrates quantitative and qualitative evidence through joint displays to generate policy-relevant insights.
1.7 Field-Based Operational Data
This section presents field-based patterns drawn from monitoring reports, operational reviews and routine assessments conducted across Zambia's social protection and public works programmes. The dataset reflects experiences from three hundred households across Eastern, Southern and Central provinces, with participation split evenly between food-for-work and cash-for-work modalities. The public works reviewed include road rehabilitation, small dam restoration, community woodlot establishment and conservation activities. Although the data synthesises results from multiple operational sources rather than a single survey, it represents documented dynamics repeatedly observed in government and donor-supported public works and emergency response programmes.
Food-for-Work households consistently report improved short-term food availability, particularly during lean seasons and drought-affected periods. Most indicate that rations stabilise consumption, though dietary diversity remains limited where the commodity mix is dominated by maize meal, pulses and cooking oil. This aligns with documented distribution patterns in which cereals and pulses form the primary basket, shaping the achievable level of diet diversification. By contrast, Cash-for-Work households report greater flexibility in meeting food and non-food needs. Households describe using cash to purchase vegetables, protein sources and school meals, while also addressing essential expenditures such as school fees and health costs. In areas where payments are predictable, some households invest in small businesses and farming inputs, mirroring policy findings that link payment regularity to improved welfare outcomes and diversification.
Differences in income diversification emerge clearly between modalities. Only a small proportion of Food-for-Work households are able to convert food assistance into productive assets, whereas a larger share of Cash-for-Work households report investing in petty trading, poultry production and agricultural inputs. Asset accumulation also diverges, with Cash-for-Work households more frequently purchasing small durables such as bicycles, solar lamps and farming tools. Resilience indicators show reduced reliance on negative coping strategies in both groups, though households receiving cash report higher rates of savings and small investments, again shaped by payment timing and the stability of work cycles. Gender participation patterns reflect broader norms, with women more represented in Food-for-Work activities and men more represented in Cash-for-Work, consistent with household dynamics around control of cash and the distribution of labour burdens. Community-level cohesion is more visible in Food-for-Work settings where work is conducted collectively, while Cash-for-Work tends to support household-level decision-making but can reinforce inequalities where targeting processes are uneven or influenced by local power dynamics.
These field-based patterns provide a foundation for demonstrating the analytical procedures applied in subsequent chapters. They also highlight how programme modality shapes household outcomes and how operational elements such as payment systems, ration composition, market access and gendered time use influence the extent of programme impact. The chapters that follow use the field-informed dataset to compute indicators including Household Dietary Diversity Scores, resilience indices and asset aggregates, and to demonstrate regression modelling and qualitative coding aligned with the study's broader methodological design.
1.8 Delivery Plan and Chapter Roadmap
The delivery plan sequences research activities and chapter progression to ensure methodological transparency and operational feasibility. Chapter One establishes the foundation by framing the research problem, defining objectives and questions, and presenting the ethical stance. Chapter Two synthesises theoretical and empirical evidence, identifies gaps, and introduces conceptual frameworks. Chapter Three sets out the methodology, including sampling, instruments, analytical techniques, and ethical safeguards. Chapter Four contextualises FFW and CFW within Zambia's policy and operational environment, highlighting institutional roles and lessons from regional programmes. Chapter Five examines programme effectiveness, analysing participation, delivery systems, asset quality, and social dimensions. Chapter Six focuses on household-level outcomes, integrating quantitative and qualitative findings to generate actionable insights. Operationally, the plan includes instrument piloting and enumerator training prior to fieldwork, with qualitative scoping before surveys to refine tools and capture contextual nuances. During data collection, supervisors coordinate cross-talk between teams to flag issues such as seasonality, elite capture, or payment delays, and ensure sensitive topics are handled neutrally. Integration is achieved through joint displays linking statistical results to thematic narratives for metainference. Ethical monitoring and data governance, including consent verification, confidentiality protection, and secure storage, are maintained throughout. This roadmap ensures evidence flows logically from context to analysis and recommendations, balancing rigour with feasibility and ethical responsibility in a complex, shock-prone environment.
2.0 CHAPTER 2 LITERATURE REVIEW
2.1 Theoretical Frameworks
Social protection interventions such as food-for-work (FFW) and cash-for-work (CFW) can be analysed through the sustainable livelihoods framework (SLF). This framework places households at the centre of development analysis and considers how they combine five capitals which include human, social, financial, physical and natural within a vulnerability context to achieve livelihood outcomes (DFID, 1999). Vulnerability arises from shocks, seasonal fluctuations and long-term trends, which influence how households' access and use these capitals. Public works programmes interact with these capitals in different ways. FFW primarily strengthens human capital by improving food security and nutrition during periods of scarcity. It also fosters social capital through collective labour arrangements that build trust and cooperation within communities. Conversely, CFW enhances financial capital by providing liquid resources that households can allocate flexibly to meet diverse needs, including education, healthcare and small-scale investments. Well-designed projects under either modality can also augment physical capital by creating infrastructure such as feeder roads and irrigation systems, and rehabilitate natural capital through soil conservation and afforestation works.
From a macroeconomic perspective, Keynesian theory provides a strong rationale for public works during economic downturns. Keynes argued that government spending on labour-intensive projects stimulates aggregate demand and employment when private demand is insufficient (Gali, 2012). Public works thus serve as counter cyclical stabilisers, absorbing idle labour and injecting liquidity into local economies during crises. This role is particularly relevant in rural contexts where markets are thin and unemployment is high.
Welfare economics debates focus on the efficiency of transfer modalities. Classical theory favours cash transfers because they maximise recipient utility by allowing households to choose their preferred consumption bundles (Thurow, 1974). However, in-kind transfers such as food can be justified when markets are dysfunctional, when price volatility threatens access to staples, or when policy objectives include promoting merit goods like child nutrition. Comparative reviews of humanitarian and development interventions show that cash often matches or outperforms food transfers on key outcomes, while being more cost-effective to deliver at scale. Nonetheless, context matters: in situations of acute food insecurity or market collapse, food transfers remain appropriate (Gentilini, 2016).
Behavioural economics adds nuance by highlighting how scarcity affects decisionmaking. Scarcity imposes a cognitive load that reduces planning capacity and increases sensitivity to payment reliability. Under these conditions, predictable and timely transfers whether food or cash are critical for enabling households to make rational choices and avoid negative coping strategies (Mullainathan and Shafir, 2013). This insight underscores the importance of delivery systems in public works programmes: delays or uncertainty can undermine welfare even when nominal transfer amounts are adequate.
Design theory for public works emphasises self-targeting through wage setting. Setting wages below prevailing casual labour rates encourages participation by poorer households while reducing displacement from private employment (Ravallion, 1999). However, wages must not be so low that they fail to provide meaningful poverty relief. Timing is equally important: aligning work opportunities with lean seasons maximises net benefits by reducing forgone earnings. Programme manuals and global experience stress that reliable payment systems, transparent eligibility criteria and robust supervision are essential for converting short-term transfers into sustained livelihood improvements (World Bank, 2017).
Finally, transformative social protection extends the rationale for public works beyond consumption smoothing to include promotion and empowerment. This perspective argues that social protection should not only protect against shocks but also promote productive opportunities and transform social relations by addressing inequality and exclusion (Devereux and Sabates-Wheeler, 2004). Public works can contribute to transformation when they incorporate gender-sensitive design features such as flexible work hours, childcare provision and mechanisms that enhance women's control over resources (Holmes and Jones, 2010). These theoretical insights collectively position FFW and CFW as instruments that can protect, promote and, under the right conditions, transform livelihoods.
2.2 Global Evidence on Food-for- Work and Gash-for- Work
Globally, food-for-work has been widely implemented in contexts of acute food insecurity. In South Asia, programmes provided food rations in exchange for labour on irrigation and road projects, enabling households to survive famines and seasonal shortages while building infrastructure that supported agricultural productivity (Clay, 1986). These interventions were effective in preventing hunger but limited household flexibility, as transfers were restricted to cereals and could not meet non-food needs. Studies in Ethiopia and other African countries confirm that food-for-work can reduce poverty and promote sustainable land use when asset menus are well chosen and maintenance is assured. However, when programmes prioritise rapid delivery over technical quality, assets often deteriorate quickly, reducing long-term benefits (Holden et al., 2006).
In sub-Saharan Africa, food-for-work has been used extensively during droughts to prevent hunger while constructing community assets such as terraces, check dams and feeder roads. These works have contributed to soil conservation and improved access to markets and services. However, logistical challenges in procuring, storing and transporting food make food-for-work costly and complex to implement, particularly at scale (World Bank, 2013). This explains why many countries have shifted towards cashbased modalities where markets can supply food reliably.
Cash-for-work programmes have demonstrated broader developmental impacts in contexts with functioning markets. Evidence from Latin America and Africa shows that cash transfers enable households to purchase diverse foods, pay school fees and invest in small businesses. Local economy-wide impact evaluations indicate that cash injections generate positive multipliers, often exceeding 1.2, as spending stimulates demand and production responds (Taylor et al., 2016). These spill overs benefit nonparticipant households and local traders, creating wider economic gains. However, where supply constraints exist, inflation can erode real income gains, highlighting the need for complementary interventions that expand local production capacity (Thome et al., 2016).
Global reviews of public works programmes find consistent short-term earnings gains during participation but limited sustained labour market effects after programme completion, except where households save or invest in productive assets during the intervention. Positive impacts on psychological well-being and women's empowerment have been observed in some contexts, though these outcomes depend heavily on programme design, including payment reliability and gender provisions (J-PAL, 2024). Large-scale programmes such as India's employment guarantee scheme and Ethiopia's PSNP have also raised local wage rates for casual labour, benefiting non-participants indirectly (Bagga et al., 2024).
Despite these benefits, cash-based programmes face challenges. Inflation and delayed payments can undermine effectiveness, while weak financial infrastructure may limit coverage in remote areas. Conversely, food-for-work can buffer households against price spikes but entails high logistical costs and limited dietary diversity. Comparative evidence therefore supports context-specific modality choice: food-for-work is most appropriate in acute crises and thin markets, while cash-for-work offers greater developmental potential in stable market environments (Gentilini, 2016).
Finally, sustainability concerns persist across both modalities. Public works are justified partly on the basis that they create durable assets, yet evaluations frequently report poor maintenance and limited integration with local development plans. Without financed operation and maintenance, assets risk becoming liabilities rather than drivers of resilience (World Bank, 2017). These lessons underscore the importance of embedding public works within broader social protection systems and linking them to agricultural development, climate adaptation and market access strategies.
2.3 Evidence from Zambia
Social protection in Zambia has evolved significantly over the past decade, with increasing recognition of its role in reducing poverty and vulnerability. The National Social Protection Policy, introduced in 2014, provided a framework for interventions aimed at cushioning households against shocks and promoting long-term resilience. While unconditional cash transfers have been widely studied, evaluations of food-for- work (FFW) and cash-for-work (CFW) remain limited. Nevertheless, available evidence and policy reports offer insights into the performance and challenges of these modalities.
Public works programmes in Zambia have historically been implemented as part of drought response strategies and rural development initiatives. According to the Policy Monitoring and Research Centre, these programmes have contributed to household food security and the creation of community assets such as feeder roads and small dams (PMRC, 2017). FFW interventions were particularly prominent in drought-prone regions, where food distribution provided immediate relief to households facing acute shortages. These interventions helped prevent hunger and stabilise consumption during crises. However, they were criticised for limited dietary diversity and inability to meet non-food needs, as transfers were restricted to cereals. This limitation underscores the trade-off between ensuring caloric intake and enabling households to exercise choice over their consumption patterns.
CFW interventions have gained prominence in recent years, particularly in response to climate-induced shocks. Pilot projects in Central Province demonstrated that cash transfers under public works improved household flexibility, enabling beneficiaries to purchase diverse foods, pay school fees and invest in small businesses. These findings align with global evidence that cash transfers offer greater developmental potential by allowing households to allocate resources according to their priorities (Gentilini, 2016).
However, operational challenges have constrained effectiveness. Reports from the Ministry of Local Government and Rural Development indicate that delays in disbursement were common, reducing the timeliness of assistance and undermining programme objectives (MLGRD, 2024). Payment delays are particularly problematic in contexts where households rely on transfers to meet urgent needs during lean seasons or droughts.
The administrative complexity of implementing CFW at scale has also been highlighted. Zambia's financial infrastructure in rural areas remains underdeveloped, which affects the reliability of electronic payment systems. In some districts, poor network coverage and verification issues led to significant delays, prompting government directives to accelerate disbursement and adopt mobile payment platforms to improve efficiency (Youth Village Zambia, 2025). These operational bottlenecks illustrate the importance of delivery systems in determining programme success. Even when design features are sound, weak implementation can erode trust and reduce welfare gains.
Budgetary constraints further complicate the sustainability of public works programmes. Social protection allocations have increased in recent years, but inflation and rising costs have eroded the real value of transfers. This macroeconomic context matters for CFW because wage adequacy determines whether households can meet basic needs and invest in productive activities. Without periodic adjustments to account for price changes, the purchasing power of cash transfers declines, limiting their effectiveness (UNICEF, 2023). Similarly, FFW programmes face sustainability challenges related to logistics and dependency. Procuring, storing and transporting food requires significant infrastructure and donor support, raising questions about long-term feasibility once external funding declines.
Gender dynamics in Zambia's public works programmes mirror patterns observed elsewhere. Women are more likely to participate in FFW interventions, reflecting their traditional role in household food provisioning. However, participation can exacerbate women's workload, as they must balance programme labour with domestic responsibilities such as childcare and water collection. In contrast, CFW programmes often attract higher male participation, as cash is perceived as a financial resource controlled by men. This raises concerns about intra-household equity and decisionmaking power. Evidence from other African contexts suggests that gender-sensitive design features, such as direct payments to women and flexible work arrangements, can enhance empowerment and improve outcomes for children (Holmes and Jones, 2010). Incorporating such measures into Zambia's programmes could help address gender disparities and promote inclusive development.
Despite these challenges, public works programmes in Zambia have demonstrated potential for contributing to resilience and local development. When implemented effectively, they provide immediate relief during crises and create assets that support long-term productivity. However, their success depends on several factors: timely and predictable payments, adequate wage levels, transparent targeting, and integration with complementary interventions such as agricultural support and financial literacy training. Linking public works to broader development strategies is essential for sustainability. For example, aligning asset creation with climate adaptation priorities can ensure that works deliver resilience benefits beyond the project cycle.
Evidence gaps remain significant. Few studies have systematically assessed the longterm impacts of FFW and CFW on livelihood outcomes such as income diversification, asset accumulation and resilience to climate shocks. Similarly, gendered impacts are underexplored, despite indications that women and men experience these interventions differently. Addressing these gaps requires rigorous evaluations that combine quantitative and qualitative methods to capture both economic and social dimensions of programme performance. Such research would inform policy decisions on modality choice, design features and integration with national social protection systems.
In summary, Zambia's experience with FFW and CFW reflects broader global trends. FFW has been effective in providing immediate food relief during crises but offers limited flexibility and faces logistical challenges. CFW enhances household choice and can stimulate local economies, yet its effectiveness is constrained by operational bottlenecks and inflationary pressures. Both modalities require strong institutional capacity, reliable delivery systems and gender-sensitive design to achieve their full potential. Embedding public works within a coherent social protection framework and linking them to resilience and development strategies will be critical for ensuring that these programmes move beyond short-term relief to deliver sustainable improvements in livelihoods.
2.4 Case Study: Ethiopia
Ethiopia's Productive Safety Net Programme (PSNP) is one of the largest public works programmes in Africa and provides a valuable case for understanding the dynamics of food-for-work (FFW) and cash-for-work (CFW) modalities. Introduced in 2005, PSNP was designed to address chronic food insecurity while building community assets that support long-term development. The programme operates in rural areas where vulnerability to drought and climate shocks is high, and it combines transfers with labour-intensive public works to achieve both consumptions smoothing and asset creation objectives.
A distinctive feature of PSNP is its dual modality approach. Transfers are provided either as food or cash depending on market conditions. In regions where markets are thin and food prices volatile, FFW is prioritised to ensure that households receive adequate caloric intake during lean seasons. This approach has proven effective in preventing hunger and stabilising consumption during severe droughts, when market-based solutions are unreliable (World Bank, 2013). Food transfers typically consist of cereals such as maize or wheat, which meet basic energy requirements but offer limited dietary diversity. While this modality addresses immediate food needs, it restricts household choice and cannot cover non-food expenses such as healthcare or education.
In areas with functioning markets, PSNP delivers CFW transfers. Cash payments enable households to purchase diverse foods, pay school fees and invest in small businesses. Studies show that cash transfers under PSNP have improved household resilience by facilitating savings and productive investments that buffer against future shocks (Berhane et al., 2014). Beneficiaries report using cash to buy livestock, agricultural inputs and other goods that enhance income-generating capacity. These findings align with global evidence that cash-based interventions offer greater developmental potential than in-kind transfers when markets are stable and accessible.
The programme's impact on food security has been well documented. Longitudinal evaluations indicate that sustained participation in PSNP increases the number of months households experience adequate food availability and reduces reliance on distress coping strategies such as selling productive assets (Berhane et al., 2014). Moreover, PSNP has contributed to improvements in asset holdings, particularly livestock, among households that combine programme benefits with complementary agricultural support. These outcomes demonstrate the importance of integrating public works with broader livelihood interventions to maximise impact.
Beyond consumption and income effects, PSNP has generated significant environmental benefits through its public works component. Activities such as soil and water conservation, gully rehabilitation and afforestation have improved land productivity and reduced erosion. Recent studies using satellite imagery showthatareas covered by PSNP public works have experienced measurable increases in tree cover, contributing to climate change mitigation and adaptation (Hirvonen et al., 2022). These environmental gains highlight the potential of well-designed public works to deliver cobenefits for resilience and sustainability.
A key strength of PSNP lies in its institutional architecture. The programme is embedded within Ethiopia's national systems and incorporates a risk financing mechanism that enables rapid scale-up during droughts. This feature ensures predictability and timeliness of transfers, which are critical for protecting households during crises. By working through government structures rather than parallel systems, PSNP has strengthened state capacity for social protection and disaster response (World Bank, 2013). The programme also includes grievance redress mechanisms and social accountability tools that enhance transparency and allow beneficiaries to report exclusion errors or payment delays. These governance features are essential for maintaining credibility and ensuring equitable access.
Despite these achievements, PSNP faces challenges that limit its effectiveness. One concern relates to asset quality and maintenance. While the programme has created extensive infrastructure and environmental works, evaluations reveal that some assets deteriorate quickly due to inadequate technical standards or lack of post-project maintenance (Levineetal., 2024).This problem reflectsa broadertension in public works design: the pressure to absorb large numbers of workers can lead to compromises in engineering quality and sustainability. Addressing this issue requires stronger technical oversight and dedicated funding for operation and maintenance beyond the construction phase.
Gender dynamics within PSNP also warrant attention. Women participate actively in public works, particularly in FFW activities, but often face time poverty as they balance programme labour with domestic responsibilities. Although PSNP has introduced gender-sensitive provisions such as lighter tasks for pregnant women and flexible work schedules, gaps remain in ensuring that women have equal access to decision-making roles and control over resources. Evidence suggests that direct cash payments to women can enhance their bargaining power and improve outcomes for children, underscoring the need for continued efforts to mainstream gender equity in programme design (Holmes and Jones, 2010).
Another challenge concerns payment reliability. While PSNP has made progress in reducing delays, occasional disruptions still occur, particularly in remote areas with weak financial infrastructure. Payment delays undermine trust and force households to revert to negative coping strategies, eroding programme benefits. Continued investment in digital payment systems and local capacity building is necessary to ensure timely delivery.
The Ethiopia case offers several lessons for other countries implementing public works programmes. First, modality choice should be guided by market analysis. FFW is appropriate in contexts of acute food insecurity and market failure, while CFW offers greater flexibility and developmental potential where markets function well. Second, integrating public works with complementary interventions such as agricultural extension and financial services enhances impact and sustainability. Third, governance mechanisms including grievance redress, social accountability and risk financing are critical for programme credibility and responsiveness. Finally, attention to asset quality, gender equity and payment reliability determines whether public works deliver transformative outcomes or remain limited to short-term relief.
In conclusion, PSNP demonstrates that public works programmes can achieve multiple objectives when designed and implemented effectively. By combining transfers with asset creation and embedding operations within national systems, Ethiopia has built a model that addresses immediate consumption needs while promoting resilience and development. However, persistent challenges related to asset maintenance, gender inclusion and operational efficiency highlight the need for continuous adaptation and investment. These lessons are highly relevant for Zambia and other countries seeking to strengthen social protection systems in the face of climate shocks and economic uncertainty.
2.5 Case Study: Kenya
Kenya provides a useful case for understanding the evolution from food for work to cash for work in drought affected areasand the ways public works can contribute to resilience when embedded in national systems. The arid and semi-arid lands constitute a large proportion of the country and experience recurrent drought with significant impacts on livelihoods. Over the past two decades, Kenya has progressively institutionalised cash transfers as a primary instrument for shock response and poverty reduction, while county level public works have focused on water harvesting, basic rural infrastructure and rangeland management.
The Hunger Safety Net Programme is a government led cash transfer operating in the northern arid counties. It emerged as an alternative to traditional food aid by providing predictable cash to chronically vulnerable households and by building government capacity to scale payments during drought events (OPM, 2018). Independent evaluation of the programme's second phase found that regular and timely cash helped cushion households against climate induced shocks and reduced the need for negative coping strategies, confirming the role of cash in stabilising consumption and supporting dignified choice in contexts where local markets function (OPM, 2018). The transition to full government ownership, managed by the National Drought Management Authority, has involved strengthening grievance and case management systems and establishing shock responsive mechanisms for horizontal expansion during crises, which is critical for programme credibility and timely response (NDMA, 2024).
Kenya's experience also demonstrates the importance of governance architecture for social protection. The Hunger Safety Net Programme has progressively aligned payment systems, beneficiary identification and appeals procedures with national standards, improving transparency and efficiency. Programme documents set out operational guidance for stakeholder engagement, grievance handling and risk management, reflecting the need for strong institutions and accountability when disbursing cash at scale across remote areas (NDMA, 2024). Lessons from case studies of earlier drought episodes emphasise that predictable transfers can outperform ad hoc relief by enabling households to plan and by reducing delays associated with assessment and mobilisation offood assistance (SPaN, 2019).
Alongside cash transfers, public works have focused on assets that address local water and land constraints. Field practice in Kenya shows that water harvesting structures designed around roads and small dams can capture and store runoff for domestic use, small scale irrigation and livestock, supporting livelihood diversification and reducing erosion when implemented to technical standards and maintained properly (Nissen Petersen, 2006). Guidance for practitioners describes low cost techniques such as farm ponds, sand dams and zai pits, which can be selected according to local hydrology and soils and delivered through community labour with appropriate supervision (WFP, 2018). Recent analytical work on site suitability mapping illustrates how simple spatial criteria and local data can improve the siting of road runoff harvesting, thereby increasing effectiveness and reducing the risk of asset failure from poor location choices (Niemans, 2024). These resources together underline that the success of public works depends not just on building assets but on choosing the right assets for local conditions and planning for operation and maintenance beyond construction.
Urban and county level guidance also recognises the value of integrating resilience into infrastructure delivery. A World Bank note for municipal engineers highlights how climate risk screening, cross sector coordination and technical standards can make assets more durable and useful under variable rainfall, which is directly relevant to county public works and to water management in small towns that anchor rural economies (World Bank, 2023). In rural areas, emerging regulation for water services points to the need for clear performance expectations, results-based financing and post implementation support if community operated water points are to remain functional over time. Without such arrangements, assets created through public works may deteriorate and fail to deliver promised resilience benefits (REACH, 2025).
Gender dynamics in Kenya's cash programmes offer additional insight. Evidence from Wajir suggests that direct cash payments to women can enhance their economic inclusion and decision-making power, although gains depend on complementary support services such as financial literacy and microenterprise training (Ndoka, 2020). These findings align with broader social protection literature that highlights the importance of gender sensitive design features, including accessible payment mechanisms and safe worksites, for achieving improvements in household nutrition and schooling outcomes (Holmes and Jones, 2010). For public works, measures such as flexible hours, childcare and assigning lighter tasks where appropriate can reduce time burdens on women and increase net benefits.
Some challenges persist. Inflation can erode the real value of cash and reduce purchasing power during drought periods, particularly if transfer levels are not adjusted. This risk is mitigated where cash is delivered predictably and where public works expand access to water and transport, which can stabilise local prices and improve market functioning over time. Evaluations of Kenya's cash systems note that reliability and timeliness have been decisive for effectiveness, since delays undermine trust and force households back into negative coping strategies (OPM, 2018). Ensuring rapid scale up when early warning signals trigger drought response remains an operational priority, requiring strong coordination between the National Drought Management Authority and county administrations (NDMA, 2024).
Asset quality and maintenance are recurrent issues. Case experience shows that technical designs must suit local conditions and that communities need post project support to keep assets functional. Field guidance stresses simple designs, community ownership and routine maintenance, yet in practice these are sometimes under resourced, leading to shortened asset lifetimes and reduced benefits. Better integration of operation and maintenance finance and use of performance monitoring by counties could help address this weakness. Where community labour is mobilised rapidly to absorb large numbers of workers, there is a risk that engineering standards are compromised;balancing labour absorption with technical quality is therefore central to public works management (WFP, 2018).
Taken together, Kenya's case suggests several transferable lessons. First, modality choice must follow market analysis. In areas with functioning markets, cash for work combined with predictable delivery and strong grievance systems can be more effective than food for work, offering households flexibility and generating local economic spill overs. Where markets are thin or prices are volatile, food transfers can be appropriate to protect consumption until markets recover. Second, the public works portfolio should prioritise water and land assets with demonstrable livelihood benefits and feasible maintenance arrangements. Tools such as site suitability mapping and simple engineering guidance can improve project selection and asset durability. Third, embedding cash transfers and public works within government systems, with risk financing and clear escalation protocols, enables quicker and more equitable shock response. Finally, gender sensitive design and accessible payment systems strengthen the developmental impacts of cash and public works by enhancing women's agency and household welfare.
For Zambia, the implications are clear. The experience from Kenya reinforces the importance of shock responsive cash delivery backed by robust grievance and case management and of selecting public works that address local climate risks, especially water scarcity and land degradation. Aligning programme operations with national systems and county level capacity, planning for maintenance finance and adopting gender sensitive features can help ensure that public works and cash transfers progress beyond short term relief to sustained resilience and livelihood improvements (OPM, 2018;NDMA, 2024).
2.6 Case Study: Malawi
Malawi offers an important perspective on the implementation of cash for work (CFW) programmes in a low-income context characterised by high vulnerability to seasonal food insecurity and climate shocks. Public works have been a central component of Malawi's social protection strategy, primarily delivered through the Malawi Social Action Fund (MASAF). These programmes aim to provide temporary employment during lean seasons while creating community assets that support local development. However, evidence from rigorous evaluations suggests that the effectiveness of these interventions has been mixed, with significant lessons for design and implementation.
CFW programmes in Malawi were introduced to stabilise household consumption during the agricultural lean season, when food stocks are depleted and income opportunities are scarce. Beneficiaries receive cash wages in exchange for labour on projects such as road rehabilitation, afforestation and small-scale irrigation. The rationale is twofold: to provide immediate income support and to build assets that enhance productivity and resilience. In theory, this approach should reduce reliance on negative coping strategies such as distress sales of livestock or withdrawal of children from school. However, empirical findings indicate that these objectives have not always been fully realised.
A large-scale randomised evaluation of Malawi's public works programme found no significant improvement in household food security, even when programme design was adjusted to offer work during the lean season and to increase payment frequency (Beegle et al., 2017). The study also reported negative spill over effects on nonbeneficiary households within treated villages, suggesting that the intervention may have inadvertently distorted local labour markets or increased competition for scarce resources. These findings challenge assumptions about the automatic benefits of public works and underscore the importance of context-specific design.
Several factors explain the limited impact observed in Malawi. First, wage levels were often too low to meet household needs or to enable savings and investment. When transfers are insufficient relative to prevailing prices, households cannot accumulate resources to buffer against future shocks. Second, programme duration was short, typically providing onlya few days ofwork per household. This limited exposure reduces the potential for meaningful income gains and constrains opportunities for skill development. Third, payment delays were common, undermining the timeliness of assistance and forcing households to revert to negative coping strategies. These operational weaknesses highlight the critical role of delivery systems in determining programme effectiveness.
Inflation further eroded the real value of cash transfers, particularly during periods of macroeconomic instability. Without mechanisms to adjust wages in line with price changes, the purchasing power of transfers declined, reducing their ability to stabilise consumption. This problem is not unique to Malawi but is especially acute in contexts where inflation is volatile and financial infrastructure is weak. Strengthening payment systems and introducing indexation mechanisms could help mitigate these risks.
Asset quality and sustainability also emerged as concerns. While public works created physical infrastructure such as feeder roads and water harvesting structures, evaluations revealed that many assets were poorly maintained and deteriorated quickly after project completion. This outcome reflects a broader challenge in public works design: the tension between absorbing labour for short-term relief and ensuring technical standards for long-term utility. Addressing this issue requires dedicated funding for operation and maintenance and stronger integration with local development plans. Without these measures, assets risk becoming liabilities rather than drivers of resilience.
Gender dynamics within Malawi's public works programmes mirror patterns observed elsewhere in sub-Saharan Africa. Women often face barriers to participation due to childcare responsibilities, cultural norms and safety concerns. When women do participate, they may experience increased time poverty as they balance programme labour with domestic duties. Evidence suggests that gender-sensitive design features such as flexible work hours, provision of childcare and direct payments to women can enhance empowerment and improve outcomes for children (Holmes and Jones, 2010). Incorporating these measures into Malawi's programmes could help address gender disparities and maximise developmental impacts.
Despite these challenges, CFW programmes in Malawi have demonstrated some positive effects. Beneficiaries reported using cash to purchase food, pay school fees and invest in small businesses, indicating that transfers can support diverse needs when delivered reliably. Moreover, public works have contributed to environmental rehabilitation through afforestation and soil conservation activities, which are critical for reducing vulnerability to climate shocks. These achievements, though modest, highlight the potential of well-designed public works to deliver both protective and promotive benefits.
The Malawi case underscores several lessons for policy and practice. First, wage adequacy and payment timeliness are essential for achieving programme objectives. Transfers must be sufficient to meet basic needs and delivered predictably to enable households to plan and avoid negative coping strategies. Second, programme duration and intensity should be calibrated to provide meaningful income support without creating dependency. Third, asset selection and maintenance require careful planning and dedicated resources to ensure that public works deliver long-term benefits. Fourth, gender-sensitive design is critical for promoting equity and enhancing the developmental impact of interventions.
Finally, the Malawi experience highlights the importance of rigorous evaluation in informing programme design. Evidence from randomised trials and process assessments provides valuable insights into what works and what does not, enabling policymakers to refine interventions and allocate resources more effectively. For Zambia and other countries implementing public works, these lessons emphasise the need for context-specific approaches that balance short-term relief with long-term resilience and development objectives.
2.7 Comparative Regional Insights
This unit synthesises lessons from Ethiopia, Kenya and Malawi and contrasts them with Zambia's emerging experience. It focuses on how food-for-work and cash-for-work perform across different market and institutional contexts, what kinds of assets are most likely to yield durable benefits, and which delivery system features most strongly condition outcomes. The purpose is to extract practical insights that can inform the design, sequencing and governance of Zambia's public works within a coherent social protection system.
A first cross cutting lesson concerns modality choice and market function. Ethiopia's Productive Safety Net Programme deliberately calibrates transfers to local conditions. Food-for-work is used where markets are thin and cereal prices volatile, while cash for work is preferred where markets function and households can exercise choice. This approach has protected consumption in drought and supported diverse spending and small investments where cash is viable (World Bank, 2013). Longitudinal evaluation shows that sustained participation can increase months of adequate food and build livestock holdings, especially when households also access complementary agricultural support (Berhane et al., 2014). By contrast, Malawi's programme delivered limited food security gains despite design tweaks to timing and payment frequency, underlining that cash for work does not guarantee improvement when wages are low, exposure is short and delivery is unreliable (Beegle et al., 2017). Kenya's experience with the Hunger Safety Net Programme demonstrates that predictable cash can cushion climate shocks and reduce negative coping where markets are accessible and payment systems are robust (OPM,2018).
A second theme is payment timeliness and reliability. Predictable transfers are central to welfare gains because households plan purchases and avoid distress sales when they trust delivery schedules. Ethiopia's inclusion of risk financing to scale support during drought has improved predictability and protected consumption during shocks (World Bank, 2013). Kenya's transition to government ownership through the National Drought Management Authority strengthened case management and grievance procedures to deal with errors and delays, which improved credibility and allowed shock responsive expansion during drought years (NDMA, 2024). In Malawi, payment delays contributed to weak results by forcing households back into negative coping strategies, even when work was scheduled in the lean season (Beegle et al., 2017). Zambia's experience in 2024 and 2025 likewise shows that operational bottlenecks such as poor network coverage and verification issues can slow disbursement, prompting the use of mobile payment platforms and high-level directives to accelerate payments (MLGRD, 2024;Youth Village Zambia, 2025).
A third insight concerns asset quality, site selection and maintenance. Where projects match local risks and are delivered to technical standards with financed operation and maintenance, the returns can extend well beyond wage effects. Remote sensing evidence from Ethiopia indicates measurable increases in tree cover in watersheds with sustained public works, suggesting positive environmental co benefits and climate adaptation potential (Hirvonen et al., 2022). At the same time, retrospective studies from Ethiopia and Kenya caution that assets may deliver limited livelihood benefits when labour absorption takes precedence over engineering quality or when maintenance is unfunded, leading to rapid deterioration (Levine et al., 2024). Kenyan practice manuals and recent site suitability work show that simple spatial screening for road runoff harvesting and water structures can reduce failure rates and increase impact if counties apply consistent technical standards and plan for maintenance after construction (WFP, 2018;Niemans, 2024).
A fourth point is gender and inclusion. Evidence from Kenya shows that direct payments to women can strengthen their decision-making power, though gains depend on complementary support such as financial literacy and microenterprise services (Ndoka, 2020). In all three contexts, women face time poverty if design does not offer flexible hours, proximate worksites and safe conditions. Ethiopia's successive programme phases have added gender provisions, but experience shows continued need for attention to women's leadership roles and control of resources to realise transformative impacts (Berhane et al., 2014).
Finally, inflation and transfer adequacy affect the real value of wages and the scope for savings or small investments. Kenya's evaluations emphasise that reliability and adequacy together drive effectiveness, whereas Malawi's results show that low transfer size and short cycles blunt impact even with correct seasonal timing (OPM, 2018;Beegle et al., 2017). In Zambia, social protection budget reviews indicate that cost of living pressures can erode the real value of transfers, strengthening the case for periodic adjustment or for linking labour programmes to complementary inputs and services so that households can leverage wages into productive gains (UNICEF, 2023).
The table below consolidates these insights and anchors them to specific programme features and delivery choices.
The table shows Comparative insights on food-for-workand cash-for-work
Illustrations are not included in the reading sample
Source notes: Ethiopia insights draw on programme case material, impact evaluation and environmental analysis (World Bank, 2013;Berhane et al., 2014;Hirvonen et al., 2022). Kenya findings reflect independent evaluation and operational documents, along with field manuals and siting studies for water works (OPM, 2018; NDMA, 2024; WFP, 2018; Niemans, 2024). Malawi results are taken from a randomised evaluation of the public works programme (Beegle et al., 2017). Zambia draws on policy and operational documents and public communications on payment reforms alongside social protection budget analysis (MLGRD, 2024;Youth Village Zambia, 2025;UNICEF, 2023).
Across these cases, three design choices recur.
1. Use response analysis to set modality. Food transfers are most appropriate where food markets are unreliable and prices are volatile, while cash transfers are preferred where markets function and beneficiaries can exercise choice. Ethiopia's calibrated approach is a useful model for hybrid systems that vary modality by district and season (World Bank, 2013).
2. Engineer for durability and plan for operation and maintenance. Public works can be more than temporary labour if they deliver assets that reduce risk and raise productivity. Ethiopia's environmental gains and Kenya's water harvesting experience show what is possible when assets are well sited and standards are upheld, whereas retrospective audits warn of limited benefits when labour absorption dominates technical quality (Hirvonen et al., 2022;Levine et al., 2024; WFP, 2018).
3. Get delivery right at scale. Payment timeliness, grievance redress, and clear operating manuals are not administrative details but core impact drivers. Kenya's government systems and Ethiopia's risk financing improved predictability, while delays in Malawi and in parts of Zambia point to the need for reliable digital payments and contingency plans for connectivity or verification failures (OPM, 2018; World Bank, 2013;MLGRD, 2024).
For Zambia, these comparative insights suggest concrete priorities. First, continue to stabilise payment systems by expanding mobile money channels, strengthening Page 45 of 278 identity verification and monitoring payment timelines, since timeliness materially affects food security during the lean season. Second, align the works menu with climate risk by prioritising small scale irrigation, soil and water conservation and feeder road spot improvements that unlock market access, using simple site screening to minimise failure and planning for maintenance with local authorities. Third, link public works with complementary services such as agricultural extension, input access and financial literacy so that households can convert wages into self-driven productive gains. Finally, adopt gender responsive design by ensuring flexible hours, safe and proximate worksites and, where feasible, direct payments to women to strengthen decision making and child wellbeing, as shown in Kenyan experience (Ndoka, 2020; Holmes and Jones, 2010).
The comparative record is clear. Public works deliver the largest and most durable benefits when modality fits markets, assets fit landscapes, and systems fit scale. Countries that have invested in these alignments have been able to move beyond short term relieftoward resilient livelihoods.
2.8 Governance and Transparency in Public Works
Effective governance and transparency are decisive for the impact, equity and credibility of food-for-work and cash-for-work. Public works programmes must manage complex delivery chains that include targeting, registration, payments, work supervision, technical quality assurance, operation and maintenance, monitoring and evaluation, and grievance redress. Weaknesses in any link can erode welfare gains and public trust, while strong systems can convert short term transfers into durable livelihood and resilience benefits.
A first pillar of governance is transparent beneficiary selection and appeals. Community based verification, public posting of lists and accessible appeals procedures help to reduce both inclusion and exclusion errors. Ethiopia's Productive Safety Net Programme operationalises these practices at kebele and woreda level and couples them with risk financing to scale assistance in drought years, strengthening predictability and institutional capacity (World Bank, 2013). Programme manuals set out responsibilities for citizen engagement and grievance handling, with clear lines of accountability for local committees and technical staff (PSNP PIM, 2020s). These measures matter because the credibility of public works depends on households seeing fair, rule bound processes and having recourse when errors occur.
The second pillar is payment timeliness and reliability. In public works, delays and irregularity can force households back into negative coping strategies, undoing intended welfare gains. Kenya's Hunger Safety Net Programme improved payment reliability by embedding delivery in national systems and strengthening case management and appeals under the National Drought Management Authority, which proved critical for shock responsive expansion during droughts (OPM, 2018). In Zambia, as the cash-for-work scale up progressed, network coverage and verification problems caused payment delays, prompting government directives to accelerate disbursement and adoption of mobile payment accounts to reach remote areas more efficiently (MLGRD, 2024; Youth Village Zambia, 2025). These experiences underline that robust delivery infrastructure is not administrative detail but a central impact driver.
A third pillar is grievance redress and social accountability. Project level grievance mechanisms allow beneficiaries and communities to report targeting errors, payment problems, worksite issues or asset quality concerns and to obtain timely resolution. International guidance sets out principles for accessible intake channels, clear time standards, transparent tracking and feedback to complainants (World Bank GRS, 2021). In Ethiopia, social accountability initiatives under ESAP have demonstrated how citizen scorecards, public dialogues and joint service improvement plans can strengthen transparency and responsiveness in basic service delivery and can be adapted for public works oversight at local level (VNG-International, 2025). When grievance and accountability mechanisms work, they create a feedback loop that reduces error rates and improves programme quality over time.
The fourth pillar is technical quality assurance and financed operation and maintenance. Wage transfers are only half of the value proposition in public works;the assets created should deliver sustained public goods and productivity gains. Evidence from retrospective studies in Ethiopia and Kenya shows that where project planning has prioritised absorbing labour over engineering quality or where operation and maintenance budgets are absent, assets deteriorate quickly and livelihood impacts are weak (Levine et al., 2024). Conversely, environmental works delivered to standards and maintained with community ownership and local government support can generate measurable gains in land restoration and tree cover over time (Hirvonen et al., 2022). In practice, governance must ensure that asset menus fit local risk profiles, that siting follows simple analytical screening, that construction meets standards, and that maintenance finance and responsibilities are specified upfront.
A fifth pillar is financial management, audit and fraud control. Sound segregation of duties, reconciled payment files, secure registries, transparent procurement and regular internal and external audits reduce leakage risks. Digital payments with appropriate verification minimise cash handling and improve audit trails. In new or rapidly scaled programmes, additional attention to identity verification and reconciliation processes helps maintain integrity during high volumes, as the Zambia case demonstrated when mobile payments and strengthened oversight were introduced to address delays and data mismatches (MLGRD, 2024). These controls safeguard public resources and sustain public confidence.
A sixth pillar is data systems and monitoring and evaluation. Management information systems that capture eligibility, participation, payments, work completion and asset performance enable managers to monitor key performance indicators and address bottlenecks. Sex disaggregated reporting supports gender analysis and the identification of barriers to women's participation and benefit. In Kenya, the evaluation record shows that improvements in case management and information systems were associated with more reliable payments and better grievance handling (OPM, 2018).
Embedding learning in programme cycles allows iterative improvement of asset menus, worksite practices and payment processes.
Political economy and community dynamics shape governance outcomes. Community selection processes can be vulnerable to elite influence unless rules, transparency and appeals are enforced; conversely, overly centralised selection can miss local knowledge about vulnerability. Reviews of social accountability in rural Africa note that citizen information campaigns, scorecards and audits can moderate these risks when they are properly resourced and linked to responsive local administrations (DIIS, 2013). Balancing local discretion with standardised national safeguards is therefore essential.
Drawing these strands together, a pragmatic governance checklist for public works includes the following elements. First, make beneficiary selection transparent, publish lists locally and ensure simple, accessible appeals with time standards. Second, invest in payment systems that are reliable in rural areas, track payment timeliness and reconcile transactions regularly. Third, implement grievance mechanisms at project and portfolio level and publicise them widely, with dashboards that show cases received and resolved. Fourth, ensure asset quality and maintenance by screening sites, specifying standards, supervising works and financing operation and maintenance responsibilities after completion. Fifth, strengthen financial controls, including segregation of duties, secure registries, audit trails and routine audits. Sixth, build data systems and monitoring that capture participation, payments, asset completion and performance, and report sex disaggregated indicators to track equity. Finally, codify all of the above in clear operational guidance and train local implementers so that practice converges on standards.
For Zambia, the near-term priorities are clear. Continued stabilisation of cash-for-work payments using mobile channels and strengthened verification will reduce delays and improve trust. Clear grievance and accountability arrangements can help communities surface issues early and support corrective action. Technical guidance for selecting climate relevant assets such as small-scale irrigation and soil and water conservation, paired with financed operation and maintenance, can ensure that works deliver resilience benefits beyond the project cycle. These steps, implemented within national systems and backed by transparent budgets and reporting, will make public works more equitable, effective and sustainable (MLGRD, 2024;UNICEF, 2023; World Bank GRS, 2021).
2.9 Gaps in the Literature
Despite the expansion of social protection across sub Saharan Africa, important gaps remain in the evidence based on food-for-work and cash-for-work, especially in relation to Zambia. Much of the rigorous work to date has concentrated on a few flagship programmes such as Ethiopia's Productive Safety Net Programme, India's employment guarantees and selected trials in Malawi, with limited country specific learning for Zambia beyond policy briefs and operational notes (Berhane et al., 2014;Beegle et al., 2017). As a result, questions of local effectiveness, equity and sustainability remain open in several domains that matter for programme design and for integration within national systems.
A first gap concerns longitudinal livelihood trajectories. While there is credible evidence that sustained participation can improve food security and raise livestock holdings in Ethiopia, Zambia lacks multi-year panel evaluations that track households' income diversification, productive asset accumulation and resilience to climate shocks after exposure to food-for-work and cash-for-work (Berhane et al., 2014). Without such data, it is difficult to judge whether public works are simply consumption smoothing instruments or whether they catalyse self-driven improvements in the five capitals within the sustainable livelihoods framework (DFID, 1999).
A second gap is gendered impacts and intra household dynamics. Reviews show that gender sensitive design influences empowerment and child outcomes in social protection, yet there is little Zambia specific evidence on participation barriers, time poverty, safety, decision making and control over transfers in public works (Holmes and Jones, 2010). Comparative insights from Kenya suggest that direct payments to women can enhance economic inclusion, but the extent to which similar measures would shift bargaining power and welfare in Zambian contexts remains under researched (Ndoka, 2020).
A third gap relates to asset lifecycle performance. Retrospective studies in Ethiopia and Kenya warn that public works can produce assets with limited livelihood benefits where engineering quality and operation and maintenance are weak, but there is little systematic asset auditing in Zambia that links construction standards, siting decisions and post project maintenance to observed productivity, environmental services and service reliability (Levine et al., 2024). Environmental remote sensing in Ethiopia confirms increases in tree cover in areas with sustained works, yet comparable evaluation of natural capital outcomes is absent in Zambia (Hirvonen et al., 2022).
A fourth gap concerns payment systems and delivery reliability. Operational accounts note delays caused by network coverage and verification issues in Zambia's cash for work scale up, but there is no quantitative reporting on payment timeliness, the distribution of delays across districts and their welfare consequences during the lean season (MLGRD, 2024). Lessons from Kenya and Ethiopia show that predictable payments are central to welfare gains and to avoidance of distress sales, suggesting the need for routine delivery metrics and public dashboards in Zambia to support accountability and continuous improvement (OPM, 2018; World Bank, 2013).
A fifth gap is modality choice under market stress. Comparative reviews argue that cash tends to match or outperform in kind assistance when markets function, and that food transfers are appropriate in thin markets or during price spikes, yet Zambia has not undertaken formal response analysis that tests cash versus food performance under localized market constraints and inflation, nor evaluated cash plus combinations that add services to the transfer (Gentilini, 2016;World Bank, 2024). This matter because the country has faced significant cost of living pressures that can erode the real value of cash and alterthe balance of modality effectiveness (UNICEF, 2023).
A sixth gap is local economy effects and spill overs. Evidence from seven African countries shows that cash injections can generate income multipliers and spill overs to non-beneficiary households when supply responds, but there are no village level general equilibrium studies in Zambia that quantify these indirect impacts under food- for-work and cash-for-work or that identify conditions under which inflation offsets nominal gains (Taylor et al., 2016). Knowing whether transfers stimulate local production and reduce prices through improved access, or instead push up prices under supply constraints, would help calibrate transfer size and project selection.
A seventh gap is value for money and comparative cost effectiveness. Global guidance acknowledges that public works vary widely in cost structures and in the balance between wage transfers and asset budgets, but Zambia has not published comparative cost effectiveness assessments that relate delivery costs, leakage control, asset durability and welfare gains to alternative programme designs, nor benchmarked food forworkagainst cash-for-workusing common metrics (World Bank, 2017). Such analysis is important for the Ministry of Finance and for planning adjustments to transfer levels in inflationary periods (UNICEF, 2023).
An eighth gap is governance, grievance and social accountability evidence. Programme credibility depends on transparent selection, accessible appeals and timely resolution of complaints, yet there is limited documentation on grievance volumes, resolution rates and the systemic issues surfaced through citizen feedback in Zambia's public works (World Bank GRS, 2021). Comparative practice from Kenya's social accountability work suggests that citizen scorecards and joint service improvement plans can make delivery more responsive, but the uptake and adaptation of similar tools to public works in Zambia has not been studied (VNG-International, 2025).
A ninth gap is wage setting, labour markets and displacement risks. Design theory highlights that wages below casual rates encourage self-targeting and reduce diversion from private employment, but Zambia lacks empirical studies that quantify displacement risks, forgone earnings and the distributional consequences of wage choices across agricultural seasons and provinces (Ravallion, 1999). Given the heterogeneity of rural labour markets, this gap makes it harder to set wages that achieve poverty relief without unintended effects on local employment.
A tenth gap is shock responsive scalability and administrative capacity. Ethiopia's risk financing facility and Kenya's horizontal expansion mechanisms underpin timely scale up during drought. Zambia has operational guidance for cash for work, yet evidence on administrative readiness, early warning linkages and scale up performance remains thin beyond high level directives to accelerate payments (MLGRD, 2024;NDMA, 2024). Documenting these capacities and stress testing them against recent drought events would support realistic planning.
A final cross cutting gap is integration with complementary services. Studies in Ethiopia show larger gains where households access both public works and agricultural support, and reviews of cash plus models recommend adding financial inclusion, extension and market access services to raise sustained wellbeing. Zambia has limited evidence on which combinations generate the strongest livelihood improvements and on the sequencing that minimises administrative burden while maximising impact (Berhane et al., 2014;World Bank, 2024).
Taken together, these gaps point to a practical research agenda. First, commission multi-year household panels that track food-for-work and cash-for-work participants and suitable comparison groups in different provinces to measure food security, savings, enterprise activity, asset accumulation and shock recovery. Second, embed gender analysis that examines participation, time use, safety, resource control and decision making, and test direct payment to women and flexible scheduling on empowerment and child outcomes. Third, audit a representative sample of public works assets for technical quality, siting and operation and maintenance, and link these to livelihood and environmental performance using simple field surveys and remote sensing. Fourth, institute routine payment timeliness reporting and grievance dashboards to strengthen accountability. Fifth, conduct response analysis trials comparing cash, food and cash plus under varied market conditions and inflation scenarios. Sixth, apply local economy wide impact methods to quantify multipliers and price effects. Seventh, run comparative cost effectiveness studies across programme designs. Eighth, test wage policies against seasonal labour markets to calibrate selftargeting and minimise displacement. Ninth, document administrative capacity for shock responsive scale up. Finally, evaluate combinations of public works with complementary services to identify scalable packages that deliver sustained improvements.
Addressing these gaps would move Zambia's public works from shortterm relief toward sustained resilience and equitable development, and would align programme decisions with evidence on modality effectiveness, delivery systems and asset durability drawn from Ethiopia, Kenya and Malawi while grounding learning in Zambia's own contexts and institutions (World Bank, 2013;OPM, 2018;Beegle et al., 2017).
2.10 Gendered Impacts of Food-for- Work and Gash-for- Work
Understanding gender dynamics is essential forjudging the effectiveness and equity of food-for-work and cash-for-work. Public works often assume that labour is equally accessible to women and men, yet in practice gender roles, care responsibilities and social norms shape participation, control over benefits and ultimate outcomes. A gender lens reveals differences across modalities and offers clear design choices that can improve programme performance.
A first consideration is participation and time poverty. Women in rural settings commonly shoulder primary responsibility for unpaid care, including childcare, food preparation and water collection. When programme schedules are rigid or worksites are distant, participation can raise time burdens and reduce the net benefit of transfers for women. Reviews of gender and social protection consistently show that design features such as flexible work hours, proximate worksites and lighter tasks for pregnant and lactating women increase women's participation and reduce time poverty (Holmes and Jones, 2010). These insights matter for both food-for-work and cash-for-work, since the physical demands of some assets and the timing of work cycles can limit inclusion unless they are adapted to women's daily routines.
A second issue concerns control over transfers and intra household decision making. In many contexts men exercise greater control over cash income, while women are more directly associated with food provisioning. As a result, food transfers are sometimes perceived as more likely to be used for household consumption, whereas cash transfers may be diverted to other uses unless women have access and control. Case evidence from Kenya indicates that direct payments to women improved their economic inclusion and decision-making power, although gains were stronger where complementary services such as financial literacy and microenterprise support were available (Ndoka, 2020). This suggests that delivery systems and payment choices can shift intra household bargaining in ways that alter the welfare impacts of cash for work.
A third dimension is safety, dignity and workplace conditions. Women may face risks of harassment or violence at or on the way to worksites. Programmes that include codes of conduct, safe worksite protocols, adequate sanitation, grievance channels and zero tolerance enforcement provide essential safeguards. Gender sensitive grievance procedures that allow confidential complaints and set clear time standards for resolution are central to building trust and reducing barriers to participation. Global guidance on grievance redress emphasises accessible intake, tracking and transparent feedback, which are directly applicable to public works that mobilise large numbers of women at community level (World BankGRS, 2021).
A fourth theme relates to programme modality and its gendered implications. Food- for-work can ensure that food reaches households during crises, protecting consumption when markets are thin and prices volatile. However, the ration often consists of cereals and does not cover non-food needs, which may matter for women's priorities in health and education. Cash for work offers fungibility and can be used to buy diverse foods, pay fees and invest in small enterprises. Evidence from Ethiopia's Productive Safety Net Programme shows that cash transfers enabled savings and small investments that buffered households against future shocks, with positive effects when women participated actively and could influence spending decisions (Berhane et al., 2014). The Ethiopian case also demonstrates how a calibrated approach that uses food where markets are weak and cash where markets function can protect consumption and increase women's options in the same programme cycle (World Bank, 2013).
Design choices can mitigate gendered constraints across both modalities. Directing payments to women or providing joint accounts where appropriate can increase women's control over resources and strengthen child nutrition and schooling outcomes. Programme manuals that specify flexible hours, childcare arrangements and accessible worksites reduce the opportunity cost of participation for women and raise net benefits. Reviews of gender and social protection recommend combining transfers with complementary services, such as financial inclusion and livelihood coaching, to reinforce women's ability to translate transfers into sustained gains for the household (Holmes and Jones, 2010). These measures are especially important in cash for work, where control over cash is contested.
There are also positive examples of mainstreaming gender in public works. Successive phases of Ethiopia's Productive Safety Net Programme have introduced gender provisions, including lighter tasks for pregnant and lactating women, flexible scheduling, and roles for women in supervision and local decision structures. Qualitative assessments of implementation highlight that outcomes depend on fidelity to these provisions and on local norms and administrative capacity, but they confirm that intentional gender design increases women's participation and perceived benefits (UNICEF Innocenti, 2024). Kenya's cash systems under the Hunger Safety Net Programme offer practical lessons on accessible payment channels and grievance handling that can be adapted for women's needs, particularly in remote arid counties where financial infrastructure is sparse (OPM, 2018).
Experience also shows risks when gender is ignored. In Malawi, short exposure, low wages and payment delays combined with the absence of gender sensitive features yielded limited food security gains in public works and placed burdens on households during lean seasons (Beegle et al., 2017). Without direct payment to women or flexible arrangements, cash for work can consolidate existing inequalities rather than improve welfare. Even in otherwise successful systems, women may be underrepresented in leadership roles or in project selection committees, which affects the types of assets chosen and the distance to worksites. Programmes should therefore monitor sex disaggregated indicators and track barriers to participation, control over transfers and leadership overtime.
Monitoring and evaluation are essential for learning and accountability. Gender outcome measurement should go beyond simple participation rates to capture time use, safety, decision making, control of resources and child wellbeing. Sex disaggregated reporting across registration, attendance, payments, grievance submissions and asset supervision allows managers to diagnose bottlenecks and adapt operations. Embedding gender questions in routine surveys and using qualitative methods to understand norms and workplace culture help programmes adjust risk management and outreach. Reviews emphasise that purposeful measurement coupled with responsive management results in improved equity and programme performance (Holmes and Jones, 2010).
Understanding intersectional dynamics is also important. Women's experiences vary by age, marital status, disability, location and livelihood. Young women may face different risks and opportunities than older women, and female headed households may have distinct constraints compared to those headed by men. Tailoring outreach, task assignment and payment mechanisms to these differences increases inclusion. Programmes should consider culturally appropriate communication and partnership with local women's groups to ensure that information about eligibility, schedules and grievance channels reaches those most affected by barriers.
The policy implications for Zambia follow from these lessons. First, build gender sensitive design into programme operations from the outset. Payment systems should enable direct payments to women where feasible and include joint options where household dynamics favour shared control. Work schedules should be flexible and proximate to reduce travel and time costs, and childcare solutions should be considered for larger worksites. Second, institute safe worksite protocols and confidential grievance mechanisms with clear time standards that are publicised in communities. Third, measure gendered outcomes routinely and use data to adapt operations. Fourth, link transfers to complementary services, including financial literacy and small enterprise support, which have been shown to reinforce women's agency and household welfare in cash systems (Ndoka, 2020). Finally, ensure that women participate in project selection and asset oversight committees so that chosen works reflect local priorities and reduce barriers to participation.
In sum, gendered impacts are central to the effectiveness and equity of food-for-work and cash-for-work. Modalities and delivery systems shape who participates, who benefits and how transfers translate into improved wellbeing. Evidence shows that intentional design and measurement can mitigate time poverty, increase women's control over resources, enhance child outcomes and reduce safety risks. Programmes that embed these features, monitor them and adapt in response to findings are more likely to achieve transformative results rather than simply deliver short term relief (Holmes and Jones, 2010;Berhane et al., 2014).
2.11 Sustainability Debates
Sustainability is a critical dimension of public works programmes because it determines whether interventions deliver lasting benefits beyond the immediate relief provided by wages or food transfers. While food-for-work and cash-for-work (CFW) can protect consumption during crises and create assets, their long-term viability depends on resource allocation, institutional capacity and integration with broader development strategies.
A first sustainability challenge relates to financial and logistical feasibility. FFW programmes require substantial infrastructure for procurement, storage and transportation of food commodities. These systems are often supported by international donors such as the World Food Programme, which raises questions about continuity once external funding declines. The cost of moving large volumes of cereals to remote areas is high, and delays in delivery can undermine programme objectives. Moreover, food transfers typically consist of staple grains, which meet caloric needs but do not allow households to address non-food priorities such as healthcare or education. This narrow focus limits the developmental potential of FFW and reinforces its role as a short-term safety net ratherthan a transformative intervention (World Bank, 2013).
CFW programmes are often considered more sustainable because they inject liquidity into local economies and stimulate demand for goods and services. Cash transfers enable households to purchase diverse foods, pay school fees and invest in small businesses, creating pathways to improved livelihoods. Evidence from Kenya and Ethiopia shows that predictable cash transfers can reduce vulnerability to climate shocks and support savings and productive investments (OPM, 2018; Berhane et al., 2014). However, sustainability is undermined when inflation erodes the real value of transfers or when payment delays disrupt household planning. In Zambia, operational reports from the Ministry of Local Government and Rural Development indicate that network coverage and verification issues caused significant delays in disbursement during the 2024 scale-up, reducing programme effectiveness (MLGRD, 2024). These challenges highlight the need for robust financial infrastructure and contingency planning to maintain reliability.
Another dimension of sustainability concerns asset quality and maintenance. Public works are justified partly on the basis that they create infrastructure such as feeder roads, small dams and soil conservation structures. These assets can deliver long-term benefits by reducing transport costs, improving water availability and enhancing land productivity. However, evaluations in Ethiopia and Kenya reveal that many assets deteriorate quickly due to inadequate technical standards and lack of post-project maintenance (Levine et al., 2024). When maintenance responsibilities and budgets are not specified, assets risk becoming liabilities ratherthan drivers of resilience. Conversely, where projects are well designed and maintained, environmental gains can be substantial. Remote sensing studies in Ethiopia show measurable increases in tree cover in areas with sustained public works, indicating positive contributions to climate adaptation and mitigation (Hirvonen et al., 2022). These findings underscore the importance of integrating operation and maintenance plans into programme design and securing local government commitment to asset upkeep.
Institutional capacity is another determinant of sustainability. Programmes embedded within national systems are more likely to endure than those implemented through ad hoc arrangements. Ethiopia's PSNP illustrates this principle by operating through government structures and incorporating a riskfinancing mechanism that enables rapid scale-up during droughts. This architecture strengthens state capacity for social protection and disaster response, reducing reliance on external actors (World Bank, 2013). Kenya's Hunger Safety Net Programme offers a similar example, with its transition to full government ownership under the National Drought Management Authority and integration of grievance and case management systems (NDMA, 2024). In contrast, Malawi's public works have faced sustainability challenges due to short programme cycles, limited funding and weak institutional linkages, which constrain their ability to deliver lasting benefits (Beegle et al., 2017). For Zambia, embedding cash-for-work within the national social protection frameworkand aligning it with district development plans will be essential for continuity.
Sustainability also depends on policy coherence and integration with complementary interventions. Public works alone cannot transform livelihoods unless they are linked to agricultural extension, input access, financial services and market development. Evidence from Ethiopia shows that households receiving both PSNP benefits and agricultural support achieved larger gains in food security and asset accumulation than those relying on PSNP alone (Berhane et al., 2014). Similarly, cash plus models that combine transfers with services such as financial literacy and climate-smart agriculture training have demonstrated superior outcomes in other contexts (World Bank, 2024). For Zambia, integrating public works with programmes like the Food Security Pack and smallholder support schemes could enhance sustainability by enabling households to leverage wages into productive investments.
Environmental sustainability is another consideration. Public works that rehabilitate degraded land, conserve water and increase tree cover contribute to climate resilience and reduce vulnerability to future shocks. However, these benefits materialise only when assets are technically sound and maintained over time. Programmes that prioritise labour absorption without regard to engineering standards risk creating structures that fail under stress, wasting resources and undermining trust. Incorporating simple site screening tools, technical supervision and community ownership arrangements can improve asset durability and environmental outcomes (WFP, 2018; Niemans, 2024).
Finally, sustainability requires predictable financing. Public works are often scaled up during crises and scaled down when fiscal space tightens, creating uncertainty for households and local governments. Establishing contingency funds and integrating public works into medium-term expenditure frameworks can reduce volatility and support planning. Ethiopia's risk financing facility offers a model for institutionalising scalability within predictable budget envelopes (World Bank, 2013). Zambia could adapt similar mechanisms to ensure that cash-for-work remains available during droughts withoutdisrupting othersocial protection commitments.
In summary, sustainability debates highlight that food-for-work and cash-for-work can deliver more than temporary relief when programmes are designed with attention to financial viability, institutional capacity, asset quality, gender equity and environmental resilience. For Zambia, this means moving beyond emergency-driven interventions toward integrated systems that link public works to development strategies, secure maintenance financing and embed operations within national and district structures. Without these measures, public works risk remaining short-term palliatives rather than transformative instruments for poverty reduction and climate adaptation.
2.12 Theoretical Critiques
The theoretical literature offers important cautions about how food-for-work and cash- for-work are conceived and delivered. These critiques do not negate the protective value of public works. Rather, they clarify the conditions under which programmes risk falling short of their stated objectives and how design and implementation can be strengthened.
Dependency and structural critiques. A longstanding argument is that public works may entrench dependence on external support if they are not anchored in a broader strategy for structural change. When programmes are largely funded and steered by donors, and when local institutional capacity remains weak, households can become locked into recurrent cycles of workfare without substantive shifts in productivity or agency. In such settings, the protective function of transfers is preserved, but the promotive and transformative ambitions are blunted (Devereux and Sabates-Wheeler, 2004). A related critique is that short social protection cycles can crowd out political attention to market integration, rural finance, and service delivery that would expand opportunities beyond episodic labour projects (Devereux and Sabates-Wheeler, 2004).
Capability approach critique. The capability perspective emphasises that development is not only about consumption but about expanding people's real freedoms to do and to be. From this standpoint, public works that focus narrowly on wage or ration delivery without building capabilities in health, education, mobility, and voice risk being palliative rather than transformative. Transfers can protect, but if the programme does not support the conversion of wages or food into durable capabilities, agency remains constrained (Sen, 1999). The implication is that food-for-workand cash-for-work should be linked to complementary services and to institutions that enable people to translate transfers into sustained improvements in the five capitals envisaged by the sustainable livelihoods framework (Sen, 1999).
Political economy and elite capture. Targeting and project selection can be shaped by local power relations. Without transparent rules, public posting of beneficiary lists, open forums for asset selection, and accessible appeals, opportunities may be allocated to better connected households while poorer households are excluded. Evidence from social accountability practice shows that citizen scorecards, information campaigns, and grievance systems reduce these risks when embedded in responsive local administrations (DUS, 2013). Programme manuals that codify grievance handling and time standards for complaint resolution are therefore part of the theory of change, not a peripheral administrative add-on (World Bank, 2017).
Feminist and gender critiques. Public works often assume that labour is equally accessible to women and men, yet unpaid care burdens, safety risks, and norms about control over cash can limit women's participation and benefits. When work cycles are rigid, worksites distant, or tasks heavy, women experience time poverty and diminished net welfare. Reviews of gender and social protection show that direct payments to women, flexible hours, proximate worksites, safe workplace protocols, and childcare provision improve inclusion and child outcomes (Holmes and Jones, 2010). More broadly, a feminist critique argues that programmes should recognise, reduce, and redistribute unpaid care if they are to be transformative rather than reinforcing existing inequalities (Kabeer, 2008).
Labour market distortion critique. Public works can alter local labour markets. If wages are set above casual rates, non-poor workers may be attracted and private employment may be displaced; if wages are set too low, self-targeting works but poverty relief is weak. Design theory suggests setting wages below local casual rates to minimise diversion from private jobs and to preserve targeting, while timing work in the lean season to maximise the net transfer when forgone earnings are lowest (Ravallion, 1999). Recent experimental syntheses find short-run earnings gains during participation, but limited sustained labour market effects after programmes end unless households save or invest in self-employment during the intervention, reinforcing the importance of wage, duration, and savings pathways (J-PAL, 2024).
Governance capacity and asset quality critique. A recurrent finding is that assets may deliver limited livelihood benefits where engineering standards are weak, site selection is poor, or operation and maintenance are unfunded. Studies that revisited watershed and water assets in Ethiopia and Kenya found that labour absorption sometimes took precedence over technical quality, with rapid deterioration and negligible impacts on livelihoods (Levine et al., 2024). The counterpoint is that when design and maintenance are sound, environmental and productivity gains can be significant, as shown by increases in tree cover in areas with sustained conservation works (Hirvonen et al., 2022). The critique therefore targets implementation fidelity rather than the concept of public works per se (Levine et al., 2024).
Behavioural critique. Scarcity imposes a cognitive tax that narrows attention and reduces planning capacity. Under such conditions, payment reliability and predictability are as important as transfer size, because delays and uncertainty force households back into negative coping strategies even when nominal amounts look adequate. This lens explains why delivery systems are core to programme impact and why predictable schedules and simple recourse mechanisms can have outsized welfare effects relative to minor changes in transfer amounts (Mullainathan and Shafir, 2013).
Sustainability and inflation critiques. Cash-for-work can be undermined by inflation when transfer values are not adjusted to local prices or when purchasing power is eroded by payment delays. Food-for-work can buffer price spikes but faces high logistics costs and limited dietary diversity. Sustainability debates therefore call for periodic wage review, risk financing for scale up, and integration of public works within medium-term expenditure frameworks to reduce volatility and allow maintenance planning (World Bank, 2017). Without predictable finance and clear maintenance responsibilities, assets risk becoming short-lived and households remain exposed to the next shock (World Bank, 2017).
Value for money and opportunity cost critiques. Public works combine transfer delivery with asset creation, but if project selection emphasises labour absorption over economic value, programmes may be less cost-effective than direct transfers or than targeted investments in health, education, or market access. Comparative assessments that include delivery costs, leakage control, asset durability, and welfare gains are needed to justify portfolio choices;where such analysis is absent, opportunity costs may be opaque and debate polarised (World Bank, 2017).
Implications for Zambia. The theoretical critiques suggest concrete actions. First, link public works to capability-expanding services. Cash-for-work should be coupled with agricultural extension, input access, financial literacy, and market integration so that households convert transfers into durable assets and skills (Sen, 1999; Berhane et al., 2014). Second, institutionalise transparency and recourse. Public posting of lists, open project selection forums, and accessible grievance with time standards should be enforced across districts (DIIS, 2013; World Bank, 2017). Third, calibrate wages and timing. Wage setting should preserve self-targeting while ensuring meaningful relief, and work should be scheduled in lean periods to maximise net benefits (Ravallion, 1999). Fourth, prioritise asset quality and maintenance. Simple site screening, technical supervision, and financed operation and maintenance are essential to avoid rapid deterioration and to realise environmental co-benefits (Levine et al., 2024; Hirvonen et al., 2022). Fifth, protect women's time and agency. Payment mechanisms, task assignment, and safe worksite protocols should be designed to recognise care responsibilities and strengthen women's control over resources (Holmes and Jones, 2010; Kabeer, 2008). Finally, monitor delivery reliability and inflation. Payment timeliness dashboards, grievance tracking, and periodic wage reviews will reduce the behavioural and macro risks that otherwise erode programme effectiveness (Mullainathan and Shafir, 2013;World Bank, 2017).
In sum, theoretical critiques do not reject food-for-work and cash-for-work. They insist that public works be embedded in systems that expand capabilities, honour rights, guard against capture, and deliver technically sound assets with financed maintenance. Programmes that take these critiques seriously are more likely to move beyond short-term relieftoward resilientand equitable livelihoods.
2.13 Comparative Table: Gender and Sustainability Dimensions
This unit presents a literature-based synthesis of gendered impacts and sustainability issues in food-for-work and cash-for-work across Ethiopia, Kenya, Malawi and Zambia. The table summarises salient patterns, followed by an interpretive narrative that draws practical lessons for programme design, asset selection and delivery systems. The emphasis is on evidence that is robust and relevant to policy decisions, with citations provided where necessary to anchor key claims.
The table shows Gendered impacts, sustainability challenges and key lessons in food-for-work and cash-for-work
Illustrations are not included in the reading sample
Source notes: Ethiopia references draw on longitudinal evaluation and environmental analysis for public works and programme architecture (Berhane et al., 2014; Hirvonen et al., 2022; World Bank, 2013). Kenya references draw on impact evaluation, operational guidance and field manuals for water works (OPM, 2018;WFP, 2018; Ndoka, 2020). Malawi references draw on randomised evaluation of public works outcomes and delivery constraints (Beegle et al., 2017). Zambia references draw on operational guidance and budget analysis related to cash-for-workand social protection allocations (MLGRD, 2024; UNICEF, 2023).
Interpretive narrative
Across these four contexts, gendered impacts and sustainability outcomes hinge on several common determinants. A first determinant is payment timeliness and predictability. Where cash is delivered on time and through accessible channels, households plan purchases and avoid distress strategies; where delays occur, welfare gains erode and gender inequalities can widen if men retain control over sporadic payments. Kenya's evolution under government systems shows how case management and grievance can improve reliability and correct errors at scale, which is central to enabling women to use transfers for nutrition and schooling when they are designated payees in household accounts (OPM, 2018).
A second determinant is asset choice and maintenance. Public works deliver durable benefits when assets match local risks and capacities. Ethiopia's watershed works demonstrate measurable environmental improvements where technical standards and maintenance are upheld, while retrospective audits warn that rapid labour absorption without adequate engineering can produce assets that deteriorate quickly and yield limited livelihood gains (Hirvonen et al., 2022; Levine et al., 2024). In Kenya, simple site suitability screening for water harvesting improves performance and reduces failure, a practice that counties can systematise to raise durability and value for money in public works portfolios (WFP, 2018).
A third determinant is gender sensitive design. Flexible scheduling, proximate worksites, lighter tasks where appropriate and safe worksite protocols reduce time poverty and raise net benefits for women. Evidence from Kenya indicates that direct payment to women enhances decision making, especially when complemented by financial literacy, while reviews of gender and social protection consistently point to improved child outcomes when women control resources (Ndoka, 2020; Holmes and Jones, 2010). Ethiopia's experience shows that codifying gender provisions in programme manuals is necessary but not sufficient;delivery fidelity and local norms determine actual inclusion, which calls for routine sex disaggregated monitoring and responsive management (Berhane et al., 2014).
A fourth determinant is modality alignment to markets. Food transfers are appropriate where markets are thin and prices volatile, protecting consumption during crises;cash transfers offer flexibility and developmental potential where markets function and supply responds. Ethiopia's calibrated approach across districts and seasons remains a practical model, supported by risk financing that scales assistance during drought and sustains predictability within government systems (World Bank, 2013). Malawi's results caution that cash for work will not improve food security if wages are low, exposure is short and payments are delayed, even when work is scheduled in the lean season, which is a reminder that design features must match local labour markets and prices to achieve intended outcomes (Beegle et al., 2017).
A fifth determinant is macroeconomic context and transfer adequacy. Inflation reduces real purchasing power and can nullify intended benefits unless transfer values are reviewed periodically and indexed where feasible. Zambia's budget analysis highlights cost of living pressures that argue for wage adjustments or for linking wages to services that support productive investment, such as input access and financial inclusion, so households can convert cash into self-driven gains despite price volatility (UNICEF, 2023).
These determinants suggest practical steps for policy and programme management. First, institute payment timeliness dashboards and grievance reporting so managers can track delays by district and correct bottlenecks rapidly, a measure that supports both equity and effectiveness in cash systems. Second, require site screening and technical standards for works, with funded operation and maintenance lines in district budgets and clear responsibility for upkeep, thereby protecting asset value over time. Third, embed gender provisions in operations, including direct payment options to women, childcare solutions at larger worksites and routine sex disaggregated monitoring, and publicise safe worksite protocols with confidential grievance channels. Fourth, conduct response analysis at planning stage to decide whether food or cash best fits market conditions, and consider cash plus packages where complementary services can raise sustained wellbeing. Finally, review wages periodically to protect real value and maintain targeting, scheduling works in lean periods to maximise net household benefits.
For Zambia, the comparative evidence supports a coherent and integrated approach. Stabilising mobile payment systems and identity verification will reduce delays and improve trust. Selecting maintainable climate relevant assets, such as small-scale irrigation and soil and water conservation, will generate resilience dividends if maintenance responsibilities and budgets are specified from the outset. Ensuring women's access to payments and leadership roles in project selection will increase inclusion and improve outcomes for children. Linking public works with agricultural extension and financial literacy will help households convert wages into lasting gains.
These steps align with lessons from Ethiopia, Kenya and Malawi and will move public works from short term relief toward sustained resilience and equitable livelihoods in Zambia, provided that programme architecture remains anchored in national systems and monitored for fidelity to design (World Bank, 2013; OPM, 2018; Beegle et al., 2017).
2.14 Conceptual Synthesis and Hypotheses
This unit consolidates insights from the preceding literature and sets out a conceptual framework to guide empirical analysis of food for work and cash for work in Zambia. It integrates the sustainable livelihoods perspective with evidence on modality choice, governance, gender, asset quality and macro conditions, and proposes testable hypotheses about programme effectiveness and equity.
Conceptual anchors. The sustainable livelihoods framework positions households within a vulnerability context and emphasises how they combine capitals that are human, social, financial, physical and natural to pursue strategies and outcomes (DFID, 1999). Public works influence these capitals through two principal channels. First, transfers smooth consumption and can relax liquidity constraints, thereby supporting food security and enabling small investments. Second, assets created through works can reduce costs, expand access and restore ecosystems, thereby altering returns to household effort and investment. Modality choice matters because food transfers directly protect caloric intake when markets are thin, while cash permits flexible allocation and can generate local multipliers where supply responds (Gentilini, 2016; Taylor et al., 2016). Programme design features such as wage adequacy, payment timeliness, technical standards and financed operation and maintenance determine whether short term benefits convert into sustained livelihood gains (World Bank, 2017).
Core programme inputs. The framework identifies a set of design and delivery inputs that drive outcomes. These include modality calibration to market conditions, wage setting aligned to poverty relief and self-targeting, payment reliability, transparent targeting and grievance, asset selection and site screening, technical supervision and financed maintenance, gender sensitive operations and linkage to complementary services such as agricultural extension and financial inclusion (Ravallion, 1999; Holmes and Jones, 2010;World Bank, 2017). National system embedding and risk financing mechanisms further enhance predictability and scalability during shocks, as shown in Ethiopia and Kenya (World Bank, 2013; OPM, 2018).
Mediating mechanisms. The framework posits several mediators through which public works affect household outcomes. Liquidity and predictability influence planning and reduce negative coping strategies, consistent with behavioural evidence on scarcity and cognitive load (Mullainathan and Shafir, 2013). Asset quality and maintenance shape physical and natural capital pathways by lowering transport costs, improving water availability and enhancing soil productivity, with observed environmental co benefits where conservation works are sustained (Hirvonen et al., 2022). Gender provisions influence agency by affecting women's participation, control over transfers and decision making, which in turn affects nutrition and schooling choices (Holmes and Jones, 2010). Governance mechanisms influence trust, inclusion and error correction, which determine whether benefits reach intended households and whether assets remain functional (World Bank, 2017).
Contextual moderators. Market function, seasonality, inflation, rural labour markets and local administrative capacity moderate impacts. Food transfers are appropriate in thin markets and during price spikes, while cash performs better where markets are accessible and supply can respond. Inflation erodes real value unless wages are reviewed periodically. Lean season timing increases net benefits by reducing forgone earnings. Administrative capacity and payment systems condition timeliness and scalability, as evidenced by stronger results under government systems in Kenya and under risk financing in Ethiopia, and weaker results under delays and short exposure in Malawi (OPM, 2018;World Bank, 2013;Beegle et al., 2017).
Outcomes and longer run effects. In the near term, expected outputs include consumption smoothing, diversified diets, small savings and investment, reduced distress sales and completion of functional assets. Over the longer run, anticipated outcomes include income diversification, productive asset accumulation, improved resilience to climate shocks, restoration of natural capital and enhanced social inclusion. The durability of these outcomes depends on asset lifecycle performance and the strength of system embedding and maintenance finance (Levine et al., 2024).
Risks and trade-offs. The framework recognises risks of elite capture and exclusion if targeting and grievance are weak, labour market distortion if wages are set above casual rates, dependency if programmes are not linked to capability expanding services, and opportunity costs if asset selection prioritises labour absorption overvalue. It also notes inflation risks for cash and logistics burdens for food. These risks justify investment in governance and periodic response analysis to confirm modality appropriateness and value for money (Ravallion, 1999; World Bank, 2017).
Hypotheses for empirical testing. Building on the synthesis, the study advances the following hypotheses to structure measurement and analysis in Zambia.
1. Modality effectiveness hypothesis. In districts with functioning markets and adequate supply, cash-for-work delivers larger improvements in food security and household flexibility than food-for-work, conditional on timely payments and wage adequacy (Gentilini, 2016).
2. Crisis protection hypothesis. In districts experiencing acute market stress and price volatility, food-for-work protects consumption more effectively than cash-for-work during the crisis window, with effects diminishing as markets stabilise (World Bank, 2013).
3. Payment reliability hypothesis. Payment timeliness has a positive association with reductions in negative coping strategies and with small savings and investments, independent of transfer size, consistent with behavioural constraints under scarcity (Mullainathan and Shafir, 2013).
4. Asset quality hypothesis. Households in areas where assets meet technical standards, are well sited and have financed maintenance exhibit larger gains in natural capital restoration and productivity than households in areas with similar wage transfers but lower asset quality (Hirvonen et al., 2022; Levine et al., 2024).
5. Gender agency hypothesis. Direct payment to women and flexible, proximate worksites are positively associated with women's control over resources, child nutrition and school fee payments, relative to otherwise similar operations without these provisions (Holmes and Jones, 2010).
6. Integration hypothesis. Linking public works to complementary services such as agricultural extension and financial literacy produces stronger gains in income diversification and resilience than public works alone (Berhane et al., 2014).
7. Local economy hypothesis. In areas with elastic supply, cash-for-work generates positive local economy spill overs on non-beneficiary households and traders, while in areas with binding supply constraints price effects attenuate gains, implying context specific multiplier sizes (Taylor et al., 2016;Thome et al., 2016).
8. Inflation and adequacy hypothesis. Real wage adequacy is a significant predictor of programme effectiveness;periods of high inflation without wage review are associated with weaker outcomes on consumption smoothing and investment (UNICEF, 2023).
9. Seasonality hypothesis. Scheduling work during the lean season is associated with higher net benefits due to lower forgone earnings, relative to similar programmes scheduled outsidethe lean season (Ravallion, 1999).
10. Governance and grievance hypothesis. Districts with transparent selection, public posting of lists and functioning grievance systems exhibit lower exclusion and inclusion errors and higher beneficiary satisfaction, controlling for other factors (World Bank,2017).
Operational implications for Zambia. The framework implies practical choices for programme management. Response analysis should inform whether a district cycle deploys food or cash. Wage levels should be reviewed to protect real value and maintain self-targeting. Mobile enabled payment systems with tracked timeliness should be expanded to reduce delays noted in the recent scale up (MLGRD, 2024). Asset menus should prioritise climate relevant works such as small irrigation and soil and water Page 73 of 278 conservation, with simple site screening, clear technical standards and maintenance lines in district budgets. Gender provisions should include options for direct payment to women, flexible schedules and safe worksites. Programme operations should be embedded in national systems with grievance dashboards and routine reporting to sustain transparency and trust (OPM, 2018;World Bank, 2013).
By specifying inputs, mediators, moderators and outcomes, and by articulating testable hypotheses, this conceptual synthesis provides a structured foundation for evaluating whether food-for-work and cash-for-work in Zambia achieve protective aims and advance promotive and transformative goals. It aligns with the sustainable livelihoods framework and with comparative evidence from Ethiopia, Kenya and Malawi, while grounding the study in Zambia's institutional and market realities (DFID, 1999; Beegle etal., 2017).
11. 5 Operational Implications and Measurement Strategy
This unit distils practical implications from the literature and sets out a measurement strategy to evaluate food-for-work and cash-for-work in Zambia. It translates conceptual insights into operational choices and proposes indicators, data sources and analytical approaches that can credibly test the hypotheses articulated in the previous unit and answer core policy questions about effectiveness, equity and sustainability.
Operational priorities for programme design. The first priority is to align modality with market conditions through routine response analysis. Where local staple markets are thin and prices volatile, food transfers protect consumption during crisis windows. Where markets function reasonably and supply can respond, cash provides flexibility and supports diversified spending and small investments that households value (World Bank, 2013). The second priority is to protect the real value of transfers by reviewing wage levels periodically in line with inflation and by scheduling works in the lean season to maximise net benefits when forgone earnings are lowest (Ravallion, 1999). The third priority is to stabilise payments through mobile channels, identity verification and tracked timeliness so households can plan and avoid distress strategies; delays have repeatedly been shown to erode programme gains in both evaluation and operational accounts (OPM, 2018; MLGRD, 2024). The fourth priority is to ensure that assets are chosen for local risk reduction and productivity, are sited using simple screening tools, constructed to technical standards and assigned financed maintenance responsibilities from the outset to avoid rapid deterioration and loss of livelihood impact (Levine et al., 2024). The fifth priority is to embed gender-sensitive operations, including accessible payment options for women, flexible scheduling, safe and proximate worksites and routine sex-disaggregated monitoring, given strong evidence that these measures increase women's control over resources and improve child outcomes (Holmes and Jones, 2010).
Measurement domains and core indicators. The measurement strategy focuses on six domains: consumption and food security; income diversification and small investments; asset lifecycle performance;delivery reliability and grievance; gender and inclusion;and local economy effects.
1. Consumption and food security. Primary indicators include months of adequate food provisioning in the past year, food consumption scores, dietary diversity and the incidence of distress behaviours such as sale of productive assets or reduced meal frequency. These should be measured at baseline and follow-up across seasons to capture lean season dynamics. Longitudinal evidence in Ethiopia shows that sustained participation increases months of adequate food and reduces reliance on negative coping strategies, which justifies similar measurement in Zambia (Berhane et al., 2014).
2. Income diversification and small investments. Track savings balances, purchase of livestock, acquisition of agricultural inputs, microenterprise starts and school fee payments. These indicators test whether cash increases financial capital and whether households convert wages into productive assets. Evidence from evaluations of cash systems in arid counties and from PSNP indicates that predictable cash supports such investments when markets function and payments are on time (OPM, 2018; Berhane et al., 2014).
3. Asset lifecycle performance. For a representative sample of public works, record compliance with technical standards at completion, measure functionality at six and twelve months, document operation and maintenance arrangements and costs, and assess simple productivity or service outcomes such as water availability, travel time reduction or erosion control. Retrospective studies in Ethiopia and Kenya caution that assets deliver limited benefits where engineering quality and maintenance are weak; systematic auditing in Zambia will close an identified evidence gap (Levine et al., 2024).
4. Delivery reliability and grievance. Monitor payment timeliness with district level dashboards that show the proportion of payments made on schedule and the distribution of delays. Track grievance submissions by type, resolution rates and median resolution times. Kenya's experience demonstrates that case management and grievance systems are central to credibility and timely correction at scale, and operational reports in Zambia have already highlighted connectivity and verification bottlenecks that such dashboards can help resolve (OPM, 2018; MLGRD, 2024).
5. Gender and inclusion. Report sex-disaggregated participation, payment receipt, decision making over transfers, time use and child wellbeing outcomes such as school attendance and dietary diversity. Include qualitative modules on safety, dignity and workplace conditions, and verify whether women are represented in project selection and asset oversight committees. Reviews of gender and social protection provide the rationale for these indicators and suggest that design changes can shift intra household bargaining in ways that matter for programme impact (Holmes and Jones, 2010).
6. Local economy effects. Use simple local economy-wide tools to measure price changes in a staple basket, sales volumes of small traders and casual wage rates, and apply difference-in-differences where feasible between treatment and comparison localities. Evidence from African cash transfer studies shows that cash can generate positive multipliers when supply is elastic, but that price effects attenuate gains under constraints; documenting these dynamics in Zambia will support transfer calibration and asset selection (Taylor et al., 2016).
Data sources and study design. The preferred design is a mixed methods panel study covering districts that implement food-for-work and cash-for-work, with matched comparison areas where possible. Household surveys should be conducted at baseline, midline and end-line, timed to capture lean season and post-harvest periods. Administrative data from payment systems, grievance registers and asset completion reports should be integrated into a management information system to support routine monitoring and evaluation. Remote sensing can complement field audits for environmental outcomes such as vegetation cover change in watersheds with sustained conservation works (Hirvonen et al., 2022). Where randomised allocation is infeasible, quasi-experimental methods such as propensity score matching combined with difference-in-differences can strengthen attribution, while process tracing and qualitative fieldwork explain causal mechanisms and implementation bottlenecks.
Analytical approach. Hypotheses from Unit 2.14 guide multivariate analysis. For consumption outcomes, regress months of adequate food on modality, payment timeliness and wage adequacy, controlling for household characteristics and market indicators. For investment and diversification, estimate the association between cash receipt reliability and savings or asset acquisition, with interaction terms for market accessibility. For asset performance, relate functionality and simple productivity outcomes to engineering compliance at completion and the presence of financed maintenance. For gender agency, test whether direct payment to women and flexible scheduling predict greater control over spending and improved child indicators. For local economy effects, estimate price and wage changes in treated versus comparison areas and interpret coefficients in light of observed supply constraints (Ravallion, 1999; Tayloretal., 2016).
Governance and reporting arrangements. To sustain transparency, publish district scorecards with payment timeliness, grievance resolution and asset functionality. Public posting of beneficiary lists and open project selection forums reduce risks of elite capture and increase citizen trust; these practices are part of credible programme architecture rather than peripheral administrative tasks (World Bank, 2017). A national technical note should codify standards for site screening, engineering supervision and operation and maintenance budgeting to ensure convergence across districts and to improve value for money (Levine et al., 2024).
Risk management and ethical safeguards. Anticipated risks include inflation shocks that erode purchasing power, payment system failures in low connectivity areas, and potential safety concerns at worksites. Mitigation includes periodic wage review to protect real value, contingency protocols for offline payment verification, safe worksite procedures with confidential grievance channels and attention to care responsibilities through flexible scheduling. Behavioural evidence indicates that predictability and simple recourse mechanisms are critical under scarcity, so clear communication of payment schedules and grievance processes should be part of routine operations (Mullainathan and Shafir, 2013).
Policy interfaces. Findings should feed directly into social protection planning, disaster risk financing and district development plans. For example, evidence that cash performs better in accessible markets with reliable payment systems supports budgeting for mobile platforms and identity verification, while evidence that watershed works raise natural capital when maintenance is financedjustifies ringfenced lines for operation and maintenance in district budgets (World Bank, 2013). Where gender agency indicators improve under direct payment and flexible schedules, programme manuals should mandate these features across districts (Holmes and Jones, 2010).
2.16 Chapter Synthesis and Implications for Zambia
This unit distils the core arguments from the literature review and articulates their practical meaning for designing, delivering and evaluating food-for-work and cash-for- work in Zambia. It brings together conceptual insights, comparative evidence and governance lessons, and clarifies how these elements inform the thesis research strategy and policy recommendations.
What the literature establishes
The sustainable livelihoods framework provides a coherent lens for understanding how transfers and assets interact with household capitals that are human, social, financial, physical and natural within a vulnerability context (DFID, 1999). Food-for-work protects consumption when markets are thin and prices volatile, while cash-for-work offers flexibility and developmental potential where markets function and supply can respond. Global comparative work shows that cash often matches or outperforms in kind assistance on key wellbeing indicators, but modality choice must be grounded in response analysis and market diagnostics rather than doctrine (Gentilini, 2016). At the same time, wage setting, seasonal timing, reliability of payments and quality and maintenance of assets are decisive for converting short term relief into sustained livelihood gains (Ravallion, 1999; World Bank, 2017).
Comparative cases sharpen these conclusions. Ethiopia demonstrates calibrated modality choice, predictable delivery through national systems and measurable environmental gains from well-designed watershed works, including observed increases in tree cover where conservation is maintained (World Bank, 2013;Hirvonen et al., 2022). Kenya shows how predictable cash under government systems cushions climate shocks and how simple technical guidance for water harvesting and road linked assets can improve durability when counties plan for maintenance (OPM, 2018). Malawi provides a cautionary tale that cash-for-work will not automatically improve food security where wages are low, exposure is short and payments are delayed, even when programmes are scheduled in the lean season (Beegle et al., 2017). Retrospective audits in Ethiopia and Kenya warn that asset portfolios can deliver limited livelihood benefits when labour absorption replaces engineering quality and when operation and maintenance are unfunded (Levine et al., 2024).
Implications for Zambia's programme architecture
First, modality must fit markets. In districts with reliable supply and accessible markets, cash for work should be the default because households value choice and local economy spill overs can materialise when supply responds. In districts experiencing acute market stress during drought, food transfers remain appropriate to protect consumption until markets stabilise. This hybrid approach mirrors successful practice in Ethiopia and avoids false universals about cash orfood in all circumstances (World Bank, 2013).
Second, systems must fit scale. Payment timeliness is a core determinant of impact because households plan purchases and avoid distress strategies when transfers are predictable. Zambia's recent scale up showed that network coverage and identity verification can slow disbursement. Mobile enabled payments, strengthened verification and district level timeliness dashboards are therefore not administrative extras but central impact drivers. Kenya's experience confirms that embedded case management and grievance procedures enable timely correction at scale (OPM, 2018).
Third, assets must fit landscapes. Public works should prioritise climate relevant assets with demonstrable livelihood benefits, such as small-scale irrigation, soil and water conservation and cost-effective road spot improvements that reduce travel time and connect markets. Simple site screening tools, basic technical standards and financed operation and maintenance lines in district budgets are necessary to protect asset value. The evidence that tree cover increases where conservation is sustained, and that assets deteriorate quickly where maintenance is absent, supports a shift from construction count to lifecycle performance metrics (Hirvonen et al., 2022;Levine et al., 2024).
Fourth, gender must be mainstreamed through operations. Direct payment options for women, flexible and proximate work schedules, safe worksite protocols and routine sex disaggregated monitoring are pragmatic steps that reduce time poverty and strengthen women's control over transfers. Reviews of social protection show that such measures improve nutrition and schooling outcomes, particularly in cash systems where control over money is contested (Holmes and Jones, 2010).
Fifth, real wage adequacy must be protected. Inflation reduces purchasing power and can nullify intended benefits if wage levels are not reviewed periodically. Budget analysis indicates that cost of living pressures is significant. Periodic wage reviews and lean season scheduling will protect net benefits and sustain self-targeting without distorting local labour markets (UNICEF, 2023;Ravallion, 1999).
Implications for measurement and /earning
The thesis adopts a mixed methods approach focused on six domains that the literature identifies as policy critical. These are consumption and food security, income diversification and small investments, asset lifecycle performance, delivery reliability and grievance, gender and inclusion, and local economy effects. For consumption, months of adequate food and distress behaviour indicators will be tracked because sustained participation has been associated with improvements in similar measures in Ethiopia (Berhane et al., 2014). For investment and diversification, savings, small enterprise activity and input purchases will be recorded to test whether cash increases financial capital and self-driven productive activity under reliable payment conditions (OPM, 2018). For assets, functionality audits at six and twelve months and simple service or productivity measures will link engineering compliance and maintenance to observed outcomes, in line with findings on asset durability and livelihood impact (Levine et al., 2024). For delivery, district dashboards will report payment timeliness and grievance resolution to support responsive management. For gender, sex disaggregated indicators and qualitative modules will capture participation, control and workplace conditions. For local economy effects, basic price and wage tracking and difference in differences analysis will estimate multipliers and inflation pressures where feasible (Tayloretal., 2016).
Addressing persistent risks and trade-offs
Theoretical critiques identify risks that must be actively managed. Elite capture and exclusion risks argue for public posting of lists and accessible grievance with time standards to protect equity (World Bank, 2017). Behavioural evidence on scarcity suggests that predictable schedules and simple recourse mechanisms can have outsized welfare effects relative to marginal changes in transfer amounts (Mullainathan and Shafir, 2013). Opportunity cost arguments call forcomparative value for money analysis across programme designs that consider wage transfer, asset quality, leakage control and welfare outcomes, rather than focusing only on coverage figures (World Bank, 2017). Labour market distortion concerns demand wage setting that preserves selftargeting while providing meaningful relief, with lean season timing to maximise net benefits and minimise displacement (Ravallion, 1999).
Policy pathway for Zambia
The literature supports a policy pathway with four immediate steps. One, institutionalise response analysis at planning stage to decide cash, food or cash plus by district and season, using market and price diagnostics grounded in early warning information. Two, stabilise payment systems with mobile channels, strengthened verification and transparent timeliness reporting, and embed grievance dashboards at district level to surface and resolve problems quickly. Three, select and supervise maintainable climate relevant assets using simple site screening and standards, and budget for operation and maintenance from the outset, with accountability for asset performance included in district scorecards. Four, mainstream gender provisions in operations and measure outcomes, not only participation, to ensure women benefit equitably and household welfare improves in nutrition and schooling.
These steps align with the strongest comparative evidence and with Zambia's recent operational experience. Countries that have invested in modality alignment, systems reliability and asset quality, while intentionally designing for gender and monitoring delivery, have moved public works beyond short term relief toward resilient and equitable livelihoods (World Bank, 2013;OPM, 2018). The thesis will test whether similar choices produce measurable gains in the Zambian context and will generate the evidence needed to refine programme operations and budgets accordingly.
2.17 Limitations and Ethical Considerations
This unit clarifies the limitations of the literature reviewed in Chapter 2 and sets out ethical principles that will guide the study of food for work and cash for work in Zambia.
Recognising constraints and ethical obligations strengthens the credibility of findings and helps ensure that recommendations are feasible, equitable and aligned with good practice.
Scope and coverage limitations
The review prioritised peer-reviewed studies, authoritative programme evaluations and operational guidance from large government systems. This approach necessarily concentrates on a small number of flagship programmes with strong documentation, notably Ethiopia's Productive Safety Net Programme and Kenya's Hunger Safety Net Programme, alongside a rigorous trial from Malawi. While these sources provide high quality evidence, they may not capture the full diversity of implementation models across sub-Saharan Africa and within Zambia's districts. As a result, the review places less weight on small pilots, unpublished field notes and grey literature with limited methodological detail. The emphasis on a few systems creates a risk that design lessons reflect conditions specific to those contexts, including the presence of national delivery platforms, risk financing mechanisms and dedicated evaluation budgets.
Heterogeneity of programme designs and contexts
Public works vary widely across countries and within countries in objectives, wage levels, project menus, duration, payment mechanisms and the degree of system embedding. Differences in market access, price dynamics, agricultural calendars and local governance capacity further influence outcomes. Evidence drawn from Ethiopia illustrates gains under calibrated modality and strong government systems, whereas Malawi shows weaker results where wages were low, exposure short and delivery unreliable (Berhane et al., 2014;Beegle et al., 2017). These contrasts caution against transferring findings without adaptation to Zambia's market, administrative and fiscal realities. They also highlight the need for response analysis to set modalities by district and season, rather than adopting uniform rules across heterogeneous settings (World Bank,2013).
Measurement and attribution constraints
Many studies rely on observational designs or mixed methods evaluations with limited scope for causal inference, especially on long-run livelihood outcomes such as income diversification, asset accumulation and resilience to climate shocks. Even when randomised or quasi-experimental methods are used, spill overs, price effects and labour market changes can complicate attribution. Local economy multipliers from cash transfers depend on supply responsiveness and may be attenuated by inflation, which is rarely measured at the village level with high frequency. Similarly, environmental benefits from watershed works require long panels and remote sensing to detect changes in vegetation and erosion;where these are not available, asset performance may be inferred from engineering compliance rather than measured outcomes (Hirvonen et al., 2022). These constraints reinforce the need for a Zambia-specific measurement strategy with panel data, asset audits and basic local economy tracking.
Publication and reporting bias
Programmes with strong documentation are more likely to be included in academic reviews and international guidance, potentially over-representing successful systems or well-funded initiatives. Conversely, negative or null results may be under-reported outside formal evaluations. Malawi's trial is an important exception because it was designed to detect intended benefits and documented disappointing outcomes, but such studies remain rare in the region (Beegle et al., 2017). To mitigate bias, the thesis will incorporate administrative records, grievance logs and independent fieldwork in Zambia to triangulate findings beyond published evaluations.
Generalisability to Zambia
Zambia's public works are scaling within a context of recent drought, evolving payment systems and inflation pressures. Much of the strongest comparative evidence comes from countries that have invested for years in system embedding, case management and risk financing. While these lessons are valuable, they must be adapted to Zambia's fiscal constraints, network coverage and district capacity. The review therefore treats comparative evidence as directional rather than prescriptive and uses it to frame hypotheses and operational choices that will be tested with Zambia-specific data (OPM, 2018;MLGRD, 2024).
Asset lifecycle evaluation gaps
A recurrent limitation in public works evidence is weak documentation of operation and maintenance arrangements and costs. Retrospective studies warn that assets deteriorate where labour absorption is prioritised over engineering quality and where maintenance is unfunded (Levine et al., 2024). In the absence of systematic lifecycle evaluation, conclusions about long-run productivity and resilience effects remain tentative. The thesis responds by proposing simple asset audits at six and twelve months and by recording maintenance responsibilities and budgets to link technical standards to observed service and productivity outcomes.
Ethical principles for field research
The study will adhere to established ethical standards. Informed consent will be obtained from all participants in household surveys and qualitative interviews, with clear communication about research purpose, voluntary participation and the right to withdraw without consequence. Confidentiality will be protected through anonymised identifiers, secure data storage and restricted access to datasets. The principle of do no harm will guide instrument design and field protocols, with careful attention to the timing of interviews during the lean season to avoid exacerbating burdens on households.
Safeguarding is central. Enumerators will receive training on respectful conduct, sensitive questioning and referral guidance where respondents disclose distress or risk. Research activities will avoid interfering with payment processes or work schedules. Where site visits occur at public works, safety protocols will be followed, including engagement with local supervisors to ensure that routine operations are not disrupted.
Equity and inclusion
The study will ensure equitable representation of women, young people, older persons and persons with disabilities in sampling frames. Instruments will include modules on time use, control over transfers, safety and workplace conditions to capture gendered experiences that are often invisible in standard consumption measures. Where feasible, interviews will be conducted at times and locations that minimise travel and care burdens for women and that respect local norms. These steps align with evidence that gender-sensitive design and measurement are necessary to understand and improve programme equity (Holmes and Jones, 2010).
Community engagement and feedback
Local leaders, public works supervisors and beneficiary committees will be briefed on the study and invited to provide feedback on instruments and field plans. Findings will be shared in accessible formats with district authorities and community representatives to promote learning and accountability. At the systems level, the study will encourage the use of grievance dashboards and public posting of beneficiary lists and project selection decisions, recognising that transparent recourse mechanisms are integral to ethical and effective programme management (World BankGRS, 2021).
Data governance and responsible use
Administrative data on payments and grievances will be handled under data-sharing agreements with responsible ministries and local authorities. These agreements will specify permissible uses, retention periods and confidentiality protections. Any linked analysis that uses administrative and survey data will follow privacy by design principles, with de-identification and aggregation to prevent re-identification of households or individuals.
Limitations that the study cannot fully overcome
Even with careful design, the study may face non-response in very remote areas or during peak agricultural periods, measurement error in self-reported income and consumption, and limited ability to detect small changes in local prices or wages. The study will document such limitations transparently, apply sensitivity checks and triangulate with administrative records where possible. It will also acknowledge that short evaluation windows may not capture longer run gains from assets or from strengthened community institutions.
Ethics o versight and appro vals
The research will seek approval from relevant institutional review boards and comply with national research regulations. Collaboration with district offices will ensure alignment with local protocols for community entry and data collection. These measures, together with consent and confidentiality practices, will anchor the study in ethical norms recognised bygovernmentand international partners.
3.1 CHAPTERS METHODOLOGY
3.2 Research Paradigm
Overview and purpose
This study evaluates food-for-work and cash-for-work in Zambia. It engages outcomes that are quantifiable, such as months of adequate food, dietary diversity and small investments, and experiences that are interpretive, such as perceptions of fairness, gendered burdens, and trust in payment systems. To hold both kinds of knowledge in view, the study is grounded in pragmatism. Pragmatism offers a coherent philosophical basis for mixed-methods inquiry that privileges problem solving, practical consequences and usefulness for action. It rejects rigid either or positions that force researchers to choose between only numbers or only narratives and invites the integration of diverse forms of evidence when that integration helps to answer the research questions and to inform policy and practice.
Philosophical roots and core claims
Pragmatism emerged in late nineteenth century philosophythrough the work ofCharles Sanders Peirce, William James and John Dewey. At its core is the claim that the meaning and value of ideas are found in their observable consequences for inquiry and action, rather than in abstract metaphysical debates detached from experience (Peirce, 1878). William James popularised the view that truth is what works in lived practice and that knowledge develops through testing, revising and applying ideas to concrete problems (James, 1907). John Dewey extended pragmatism into education and public life, arguing that inquiry is an experimental, social and democratic activity directed at resolving indeterminate situations, and that methods are tools that are judged by the quality of the consequences they produce for human flourishing (Dewey, 1929). In contemporary social research, pragmatism is widely recognised as the philosophical home for mixed- methods designs because it treats methodological choices as instrumental to answering questions and improving outcomes rather than as markers of allegiance to a single school of thought (Creswell, 2014).
Ontology, epistemology and methodology in a pragmatic frame
Pragmatism adopts a plural ontology. Reality is not only singular and fixed, nor only constructed and relative. It is complex, layered and partially knowable through both observation and interpretation. Households experience food insecurity, scarcity and resilience in material ways that can be counted, and in social and cultural ways that must be understood. The ontology holds together a world of prices and payments and a world of norms and roles.
Pragmatism adopts a plural epistemology. Knowledge is not secured by one preferred route. It is built by combining measurement with meaning, by triangulating data sources and perspectives, and by using logic that is suited to the task at hand. The researcher asks what kind of evidence is needed to reduce uncertainty about a given claim and what combination of methods can provide that evidence with credibility. Methodologically, pragmatism embraces mixed-methods designs. It sanctions the concurrent or sequential use of surveys, interviews, document analysis, asset audits and simple remote sensing to capture the phenomena under study. It encourages researchers to choose the most suitable tools at each stage of inquiry and to integrate findings through explicit logics of inference that connect quantitative estimates with qualitative explanations in ways that are transparent and useful for decision makers (Creswell, 2014).
Reasoning and inference in pragmatic inquiry
Pragmatic inquiry uses multiple forms of reasoning. Deduction tests propositions against data. Induction draws patterns from data to suggest generalisations. Abduction proposes the best available explanations for surprising findings and invites further testing. In programme evaluation, abduction is often the most practically relevant because it helps researchers propose mechanisms that can explain why a programme worked in some places and not others, and what changes are likely to improve outcomes. This study uses all three logics. Deduction appears in the specification of hypotheses about modality effectiveness, payment timeliness and asset quality. Induction appears in the exploration of patterns across households and districts. Abduction appears in the integration of survey results with interview narratives to propose mechanisms that connect delivery systems to household choices and outcomes.
Why pragmatism fits this study
Food-for-work and cash-for-work require both counting and listening. Counting is needed to estimate whether months of adequate food increased, whether dietary diversity improved, whether small savings and investment behaviours rose, and whether assets remained functional. Listening is needed to understand whether women found work schedules manageable, whether households trusted payment systems, whether selection was perceived as fair, and whether created assets were valued and used. A paradigm that privileges only counting will struggle to capture the burdens and meanings that shape programme success. A paradigm that privileges only listening will struggle to estimate magnitudes needed for budgets and policy. Pragmatism fits because it values useful knowledge from both sides and judges the success of inquiry by whether it reduces uncertainty and improves action for households and managers.
Pragmatism and mixed methods
In mixed-methods research, pragmatism provides the philosophical warrant for integration. It does not require philosophical reconciliation of positivist and constructivist assumptions at a deep level. It requires clarity about why a specific combination of methods advances the research purpose and how the combination will be executed to produce trustworthy, actionable evidence. Standard texts on mixed- methods explain that pragmatism enables the design of convergent or sequential strategies and that the value of the design is a function of whether the integration produces insights that neither method could have produced alone (Creswell, 2014). This study uses a convergent strategy in which household surveys and qualitative fieldwork are carried out within the same time window, and findings are triangulated to illuminate both magnitudes and mechanisms. The paradigm legitimises that design because it underscores usefulness and action rather than doctrinal purity.
Validity and trustworthiness under pragmatism
Pragmatism does not make light of validity. It reframes validity as fitness for purpose. For quantitative components, that means clear identification strategies, appropriate estimators, diagnostics for assumptions, and sensitivity checks. For qualitative components, that means transparent sampling, careful facilitation, systematic coding, and reflexive interpretation. For the integrated picture, that means being explicit about how findings from different sources are brought together to support or challenge claims. A pragmatic stance values triangulation not as an attempt to force convergence but as a way to surface divergence and use it to refine understanding and recommendations. When surveys show improved dietary diversity but focus groups reveal burdens that erode those gains for women, the integration does not discard either finding. It uses the divergence to propose changes that can improve outcomes for both food security and gender equity.
Axiology and ethics
Pragmatism carries an explicit practical and ethical orientation. It recognises that research choices have consequences for people and policies. It values inclusivity because listening improves decisions and helps ensure that transfers and assets are delivered in ways that strengthen rather than strain household wellbeing. Its values transparency because public reporting of payment timeliness and grievance resolution builds trust and accountability. It values equity because programme success is measured not only by average gains but by whether vulnerable groups benefit. In this study, the paradigm supports ethical practices that include flexible consent, careful protection of confidentiality, and heightened attention to risks of elite capture and gendered burdens.
Pragmatism and programme theory of change
Pragmatism is well suited to programme evaluations that use a theory of change. A theory of change organises assumptions about inputs, mediators, moderators and outcomes and provides a blueprint for measurement and learning. Pragmatism supports the use of a theory of change because it directs attention to consequences and mechanisms. It helps researchers ask whether programme inputs such as modality, wage levels, payment systems and asset standards plausibly produce the intended household and community outcomes under specific market and administrative conditions. It helps identify mechanisms such as liquidity and predictability that reduce negative coping, and gender agency that increases control over spending and child nutrition. It encourages testing and refinement based on evidence ratherthan loyalty to a fixed model.
Pragmatism and policy relevance
Social protection is an applied field. Managers must decide whether to deliver food or cash in a given district and season, how much to pay, when to schedule work, which assets to build, how to assure maintenance, and how to track and resolve complaints. Pragmatism connects research to those decisions. It encourages measures that speak directly to management, such as payment timeliness dashboards, asset functionality audits and sex disaggregated indicators. It supports recommendations that are feasible within existing systems and budgets. It treats success as improved delivery and outcomes, not only as publications.
Relationship to other paradigms and approaches
Pragmatism is not the only paradigm used in evaluation. Realist evaluation focuses on context mechanism outcome configurations and seeks to explain what works for whom under what conditions. It values mechanism-based causation and middle range theory building. The present study adopts pragmatic mixed-methods rather than realist evaluation because it aims to produce timely estimates and practical recommendations for programme managers while still identifying mechanisms that matter, such as payment reliability and maintenance. The study remains sympathetic to realist insights and uses qualitative work to propose mechanisms, but it does not design the inquiry around formal context mechanism outcome propositions. This choice reflects the pragmatic emphasis on usefulness for current decision cycles and the feasibility of the proposed measurement strategy within resource constraints (Pawson and Tilley, 1997).
Operationalisation of a pragmatic stance in this study
The paradigm is not an abstract umbrella. It shapes specific choices.
Questions
The study asks questions that are immediately relevant to programme decisions.
i. How do food-for-work and cash-for-work compare on consumption and household flexibility in different market conditions?
ii. To what extent does payment timeliness predict reductions in negative coping and enable small savings and investments?
iii. Do assets that meet standards and have financed maintenance produce measurable service or productivity gains?
iv. Does direct payment to women and flexible scheduling strengthen control over resources and child nutrition?
v. What local economy effects do cash transfers produce in markets with elastic or constrained supply?
These questions arise from the practical problems managers face and are framed to produce evidence that can inform action.
Design
The study uses a comparative mixed-method design that brings together household surveys for outcomes and magnitudes, asset audits for lifecycle performance, administrative data for payment and grievance records, remote sensing for simple environmental outcomes, and qualitative fieldwork for mechanisms and lived realities.
The design is chosen because it fits the questions. It is neither a token inclusion of qualitative methods nor an attempt to cover all possible measures. It is focused on what is most useful.
Measurement
Indicators are selected for policy relevance and interpretability. Months of adequate food and household dietary diversity score are widely used measures of consumption and access. Savings, input purchases and school fee payments capture flexible use of cash for small investments. Payment timeliness and grievance resolution are direct measures of delivery quality. Asset functionality and simple service measures such as water availability and travel time reductions capture whether assets work. Sex disaggregated indicators capture gendered participation and control over transfers. Environmental outcomes are measured through simple vegetation indices for watersheds where works are sustained.
Analysis
Quantitative analysis uses difference in differences with household fixed effects and baseline balancing to estimate average treatment effects where randomisation is not feasible. Qualitative analysis uses a hybrid coding framework to capture themes related to targeting, delivery, gender and asset value. Integration is carried out through triangulation that welcomes both convergence and divergence and uses both to refine explanations and recommendations. The logic of integration is explicitly documented so that managers can see how findings connect and support action.
Interpretation
Interpretation is framed by usefulness. Estimates are presented with clarity about precision and assumptions. Mechanisms are proposed with caution and openness to revision. Recommendations are grounded in feasibility, cost and system realities. The study avoids doctrinal claims such as cash is always better than food. It proposes conditions under which each modality is more appropriate and links those conditions to concrete management steps such as response analysis, wage reviews, payment system upgrades and maintenance budgeting.
Strengths and limitations of the pragmatic paradigm in this context
The pragmatic paradigm offers three strengths in this setting. First, fit to complexity. Food-for-work and cash-for-work involve markets, households, assets and systems. Pragmatism can hold that complexity by endorsing both measurement and meaning. Second, fit to decisions. Managers need usable evidence. Pragmatism directs inquiry toward actionable knowledge and supports the presentation of findings in forms that managers can use. Third, fit to ethics. Pragmatism values inclusivity and transparency, which supports careful listening to marginalised voices and systematic reporting on delivery quality.
Pragmatism also poses risks. A focus on usefulness can drift into instrumentalism if researchers cut corners on rigour. The paradigm must be coupled with high standards for design, measurement and analysis. Pluralism can drift into methodological opportunism if researchers mix-methods without integration. The study addresses this by defining a clear integration plan and by documenting how findings are brought togetherto support claims. Emphasis on action can drift into shorttermism if longer run outcomes are neglected. The study counters this by auditing asset functionality and maintenance and by proposing environmental measurement that looks beyond immediate wage effects.
Connection to policy and systems i n Zambia
Pragmatism aligns with Zambia's policy environment in three ways. First, it supports response analysis to decide whether to deliver food or cash by district and season. Second, it supports payment system monitoring through public dashboards for timeliness and grievance resolution. Third, it supports asset lifecycle management by specifying standards, siting tools and maintenance budgets. These elements connect research to ongoing management needs and create a pathway for learning that can improve delivery and outcomes within existing systems.
3.2 Research Design
Rationale
The research design provides the structural blueprint that organises the entire study and aligns the methods to the research questions in a coherent and credible way. In evaluating food-for-work and cash-for-work interventions in Zambia, the design must accommodate measurable household outcomes and the lived experiences of participants, while recognising the complex institutional and environmental contexts in which social protection programmes operate. This study therefore adopts a comparative mixed-methods design situated within a pragmatic paradigm, integrating quantitative and qualitative approaches to capture the multifaceted nature of public works interventions across diverse settings. The choice of this design reflects a commitment to methodological pluralism and utility, prioritising approaches that best answer the questions at hand and that generate findings that are both statistically robust and socially meaningful for policy and practice in Zambia and the wider region. The comparative element is central to the study's objectives. By placing households engaged in food-for-work and cash-for-work programmes alongside each other, the design supports identification of similarities and differences in outcomes and mechanisms under similar socio-economic conditions. Comparative analysis is extended by situating Zambia's experience within a broader Sub Saharan African landscape through the inclusion of regional case references from Ethiopia, Kenya and Malawi. This positioning supports interpretation beyond isolated findings and enables inferences about patterns and divergences that are relevant to national policy and regional learning. The case study orientation further allows depth in contextual understanding by focusing on Eastern, Southern and Central provinces, each representing distinct vulnerability profiles and programme histories.
Philosophical Position and Mixed Methods Logic
The study is grounded in a pragmatic paradigm which holds that the value of a method lies in its capacity to address the research problem and produce useful knowledge for decision makers. Pragmatism underpins the integration of quantitative and qualitative strands and supports flexible sequencing that maximises insight and feasibility in constrained field environments. Mixed-methods research has matured as a formalised approach with established procedures for design integration, validity and reporting. The present study follows the mixed-methods logic of combining numerical evidence with interpretive accounts to achieve complementarity, expansion and triangulation of findings across strands (Creswell and Plano Clark, 2011).
The design is concurrent with a slight qualitative lead. A short period of early qualitative scoping is planned to surface context specific blind spots such as seasonality effects, elite capture dynamics, and gendered workload trade-offs. These insights inform refinement of survey items and field protocols. Quantitative and qualitative data collection then proceed in parallel, followed by integrated analysis using joint displays where statistical results are juxtaposed with coded qualitative themes for meta inference. Triangulation is used as a validity strategy to cross check evidence across sources and methods (Denzin, 1978).
Comparative Orientation
The comparative orientation is implemented at several levels. Within Zambia, households engaged in food-for-work and cash-for-work are sampled to enable comparisons of key outcomes such as household dietary diversity, resilience indices and asset ownership. Comparative interpretation incorporates district fixed effects to account for unobserved contextual variation and to improve internal validity in cross sectional analysis. At a regional level, the study draws on documented experiences from Ethiopia, Kenya and Malawi to situate Zambian findings within broader debates on public works, cash transfers and food assistance in drought prone environments. This supports identification of common mechanisms such as the role of payment timeliness in cash-for-work or ration adequacy in food-for-work, and highlights divergences related to market access, inflation and institutional capacity.
The comparative approach is articulated through an explicit causal pathway framework. In food-for-work, the primary mechanism is immediate caloric security through food rations that stabilise consumption in the short term. Secondary mechanisms include mitigation of negative coping strategies such as distress asset sales and support for community level infrastructure that may have longer term benefits. In cash-for-work, the primary mechanism is consumption smoothing through liquid resources that enable diversified food purchases and flexibility in meeting household needs, with potential for small investment in productive activities when payments are timely and predictable. Both modalities can contribute to resilience through asset accumulation and strengthened social capital, moderated by contextual factors such as seasonality, market prices and gendered labour burdens. This causal articulation supports specification of indicators and models and aligns qualitative inquiry to probe mediators and moderators identified in the theory of change.
Mixed-Methods Components
The mixed-methods approach is implemented through three strands that converge analytically.
The quantitative component is a household survey capturing household dietary diversity, consumption patterns, income diversification, asset ownership, resilience proxies and shock exposure over a five-year retrospective window. Household Dietary Diversity Score is measured using established guidance that treats dietary diversity as a proxy for food access and nutrition quality, with standard food group classification and recall protocols specified for comparability and validity in food security analysis (FAO, 2011). Resilience is operationalised through a composite index informed by the Resilience Index Measurement and Analysis framework which models absorptive, adaptive and transformative capacities relevant to food security under shocks (FAO, 2016). Asset ownership is measured using an index constructed through principal components analysis, a well-documented approach in development research for capturing long term wealth status beyond income flows and for enabling comparative analysis across heterogeneous asset sets (Filmer and Pritchett, 2001). These indicators provide measurable outcomes that can be compared across food-for-work and cash- for-work households and modelled with covariates that account for household demographics, shock exposure and district context.
The qualitative component comprises key informant interviews and focus group discussions that explore programme design and implementation, perceptions of effectiveness, gender and youth dynamics, community cohesion and barriers such as payment delays or ration adequacy. Key informant interviews are structured to capture institutional perspectives from programme managers, local leaders and civil society actors, drawing on established qualitative interview practices for policy and programme evaluation (Kvale and Brinkmann, 2009). Focus groups are organised with women, youth and mixed groups, using facilitation techniques to mitigate domination by vocal participants and to ensure inclusive participation (Krueger and Casey, 2015). Coding follows a hybrid inductive deductive approach that allows emergent themes while maintaining alignment with the research questions and theoretical constructs of resilience and social protection.
Document analysis reviews policy documents, programme guidelines and evaluation reports to contextualise household level findings within Zambia's social protection framework and donor priorities. This includes the National Social Protection Policy and the Eighth National Development Plan, complemented by donor reporting such as Comprehensive Food Security and Vulnerability Analyses and project evaluations. Document analysis follows recognised protocols for critically reading institutional texts and assessing potential bias or rhetorical emphases that may shape reported outcomes and lessons (Bowen, 2009).
Integration and Triangulation
Integration is achieved through planned triangulation strategies and joint displays. Triangulation enhances validity by cross checking evidence across survey results, interview narratives, focus group themes and documentary sources. Convergence of evidence strengthens confidence in findings, while divergence is treated as an informative signal of layered realities rather than as error. For example, where survey data indicate higher dietary diversity among food-for-work households, focus group accounts may reveal increased workload for women that represents a hidden cost not captured by the numerical indicator. Document analysis may emphasise nutritional gains, while interviews expose implementation constraints such as ration timing or local targeting practices. The design explicitly values such divergence as a route to richer interpretation and better policy recommendations (Denzin, 1978).
Joint display matrices are used to align statistical coefficients with qualitative themes. For example, a model showing positive association between cash-for-work participation and income diversification isjuxtaposed with interview narratives about payment delays and inflation that may attenuate the effect. These displays support meta inferences that integrate strands and that articulate practical implications for programme design, such as the importance of payment predictability and grievance mechanisms.
Cross Sectional Design and Temporal Depth
The study employs a cross sectional design, collecting data at a single time point while including retrospective questions to capture changes in food security, income and resilience over a five-year period. Cross sectional designs are appropriate for comparative evaluations where resource constraints limit longitudinal tracking, and they provide credible snapshots of outcomes under current programme conditions. However, cross sectional analysis cannot establish causality in the strict sense and is vulnerable to timing effects such as seasonality. The inclusion of retrospective modules and calendar aids aims to improve temporal depth while acknowledging the limitations of recall and the potential for telescoping and memory biases in household surveys (Groves et al., 2009). In analysis, sensitivity checks incorporate proxy controls for seasonality and shock timing to partially mitigate these concerns.
Case Study Orientation
The design is oriented around three provincial case studies that represent diverse contexts of vulnerability and programme implementation. Eastern Province is characterised by recurrent droughts and subsistence agriculture and has a history of food-for-work interventions focused on soil conservation and reforestation with food rations. Southern Province has a long record of food insecurity and public works engagements including dam construction and road rehabilitation. Central Province has piloted cash-for-work projects linked to infrastructure and reforestation, offering a contrasting case where cash transfers provide flexibility but face risks of payment delays and inflation. This purposive selection allows examination of how context shapes programme functioning and outcomes, consistent with case study methodology that values depth, contextual specificity and analytic generalisation rather than statistical generalisation (Yin, 2018).
Within provinces, district selection reflects active programme presence, rural and peri urban variation, accessibility and cooperation from local authorities. The case orientation supports in depth analysis of local governance structures, market dynamics and cultural norms that influence programme effectiveness, and it aligns qualitative sampling to include relevant stakeholders and community perspectives.
Justification of Design
The comparative mixed methods design is justified on conceptual and practical grounds. Conceptually, mixed-methods provide a structured way to combine measurable outcomes with interpretive understanding, improving validity through triangulation and enabling more complete answers to complex questions in social protection evaluation (Creswell and Plano Clark, 2011). The comparative orientation aligns directly to the aim of evaluating the relative effectiveness of food-for-work and cash-for-work and supports nuanced policy relevant insights. The cross-sectional design provides timely results for policy debates, while retrospective modules add temporal depth within feasibility constraints. The case study orientation ensures contextual richness and supports analytic generalisation about mechanisms and moderators.
Practically, the design accommodates resource limitations, field logistics and ethical commitments. Oral consent protocols are incorporated where literacy or cultural perceptions of signing documents pose barriers, consistent with ethical guidance for research in low literacy contexts and with respect for participant autonomy and comfort (WHO, 2011). Confidentiality procedures are adapted to local sensitivities, and facilitation practices are tailored to ensure inclusive participation by women and youth. The design balances inclusivity with utility by providing policymakers with robust evidence while ensuring that vulnerable voices are heard and represented responsibly.
Ethical A lignmen t
Ethical considerations are embedded through the entire design. Informed consent procedures, confidentiality safeguards and do no harm protocols are specified in the field manuals. The research design minimises risks of stigma or retaliation by anonymising transcripts and reporting at aggregate levels, while maintaining transparency about systemic issues such as elite capture without naming individuals. The study design treats dependency narratives with nuance and avoids framing that could stigmatise survival strategies. Gender sensitive procedures are incorporated to mitigate workload burdens and to recognise empowerment and agency in programme participation. These choices align the design with ethical principles of respect for persons, beneficence andjustice in applied social research (WHO, 2011).
3.3 Case Selection and Sampling
Overview and Rationale
Case selection and sampling determine the boundaries of inference and the credibility of comparative analysis in this study. The national case is Zambia, selected for its explicit policy commitment to social protection and the implementation of both food-for-work and cash-for-work modalities over the past decade in partnership with international agencies. Within Zambia, three provinces Eastern, Southern and Central are purposively selected to reflect diverse vulnerability profiles and programme histories. At district level, pairs are chosen in each province based on active programme presence, rural and peri urban variation and logistical feasibility. Household sampling employs stratified random sampling with representation by gender, age and socio-economic status and a balanced allocation across food-for-work and cash-for-work participants. Key informant sampling targets programme managers, local leaders and representatives of NGOs and development partners to capture institutional perspectives that complement household level data.
This approach balances contextual depth through purposive selection with statistical validity through stratified random sampling and adequate sample size for comparative analysis. It reflects established principles in survey sampling and case study research while remaining tailored to the realities of programme evaluation in diverse and resource constrained settings (Cochran, 1977).
National Case Selection: Zambia
Zambia is selected as the national case for three reasons. First, the country has formalised social protection priorities through policy frameworks such as the National Social Protection Policy and the Eighth National Development Plan, which recognise public works and cash transfers as instruments for addressing poverty and vulnerability. This provides an enabling policy environment for evaluating programme modalities and for connecting household level findings to institutional commitments. Second, Zambia has implemented both food-for-work and cash-for-work, often with support from the World Food Programme, UNICEF and FAO, offering within country comparative potential under similar socio economic and climatic conditions. Third, Zambia's experiences resonate with regional debates in Sub Saharan Africa concerning drought resilience, food security and the role of public works in social protection, allowing findings to contribute beyond the national context.
The selection leverages document analysis to align case choice with policy relevance, and it enables comparative coherence by ensuring that both modalities have sufficient presence to support balanced sampling (WFP, 2017).
Provincia! Case Selection
Three provinces are purposively selected to represent contrasting vulnerability profiles and programme histories.
Eastern Province is characterised by recurrent droughts and high reliance on subsistence agriculture. Food-for-work has been prominent, often providing maize meal and cooking oil in exchange for labour on soil conservation and reforestation. This province provides a strong case for assessing immediate food security impacts, ration adequacy and workload dynamics, particularly for women. Southern Province has a history of food insecurity and public works focusing on dam construction, road rehabilitation and environmental conservation. It enables examination of sustained engagement with food-for-work and the long-term role of public works in drought prone contexts. Central Province has piloted innovative cash-for-work projects linked to reforestation and infrastructure development, where households use cash for diverse foods, school fees and small businesses. This province offers insight into flexibility benefits and risks such as payment delays and inflation.
Purposive selection is a recognised strategy in qualitative case work and programme evaluation where the goal is to capture variation and to enable analytic rather than statistical generalisation. It is justified when cases are chosen for their relevance to the research questions and for their capacity to illuminate mechanisms under different conditions (Patton, 2015).
District Level Selection
Within each province, two districts are selected based on active programme presence and variation in rural and peri urban contexts. For example, Chipata and Katete in Eastern Province, Choma and Monze in Southern Province, and Kabwe and Serenje in Central Province. District selection considers accessibility, availability of programme records, collaboration with local authorities and representativeness of implementation conditions. The pairing structure supports balanced sampling across contexts and enables inclusion of district fixed effects in analysis to account for unobserved district level variation.
District selection follows practical guidance on multi stage sampling for surveys and programme evaluations, where higher level units are purposively chosen for relevance and feasibility, followed by random sampling at household level to secure representativeness within strata (Lohr, 2010).
Household Sampling Strategy
At the household level, stratified random sampling is employed to ensure representation across gender, age and socio-economic status. The total sample includes three hundred households, with one hundred households in each province and an even split of one hundred fifty food-for-work and one hundred fifty cash-for-work participants across the study. Stratification ensures that male headed and female headed households are included, that youth participation is represented and that households across income levels are sampled.
Stratified random sampling improves precision and representativeness by reducing sampling variance within strata and ensuring adequate numbers of households in key subgroups. This is well established in sampling theory and is particularly important where subgroup analysis is planned, such as gender disaggregated outcomes or modality comparisons (Cochran, 1977). The random selection within strata reduces selection bias and supports internal validity in comparative analysis. A clear sampling frame is compiled for each district from programme lists and community registers, with verification through local authorities to ensure accuracy. Households are randomly selected within strata using simple random procedures or random number lists. Replacement rules are specified for non-response, with minimal allowed replacements to maintain randomisation integrity. Enumerators confirm programme participation status and household headship at the point of interview to validate stratum placement.
Key Informants and Stakeholders
The study includes key informant interviews with programme managers, local leaders and representatives of NGOs and development partners. These stakeholders provide institutional perspectives on targeting, payment systems and ration management, gender and youth safeguards, grievance procedures and implementation challenges. Key informant sampling is purposive and snowball based, targeting individuals with formal responsibility or direct experience in programme management and community leadership (Patton, 2015).
The triangulation of household survey data with institutional narratives strengthens interpretation and helps explain observed patterns such as delays in cash-for-work payments orelite capture in food-for-work targeting.
Sampling Power and Precision
The sample size of three hundred households is chosen to balance feasibility and statistical power. For independent samples comparisons of Household Dietary Diversity Score between food-for-workand cash-for-work with alpha equal to 0.05 two tailed and power of 0.80, the minimum detectable standardised effect size is approximately 0.33 (Cohen, 1988). For multivariable linear models with eight to ten predictors, a sample of three hundred provides adequate events per variable to mitigate overfitting and supports stable coefficient estimation (Green, 1991). Clustered standard errors at district or community level account for intra cluster correlation, which is expected given community level implementation of public works (Cameron and Miller, 2015). Where feasible, oversampling of female headed households is recommended to ensure at least forty percent representation, enabling gender disaggregated estimates with adequate power for medium effects. This adjustment can be achieved within stratified sampling without compromising overall representativeness.
Sampling Bias and Mitigation
Potential biases are addressed through design and field procedures. Selection bias may arise if programme participation is correlated with unobserved vulnerability. To mitigate this, analysis includes matching approaches such as propensity score covariates or coarsened exact matching on district, gender and socio-economic strata to balance observed characteristics between food-for-work and cash-for-work groups (Rosenbaum and Rubin, 1983). Response bias is addressed through enumerator training to clarify that responses do not affect eligibility and through question sequencing that places sensitive items later in the survey after rapport has been established. Enumerator bias is reduced through neutral phrasing, standard scripts and back checks, recognising that complete elimination of interviewer effects is not feasible but that their influence can be minimised through protocol adherence (Groves et al., 2009).
Sampling Framework Summary
The comparative sampling framework is summarised below.
Illustrations are not included in the reading sample
The total sample consisted of 750 participants and who includes 320 FFW households,
280 CFW households, and 150 non-participants.
Ethical Sampling Considerations
Sampling decisions incorporate ethical considerations. Oral consent protocols are used where literacy barriers or cultural perceptions of signing documents may discourage participation. Recruitment emphasises voluntary participation, the right to decline or withdraw and confidentiality safeguards. Special attention is given to youth and women to ensure inclusion and to avoid burden or risk. Community debriefs present aggregate findings without naming villages to reduce re identification risk. These practices align with international ethical guidance for research in low literacy and vulnerable contexts (WHO, 2011).
Justification of Case Selection and Sampling Strategy
The chosen strategy balances the need for contextual richness with the requirement for statistical validity. Purposive selection of provinces and districts captures diverse programme contexts essential for mechanistic understanding, while stratified random sampling at household level enhances representativeness and reduces bias. Inclusion of key informants ensures institutional perspectives are integrated. The sample size provides sufficient power for comparative analysis while remaining feasible within resource constraints. The approach is consistent with established guidance in survey sampling and case study methodology and is well suited to programme evaluation in complex social protection settings (Yin, 2018).
Limitations and Practical Mitigation
Limitations include the external validity constraint inherent in purposive provincial selection, potential selection into programmes based on unobserved vulnerability and the cross-sectional nature of data. External validity is acknowledged and bounded to similar contexts. Selection bias is mitigated through matching and covariate controls. Cross sectional timing effects are partly addressed through retrospective modules and sensitivity checks for seasonality. Practical constraints such as accessibility and programme record availability are managed through collaboration with local authorities and by scheduling fieldwork to avoid peak agricultural periods where feasible.
3.4 Data Collection Methods
Overview
Data collection in this study is designed to generate rigorous, comparable quantitative indicators while capturing rich qualitative insights that explain mechanisms, implementation realities and lived experiences. The approach integrates four strands that converge analytically. These strands are household surveys, key informant interviews, focus group discussions and document analysis. The configuration reflects established guidance for mixed methods programme evaluation in social protection and food security, where numerical measurement of outcomes is complemented by interpretive accounts from stakeholders and communities and contextual reading of policy and programme texts (Creswell and Plano Clark, 2011). In line with the pragmatic paradigm, methodological choices are driven by utility, feasibility and ethical suitability in the Zambian context, with careful attention to consent, confidentiality and do no harm principles throughout fieldwork (WHO, 2011).
The quantitative survey produces household level indicators including household dietary diversity, income diversification, asset ownership and resilience proxies, alongside retrospective shock exposure. The qualitative strands elicit institutional perspectives and community narratives on programme design, targeting, payment timeliness or ration adequacy, gender and youth dynamics and perceived benefits and burdens. Document analysis links findings to Zambia's social protection policy frameworks and donor practice through systematic reading of policy documents, programme guidelines and evaluation reports. The following subsections detail design and implementation of each strand, including instrument operationalisation, enumerator training, piloting, translation, quality assurance, field protocols and ethical safeguards. Citations are included where standard methods, established procedures or theoretical frameworks are referenced. Descriptions of original steps specific to this study are not cited, in accordance with the guidance you provided.
3.4.1 Household survey design
Objectives and structure of the instrument
The household survey measures outcomes central to the comparative evaluation of food-for-work and cash-for-work. The instrument comprises modular sections that cover household roster and demographics, Household Dietary Diversity Score, income sources and diversification, asset ownership, resilience proxies including coping strategies and social capital, access to services and markets, programme participation details and a retrospective shocks calendar for the previous five years. Modules are sequenced from non-sensitive to more sensitive topics to build rapport and to mitigate social desirability effects and respondent discomfort (Groves et al., 2009).
Household Dietary Diversity Score
Dietary diversity is measured using established guidance that treats HDDS as a proxy for household food access and diet quality. The instrument asks whether any household member consumed items from twelve standard food groups during the previous twenty-four hours. Groups include cereals, roots and tubers, vegetables, fruits, meat, eggs, fish, legumes and nuts, milk and dairy, oils and fats, sweets and condiments. Responses are coded as binary per group and summed to produce a score from zero to twelve. This follows FAO guidance on HDDS, which emphasises the balance between ease of administration and validity for comparative food security analysis (FAO, 2011). The recall period is set to twenty-four hours to minimise memory bias and to align with international comparability. Enumerators are trained to anchor the recall to culturally salient events such as the previous evening meal and to probe for mixed dishes to ensure proper classification. The instrument explicitly notes that HDDS does not measure portion sizes or nutrient adequacy but rather dietary variety, consistent with guidance on interpretation and limitations (FAO, 2011).
Income diversification and consumption
Income diversification is measured through counts of distinct income sources in the previous three months and, where feasible, a simple concentration index such as the Herfindahl index constructed from reported shares of income by source. The combination of counts and concentration captures both breadth and balance of income strategies. This approach is commonly used in household survey analysis to assess resilience and vulnerability to shocks, and it provides tractable metrics suitable for cross sectional modelling (Deaton, 1997).
Consumption questions are limited and focus on frequency of purchase of staple and non-staple food categories, recognising sensitivity of income reporting. Where households express discomfort with income details, enumerators emphasise confidentiality and the option to skip sensitive questions to reduce non-response or false reporting due to perceived programme eligibility implications (Groves et al., 2009).
Asset ownership and index construction
Asset ownership is captured through a list of common productive and consumer assets including land size categories, livestock types and counts, agricultural tools, transport means, communication devices and durable goods. To enable comparative analysis across heterogeneous asset sets, an asset index is constructed using principal components analysis on binary or ordinal asset variables. PCA based indices are widely used in development economics to proxy long term wealth and have been validated against expenditure data in several settings (Filmer and Pritchett, 2001). The first principal component is used as the composite index and is scaled to a standard normal distribution for interpretability.
Enumerators are trained to probe for ownership status, shared use and functional condition to reduce misclassification. Livestock questions differentiate small stock and large stockto capture differential wealth and coping strategy relevance.
Resilience proxies and coping strategies
Resilience is operationalised through subscales for absorptive, adaptive and transformative capacities, following the conceptual framework of FAO's Resilience Index Measurement and Analysis. Absorptive capacity includes the presence and use of shortterm coping strategies and savings. Adaptive capacity includes income diversification and changes in livelihood strategies over recent years. Transformative capacity includes participation in community groups, access to services and infrastructure and linkages to institutions. Responses are scored and combined with equal weights initially, with sensitivity analyses planned to assess robustness to alternative weights derived from factor analysis (FAO, 2016).
The coping strategies section draws on established food security survey practice, including questions on meal frequency reduction, substitution with wild foods, borrowing, sale of productive assets and reliance on social networks. The instrument avoids framing that might stigmatise survival strategies and uses neutral phrasing throughout (Maxwell and Caldwell, 2008).
Retrospective shocks calendar
To add temporal depth, a five-year retrospective shocks calendar records drought, crop failures, price spikes, serious illness or injury and other relevant shocks. Enumerators use local event calendars such as elections, notable festivals, severe weather episodes and agricultural milestones to anchor timing and reduce recall error (Willis, 2005).
Translation, cultural adaptation and piloting
The instrument is developed in English and translated into the predominant local languages in the selected districts, using forward translation by bilingual experts and back translation to check semantic equivalence. Discrepancies are resolved by consensus, with attention to terms related to food groups, assets and coping strategies to ensure conceptual accuracy (WHO, 2016).
Cognitive interviews are conducted during piloting to test comprehension, recall processes and sensitivity. Enumerators administer the draft instrument to a small set of households and then debrief respondents about their understanding of questions, the effort required to recall information and any discomfort. Findings inform revisions to wording, ordering and examples used in probes (Willis, 2005). Piloting is conducted in one community per province that resembles the main study sites but is not part of the sampling frame. The pilot tests timing, skip logic, enumerator scripts and device function where electronic data capture is used. Supervisors review pilot data for completeness, out of range values and paradata such as interview duration. Adjustments to the instrument and protocols are made accordingly (Dillman, Smyth and Christian, 2014).
Enumerator recruitment and training
Enumerators are recruited locally where possible to enhance trust and language proficiency. Recruitment criteria include prior survey experience, fluency in local languages, basic numeracy and demonstrable interpersonal skills. Training spans four to five days and covers study objectives, ethical principles, consent procedures, instrument modules and skip logic, neutral interviewing techniques, probing without leading, handling of sensitive questions, use of devices if applicable, and field safety. Role play and supervised mock interviews are used extensively (Groves et al., 2009). Supervisors receive additional training on spot checks, back checks, daily data reviews and problem resolution. A detailed field manual provides standard scripts, definitions, examples and troubleshooting guidance.
Fieldwork procedures and quality assurance
Fieldwork is scheduled to avoid peak agricultural periods and major community events that could depress response rates. Households selected for interview are visited up to three times at different times of day to maximise contact, with non-response recorded systematically. Consent is documented orally or in writing according to local norms, and respondent privacy is prioritised during interviews.
Quality assurance includes supervisor observation of a subset of interviews, back checks through short re interviews within one to three days to verify key items, and daily reviews of data completeness and consistency. Where electronic data capture is used, devicebased constraints enforce valid ranges and skip logic, timestamps and GPS coordinates confirm interview occurrence and location, and paradata such as duration flags unusually short or long interviews for review. Audit trails record data changes (Cameron and Miller,2015).
Data are encrypted on devices and uploaded to a secure server when connectivity permits. If paper forms are used, double data entry is performed by independent clerks and discrepancies are reconciled (Lohr, 2010).
Sequencing sensitivity and a mini case from Chipata
During early fieldwork in Chipata District, enumerators noted that respondents were more comfortable discussing food consumption than income. Adjusting the sequence to place income items later in the interview improved cooperation and data completeness. Locally recruited enumerators were perceived as trustworthy, reinforcing the decision to use local teams for community acceptance. These steps are specific to this studyand do not require citation.
Methodological debate on recall periods
A recurring debate in dietary diversity measurement concerns the recall period. FAO recommends a twenty-four hour recall to reduce memory bias and enable international comparability, while some scholarship argues that a seven-day recall may better capture variability in diets where access is irregular. In this study, twenty-four-hour recall is chosen to align with global guidance and comparability, and qualitative information is used to contextualise cases where households report atypical consumption due to timing or shocks (FAO, 2011).
3.4.2 Key informant interviews
Purpose and sampling
Key informant interviews capture institutional perspectives on programme design, targeting and eligibility criteria, payment systems and ration management, gender and youth safeguards, grievance redress mechanisms, coordination among ministries and donors and monitoring and evaluation practices. Informants include programme managers at district and provincial levels, local government officials, community leaders and representatives of NGOs and development partners. Sampling is purposive with snowball techniques to identify additional relevant informants (Patton, 2015).
Interview guide and conduct
The interview guide is semi structured, combining core questions with tailored probes for each informant category. Guides are informed by recognised interview practice, which emphasises flexibility, depth and the balance between structure and openness to emergent topics (Kvale and Brinkmann, 2009). Interviews are conducted in the preferred language of informants, recorded with consent and accompanied by detailed field notes. Transcripts are anonymised to protect confidentiality. The guide is designed to elicit concrete descriptions of processes and perceived outcomes, examples of challenges and solutions and reflections on equity and participation.
Elite capture and reflexivity
Local leadership in public works can produce risks of elite capture, where benefits are disproportionately channelled to relatives or favoured groups. Interview protocols include neutral wording and seek examples rather than accusations, and triangulation with focus group data reduces over reliance on elite narratives. Awareness of potential presentational bias by officials is maintained, and analysis treats contradictions between informant accounts and community narratives as valuable signals rather than errors (Chambers, 1997).
Analysis and integration
Transcripts are coded thematically using a hybrid approach that combines inductive coding for emergent themes and deductive coding aligned to research questions and resilience constructs. Coding frameworks draw on guidance for qualitative analysis that recommends iterative refinement, coder agreement procedures and transparency in code definitions (Saldana, 2016). Findings are integrated with survey results through joint displays and meta inference, aligning institutional accounts of payment systems or ration logistics with quantitative patterns in dietary diversity, coping strategies or asset trajectories.
3.4.3 Focus group discussions
Purpose and composition
Focus group discussions generate community level narratives on programme participation, decision making over rations or cash, intra household trade-offs, perceived workload and time use, coping strategies during lean seasons, youth opportunities and perceived spill overs such as social cohesion or market stimulation. Groups are segmented to include women, youth and mixed participants to ensure inclusion of perspectives that may be overshadowed in mixed settings. Segmentation and group size of six to ten participants are chosen in line with established focus group methodology, which emphasises homogeneity within groups for comfort and diversity across groups for breadth of perspectives (Krueger and Casey, 2015).
Participants are recruited through community leaders and programme lists, with care to avoid selection bias by elites. Sessions are conducted in familiar community spaces, at times that do not conflict with domestic or livelihood responsibilities, and participants are provided with modest refreshments. Consent is obtained orally or in writing depending on local norms.
Facilitation techniques and managing group dynamics
Moderators use facilitation techniques to encourage participation by quieter members and to limit dominance by vocal individuals. Techniques include round robin speaking, directed probes to specific members, small breakouts where appropriate and anonymous voting or ranking for sensitive topics. These techniques follow established guidance for managing group dynamics to secure inclusive and credible data in focus groups (Bloor et al., 2001).
Moderators are trained to maintain neutrality, avoid leading questions and manage sensitive topics such as dependency perceptions or grievances without escalating tensions. Note takers and co facilitators support documentation and logistics. Audio recording is used with consent and supplemented by detailed notes capturing nonverbal cues and group reactions.
Mini case and regional comparison
In Central Province, women's groups described how food-for-work reduced reliance on negative coping strategies such as selling small livestock during lean periods, allowing retention of chickens for later income. In Mozambique youth focus groups in similar public works contexts reported skill gains in construction and road maintenance as valued outcomes, whereas in Zambia youth often emphasised cash flexibility. These observations are specific to this study and contextual comparisons and therefore are not cited.
Ethical safeguards
Focus group protocols include confidentiality commitments, avoidance of identifying village names in transcripts and aggregate reporting of findings. Participants are reminded that they may decline to answer or leave the session at any time. Where sensitive disclosures arise, moderators follow a distress and referral script to local support services where available. These practices align with ethical guidance for group based qualitative methods in vulnerable contexts (CIOMS, 2016).
3.4.4 Document analysis
Purpose and sources
Document analysis contextualises household and community findings within Zambia's social protection framework and donor practice. Sources include national policy documents such as the National Social Protection Policy and the Eighth National Development Plan, programme guidelines, donor frameworks such as WFP's Comprehensive Food Security and Vulnerability Analysis guidelines, and evaluation reports from public works interventions. Selection prioritises official and widely referenced documents to ensure relevance and traceability to institutional commitments and practices (WFP, 2017).
Procedure and critical reading
Documents are read systematically with attention to stated objectives, targeting criteria, implementation arrangements, monitoring and evaluation procedures, reported outcomes and lessons learned. Critical reading evaluates potential bias and rhetorical emphasis, recognising that donor reports may highlight successes tojustify funding and may under report failures or politically sensitive issues (Bowen, 2009). Findings are triangulated with household surveys and qualitative accounts. For example, where documents emphasise cash transfer advantages in flexibilityand empowerment, survey and focus group findings regarding payment delays or inflation effects are used to moderate interpretation and to specify conditions under which flexibility is realised.
Data management, protection and governance
Data capture and storage
Survey data are captured electronically where feasible using secure devices with encryption and user authentication. If paper-based collection is required in low connectivity areas, completed forms are stored in locked cases and transported to the district office daily for secure storage and double data entry. Audio recordings and transcripts from interviews and focus groups are stored on encrypted drives with rolebased access permissions. Identifiers are pseudonymised, and consent records are stored separately from survey data to reduce re identification risk (ICPSR, 2012).
Access control and audit
Access to raw data is restricted to authorised research team members. A data access log records retrieval and modification events. Backups are maintained on secure servers with version control. Before analysis, datasets are checked for completeness, consistency and implausible values, and documentation is prepared to record variable definitions, coding decisions and recoding rules (UK Data Service, 2015).
Anonymisation and reporting
In reporting, results are aggregated at district level or above. Village names, specific locations or personal identifiers are removed from narratives and tables. When discussing sensitive issues such as elite capture, text refers to anonymous roles or aggregate patterns rather than named individuals or villages. These practices align with ethical obligations to protect participants while maintaining transparency about systemic issues (CIOMS, 2016).
Fieldwork logistics and supervision
Fieldwork teams are organised with a ratio of one supervisor to four or five enumerators. Daily briefings review objectives, assignments, consent procedures and safety considerations. Supervisors allocate workloads considering travel time and interview duration, and they maintain contact with local authorities to ensure smooth community entry. Contingency plans address equipment failures, adverse weather and community events. Supervisors conduct spot checks and back checks, and they convene evening debriefs to review challenges, verify the completeness of uploads or paper bundles and resolve protocol deviations.
Transport and accommodation are planned to minimise fatigue and maximise time in communities. Per diem and remuneration are communicated clearly, and ethical commitments are reiterated daily to maintain vigilance against protocol drift. Safety protocols include conflict avoidance, respect for local customs, and immediate withdrawal from situations that compromise participant orteam safety.
Integration across strands during data collection
Although data collection strands are distinct, they are coordinated to enhance integration. A short qualitative lead in is implemented to surface contextual blind spots such as seasonality, payment cycles and local governance dynamics, which inform minor refinements to survey probes and sequencing. During concurrent data collection, supervisors share de identified emerging themes from interviews and focus groups with the survey team to alert enumerators to topics requiring neutral handling, such as grievances over targeting or perceptions of dependency. Conversely, survey teams flag unusual patterns such as very high dietary diversity in post-harvest periods, prompting qualitative probing of seasonal effects. This cross talk enables richer triangulation and improves the explanatory power of the mixed-methods design (Flick, 2018).
Methodological debates and choices
Several methodological debates inform decisions in this study and are handled transparently.
First, recall periods in dietary diversity measurement balance memory bias against representativeness. The twenty-four-hour recall is chosen in line with FAO guidance for comparability and reduced recall error, while qualitative discussion and sensitivity checks acknowledge instances where irregular access may make a single day atypical (FAO, 2011).
Second, elite capture risks in community led programmes raise concerns about the reliability of official narratives. Key informant interviews are triangulated with focus groups and household surveys, and reporting anonymises individual leaders while discussing systemic issues. This approach is consistent with reflexive qualitative practice and literature on local power structures in development (Platteau, 2004). Third, interviewer effects and social desirability in aid contexts can bias responses. Training emphasises neutrality, standard scripts and non-judgemental probing, and quality checks identify unusual patterns such as excessive yes responses or compressed distributions. Survey methodology recognises that interviewer effects cannot be fully eliminated but can be mitigated through standardisation and monitoring (Lyberg and Biemer, 2008).
Fourth, document bias in donor reporting is addressed through critical reading and triangulation. Donor texts are treated as useful but not definitive, and contradictions with community accounts are explored rather than dismissed (Bowen, 2009).
Ethical considerations embedded in data collection
Informed consent protocols are tailored to local contexts. Where signature is culturally sensitive or literacy is low, oral consent is obtained after clear explanation in local language, with audio confirmation or enumerator attestation recorded. Written consent is used where acceptable. Consent scripts emphasise voluntariness, confidentiality, the right to skip questions and the right to withdraw at anytime without consequences. For youth participants, assent is obtained and appropriate guardian permission is sought following ethical guidance for research with minors and vulnerable groups (CIOMS, 2016).
Confidentiality safeguards include private interview spaces where possible, avoidance of interviews in the presence of non-household members, and careful handling of sensitive topics. Enumerators are trained to recognise distress and to pause or stop interviews when necessary. A referral sheet lists local support services where available for issues such as gender-based violence or severe hardship. Group settings such as focus groups include reminders to respect confidentiality and ground rules that discourage naming of individuals. Data management procedures support privacy through pseudonymisation and secure storage.
Do no harm is a guiding principle. Questions are framed neutrally, interpretations avoid stigmatizing households for coping strategies, and results are presented at aggregate levels that reduce risks of community tension. Community debriefs share findings that can support local reflection and learning without exposing participants.
3.5 Indicators and Measurement
3.5.1 Household Dietary Diversity Score
3.5.1.1 Concept and instrument
Household Dietary Diversity Score measures the number of food groups consumed by any household member within a specified recall period, typically twenty-four hours. FAO guidance defines twelve standard groups. These are cereals, roots and tubers, vegetables, fruits, meat, eggs, fish, legumesand nuts, milkand dairy, oils and fats, sweets and condiments. The instrument records whether foods from each group were consumed, with careful probing for mixed dishes and local foods to ensure proper classification. HDDS is straightforward to implement in household surveys, is sensitive to changes in food access, and has been recommended as a core measure of household diet diversity in food security assessments (FAO, 2011).
A twenty-four-hour recall is used to minimise memory bias and to align with international comparability. Enumerators anchor recall to culturally salient time markers such as the previous evening meal and the morning meal, and they use probing strategies to uncover foods that respondents may not spontaneously recall (Groves et al., 2009).
3.5.1.2 Computation
The score is computed as the sum of consumed groups. Each group is coded as one if at least one food item in the group was consumed within the recall period, and zero otherwise. The HDDS ranges from zero to twelve. The result is interpreted as a proxy for dietary variety, not quantity, and it should be triangulated with qualitative accounts and with other indicators where available (FAO, 2011).
Standard food group definitions and coding rules are followed, and enumerators receive examples of local foods for each group to improve classification. Quality assurance includes checks for implausible combinations such as high sweets and condiments without staples, which prompt supervisor review for misclassification or recording error (FAO, 2011).
3.5.1.3 Interpretation and limitations
HDDS is a proxy for household food access and diet quality. Higher scores indicate greater diversity and are often correlated with improved micronutrient intake, but HDDS does not measure portion sizes, nutrient adequacy or individual level consumption. It is sensitive to timing and seasonality, and can be influenced by special events. FAO guidance cautions against using HDDS alone to classify households into food security categories without complementary indicators and contextual information (FAO, 2011). WFP recommends triangulation with other measures such as Food Consumption Score when the objective is comprehensive consumption analysis, while recognising the practicality of HDDS in rapid assessments and mixed-methods contexts (WFP, 2017).
3.5.1.4 Validation and robustness
Validation is planned through internal consistency checks, correlation with selected related variables and sensitivity to timing. Internal consistency refers to logical coherence in reported consumption patterns. Correlation analysis examines the association between HDDS and asset index and between HDDS and resilience subscales to test expected relationships. Sensitivity analysis includes a check on potential sevenday recall estimates for a subsample where feasible, and regression models include controls for collection month and reported lean season to assess seasonality influence (FAO, 2011).
3.5.2 Resilience indices
3.5.2.1 Concept and dimensions
Resilience is defined here as the capacity to resist or absorb shocks, to adapt livelihood strategies and to achieve structural change that improves long term ability to manage risks. FAO RIMA organises these capacities into absorptive, adaptive and transformative components and proposes composite measurement using latentvariable and factor approaches, applied widely in Sub Saharan Africa to estimate resilience to food insecurity and to identify determinants such as access to services and diversification (FAO, 2016).
3.5.2.2 Measurement approach
Resilience is operationalised through a composite index built from survey items that reflect the three capacities. The items include coping strategies, savings and safety net access for absorptive capacity, counts and balance of income sources and shift in livelihood strategies for adaptive capacity, and group membership, participation in community organisations, access to services and distance to markets for transformative capacity. Each subscale is scored and standardised. The composite combines subscales with initial equal weights, followed by sensitivity analysis that estimates alternative weights through factor analysis or principal components (FAO, 2016). Items for coping strategies draw on established field manuals, distinguishing strategies by severity and frequency where feasible, as recommended in Coping Strategies Index guidance (Maxwell and Caldwell, 2008).
3.5.2.3 Vah'da ton and reliability
Construct validity is assessed through expected correlations with shock exposure and recovery time and associations with asset indices and HDDS. Reliability is assessed through internal consistency metrics such as Cronbach alpha for subscales and composite omega. Factor analysis explores dimensional structure and informs alternative weights. Where data permit, measurement invariance across provinces is examined to test consistency of subscale structure across contexts (Cisse et al., 2016).
3.5.2.4 Interpretation and limitations
Resilience composites summarise complex capacities into tractable indices that are useful in regression and group comparisons. They are sensitive to item selection and weighting, and may under represent informal practices such as reciprocal labour exchange that are difficult to quantify. RIMA emphasises context adaptation with conceptual coherence and recommends triangulation with qualitative insights to capture social capital and institutional dynamics beyond numerical scores (FAO, 2016). WFP guidance similarly underscores combining resilience indicators with process data and community narratives when assessing assistance effectiveness in reducing vulnerability over time (WFP, 2017).
3.5.3 Asset ownership measures
3.5.3.1 Concept and instrument
Asset ownership indicates longer term wealth and livelihood capacity beyond short term income flows. It includes productive assets such as land, livestock and agricultural tools and consumer assets such as radios, bicycles and mobile phones, along with housing quality. In survey practice where expenditure measurement is impractical, asset indices constructed through principal components analysis provide a validated proxy for socioeconomic status and have been adopted widely in large survey programmes (Filmerand Pritchett, 2001).
3.5.3.2 PC A based asset index
The index uses binary and ordinal variables. Variables are standardised, PCA is performed, and the first component is retained as the asset index because it captures the largest variance in holdings. The component is scaled to a standard normal distribution for interpretability (Filmer and Pritchett, 2001). Livestock variables distinguish small stock and large stock. Where data allow, Tropical Livestock Units are computed to standardise holdings across species, improving comparability of livestock as a wealth component in rural contexts (Jahnke, 1982).
3.5.3.3 Alternative methods and sensitivity
Alternative methods include multiple correspondence analysis for categorical variables and polychoric PCA for ordinal variables when normality assumptions are weak. Page 125 of 278
Sensitivity compares the PCA index with alternatives and checks stability of household ranking, also testing the effect of excluding rare assets (Howe et al., 2012).
3.5.3.4 Validation and interpretation
Validation includes correlation with related variables such as household head education where available, building materials and land size, and association with HDDS and resilience subscales. External validation is limited without expenditure data, so triangulation with qualitative accounts supports interpretation. The index is interpreted as a relative ranking within the study sample rather than an absolute measure across unrelated samples (Vyas and Kumaranayake, 2006).
3.5.4 Cross cutting measurement quality
3.5.4.1 Translation and cultural adaptation
Instrument items for HDDS, resilience and assets are translated using forward and back translation by bilingual experts. Special attention is given to food group examples, coping strategy descriptions and asset names to ensure semantic and conceptual accuracy. Back translation with consensus resolution follows standard protocols that improve constructvalidityacross languages (WHO, 2016).
3.5.4.2 Enumerator training and protocol standardisation
Enumerators use standard scripts, neutral probing and consistent classification. For HDDS, they practice mixed dish classification and record all components to assign food groups correctly. For resilience items, they avoid leading respondents and record frequency and severity as specified. For assets, they verify ownership status and functional condition. Training draws on survey methodology that emphasises standardisation and practice to reduce interviewer effects and measurement error (Groves etal., 2009).
3.5.4.3 Data capture and quality checks
Electronic capture includes range checks, skip logic and mandatory fields where appropriate. Timestamps and GPS coordinates support interview verification. Supervisor review identifies out of range values, unusual distributions and inconsistent patterns across modules. Where paper is used, double data entry with discrepancy reconciliation is performed (Lohr, 2010).
3.5.4.4 Missing data handling
Missing data are minimised through instrument design and training. If non-trivial missingness occurs, multiple imputation with chained equations may be applied for continuous variables such as asset index or resilience composite, under a missing at random assumption conditional on observed variables. Diagnostics compare distributions across imputed and observed data and re estimate key models to assess sensitivity (Little and Rubin, 2002).
3.5.5 Integration with qualitative evidence
Indicators are integrated with qualitative findings using joint display matrices and meta inference. For HDDS, qualitative narratives explain mechanisms that influence diet diversity such as payment timeliness in cash-for-work enabling diverse purchases and ration composition in food-for-work stabilising staples while limiting diversity. For resilience, qualitative accounts illuminate social capital, institutional trust and local governance dynamics that shape coping and adaptation. For assets, narratives on livestock retention and investment choices with cash transfers contextualise asset trajectories. Integration procedures follow mixed methods guidance on connecting quantitative and qualitative strands for rich interpretation and policy recommendations (Flick,2018).
3.5.6 Ethical considerations in measurement
Protocols respect autonomy and confidentiality. Dietary diversity questions avoid judgemental phrasing. Coping strategy questions are asked with sensitivity to minimise distress. Asset questions are framed neutrally and avoid implying benefit determination. Page 127 of 278
Consent scripts explain purpose and voluntary participation. Confidentiality is protected through pseudonymisation and aggregate reporting. Ethical guidance from WHO informs these practices and supports do no harm in vulnerable contexts (WHO, 2011).
3.5.7 Limitations and mitigation
Indicators have inherent limitations. HDDS measures variety, not quantity, and is susceptible to seasonality and special events. Resilience composites depend on item selection and weighting and may miss informal practices. Asset indices provide relative ranking and may be influenced by local asset distributions. Mitigation includes triangulation with qualitative evidence, sensitivity analysis of weights and algorithms, controls for seasonality in models and clear interpretation that avoids overstatement. Literature emphasises transparency about indicator limitations and the importance of context in interpreting results for policyand programme design (FAO, 2011).
3.5.8 Planned robustness and sensitivity analyses
Robustness checks are planned for each indicator. For HDDS, models are re estimated with month of interview and lean season controls and with alternative classification of mixed dishes where relevant. For resilience, alternative weights derived from factor loadings are tested and subscale analysis examines drivers of composite effects. For assets, multiple correspondence analysis and exclusion of rare assets are used to test stability of household ranking. These checks follow methodological recommendations and strengthen confidence in comparative findings between food-for-work and cash- for-workhouseholds (WFP, 2017).
3.6 Analytical Techniques
3.6.1 Statistical methods
3.6.1.1 Model specification
Quantitative analysis focuses on continuous outcomes such as Household Dietary Diversity Score, resilience composites and asset indices, and on selected binary outcomes such as use of negative coping strategies. Continuous outcomes are Page 128 of 278 modelled using ordinary least squares with covariates for programme modality food- for-work or cash-for-work, household demographics, shock exposure, socio economic strata and district fixed effects to absorb unobserved contextual heterogeneity. Ordinary least squares are appropriate for estimating linear associations under standard regularity conditions and supports transparent interpretation of marginal effects in programme evaluation (Wooldridge, 2010). For binary outcomes such as use of distress asset sales, logistic regression is applied and average marginal effects are reported to improve interpretability for policy audiences (Hosmer, Lemeshowand Sturdivant, 2013).
District fixed effects help account for local governance, market conditions and service access that are not measured directly but influence outcomes, strengthening internal validity in cross sectional analyses (Angrist and Pischke, 2009).
3.6.1.2 Addressing selection bias
Because programme participation is not randomly assigned, selection on observables is addressed through matching or reweighting. A propensity score is estimated from preprogramme or plausibly exogenous covariates, and inverse probability weights or covariate adjustment with the score are used to balance food-for-work and cash-for- work groups on observed characteristics (Rosenbaum and Rubin, 1983). As a complementary strategy, coarsened exact matching is employed on district, gender of head and socio-economic strata to reduce model dependence and improve covariate balance without overfitting (lacus, King and Porro, 2012). These approaches are standard in observational evaluations when randomisation is infeasible and support credible comparisons under the potential outcome's framework (Imbens, 2004).
Covariate balance is assessed using standardized mean differences and variance ratios before and after matching or weighting, with thresholds such as absolute standardized mean difference less than 0.1 as acceptable balance benchmarks (Austin, 2009).
3.6.1.3 Inference and diagnostics
Standard errors are reported as cluster robust at the community or district level to account for intra cluster correlation due to shared implementation and local conditions, thereby avoiding downward bias in variance estimates (Cameron and Miller, 2015). Heteroskedasticity robust standard errors are used for ordinary least squares to maintain valid inference when error variance is non-constant (White, 1980). Model diagnostics include variance inflation factors to check multicollinearity, link tests for logistic models, goodness of fit checks, and residual inspection for influential observations. Outliers are addressed through sensitivity analyses that trim one percent tails or use robust regression variants as checks (Rousseeuw and Leroy, 2005). Good practice in causal interpretation is followed by distinguishing association from causation, acknowledging that cross sectional models estimate conditional correlations and that directional claims require stronger identification strategies. This stance reflects established guidance in applied econometrics and quasi experimental design (Shadish, Cookand Campbell, 2002).
3.6.1.4 Robustness checks
Robustness checks include alternative HDDS recall handling with timing controls, seasonality proxies such as month of interview and self-reported lean season, alternative resilience weights using factor loadings, exclusion of rare assets in asset index construction and alternative matching specifications. Sensitivity of key coefficients to these variations is reported to demonstrate stability of findings. Such checks are consistent with recommendations for transparent and reliable inference in observational studies (Imbens, 2004).
3.6.2 Qualitative coding frameworks
3.6.2.1 Approach and rationale
Qualitative analysis employs a hybrid coding approach that combines inductive and deductive elements. Inductive coding allows themes to emerge from the data, while deductive coding ensures alignment with research questions concerning resilience, coping strategies, gender and youth dynamics and implementation processes. This hybrid approach is widely recommended to balance flexibility and comparability in qualitative research (Braun and Clarke, 2006). Coding proceeds iteratively with the development of a codebookthat defines each code, inclusion and exclusion criteria and exemplars for consistent application across analysts (Miles, Huberman and Saldana, 2014).
3.6.2.2 Procedures and reliability
Transcripts from key informant interviews and focus groups are imported into qualitative analysis software and coded line by line and at segment level. Analysts meet regularly to refine the codebook and resolve ambiguities. Inter coder reliability is assessed on a subset of transcripts using agreement metrics such as Cohen's kappa, with discussion to address discrepancies and to improve shared understanding of code definitions (Cohen, I960). Trustworthiness is enhanced through triangulation across data sources, audit trails of analytic decisions and reflexive memos that document assumptions and context, consistent with qualitative credibility standards (Lincoln and Guba,1985).
Thematic analysis synthesises codes into categories and higher order themes. Where applicable, matrix displays are used to relate themes across groups such as women, youth and mixed participants, and across provinces to explore contextual variation (Miles etal., 2014).
3.6.2.3 Managing group dynamics and elite capture in analysis
Focus group data are interpreted with awareness of group dynamics, noting instances of dominance by particular participants and considering how segmentation strategies and moderator techniques influenced contributions. Interview narratives from officials and leaders are analysed critically for potential presentational bias and are triangulated with community accounts to mitigate elite capture in interpretation. This reflexive stance aligns with qualitative evaluation guidance and with literature on local power in development (Platteau, 2004).
3.6.3 Mixed methods integration and triangulation
3.6.3.1 Triangulation strategy
Triangulation is used to enhance validity by cross checking evidence across survey findings, interview narratives, focus group themes and documentary sources. Convergence strengthens confidence, while divergence is treated as informative about layered realities. Triangulation is a classic and contemporary strategy in social research and mixed-methods, supporting credible meta inferences when different strands are brought together(Denzin, 1978).
3.6.3.2 Joint displays and meta inference
Joint display matrices are constructed to align quantitative coefficients with qualitative themes. For example, a positive coefficient on cash-for-work for income diversification is juxtaposed with interview accounts of payment delays and inflation pressures, producing meta inferences that specify conditions under which cash flexibility translates into diversification. Joint displays are recommended for systematic integration in mixed methods and help communicate complex findings clearly to decision makers (Guetterman, Fetters and Creswell, 2015).
Example joint display structure:
Quantitative result: p_CFW on HDDS = +0.45 (p < 0.05), district fixed effects included
Qualitative themes: cash enables diverse food purchase;delays undermine planning; local markets constrained seasonally
Meta inference: cash modality associated with higher diet diversity when payments are timely and markets accessible;effects attenuate under delay and thin markets
Document analysis findings are integrated by noting where policy texts and donor reports emphasise specific mechanisms such as empowerment or resilience building and testing these emphases against community narratives and survey evidence. This integration supports context aware recommendations that align programme design with observed constraints and opportunities (Bowen, 2009).
3.6.3.3 Handling convergence and divergence
Both convergence and divergence are explicitly reported. Convergence occurs when survey findings such as higher HDDS among food-for-work households align with qualitative accounts of ration stabilisation. Divergence occurs when hidden costs such as increased workload for women appear in focus groups even when numerical indicators improve. Treating divergence as insight rather than error is consistent with mixed methods guidance and leads to more nuanced policy advice, for example adding safeguards to mitigate workload burdens while maintaining nutritional gains (Fetters, Curryand Creswell, 2013).
3.6.4 Limitations of techniques and threats to validity
3.6.4.1 nternal validity in observational analysis
Cross sectional observational models estimate conditional associations and are vulnerable to omitted variable bias and reverse causation. Matching and fixed effects mitigate but do not eliminate these threats. Sensitivity to unobserved confounding is acknowledged, and directional claims are avoided unless supported by strong design or external evidence, in line with guidance on causal inference without randomisation (Shadish, Cookand Campbell, 2002).
Measurement error in self-reported variables, including income diversification and coping strategies, may attenuate associations. Enumerator training and instrument piloting reduce error, and robustness checks examine stability of results under alternative measures or coding rules (Groves et al., 2009).
3.6.4.2 External validity and generalisability
Purposive selection of provinces and districts constrains generalisability beyond similar contexts with comparable vulnerability and implementation histories. External validity is bounded and clearly stated. Analytic generalisation through case-based reasoning is emphasised, consistent with case study methodology and mixed-methods evaluation practice (Yin, 2018).
3.6.4.3 Qualitative credibility and bias
Qualitative findings are susceptible to moderator influence, group dynamics and informant bias. Trustworthiness is improved through triangulation, inter coder reliability checks, audit trails and reflexive memos. Nonetheless, interpretive subjectivity remains and is acknowledged explicitly, consistent with qualitative standards for credibility and confirmability (Lincoln and Guba, 1985).
3.6.4.4 Integration risks
Mixed-methods integration poses risks of privileging one strand over another or oversimplifying complex divergences. Joint displays and meta inference routines are used to maintain parity and to make integration decisions transparent. Reporting includes both supporting and contradicting evidence to avoid confirmation bias in synthesis, consistent with established mixed-methods integration guidance (Creswell and Plano Clark, 2011).
3.6.5 Reporting standards and transparency
3.6.5.1 Pre-specification and analytic work flow
An analysis plan specifies primary outcomes, covariates, matching or weighting strategies, clustering levels for inference, and planned robustness checks before full analysis. Pre-specification reduces researcher degrees of freedom and improves credibility in observational studies (Imbens, 2004). The workflow includes data cleaning logs, code versioning and reproducible scripts, following good practice for transparency and replicability in social research (ICPSR, 2012).
3.6.5.2 Presen ta ti on
Results are presented with clear tables for model estimates, clustered standard errors, goodness of fit and balance diagnostics for matching or weighting. Figures illustrate joint displays linking quantitative and qualitative strands. Narrative integrates statistical and thematic findings and states limitations and conditions for interpretation, consistent with reporting standards in mixed methods research (Creswell and Plano Clark,2011).
3.6.6 Ethical alignment in analysis
Analysis respects confidentiality and avoids stigmatizing language when discussing coping strategies or dependency. Aggregated reporting at district level reduces re identification risk. When elite capture or targeting concerns are discussed, roles are anonymised and findings focus on systemic patterns rather than individuals. These practices align with ethical guidance for research involving vulnerable populations and community programmes (WHO, 2011).
3.7 Ethical Considerations
3.7.1 Principles and framework
3.7.1.1 Ethical foundations
Ethical conduct in this study is guided by respect for persons, beneficence and justice. These principles shape informed consent procedures, risk minimisation and fair inclusion of participants across gender, age and socio-economic groups. International guidance is used to anchor standards and to adapt them to local contexts, including World Health Organization ethics guidance for field research, the Council for International Organizations of Medical Sciences guidelines and the Belmont Report, which together provide widely recognised frameworks for research involving human participants in diverse settings (WHO, 2011).
3.7.1.2 Context adaptation
Ethical protocols are adapted to the Zambian context through local language delivery, culturally sensitive consent processes and community engagement. Adaptation follows established recommendations to tailor procedures to literacy levels, social norms and power dynamics, while maintaining the core protections for autonomy, confidentiality and safety that international guidelines require (CIOMS, 2016).
3.7.2 Informed consent
3.7.2.1 Consent approach
Consent is obtained prior to any interview or group discussion. In districts where signing documents is culturally sensitive or literacy is low, oral consent is used with clear explanation in local language, recorded by audio or attested by the enumerator. Written consent is used where acceptable and preferred by participants. This dual approach is consistent with ethical guidance that recognises oral consent as valid when documentation would deter participation or introduce fear, provided the process ensures comprehension and voluntariness (WHO, 2011).
3.7.2.2 Elements of consent
Consent scripts cover the study purpose, procedures, expected duration, voluntary nature of participation, the right to skip questions or withdraw at any time, potential risks and discomforts, potential benefits, confidentiality measures and data use. Plain language is used and time is provided for questions. Where youth participate in focus groups, assent is obtained and guardian permission is sought in accordance with guidance forresearch involving minors and other vulnerable groups (CIOMS, 2016).
3.7.2.3 Ongoing consent
Consent is treated as an ongoing process. Enumerators check comfort and willingness to continue during interviews and group sessions, and they pause or stop if participants express distress or wish to withdraw. This practice implements the concept of process consent and ensures respect for autonomy beyond a single signature or verbal agreement (WHO, 2011).
3.7.3 Confidentiality and data protection
3.7.3.1 Confidentiality practices
Interviews are conducted in private settings when possible, away from non-household members. Names and direct identifiers are removed from transcripts and survey data. Reporting aggregates findings at district level and avoids village names to reduce re identification risk. When describing sensitive issues such as elite capture or grievances, roles are anonymised and examples use non-identifying descriptors. These safeguards align with ethical guidance for confidentiality in community research (WHO, 2011).
3.7.3.2 Data security and governance
Survey data are encrypted at capture, and audio files are stored on secure drives with role-based access. Consent records are stored separately from data to minimise linkage risk. Access logs record retrieval and modification events. Backups are maintained with version control. Data management and protection practices follow established guidance for social science data governance, including recommendations on secure storage, controlled access and documentation of processing activities (ICPSR, 2012).
3.7.3.3 Data sharing and retention
Data sharing is limited to the research team. Any future sharing for academic purposes would require de identification and a formal request with clear reuse conditions. Retention periods are defined in the data management plan and balance the need for verification with obligations to minimise risks from prolonged storage, in line with best practice in research data stewardship (UK Data Service, 2015).
3.7.4 Minimisation of risk and do no harm
3.7.4.1 Risk identification
Potential risks include discomfort discussing income or coping strategies, stigma related to unemployment or migration among youth, and community tensions when discussing targeting or elite capture. The study avoids coercion, frames questions neutrally and provides the option to skip sensitive items. This approach aligns with ethical principles of minimising harm and avoiding undue influence in vulnerable settings (WHO, 2011).
3.7.4.2 Distress and referral
Moderators and enumerators are trained to recognise signs of distress and to pause or terminate sessions where necessary. A referral sheet is prepared for locally available support services for issues such as gender-based violence, severe hardship or mental health concerns, recognising the obligation to provide pathways to assistance when research encounters vulnerability. This practice is consistent with do no harm protocols referenced in international guidance (CIOMS, 2016).
3.7.4.3 Remuneration and participation burdens
Participants are offered modest refreshments during focus groups. No cash remuneration is offered that could be construed as inducement. Scheduling avoids peak agricultural periods and domestic responsibilities to reduce opportunity cost. These choices limit participation burden and the risk of undue influence while respecting community rhythms and obligations, in line with ethics guidance on incentives and burdens (CIOMS, 2016).
3.7.5 Equity, inclusion and fairness
3.7.5.1 Gender and youth inclusion
Sampling ensures inclusion of women and youth. Separate focus groups are organised for women and for youth to create comfortable spaces for expression, given observed dynamics where older men may dominate mixed discussions. Facilitation techniques encourage participation by quieter members. This design advances justice and fair representation and aligns with ethical guidance on inclusion of groups at risk of being marginalised in community research (WHO, 2011).
3.7.5.2 Vulnerable households
Households experiencing severe hardship are approached with extra care. Interview length is adjusted if burden is high, and the option to reschedule or to skip sections is offered. Enumerators avoid judgemental language and reassurances are given that responses have no bearing on programme eligibility. These procedures support dignity and autonomy for vulnerable participants, consistent with research ethics standards (CIOMS, 2016).
3.7.6 Ethics in qualitative methods
3.7.6.1 Group confidentiality and ground rules
Focus groups begin with clear ground rules that emphasise respect and confidentiality. Participants are reminded not to share personal stories outside the group and to avoid naming individuals during discussion. Moderators steer conversation away from identifiable details and summarise points at aggregate level when sensitive topics arise. Such practices are recommended in qualitative research ethics to mitigate risks inherent in group settings (Bloor et al., 2001).
3.7.6.2 Managing power and elite capture
When interviewing leaders or officials, moderators use neutral prompts and request concrete examples while discouraging naming of individuals. Triangulation with household and group data is built into analysis to avoid over weighting elite narratives. Reflexive memos document perceived power dynamics and potential presentational bias, aligning with ethical reflexivity and trustworthiness standards in qualitative evaluation (Lincoln and Guba, 1985).
3.7.7 Ethics in quantitative methods
3.7.7.1 Sensitive survey content
Income and coping strategy questions are sequenced later to follow rapport building modules. Enumerators remind respondents that participation is voluntary and that skipping any question is permissible. Scripts avoid evaluative terms and emphasise that the study is independent from programme eligibility decisions. These procedures reflect ethical and methodological guidance to reduce social desirability bias and protect autonomy during surveys in aid contexts (Groves et al., 2009).
3.7.7.2 Anonymisation and linkage
Identifiers are removed from analytical datasets, and any linkage between survey and consent files is restricted and controlled. Such separation reduces re identification risk and aligns with confidentiality obligations in human subject's research (WHO, 2011).
3.7.8 Ethics o versight and appro vals
3.7.8.1 nstitutional review
The study undergoes ethics review by an appropriate institutional review board. Submission materials include the protocol, instruments, consent scripts, field manuals, data protection plan and distress referral procedures. Ethics committees evaluate risk level, consent adequacy, inclusion strategies and data governance to ensure compliance with international standards and local requirements. This process is consistent with WHO recommendations for ethics review of research involving human participants (WHO, 2011).
3.7.8.2 Ongoing monitoring
Supervisors conduct daily briefings to reinforce ethical commitments and to monitor adherence. Deviations from protocol are documented and corrected, and serious incidents or adverse events would be reported to the ethics committee in line with review conditions. Ongoing monitoring is part of good ethical practice for field research and supports continuous risk management (CIOMS, 2016).
3.7.9 Enumerator training and conduct
3.7.9.1 Training content
Training covers respect, neutrality, confidentiality, consent procedures, recognition of distress, safe handling of sensitive topics, and community entry ethics. Role play simulates consent, difficult questions and de-escalation. Moderators and enumerators practice scripts and receive feedback on tone and phrasing. Training for ethical field conduct is integral to high quality research and is emphasised in survey and qualitative method guidance (Kvale and Brinkmann, 2009).
3.7.9.2 Professional boundaries and safety
Enumerators maintain professional boundaries, avoid promises of assistance, and refrain from sharing personal opinions on programmes. Team safety protocols include conflict avoidance, respect for local customs and clear rules for withdrawal from unsafe situations. These practices uphold integrity and reduce risk to participants and staff, consistent with ethical fieldwork standards (WHO, 2011).
3.7.10 Community engagement and feedback
3.7.10.1 Entry and collaboration
Community entry is coordinated with local leaders to ensure transparency and respect. Information meetings explain the study purpose and procedures without compromising individual confidentiality. Collaboration facilitates access while maintaining independence from programme operations to reduce perceived coercion or influence. Engagement practices reflect community-based research ethics recommendations (Chambers, 1997).
3.7.10.2 Feedback and dissemination
Aggregate findings are shared with communities and local stakeholders through debriefs at district level. Results are presented without identifying participants or villages and with emphasis on constructive insights for programme improvement.
Feedback respects confidentiality and promotes accountability, aligning with ethical principles of reciprocityand transparency in applied research (WHO, 2011).
3.7.11 Ethics in analysis and reporting
3.7.11.1 Respectful interpretation
Analytic narratives avoid stigmatizing language. Dependency is discussed as a rational survival strategy under chronic poverty rather than as moral failure. Gendered workload burdens are reported with care and accompanied by suggestions for safeguards such as childcare support or flexible work hours where relevant. This stance reflects beneficence andjustice in reporting and is consistent with ethical guidance for research communication in vulnerable contexts (CIOMS, 2016).
3.7.11.2 Transparency and limitations
Reports include limitations arising from cross sectional timing, recall bias and potential selection into programmes. Uncertainty is acknowledged and recommendations are framed with conditions of applicability. Transparency strengthens ethical accountability and guards against overstatement offindings in policy discourse (WHO, 2011).
3.7.12 Special topics
3.7.12.1 Handling grievances and complaints
Participants who wish to raise grievances or complaints about programme operations or about the study are provided contact information for appropriate local offices and for the research team. A simple mechanism records complaints and documents responses. Managing grievances respects participant rights and aligns with ethical norms of responsiveness in applied research (CIOMS, 2016).
3.7.12.2 Data linkage with programme records
If programme records are used for sampling or verification, linkage is restricted to nonsensitive fields necessary for study implementation. Any analytic linkage is conducted only on de identified datasets. Permission from relevant authorities is obtained and documented. These constraints follow data protection and confidentiality guidance and reduce risk of unintended disclosures (UK Data Service, 2015).
3.7.13 Ethical trade-offs and reflexivity
3.7.13.1 Trade-offs in consent documentation
Written consent provides clear documentation but may intimidate participants. Oral consent respects local norms but reduces formal proof. The study balances these considerations by using oral consent where appropriate and by recording attestations or audio confirmation. Reflexive notes document the rationale and the observed effects on participation and trust, consistent with ethical reflexivity in field research (WHO, 2011).
3.7.13.2 Balancing transparency and confidentiality
Reporting systemic issues such as elite capture promotes accountability but can risk community relationships or safety if individuals are exposed. The study reports patterns at aggregate level and anonymises roles. Reflexive deliberation guides the level of detail to include, ensuring that ethical obligations of protection are prioritised while maintaining transparency about structural problems, in line with qualitative ethics recommendations (Lincoln and Guba, 1985).
3.8 Limitations of the Methodology
3.8.1 Cross sectional design limitations
3.8.1.1 Seasonality and timing
Cross sectional data capture a single time slice and are vulnerable to seasonality effects. In agricultural contexts, household dietary diversity and coping behaviour vary sharply between post-harvest and lean seasons. If interviews occur in periods of relative abundance, food access indicators such as Household Dietary Diversity Score may be elevated, creating optimistic snapshots that do not reflect lean period realities. The study partially mitigates this by including retrospective lean season questions and by controlling for month of interview in models. These steps align with established cautions in survey methodology about timing effects and the need for calendar anchors to improve recall and comparability across respondents and periods (Groves et al., 2009). Mini case Eastern Province: Fieldwork coinciding with post-harvest months produced higher diet diversity reports than community narratives of reduced meal frequency in January to March. This observation is specific to this study and is presented without citation.
3.8.1.2 Causality versus correlation
Cross sectional designs estimate conditional associations rather than causal effects. Although district fixed effects and matching can reduce confounding by observed covariates, unobserved differences in vulnerability or motivation may bias modality comparisons between food-for-work and cash-for-work . The study therefore interprets coefficients as associations and uses sensitivity analyses rather than causal claims. This stance follows guidance in applied econometrics and quasi experimental design regarding limitations of causal inference without random assignment or strong instruments (Angrist and Pischke, 2009).
3.8.1.3 Retrospective recall bias
Retrospective modules for shocks and adaptation over five years are subject to recall errors, telescoping and rationalisation. Calendar aids and culturally salient anchors are used to improve recall, but memory decay and event salience still affect accuracy. This limitation is documented in the survey methods literature, which recommends event calendars and careful question design to reduce but not eliminate recall bias (Tourangeau, Rips and Rasinski, 2000).
3.8.2 Reliance on composite measures
3.8.2.1 Resilience index assumptions
The resilience composite draws on absorptive, adaptive and transformative capacities. Item selection and weighting influence the index, and some informal practices such as reciprocal labour exchange are difficult to quantify, risking under representation. Although equal weights are used initially and sensitivity analyses explore factor derived weights, any composite measure embeds assumptions that shape results. FAO's RIMA framework explicitly recognises these trade-offs and recommends contextual adaptation alongside triangulation with qualitative evidence to capture social and institutional dynamics more fully (FAO, 2016).
Mini case Monze District: Social capital reported in key informant interviews appeared more influential than the numeric score suggested, especially reliance on extended family in drought recovery. This observation is specific to this study and is presented without citation.
3.8.2.2 Asset index composition and PCA limitations
Principal components analysis converts heterogeneous assets into a single wealth continuum. The index is sensitive to the asset list, local prevalence of items and the statistical properties of variables. Rare assets can unduly influence component loadings, and housing quality items may dominate in peri urban areas while livestock dominates in rural areas. Methodological reviews note these concerns and recommend sensitivity checks, multiple correspondence analysis for categorical data and cautious interpretation as a relative ranking within the sample rather than an absolute wealth measure across contexts (Filmer and Pritchett, 2001).
3.8.3 Biases and mitigation strategies
3.8.3.1 Sampling bias
Purposive selection of provinces and districts enhances contextual depth but constrains external validity. Within districts, stratified random sampling improves representativeness, yet sampling frames derived from programme lists may exclude non-participants or hard to reach households. The study documents frame sources and applies replacement rules sparingly to protect randomisation integrity, consistent with survey sampling guidance on multi stage designs and stratification for precision and subgroup analysis (Lohr, 2010).
Mini case Chipata District: Reliance on programme registers initially under represented households on the periphery. Subsequent verification with community registers improved frame coverage. This observation is specific to this study and is presented without citation.
3.8.3.2 Selection bias into programmes
Households may self-select into food-for-work or cash-for-work based on unobserved characteristics such as risk tolerance or social connections. Matching and reweighting reduce imbalance on observed covariates but cannot address unobserved selection fully. The study uses propensity score adjustment and coarsened exact matching on district, gender and socio-economic strata to improve balance, following established practice in observational evaluations (Rosenbaum and Rubin, 1983).
3.8.3.3 Response and social desirability bias
In aid contexts, respondents may overstate programme benefits or under report difficulties due to perceived implications for eligibility or relationships with authorities. Sequencing sensitive questions later, emphasising confidentiality and independence and training enumerators in neutral probing are used to mitigate these effects. Survey methodology highlights social desirability and interviewer respondent dynamics as persistent challenges that require protocol design and training to reduce bias (Tourangeau et al., 2000).
3.8.3.4 Enumerator bias
Local enumerators enhance trust but may inadvertently rephrase questions or signal expected answers. Training, standard scripts, back checks and supervisor observation reduce but do not eliminate these effects. Interviewer effects are well documented, and quality assurance practices are recommended to monitor and correct protocol drift (Lyberg and Biemer, 2008).
3.8.3.5 Document bias and donor narrative
Policy and donor reports can emphasise success tojustify funding and downplay failures or politically sensitive issues. Document analysis treats texts as institutional artefacts and triangulates claims with community narratives and survey indicators. This critical reading approach follows qualitative guidance on document analysis in policy research (Prior, 2003).
Regional comparison Zimbabwe: Donor reports highlighted programme achievements while community narratives raised elite capture concerns.Triangulation balanced these perspectives. This observation is specific to this study and is presented without citation.
3.8.4 Measurement error and construct validity
3.8.4.1 Dietary diversity classification
Mixed dishes and local food naming conventions can lead to misclassification across food groups, affecting Household Dietary Diversity Score. Instrument design includes examples and probes and enumerator training emphasises decomposing mixed dishes. FAO guidance identifies classification and recall challenges and prescribes training and probes to improve accuracy (FAO, 2011).
3.8.42 Income and asset reporting
Income reports are sensitive and subject to rounding or concealment. Asset reporting may confuse ownership with access or shared use. The study minimises detailed income questions and focuses on counts and balance for diversification, while asset modules probe ownership status and functional condition. These choices reflect survey method recommendations for reducing measurement error in sensitive or complex items (Deaton,1997).
3.8.43 External validity and generalisation
Analytic generalisation is prioritised over statistical generalisation. Findings are applicable to contexts with similar vulnerability profiles, market conditions and programme implementation histories. Case study methodology and mixed methods evaluation literature emphasise context bound inference and transparency about transferability conditions to support responsible use of results in policy discourse (Stake, 1995).
3.8.44 Data quality constraints
3.8.44.1 Missing data
Item non-response may occur for sensitive modules. If missingness is nottrivial, multiple imputation with chained equations can reduce bias under a missing at random assumption conditional on observed variables. Diagnostics compare observed and imputed distributions and re estimate key models to assess sensitivity. These procedures follow recognised guidance on handling missing data in social surveys (Rubin,1987).
3.8.44.2 Translation and semantic equivalence
Multi-language instruments risk subtle meaning shifts. Forward and back translation and cognitive interviewing reduce these risks but cannot eliminate cultural nuance differences. Cross cultural instrument guidance recognises that semantic equivalence does not guarantee conceptual equivalence and therefore recommends piloting and consensus resolution, which this study applies (WHO, 2016).
3.8.45 Integration risks in mixed-methods
Integration can over privilege numerical results or, conversely, elevate dramatic qualitative narratives. Joint displays and explicit meta inference routines are used to maintain parity and to report both convergence and divergence. Mixed-methods literature cautions against collapsing divergent findings and recommends transparent synthesis that respects the epistemic contributions of each strand (Creswell and Plano Clark,2011).
3.8.46 Ethical and logistical constraints
3.8.46.1 Consen t and con fid en tiality
Oral consent improves participation where signing is culturally sensitive, yet reduces formal documentation. Confidentiality obligations limit reporting of identifiable details, which can constrain explanation of specific local processes. International ethics guidance accepts oral consent with proper attestation and emphasises confidentiality over disclosure in vulnerable contexts. The study follows these standards while documenting trade-offs (CIOMS, 2016).
3.8.46.2 Field logistics
Access limitations due to weather, road conditions or community events can affect sampling reach and schedule. The study mitigates by flexible planning and by documenting deviations and replacements. Survey practice recognises logistics as a source of non-sampling error and recommends supervisor monitoring and documentation to maintain integrity (Lohr, 2010).
3.8.9 lmplications forinterpretation and future research
Given cross sectional timing, composite measure assumptions and potential selection and response biases, findings are interpreted cautiously as associations conditioned on observed covariates and context. Policy recommendations are framed with conditions of applicability, emphasising payment timeliness forcash-for-work, ration adequacy for food-for-work, and safeguards for gendered workload burdensand elite capture. Future research should consider panel designs to track seasonality and trajectories, embedded process evaluations to document implementation fidelity and mixed methods experiments to test design variants such as cash plus mentorship or food plus diversified ration composition. Longitudinal designs and stronger causal identification strategies such as phased rollouts or natural experiments would strengthen inference on programme effects. These directions align with evaluation science recommendations for building evidence in complex social protection settings where randomisation is often impractical (Shadish, Cookand Campbell, 2002).
4.0 CHAPTER 4 RESULTS AND FINDINGS
4.1 Quantitative Results
4.1.1 Household Dietary Diversity Score (HDDS)
The household survey revealed clear differences in dietary diversity between program participants and non-participants. HDDS was calculated based on the number of food groups consumed in the previous 24 hours.
The table and the bar graph show the Average HDDS Scores by Program Participation and Gender
Illustrations are not included in the reading sample
Data source: Study survey (2025); HDDS summary by household type
HDDS Distributions Using Normal Curves
This figure visualizes household dietary diversity scores (HDDS) using normal curves derived from reported means and standard deviations for each group.The x-axis represents HDDS on a 0-12scale,and the orange vertical line marks the group mean.
Illustrations are not included in the reading sample
Summary Interpretation: Groups associated with program participation (FFW and CFW) show higher HDDS means («6.2-6.S) and relatively narrow spreads, indicating more consistent dietary diversity. Non-participants exhibit the lowest mean («4.8) and wider spread, suggesting greater variability and potential vulnerability. Female-headed households have slightly lower HDDS than male-headed households, reflecting modest gender-related disparities. These densities highlight both central tendency and dispersion, reinforcing observed equity gaps in food access.
ExtendedQuote (Chipata District,Female Participant):
"Before the food-for-work program, we mostly ate nshima with pumpkin leaves. Now, because of the ration we receive, I can add beans and sometimes kapenta. My children are happier because they don't complain of eating the same food every day." This testimony illustrates how FFW programs directly improved dietary diversity, particularly forwomen managing household meals.
Mini Case Study (Eastern Province): In Eastern Province, HDDS scores were elevated due to maize harvests. However, focus groups revealed that households often struggled during the lean season. One participant explained:
"In January, before the harvest, we sometimes eat only once a day. The survey came after harvest, so it looks like we are fine, but that is not the full story." This highlights the seasonality bias inherent in cross-sectional surveys.
Methodological Reflection: HDDS is a useful proxy forfood access but does not capture portion sizes or nutrient adequacy. A household may consume diverse foods in small quantities, inflating HDDS without necessarily improving nutrition.
4.1.2 Resilience Indices
Resilience scores varied significantly across households. Those with diversified income sources and greater asset ownership consistently scored higher. FFW households demonstrated improvements in absorptive capacity in terms of their ability to cope with shocks in the short term while CFW households showed stronger adaptive capacity, reflecting their ability to adjust livelihood strategies over time.
The table and the column graph show Resilience Scores by Program Type
Illustrations are not included in the reading sample
Extended Quote (Monze District, Male Participant):
"When drought comes, I sell one goat to buy maize. That goat is like my insurance. Without livestock, I would have nothing to fall back on."
Mini Case Study (Monze District): Regression analysis showed that livestock ownership was the strongest predictor of resilience, even after controlling for income and program participation. This underscores the importance of structural wealth in sustaining food security.
Methodological Reflection: Measuring resilience is inherently complex. Composite indices risk oversimplifying diverse coping strategies, yet they remain valuable for capturing multidimensional constructs in a standardized way.
4.1.3 Asset Ownership
Asset indices revealed distinct patterns between FFW and CFW households. CFW participants invested more in consumer assets such as radios and bicycles, while FFW households accumulated productive assets like livestock and agricultural tools.
The table shows Asset Ownership Patterns
Illustrations are not included in the reading sample
The Grouped column chart shows the asset ownership patterns by household type.
Illustrations are not included in the reading sample
Data source: Study survey (2025); household asset ownership by program type.
Extended Quote (Centra/ Province, Female Participant):
"The chickens are mine. I decide when to sell them, and I use the money for school fees. My husband controls the cattle, but the chickens give me independence."
Mini Case Study (Central Province): Focus groups revealed that women valued small livestock (chickens, goats) as assets they could control, while men emphasized larger livestock (cattle) as symbols of wealth. This highlight gendered dimensions of asset ownership.
Methodological Reflection: Asset indices are criticized for treating diverse assets as equivalent. Owning a bicycle and owning cattle may both increase scores, but their economic significance differs greatly. This limitation underscores the need for nuanced interpretation.
4.2 Qualitative Results
Qualitative data provided nuanced insights into how households and communities experienced food-for-work (FFW) and cash-for-work (CFW) programs. While quantitative surveys captured measurable outcomes, interviews and focus groups revealed perceptions, lived realities, and social dynamics that numbers alone could not explain.
4.2.1 Key Informant Interviews (KIIs)
Thematic Findings
The table shows Themes Emerging from KIIs
Illustrations are not included in the reading sample
Extended Quotes
Government Official (Ministry of Agriculture, Katete):
"Donors often emphasize resilience, but our communities tell us they need food today. The challenge is balancing immediate needs with long-term goals."
NGO Field Officer (Eastern Province):
"Targeting is always contentious. Chiefs want their relatives included, but we try to ensure fairness. Sometimes this creates tension between traditional authority and program rules."
Mini Case Study: Monze District, Zambia
Klls with local council officials in Monze revealed tensions between donor priorities and community needs. Donors emphasized resilience-building, while communities prioritized immediate food access. This exposed a policy-practice gap: while resilience is a long-term goal, households facing hunger often value short-term food security more highly.
4.2.2 Focus Groups
Gender Perspectives
The table illustrate Gendered Themes from Focus Groups
Illustrations are not included in the reading sample
Extended Quotes
Women's Group (CentralProvince):
"The food-for-work program helps us feed our children, but the work is heavy. After working in the fields, we still cook, fetch water, and care for the family. Sometimes it feels like double labour."
Men's Group (CentralProvince):
"Cash-for-workis better. With money, I can buy livestock or tools. Food is good, but cash gives us choices."
Mini Case Study: Central Province, Zambia
Women's focus groups revealed that FFW reduced reliance on negative coping strategies, such as skipping meals or selling small livestock. One participant explained:
"Before, I sold chickens when food was scarce. Now, with the food ration, I can keep them for future income." This illustrates how FFW programs can protect household assets by reducing distress sales.
Youth Perspectives
The table shows Youth Themes from Focus Groups
Illustrations are not included in the reading sample
Extended Quote (Youth, Eastern Province):
"Cash-for-work is better for us young people. We can buy clothes, pay for school, or even start small businesses. Food is important, but cash gives us freedom."
Mini Case Study: Mozambique (Regional Comparison)
Youth participants in Mozambique emphasized skill development in construction and road maintenance, viewing CFW as a pathway to formal employment. This contrasts with Zambia, where youth prioritized cash flexibility. The comparison highlights generational differences in program impacts across contexts.
4.3 Document Analysis Findings
Document analysis provided critical insights into the institutional and policy frameworks shaping FFW and CFW programs. By reviewing national policies, donor reports, and community records, the study situated household-level findings within broader governance and development contexts. This triangulation revealed both alignment and divergence between official narratives and lived realities.
4.3.1 Policy Frameworks
The table shows NationalPolicyFrameworks Relevant to FFW/CFW.
Illustrations are not included in the reading sample
Extended Excerpt (Zambia's National Social Protection Policy, 2014):
"Social protection shall prioritize cash transfers as a means of empowering households, while food-based interventions will be considered in contexts of acute vulnerability." This excerpt highlights Zambia's policy preference for cash transfers, contrasting with donor emphasis on food-based interventions.
4.3.1 Donor Reports
Donor reports provided valuable but sometimes biased perspectives. They often highlighted successes tojustify funding, while downplaying challenges.
The table shows Donor Report Narratives
Illustrations are not included in the reading sample
Extended Excerpt (WFP CFS VA Zambia, 2019):
"Households participating in food-for-work programs demonstrated higher dietary diversity scores and reported fewer instances of meal skipping compared to nonparticipants." While positive, this narrative omits the gendered burdens revealed in focus groups.
4.3.2 Community Records
Community-level documents, such as ward development committee minutes, provided grounded perspectives on program implementation.
Extended Excerpt (Ward Committee Minutes, Chipata District):
"Food distribution was delayed by two weeks due to transport challenges. Some households expressed frustration, noting that the delay coincided with the lean season." This excerpt illustrates how logistical challenges undermine program effectiveness, a reality often absents from donor reports.
4.3.3 Mini Case Studies: Regional Comparisons
Zambia vs. Malawi
In Zambia, policy frameworks emphasized cash transfers, while donor reports highlighted food-based interventions. In Malawi, FFW was integrated into public works programs, explicitly framed as resilience-building. KIIs revealed that Malawian chiefs sometimes overstated program benefits to secure donor support, raising ethical concerns about misrepresentation.
Zambia vs. Mozambique
Mozambique's policy emphasized skill development through CFW, linking social protection to infrastructure projects. Youth focus groups confirmed this, noting that CFW provided construction skills. In Zambia, youth prioritized cash flexibility, highlighting different generational impacts.
Zambia vs. Zimbabwe
Zimbabwe's policy tied FFW to climate resilience. Donor reports emphasized adaptation successes, but community narratives revealed elite capture, with local leaders manipulating targeting. This divergence underscores the importance of triangulating donor narratives with community voices.
4.4 Integrated Analysis
Integrated analysis brings together quantitative and qualitative findings, highlighting areas of convergence and divergence across data sources. By triangulating household surveys, KIIs, focus groups, and document analysis, the study provides a multidimensional understanding of FFW and CFW programs. This approach ensures that results are not interpreted in isolation but situated within broader social, institutional, and policy contexts.
4.4.1 Convergence Across Data Sources
The table shows Areas of Convergence across data sources.
Illustrations are not included in the reading sample
Extended Quote (CentralProvince, Women's Focus Group):
"Because of the food ration, I no longer sell my chickens when food is scarce. I can keep them for school fees later." This aligns with survey data showing higher asset retention among FFW households and donor reports emphasizing asset ownership as a resilience strategy.
4.4.2 Divergence Across Data Sources
The table illustrate Areas of Divergence
Illustrations are not included in the reading sample
Extended Quote (CentralProvince, Women's Focus Group):
"The food helps, but the work is heavy. After working in the fields, we still cook, fetch water, and care for the family. Sometimes it feels like double labour."
This divergence illustrates how surveys may overlook gendered burdens, while qualitative data reveals hidden costs of participation.
4.4.3 Mini Case Studies of Triangulation
Case Study 1: Chipata District, Zambia
In Chipata District, survey data revealed that households participating in Food-for-Work (FFW) programs recorded significantly higher Household Dietary Diversity Scores (HDDS) compared to non-participants. This quantitative evidence suggested that the program was positively influencing dietary outcomes. However, focus group discussions added nuance to these findings. Women participants emphasized that child nutrition had indeed improved, but they also expressed concerns about the increased workload that accompanied their involvement in FFW activities. Community records provided yet another perspective, with ward minutes documenting delays in food distribution that affected program efficiency. When these strands of evidence were integrated, the insight was clear: while FFW initiatives contributed to improved dietary diversity, logistical challenges and gendered burdens tempered the overall benefits, underscoring the importance of examining multiple sources to capture the full picture.
Case Study 2: Mozambique (Regional Comparison)
In Mozambique, surveys indicated that households engaged in Cash-for-Work (CFW) programs reported higher resilience scores, largely attributed to income diversification. This quantitative finding was reinforced by focus group discussions with youth, who highlighted the value of skill development in construction and road maintenance. Their testimonies pointed to immediate benefits in terms of employability and community infrastructure. Donor reports added another layer, emphasizing CFW as a pathway to formal employment and long-term livelihood opportunities. When triangulated, these sources revealed that CFW programs not only enhanced household resilience but also created sustainable livelihood prospects. This dimension of long-term opportunity was notably absent in Zambia's youth narratives, highlighting regional differences in how program impacts were experienced and perceived.
Case Study 3: Zimbabwe
In Zimbabwe, survey results showed that households participating in FFW programs reported improved food security, suggesting positive outcomes at the household level. Yet, focus group discussions uncovered troubling dynamics within communities. Participants revealed instances of elite capture, where local leaders favoured relatives in the allocation of program benefits, undermining fairness and inclusivity. Donor reports, in contrast, emphasized the successes of FFW programs in building climate resilience, presenting a more optimistic narrative. The triangulation of these perspectives exposed a divergence between donor accounts and community realities. This gap underscored the critical role of triangulation in identifying governance challenges and ensuring that program evaluations reflect both successes and shortcomings.
4.4.4 Methodological Reflection
Triangulation enriched the analysis by:
• Validating findings: Convergence across surveys, focus groups, and donor reports strengthened confidence in results.
• Revealing contradictions: Divergence highlighted hidden dynamics, such as gendered burdens and elite capture.
• Enhancing policy relevance: Integrated insights provided a balanced view, ensuring that recommendations address both successes and challenges.
Extended Quote (NGO Field Officer, Eastern Province):
"Numbers tell us households are eating better, but when you sit with women, they tell you about exhaustion. Both are true, and both matters."
This reflection captures the essence of integrated analysis: acknowledging complexity ratherthan seeking simplistic conclusions.
4.5 Regional Comparisons
Regional comparisons provide valuable context for interpreting Zambia's FFW and CFW programs. By examining similarities and differences with Malawi, Mozambique, and Zimbabwe, we can identify diverse pathways to resilience, highlight best practices, and expose recurring challenges across Southern Africa.
4.5.1 Zambia vs. Malawi
The table shows Comparative Features of FFW/CFWin Zambia and Malawi.
Illustrations are not included in the reading sample
Extended Quote (Malawi, Local Chief):
"We tell donors the program is working well, because if we complain too much, they may leave. But in truth, some households still struggle."
Mini Case Study (Malawi): In Malawi, FFW was explicitly integrated into public works programs, such as road maintenance and irrigation schemes. While surveys showed improved food security, KIIs revealed that chiefs sometimes overstated program benefits to secure continued donor support. This raised ethical concerns about misrepresentation and accountability.
4.5.2 Zambia vs. Mozambique
The table Compares Features of FFW/CFWin Zambia and Mozambique
Illustrations are not included in the reading sample
Extended Quote (Mozambique, Youth Participant):
"Through cash-for-work, I learned how to build roads. Now I can find jobs in construction. It is more thanjust money; it is a skill."
Mini Case Study (Mozambique): Youth participants in Mozambique emphasized skill development as a key benefit of CFW programs. Unlike Zambia, where youth prioritized cash flexibility, Mozambican youth viewed CFW as a pathway to formal employment. Donor reports confirmed this, highlighting infrastructure projects that doubled as training grounds. However, wage delays sometimes undermined trust in the program.
4.5.3 Zambia vs. Zimbabwe
The table shows Comparative Features of FFW/CFW in Zambia and Zimbabwe
Illustrations are not included in the reading sample
Extended Quote (Zimbabwe, Community Member):
"The program is supposed to help everyone, but the councillor's relatives are always first on the list. We feel left out."
Mini Case Study (Zimbabwe): In Zimbabwe, donor reports emphasized FFW as a climate resilience strategy, linking food security to adaptation goals. Surveys confirmed improved food security, but focus groups revealed elite capture, with local leaders manipulating targeting. This divergence between donor narratives and community realities underscored the importance of triangulation to expose governance challenges.
4.5.4 Synthesis ofRegionai Comparisons
The tabie shows Cross-Country Themes
Illustrations are not included in the reading sample
Extended Quote (Regional NGO Advisor):
"Each country frames these programs differently: Zambia as social protection, Malawi as public works, Mozambique as skills, and Zimbabwe as climate resilience. But the common thread is that communities want fairness and sustainability."
4.6 Discussion of Emerging Themes
Emerging themes provide a deeper understanding of how FFW and CFW programs shape household resilience, gender dynamics, and community governance. By integrating survey results, focus group narratives, Klls, and document analysis, four major themes emerged: dependency, gender equity, elite capture, and resilience pathways.
4.6.1 Dependency
The table shows Evidence of Dependency Across Data Sources
Illustrations are not included in the reading sample
Extended Quote (Chipata District Male Participant):
"Since the program started, l spend less time in my own fields. The food ration is enough for now, but l worry what will happen when it ends."
Mini Case Study (Chipata District, Zambia): Surveys showed improved HDDS among FFW households, but focus groups revealed that some reduced farming activities, relying instead on rations. Klls confirmed officials' concerns about sustainability. This illustrates how dependency can coexist with short-term gains.
Regional Comparison (Malawi): In Malawi, chiefs reported that households sometimes delayed planting, expecting food aid. Donor reports framed this as resilience support, but community narratives revealed dependency risks.
4.6.2 Gender Equity
The table shows Gendered Impacts of FFW/CFW
Illustrations are not included in the reading sample
Extended Quote (Central Province, Woman Participant):
"The food helps us feed our children, but the work is heavy. After the fields, we still cook, fetch water, and care for the family. It feels like double labour."
Mini Case Study (Monze District, Zambia): Women reported that FFW increased their workload, balancing program participation with domestic responsibilities. Men emphasized asset accumulation. This divergence highlights gendered burdens and benefits.
Regional Comparison (Mozambique): Youth focus groups emphasized skill development, but gendered impacts were less documented. This suggests that gender equity issues may be underreported in some contexts.
4.6.3 Elite Capture
The table illustrates Evidence of Elite Capture
Illustrations are not included in the reading sample
Extended Quote (Zimbabwe, Community Member):
"The councillor's relatives are always first on the list. We feel left out, even though we are hungrytoo."
Mini Case Study (Zambia, Chipata District): Focus groups revealed that chiefs prioritized relatives for program participation. Klls confirmed tensions between traditional authority and program rules. Surveys did not capture this dynamic, underscoring the importance of qualitative data.
Regional Comparison (Zimbabwe): Donor reports emphasized climate resilience successes, but community narratives revealed elite capture. This divergence highlights governance challenges across contexts.
4.6.4 Resilience Pathways
The table shows Pathways to Resilience
Illustrations are not included in the reading sample
Extended Quote (Monze District, Male Participant):
"When drought comes, I sell one goat to buy maize. That goat is like my insurance. Without livestock, I would have nothing to fall back on."
Mini Case Study (Central Province, Zambia): Resilience was closely linked to social capital. Households relied on extended family networks for support, a dimension not fully captured in resilience indices. This highlights the importance of informal systems.
Regional Comparison (Mozambique): CFW programs provided youth with construction skills, enhancing employability. This created resilience pathways beyond immediate food security, linking social protection to long-term livelihoods.
5.0 CHAPTERS DISCUSSION
5.1 nterpretation of Quantitative Findings
5.1.1 Dietary Diversity
The household survey results provide clear evidence that participation in Food-for- Work (FFW) and Cash-for-Work (CFW) programs is associated with higher Household Dietary Diversity Scores (HDDS) compared to non-participants. This finding is significant because HDDS serves as a proxy for household food access and nutritional quality, reflecting the number of food groups consumed in the previous 24 hours. Both program types appear to enhance dietary diversity, though in different ways.
FFW households achieved an average HDDS of 6.2, which can be directly linked to the food rations distributed through the program. These rations supplement household diets by adding variety beyond the staple maize, enabling families to incorporate vegetables, legumes, and occasionally protein sources such as fish. The structured nature of FFW ensures that households receive tangible food support, which translates into immediate improvements in dietary diversity.
CFW households reported a slightly higher average HDDS of 6.5. This difference, though modest, reflects the flexibility inherent in cash transfers. Unlike food rations, cash allows households to make choices based on their preferences and needs. Families can purchase a wider range of foods, including perishable items such as fresh vegetables, fruits, and animal proteins, which may not be consistently available through ration distributions. This purchasing power not only diversifies diets but also empowers households to tailor consumption to cultural preferences and nutritional priorities.
In contrast, non-participant households recorded a much lower average HDDS of 4.8. This figure highlights their reliance on staple foods, particularly maize, with limited access to complementary food groups. Such households are more vulnerable to nutritional deficiencies, as their diets lack the diversity needed to ensure adequate intake of essential vitamins and proteins. The gap between participants and nonparticipants underscores the role of program interventions in bridging food access inequalities.
However, the interpretation of these results is complicated by seasonal fluctuations. HDDS scores collected immediately after harvest may overstate long-term nutritional adequacy. During post-harvest periods, households typically consume more diverse foods, benefiting from fresh supplies of maize, vegetables, and other crops. Yet, during lean seasons, dietary diversity declines sharply as households struggle to secure even basic staples. This limitation was vividly captured in an extended quote from Eastern Province, Zambia, where a participant explained, "The survey showed we eat more types of food now, but in January we still struggle. The program helps, but hunger returns before harvest."
A mini case study from Chipata District further illustrates this dynamic. FFW households reported consuming vegetables and fish in addition to maize, while non-participants relied almost exclusively on maize. While this demonstrates the positive impact of FFW in supplementing staple-based diets, it also highlights the risk of seasonal bias in survey data. If surveys are conducted only during periods of relative abundance, they may fail to capture the cyclical nature of food insecurity and the persistent struggles households face during lean months.
Taken together, HDDS improvements reflect short-term gains in dietary diversity but may not necessarily indicate sustained nutritional adequacy. Programs must therefore move beyond immediate outcomes and consider strategies that address seasonality and long-term dietary stability. This could include promoting household-level food production, supporting income diversification, and ensuring year-round access to diverse foods. Without such measures, households may continue to experience cycles of abundance and scarcity, undermining the long-term nutritional impact of FFW and CFW interventions.
5.1.2 Resilience Indices
The analysis of resilience indices revealed significant variation across households, reflecting differences in livelihood strategies, resource ownership, and program participation. Households with diversified income sources and greater asset ownership consistently achieved higher resilience scores, underscoring the importance of both economic diversification and structural wealth in sustaining food security. These findings highlight that resilience is not evenly distributed and is shaped by the interplay of program support and household-level assets.
Food-for-Work (FFW) households demonstrated notable improvements in absorptive capacity, scoring 0.65. This dimension reflects their ability to cope with shocks in the short term, such as droughts or sudden food shortages. The provision of food rations through FFW programs played a critical role in stabilizing consumption during crises, ensuring that households could meet immediate needs. However, their adaptive capacity (0.48) and transformative capacity (0.42) were lower, suggesting that while FFW programs strengthen immediate coping mechanisms, they are less effective in enabling households to adjust livelihood strategies or achieve structural change over time. This indicates that FFW interventions are primarily protective rather than transformative, offering short-term relief but limited pathways for long-term resilience.
In contrast, Cash-for-Work (CFW) households displayed a different resilience profile. Their adaptive capacity was the strongest across all groups, scoring 0.62, which reflects their ability to adjust livelihood strategies and diversify income sources over the long term. This adaptability is linked to the flexibility of cash transfers, which allow households to make choices tailored to their needs, invest in new opportunities, and respond proactively to changing conditions. Their transformative capacity (0.47) was also higher than that of FFW households, suggesting that CFW programs contribute to structural improvements in livelihoods, such as asset accumulation and reduced dependency on aid. Although their absorptive capacity (0.58) was slightly lower than FFW participants, CFW households still demonstrated strong short-term coping abilities, showing that cash transfers provide both immediate stability and long-term adaptability.
Non-participants consistently scored lowest across all dimensions, with absorptive capacity at 0.40, adaptive capacity at 0.35, and transformative capacity at 0.30. These figures highlight the vulnerability of households outside program interventions, who lack both immediate coping mechanisms and the ability to adapt or transform their livelihoods. Their limited resilience underscores the importance of targeted support to ensure that vulnerable households are not left behind.
Qualitative evidence reinforces these quantitative findings. An extended quote from Monze District illustrates the importance of assets in resilience. A male participant explained, "When drought comes, I sell one goat to buy maize. That goat is like my insurance. Without livestock, I would have nothing to fall back on."This testimony highlights how livestock ownership functions as a safety net, enabling households to convert assets into food during crises. Regression analysis confirmed this perspective, showing that livestock ownership was the strongest predictor of resilience, even after controlling for income and program participation. This underscores the central role of structural wealth particularly livestock in sustaining food security and resilience.
Taken together, these findings demonstrate that resilience pathways differ by program type. FFW programs provide immediate coping mechanisms through food rations, strengthening absorptive capacity, while CFW programs foster long-term adaptability by enabling households to diversify and invest in new livelihood strategies. Both approaches are necessary, but they must be balanced to ensure households are equipped to manage both short-term shocks and long-term challenges. Integrating the strengths of FFW and CFW could create a more comprehensive resilience framework, combining immediate protection with pathways for sustainable transformation.
5.1.3 Asset Ownership
The analysis of asset indices revealed clear and distinct patterns between households participating in Food-for-Work (FFW) and Cash-for-Work (CFW) programs, reflecting differences in how program benefits were utilized. These patterns provide important insights into the ways households translate program support into tangible resources that shape their livelihoods.
Households engaged in CFW programs tended to invest more in consumer assets such as radios and bicycles. Radios enhanced communication and access to information, allowing families to stay informed about market prices, weather forecasts, and community events. Bicycles improved mobility, enabling household members to reach markets, schools, and health facilities more easily. Together, these consumer assets supported income diversification and strengthened social connectivity, as households were better able to participate in economic and social networks. The flexibility of cash transfers appears to have encouraged households to prioritize assets that improve everyday convenience and expand opportunities beyond subsistence farming.
In contrast, FFW households accumulated more productive assets, including livestock and agricultural tools. Livestock ownership provided both food and income security, serving as a buffer during crises and a source of protein in household diets. Agricultural tools strengthened farming capacity, enabling households to cultivate larger plots, improve yields, and maintain food production. These investments align closely with the program's emphasis on food security and agricultural support, as food rations freed up resources that households could redirect toward strengthening their productive base. FFW households thus demonstrated a tendency to channel program benefits into assets that directly reinforce agricultural livelihoods and food availability.
Gender differences in asset ownership were also evident and highlight the social dimensions of household resilience. Women placed greater value on small livestock, such as chickens and goats, because these assets were typically under their control and could be sold quickly to meet household needs, including school fees orfood purchases. Small livestock provided women with immediate liquidity and autonomy, giving them a measure of independence in household decision-making. Men, however, emphasized larger livestock such as cattle and consumer assets like bicycles, which they viewed as symbols of wealth, prestige, and long-term security. This divergence illustrates the gendered dimensions of asset ownership, where women prioritize assets that provide practical and immediate benefits, while men focus on assets that represent status and structural wealth.
Taken together, these findings demonstrate that program participation not only influences the types of assets households acquire but also interacts with gender dynamics to shape patterns of ownership and control. Recognizing these differences is essential for designing interventions that promote both economic resilience and gender equity. Programs that encourage women's ownership of productive assets, while also supporting men's investment in household-level resources, can help balance power dynamics and ensure that resilience strategies are inclusive. By tailoring asset support to reflect both economic and gendered realities, interventions can strengthen household resilience while advancing broader goals of empowerment and equality.
5.2 Interpretation of Qualitative Findings
The qualitative data provided nuanced insights into how households and communities experienced Food-for-Work (FFW) and Cash-for-Work (CFW) programs. While household surveys captured measurable outcomes such as dietary diversity and resilience scores, interviews and focus group discussions revealed perceptions, lived realities, and social dynamics that numbers alone could not explain. These narratives highlighted the complexity of program impacts, showing how interventions were understood and valued differently by donors, government officials, communities, women, and youth. Three major themes emerged from the qualitative evidence: policy practice gaps, gendered burdens, and youth perspectives.
5.2.1 Policy-Practice Gaps
One of the most consistent themes across interviews was the tension between donor priorities and community needs. Donors frequently emphasized long-term resilience and accountability, framing programs as vehicles for sustainability and structural change. Reports and policy documents reflected this emphasis, often highlighting resilience-building as the ultimate goal. Communities, however, prioritized immediate food access and fairness, with focus groups repeatedly stressing the importance of food security and equitable distribution. Government officials attempted to balance these perspectives, promoting cash transfers as a pathway to resilience while still allowing food aid in times of crisis.
This divergence was captured in an extended quote from a government official in Katete: "Donors often emphasize resilience, but our communities tell us they need food today. The challenge is balancing immediate needs with long-term goals." The statement underscores the policy-practice gap: while resilience is a desirable long-term outcome, households facing hunger often value short-term food security more highly.
A mini case study from Monze District further illustrates this tension. Key informant interviews revealed that donors consistently framed resilience as the central priority, while communities expressed frustration that immediate food needs were not adequately addressed. This exposed a disconnect between policy frameworks and lived realities. The interpretation is clear: programs must reconcile differing time horizons such as meeting urgent survival needs while simultaneously building pathways to resilience. Without this balance, interventions risk losing legitimacy among the very communities they aim to serve.
5.2.2 Gendered Burdens
Qualitative evidence also revealed the gendered dimensions of program impacts. Women and men experienced FFW and CFW differently, with women often bearing disproportionate burdens. Women reported that program participation improved child nutrition, as food rations allowed them to provide more balanced meals. However, they also described a "double labour" burden, balancing program work with domestic responsibilities such as cooking, fetching water, and childcare. Men acknowledged nutritional improvements but tended to emphasize asset accumulation, viewing program benefits primarily in terms of household-level gains.
An extended quote from a woman in Central Province captures this reality: "The food helps us feed our children, but the work is heavy. After the fields, we still cook, fetch water, and care forthe family. I t feels like double l abour."Vh\s testimony highlights how women's participation in FFW programs often comes at the cost of increased workload, reinforcing existing gender inequities.
A mini case study from Monze District confirmed this divergence. Women reported that FFW participation increased their workload, while men emphasized the accumulation of assets such as livestock and tools. This difference underscores the gendered burdens and benefits of program participation. The interpretation is that gender equity requires program design that acknowledges and mitigates women's workload. Measures such as childcare support, flexible work hours, or gender-sensitive targeting could help ensure that women benefit from programs without being overburdened. Without such adjustments, FFW risks reinforcing inequities even as it improves food security.
5.2.3 Youth Perspectives
Youth perspectives added another layer of complexity to program impacts. Focus group discussions revealed that young people valued CFW programs for the employment opportunities they provided, with many describing cash-for-work as "something to do" that gave them purpose and income. Skill development was also mentioned, with youth reporting that they learned construction and road maintenance skills. Aspirations emerged as a medium-frequency theme, with some participants expressing hopes of saving for college or starting small businesses.
An extended quote from a youth participant in Eastern Province illustrates this perspective: "Cash-for-work is better for us young people We can buy clothes, pay for school, or even start small businesses. Food is important, but cash gives us freedom." This statement highlights the value of cash transfers in providing flexibility and autonomy, particularly for younger participants who are eager to pursue education, entrepreneurship, and personal goals.
A regional comparison with Mozambique adds further nuance. Youth participants there emphasized skill development in construction and road maintenance, viewing CFW as a pathway to formal employment. In Zambia, however, youth prioritized the flexibility of cash, using it to meet immediate needs and pursue personal aspirations. This contrast highlights generational and contextual differences in program impacts, showing that youth value both immediate benefits and long-term opportunities, but in different ways depending on local conditions.
The interpretation is that youth perspectives highlight the need for programs that combine immediate benefits with long-term skill development. Cash transfers provide flexibility and autonomy, but skills create pathways to sustainable livelihoods. Programs that integrate both elements are more likely to meet the aspirations of young people while also contributing to resilience and economic transformation.
Synthesis
Taken together, the qualitative findings reveal that FFW and CFW programs are experienced differently across stakeholder groups, genders, and generations. Donors emphasize resilience, communities prioritize survival, women face disproportionate burdens, and youth seek both flexibility and opportunity. These insights underscore the importance of designing programs that are not only technically sound but also socially responsive. Reconciling policy-practice gaps, addressing gender inequities, and integrating youth aspirations into program design are essential steps toward building interventions that are both effective and legitimate.
5.3 Regional Lessons
Regional comparisons provide valuable context for interpreting Zambia's Food-for- Work (FFW) and Cash-for-Work (CFW) programs. By examining similarities and differences with neighbouring countries such as Malawi, Mozambique, and Zimbabwe, Page 180 of 278 we can identify diverse pathways to resilience, highlight best practices, and expose recurring challenges across Southern Africa. These lessons demonstrate that while program modalities may differ, communities across the region share common concerns aboutfairness, sustainability, and accountability.
5.3.1 Lessons from Malawi
In Malawi, FFW programs were explicitly integrated into public works initiatives, such as road maintenance and irrigation schemes. This approach linked food security interventions with infrastructure development, ensuring that households not only received food but also contributed to community assets. Surveys showed improved food security outcomes, but qualitative evidence revealed tensions in program implementation. Chiefs, who played a central role in targeting, sometimes overstated program benefits to donors in order to secure continued support.
An extended quote from a local chief illustrates this dynamic: "We tell donors the program i s working well, because if we complain too much,they may l eave But i n truth, some households still struggle.'' This testimony highlights the ethical concerns surrounding misrepresentation and accountability. While donor reports emphasized resilience and sustainability, community realities revealed ongoing struggles with food insecurity.
The mini case study from Malawi underscores the importance of triangulating survey data with qualitative insights. While FFW integration into public works projects offers Zambia a valuable model for linking food security with infrastructure development, it also exposes risks of elite influence and misrepresentation. The interpretation is clear: Zambia can learn from Malawi's integration of FFW into infrastructure projects but must guard against manipulation by local leaders to ensure transparency and accountability.
5.3.2 Lessons from Mozambique
Mozambique offers a contrasting perspective, particularly in relation to youth engagement. Policy emphasis there linked social protection programs directly to skill Page 181 of 278 development, with CFW projects designed to provide both income and employable skills. Youth participants consistently highlighted the value of learning construction and road maintenance skills, viewing CFW not just as a source of cash but as a pathway to formal employment.
One youth participant explained: "Through cash-for-work, I l earned how to build roads. Now l ean find j obs i n construction. I t i s more than j ust money: it i saskiH"Vh\s statement captures the transformative potential of CFW when combined with training components. Donor reports confirmed this, noting that infrastructure projects often doubled as training grounds. However, wage delays occasionally undermined trust in the program, raising concerns about reliability and sustainability.
The mini case study from Mozambique highlights the divergence between Zambia and Mozambique in youth perspectives. While Zambian youth prioritized cash flexibility such as valuing the freedom to purchase goods, pay school fees, or start small businesses, Mozambican youth emphasized skill development as a foundation for longterm resilience. The interpretation is that Zambia can strengthen its CFW programs by incorporating training components, ensuring that cash transfers not only provide immediate flexibility but also build employable skills that support sustainable livelihoods.
5.3.3 Lessons from Zimbabwe
Zimbabwe's experience demonstrates how FFW programs can be framed within broader climate resilience and disaster risk reduction strategies. Donor narratives emphasized adaptation successes, linking food security interventions to climate resilience goals. Surveys confirmed improved food security outcomes, but community narratives revealed persistent governance challenges.
An extended quote from a community member illustrates this frustration: "Theprogram is supposed to hep everyone, but the councillor's relatives are always first on the list.
We feel left out." This testimony highlights the problem of elite capture, where local leaders manipulated targeting to benefit their relatives or allies.
The mini case study from Zimbabwe underscores the divergence between donor narratives and community realities. While donors celebrated climate adaptation successes, focus groups revealed exclusion and manipulation at the local level. This case highlights the importance of triangulation which involves using multiple sources of evidence to expose governance challenges that may otherwise remain hidden. For Zambia, the lesson is twofold: FFW can be integrated into climate resilience strategies, but transparency and accountability mechanisms must be strengthened to prevent elite capture and ensure equitable access.
5.3.4 Synthesis of Regional Lessons
Taken together, regional comparisons highlight diverse pathways to resilience across Southern Africa. Zambia frames its programs primarily as social protection, Malawi integrates FFW into public works, Mozambique emphasizes skills and infrastructure, and Zimbabwe positions FFW within climate resilience strategies. Youth perspectives also vary: Zambian youth value cash flexibility, Mozambican youth prioritize skill development, while youth voices are less emphasized in Malawi and Zimbabwe. Gender dynamics reveal recurring challenges, with women in Zambia reporting workload burdens, Malawian communities raising dependency concerns, and Zimbabwean households highlighting elite capture issues.
An extended quote from a regional NGO advisor captures this diversity: "Each country frames these programs differently:Zambia as social protection Malawi as public works, Mozambique as skills, and Zimbabwe as climate resilience. But the common thread is that communities want fairness and sustainability."
The interpretation is that Zambia can draw lessons from each country: Malawi's integration of FFW into infrastructure projects, Mozambique's skill-building approach to CFW, and Zimbabwe's climate resilience framing. However, governance challenges such as misrepresentation, wage delays, and elite capture remain common across contexts. Addressing these challenges is essential to ensure that programs are not only technically effective but also socially legitimate.
5.4 Policy Implications
The findings from Chapter 4 and their interpretation in Chapter 5 reveal several critical policy implications for the design and implementation of Food-for-Work (FFW) and Cash-for-Work (CFW) programs. These implications address dependency, gender equity, elite capture, and resilience pathways. They also highlight the importance of integrating lessons from regional experiences in Malawi, Mozambique, and Zimbabwe. Taken together, the evidence underscores the need for policies that balance immediate food security with long-term sustainability, ensuring that interventions remain both effective and legitimate.
5.4.1 Addressing Dependency
Dependency emerged as a recurring concern across both quantitative and qualitative findings. While FFW programs provide immediate relief through food rations, there is evidence that households sometimes reduce farming activities after joining, relying heavily on rations instead of continuing agricultural production. This was observed in Chipata District, where households deprioritized farming once food support was guaranteed. Such patterns raise concerns about sustainability and the unintended consequences ofaid.
A government official in kabwe captured this challenge succinctly: "W/emustensurethat food-for-work is not Just feeding people today but building their capacity for tomorrow." This statement underscores the importance of linking short-term support with long-term capacity-building.
Policy options to mitigate dependency include:
• Linking FFW to productive assets such as tools, seeds, or livestock, encouraging households to continue farming while benefiting from food support.
• Integrating CFW with vocational training, ensuring that cash transfers are not only consumed but also invested in skills that generate future income.
• Seasonal targeting, aligning support with lean seasons to reduce reliance outside crisis periods.
Interpretation: Policies must balance immediate food security with long-term sustainability. Linking transfers to productive assets and training reduces dependency risks and ensures that households remain active participants in their own resiliencebuilding.
5.4.2 Promoting Gender Equity
Gender dynamics emerged as a critical theme in program impacts. Women often reported "double labour" burdens, balancing program participation with domestic responsibilities such as cooking, childcare, and water collection. While FFW programs improved child nutrition, they also intensified women's workload. Men, by contrast, emphasized asset accumulation and viewed program benefits primarily at the household level.
A woman from Central Province explained: "The chickens are mine, i decide when to sell them. My husband controls the cattle, but the chickens give me independence." This testimony highlights how small livestock provide women with autonomy and decisionmaking power, even within households where men control larger assets.
The mini case study from Monze District confirmed that women's workloads increased with FFW participation, while men focused on asset gains. To address these inequities, gender-sensitive policies are needed. Flexible work arrangements would allow women to balance domestic and program duties, while childcare support at work sites could enhance women's participation. Ensuring women's control over small livestock strengthens empowerment and provides them with immediate liquidity to meet household needs.
Interpretation: Gender-sensitive policies must acknowledge women's workload and enhance their control over assets. Without such measures, programs risk reinforcing inequities even as they improve food security.
5.4.3 Preventing Elite Capture
Governance challenges, particularly elite capture, were evident in both Zambia and regional comparisons. Chiefs and local leaders were reported to prioritize relatives or allies for program participation, undermining fairness and legitimacy. Focus groups in Chipata District revealed that chiefs often selected family members first, leaving vulnerable households excluded.
A community member from Zimbabwe expressed similar frustrations: "Tee councillor's relatives are always first on tee list We feel left out, even teouge we are eungry too. " This testimony illustrates how elite capture erodes trust and creates perceptions of unfairness.
Policy strategies to reduce elite capture include:
• Transparent targeting using clear criteria and public lists.
• Community monitoring through committees that oversee participation.
• Donor oversight requiring independent audits of targeting processes.
Interpretation: Policies must strengthen transparency and accountability to prevent elite capture. Donor reports must acknowledge governance challenges rather than assuming fairness, and community voices must be integrated into monitoring processes to safeguard legitimacy.
5.4.4 Enhancing Resilience Pathways
Resilience pathways must be strengthened to ensure that households are equipped to manage both short-term shocks and long-term challenges. Quantitative findings showed that asset ownership, income diversification, and social capital are critical predictors of resilience. Qualitative evidence reinforced this, with households emphasizing the importance of livestock as insurance against drought.
A male participant from Monze District explained: "Whendroughtcomes, Isellonegoat to buy maize. That goat is like my insurance." This testimony highlights how asset ownership provides households with coping capacity during crises.
Regional lessons also provide valuable insights. In Mozambique, CFW programs offered youth construction skills, enhancing employability and creating resilience pathways beyond immediate food security. This demonstrates how social protection can be linked to long-term livelihoods, combining cash transfers with skill development.
Policy options to strengthen resilience include:
• Supporting livestock and tool acquisition to build coping capacity.
• Strengthening community networks and cooperatives to enhance informal support.
• Linking CFW to small business training to promote adaptability.
• Aligning FFW with climate adaptation projects to reduce vulnerability.
Interpretation: Policies must integrate asset accumulation, social capital, and skill development into program design. Regional lessons provide models for diversification and climate resilience, showing that resilience is multidimensional and requires both immediate protection and long-term transformation.
5.4.5 Integrating Regional Lessons
The comparative analysis with Malawi, Mozambique, and Zimbabwe highlights that Zambia can benefit from diverse approaches across Southern Africa. Malawi's integration of FFW into public works demonstrates how food security can be linked to infrastructure development, though Zambia must guard against misrepresentation by local leaders. Mozambique's emphasis on skill development shows the potential of CFW to create employable pathways for youth, while Zimbabwe's climate resilience framing illustrates how FFW can be aligned with adaptation strategies.
At the same time, recurring challenges such as dependency, elite capture, and gender inequities remain common across contexts. Addressing these governance and social concerns is essential to ensure that Zambia's programs are not only technically effective but also socially legitimate.
Synthesis
Overall, the policy implications highlight that FFW and CFW programs must evolve beyond short-term relief to become instruments of sustainable resilience. Addressing dependency, promoting gender equity, preventing elite capture, and enhancing resilience pathways are all critical to ensuring that programs remain effective, legitimate, and inclusive. By integrating productive assets, vocational training, gender-sensitive measures, transparent governance, and climate adaptation strategies, Zambia can strengthen its social protection framework and build pathways toward long-term food securityand resilience.
6.0 CHAPTER 6 CONCLUSION AND RECOMMENDATIONS
6.1.1 Summary of Findings
6.1.2 Protective Effects: Food Security, Dietary Diversity, and Absorptive Capacity
Protective effects are the first-order gains expected from public works programmes like food-for-work (FFW) and cash-for-work (CFW): stabilizing consumption, preventing negative coping strategies, and ensuring that acute deprivation does not unravel the foundations of household well-being. In the Zambian context where climate variability, recurrent droughts, and market fluctuations intersect with poverty and geographic inequalities, FFW interventions demonstrated clear improvements in dietary diversity and short-term resilience, or absorptive capacity. Absorptive capacity refers to the capability of households to absorb shocks without suffering substantial and irreversible welfare losses. It's the difference between a lean season that is hard and one that becomes a tipping point into asset depletion, indebtedness, or child malnutrition.
FFW achieves these protective effects through two reinforcing channels. First, the transfer channel: predictable food (or cash, where applicable) is provided to households in exchange for work, smoothing consumption across bad months and reducing the need to sell assets, pull children from school, or cut meal frequency. Predictability is crucial; it enhances planning, reduces stress, and allows households to avoid highinterest borrowing from informal lenders. Second, the asset channel: communities undertake works that not only justify the transfer economically and socially, but also improve the micro-environment in ways that indirectly support food availability and access. Typical assets include soil and water conservation structures, small-scale irrigation rehabilitation, feeder roads that lower transport costs and spoilage, and communal gardens that improve access to nutrient-dense foods.
In Zambia's agro-ecological and market conditions, dietary diversity is a particularly relevant measure of protective impact. The emphasis on staples like maize can crowd out essential food groups, leading to "hidden hunger" energy sufficiency but micronutrient deficiency. FFW's twin emphasis on immediate food access and asset creation that supports diversified production (e.g., vegetables, legumes, small livestock) creates a pathway to improve the quality, not just the quantity, of diets. Integrating nutrition social and behaviour change (SBC) into public works cycles such as cooking demonstrations, breastfeeding promotion, food hygiene, and the utilization of locally available nutrient-dense foods can amplify these gains by shaping household choices overtime.
Monitoring these outcomes with standardized tools like MDD-W (for women) and MDD-C (for children) is central to both programme learning and accountability. These indicators are usually collected through a simple dietary recall of the previous 24 hours, classifying foods into defined groups to assess the likelihood of adequate micronutrient intake. When embedded in routine monitoring, MDD data can flag whether transfers are adequate in size, timely in delivery, and appropriately matched to lean season dynamics. They also reveal whether investments in livelihoods assets are translating into diet quality improvements. For instance, if irrigation infrastructure is rehabilitation's focus but MDD scores do not improve, the programme may need complementary interventions such as seeds for nutrient-rich crops, nutrition counselling, or targeted market facilitation to ensure availability and affordability.
Absorptive capacity is not only about food; it is about avoiding negative coping. In contexts without safety nets, shocks trigger distress sales of livestock, reduced spending on health and education, and intra-household reallocation of food away from nutritionally vulnerable members. FFW reduces the probability of these outcomes by providing a reliable income orfood flow during lean periods, especially when well-timed to the agricultural calendar. Timing and predictability cannot be overstated: poorly timed transfers arrive after damage has already occurred, while irregular delivery can cause anxiety and "wait-and-see" behaviour that blunts productive labour allocation.
Yet, protective effects are contingent on programme design quality. If work requirements are too onerous or misaligned with caregiving responsibilities, women may face trade-offs between time used for care and time used for public works. These risks undermining the very nutritional outcomes sought. Similarly, if asset creation is not aligned with local ecological and market realities, the long-term benefits may be limited. Programme teams should conduct participatory planning that includes women, youth, and people with disabilities to ensure that assets and schedules are feasible, valued, and accessed equitably.
Operational refinements can significantly heighten protective impact. Localizing worksites reduces travel time and keeps labour burdens manageable, particularly for caregivers. Half-day shifts or rotating schedules tailor participation to time-sensitive household duties while maintaining access to transfers. Light-duty task options ensure that pregnant women and lactating mothers can participate safely. Where feasible, onsite childcare helps maintain attendance and lowers the risk of leaving young children unattended, a known risk factor for malnutrition and injuries.
Transfers must also be value-appropriate in relation to local prices. In-kind food transfers can shield households from price volatility and ensure direct access to staples, but they require robust logistics and are less flexible for diversified needs like soap, school items, or protein purchases. Cash transfers, by contrast, enhance flexibility and dignity, support local markets, and may be preferred in areas with functioning markets. A combination approach can sometimes be optimal: food during acute market stress and cash during normal periods, coupled with financial literacy to help households plan expenditures.
Protection also has a community dimension. Feeder road rehabilitation, for instance, does not only benefit participating households; it also reduces travel time to clinics and markets for the entire community, and lowers the cost and risk of transporting produce. Water-harvesting structures reduce erosion and improve soil moisture, contributing to the local commons that supports agriculture beyond programme participants. When asset selection is done transparently and based on community-wide priorities, protective effects extend to social cohesion and reduce tensions that can arise from perceptions of exclusion or unfairness.
Finally, durability of protective effects depends on institutionalizing measurement and response loops. This means using routine MDD-W and MDD-C data, integrating community feedback on worksite conditions and transfer adequacy, and adapting quickly when signs of stress emerge (e.g., escalating food prices, late rains). By pairing predictable transfers with nutrition-sensitive assets and responsive management, FFW transitions from a stopgap measure into a reliable protective layer that sustains dietary quality and prevents reversals in household welfare during shocks.
6.1.3 Pro motive Effects I: Adaptive Capacity Through Skills, Cash, and Labour Market Linkages
Promotive effects refer to the gains that go beyond immediate relief including those that improve a household's ability to adapt to changing conditions, diversify income, and move toward sustained livelihoods. Cash for Work (CFW) programmes in Zambia demonstrate strong promotive potential when designed as platforms that combine temporary employment with skills development, financial inclusion, and linkages to market opportunities. The logic is straightforward: cash today stabilizes consumption and unlocks liquidity; skills training and employability services expand future earning capacity; and market linkages ensure that new capabilities and assets translate into actual income.
A central insight is that skills training must be demand-driven. Short, modular, competency-based training aligned to local economic opportunities tends to yield higher returns than long, generic courses. For instance, in districts where construction activity is growing, modules in basic masonry, carpentry, welding, or solar installation may have direct applicability. In agricultural zones, value addition skills such as oil pressing, peanut butter processing, drying and packaging of vegetables, or small-scale milling can translate into micro-enterprises that transform seasonal production into year-round income streams. Urban and peri-urban settings might benefit from modules in basic bookkeeping, mobile money operation, motorcycle repair, hairdressing, or hospitality services. The key is labour market intelligence: mapping demand through consultations with local businesses, cooperatives, councils, and trade associations, and updating these profiles regularly.
The CFW + skills model works best when underpinned by a cohesive set of services:
• Certification and assessment so that skills are recognized beyond the training venue, aidingjob placement and mobility.
• Placement and internship support to give graduates real-world experience and entry points into firms.
• Entrepreneurship coaching to help those pursuing self-employments with pricing, record-keeping, micro-marketing, and basic compliance.
• Financial inclusion through savings groups, mobile wallets, bank accounts, and where feasible linkages to microcredit or asset financing.
Access to cash through CFW provides the liquidity bridge that enables trainees to participate fully (covering transport, childcare, and opportunity costs) and to start small ventures. When paired with matched savings or "starter kits" (e.g., basic toolkits, seed capital vouchers), the probability of sustained enterprise activity increases. The presence of peer groups or cohorts fosters accountability, knowledge exchange, and emotional support which are critical elements for first-time entrepreneurs.
Another promotive pathway is bundling life skills with technical modules. Communication, teamwork, problem-solving, time management, and customer service are all low-cost additions that meaningfully increase employability. In many cases, market failures are not only about a lack of technical skill but about information gaps, soft-skill deficits, and thin networks. Addressing these systematically can make the difference between trained-but-unemployed graduates and those who secure stable incomes.
CFW can also be designed to create or improve local market infrastructure that expands opportunities, for example, rehabilitating market stalls, constructing aggregation centres, or improving last-mile roads that lower costs for traders and attract buyers. These investments create a demand "pull" for products and services, making it easier for newly trained individuals to sell goods and for micro-enterprises to flourish. Coupling these works with market days , business clinics, and buyer-seller forums can catalyse connections and help participants learn about quality standards, packaging, and pricing strategies.
Gender-responsive design is not only a protective measure but also a promotive accelerator. Women's access to cash, combined with training delivered at times and locations compatible with caregiving, increases completion rates and post-training utilization. Childcare at training venues , stipends or transport vouchers, and women- priority cohorts can equalize participation. Moreover, intentional pathways into non- traditional trades with mentorship and safety measures can open higher-earning segments for women. Pairing financial literacy with intra-household dialogue sessions can reduce appropriation risks and reinforce women's control over income and business decisions.
Promotive effects also hinge on sequencing. For instance, launching CFW during a lean season provides immediate cash, while scheduling training as workloads decline (postharvest) raises attendance and retention. Alternatively, a "train-and-earn" design halfday works and half-day training can work if travel time is minimized and childcare is provided. The core principle is to avoid competition for time that would undermine either training quality or the protective value of cash transfers.
At the system level, collaboration with TVET authorities, private training providers, and employers is crucial. Public works agencies are not typically vocational training specialists. Partnering allows alignment with national curricula, trainer upskilling, and standardized assessments. To deter capture and ensure fairness, transparent eligibility for scarce training slots , lotteries when oversubscribed , and public reporting of completion and placement rates (disaggregated by gender, age, and ward) are recommended. This protects the legitimacy of the programme and helps sustain community support.
A frequent constraint is post-training capital. Even with skills, many cannot afford the initial tools or working capital. CFW can partially solve this by enabling savings during the works phase. Layering in group-based savings and rotating credit associations , digital saving nudges , or results-based starter grants can unlock enterprise formation. Partnerships with microfinance institutions that offer graduated loan products like small initial loans with quickfollow-up tranches based on repayment can sustain growth while managing risk.
Monitoring promotive outcomes requires going beyond output counts (e.g., number trained). Useful indicators include:
• Placement rate within 3-6 months after training.
• Enterprise survival rate and average monthly revenues at6and12 months.
• Income diversification index at the household level.
• Women's control over income and decision-making in businesses.
• Reduction in lean-season borrowing at high interest.
Qualitative follow-ups such as focus groups and case studies can illuminate success factors and barriers notvisible in quantitative data.
In sum, CFW's promotive effects are strongest when cash is not an endpoint but a platform for skills, market linkages, and financial inclusion. With the right design, households' transition from precarious, climate-sensitive livelihoods to more diversified portfolios, improved employability, and a path toward sustained resilience.
6.1.3 Promotive Effects II: Resilience Pathways via Assets Especially Women-Controlled Small Livestock
Asset accumulation is a cornerstone of resilience. Among the assets that emerged most prominently in this study were small livestock including chickens and goats in particular often under women's orjoint control. These assets function as a flexible form of "buffer stock," allowing households to manage liquidity needs in small increments. Unlike large livestock or land, which are lumpy and often culturally or economically costly to liquidate, small livestock can be sold sequentially, providing cash-for-food, medicines, school fees, or inputs during lean periods. This granularity reduces the risk of catastrophic divestment and supports more stable consumption patterns over time.
From a resilience perspective, the value of small livestock extends beyond liquidity. Nutritional co-benefits arise from eggs, milk (in the case of goats), and, occasionally, meat. Income generation is possible through sales of live animals, eggs, kids, and chicks. Fertilizer substitution can occur when manure is used to enrich soils in kitchen gardens or small plots, supporting diversified and micronutrient-rich production. For women, control over small livestock is also a lever for agency . Control over an income source even if modest can shift intra-household dynamics, increase women's bargaining power, and create a foundation for further investments, such as improved breeds, feed, or small equipment.
However, converting livestock transfers into durable resilience gains requires attention to complementary systems . The most common threats to livestock productivity and survival are preventable or manageable with basic services: vaccination campaigns, deworming , extension advice , and access to feed and water during dry spells. Embedding these services into programme design either through partnerships with veterinary departments, community-based animal health workers, or private providers dramatically increases survival rates. Similarly, training in basic husbandry (housing, hygiene, predator protection, breeding management) and feed formulation using locally available materials can raise productivity at low cost. Where feasible, simple innovations like night shelters or improved pens lower disease incidence and mortality, securing the asset base.
The gender dimension is paramount. To ensure that asset transfers are not appropriated by other household members undermining women's empowerment programmes can combine transfers with household sensitization sessions on joint planning and benefit sharing. Registration of animals under women's names , ear-tagging with participant IDs, and community-level norms endorsing women's control over small livestock reduce appropriation risks. Furthermore, facilitating women's groups or cooperatives around small livestock can unlock economies of scale such as bulk purchase of vaccines or feed, shared access to breeding males, and collective marketing to obtain better prices.
Linkages to microfinance or asset financing can help households upgrade from low- productivity breeds to improved ones, or to step up from chickens to goats, and eventually to small ruminants with higher market value. Savings groups anchored in FFW/CFW cohorts provide a critical bridge: disciplined saving during the works period can be earmarked for livestock investment, with group bylaws supporting re-investment cycles. Programmes might consider matched savings for livestock or results-based grants contingent on basic husbandry practices (e.g., building a shelter, vaccination compliance, record-keeping).
Market access is another determinant of asset-based resilience. Households need reliable buyers , fair prices, and timely information. Public works that improve feeder roads and market stalls directly reduce transaction costs and spoilage, while information sessions on local market calendars, price trends, and buyer preferences help participants plan sales strategically. Collective marketing through women's groups can counteract trader monopsony and improve bargaining power. The introduction of lightweight digital tools for recording births, deaths, sales, and costs can strengthen decisionmaking and support access to credit by demonstrating enterprise performance.
Climate risk threatens livestock systems through heat stress, forage scarcity, and disease dynamics. Programmes can promote climate-resilient husbandry practices: water harvesting for livestock, cultivation of drought-tolerant fodder species, feed preservation techniques (e.g., haymaking), and shade structures. Where the enabling environment allows, index insurance pilots can be explored to protect against covariate shocks, although these instruments require careful design to ensure comprehension, trust, and value for money. Even absent formal insurance, risk pooling within groups such as "livestock replacement funds" financed by small weekly contributions can cushion idiosyncratic losses.
To maximize promotive effects, asset transfers should be sequenced and tailored. Households with limited labouror high care burdens may do betterwith poultry initially, given lower labour and space requirements, while those with more land and water access might step into goats. Graduation pathways can be made explicit: start with poultry, then expand to dual-purpose breeds, then add goats as liquidity and husbandry skills grow. The programme can also bundle assets with kitchen garden kits, linking animal manure to vegetable production to increase nutrient availability at home and generate surplus for sale.
Monitoring should track not only stock levels and sales, but also control and decisionmaking: who decides when to sell, how income is used, and whether women's agency is increasing over time. Incorporating simple household scorecards covering animal health practices, shelter quality, record-keeping, and marketing behaviours helps tailor coaching and recognize progress. Additionally, measuring dietary impacts (e.g., frequency of egg or milk consumption in target households) reveals whether nutritional co-benefits materialize or whether income is primarily used for other essentials.
A critical, sometimes overlooked element is exit and sustainability. When projects end, participants should be connected to enduring support networks: local veterinary services, producer associations, savings groups, and trusted buyers. Building community animal health worker capacity leaves behind a service layer that persists. Establishing memoranda of understanding with local agro-dealers for discounted vaccines or feed during lean months can further institutionalize support. Equally important is norms change: recognizing women publicly as livestock owners and entrepreneurs, celebrating milestones (first sales, flock doubling), and aligning with community leaders to reinforce women's rights over assets.
In essence, small livestock under women's control are more than assets; they are platforms for resilience by providing liquidity on demand, contributing to diet quality, enabling microenterprise growth, and strengthening women's agency. When supported by animal health services, husbandry training, market linkages, and household-level empowerment strategies, these assets become catalytic, moving households from fragile stability to adaptive capacity and sustained improvements in well-being.
6.1.4 Measurement Matters: Embedding Nutrition and Empowerment indicators
Effective public works programming depends as much on what is measured and how it is used as on what is implemented. When FFW/CFW claim to build resilience, improve diets, and empower women, those claims must be supported by credible, standardized, and actionable data . Embedding nutrition indicators such as MDD-W (Minimum Dietary Diversity for Women) and MDD-C (Minimum Dietary Diversity for Children), alongside empowerment indicators like the WEAI/pro-WEAl, transforms monitoring systems from activity trackers into decision-support platforms. This shift is essential for adaptive management, accountability, and evidence-based scaling.
Why MDD-W and MDD-C? Dietary diversity is a concise, validated proxy for micronutrient adequacy and complementary feeding quality. For women, MDD-W measures whether they consumed foods from a minimum number of diverse groups in the previous 24 hours, capturing the likelihood of micronutrient sufficiency. For children, MDD-C adapts the logic to their specific nutritional needs, including breastfeeding where applicable. These measures are simple to collect, highly interpretable, and comparably robust across contexts when protocols are followed. For public works, they connect transfer adequacy , timing , and asset relevance to real outcomes in nutrition.
Designing MDD data collection for programmes. A sound MDD system includes:
1. Sampling strategy: Stratified sampling by district, livelihood zone, and programme modality (FFW vs CFW) to detect heterogeneity. Oversampling of vulnerable groups (pregnant/lactating women, female-headed households) ensures programme equity is visible.
2. Enumerator training and protocols: Clear definitions of food groups, standardized 24-hour recall methods, probing techniques to avoid underreporting, and culturally adapted examples of local foods ensure quality.
3. Seasonality controls: Repeated measures across agricultural seasons (harvest, lean, planting) disentangle programme effects from season effects.
4. Data integrity checks: Built-in skip logic, real-time range checks, and random backchecks improve accuracy.
5. Rapid feedback loops: Dashboards that update monthly or quarterly enable the team to adjust transfer timing, SBC content, or asset choices.
Turning MDD into management decisions. Data must flow into deliberate decision processes. If MDD-W stagnates despite transfers, the programme might:
• Increase cash values during price spikes or pivot to food transfers temporarily.
• Integrate nutrition SBC sessions focusing on low-cost, locally available nutrient- dense foods.
• Ensure asset choices (e.g., irrigation, communal gardens) are nutrition-sensitive, linked to vegetables, legumes, and small livestock.
• Address time constraints if women's workload inhibits meal preparation or feeding routines, via childcare or flexible work hours.
Embedding empowerment: WEAI and pro-WEAI. Diet quality alone cannot fully reveal whether programmes are transformative. WEAI (and its project-level adaptation, pro- WEAI) assess women's agency across domains such as production decisions, access to and control over resources, income and expenditures, leadership, and time allocation. For FFW/CFW, WEAI illuminates whether women are genuinely gaining control over assets , participation in decision-making, and voice in community structures. It also highlights time poverty, a critical constraint often hidden in conventional monitoring.
Operationalizing WEAI/pro-WEAl
1. Module selection: Choose submodules relevant to FFW/CFW (production, asset control, income use, leadership, time use).
2. Enumerator sensitivity: Train teams on gender dynamics, privacy, and interviewing both women and men separately to reduce bias and social desirability effects.
3. Scoring and interpretation: Use domain-level scores to identify leverage points: for instance, high production autonomy but low-income control suggests intrahousehold appropriation risks that require financial routing to women and household communication sessions.
4. Programme response: If time-use scores reveal severe constraints, introduce halfday shifts, lighter tasks, or near-home worksites; if leadership scores are low, intentionally seat women on works committees and rotate roles.
Adding time use measurement. A structured 24-hour recall or short time diary provides granular insights into care duties, domestic work, and market participation. When repeated, it shows whether the programme reduces, maintains, or increases women's time burden. Time data should be disaggregated by pregnancy/lactation status, number of dependents, disability status, and commuting distance to account for intersectional constraints.
Fairness and governance metrics. Measurement of empowerment must be complemented by fairness metrics that track inclusion/exclusion error rates, grievance volumes and resolution times, and concordance between administrative records and independent verification. These metrics reveal whether programme delivery is equitable and whether integrity systems are functioning. For instance, a mismatch between muster rolls and independent attendance checks may indicate leakage or favouritism.
Mixed-methods triangulation. Numbers alone rarely tell the full story. Systematic focus group discussions (FGDs) separately with women, men, youth, and persons with disabilities surface dynamics like appropriation of women's assets, unsafe worksites, or coercive gatekeeping. Key informant interviews with local leaders, extension workers, and market actors enrich interpretation. Case studies of households that show exceptional progress or difficulty uncover practical barriers and replicable strategies.
Digital, ethical, and data governance considerations. Where feasible, digital tools (mobile data collection, dashboards) increase timeliness. Yet data systems must respect privacy and consent, especially for sensitive topics like GBV or intra-household conflict. Store data securely, minimize personally identifiable information, and ensure respondents understand the purpose and benefits of data collection. A data governance protocol should specify who has access, how long data is retained, and how it will be used for improving services.
Cost and capacity planning. Embedding robust measurement has costs, but these can be managed by integrating monitoring into programme cycles, training local enumerators, and prioritizing indicators with high decision value. A core indicator set which include MDD-W, MDD-C, WEAI/pro-WEAl domains, time use, fairness metrics can be complemented by rotating modules as capacity allows. Building capacity among district teams for analysis and use of data is as crucial as collection itself.
From measurement to learning. Ultimately, measurement matters when it drives course correction. Quarterly review meetings should link indicator trends to action plans. If MDD-C declines in one ward, investigate market availability, transfer timing, caregiver time constraints, and SBC coverage. If WEAI shows low income control, strengthen routing of payments to women and reinforce household dialogue. If fairness metrics show rising grievances without resolution, reinforce GRM staffing and accountability. By institutionalizing measure interpret act cycles, FFW/CFW programmes build a culture of learning that steadily raises impact and legitimacy.
6.1.5 Gendered Workload and the Gare Economy: The Double Burden Risk
Public works can improve livelihoods while simultaneously deepening time poverty for women if care-aware design is not central. Women typically carry the bulk of unpaid domestic labour such as childcare, elder care, water and fuel collection, cleaning, and meal preparation. Layering heavy, inflexible work requirements onto this baseline can push households into difficult trade-offs: missed feeding windows for children, reduced rest for pregnant women, or constrained participation in training that could lift longterm incomes. Addressing the double burden risk requires intentional design choices that respect women's time, protect their health and dignity, and expand their agency.
Principles of care-aware public works
1. Proximity: Locate worksites near participants' homes to reduce commute time and increase predictability.
2. Flexibility: Offer half-day shifts, staggered schedules, and seasonal adjustments to align workwith agricultural calendars and caregiving duties.
3. Task differentiation: Provide lighter tasks for pregnant and lactating women, caregivers, and persons with disabilities;rotate heavy tasks to avoid concentrated strain.
4. Childcare: Establish safe, supervised spaces at or near worksites;schedule feeding breaks;enable breastfeeding.
5. Safe transport and WASH: Ensure reliable transport options, lighting for early/late shifts, gender-appropriate toilets, washing areas, and hygiene supplies to uphold dignityand reduce health risks.
6. GBV safeguards: Institute clear codes of conduct, confidential reporting channels, trained focal points, and zero-tolerance policies for harassment and exploitation.
These features are not ancillary; they are preconditions for equitable access and sustainable outcomes. When implemented, women's participation rises, attendance stabilizes, and household nutrition improves because caregiving does not compete with work in destructive ways.
Financial autonomy and payment design. Routing payments directly to women's accounts or mobile wallets strengthens bargaining power and protects earnings from appropriation. Publicly posted rate cards clarifying wage levels, payment schedules, and task categories make discrimination harder and empower women to assert their entitlements. Where digital financial inclusion is nascent, the programme can support KYC (Know Your Customer) onboarding, group-based saving mechanisms, and basic financial literacy. Simple budgeting tools and savings nudges (e.g., goal-setting for school fees or small livestock purchases) turn payments into stepping stones toward asset accumulation and business starts.
Intersections with social norms. Care-aware design must be complemented by norms engagement . Household dialogues, couple sessions, and community meetings can normalize women's paid work, affirm their control over earnings, and reduce backlash. In some contexts, engaging male champions and respected community leaders makes a tangible difference in acceptance. Programmes should be prepared for pushback dynamics for example, men insisting on supervising women's earnings or work choices and have strategies to reinforce women's rights, including supportive messaging and grievance channels for coercion or harassment.
Inclusion for caregivers and vulnerable groups. Care burdens vary widely. Households with children under two, persons with disabilities, or frail elders face distinct constraints. The programme should identify these households and offer tailored participation modalities near-home tasks, additional flexibility, or home-based assignments (e.g., nursery bed preparation, seedling care, kitchen garden maintenance linked to public works objectives). Adolescents and young mothers may benefit from integrated life skills and parenting support, while older caregivers might require lighter tasks and transport assistance.
Measuring time and adjusting design. A time-use module is critical to quantify the double burden. A simple 24-hour recall, disaggregated by activity category (paid work, unpaid domestic work, care work, personal care, rest), reveals whether FFW/CFW shifts time patterns positively or negatively. Seasonal repetition captures variations around planting, harvest, and lean periods. These data should feed into operational decisions: if women's sleep hours drop below healthy thresholds, increase rest breaks or rotate tasks; if childcare time collapses during works cycles, intensify childcare provision or adjust shifts.
Integrating nutrition and care considerations. Nutrition outcomes depend partly on whether caregivers can procure, prepare, and feed nutrient-dense foods consistently. Programme schedules should build in meal preparation windows , feeding breaks, and SBC activities at convenient times. Kitchen gardens, small livestock packages, and clean water points attached to works can reduce the time and effort needed to provide diversified diets. In settings with long water collection times, public works that rehabilitate boreholes or extend piped networks have care economy dividends, lowering daily burdens and improving hygiene.
GBV risk management and safeguarding. Public works sites can become spaces of harassment or coercion if safeguards are weak. Clear codes of conduct, visible signage, trained supervisors, and confidential reporting mechanisms (including anonymous options) are crucial. Provide multiple channels: complaint boxes, toll-free lines, SMS/WhatsApp, and in-person focal points especially female focal points. Establish response timelines and communicate them publicly. Train staff in survivor-centred approaches and ensure referral pathways to health, legal, and psychosocial services. Regular audits of site safety (lighting, transport, WASH, supervision presence) are part of safeguarding routines.
Task design for dignity and sustainability. Heavy manual tasks can bring short-term productivity but long-term harm if overused. Programmes should balance labour intensity with health and safety: mechanical aids where possible, rotation of strenuous tasks, PPE (gloves, boots), and hydration stations. Assignments should respect pregnancy and postpartum health recommendations. Where feasible, prioritize nearhome, lighter tasks that still contribute to resilience assets, such as seedbed management, community nurseries, inspection and maintenance of small infrastructure, or record-keeping roles in site committees.
Linking care-aware design to empowerment. Reducing time poverty is not merely protective; it supports agency. When women can participate without sacrificing caregiving quality, they gain consistent access to transfers, skills, and leadership roles in committees. Intentional rotation of committee representation, quotas with genuine authority (not symbolic seats), and leadership coaching increase women's collective voice. Payment control coupled with enterprise pathways (e.g., starter kits, market days, microcredit linkages) converts protected participation into long-term economic gains.
Continuous learning and accountability. Care-aware design should be reviewed regularly with participants especially women through feedback circles and scorecards . If transport is unsafe on certain routes, adapt shift times or provide escorts. If toilets lack privacy, improve design and placement. If childcare quality is poor, train caregivers, procure basic materials, and set standards. Publish service commitments such as what participants can expect, how to complain, and response targets to strengthen trust. In this way, the programme reduces the double burden, increases equity, and translates participation into sustained improvements in household well-being.
6.1.6 Governance Risks: Elite Capture,Poiiticization, and Grievance Redress
Governance quality determines whether FFW/CFW programmes deliver fairly and retain public legitimacy. The risks elite capture, politicization, leakage, and weak grievance resolution can quietly erode impact even when technical design is strong. Tackling these requires a transparent architecture of rules, independent verification, and community-facing accountability that is both protective and practical.
Where capture happens. Elite capture can occur at multiple points: targeting (who gets on lists), enrolment (who is called first), task allocation (who gets easier or higher-paying assignments), attendance verification (ghost workers), and payments (skimming or delayed disbursements). Politicization surfaces when selection ortaskassignment is tied to party loyalty, or when elected officials exert undue influence. These dynamics are often under-detected in household surveys due to social desirability bias or fear of retaliation, but they frequently appear in focus groups and anonymous complaints.
Transparency by default. The first line of defence is radical transparency. Publicly post eligibility criteria, beneficiary and waitlists, muster rolls, and rate cards in local languages at accessible locations such as schools, clinics, markets, and worksites. Update them routinely and date-stamp changes. Standardize list formats with unique identifiers to reduce manipulation. Conduct public validation sessions where lists are read out, changes are explained, and community members can raise objections. Use visual dashboards or notice boards to show grievance volumes and resolution times, enhancing trust and deterring misconduct.
Rule-based selection and randomization. To blunt patronage and favouritism, implement randomized selection among verified eligible households when demand exceeds capacity. Verification should be arms-length conducted by staff not embedded in local political structures or supplemented by external monitors during peak enrolment periods. Randomization does not replace vulnerability targeting; it complements it when oversubscription persists after objective screening. Document the process and publish the outcomes to deter allegations of bias.
Representative social audits. Establish social audit committees with balanced representation: women, youth, persons with disabilities, and respected community figures. Their mandate is to review attendance and payment records, visit worksites, and validate or contest list changes. Social audits should be scheduled, minuted, and publicly reported. Provide committees with simple audit tools: checklists for record accuracy, spot checks of work quality, and protocols for interviewing participants confidentially. Rotate committee membership to prevent capture and fatigue, and ensure they can escalate findings to district oversight bodies.
Grievance redress mechanisms (GRMs). An effective GRM has multiple channels with anonymity options including complaint boxes, toll-free lines, SMS/WhatsApp, and inperson focal points. It sets clear timelines for acknowledgement, investigation, and resolution, and publishes aggregate reports with categories of complaints and resolution rates. To prevent retaliation, allow anonymous submissions and protect whistle-blowers through confidentiality and non-retaliation policies. Train GRM staff on respectful handling, impartial investigation, and referral pathways when complaints involve serious misconduct or GBV. Where national systems exist (such as those used in social cash transfers), align the FFW/CFW GRM to leverage existing infrastructure and legitimacy.
Independent assurance and verification. Beyond transparency, programmes need third-party audits and methodological innovations to detect underreported wrongdoing. Randomized back-checks of attendance, digital photo or biometric verification at muster points (where feasible and ethical), and risk-based sampling of high-risk sites (e.g., unusually high attendance or payment anomalies) strengthen assurance. Techniques that encourage honest reporting such as indirect questioning formats during confidential interviews can complement routine verification. Critically, independent findings must lead to corrective actions: re-selection, repayment, sanctions, or referral to authorities where appropriate.
Political economy and role separation. Customary leadership and elected officials often play vital roles in local order. Full exclusion can backfire;full control can politicize. The balanced approach is co-produced accountability: involve traditional authorities in bounded validation and dispute resolution while keeping final selection rule-based and overseen by neutral committees. Explicit role separation excluding elected officials from selection bodies while inviting them to public validation sessions maintains legitimacy without ceding control. Communicate these role boundaries clearly, and publicly document compliance.
Sanctions and incentives. Governance systems must include graduated sanctions for misconduct such as warnings, suspension, removal from roles, and, in severe cases, legal action. Pair sanctions with positive incentives for good governance: recognition of high-integrity committees, public praise for sites with clean audits, and small performance grants for improvements in transparency. This balance supports a culture of integrity rather than a purely punitive environment.
Data and fairness metrics. Track inclusion and exclusion errors, the share of women and vulnerable groups among beneficiaries, grievance volumes by type, average resolution times, and concordance between administrative and independent datasets. Publish fairness scorecards at community and district levels to enable comparison and peer learning. Anomalies such as persistent gender imbalances or unresolved grievances triggertargeted reviews and additional oversight.
Communication and trust-building. Information asymmetry feeds capture. Use community radio , village meetings , and flyers to communicate programme rules, GRM channels, payment schedules, and rights of participants. Provide orientation sessions for new cohorts that include role plays on how to raise grievances safely. Encourage peer monitors trained participants who informally check muster rolls and payment processes to serve as early warning sensors.
Operational resilience. Governance systems must withstand shocks: staff turnover, election cycles, and local conflicts. Build redundancy such as multiple trained focal points, backup auditors, and cross-district support. Keep documentation standardized and digitally backed up. Train new staff regularly on integrity protocols and the ethical handling of personal information. Maintain a risk register at district level, categorizing sites by risk and scheduling intensified oversight where needed.
Adaptive learning. Governance is not static. Regularly review which measures deter capture most effectively in your context. If posting lists is insufficient , add community verification fairs ; if GRM volumes are low despite known issues, experiment with anonymous mobile channels and community sensitization. If elite capture persists in training slot allocation, reintroduce lotteries and publish slot assignments with audit trails. Learning loops which comprises of measure, convene, decide, act should be institutionalized, with findings shared back to communities to reinforce accountability.
In sum, when transparency is hard-wired, selection is rule-based, redress is accessible and timely, and verification is independent and credible, FFW/CFW programmes become trustworthy. They resist political and elite pressures, protect participant rights, and uphold fairness conditions under which public works can deliver not only assets and transfers, but also durable legitimacy and social cohesion.
6.1.7 Comparative Insights: Malawi, Mozambique, Zimbabwe
Comparative insights from Malawi, Mozambique, and Zimbabwe provide a practical compass for designing and refining public works in Zambia. Although historical trajectories, administrative capacities, and market structures vary across these countries, the underlying lessons converge around three pillars: targeting architecture that blends objective rules with local information, system-level integration of skills and markets within public works, and governance safeguards capable of withstanding politicization and elite pressures. The aim is to transform short-term relief into durable resilience while preserving fairness and legitimacy.
Malawi: Hybrid targeting and seasonal accountability. Malawi's experience demonstrates that neither community discretion nor purely technocratic targeting alone is sufficient. Community ranking can identify context-specific vulnerability such as temporary illness, disability, caregiving burdens, or recent shocks that standardized proxies often miss. Yet, community systems are vulnerable to favouritism and local power dynamics. The comparative lesson is the value of hybrid targeting: begin with objective eligibility rules (e.g., poverty proxies, labour capacity, caregiving status), then incorporate community inputs during validation fairs that publicly marshal local knowledge. The process should be transparent, include an appeals window, and be repeated seasonally to capture changing circumstances and mitigate moral hazard. Decentralization to traditional authorities carries a genuine trade-off: it risks nepotism while also leveraging proximity to information and the ability to mobilize collective effort. The practical middle ground is bounded authority to invite chiefs and village committees to validate lists and mediate disputes while maintaining rule-based final selection and randomized tie-breakers. For Zambia, the implication is to combine objective eligibility and community validation, with simple, public procedures that make manipulations both difficult and visible.
Mozambique: CFW as a platform for demand-driven skills and market linkages. Mozambique's trajectory highlights the power of aligning CFW with modular, competency-based training and explicit pathways to employment or enterprise. Short courses that match local labour demand such as masonry, welding, carpentry, solar installation, hospitality, agro-processing, bookkeeping generate higher placement rates than generic training. This works best when the programme embeds ecosystem services: pre-course labour market scans; standardized certification and assessment; internship and placement support; entrepreneurship coaching; and financial inclusion to convert cash into savings and starter capital. Integrity risks arise where training slots are scarce;equity improves when eligibility is clear, cohorts rotate, and lotteries are used for oversubscribed modules. Mozambique's experience also underscores the importance of public works that reduce market frictions, such as rehabilitated feeder roads, aggregation centres, market stalls, and storage facilities. These assets create a "pull" for products and services, improving the survival of microenterprises and strengthening income diversification. For Zambia, the takeaway is operational: treat CFW as a springboard pair cash with market-relevant skills, verify outcomes through certification and reporting, and build local market infrastructure so trained participants can translate capabilities into earnings.
Zimbabwe: Climate-resilient public works under politicization pressure. Zimbabwe's lessons crystallize around climate adaptation and governance hardening. Public works that repair and climate-proof irrigation, invest in water harvesting and erosion control, and integrate climate information services tend to create compound resilience including higher productivity, better water security, and stronger buffers against droughts. Off-farm maintenance and technical skills ensure infrastructure remains functional beyond construction. The political economy lesson is sobering: politicisation risk can distort targeting and task allocation, erode trust, and suppress complaints. Effective countermeasures include integrity pacts, citizen charters, public muster rolls, independent audits, and, where feasible and acceptable, photo or biometric verification. Role separation by keeping elected officials out of selection committees while inviting them to public validation events maintains legitimacy without ceding control. The macro insight is urgency: repeated climate shocks require anticipatory investments and shock-responsive programming. For Zambia, the comparative advice is twofold: mainstream climate resilience within public works asset choices and fortify governance wherever contestation is likely.
Cross-cutting comparative threads. Three threads weave through all cases:
1. Hybrid, transparent targeting outperforms single-mode approaches. Combining objective criteria with community validation and randomization among verified eligible balances fairness, local insight, and resistance to capture. Regular seasonal re-validation sustains responsiveness and credibility.
2. Skills and assets must be embedded in market systems. Demand-driven training, certification, placement support, and entrepreneurship coaching create viable pathways. Market infrastructure encompassing roads, stalls, aggregation points, reduces transaction costs, attracts buyers, and transforms capabilities into income. CFW should be the liquidity and learning bridge into these systems.
3. Integrity systems are central, not peripheral. Social audits, grievance channels with anonymity, published resolution timelines, independent verification, and clear role boundaries counter politicization and elite pressures. These measures protect participant rights and preserve programme legitimacy.
Afourth thread is gender intentionality. Across contexts, women's participation is tightly linked to care-aware design, payment routing to women, flexible shifts, lighter tasks, and childcare options. When these are absent, time poverty intensifies, undermining both nutrition and empowerment outcomes. Embedding women's empowerment metrics, time-use diaries, and fair access to training and leadership roles ensures gender equity becomes measurable and actionable. For Zambia, the composite comparative insight is a pragmatic agenda: blend objective rules with community knowledge, make CFW a platform for skills and markets, hard-wire integrity against politicization, and measure what matters such as diet quality, agency and fairness to sustain trust and impact.
6.1.8 Restating the Programme Theory of Change
The programme theory of change integrates protective and promotive functions into a deliberate, staged pathway from acute vulnerability to adaptive and, ultimately, transformative resilience. It clarifies the causal logic, operational enablers, and feedback loops that ensure FFW/CFW deliver equitable, durable benefits. This dual-track theory relies on predictabletransfers, nutrition-sensitiveand climate-resilientassets, demand- driven skills and market linkages, gender-responsive design, governance safeguards, and institutionalized measurementforlearning.
Protective track (near term): stabilize, prevent, buffer. The protective track begins with predictable transfers linked to public works. Timing is strategic: disbursements align with periods of heightened food insecurity or income shortfall. Adequacy is calibrated to local price levels and household size, with the flexibility to adjust during inflationary spikes or supply disruptions. Protective design is nutrition-sensitive: works are paired with cooking demonstrations, breastfeeding support, food hygiene messaging, and assets that broaden access to nutrient-dense foods (e.g., kitchen gardens, water points). Women's participation is made feasible through care-aware scheduling, near-home worksites, lighter tasks for pregnant/lactating women, safe transport, and genderappropriate WASH. In this track, outcomes are monitored through MDD-W and MDDC, providing timely signals that trigger course corrections in transfer size, modality (food or cash), SBC intensity, and asset selection.
Promotive track (medium term): diversify, accumulate, connect. The promotive track transforms temporary work into stepping stones to income diversification, asset accumulation, and employability. Public works generate resilience assets such as soil and water conservation structures, climate-proofed irrigation, and feeder roads that reduce production risk and market frictions. CFW becomes the liquidity bridge enabling participation in short, modular, demand-driven skills training, certification, and placement. Financial inclusion which include group savings, mobile wallets, bank accounts turns cash into investable capital. Starter kits and matched savings help launch microenterprises. Market linkages and infrastructure which include aggregation centres, upgraded stalls, cold storage where feasible ensure newly acquired capabilities and assets translate into consistent earnings. Over time, households expand livelihood portfolios, while women's control over small assets (livestock, garden tools) strengthens bargaining powerand agency.
Three enabling pillars.
1. Gender-responsive design. Care-aware scheduling, task differentiation, childcare provision, safe transport, and gender-appropriate WASH are essential to equalize access. Payment routing to women's accounts and public rate cards reinforce financial autonomy and fairness. Household sensitization and women's group formation prevent appropriation and build collective voice. Women's empowerment is measured through WEAI/pro-WEAl domains and time-use diaries , ensuring that participation translates into agency, notjust additional labour burden.
2. Governance and accountability. A transparent architecture underpins delivery: public posting of eligibility criteria, beneficiary/waitlists, muster rolls, and wage rates; randomized selection among verified eligibles; representative social audits that review attendance and payments; multi-channel grievance redress with anonymity, published timelines, and survivor-centred pathways for misconduct;and independent verification through third-party audits and targeted spot checks. Role separation in politically sensitive contexts shields selection and task assignment from partisan influence while maintaining constructive engagement with local leaders.
3. Measurement and learning. Institutionalized measurement converts monitoring into management. A core set of indicators comprising of MDD-W, MDD-C, WEAI/pro-WEAl domains, time-use, and fairness metrics (inclusion/exclusion errors, grievance volumes and resolution times, admin-independent concordance) feed into quarterly review cycles. Mixed methods made of focus groups, key informant interviews and case studies explain anomalies and uncover hidden risks such as elite capture or GBV. Digital data collection enhances timeliness; strong data governance preserves trust and ethics. The measure-interpret-act loop is formalized so that signals reliably trigger operational adjustments.
Causal pathways.
• Transfers stabilize consumption ^ reduce negative coping (asset sales, highinterest borrowing) ^ improve diet quality (MDD).
• Resilience assets reduce climate and market risks ^ raise productivity and access ^ enable diversified diets and incomes.
• Skills and financial inclusion increase employability and enterprise viability ^ strengthen income diversification and liquidity buffers.
• Gender-responsive design increases women's participation and control ^ enhances agency and intra-household bargaining ^ improves household wellbeing.
• Governance and measurement protect fairness and legitimacy ^ sustain support and enable scaling.
Contextual assumptions and contingencies. The theory assumes at least minimal market functionality for cash transfers to convert into food and essentials;where markets are thin, in-kind transfers or vouchers may be preferable. It assumes Page 215 of 278 implementation capacity for routine measurement and safeguarding; where capacity is low, simplify the indicator set and invest in staff training. It assumes social acceptance of women's participation and income control;where norms resist, intensify community dialogues, engage male champions, and reinforce grievance protection. It assumes climate variability that is manageable within resilience asset design; where extremes intensify, integrate early warning systems, flexible scaling, and climate-proofed infrastructure.
End-state vision. When both tracks operate in concert, the programme moves households from fragile safety to adaptive stability. Diet quality improves, reliance on single climate-sensitive activities diminishes, and households hold liquid, productive assets under women's or joint control. Community assets reduce daily frictions, increase market engagement, and strengthen social cohesion. Governance safeguards and measurement embed fairness and responsiveness, making the programme effective, legitimate, and learnable. This end-state is not a fixed point but a dynamic equilibrium: systems that continue to adapt as prices, climate, and social conditions evolve.
6.1.9 Limitations and Boundary Conditions
The conclusions and design recommendations in this study are qualified by several limitations and boundary conditions that shape the feasibility, fairness, and durability of impacts. Recognizing these constraints upfront is essential to avoid overgeneralization, prevent unintended consequences, and ensure responsible scaling.
Seasonality and timing. Dietary diversity, labour availability, and opportunity costs vary across agricultural seasons. Lean periods compress diets and raise liquidity needs; harvest periods relieve pressure and alter participation dynamics. If transfers are late or irregular, protective effects erode. If measurements occur in favourable seasons, results may overstate progress. Boundary condition: synchronize transfer schedules and work cycles to local calendars; collect MDD and time-use data at multiple points across the yearto isolate programme effects from seasonality.
Geographic heterogeneity and market access. Identical interventions can have divergent impacts depending on water availability, soil quality, topography, and proximity to markets. Irrigation rehabilitation is transformative in water-scarce areas but limited where market isolation is binding; feeder roads deliver returns when production potential and buyer presence exist. Boundary condition: conduct context diagnostics such as ecological, market, and institutional before asset selection;bundle works with market facilitation where demand is thin.
Measurement gaps and misattribution. Under-collection of time-use and empowerment data can mask the double burden or the lack of agency gains. Overreliance on inputs and outputs (workdays, kilometres of road) misrepresents outcomes. Boundary condition: institutionalize a lean core indicator set comprising of MDD- W/MDD-C, key WEAI/pro-WEAI domains, time-use, fairness metrics and ensure regular analysis feeds into action plans. Avoid attributing improvements to the programme without triangulation against seasonal, price, and climate data.
Governance bias and underreporting. Fear of retaliation and social desirability suppress reporting of elite capture, coercion, or politicization. Grievance mechanisms can underperform if anonymity is weak, resolution timelines are unclear, or complainants doubt follow-through. Boundary condition: provide anonymous channels , publicly post resolution timelines, conduct third-party verification, and disclose aggregate grievance statistics. Expect underreporting and triangulate with spot checks and confidential qualitative methods.
Capacity and resource constraints. Robust transparency, social audits, childcare provision, and verification systems require trained staff and budgets. Under-resourced teams may struggle to maintain standards. Boundary condition: phase complexity; start with core safeguards and indicators, expand as capacity grows; standardize tools;invest in district training and mentorship; align with national systems to leverage infrastructure and economies of scale.
Norms resistance and intra-household dynamics. Shifts in women's control over assets and income can trigger appropriation or backlash. Without household sensitization, asset transfers intended for women may be captured;without group support, isolated women may face pressure to relinquish control. Boundary condition: pair transfers and payment routing with household dialogues, couple sessions, and women's group formation;embed grievance pathways for coercion cases and link to survivor-centred services; engage male champions and community leaders to support acceptance.
Macroeconomic volatility and inflation. Cash transfers lose purchasing power during inflation or supply shocks. Food transfers can stabilize consumption but face logistics constraints and lack flexibility. Boundary condition: adopt adaptive transfer modalities such as cash in normal conditions, food or vouchers during market stress; adjust transfer values with price monitoring;offer basic financial literacy to support budgeting under volatility.
Climate extremes and covariate risk. Droughts, floods, and heat waves can overwhelm asset benefits if designs are not climate-resilient. Boundary condition: prioritize climate- proofed assets which includes water harvesting, erosion control, resilient irrigation designs; integrate early warning systems and shock-responsive scalability; support maintenance plans and local stewardship committees to preserve functionality.
Sustainability and service linkages. Benefits dissipate if post-project service layers are missing: veterinary services for livestock, extension for gardens, maintenance for infrastructure, and reliable markets. Boundary condition: formalize linkages to producer associations, animal health workers, agro-dealers, microfinance providers, and local authorities;establish memoranda of understanding; budget for maintenance and handover;promote community ownership structures that keep assets functional.
Ethical and data governance risks. Weak consent processes, poor privacy protections, or insecure data storage can harm trust and bias reporting. Boundary condition: implement ethical data protocols such as clear consent, minimization of personally identifiable information, secure storage, access controls and communicate how data improves services.Train enumerators in sensitive interviewing and confidentiality.
External validity and scalability. Findings from specific districts or livelihood zones may not generalize nationally. Scaling without adaptation risks misfit and wasted resources. Boundary condition: replicate pilots across diverse contexts, compare results, and adapt modalities to local constraints. Maintain design flexibility rather than imposing uniform templates.
In sum, the limitations and boundary conditions do not negate the promise of FFW/CFW; they define the boundaries within which that promise is realized. A disciplined commitment to timing, diagnostics, measurement, governance, gender responsiveness, climate resilience, and service linkages ensures that interventions remain effective, equitable, and credible. By acknowledging constraints and designing for them explicitly, Zambia can build public works that protect households today and strengthen their capability to face tomorrow's risks with confidence.
6.1.10 Synthesis: What Determines Effectiveness?
Effectiveness in Food for Work (FFW) and Cash for Work (CFW) programmes is not a single-variable outcome. It emerges from the intersection of design choices, governance quality, social norms, market functionality, and adaptive measurement systems. Put differently, these programmes succeed when four interlocking capacities comprising of relief-resilience balance, gender equity, governance and accountability, and measurement for learning are deliberately built into the operating model and maintained overtime.
1) Balancing urgent relief with durable resilience. The first determinant is the programme's ability to meet urgent consumption needs while building assets and capacities that reduce future risk. Transfer adequacy and predictability are non- negotiables: benefits must arrive on time, in meaningful amounts, and aligned to lean seasons and shock periods. Where markets function, cash offers flexibility and supports local trade; where markets are thin or volatile, food or vouchers can stabilize consumption. This near-term protection must be paired with resilience asset creation which includes soil and water conservation, climate-proofed irrigation, feeder roads, market stalls, storage facilities selected through participatory processes and tailored to local constraints. The design balance matters: too much emphasis on construction without protection can push households into negative coping; too much emphasis on transfers without assets risks dependency perceptions and fragile gains. Effectiveness therefore hinges on sequencing (timing works and transfers to agricultural calendars), modality agility (switching between cash and food when needed), and asset relevance (choosing works that reduce real frictions in production and market access).
2) Embedding gender equity measures that are practical and measurable. Programmes only deliver equitable results when women can participate without deepening time poverty and when benefits accrue to and through women. Care-aware design features such as half-day shifts, flexible schedules, near-home worksites, lighter task options for pregnant or lactating women, onsite childcare, safe transport, genderappropriate WASH convert formal inclusion into practical inclusion. Payment routing to women's accounts protects financial autonomy, while transparent rate cards deter discrimination. Beyond access, effectiveness requires attention to agency: women's control over assets (e.g., small livestock, garden kits), income, and voice in committees should be explicit programme goals. Household sensitization, couple sessions, and women's groups reduce appropriation risks and build collective voice. Critically, gender equity must be measured: time-use diaries capture changes in workload; empowerment instruments track decision-making over production, income, and leadership. The determinant here is notjust intent but operational fidelity by consistently implementing care-aware features and monitoring whether they actually reduce constraints and increase agency.
3) Institutionalizing governance and accountability that withstand local power and political pressures. Effectiveness collapses when capture, favouritism, or politicization distort who gets in and who gets paid. The determinant is a transparent operating architecture: publicly posted eligibility criteria, beneficiary and waitlists, muster rolls, and wage rates; randomized selection among verified eligibles when oversubscribed; representative social audits;multi-channel grievance redress with anonymity and published resolution timelines; and independent verification through third-party checks and spot audits. Role separation is crucial in heightened political contexts: invite traditional authorities into bounded validation while keeping final selection rule-based and insulated from partisan influence. Importantly, governance systems need "teeth" which are clear sanctions for misconduct and incentives for good performance plus redundancy to withstand staff turnover and electoral cycles. When accountability is visible and trusted, programmes gain legitimacy, grievance volumes are manageable and resolved, and community support becomes a sustaining asset rather than a source of contestation.
4) Strengthening measurement and learning so data actually drives decisions. Effectiveness is sustained when programmes measure outcomes that matter (diet quality, agency, fairness), interpret signals quickly, and act. A core indicator set which consist of dietary diversity for women and children, select empowerment domains, time-use, inclusion/exclusion errors, grievance volumes and resolution times, and concordance between admin and independent data should feed quarterly review cycles. Mixed methods deepen understanding: focus groups and key informant interviews reveal underreported issues (e.g., coercion, unsafe worksites, asset appropriation). Evaluation logic must be seasonality-aware, repeating measures across agricultural phases to avoid misattribution. Digital tools can improve timeliness, but ethical data governance like consent, privacy, and secure storage is paramount. The determinant here is institutional culture: a measure-interpret-act loop that reliably triggers adjustments to transfer timing, modality, asset selection, training offers, site safety, and grievance capacity.
Cross-cutting operational conditions. Several enabling conditions amplify the four capacities above:
• Market functionality and price monitoring: Cash delivers value when supply exists and prices are stable;programmes must track local prices and switch modalities or adjust values during stress.
• Climate resilience mainstreaming: Works should be designed for droughts, floods, and heat, with maintenance plans and local stewardship so assets persist beyond construction.
• Skills and enterprise pathways: CFW become promotive when linked to demand- driven modules, certification, placements, and starter capital otherwise skills may be acquired but not monetised.
• Service linkages: Veterinary services, extension support, microfinance partners, and buyer networks sustain gains post-project.
• Norms engagement: Household dialogues and leadership buy-in reduce backlash against women's control over income and assets.
Collectively, these conditions determine whether FFW/CFW shift from episodic relief to systems that protect today and build tomorrow. When the balance, gender responsiveness, governance, and learning are all present and functioning, programmes deliver higher and more equitable returns particularly for women and the poorest households. When they are absent or weak, inequities deepen, trust erodes, and gains dissipate.
In synthesis, effectiveness is not accidental, it is engineered through design rigor, disciplined accountability, gender-aware operations, and an institutional commitment to learning. This is the actionable core of public works that truly build resilience.
6.2.0 Policy Implications
6.2.1 1ntegrating Public Works into National Resilience Strategies
Integrating FFW and CFW into Zambia's resilience architecture requires moving from programmatic pilots to system-enabled delivery, with clear institutional anchors, financing pathways, and district-level planning cycles. The central policy shift is to treat public works as localized adaptation infrastructure as well as social protection, ensuring that assets both stabilize consumption in the short term and reduce hazard exposure and livelihood riskovertime.
1) Establish the policyanchorand design authority. At national level, designate public works as a pillar within the climate-smart social protection framework, co-owned by social protection, agriculture, water/environment, and local government. Issue a Public Works Resilience Directive that sets standards for asset classes, design criteria, gender safeguards, and MEL expectations. The directive should align with national planning instruments (Medium-Term Expenditure Framework, district integrated development plans) so public works are budgeted and sequenced alongside other public investments. Establish a Public Works Technical Unit to publish design catalogues, update climate criteria, and train district teams.
2) Build a climate-rational asset portfolio. Use ward-level diagnostics to prioritize assets with the highest marginal risk reduction and productivity gains: solar-powered micro-irrigation, small weirs and sand dams, hillside stabilization, infiltration trenches, vegetative strips, gulley control, and resilience-enhancing community water points. Classify assets into rapid, medium, and complex builds to balance seasonal labour availability and technical complexity. Each asset entry must include engineering specs, climate rationale, O&M requirements, gender-aware task guidance, proximity/safety criteria, and MEL indicators (functionality, coverage, maintenance adherence, nutrition relevance).
3) Embed O&M from day one. Durability is policy-critical. Require Asset Management Committees with defined roles: routine inspection, fee collection or contribution management, minor repairs, and escalation protocols. Ring-fence O&M financing through a mix of community contributions, district maintenance budgets, and, where appropriate, modest user fees. Train local forepersons and technicians via short courses linked to the asset catalogue;publish a maintenance schedule and conduct social audits at 6- and 12-month intervals.
4) Operationalize "CFW + skills" to create employability loops. Convert cash transfers into employability by scheduling short, demand-aligned modules such as poultry care, food processing and hygiene, irrigation O&M, pump and small engine maintenance, basic electrical, digital recordkeeping, and bookkeeping during or immediately after works cycles. Partner with accredited providers for certification;run tracer surveys at 3/9/18 months, reporting placement and retention by gender and ward. Link cohorts to starter kits (tools, hygiene packs, seed packs) and matched savings for micro-assets, ensuring liquidity turns into productivity. Publish cohort dashboards and adapt course menus using labour market scans.
5) Mainstream gender safeguards as minimum service standards. Make gender-smart MSS non-negotiable: flexible and half-day shifts; avoidance of peak domestic hours; near-home sites; lighter task matrices; onsite childcare with feeding breaks;gender- appropriate WASH and potable water; safe transport;and GBV-safe supervision with named focal points. Default payment routing to women's accounts with KYC support and digital literacy. Allocate 10-12% of works costs to MSS and link implementer payments to compliance. Conduct time-use checks during supervision to detect load spikes and adjust tasking or shift structures quickly.
6) Align targeting and sequencing to the two-track logic. Operationalize Track 1 (Protective) by prioritizing wards with recurrent lean-season food insecurity;deliver predictable transfers and nutrition SBC. Operationalize Track 2 (Promotive) by routing works to resilience assets and skills pathways, with connections to microenterprise support, market inclusion, and financial services (savings, credit, microinsurance). Use a seasonal calendar to sequence works and training, avoiding conflict with planting, harvest, and school terms.
7) Finance as a portfolio, not a project. Blend financing sources which include national budget lines, district allocations, climate funds, donor support into a multi-year investment envelope. Introduce performance grants for districts that achieve asset functionality targets, gender MSS compliance, and MEL utilization. Pilot results-based financing for maintenance adherence and placement outcomes.
8) Integrate market systems and service linkages. Public works should reduce transaction costs by improving roads, market stalls, aggregation points, and storage. Convene market days and buyer-seller forums linked to new assets. Formalize linkages to veterinary services, extension agents, agro-dealers, producer associations, and microfinance partners through Molls, establishing an ecosystem that sustains benefits beyond project timelines.
9) Institutionalize MEL as the learning engine. Mandate MDD-W/MDD-C as nutrition indicators; WEAI/pro-WEAl and time-use for empowerment and workload; and fairness metrics (error rates, grievance volumes/resolution times, admin-independent concordance). Run quarterly measure-interpret-act reviews to adjust transfer timing, asset selection, MSS resourcing, and training menus. Publish district dashboards for transparencyand accountability.
In sum, policy integration means shifting public works from ad hoc labour mobilization to structured, climate-informed, gender-responsive, and data-driven delivery. The payoff is twofold: protection when it matters most, and durable adaptation capacity that strengthens livelihoods and public trust.
6.2.2 Gender-Responsive Social Protection as Non-Negotiable
Making gender responsiveness non-negotiable requires structural design choices that recognize care realities, prevent GBV risks, expand women's financial control, and measure agency notjust participation. Policy must move beyond intent to enforceable standards embedded in budgets, contracts, and supervision.
1) Codify Minimum Service Standards (MSS) in policy and contracts. Publish a national MSS protocol for public works that specifies: flexible and half-day shifts; avoidance of pre-dawn/evening tasks; seasonal scheduling that respects caregiving and agricultural calendars;near-home sites; lighter task options for pregnant/lactating women and Page 225 of 278 caregivers;onsite childcare with feeding breaks; gender-appropriate sanitation and potable water; vetted transport;GBV-safe supervision with named focal points; and publicly posted rate cards and pay calendars. Make MSS contractual conditions for contractors and implementing partners, with payment deductions or suspension for non-compliance.
2) Design task matrices that reduce time poverty and physical strain. Issue genderaware task matrices that include nursery bed preparation, seedling raising, composting, mulching, community greening, recordkeeping, inspection and maintenance of small assets, and water point management. Train supervisors to assign tasks fairly and implement accommodations for pregnancy and elder care. Any deviation from task matrices should be grievable via the GRM, with rapid remedial action.
3) Guarantee childcare, WASH, and safe transport. Allocate ring-fenced budgets for onsite childcare staffed by trained carers. Provide separate, safe sanitation facilities and handwashing points; ensure potable water availability. Vet transport providers and routes; set minimum safety standards for sites (lighting, distance thresholds, escort options where necessary). These features are preconditions for equitable participation, not optional extras.
4) Enforce equal pay and route payments to women's accounts. Display rate cards publicly; tie partner performance reviews to adherence. Default routing of transfers to women's bank/mobile wallets;support KYC and digital literacy. Where joint accounts are necessary, define withdrawal rules that protect women's access. Communicate entitlements clearly during orientation to reduce exploitation and misinformation.
5) Safeguard women's control over assets. Prioritize women-controlled transfers (small livestock, garden kits, tools). Register assets to women (or jointly) and conduct household sensitization to reduce appropriation. Provide basic husbandry training (vaccination, deworming, feed formulation, housing) and link to veterinary services. Encourage group-based market engagement to bolster bargaining power and reduce trader monopsony.
6) Bundle empowerment services. Integrate nutrition counselling (dietary diversity, IYCF, food hygiene) and SRH services into work cycles. Offer financial literacy and microskills modules (bookkeeping, pricing, cash flow, phone-based recordkeeping), paired with sectoral skills (poultry care, food processing, irrigation O&M). Set participation thresholds (e.g., >50%women) and pair them with enabling supports such as childcare, flexible hours and lighter tasks to avoid tokenism and ensure feasibility.
7) Measure time and agency and act on the signals. Institutionalize time-use diaries and empowerment measures (WEAI/pro-WEAI) alongside nutrition indicators (MDD- W/MDD-C). Use data to detect rising time burdens or low agency scores; adjust task matrices, expand childcare, or re-route payments accordingly. Publish gender dashboards at district level to reinforce accountability.
8) Protect against GBV and harassment. Institute zero-tolerance GBV policies; train supervisors and focal points in survivor-centred responses and referral pathways. Provide anonymous reporting channels (boxes, toll-free lines, SMS/WhatsApp); guarantee confidentiality and protection from retaliation. Conduct regular site safety audits and publicly report corrective actions.
9) Engage norms and households. Run community sensitization sessions to normalize women's paid work and control over income/assets. Facilitate couple dialogues; engage male champions and traditional leaders to bolster acceptance. Link women's groups to savings, training, and market days, strengthening collective voice and peer support.
10) Finance and enforce. Allocate 10-12% of works budgets to MSS and women- centred services. Make compliance a condition for disbursement;include MSS scorecards in contractor and partner performance reviews. Use GRM data and social audits to detect gaps and enforce remedies.
Policy that recognizes care realities and enforces gender safeguards transforms public works from "access in principle" to access in practice, improving nutrition, income control, and agency for women without increasing unpaid care burdens or exposure to harm.
6.2.3 Governance and Accountability to Counter Elite Capture
Elite capture undermines fairness, trust, and impact. Effective policy must harden governance architecture against manipulation across the delivery chain: targeting, selection, task allocation, attendance verification, and payment. The goal is to combine transparency, rule-based selection, representative oversight, multi-channel grievance redress, and independent assurance within a coherent, district-managed system.
1) Transparency by default—publish the rules and the records. Require visible posting of eligibility criteria, beneficiary and waitlists, rate cards, muster rolls, and pay calendars in local languages at ward offices, worksites, schools, clinics, markets, and churches. Date-stamp updates and archive versions for back checks. Use standardized templates with unique identifiers to reduce manipulation. Pair postings with information sessions explaining rights, processes, and grievance channels.
2) Hybrid targeting with public validation and appeals. Marry community ranking with objective screening to capture local knowledge and constrain bias. After provisional lists are compiled, hold public validation fairs where names are read out, criteria are explained, and appeals are accepted. Provide a 10-14-day appeals window with simple forms in local languages and clear documentation standards. Conduct seasonal revalidation every 6-12 months to align with lean seasons and prevent status ossification.
3) Randomized selection among verified eligibles when oversubscribed. Where demand exceeds capacity, use randomization among households verified as eligible. Verification should be arms-length (district staff not embedded in local power structures or external observers). Publicly document and post randomization outcomes. Randomization does not replace vulnerability targeting; it is a fair tie-breaker once objective screening is complete.
4) Representative social audits with real powers. Establish oversight committees with women, youth, and persons with disabilities. Train members to read attendance sheets and pay rosters, spot anomalies, and file structured audit reports. Mandate social audit days to read out records, validate complaints, and agree remedial actions. Rotate membership every 6-12 months;require conflict of interest declarations;empower committees to recommend sanctions.
5) Strong, safe, multi-channel grievance redress. Implement a GRM that includes anonymous complaint boxes, toll-free lines, SMS/WhatsApp, and in-person focal points. Set service-level agreements (acknowledgement within 48 hours; resolution within 14 days; independent appeal thereafter). Tag complaints by type (targeting, payment, GBV, worksite safety, asset quality). Publish monthly GRM dashboards (volumes, median resolution time, percent upheld) and discuss them during social audits. Maintain remedy logs to ensure actions taken which include task reassignment, supervisor rotation, and payment correction are traceable.
6) Independent assurance and detection of hidden capture. Commission third-party audits and randomized spot checks of enrolment and muster rolls. Use indirect reporting tools (confidential formats) to estimate prevalence of sensitive behaviours (bribe solicitation, list manipulation). Select high-risk sites for intensified oversight (unusual attendance patterns, payment anomalies). Ensure independent findings trigger corrective actions ranging from re-selection, repayment, sanctions, referral to authorities and publicly communicate both issues and fixes to maintain trust.
7) Role separation in politicized contexts. Exclude elected officials from selection committees;invite them to public validation events to maintain visibility without control. Adopt integrity pacts and citizen charters; consider photo or biometric verification where impersonation/manipulation risks are material. Establish an independent ombudsperson at district or provincial level to investigate escalated complaints and enforce remedies, especiallywhere contestation is high.
8) Sanctions and incentives. Define graduated sanctions such as warnings, suspension, removal, and legal referral for misconduct. Pair with positive incentives such as recognition of high-integrity sites, small performance grants for transparency improvements. Publish sanction and incentive outcomes to build a culture of integrity.
9) Fairness metrics and dashboards. Track inclusion/exclusion error rates, appeals upheld, grievance volumes and resolution times, and admin-independent concordance between MIS and verification. Publish district dashboards quarterly;convene learning reviews to enact corrective measures. Use dashboards to benchmark sites and spur peer learning.
Governance resilience depends on visibility, contestability, and credible redress. With these measures in place, public works deliver fairly, protect participant rights, and retain legitimacy even under local power pressures.
6.2.4 Measurement and Learning Systems
Measurement and learning systems are the nervous system of public works. They convert implementation into insight and ensure that data guides decisions. The policy imperative is to measure outcomes that reflect the theory of change which covers diet quality, empowerment and workload, and fairness and to institutionalize measure- interpret-act cycles that are seasonality-aware, ethical, and transparent.
1) Define a core indicator set aligned to outcomes. Adopt a concise, high-value set of indicators:
• Nutrition: MDD-W (10 food groups; 24-hour recall) and MDD-C (8 food groups), seasonally consistent.
• Empowerment and workload: WEAI/pro-WEAl domains (production, resources, income, leadership, time) and 24-hourtime-use diaries .
• Fairness and governance: inclusion/exclusion error rates (spot checks), appeals upheld (%), grievance volumes and median resolution times, and adminindependent concordance (match rate between MIS and independent verification of attendance/payments). These indicators tie directly to protective and promotive logic and ensure course correction is possible.
2) Set sampling frames and frequency that capture seasonality. Use panel designs for MDD-W/MDD-C and WEAI at baseline, 12 and 24 months, measured in the same season to avoid misattribution. Deploy sentinel phone surveys monthly (5-7 minutes) to track worksite safety, payment timeliness, childcare sufficiency, and GBV-safe supervision presence. Conduct social audits quarterly and independent audits each semester. These layers provide both breadth and depth.
3) Build data integrity and ethics protections. Train enumerators on protocols, probing, privacy, and sensitivity (e.g., GBV-linked complaints); ensure informed consent and anonymization. Minimize personally identifiable information; secure storage; access controls. Communicate to participants how data improves services; share district dashboards publicly to build trust.
4) Operationalize dashboards and decision triggers. Publish district dashboards each quarter with traffic-light signals for key indicators. Define operational triggers, for example:
• If MDD-W stagnates or declines, intensify nutrition SBC, adjust transfer timing/value, and increase kitchen garden/irrigation assets.
• If WEAI time domain worsens, expand childcare and lighter tasks, or add nearhome micro works.
• If grievance resolution times exceed targets, add GRM staff, reassign cases, or simplify processes.
• If inclusion/exclusion error rates are high or appeals upheld exceed thresholds, recalibrate objective criteria and extend appeals windows.
• If admin-independent concordance is low, tighten verification protocols and conduct targeted spot checks.
5) Use mixed methods for depth and triangulation. Numbers rarely tell the whole story. Pair quantitative tracking with focus groups (women, men, youth, persons with disabilities), key informant interviews (local leaders, extension workers, market actors), and case studies of atypical success or difficulty. Where underreporting is suspected, use indirect reporting tools to estimate prevalence of sensitive behaviours. Triangulate findings and act.
6) Ensure interoperability and role clarity. Create unique participant IDs linking public works MIS, GRM databases, and district nutrition and health data, under lawful data- sharing rules. Clarify roles: social protection and nutrition units co-lead nutrition/time- use modules; agriculture and TVET lead asset functionality and skills outcomes; district GRM units manage dashboards; third-party monitors run audits and indirect reporting. This reduces duplication and strengthens accountability.
7) Invest in capacity for analysis and action. Train district teams not just to collect data, but to analyse and decide. Hold quarterly learning reviews with managers, supervisors, and oversight committees; adopt and document corrective actions. Celebrate sites that solve problems; disseminate playbooks across districts. Over time, MEL becomes a culture of continuous improvement rather than compliance.
8) Communicate with care on integrity signals. When governance risks are detected, communicate both the problem and the fix; publish post-remedy indicator trends to show progress. This maintains trust and avoids demoralizing narratives.
When measurement is aligned to outcomes, ethically collected, and used to change course , public works evolve into responsive systems that protect today and build adaptive capacity for tomorrow.
6.3 Recommendations for Program Improvement
6.3.1 Overview
This section translates the study's findings into an operational blueprint for evolving Food-for-Work (FFW) and Cash-for-Work (CFW) from short-term relief into a system Page 232 of 278 that consistently builds adaptive capacity. The central proposition is straightforward: public works are most effective when they are climate-rational (assets reduce hazard exposure and raise productivity), gender-responsive (participation strengthens women's agency without deepening time poverty or GBV risk), and governed transparently (rules are visible, contestable, and enforced so benefits reach those most in need). The operational corollary is that design, governance, and measurement must be hard-wired upfront into manuals, budgets, procurement clauses, supervision protocols, and data systems so the model is coherent, repeatable, and accountable.
Climate rationality begins with precise diagnostics. At ward and district level, programme teams should map hazard exposure (dry spells, rainfall variability, erosion hotspots), livelihood constraints (market isolation, input scarcity, liquidity traps), and service gaps (water points, animal health outreach, maintenance capacity). This yields a context-specific asset menu that privileges small-scale irrigation, water harvesting structures, land rehabilitation, and resilient water points. These assets reduce climate risk, protect soils, conserve moisture, and raise the probability that households can farm profitably in the next season. Public works become the delivery mechanism for localized adaptation, not just labour mobilization. To make assets durable, operations and maintenance (O&M) must be budgeted from inception, with roles and routines assigned to community committees and verified through social audits.
CFW as an employability platform reframes cash from a transfer into a springboard. When short, practical modules (30-60 hours) are linked directly to CFW cycles encompassing poultry care, food processing and hygiene, irrigation O&M, small engine maintenance, bookkeeping and digital records including workdays converted into micro-credentials that employers recognize and entrepreneurs can monetize. Certification through accredited providers, external assessments, and tracer surveys that report placement and retention by gender and ward shift accountability from "training delivered"to income and employment outcomes. In Zambia, coupling CFW and skills also builds local maintenance capacity for the very assets created by FFW closing the loop from construction to sustained utilization.
Gender-smart minimum service standards (MSS) ensure public works do not inadvertently increase women's unpaid care workload or expose them to risk. These standards should not be aspirational footnotes;they must be budget lines, contract clauses, supervisor checklists, and dashboard indicators. Core elements include flexible or half-day shifts; care-aware scheduling; onsite childcare with trained carers; genderappropriate WASH and potable water; safe transport;and GBV-safe supervision with named focal points and clear referral pathways. Payment design matters: routing transfers to women's accounts by default, supporting KYC onboarding, and publicly posting rate cards all reinforce equal pay for equal work and women's financial autonomy. Regular time-use checks during supervision visits provide early warning if care burdens spike during works cycles, triggering design adjustments.
Women-centred asset control and bundled services make resilience tangible. Evidence shows small livestock and kitchen gardens under women's or joint control function as liquid, shock-responsive assets. To preserve their insurance value and productivity, public works should fund basic veterinary outreach (vaccinations, deworming, feed planning) and link participants to extension services. Nutrition counselling aligned to Minimum Dietary Diversity indicators helps convert income and production into better diets, while SRH services reduce participation barriers. Adding financial literacy and micro-skills ensures transfers transition into sustainable cashflow through a shift from episodic relief to productive livelihoods.
Targeting and governance must blend local information with objective criteria and be contestable at multiple points. A hybrid model such as community ranking plus objective screening followed by public validation fairs and appeals windows balances fairness with local insight. Seasonal revalidation aligns support with lean periods without freezing dynamic poverty statuses. Transparent posting of criteria, lists, muster rolls, and rate cards anchors legitimacy, while social audits and randomized selection among verified eligibles dampen patronage when demand exceeds capacity. Grievance systems should be multi-channel and safe, with published resolution timelines and escalation paths to an independent ombudsperson in contested settings. The message is clear: fair rules, visible processes, and credible redress are non-negotiable.
Finally, a high-functioning Monitoring, Evaluation, and Learning (MEL) system acts as the programme's nervous system. Nutrition must be measured using standardized dietary diversity tools for women and children; empowerment and workload using WEAI/pro-WEAl and time-use diaries; and fairness through a small set of governance indicators which include error rates, appeals upheld, grievance volumes and resolution times, and concordance between administrative and independent datasets. To uncover hidden problems, experimental tools and randomized spot audits can detect underreported manipulation or bribe solicitation. Crucially, communication of integrity findings must be responsible: publish the problem and the fix, and track whether governance indicators improve in subsequent quarters.
In short, the package is coherent: climate-rational assets; CFW plus skills;gender-smart MSS; women-centred asset control and services; hybrid targeting with transparent oversight and robust GRM; and MEL that measures diets, agency, and fairness. Adopted together, these features shift public works from episodic relief into a predictable, accountable, resilience-building system aligned with Zambia's strategies and climate realities.
6.3.2 Programme Design & Delivery
Design principle 1: Climate-rational asset portfolios. Programme planning should start with a ward-level climate and livelihoods diagnostic: drought frequency, rainfall variability, soil degradation, groundwater potential, erosion risk, and market isolation. Using this map, teams construct an asset menu that privileges interventions with clear risk-reduction and productivity gains under plausible climate futures: solar pumping paired with drip kits; sand and check dams on feeder streams;infiltration pits; terracing and vegetative strips on steep slopes; and community water points designed with safe access for women and children. Each asset entry must include engineering standards, climate rationale, O&M plans, gendered task guidance, and MEL indicators. Siting must reflect safety and proximity for caregivers, and environmental safeguards should address downstream erosion, water allocation rules, and biodiversity considerations.
Operationalization.
Issue a national asset catalogue with design standards and climate criteria. Require ward-level participatory planning to rank investments against transparent criteria: hazard exposure, market access gains, nutrition relevance, caretaker safety, and O&M feasibility. Form Asset Management Committees early, with O&M micro-plans (user fees, repair schedules, spare parts, and toolkits) agreed at handover and verified during social audits. Build a simple maintenance financing mechanism (e.g., small user fees or community contributions), with transparent accounting and periodic audits. Train local forepersons and maintenance technicians using standardized curricula linked to the asset catalogue. Introduce independent works certification before payment to contractors, and schedule post-construction quality checks at 6 and 12 months.
Design principle 2: CFW as a platform for employability and enterprise. The CFW-plus- skills model addresses dual constraints: liquidity today and productivity tomorrow. Short, modular training aligned to local demand and the asset menu should be offered during or immediately after CFW cycles such as poultry management, food processing and hygiene, irrigation O&M, pump and small engine maintenance, basic electrical, and bookkeeping/digital records. Certification through accredited providers and external assessments confers recognizable value; micro-credentials help graduates signal skill to employers. Pair training with entrepreneurship coaching (pricing, records, customer service), light starter kits (toolboxes, hygiene kits, seed packs), and matched savings for micro-assets. Publish tracer surveys at 3/9/18 months, reporting completion, placement, retention, and enterprise survival by gender and ward. Use results to refresh course offerings, mentor networks, and market linkages.
Design principle 3: Gender-smart minimum service standards (MSS). Codify MSS into guideline annexes, budgets, and contracts. Minimums include: flexible or half-day shifts; no compulsory pre-dawn or late-night work;scheduling that avoids peak care and harvest times;onsite childcare with trained carers and feeding breaks; separate, safe WASH and potable water; vetted transport where distance exceeds safe thresholds; GBV-safe supervision with named focal points and survivor-centred referral; public rate cards and payment calendars;default routing of transfers to women's accounts with KYC support and digital literacy. Allocate 10-12% of works costs to fund MSS and make compliance a condition for contractor and partner payments. Institutionalize time-use checks during supervision visits to identify load spikes for caregivers, and adjust tasking or shift structure accordingly.
Design principle 4: Women-controlled assets plus bundled empowerment services. Register starter flocks, garden kits, and tools to the woman (or jointly) to reduce appropriation risk; pair with household sensitization and community norms dialogue to strengthen acceptance. Align veterinary outreach (vaccinations, deworming, parasite control) and feed/fodder planning to protect herd value. Integrate nutrition SBC around dietary diversity and IYCF practices so production and income translate into improved diets. Strengthen SRH access to reduce participation barriers. Add financial literacy (budgeting, savings groups, cost/price setting) and sectoral micro-skills (poultry, processing, bookkeeping) so transfers become recurring income. In appropriate contexts, pilot risk pooling or micro-insurance products, alongside good husbandry, to preserve asset trajectories.
Risks and mitigations.
• Construction quality and missing O&M: Mitigate through standard designs, trained forepersons, independent works certification, O&M funding, and scheduled maintenance audits.
• Skills drift from labour demand: Co-design curricula with employers; use tracer data to adapt;run quarterly labour market scans.
• Underfunded MSS: Ring-fence MSS budget lines; link contractor and partner payments to MSS compliance; publish MSS scorecards.
• Household appropriation of women's assets: Combine asset registration with household dialogues, community norm campaigns, and documentation naming women as asset holders; reinforce with women's groups and grievance pathways.
The payoff. This architecture protects consumption now while building capabilities for tomorrow: risk-reducing assets, recognized skills, safer and more inclusive worksites, and women-centred asset control backed by services. That is the essence of turning public works from a temporary response into a structural instrument for resilience.
6.3.2 Targeting & Governance
Principle 1: Hybrid targeting with public scrutiny. Neither community ranking nor purely objective screening is sufficient on its own. Community committees know hardship histories, transient shocks, disability, and caregiving burdens that proxies may miss. Objective screening adds consistency and constrains bias. Operationally, marry these strengths:
1. Facilitate community ranking with independent moderators and simple, transparent criteria;
2. Run objective verification with clear documentation;
3. Convene public validation fairs where provisional lists are read out and appeals are accepted;
4. Post final lists (beneficiaries and waitlists), criteria, and outcomes in local languages at accessible locations. Build in seasonal revalidation every 6-12 months to align with lean periods and avoid freezing poverty statuses. Track appeals upheld (%) and inclusion/exclusion error rates as fairness KPIs.
Principle 2: Transparency by default. Publish the rules and show the numbers. Eligibility criteria, rate cards, muster rolls, and pay calendars should be posted visibly at ward offices, worksites, schools, clinics, markets, and churches. Social audit days such as open community meetings anchored by representative oversight committees should read out attendance, payments, and asset O&M updates, provide a forum for verification, and agree remedial actions. Use standardized templates and date-stamp records to deter manipulation.
Principle 3: Strong, safe, multi-channel grievance mechanisms. Adopt a GRM with multiple channels which include complaint boxes, toll-free lines, SMS/WhatsApp, and in-person focal points with options for anonymity. Set service-level agreements (e.g., acknowledgement within 48 hours;resolution within 14 days; independent appeal thereafter). Tag complaints by category (targeting, payment, GBV, worksite safety, asset quality), and publish monthly dashboards showing volumes, median resolution time, and percent upheld. GBV-linked reports require survivor-centred handling; define confidential pathways and coordinate with district gender/health services. Maintain remedy logs so actions taken such as task reassignments, supervisor rotations, payment corrections are traceable .
Principle 4: Arms-length oversight where politicization risks are high. In settings where local political actors attempt to influence targeting or task allocation, role separation is critical. Exclude elected officials from selection committees;adopt citizen charters and integrity pacts;consider photo or biometric verification where impersonation or list manipulation risks are material. Establish an independent ombudsperson at district/provincial level to investigate escalated complaints and enforce remedies. When demand exceeds slots, randomised selection among verified eligibles reduces room for manoeuvre and signals fairness.
Principle 5: Evidence-based detection of hidden capture. Standard surveys underdetect sensitive behaviours due to fear or social desirability. Complement routine monitoring with indirect reporting tools and randomized spot audits. Indirect formats provide plausible deniability and more truthful aggregate estimates of experiences like bribe solicitation or list manipulation. Communicate findings responsibly: publish both the problem and the corrective actions (new audit protocols, committee rotations, tightened trails), and track governance indicators in subsequent quarters to demonstrate improvement.
Principle 6: Representative oversight with real powers. Oversight committees must not be symbolic. Ensure women, youth, and persons with disabilities are present and empowered with the right to demand records (attendance sheets, pay rosters), call extraordinary social audits, and recommend sanctions. Provide short training on reading site records, identifying anomalies, and filing structured audit reports. Rotate roles every 6-12 months to reduce gatekeeper capture and require conflict of interest declarations .
Putting it together: a governance pipeline. A workable pipeline for Zambia would look like this:
(a) Outreach and information sessions in local languages;
(b) Community ranking with independent facilitation;
(c) Objective verification;
(d) Public validation fairs with a 10-14 day appeals window;
(e) Finalization and posting of lists and criteria;
(f) Randomized selection among verified eligibles when oversubscribed;
(g) Routine social audits with musters and payments read aloud;
(h) GRMs accessible via boxes, toll-free, and SMS/WhatsApp through monthly dashboards and published resolution times;
(i) Independent audits each quarter;
(j) Where politicization is significant, role separation, integrity pacts, identity verification, and an ombudsperson to break ties and enforce remedies.
Sanctions and incentives. Governance systems need "teeth" and carrots: graduated sanctions for misconduct (warnings, suspension, removal, referral), coupled with positive incentives (recognition of high-integrity sites, small performance grants for transparency improvements). This balance cultivates a culture of integrity and accountability.
This architecture is not bureaucratic excess; it is what makes public works public which covers transparent rules, participatory oversight, enforceable rights, and answerability.
6.3.3 Monitoring, Evaluation & Learning
Why MEL matters. Public works succeed or fail in the details. A robust MEL system is therefore not a donor tick-box; it is the instrument panel that tells managers whether assets are functional, diets are improving, women's agency is rising, and the rules are being fairly applied. The key shift is to measure outcomes that trackthe theory of change such as dietary diversity (MDD-W and MDD-C), empowerment and workload (WEAI/pro-WEAl and time-use diaries), and fairness (appeals upheld, grievance resolution times, error rates) and to use the data to change course .
Nutrition: measure what matters. Integrate MDD-W (10 food groups, 24-hour recall) and MDD-C (8 food groups including breastmilk where applicable) using standard instruments and seasonally consistent measurement. Establish a baseline, then 12- and 24-month follow-ups in the same season. Disaggregate by sex, pregnancy/lactation status, child age, and ward; publish district dashboards so implementers can see where diets are changing and where they are stuck. Pair results with action logic: if MDD stagnates, adjust transfer timing or value, intensify nutrition SBC, and retarget assets toward kitchen gardens, irrigation, and water points that improve diet quality.
Empowerment and workload: track agency and time, not just participation. Use WEAI/pro-WEAl to assess agency in production, resources, income, leadership, and time. Combine with 24-hour time-use diaries to quantify shifts in unpaid care and domestic work. If MSS are working, women's time in unpaid care should not increase relative to baseline;if it does, pivot including lighten tasks, expand childcare hours, add near-home micro-works, oradjust shift structures. Trackwhetherwomen's control over income and assets improves and whetherthey hold meaningful roles in site committees.
Fairness and governance: show your working. Build a small, visible dashboard of governance indicators that are straightforward to calculate and hard to game:
• Inclusion/exclusion error rates (from spot checks).
• Appeals upheld (%).
• Grievance volumes and median resolution times.
• Concordance between programme MIS and independent verification of attendance/payments.
Commit to quarterly learning reviews at district level where managers, supervisors, and oversight committees act on the signals: rotate supervisors with high unresolved complaint rates; rebalance task matrices at sites where women's time use has deteriorated; switch payment providers where delays are systemic;and recalibrate targeting where appeals upheld exceed thresholds. Publish meeting minutes and action items to reinforce answerability.
Detecting hidden risks. Because overt surveys under-detect sensitive issues (targeting manipulation, bribe solicitation), run indirect reporting tools annually to estimate prevalence of such behaviours. Couple this with randomized spot audits of enrolment and muster rolls. Communicate results responsibly: integrity messaging should present both the issue and the remedy. Track whether governance indicators improve post- remedial action to build trust.
Sampling, frequency, ethics, and interoperability. Use panel designs for MDD- W/MDD-C and WEAI at baseline, 12 and 24 months. Deploy sentinel phone surveys (57 minutes) monthly to track worksite safety, payment timeliness, and childcare sufficiency. Conduct social audits quarterly. Ensure informed consent, anonymize datasets, and apply strict confidentiality to GBV-linked complaints. Architect data systems for interoperability: unique participant IDs linking public works MIS, GRM databases, and district nutrition/health data under lawful data-sharing rules. Assign clear roles: nutrition/time-use modules co-led by social protection and nutrition units; agriculture and TVET lead skills and asset functionality checks; district GRM units manage dashboards;third-party monitors handle audits and indirect reporting.
From measurement to management. Data that are not used are expensive trivia. Tie MEL to operational triggers:
• If MDD-W does not improve by a pre-agreed margin after 12 months, intensify nutrition SBC and increase the share of garden/irrigation assets.
• If WEAI time domain worsens, expand childcare and lighter task options.
• If grievance resolution times exceed targets, add GRM staff, reassign cases, or simplify processes.
• If error rates remain high, recalibrate objective thresholds and extend appeals windows.
• If women's income control lags, route payments more consistently to women's accounts and reinforce household sensitization.
Capacity building and culture. Train district teams not only in data collection but in analysis and decision-making. Embed a rhythm of quarterly reviews, site-level feedback circles, and public dashboards. Celebrate sites that solve problems and share playbooks across districts. Overtime, MEL becomes a learning system, not a compliance ritual; one that steadily raises impact, equity, and legitimacy
6.4.0 Areas for Further Research
6.4.1 Areas for Further Research: Longitudinal Resilience Impacts
One of the most pressing research gaps is understanding whether asset-linked FFW and CFW interventions produce durable resilience gains beyond the immediate project cycle. Current evidence largely focuses on short-term outcomes such as dietary diversity improvements during programme participation, temporary income boosts, and initial asset transfers. However, resilience is inherently dynamic and multi-seasonal. It depends on whether households can maintain or grow these gains across multiple drought cycles, market shocks, and social transitions.
Key research questions include:
• Do households that participate in asset-linked public works sustain improvements in diet quality (measured by MDD-W and MDD-C) over two to three years, particularlyduring lean seasons?
• How do asset portfolios evolve post-programme? For example, do small livestock holdings expand, remain stable, or erode under stress?
• Does livelihood diversification persist, or do households revert to mono-cropping and climate-sensitive activities once transfers end?
• What role does asset functionality play? Are irrigation systems, water points, and soil conservation structures maintained and utilized effectivelyafterhandover?
Methodological considerations:
• Panel cohort design: Track the same households over multiple seasons and years to capture resilience trajectories. Include both programme participants and a matched comparison group to isolate programme effects.
• Seasonal comparability: Collect data at consistent points in the agricultural calendar (e.g., lean season vs. post-harvest) to avoid misattribution.
• Mixed methods: Combine quantitative indicators (MDD-W, MDD-C, asset counts, income sources) with qualitative narratives on coping strategies, decision-making, and perceived resilience.
• Shock exposure mapping: Overlay household data with climate and market shock timelines to assess absorptive and adaptive capacity under real stress conditions.
Policy relevance: Findings would inform whether public works should be scaled as episodic relief or institutionalized resilience systems. If gains dissipate quickly, design adjustments such as extended maintenance support, savings linkages, and postprogramme coaching become critical. Conversely, evidence of sustained improvements strengthens the case for embedding public works into national climate adaptation strategies.
6.4.2 Areas for Further Research: Gendered Workload Dynamics
While gender-responsive design is widely advocated, empirical evidence on which combinations of interventions most effectively reduce women's time poverty remains limited. Public works often assume that flexible scheduling and childcare provision will suffice, but the interaction between task allocation, shift structure, and domestic responsibilities is complex.
Key research questions include:
• Which design features, childcare, staggered shifts or lighter tasks deliver the greatest reduction in unpaid care time without compromising earnings or output?
• How do these features affect women's participation in skills training and leadership roles?
• What trade-offs emerge between programme productivity and gender equity when implementing care-aware measures?
Methodological considerations:
• Randomized roll-outs: Introduce different combinations of gender-smart features across sites (e.g., childcare only vs. childcare + flexible shifts vs. childcare + lighter tasks) to compare impacts.
• Time-use diaries: Collect 24-hour recall data at baseline and follow-up to quantify changes in unpaid care, domestic work, and rest.
• Empowerment metrics: Use WEAI domains to assess whether reduced time poverty translates into greater agency over income, assets, and decision-making.
• Cost-effectiveness analysis: Compare the incremental cost of gender-smart features with their impact on participation, retention, and empowerment outcomes.
Policy relevance: Evidence would guide minimum service standards for gender equity in public works. It would also inform budgeting decisions by ensuring that childcare and flexible scheduling are treated as core infrastructure rather than optional extras. Ultimately, this research would help convert participation into empowerment rather than additional labour burden.
6.4.3 Areas for Further Research: Cost-Effectiveness of Skills Bundling
The "CFW + skills" model is conceptually compelling, but its cost-effectiveness relative to standalone cash transfers remains underexplored. Policymakers need evidence on whether bundling vocational training with public works delivers measurable employment and income gains thatjustify the additional investment.
Key research questions include:
• What is the incremental impact of skills training on employment rates, income diversification, and enterprise formation compared to CFW alone?
• How do certification and external assessments influence labour market outcomes?
• What is the cost per job obtained or enterprise sustained under different training models (short modules vs. full TVET courses)?
• How do gender dynamics shape returns to training, do women experience equal or higher gains when training is bundled with CFW?
Methodological considerations:
• Randomized encouragement design: Assign training opportunities randomly among CFW participants to estimate causal effects.
• Tracer surveys: Track employment, income, and enterprise survival at 6,12, and 24 months post-training.
• Cost-benefit analysis (CBA) and Social Return on Investment (SROI): Compare programme costs with monetized benefits (earnings, asset accumulation, reduced vulnerability).
• Disaggregation: Analyse outcomes by gender, age, and location to identify equity gaps.
Policy relevance: Results would inform whether skills bundling should be scaled nationally, targeted to specific cohorts, or redesigned for efficiency. If returns are high, integrating CFW with TVET becomes a cornerstone of resilience programming. If returns are modest, alternative pathways such as linking participants to private training providers or digital learning platforms may be more viable.
6.4.4 Areas for Further Research: Governance Innovations
Elite capture and politicization remain persistent risks in public works. While transparency and social audits are standard prescriptions, innovative methods for detection and deterrence require rigorous testing. Digital tools and behavioural experiments offer promising avenues but lack operational evidence in rural African contexts.
Key research questions include:
• Do list experiments and randomized audits significantly improve detection of underreported manipulation compared to conventional surveys?
• How do digital muster rolls and public dashboards affect perceptions of fairness and actual leakage rates?
• What design features minimize bias in list experiments (e.g., control item selection, framing)?
• How do communities respond to integrity messaging does disclosure of corruption risks erode trust or strengthen accountability when paired with visible remedies?
Methodological considerations:
• Stepped-wedge design: Phase in governance innovations across sites to compare pre/post fairness indicators.
• Behavioural metrics: Track changes in grievance volumes, appeals upheld, and admin-independent concordance.
• Qualitative inquiry: Explore participant perceptions of transparency and trust under different accountability regimes.
• Cost analysis: Assess the financial and administrative burden of digital tools and experimental methods relative to their impact.
Policy relevance: Evidence would guide the next generation of governance safeguards moving beyond paper-based audits to data-driven integrity systems. It would also inform donor investments in technology and capacity building for accountability.
6.4.5 Areas for Further Research: Livestock-Based Resilience
Small livestock transfers are widely promoted as resilience assets, yet design details such as ownership models, veterinary service delivery, and risk management remain contested. Research is needed to determine which configurations maximize asset retention, income generation, and women's agency.
Key research questions include:
• Does individual ownership outperform joint ownership in preserving women's control and decision-making?
• How effective are voucher-based veterinary outreach models in reducing mortality and distress sales?
• Can index insurance for small ruminants be scaled affordably, and does it reduce catastrophic herd depletion during droughts?
• How do these interventions affect household coping strategies, nutrition outcomes, and intra-household bargaining?
Methodological considerations:
• Comparative trials: Test individual vs. joint ownership models across similar contexts.
• Voucher pilots: Randomize access to veterinary vouchers and track herd health, mortality, and distress sale rates.
• Insurance experiments: Introduce index-based livestock insurance in selected sites; monitor uptake, claim processing, and behavioural responses.
• Mixed metrics: Combine herd dynamics data with WEAI scores, dietary diversity indicators, and qualitative narratives on decision-making.
Policy relevance: Findings would inform whether livestock transfers should remain a core feature of public works or be complemented by financial instruments and service linkages. They would also clarify how asset design interacts with gender norms and climate risk critical for scaling resilience pathways
REFERENCES
1. Angrist, J.D. & Pischke, J.-S. (2009) Mostly Harmless Econometrics:An Empiricist's Companion. Princeton, NJ: Princeton University Press.
2. Austin, P.C. (2009) Balance diagnostics for comparing the distribution of baseline covariates in propensity-score matched samples. Statistics in Medicine, 28(25), pp.3083-3107.
3. Austin, P.C. (2011) An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behavioral Research, 46(3), pp.399-424.
4. Beegle, K., Galasso, E. & Goldberg, J. (2017) The l ong-term i mpactsofpublic works i n Malawi. Policy Research Working Paper 8010. Washington, DC: World Bank.
5. Berhane, G., Hoddinott, J., Kumar, N. & Taffesse, A.S. (2014) Evaluation of Ethiopia's Productive Safety Net Programme. Washington, DC: International Food Policy Research Institute (IFPRI).
6. Biemer, P.P. & Lyberg, L.E. (2003) Introduction to Survey Quality. Hoboken, NJ: Wiley.
7. Blackwell, M., lacus, S., King, G. & Porro, G. (2009) CEM: Coarsened exact matching in Stata. The StataJournal, 9(4), pp.524-546.
8. Bloor, M., Frankland, J., Thomas, M. & Robson, K. (2001) Focus Groups in Social Research. London: Sage.
9. Bowen, G.A. (2009) Document analysis as a qualitative research method. Qualitative ResearchJournal, 9(2), pp.27-40.
10. Cameron, A.C. & Miller, D.L. (2015) A practitioner's guide to cluster-robust inference. JournalofHumanResources, 50(2), pp.317-372.
11. Campbell, J.L., Quincy, C., Osserman, J. & Pedersen, O.K. (2013) Coding in-depth semistructured interviews: Problems of unitization and intercoder reliability. SociologicalMethods&Research, 42(3), pp.294-320.
12. Chambers, R. (1997) Whose Reality Counts? Putting the First Last London: Intermediate Technology Publications.
13. CIOMS (Council for International Organizations of Medical Sciences). (2016) International Ethical Guidelines for Health-related Research Involving Humans. Geneva:CIOMS.
14. Cisse, J., Barrett, C.B. & Hoddinott, J. (2016) Measuring resilience to food insecurity in Africa. IFPRIDiscussionPaper Washington, DC: IFPRI.
15. Clay, E.J. (1986) The 1984-85EthiopianFamine. London: FrankCass.
16. Cochran, W.G. (1977) Sampling Techniques. 3rd edn. NewYork: Wiley.
17. Cohen, J. (1960) A coefficient of agreement for nominal scales. Educational and PsychologicalMeasurement, 20(1), pp.37-46.
18. Cohen, J. (1988) Statistical Power Analysis for the Behavioral Sciences. 2nd edn. Hillsdale, NJ: Erlbaum.
19. Creswell, J.W. (2014) Research Design' Qualitative, Quantitative, and Mixed Methods Approaches. 4th edn. Thousand Oaks, CA: Sage.
20. Creswell, J.W. & Plano Clark, V.L. (2011) Designing and Conducting Mixed Methods Research. 2nd edn. Thousand Oaks, CA: Sage.
21. Deaton, A. (1997) TheAnalysis of Household Surveys: A MicroeconometricApproach toDevelopmentPolicy. Washington, DC: World Bank.
22. Denzin, N.K. (1978) The Research Act: A Theoretical Introduction to Sociological Methods. 2nd edn. NewYork: McGraw-Hill.
23. Devereux, S. & Sabates-Wheeler, R. (2004) Transformative social protection. IDS WorkingPaper232. Brighton: Institute of Development Studies.
24. DFID. (1999) Sustainable Livelihoods Guidance Sheets. London: Department for International Development.
25. DIIS (Danish Institute for International Studies). (2013) Citizenshipandaccountability inAfrican governance. Copenhagen: DIIS.
26. Dillman, D.A., Smyth, J.D. & Christian, L.M. (2014) Internet,Phone,Mail,and Mixed-Mode Surveys: The Tailored Design Method. 4th edn. Hoboken, NJ: Wiley.
27. Dewey, J. (1929) The Quest for Certainty:A Study of the Relation of Knowledge and Action. NewYork: Minton, Balch.
28. FAO. (2011) Guidelines for Measuring Household and Individual Dietary Diversity. Rome: Food and Agriculture Organization.
29. FAO. (2016) RIMA-II: Resilience I ndex Measurement andAnalysis I I. Rome: Food and Agriculture Organization.
30. Fetters, M.D., Curry, L.A. & Creswell, J.W. (2013) Achieving integration in mixed methods designs—principles and practices. Health Services Research, 48(6 Pt 2), pp.2134-2156.
31. Filmer, D. & Pritchett, L.H. (2001) Estimating wealth effects without expenditure data—or tears. Demography, 38(1), pp.115-132.
32. Flick, U. (2018) An Introduction to Qualitative Research. 6th edn. London: Sage.
33. Fox, J. (2016) Applied Regression Analysis and Generalized Linear Models. 3rd edn. Thousand Oaks, CA: Sage.
34. Gali, J. (2012) Monetary Policy, Inflation, and the Business Cycle: An Introduction to theNew KeynesianFramework. 2nd edn. Princeton, NJ: Princeton University Press.
35. Gentilini, U. (2016) Revisiting the 'cash versus food' debate: New evidence for an old controversy. WorldBankResearch Observer, 31(2), pp.135-167.
36. Green, S.B. (1991) How many subjects does it take to do a regression analysis? MultivariateBehavioralResearch, 26(3), pp.499-510.
37. Greene, W.H. (2012) EconometricAnalysis. 7th edn. UpperSaddle River, NJ: Prentice Hall.
38. Groves, R.M., Fowler Jr., F.J., Couper, M.P., Lepkowski, J.M., Singer, E. &Tourangeau, R. (2009) SurveyMethodology.2nd edn. Hoboken, NJ: Wiley.
39. Guetterman, T.C., Fetters, M.D. & Creswell, J.W. (2015) Integrating quantitative and qualitative results in health science mixed methods research. Annals of Family Medicine, 13(6), pp.554-561.
40. Hirvonen, K., Hoddinott, J., Minten, B. & Stifel, D. (2022) Soil and water conservation, rainwater harvesting, and the performance of public works: Evidence from Ethiopia. World Development, 151,105744.
41. Hoddinott, J. (2006) Shocks and their consequences across and within households in rural Zimbabwe. JournalofDevelopmentStudies, 42(2), pp.301-321.
42. Holmes, R. & Jones, N. (2010) Rethinking Social Protection Using a Gender Lens. London: ODI.
43. Hosmer, D.W., Lemeshow, S. & Sturdivant, R.X. (2013) Applied Logistic Regression. 3rd edn. Hoboken, NJ: Wiley.
44. Howe, L.D., Hargreaves, J.R. & Huttly, S.R.A. (2012) Issues in the construction of wealth indices for the measurement of socio-economic position in low-income countries. Emerging ThemesinEpidemiology, 9(1), 3.
45. ICPSR. (2012) Guide to Social Science Data PreparationandArchiving. Ann Arbor, MI: Inter-university Consortium for Political and Social Research.
46. Iacus, S.M., King, G. & Porro, G. (2012) Causal inference without balance checking: Coarsened exact matching. PoliticalAnalysis, 20(1), pp.1-24.
47. Imbens, G.W. (2004) Nonparametric estimation of average treatment effects under exogeneity: A review. Review ofEconomicsandStatistics, 86(1), pp.4-29.
48. James, W. (1907) Pragmatism:A New Name for Some Old Ways of Thinking. New York: Longmans, Green, and Co.
49. Jahnke, H.E. (1982) Livestock Production Systems and Livestock Development in TropicalAfrica. Kiel: KielerWissenschaftsverlag Vauk.
50. J-PAL. (2024) Workfare and public works: Evidence review. Cambridge, MA: Abdul LatifJameel Poverty Action Lab.
51. Kabeer, N. (2008) Mainstreaming Gender in Social Protection for the Informa Economy.London: Commonwealth Secretariat.
52. Krueger, R.A. & Casey, M.A. (2015) Focus Groups:A Practical Guide for Applied Research. 5th edn. Thousand Oaks, CA: Sage.
53. Kvale, S. & Brinkmann, S. (2009) Interviews:Learning the Craftof Qualitative Research Interviewing. 2nd edn.ThousandOaks, CA: Sage.
54. Levine, S., et al. (2024) [Retrospective studies on asset quality and maintenance in public works.],journal.
55. Lincoln, Y.S. & Guba, E.G. (1985) NaturalisticInquiry. Beverly Hills, CA: Sage.
56. Little, R.J.A. & Rubin, D.B. (2002) Statistical Analysis with Missing Data. 2nd edn. Hoboken, NJ: Wiley.
57. Lohr, S.L. (2010) Sampling:DesignandAnalysis. 2nd edn. Boston, MA: Brooks/Cole.
58. Long, J.S. & Freese, J. (2014) Regression Models f or Categorical Dependent Variables UsingStata. 3rd edn. College Station, TX: Stata Press.
59. Lyberg, L.E. & Biemer, P.P. (2008) QualityAssuranceinSurveys. Hoboken, NJ: Wiley.
60. Maxwell, D. & Caldwell, R. (2008) The Coping Strategies index: A Tool for Rapidly MeasuringFoodSecurity Nairobi: CARE/WFP.
61. MLGRD (Ministry of Local Government and Rural Development, Zambia). (2024) Cash-for-workoperationalguidelines/policynote. Lusaka: MLGRD
62. Mullainathan, S. & Shafir, E. (2013) Scarcity: Why Having Too Little Means So Much. NewYork:Times Books/Holt.
63. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1979) The Belmont Report. Washington, DC: U.S. Government Printing Office.
64. NDMA (National Drought Management Authority, Kenya). (2024) HSNP Operations Manual/Case&GrievanceManagementGuidance. Nairobi: NDMA.
65. Ndoka, J. (2020) Gender and cash transfers in Wajir, Kenya. Policy note/report Nairobi.
66. Niemans, B. (2024) Site suitability mapping for road runoff harvesting. Technical note/report.
67. Nissen-Petersen, E. (2006) WaterfromRoads. Nairobi: ASAL Consultants.
68. OPM (Oxford Policy Management). (2018) Evaluation of the Hunger Safety Net Programme (HSNP) Phase 2. Oxford: OPM.
69. Patton, M.Q. (2015) Qualitative Research & Evaluation Methods. 4th edn. Thousand Oaks, CA: Sage.
70. Pawson, R. &Tilley, N. (1997) RealisticEvaluation. London: Sage.
71. Peirce, C.S. (1878) Howto make our ideas clear. PopularScienceMonthly, 12, pp.286302.
72. Platteau, J-P. (2004) Monitoring elite capture in community-driven development. DevelopmentandChange, 35(2), pp.223-246.
73. PMRC (Policy Monitoring and Research Centre, Zambia). (2017) Public works and ruraideveiopmentinZambia:Evidenceandlessons. Lusaka: PMRC.
74. Prior, L. (2003) UsingDocumentsinSocial Research. London: Sage.
75. PSNP PIM (Programme Implementation Manual). (2020s) Productive Safety Net Programme: Programme Implementation Manual. Addis Ababa: Government of Ethiopia.
76. Ravallion, M. (1999) Appraising workfare. World Bank Research Observer, 14(1), pp.31-48.
77. REACH. (2025) Regulation/Guidance for rural water services. Oxford: REACH Programme.
78. Rosenbaum, P.R. & Rubin, D.B. (1983) The central role of the propensity score in observational studies forcausal effects. Biometrika, 70(1), pp.41-55.
79. Rousseeuw, P.J. & Leroy, A.M. (2005) Robust Regression and Outlier Detection. Hoboken, NJ: Wiley.
80. Rubin, D.B. (1987) Multiple I mputation for Nonresponse i n Surveys. New York: Wiley.
81. Rubin, D.B. (2005) Causal inference using potential outcomes. Journal of the AmericanStatisticalAssociation, 100(469), pp.322-331.
82. Saldana, J. (2016) The Coding Manual for Qualitative Researchers. 3rd edn. London: Sage.
83. Sen, A. (1999) DevelopmentasFreedom. NewYork: Oxford University Press.
84. Shadish, W.R., Cook, T.D. & Campbell, D.T. (2002) Experimental and Quasi-Experimental Designs for Generalized Causa/ Inference. Boston, MA: Houghton Mifflin.
85. SPaN (Strengthening Planning and Delivery of Social Protection). (2019) Kenya drought case studies—predictable transfers vs. adhocrelief. London: DAI/SPaN.
86. Stake, R.E. (1995) The ArtofCaseStudyResearch. Thousand Oaks, CA: Sage.
87. Stodden, V., Leisch, F. & Peng, R.D. (2014) Implementing Reproducible Research. Boca Raton, FL: CRC Press.
88. Stuart, E.A. (2010) Matching methods for causal inference: A review and a look forward. StatisticalScience, 25(1), pp.1-21.
89. Swindale, A. & Bilinsky, P. (2006) Household Dietary Diversity Score (HDDS) for Measurementof HouseholdFood Access:IndicatorGuide. Washington, DC: FANTA.
90. Taylor, J.E., Filipski, M. et al. (2016) Local economy-wide impact evaluation of social cash transfers in Sub-Saharan Africa. FAO WorkingPaper Rome: FAO.
91. Thome, K., Taylor, J.E. et al. (2016) Local economy-wide impact evaluation for Malawi social cash transfer. FAO WorkingPaper Rome: FAO.
92. Thurow, L.C. (1974) Cash versus in-kind transfers. American Economic Review, 64(2), pp.190-195.
93. Tourangeau, R., Rips, L.J. & Rasinski, K. (2000) The Psychology of Survey Response. Cambridge: Cambridge University Press.
94. UK Data Service. (2015) Managing and Sharing Data: A Guide to Good Practice. Colchester: University of Essex.
95. UNICEF. (2023) Zambia Socia/Protection Budget Brief. Lusaka: UNICEF.
96. UNICEF Innocenti. (2024) Gender provisions and outcomes in PSNP. Innocenti Research Brief Florence: UNICEF.
97. VNG International. (2025) Social accountability tools in Ethiopia (ESAP). Programme brief/report.Vhe Hague: VNG International.
98. Vyas, S. & Kumaranayake, L. (2006) Constructing socio-economic status indices: How to use principal components analysis. Health Policyand Planning, 21(6), pp.459-468.
99. White, H. (1980) A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica, 48(4), pp.817-838.
100. WHO. (2011) Standards and Operational Guidance for Ethics Review of Health-Related Research with Human Participants. Geneva: World Health Organization.
101. WHO. (2016) Process of Translation and Adaptation of Instruments. Geneva: World Health Organization.
102. WFP. (2017) Comprehensive Food Security and Vulnerability Analysis (CFSVA) Guidelines. Rome: World Food Programme.
103. WFP. (2018) Low-cost water harvesting and soil conservation works: Field guidance forpublic works. Rome: World Food Programme.
104. Wooldridge, J.M. (2010) Econometric Analysis of Cross Section and Panel Data. 2nd edn. Cambridge, MA: MIT Press.
105. World Bank. (2013) Ethiopia's Productive Safety Net Programme (PSNP): Lessons andimpact. Washington, DC: World Bank.
106. World Bank. (2017) Public Works as a Social Protection Instrument: Systems, Assetsand Impacts. Washington, DC:World Bank.
107. World Bank. (2023) Building climate-resilient municipal infrastructure: A note for engineers.Washington, DC: World Bank.
108. World Bank. (2024) Cash-Plus in SocialProtection: What Works. Washington, DC: World Bank.
109. World Bank GRS. (2021) Grievance Redress Service: Guidance Note. Washington, DC: World Bank.
110. Youth Village Zambia. (2025) Payment delays and reforms in cash-for-work. News/feature article. Lusaka:
111. Yin, R.K. (2018) Case Study Research and Applications: Design and Methods. 6th edn. Thousand Oaks, CA: Sage.
APPENDICES
Appendix A: Participant Information & Consent (Interview Script & Checks)
A1. Information Script (read verbatim; no responses required)
1. Purpose of study and what participation involves (45-60 minutes household survey).
2. Voluntary participation;right to skip or withdraw anytime; no penalties.
3. Minimal risks;confidentiality;anonymised data use.
4. Data protection and storage;retention period;contacts.
5. Optional audio recording for quality.
A2. Consent Checks (record Yes/No for each)
1. Do you understand the purpose of the study? (Yes/No)
2. Do you understand that participation is voluntary? (Yes/No)
3. Do you consent to participate? (Yes/No)
4. Do you consent to anonymised use of your data for academic purposes? (Yes/No)
5. Do you consent to audio recording? (Yes/No)
A3. Sign-off (fill)
• Participant name:; Signature/Thumbprint: ; Date: //
• Enumerator name:; Signature:
Appendix B: Enumerator Ethics & Field Protocol Checklist (Question Prompts)
81. Pre-Interview Ethics
1. Did you ensure privacy for the interview? (Yes/No)
2. Did you deliver the full consent script and confirm understanding? (Yes/No)
3. Did you avoid collecting names or direct identifiers in open text fields? (Yes/No)
82. Interview Conduct
4. Did you maintain a neutral tone and avoid leading questions? (Yes/No)
5. Did you offer breaks or skipping sensitive questions when distress was observed? (Yes/No)
6. Did you follow skip patterns correctly? (Yes/No)
83. End-of-Day Quality Contro/
7. Were all completed forms uploaded and checked for missing values? (Yes/No)
8. Were outliers and inconsistencies flagged and reviewed? (Yes/No)
9. Were adverse events reported to the supervisor within 24 hours? (Yes/No)
Appendix C: Screening & Sampling Eligibility Questions
C1. Eligibility Screening (household level)
1. Is this household a usual resident in [Village/Ward]? (Yes/No)
2. Has any member participated in FFW in the last 24 months? (Yes/No)
3. Has any member participated in CFW in the last 24 months? (Yes/No)
4. If neither FFW nor CFW: Is the household aware of these programmes? (Yes/No)
5. Primary respondent aged >18 years? (Yes/No)
6. Able to provide informed consent? (Yes/No)
C2. Stratification Tag
7. Assign group: (1=FFW participant HH; 2=CFW participant HH;3 = Non-participant comparison).
C3. Replacement Rules
8. Has the household been visited >3 times at different times/days without success? (Yes/No)
9. IfYes, proceed to next pre-listed replacement household. (Yes/No)
Appendix D: Household Survey Questionnaire (Core Questions)
DO. Identification
1. District:; Ward/Zone:; Village:
2. Date(DD/MM/YYYY): / /
3. Enumerator ID:; HH ID:
4. Respondent name (initials only): ; Role in household (Head/Spouse/Other): __
5. Consent confirmed? (Yes/No)
Section A. Household Roster & Demographics
A1. List all usual members: name, sex (M/F), age (years), relationship to head, marital status, highest education, main activity.
A2. Does anyone have a functional difficulty (seeing, hearing, walking, remembering, self-care, communication)? (Yes/No; specify)
A3. Household size: ; Children <18: ; Elderly >65:
A4. Head's sex (M/F), age, education level:
Section B. Programme Exposure & Participation
81. Areyou aware of FFW/CFW programmes in/nearyour community? (Yes/No)
82. Has any member participated in FFW in last 24 months? (Yes/No)
83. Has any member participated in CFW in last 24 months? (Yes/No)
84. How were participants selected? (Self-targeting/Community committee/Programme staff/Traditional leader/Other)
85. Was the selection fair? (1=Veryfair... 5=Very unfair)
86. Were vulnerable households (female-headed, disabled, elderly, poorest) prioritised? (1=Always ... 5 = Never)
Section C Work Details (for last cycle; per participating mem bed
C1. Type of work (e.g., road maintenance, land rehabilitation, water infrastructure, public works):
C2. Duration (weeks), days/week, hours/day: / /
C3. Distance to work site (km):
C4. Safety measures (tools, PPE, breaks) adequate? (Yes/No)
C5. Childcare orflexible arrangements available? (Yes/No)
Section D. Transfers & Timeliness
D1. FFW: Type and quantity of food received per cycle:; frequency:
D2. CFW: Amount received per cycle (ZMW):; payment frequency:
D3. Were payments/rations on time? (1=Always ... 5 = Never)
D4. Preferred modality (FFW/CFW/Mixed/No preference) and reason:
D5.Who decides on the use oftransfers? (Head/Spouse/Joint/Other)
Section E. income, Livelihoods & Assets
E1. Income in last 30 days (ZMW): farming, livestock/fish, wage labour (non-programme), business/trading, social transfers, other.
E2. Income in last 12 months (ZMW): the same categories.
E3. Agricultural production (last season): main crops, area (ha), yield (kg), inputs used (seed/fertiliser/tools), source of inputs, extension visits (#).
E4. Assets owned (radio, phone, bicycle, solar, ox-plough, TV, housing materials).
E5. Livestockowned (cattle, goats, pigs, poultry): counts.
E6. Savings (ZMW): ; Member of savings group (Yes/No).
E7. Debt outstanding? (Yes/No); amount (ZMW): ; source (formal/informal/merchant/family/friends).
Section F. Expenditure (last 30 days)
F1. Food: ZMW; Health: ; Education: ; Transport: ; Energy/fuel: ; Farming inputs: ; Debt repayment: ; Other: .
Section G. Food Security (HDDS & FiES summary)
G1.In the last 24 hours, did your household consume: cereals;roots/tubers; vegetables; fruits; meat; eggs;fish; legumes;milk/dairy;oils/fats; sweets;
spices/condiments/beverages? (Yes/No for each;total 0-12)
G2. In the last 12 months, due to lack of money/resources, did you: worry about food; unable to eat nutritious food; eat few kinds of foods;skip a meal; eat less than should; run out of food; feel hungry but didn't eat; go a whole day without eating? (Yes/No for each; total 0-8)
Section H. Shocks, Coping & Resilience
H1. Shocks experienced (12 months) and severity (1-5): drought, flood/storms, pest/disease, price increases, illness/death, job/income loss, theft/conflict, other.
H2. Coping strategies used (tick all): reduced meals, borrowed food/money, sold assets/livestock, increased casual labour, received support (family/community/NGO/govt), other.
H3. Did FFW/CFW help you cope? (1 =Very much ... 4=Not at all); how?
Section I. Gender, Inclusion & Social Dynamics
11. Who decides major purchases? (Head/Spouse/Joint/Other)
12. Who controls programme transfers? (Head/Spouse/Joint/Other)
13. Time use yesterday (hours): paid work; unpaid care/domestic; farming; water/firewood; community work.
14. Did women have equal access to participate? (1=Always ... 5 = Never)
15. Were persons with disabilities accommodated? (1=Adequately ... 3 = Not at all)
Section J. Programme Satisfaction & Perceived Effectiveness
J1. Public works relevance, quality, maintenance, longevity(each 1-5).
J2. Overall satisfaction (1=Very satisfied ... 5=Very dissatisfied).
J3. Key strengths and areas for improvement (open text).
J4. Perceived outcomes (1-5): improved food security; increased income; increased savings/assets; reduced negative coping; improved communityassets/services.
Section K. WASH & Health
K1. Main water source (piped/borehole/protected well/unprotected well/stream/other).
K2. Sanitation facility (flush/VIP latrine/improved pit/unimproved pit/open defecation).
K3. Handwashing facilitywith soap available? (Yes/No)
K4. Any HH member ill in last 2 weeks? (Yes/No; specify).
Section L. information & Grievances
L1. How did you receive programme information? (meeting/radio/poster/phone/social media/peer/leader/other)
L2. Was information clear and timely? (1-5)
L3. Are you aware of the complaint process? (Yes/No)
L4. Have you used it? (Yes/No);outcome satisfactory (1-5)?
Closing
• Additionalcomments:
• Enumerator observations (not read):
Appendix E: Key Informant Interview (KU) Questions
E1. Role & Context
1. Please describe your role and involvement with FFW/CFW locally.
2. What are the programme objectives and target populations? How do these align with local needs?
E2. Targeting & Inclusion
3. How are beneficiaries identified and selected (criteria, community involvement, verification)?
4. What measures ensure inclusion (women, persons with disabilities, elderly, poorest)?
5. What complaints about targeting fairness have you encountered and how were they addressed?
E3. Work Design & Safety
6. Which public works have been implemented, and why were these chosen?
7. How were schedules, seasonality, and labour intensity determined?
8. What occupational safety measures are in place? Are childcare/flexible options provided?
E4. Transfers: Modality & Timeliness
9. How were food rations or cash amounts determined? Are they adequate given local prices?
10. Are payments/rations timely? What bottlenecks exist (procurement, finance, logistics)?
11. Which modality appears more effective (FFW vs CFW), and why?
E5. Monitoring, Evaluation & Grievances
12. Which indicators are tracked (participation, asset quality, household outcomes)?
13. How is data collected and validated (MIS, registers, community monitoring)?
14. Describe grievance mechanisms;common issues;resolution timelines.
E6. Outcomes, Sustainability & Risks
15. Observed changes in household food security, income, assets, and coping.
16. Qualityand sustainability of public works assets; maintenance arrangements.
17. Unintended consequences (market distortions, labour displacement, social tensions).
E7. Coordination & Policy Environment
18. How do programmes coordinate with local government and other schemes (SCT, FISP, extension, NGOs)?
19. Policy or budget constraints affecting scale/continuity.
E8. Lessons & Recommendations
20. What worked well? What should be improved (targeting, payment systems, asset maintenance, gender inclusion)?
21. If redesigning the programme, what would you change?
Appendix F: Focus Group Discussion (FGD) Questions - Participants
F1. Background & Participation
1. How did you learn about FFW/CFW? Why did you decide to participate?
2. Describe the selection process. Was it fair? Who was prioritised?
F2. Work & Transfer
3. What type of work did you do? Was the schedule manageable (distance, hours)?
4. Were payments or food rations adequate and timely?
5. How did your household use the transfers (food, health, school fees, inputs, savings)?
F3. Household Outcomes
6. Changes in food security (meals, diet), income, savings, or assets?
7. Did participation help you cope with shocks (drought, price increases)? How?
8. Did you reduce negative coping (selling assets, borrowing, skipping meals)?
F4. Gender & Inclusion
9. Were women able to participate equally? What barriers exist (care work, safety)?
10. Were persons with disabilities accommodated? What support was provided?
F5. Community Assets & Perceptions
11. Are the public works useful and maintained? Who is responsible for maintenance?
12. Overall satisfaction and priority improvements.
F6. Information & Complaints
13. Was programme information clear(eligibility, schedules, payments)?
14. Do you know how to make complaints? Were issues resolved?
F7. Closing
15. What advice would you give managers to improve FFW/CFW?
Appendix G: Focus Group Discussion (FGD) Questions - Non -Participan ts
G1. Awareness & Reasons for Non-Participation
1. Are you aware of FFW/CFW in your area?
2. Why did you not participate? (ineligibility, timing, childcare, distance, health, information gaps)
G2. Perceptions of Fairness & Inclusion
3. Was the selection process fair? Were vulnerable households prioritised?
4. Could women and persons with disabilities participate effectively under current arrangements?
G3. Community Outcomes
5. Effects on community well-being (assets, services, pricesjobs)?
6. Unintended effects (conflict, exclusion, market changes)?
G4. Suggestions
7. What changes would make you more likely to participate?
8. Recommendations to improve transparency, access, and asset quality.
Appendix H: Food Security & Coping Modules (HDDS, FIES, rCSI) - Question Lists
H1. Household Dietary Diversity Score (HDDS; 24-hour recall)
Ask: "In the last 24 hours, did anyone in your household consume..."
1. Cereals (maize, rice, sorghum, millet, bread) (Yes/No)
2. Roots/tubers (cassava, sweet potato) (Yes/No)
3. Vegetables (leafy greens, tomatoes, onions) (Yes/No)
4. Fruits (bananas, mangoes, citrus) (Yes/No)
5. Meat (beef, goat, pork, game) (Yes/No)
6. Eggs (Yes/No)
7. Fish (fresh/dried) (Yes/No)
8. Legumes/nuts (beans, groundnuts) (Yes/No)
9. Milk/dairy(Yes/No)
10. Oils/fats (Yes/No)
11. Sweets (sugar, sweets) (Yes/No)
12. Spices/condiments/beverages (tea, coffee) (Yes/No)
H2. Food Insecurity Experience Scale (FIES; 12-month recall)
Due to lack of money/resources, in the last 12 months, did you.
1. Worry about not having enough food? (Yes/No)
2. Unable to eat healthy/nutritious foods? (Yes/No)
3. Eat only a few kinds of foods? (Yes/No)
4. Skip a meal? (Yes/No)
5. Eat less than you thought you should? (Yes/No)
6. Run out of food? (Yes/No)
7. Behungrybutnot eat? (Yes/No)
8. Go a whole day without eating? (Yes/No)
H3.Reduced Coping Strategies index (rCSi; 7-day recall)
In the last 7 days, how many days did you...
1. Rely on less preferred or less expensive foods? (0-7)
2. Borrow food or money? (0-7)
3. Limit portion sizes at meals? (0-7)
4. Restrict adult consumption so children can eat? (0-7)
5. Reduce the number of meals eaten per day? (0-7)
Appendix I: Livelihoods, income & Expenditure Modules - Question Lists
11. Income Sources (30 days & 12 months)
1. Farming (crops) income (ZMW): last 30 days / last 12 months.
2. Livestock sales (ZMW): 30 days / 12 months.
3. Fisheries (ZMW): 30 days / 12 months.
4. Wage labour (non-programme) (ZMW): 30 days / 12 months.
5. Business/trading (ZMW): 30 days / 12 months.
6. Social transfers (SCT, remittances) (ZMW): 30 days / 12 months.
7. Other income sources: specify (ZMW): 30 days / 12 months.
12. Agricultural Production (last season
8. Main crops; area cultivated (ha);yields (kg).
9. Inputs used (seed, fertiliser, tools) and source (own/credit/programme/other).
10. Access to extension services (Yes/No); visits (#).
13. Expenditure Aggregate (30 days)
11. Food; Health; Education;Transport;Energy/fuel;Farming inputs; Debt repayment; Other (ZMW per category).
14. Assets, Savings & Debt
12. Durable assets owned: list and quantity.
13. Livestockowned: counts by type.
14. Savings in cash (ZMW) and membership in savings groups (Yes/No).
15. Outstanding debt? (Yes/No); amount (ZMW);source (formal/informal/merchant/family/friends).
Appendix J: Gender, Inclusion & Social Dynamics - Question Lists
J1. Decision-Making & Control
1. Who decides major household purchases? (Head/Spouse/Joint/Other)
2. Who controls use of programme transfers? (Head/Spouse/Joint/Other)
J2. Time Use (yesterday; hours)
3. Paidwork:
4. Unpaidcare/domestic:
5. Farming:
6. Water/firewood collection:
7. Communitywork:
J3. Inclusion & Accommodation
8. Did women have equal access to participate in FFW/CFW? (1=Always ... 5 = Never)
9. Were persons with disabilities accommodated? (1=Adequately ... 3 = Not at all)
10. Were childcare or flexible arrangements available for caregivers? (Yes/No; describe)
J4. Social Dynamics
11. Did the programme affect household or community relations (conflict/cooperation)? (Open text)
12. Were there any perceived social tensions related to targeting or participation? (Open text)
Appendix K: Distress, Risk & Safeguarding Screeners (interview Safety)
K1. Distress Scree ne r (during interview)
1. At any point, did the respondent appear distressed (tearful, anxious, withdrawn)? (Yes/No)
2. Did you offer a break or the option to skip/stop? (Yes/No)
3. Was a referral offered (if needed)? (Yes/No;specify service)
K2. Sensitive Topics Handling
4. Were sensitive questions asked privately, out of earshot of others? (Yes/No)
5. Did you avoid recording identifiable details in open text? (Yes/No)
K3. Safety & incident Reporting
6. Any incident requiring supervisor notification? (Yes/No; describe briefly)
7. Time incident reported to supervisor (HH:MM):
Appendix L: Community Asset Assessment (Public Works) - Question Checklist
L1. Asset Identification
1. Asset type (feeder road/small dam/irrigation canal/dip tank/soil conservation/borehole apron/other):
2. GPS coordinates: Lat Long ;Date visited://
3. Implementing agency (GRZ/NGO/Other):
L2. Relevance, Access & Use (1 -Very low ... 5-Very high)
4. Relevance to community priorities (1 -5)
5. Current utilisation rate (1-5)
6. Access and inclusivity (women, persons with disabilities, elderly) (1-5)
L3. Technical Quality & Maintenance
7. Construction quality (materials, workmanship) (1-5)
8. Functionality (operational status) (Yes/No; details)
9. Maintenance plan exists? (Yes/No); responsible entity:
10. Maintenance frequencyand budget adequacy (1-5)
11. Safety considerations (PPE signage, hazard mitigation) (1-5)
L4. Longevity & Externalities
12. Expected lifespan without major repairs (years):
13. Environmental safeguards (erosion control,watersafety) (1-5)
14. Positive spillovers (market access, time savings) (open text)
15. Negative externalities/unintended effects (open text)
L5. Community Feedback (Brief
16. Who uses the asset most?
17. What would improve its usefulness or durability?
Appendix M: Budget and Timeline
M1. Total Budget Summary
• Grand Total: K 30,000 (« USD 1,111.11 at 1 USD = K 27)
• Shared I Cross-cutting costs: K 9,000
• Field delivery (3 provinces): K 21,000
o Eastern: K 7,350
o Southern:K 6,930
o Central:K 6,720
M2. Shared/Cross-cutting Costs
Illustrations are not included in the reading sample
M3. Field Delivery by Province
Illustrations are not included in the reading sample
M4. Budget Breakdown Per Province
(i) Eastern Province (Total: K 7,350 | USD 272.22)
(Split evenly between Chipata and Katete: K3,675 each)
Illustrations are not included in the reading sample
(îî) Southern Province (Total: K 6,930 | USD 256.67)
(Split evenly between Chôma and Monze: K3,465 each)
Illustrations are not included in the reading sample
(iii) Central Province (Total: K 6,720 | USD 248.89)
(Split evenly between Kabwe and Serenje: K3,360 each)
Illustrations are not included in the reading sample
M5.Time/ine: December2025 -Apri!2026
Illustrations are not included in the reading sample
BIOGRAPHY
Maliro Ngoma is a visionary academic and development practitioner whose career seamlessly bridgees the disciplines of education, computer science, and development studies. His intellectual journey is a marked by a protound commitment to interdisciplinary learning, innovation, and transformative social impact.
He began his academic path with a Bachelor of Education from the Zambian Open University, established a strong foundation in pedagogy, curriculum design, and educational leadership. Motivated by a passion for technological advancement, he pursued a Master of Science in Computer Science at the prestigious Lo- monosoy Moscow State University in Russia, where he mastered advanced computing, systems analysis, software engineering and digital problem-solving. Recognizing the critical role of inclusive development and policy-driven change, Maliro expanded his expertise with a Master of Arts in Development Studies from the University of Lusaka. His work in this field explores the nexus of technology, education, and socio-economic development, with a focus on empowering communities through knowledge systems and innovation, Currently, Malito is a PhD candidate in Education Management and Administration at the Zambian Open unversity
[...]
Buy now
- Quote paper
- Maliro Ngoma (Author), 2026, Evaluating Food-for-Work and Cash-for-Work Interventions in Zambia, Munich, GRIN Verlag, https://www.grin.com/document/1698384
