Why do Farmers Join Payments for Environmental Services (PES) Schemes?

Table of Contents

Summary

Resumo

List of Abbreviations

List of Tables and Figures

Introduction

1. Problem Description and Context
1.1. Defining Payment for Environmental Services (PES)
1.1.1. Environmental or Ecosystem Services (ES)
1.1.2. Payment for Environmental Services (PES)
1.2. PES Context in Brazil
1.2.1. The PES boom in Brazil
1.2.2. Brazilian PES-Water policy

2. Literature Review
2.1. Criticism and Institutional Economics on PES
2.2. Participation in Payment for Environmental Services (PES)
2.3. Participation in other agri-environmental policies

3. Materials and Methods
3.1. PES Cases: Extrema, ProdutorES and Oásis Apucarana
3.1.1. Cases selection criteria
3.1.2. Basic social, economic and ecological characteristics of study areas
3.1.3. PES projects
3.2. Qualitative analysis
3.2.1. IoS Analytical Framework
3.2.2. Data sources for the qualitative analysis
3.3. Integrating qualitative and quantitative methods
3.4. Quantitative analysis
3.4.1. Selection and operationalization of variables
3.4.2. Latent variables
3.4.3. Opportunity cost estimations
3.4.4. Data sources for the quantitative analysis

4. Results
4.1. Descriptive statistics
4.2. Logistic regression model

5. Discussion
5.1. The IoS framework in Extrema, ProdutorES and Oásis Apucarana: some elements to analyse farmer’s behaviour
5.1.1. Properties of PES transaction
5.1.2. Actors analysis
5.1.3. Governance structure analysis
5.2. Quantitative results interpretation
5.2.1. Non-influential factors
5.2.2. Relationship between participation and income
5.2.3. The role of associations and membership
5.2.4. Gender and size: significantly different, but not explanatory variables
5.2.5. The four latent variables: access to information and environmental concern as key factors
5.2.6. The role of farm structures: family labour, slope, forest cover
5.2.7. Legal environmental restrictions and participation: APP and RLs
5.2.8. Opportunity costs
5.3. Conclusions

Bibliography

Annexes

Summary

Policy instruments for conservation and rural development based on Payments for Environmental Services (PES) have been increasingly applied by different actors responsible for water resources and environmental conservation in Brazil. Even though they might be considered relatively new policy instruments – first initiatives date from early 2000s – there is a growing momentum in many regions of the country for the use of these instruments. Usually, projects involve one of the three government levels (Federal, State-level or Municipal) in partnership with private sector and civil society organisations. Brazil has a decentralized water national policy, in which PES have been attracting unique interest from public managers.

This thesis analyses farmer’s decision to participate or refuse participation in PES projects related to water in Brazil. It focuses not only on farm or farmers characteristics, but also on the process of project implementation, particularly the institutional framework, governance structures and negotiation processes; and it is built on the assumption that both characteristics and processes are relevant to farmer decision.

Two main analyses were conducted. First, a qualitative analysis of PES-water programmes under the perspective of institutional economics (Vatn, 2010; Kosoy et al., 2008), which favours a more holistic reflexion and includes elements of context observation and facilitation processes, going beyond only procedural rules. For this, the Institutions of Sustainability (IoS) framework, proposed by Hagedorn (et al. 2002; 2008), was employed. This framework considers the analysis of properties of transactions, institutions and types of rules, governance structures and characteristics of actors, in order to assist a better understanding of the interdependencies between ecological and social systems.

For the second analysis, inputs generated by the qualitative exercise were used to develop a logistic regression model of revealed preferences applied to farmers that have decided to participate or refuse participation in the schemes. Variables from the following categories were tested: i) farmer and household characteristics, ii) farm structure and practices, iii) scheme factors.

This methodological structure was then applied to scrutinize three different PSE-water projects in Brazil: Municipal projects in Extrema, in Minas Gerais State, and Apucarana, in Paraná State, and the state-level programme of Espírito Santo State. Those were selected under the following criteria: i) active projects – which have farmers receiving payments for at least 1 year; ii) active participation – which could allow a representative sample for the econometric model; iii) different governance models – in order to compare institutional performance and innovation; iv) different farming regions – in order to allow for variance in the independent variables, favouring the identification of more reliable regression model estimates; and v) similar ecological objectives – watershed protection, restoration of riparian vegetation, and increase of the forest cover. Database consisted of 24 semi-structured interviews conducted with local environmental managers, and a survey applied to 163 farmers, distributed between participants and non-participants, and across the three regions.

Results indicate five main conclusions regarding farmer participation in PES-water projects arising from this study:

1) Participation and incomprehension of properties of transactions : it was found that some properties of transactions, in particular the relationship between riparian vegetation and water flow regulation, were not completely understood or recognized by many farmers. The non-recognition of these interactions undermines efforts to engage farmers proactively in PES projects, reduces trust-building processes and might act as a conflict generation factor in the medium term;

2) Farmers’ representative bodies and governance structures : the participation of a legitimate and representative organisation of farmers’ interests during all phases of project development and execution significantly reduces transaction costs associated with farmers’ engagement and subsequent project implementation. Risk of conflicts, suspiciousness and other entry barriers are particularly less severe when farmers’ bodies are part of the PES development process;

3) Access to information and general environmental concern : these two factors were found to be most important issues to influence farmers’ decision whether to participate or not in PES schemes in the three PES cases.

Effective communication strategies are remarkably important if a greater number of farmers are expected to participate in the projects. These communication strategies shall not only provide basic information about the project, but a comprehensive understanding of policy objectives, ecological interactions, operational procedures, etc. In some cases, establishing personal links between the rural community and project managers might prove useful to facilitate information sharing in participative social networks.

4) Farmer characteristics and farm structures : in line with previous studies that analysed participation in PES projects, some significant differences in participation levels were identified between groups of farmers. Logistic regression model results indicate that these are no general explanatory factors per se, but these highly context-based factors can play a role in excluding certain farmer categories, such as women and downstream farmers. One general factor identified in this study relates to family labour intensity, indicating that farmers that rely heavier on family labour are facing bigger constrains to adapt their production systems to participate in the PES programmes.

5) Opportunity costs : even though several assumption were made to estimate opportunity costs in this study, results are consistent with economic theory and cases analysed in the literature, indicating that farmers with higher opportunity costs participate less in PES projects. Particularly in the three studied cases, opportunity costs facilitate the understanding of several non-participant decisions.

These conclusions are expected to improve efficiency in PES-water policy development in Brazil. PES policy managers might be particularly attracted by some of the findings and might tailor their programmes aiming better participation levels.

Resumo

Instrumentos de conservação e desenvolvimento rural baseados em Pagamentos por Serviços Ambientais (PSA) têm sido crescentemente empregados por diferentes atores responsáveis pela conservação ambiental e dos recursos hídricos no Brasil. Muito embora estes instrumentos possam ser considerados relativamente novos – as primeiras iniciativas datam do início dos anos 2000 – existe um interesse crescente em muitas regiões do país para o seu uso. Na maior parte dos casos, os projetos envolvem um dos três níveis governamentais (Federal, Estadual ou Municipal) em parceria com o setor privado e organizações da sociedade civil. O Brasil possui uma política de recursos hídricos descentralizada, na qual os PSA têm atraído um interesse único dos formuladores de políticas públicas.

Esta tese analisa o processo de decisão de produtores rurais em participar ou não participar em projetos de PSA-água no Brasil. Enfoque é atribuído não apenas às suas características ou de suas propriedades, mas também ao processo de implementação do projeto, em especial a estrutura institucional, as estruturas de governança e os processos negociadores. A tese leva em consideração o pressuposto de que tanto características quanto processos são relevantes para o processo de decisão dos produtores.

Duas análises foram conduzidas. Primeiramente, uma análise qualitativa dos programas de PSA-água sob a perspectiva da economia institucional (Vatn, 2010; Kosoy et al., 2008), que favorece uma reflexão holística e inclui elementos de observação do contexto e processos facilitadores, indo além de apenas regras procedimentais. Para isto, foi empregada a estrutura metodológica Institutions of Sustainabiliy (IoS), proposta por Hagedorn (et al., 2002; 2008). Esta estrutura metodológica considera uma análise das propriedades das transações, instituições e tipos de regras, estruturas de governança e características dos atores, no sentido de facilitar um melhor entendimento das interdependências dos sistemas sociais e ecológicos.

Para a segunda análise, as contribuições geradas pelo exercício qualitativo foram utilizadas para o desenvolvimento de modelo de regressão logística sobre preferências relevadas, aplicado sobre os produtores que decidiram participar ou não participar destes esquemas de PSA. Variáveis das seguintes categorias foram testadas: i) características dos produtores e das propriedades, ii) práticas e estrutura das propriedades, iii) características dos esquemas de PSA.

Esta estrutura metodológica foi então aplicada para escrutinizar três diferentes esquemas de PSA-água no Brasil: os projetos municipais de Extrema, em Minas Gerais, e Apucarana, no Paraná, e o projeto estadual do Espírito Santo. Estes casos foram selecionados de acordo com os seguintes critérios: i) projetos ativos – na qual já existem produtores recebendo os pagamentos por no mínimo um ano; ii) participação ativa – que pudesse permitir uma amostra representativa para o modelo econométrico; iii) diferentes modelos de governança – que favorecesse a comparação das performances e inovações institucionais; iv) diferentes regiões agrícolas – que permitisse uma maior variância das variáveis independentes, o que favorece estimativas mais confiáveis do modelo de regressão; v) objetivos ecológicos similares – proteção das bacias hidrográficas, restauração da vegetação ripária e aumento da cobertura florestal. A base de dados consistiu em 24 entrevistas semi-estruturadas conduzidas com os gestores ambientais locais, e uma pesquisa aplicada a 163 produtores rurais, distribuídos entre participantes e não participantes nas três diferentes regiões.

Os resultados do estudo indicaram cinco principais conclusões com respeito ao processo decisório dos produtores rurais em programas de PSA-água:

1) Participação e incompreensão das propriedades das transações : foi encontrado que algumas propriedades das transações, em particular àquelas relacionadas com a vegetação ripária e a regulações dos fluxos hidrológicos, não são completamente compreendidas ou reconhecidas por muitos produtores rurais. O não reconhecimento destas interações enfraquece os esforços de engajar proativamente os produtores rurais nos projetos de PSA, dificulta os processos de construção de confiança e pode agir como potencial gerador de conflitos no médio prazo;

2) Organismos de representação dos produtores e estruturas de governança : a participação de organismos legítimos e representativos de interesse dos produtores rurais durante todas as fases de desenvolvimento e execução dos projetos reduz significativamente os custos de transação associados com o engajamento dos produtores e implementação do projeto. Riscos de conflito, desconfiança e outras barreiras de entrada são particularmente menos severas quando os órgãos de representação dos produtores são parte do desenvolvimento dos PSAs;

3) Acesso a informação e preocupação ambiental geral : estes dois fatores foram identificados como os mais influente no processo decisório dos produtores rurais em participar ou não dos esquemas de PSA nos três casos estudados.

Estratégias de comunicação efetivas são essencialmente importantes se é esperado que um grande número de produtores participe dos projetos. Estas estratégias de comunicação devem prover não apenas informações básicas sobre os projetos, mas um entendimento abrangente dos objetivos do projeto, interações ecológicas, procedimentos operacionais, etc. Em alguns casos, o estabelecimento de relações pessoais entre a comunidade dos produtores rurais e os gestores dos projetos pode ser uma estratégia eficiente para facilitar troca de informações e a construção de redes sociais participativas.

4) Características dos produtores e das propriedades : em consonância com estudos prévios que analisaram participação em projetos de PSA, algumas diferenças significativas foram encontradas entre grupos de produtores com respeito aos níveis de participação. Os resultados dos modelos de regressão logística indicam que estas diferenças não são fatores explicativos per se, porém estes fatores contextuais podem de fato excluir algumas categorias de produtores, como mulheres e produtores dos vales, em oposição aos das cabeceiras. Um fator geral identificado no estudo se refere à intensidade de trabalho familiar, indicando que produtores que dependem mais profundamente do trabalho familiar estão enfrentando maiores dificuldades em adaptar seus sistemas produtivos aos programas de PSA.

5) Custos de oportunidade : muito embora tenha sido necessário assumir alguns pressupostos no estudo para estimar os custos de oportunidade, os resultados encontrados são consistentes com o teoria econômica e com a literatura, indicando que produtores rurais com maiores custos de oportunidade participam menos dos projetos de PSA. Particularmente nos três casos estudados, os custos de oportunidade facilitam a compreensão de muitos casos de não participação.

Espera-se que estas conclusões aumentem a eficiência do desenvolvimento de políticas de PSA-água no Brasil. Os gestores de políticas de PSA podem ser particularmente atraídos por algumas destas conclusões, ajustando seus programas no intuito de alcançar maiores níveis de participação.

List of Abbreviations

illustration not visible in this excerpt

List of Tables and Figures

Table 1. PES projects in the Brazilian Atlantic Forest region

Table 2. Key factors driving PES participation in Lacandon rainforest, Mexico

Table 3. PES case selection criteria

Table 4: Four dimensions used to describe selected PES projects

Table 5. Sub-themes and main elements of qualitative interview guide

Table 6. List of variables considered in the statistical analysis

Table 7. Latent variables and constituent indicators

Table 8. Factor analysis for 12 indicators

Table 9. Separate factor analyses of constituent indicators for each latent variable

Table 10. Opportunity Cost estimations: formulas and sources

Table 11. Population and sample in PES cases

Table 12. Summary Statistics

Table 13. Comparison of scheme factors between 3 PES cases

Table 14. Model results for first variable selection

Table 15. Model results for only non-obliged farmers

Table 16. Model results for non-obliged farmers and relevant variables

Figure 1. Extrema, ProdutorES and Apucarana locations

Figure 2. Institutions of Sustainability (IoS) framework applied to the PES context

Figure 3. Integrating qualitative and quantitative approaches

Introduction

When interviewing a farmer during the data collection phase of this Master Thesis, this young researcher found himself puzzled by an exemplifying comment coming from one of the many dwellers he met in South-eastern Brazilian rural areas:

These sentences given by Seu Antônio when asked why he thinks his neighbour had decided not to participate in a Payment for Environmental Services (PES) project illustrate the problematic issue addressed in this thesis. Advocators of PES argue that this policy instrument has been gathering enormous attention in the developing world because it can promote both environment conservation and rural development simultaneously. According to this view, if incentives are given and the process is conducted properly, farmers would abandon some environmentally hazardous agricultural practices in exchange of compensations, resulting in positive sums for farmers and society.

Yet, we still find many situations such as described by Seu Antônio. Farmers in the same geographical location, with similar social and economic characteristics and same farm structures, technology and production systems, have ultimately taken different decisions. Why? Was the way the project was implemented relevant? Or is there any other hidden behavioural difference between Seu Antônio and his neighbour that explains this apparent contradiction? Maybe different personal attitudes toward environment or even officers did not have the chance to explain in details how the project works for Seu Antônio’s neighbour?

This thesis analyses farmer’s decision to participate or refuse participation in three PES project related to water in Brazil. It focuses not only on farm or farmers characteristics, but also on the process of project implementation, particularly the institutional framework, governance structures and negotiation processes; and it is built on the assumption that both characteristics and processes are relevant to farmer decision.

As a result, two main analyses were conducted. First, a qualitative-based analysis of the PES-water projects under the perspective of institutional economics, which favours a more holistic reflexion and includes elements of context observation and facilitation processes. Second, a logistic regression model of revealed preferences, applied to farmers that have decided to participate or refused participation in these projects, looking at characteristics of farm, farmers, schemes and other elements. Substantial effort was made to coherently integrate these two approaches into a single research framework, in order to reach more consistent conclusions.

The following structure organizes the research work. Chapter 1 defines Payment for Environmental Service and explores concepts and views of environmental or ecosystem services. The first chapter also sets the context in which this policy instrument is being applied in Brazil, particularly in the water sector, justifying the motivations behind this study.

Chapter 2 provides a literature review of the concept and summarizes the main discussions and critics to PES development, focusing on the institutional economics critique that guides part of the thesis. Relevant PES studies conducted in Latin American are briefly commented, given the importance of this policy instrument to some countries in the region. Chapter 2 also reviews studies which explored farmer’s participation in PES, both as main or secondary research objectives. The relatively scarcity of this studies also required a review of articles that addressed farmer’s participation in other agri-environmental policies.

The third chapter presents the three studied cases with its geographical and economic profiles. Chapter 3 also explains the research framework previously introduced and describes the Institutions of Sustainability (IoS) framework used in the qualitative-based analysis and the regression model specifications. Data collection methodology and procedures are also described in the third chapter.

Main results and findings are reported in the forth chapter. While procedures related to data processing are also reported in Chapter 4, data analysis, interpretation, and discussions are presented in Chapter 5. This last Chapter also finalises the thesis presenting the main conclusions reached with this study.

1. Problem Description and Context

Payment for Environmental Services has become a popular concept for many rural developers and environmental managers. During the last twenty years, a series of innovative projects have emerged in developing countries using an apparent simple rationale: providing economic incentives, instead of regulating agricultural activities, would be more effective and efficient in supporting the provision of environmental services by farmers.

However, just paying farmers in exchange of an intangible service might not be so simple or straightforward as it appears. Environmental services are the result of a number of ecological interactions, many of these just modestly understood by science, and the guarantee that a payment will secure or provide this service is subject to not only these underlying ecological relations, but also the social interactions inherent to the implementation of development projects.

Regarding the social context, one of the most important elements of a successful implementation relates to the participation of farmer’s in these programmes. Since the majority of these projects have been executed under a voluntary basis, low participation levels can undermine the achievement of project objectives, even if ecological interactions are sufficiently understood and economic incentives are put into practice.

This set of problems is explored in this chapter in two sections. The first one is dedicated to present the most accepted definitions of environmental or ecosystem services and the PES rationale. Section two sets the parameters for analyse PES within the Brazilian context and focus on water-related PES in the country.

1.1. Defining Payment for Environmental Services (PES)

In order to define payment for environmental services, it is necessary to understand what environmental or ecosystem services are, and how these services can be categorized.

1.1.1. Environmental or Ecosystem Services (ES)

Boyd and Banzhaf (2006) argue for the necessity of developing a standard definition for the term ecosystem services, stating that a uniform understanding of the concept is required for advancing environmental accounting. For the authors, ecosystem services are “components of nature, directly enjoyed, consumed, or used to yield human well-being” (idem, p.8). This precise definition would be important for developing standardized methods that integrates natural elements in well-being measures, such as green gross domestic product, but the definition focus only on the directly consumed components of nature.

A huge effort to account for the value of these services was conducted by a group of important ecologists and economists in mid-1990s, being published on Nature magazine (Contanza et al., 1997). This landmark article stated that ecosystem services were produced by the natural capital stock and that they contribute to the total economic value of the planet to similar proportions of the sum of all gross national products. An important advance of the understanding of these concepts was made by the article, when two important definitions, ecosystem functions and ecosystem services, were presented: “ecosystem functions refer variously to the habitat, biological or system properties or processes of ecosystems. Ecosystem goods (such as food) and services (such as waste assimilation) represent the benefits human population derive, directly or indirectly, from ecosystem functions” (idem, p.1).

However the most important contribution to the clarification and categorization of these concepts was the series of reports published by the Millennium Ecosystem Assessment (MEA, 2005). This international initiative of more than 1,300 scientists from different academic and geographical backgrounds under the call of UN Secretary-General and with institutional support from various international organisations helped to popularize this terminology on both academic and policy-making environments. The reports explored the relationship between ecosystem services and human well-being and defined an ecosystem as a “dynamic complex of plants, animal, and microorganism communities and the nonliving environment interacting as a functional unit” (idem, p. v). Furthermore, ecosystem services were defined as “the benefits people obtain from these ecosystems” and classified into four categories: supporting, provisioning, regulating and cultural services (idem, p. vi). According to these reports, provisioning services – such as food, water, fuel, fibres, raw materials and genetic resources – are related to the capacity of ecosystems in providing tangible goods, usually with structured markets in which society is used to perform transactions. For the other three categories, the relationship between ecological interactions and well being is less explicit, and transactions are rarely performed in structured markets. Regulating services are the benefits related to the natural processes that regulate the conditions which make human life possible, like climate, flood, disease regulations and water purification. Cultural services, on the other hand, are connected to human values, identity and behaviour. On the basis of all these services there are supporting services, necessary for the provision of other ecosystem services and to the functioning of ecological systems in the long term. Three important examples of supporting services highlighted by the MEA reports are nutrient cycling, soil formation and photosynthesis (idem, p. v-vi).

Common to emerging terminology, different words are being used by authors with diverse background to describe similar concepts. In literature, we come across studies that apply both the terms ecosystem services, usually found in ecological science and ecological economics literature, and environmental services, more frequently applied in environmental economics and policy publications (Jeanneaux & Aznar, 2010). In this thesis, these terms are used interchangeably and no distinction is made between them. Environmental Services (ES) is preferred in most cases since this is the term more frequently used by project managers involved in the three studied cases.

1.1.2. Payment for Environmental Services (PES)

Broadly spoken, Payment for Environmental Services are policy instruments that combine the transfer of monetary resources directly to farmers or other rural actors in exchange of a land-use or management practice that increases or secures the provision of an environmental service. PES-like policies have already been put into practice in a number of countries since 1980s – in developed countries usually under designation of agri-environmental measures (OECD, 2005) – even though lately scholars have made efforts to differentiate PES from other policy instruments.

While reviewing the literature discussions on ES both in economic theory and practice, Gómez-Baggethun et al. (2010) point that after the conceptual consolidation of ecosystem services in mainstream political and research agendas, there was an increasing interest in investigating market-based policy instruments that deal with these services. The authors classified these attempts under two main categories: i) Market for Ecosystem Services (MES), referring to emission trading schemes of greenhouse gases and other air and water pollutants; ii) and Payment for Ecosystem Services (PES) schemes.

Further differentiation is proposed by Broughton and Pirard (2011). These authors analysed different conservation schemes described by literature as market-based instruments (MBI), criticising that these instruments constitute an extremely heterogeneous group with less in common than usually regarded. Broughton and Pirard proposed six categories of MBI, in which PES would be a category defined as Coasean type agreements in line with the most popular definition, given by Wunder (2005).

According to this definition (Wunder, 2005), Payments for Environmental Services can be described by five basic criteria:

“(i) a voluntary transaction, where (ii) a well defined ES or a land-use likely to secure that service (iii) is being ‘bought’ by a (minimum one) ES buyer (iv) from a (minimum one) ES provider (v) if and only if the ES provider secures ES provision (conditionality).” (idem, p. 9).

Wunder’s definition introduces some very important elements of differentiation of PES projects from other market-based policy instruments. The reference to the voluntarism emphasizes the existence of options from the provider’s point of view and differentiates PES from regulatory policies in which compliance is mandatory. The criteria of a well definition of the ES signifies that both parties have knowledge on the service which is being traded, even in the case where it is not the ES itself, but a land-use likely to secure the ES, that is subject to a transaction. The existence of buyers and providers is one very important element of the definition, indicating that in PES policies the parties engaged are more specifically defined than other public policies. Finally, the conditionality element strengthens the characteristics of transaction in PES, by stating that buyers are paying for a service that must be delivered by providers.

Even though Wunder’s definition has been accepted by many scholars in the field as a guiding reference for commencing a PES analysis, a number of projects that do not gather all the five criteria are also in practice referred as PES. As pointed out by Veiga, PES that follow the five criteria suggested by Wunder are sometimes referred as “pure” PES, given that they represent a model of governance arrangement which can be hard to be found in the field (Veiga, 2010, p.10). Engel et al. (2008) present an overview of main issues affecting PES evolvement, describing PES characteristics and comparing this tool with other policy instruments. They acknowledge Wunder’s contribution in defining PES, but also recognized that many of the PES-like projects do not reach all the five criteria, even though following the basic rationale of the instrument. Broughton and Pirard (2011) discuss the same problem, stating that PES can be seen both as a principle and as a policy instrument. The former case relates to Wunder’s five criteria and represents Coasean type agreements in which the two bargaining parties spontaneously reach a mutually beneficial outcome. However, other policies that consist of valuing and trading ecosystem services might also be called PES, even if they operate in a substantially different manner – for example, reverse auctions for conservation. In fact, the three cases examined in this thesis lack some elements of the Wunder’s definition of PES as well.

A review of 287 PES projects done by Landell-Mills and Porras (2002) identified four categories of environmental services that have been presenting more frequent transactions or promising markets under PES-schemes: i) carbon sequestration – usually quantity of carbon sequestered or non-emitted; ii) biodiversity conservation – typically number of species or natural habitat area under protection; iii) landscape beauty – with payments frequently related to touristic services; and iv) watershed protection – connected to the maintenance or increase of water quality and quantity. PES-schemes that combine two or more of these categories of environmental services are also recurrent, called by those authors as bundle services (idem, p.4). As it will be better explained in the results chapter, the three PES-projects analysed in this thesis deal with the watershed protection service.

From an economic theory perspective, two arguments explain market failures in ecosystem services markets and justify the existence of PES instruments in environmental management. Landell-Mills and Porras declare that “many environmental services provided by forest fall into the category of positive externalities or public goods” (idem, p.7). As explained by Cornes and Sandler (1996), the imperfect assignment or even absence of property rights might implicate an insufficient level of payments that would compensate for produced positive externalities. Similarly, when it comes to public goods, defined by their non-excludability and non-rivalry, markets typically fail to provide a socially optimum provision due to the fact that users of these goods have no incentives to pay for suppliers.

For many authors, besides other policy tools, PES seem to provide solutions to these market failures under certain situations. Engel et al. (2008) underline that PES might be a useful policy instrument, but to a narrow set of problems. They comment on some examples of situations where other solutions are more plausible, such as cases with information problems, capital market imperfections and when collective action seems more appropriate (Bulte & Engel, 2006; Engel, 2007; Ostrom, 2003). For Engel et al., the suitability of PES is better identified in those cases where “ecosystems are mismanaged because many of their benefits are externalities from the perspective of ecosystem managers” (2008, p.665).

Pagiola et al. (2010) also recognize that PES are one of the many tools to address environmental problems, but they comment three reasons that make PES particularly attractive: i) it can be implemented as a development program that has the ability of generating its own finance; ii) it can be economically efficient, since efforts can be focused where benefits of conservation are highest and costs are lower; and iii) it can be more sustainable in the long-term, since it relies more on self-interested users than on external supporters such as governments, NGOs or donors.

The prudence regarding the scope and applicability of this policy called by some important authors in the field did not inhibit PES to become a fast-growing policy instrument in many developing countries. An online inventory maintained by the Organization of American States (OAS) counted more than 350 projects in Latin America and the Caribbean alone by June 2011 (OAS, 2011) and these figures might be enormously underestimated^[1]. Indeed, the fast popularization of the concept raised concern from part of the academic community, which recalls attention that PES policies should not be considered a one-size-fits-all approach to solve environmental problems. A more detailed review of critic literature of PES theory and practice is presented in the literature review chapter.

1.2. PES Context in Brazil

Payment for Environmental Services are a rather new policy in Brazil. First initiatives date from the early 2000s, but there is a growing momentum for its development in many regions of the country by several public and private actors, even though a geographical concentration on Center-South region is noticeable. Among the main four categories delineated of environmental services by Landell-Mills and Porras (2002), carbon sequestration and watersheed protection seem to dominate. The country has a decentralized PES-water policy in place, which requires a high level of local implementation and involvelment of a diverse set of public and private actors.

1.2.1. The PES boom in Brazil

Although successfully implemented in other Latin-American countries in the 90s, PES were relatively unknown in Brazil until early 2000s. Pagiola (2011) affirms that some early initiatives related to forestry carbon sequestration were implemented in the beginning of the decade, but it was somewhere in the middle of 2000s that a number of initiatives started to take place. One milestone is the project “Conservador das Águas” from the municipality of Extrema (Pereira et al., 2010), analysed in this thesis. This PES-water project received significant attention from specialized and general media, and influenced a number of other initiatives around the country. In such dynamic policy environment, it is difficult to estimate the total number of PES projects in the whole Brazilian territory, but this number would be situated between 30 and 50.

A seminar hold in May 2011 by the State-level Environmental Agency of São Paulo to discuss PES invited 29 initiatives. An important effort to account for PES projects found 25 projects already in implementation in the Atlantic Forest region alone (Guedes & Seehusen, 2011). These authors organized a broad investigation of carbon, water and biodiversity PES projects in different levels of development, as represented in Table 1.

Table 1. PES projects in the Brazilian Atlantic Forest region

illustration not visible in this excerpt

Source: Adapted from Guedes and Seehusen (2011).

As can be seen by the figures identified by Guedes and Seehusen, carbon sequestration and water resources protection have been dominating PES project development in Brazil. Watershed protection schemes are particularly promising given their inherent local characteristics of design and implementation, which allows an easier approximation of producers and users. Some examples of services under the water category are flood control, water flow stabilization between wet and dry seasons, reduction of soil erosion and its sedimentation impacts on water quality, riparian habitat conservation and soil contamination control. Depending on the local characteristics of the site and the type of environmental hazard, these environmental services can be maintained or improved through activities supported by PES projects. Examples of these are reforestation of riparian vegetation, improvement of watershed management, establishment of buffer zones and protected areas along rivers and streams, better soil, chemicals or fertilizers management, among others (idem, p.40).

1.2.2. Brazilian PES-Water policy

Decentralization is a fundamental element of the Brazilian PES-Water policy. The Brazilian Water Agency (Agência Nacional de Águas – ANA) has a national programme named “Water Producers”, which supports, orientates and certifies PES-water initiatives. This agency provides technical and financial support to development of some phases of those projects (ANA, 2011), but the effective implementation of the PES instrument is local.

In fact, several actors from public, private and third sectors are engaged in developing PES-water in the country. Gavaldão and Veiga (2011) identified a diverse set of organisations that are coordinating these projects in the Atlantic Forest region, with a diverse distribution of responsibilities and actions. Considering only the 8 PES-water projects already under execution, 11 local administrations, 6 State-level agencies, 5 private sector organisations and 8 NGOs are directly involved in project implementation activities. In this context, a diverse set of governance structures arise and coordination between organisations is essential to the success of these PES initiatives. For the three cases examined by this thesis, these elements are discussed in details in the results and analysis sections.

2. Literature Review

An impressive literature on PES in developing countries was published during the last few years, which illustrates the dynamism of this policy instrument in different contexts. Particularly in Latin American, PES were substantially explored both in policy and academic environment, and a substantial number of papers reviewed the PES rationale and implementation in several countries of the region (Pagiola et al., 2002; Rojas & Aylward, 2003; Albán & Argüello, 2004; Rosa et al., 2004; Pagiola et al., 2005a; Robertson & Wunder, 2005; Pagiola et al., 2007; Locatelli et al., 2008; Kosoy et al., 2008; Wunder & Albán, 2008; Wunder et al., 2008; Pagiola et al., 2011).

More recently, some authors started to introduce more criticism to the concrete implementation of this policy instrument, looking at components such as transactions costs, negotiation processes, information asymmetries and power relations, contract design, among others. A significant part of these elements were also explored by institutional economics in other contexts, and to review PES under this perspective is important due to the fact that many factors explored by these analyses may influence farmer’s decision in taking part of PES schemes.

Albeit the substantial number of articles about PES, only a few of these address the issue of participation, with even fewer articles directly discussing reasons for farmer participation as the main research objective. The relatively scarcity of this specific subject on the literature contrasts with a number of articles and models that explored participation in other agri-environmental schemes, particularly those in developed countries. This chapter reviews this literature, given the progress to identify important variables to model farmer’s decision on scheme participation.

2.1. Criticism and Institutional Economics on PES

Wunder (2006) noted that an early critique to the implementation of PES projects in developing countries was in place even before this policy instruments was effectively applied in practice. He explains that main concerns with this market approach to conservation are related to: i) asymmetric power distribution, which could set powerful conservation consortia organisations threatening weak property rights hold by powerless forest communities; ii) commoditization of conservation, which could erode communitarian non commercial conservation values into a market logic of buying and selling conservation.

This critique was discussed by Karsenty (2004) who gloomily analysed market-based conservation instruments under three criteria: environmental and economic efficacy, equity impacts and legitimacy. Due to this author, the environmental and economic efficacy of these would depend on the existence of well defined private property rights on land and actors with a “commercial logic”, in other words, actors used to operate in markets. These conditions might be found in some regions of Latin America, but for Karsenty, the existence of different communal and semi-private systems of land property rights in some African and Asian regions would pose significant problems to the commercialization of these rights. Furthermore, even in cases where enough clarity about land rights would be feasible for PES development, asymmetric power distributions between rural communities and conservation organisations could undermine a fair and legitimate negotiation of payments, contracts, responsibilities, etc. These arguments proposed by Karsenty go in line with the criticism that communities that were not used to deal with powerful agents in commercial terms could have their non-for-profit conservation values substituted for a “pure instrumental logic”, which in the end would reduce total conservation outcomes (Vatn, 2010, p.1251).

Similar concerns with the equity and legitimacy dimensions of PES were developed by Corbera et al. (2006; 2007), which analysed four PES cases in Guatemala, Nicaragua, Belize and Mexico focusing on the social development goals claimed by these projects. These authors found that existing power structures and its consequently inequalities and vulnerabilities were reinforced by the design and implementation of PES projects. In practical terms, powerless groups with limited access rights to forestry and other natural resources, such as poor farmers and women, were excluded from the negotiations and designs of these policies, allowing the authors to claim that PES might not be adequately flexible to account for specific context-related factors.

Furthermore, Kosoy and Corbera (2010) apply the concept of “commodity fetishism” to ES markets, in order to explain how unequal power relations and asymmetries might be reinforced by PES policies. To these authors, the transformation of ecosystem services into objects or rights that can be traded follows a process of commoditization, divided in three phases: i) a complex multi-functional ecological process is summarized into one single ecosystem service; ii) this ES is valued; iii) buyers and providers are linked, either directly through markets, either indirectly by the intermediation of public organisations. In the context of PES, the classical Marxist terminology of commodity fetishism refers to this the process of “masking the social relations embedded in the process of ‘producing’ and ‘selling’ ecosystem services” (idem, p.1229). Thus, for these authors, the inconsideration of these social relations reproduces power inequalities in PES development, with important equity implications.

Asymmetric power relations and the governance structures that arise during PES implementation became an important subject of analysis in institutional economics. Vatn (2010) criticizes the clear-cut definition of PES as market-based instruments. He discusses three ideal types – in the Weberian sense – of governance structures: i) hierarchies, ii) markets and iii) community management. But he points out that instead of independently existing, in the real world, these three governance structure co-exist, “operating together; partly they depend on each other, partly they may act as supplements or compete” (idem, p.1246). Therefore, for Vatn, the definition of PES as market instruments is imprecise, given the strong role of intermediaries in the process of PES implementation. Existing high transaction costs in the incipient PES market may be reduced by hierarchical or community structures and social relations are important elements of these governance structures. Thus, an institutional analysis of PES would have to consider factors such as transaction costs, governance structures, power relations, contract design, negotiating processes, among others, if suitability and social performance of PES matter.

In line with Vatn’s argument that a substantial number of PES cases requires government or community facilitation, Muradian et al. (2010) developed some further elements in institutional and political economy issues that have been scarcely noted by PES literature. The authors reviewed that for a wide variety of PES cases, important context and social factors determine schemes’ characteristics which depart those cases from the theoretical Coasean approach that dominates PES definitions. Elements such as volunteerism, well definition of ES and conditionality are most of times constrained by the uncertainties or the political context in which this instrument operates, as these factors must not be neglected in PES analysis. Specially envisaging PES practitioners, they develop an alternative conceptual framework that do not consider PES necessarily as market instruments, but as “transfer of resources between social actors, which aims to create incentives to align individual and/or collective land use decisions with the social interest in the management of natural resources” (idem, p. 1205). Furthermore, they acknowledge the complex set of social interactions in which PES policies are embedded and proposed to cluster PES schemes according to: i) the importance of the economic or financial incentive, ii) degree of ES commodification, and iii) transfer directness, in other words, the degree in which buyers and providers are intermediated by external actors. To these authors, this conceptual framework is more flexible and practical to analyse PES and their heterogeneous social contexts and institutional arrangements.

Finally, one practical example of this institutional perspective on PES is given by Corbera et al. (2009), which analysed design, performance and institutional interplay of Mexico’s carbon forestry programme. The authors reaffirm some likely contributions from an institutional perspective on PES, such as the identification of social tensions, controversies around PES designs, participation and exclusion factors, and envisaged economic and social impacts. One particular element highlighted by Corbera et al. refers to the way PES influence environmental management practices and conservation values.

In line with this institutional economics approach to PES analysis, this thesis explores some of these elements that might influence farmer’s participation in the schemes in the analysis and discussion section.

2.2. Participation in Payment for Environmental Services (PES)

Although substantial literature have been published about PES, few authors addressed farmer’s participation on its analysis, and even less literature dealt with the topic as the primary research objective. According to Kosoy et al. (2008), one possible explanation for the scarce attention to this topic in literature relates to the top-down approaches in project development identified in many PES projects. In fact, the prevailing view is that issue of raising participation levels is “a function of setting the right procedural framework” (idem, p. 2074). Thus, to these authors, the general critique to the failure of top-down and technocratic approaches in project development did not take place in PES literature so far.

More specifically, the initial concern with farmer participation was directed in how to engage poor farmers in PES schemes, given the promising social and economic benefits envisaged by PES promoters (Pagiola et al., 2010, p. 373). To respond this question, Miranda et el. (2003) noted that in Costa Rica – which has one of the most promising national-wide PES schemes – initially poor farmers had faced significant concern in accessing the program. To participate in some activities of the Costa Rican scheme, such as reforestation, farmers had to co-finance part of the investment, and poor farmers faced challenges in providing collaterals given that forestry was not considered an eligible activity for funding. On the other hand, but addressing the same question, mix results were found by Grieg-Gran et al. (2005) after reviewing a series of PES schemes in Latin America. In some cases, poor households were discriminated because they lack formal land titles necessary for contract arrangements. In other cases, poor communities benefit more because the PES scheme itself established this group as a preferential target.

Wunder (2005) noted similar conclusions, stressing that PES scheme characteristics can act both as positive or negative discriminators of poor farmer’s participation. He highlights two main structural limitations to the participation of less favoured groups: i) lack of land property rights (either formal or informal), ii) the high transactions costs associated with negotiating with multiple providers. It seems logical that a project manager that wishes to maximize the environmental benefits with constrained budget prefers to deal with bigger landowners than with multiple small farmers, since transactions costs are lower. The author also noticed that in this context of lowering transaction costs, communities are more likely to offer community-based negotiations, where many farmers might be represented. Wunder also found that marginal land is more likely to be enrolled than productive land, another logical conclusion given land opportunity costs associated with PES participation.

One important contribution to the schematization of possible factors that drive or constrain farmer’s participation in PES was introduced by Pagiola et al. (2005b). The authors separated factors that might affect participation in three categories: i) eligibility factors (can be selected to participate), ii) desirability factors (want to participate), and iii) ability factors (can participate). Moreover, to these authors, these categories operate in a logical sequence, even though each factor inside a category would operate independently. In other words, the participation processes would follow a chronologically sequence, first projects managers would target their policies to a specific group, region or watershed. Later farmers would demonstrate their interests in joining the PES programme or refuse it. Finally, those farmers that would have the means to participate would sign the contracts or apply the practices necessary for receiving the payments. In fact, within each category proposed by these authors, a numbers of factors could determine final participation, such as PES scheme characteristics (PES targeting rules, payments amounts, etc), household characteristics (opportunity costs, farm locations, slope, fitness of PES activities to the current farming system, among others) or farmer’s characteristics (education levels, access to credit and technical assistance, experience, etc.).

An attempt to examine these factors in one applied case was recently published by Pagiola et al. (2010), which analysed farmer’s participation, particularly poor household participation, in a silvopastoral PES project in Colombia. On respect to three categories previously introduced, these authors focused on the ability factors, more specifically, if eligible poor household were able to participate at the same levels of better off households. One interesting characteristic of the studied Colombian PES project was that there was a broad range of options to participate; payments were not fix or only a function of the area contracted, but subject to several different silvopastoral activities that could sum “environmental points” that were converted to payment amounts. This allowed the researchers to going beyond a binary analysis of participants and non-participants, looking also at “the intensity of participation” (idem, p. 372). In a similar way as proposed by this thesis, the authors applied an econometric analysis to some possible relevant variables in technological and agroforestry adoption literature (Feder et al., 1985; Pattanayak et al., 2003; Mercer, 2004). One important finding was that “smaller and more remote farms participated at lower levels than other households” (Pagiola et al. 2010, p.391). Regarding the main research question, the authors found that poor households indeed participated on the same intensity level as other households, and were not only restricted to the less complex and less costly participation options.

Another statistical and econometric analysis was conducted by Zbinden and Lee (2004) to assess participation in the Costa Rica’s PES programme. Using a logistic regression model applied to a list of possible explanatory variables, the authors concluded that larger, better educated, better informed, and wealthier farmers were more likely to participate in the programme. They suggested that three main features determined participation: “i) farm size; ii) human capital and household economic factors, and iii) information”, particularly access to information about the scheme (idem, p. 269).

Arriagada et al. (2009) proposed a research framework that integrated qualitative and quantitative techniques to investigate farmer’s participation in the Costa Rica’s PES programme. Using a similar approach as the one applied in this thesis, the authors examined one particular relevant region through case studies and more in depth interviews, in order to previously identify possible variables that influence participation. These variables were lather incorporated into a survey and an econometric analysis, that tested the hypothesis that high land opportunity costs were significant constrains to farmer’s scheme participation. The innovative combination of exploratory qualitative and explanatory quantitative methods allowed the authors to robustly confirm the hypothesis, concluding that marginal lands were more prone to PES enrolment. Other important driving factors identified by the study – which nevertheless were not tested by the econometric analysis – were: higher importance of off-farm income, more secure tenure of land, and higher farmer’s environmental awareness.

A more holistic approach was recommended by Kosoy et al. (2008), when investigating participation of four Mexican communities in PES carbon and biodiversity schemes. In line with the institutional critics to PES analysis, these authors suggested a multi-scale approach to more precisely understand how farmers are engaged in PES policies, putting bigger efforts in understanding the complex social and political environments in which these instruments are implemented. The three-level analytical framework are composed of: i) a rules-based approach, “which considers that high participation levels are a product of cautious thought at the policy design stage”; ii) a process-based approach, which affirms that participation results from a process of “coevolution between” relevant actors; and iii) a behavioural approach, that takes farmers as independent environmental managers and focus on their personal and farm characteristics (idem, p. 2075). The authors concluded that critical factors related to community, farmer and procedural levels drive participation, as reproduced in Table 2.

Table 2. Key factors driving PES participation in Lacandon rainforest, Mexico

illustration not visible in this excerpt

Source: reproduced from Kosoy et al. (2008, p. 2081)

2.3. Participation in other agri-environmental policies

In contrast with the relatively scarce publications about farmer’s participation in PES schemes, literature related to participation in other agri-environmental policies is vast. This specific topic is here very briefly commented, given that the list of independent variables used in the statistical analysis was complemented and shaped by reviewing this literature.

Early in 1989, Brotherton suggested that both farmer’s attitudes and scheme specificities could drive farmer’s participation in the land diversions schemes that would later become part of the European agri-environmental programme (Brotherton, 1989). This approach, which was later called “behavioural approach”, was tested by a number of studies in different contexts. Wilson (1996; 1997) used this approach to identify factors driving farmer’s participation in an Environmentally Sensitive Area (ESA) programme in England. Falconer (2000) complemented the behavioural approach with elements of transaction cost theory to investigate farmer’s participation in Agri-environmental Measures (AEMs) in eleven regions in eight different European countries. Edwards-Jones (2006) published a review of modelling farmer decision-making literature, grouping the numerous factors into six categories: i) socio-demographic elements, ii) psychological factors, iii) household characteristics, iv) farm business characteristics, v) social context factors, and vi) scheme or innovation characteristics (idem, p.783). A similar review of 23 articles regarding conservation agriculture was published by Knowler and Bradshaw (2007). They also proposed to group factors into categories, in this case, four headings: i) farm and household elements, ii) farm biophysical characteristics, iii) farm financial or management factors, and iv) exogenous factors (idem, p.25).

An alternative to the process of previously defining set variables for later testing with statistical techniques and regression models was proposed by Greiner et al. (2009). They analysed adoption of best management practices (BMPs) by a group of graziers in north-eastern Australia using an open-question survey, focusing on the motivation and risk behaviour to the adoption of these conservation practices. Thus, no pre-defined variables were suggested, but factors were identified and later reduced using Principal Component Analysis (PCA). These were consequently loaded in factor analysis, searching for factor solutions. This combination of methods allowed researchers to explore motivations, attitudes, risk behaviour and other complex psychological factors that might be tricky to model using regression analysis.

Farmer’s attitudes and beliefs towards conservation were also considered in two studies, by Defrancesco et al. (2007) that looked at AEMs participation in Italy, and Toma and Mathijs (2007), which explored participation of Romanian farmers in one organic agriculture programme. In the Italian case, two multinomial models were developed, one to assess the probability participation and another to investigate more deeply the effect of farmer’s attitudes on decision-making. These were tested to a number of independent variables, some of those related to attitudes, subjective norms and past behaviour in other conservation practices. Farmers were also clustered into four groups: resistant non-adopters, conditional non-adopters, passive adopters and active adopters (Defrancesco et al., 2007, p. 125).

Additionally, Toma and Mathijs employed structural modelling equation with latent variables to model adoption of organic agriculture. The model included six latent variables measured by 17 indicators, which, besides socio-economic factors and farm characteristics, also considered two attitude latent variables – environmental risk perception, general environmental concern – with confirmed validity by the use of factor analysis (Toma & Mathijs, 2007, p.148). As will be explained in more detail in Chapter 3, this thesis employs a similar technique to the development and validation of latent variables with respect to farmers’ environmental attitudes.

3. Materials and Methods

As described in the literature review, a variety of methods were applied to investigate farmer’s participation in PES and AEMs. Some studies relied on statistical and econometric analysis, applied to a given set of variables selected through theory and literature review. Other studies complemented this approach developing exploratory case studies, in order to generate a more profound understanding of the underlying social and political context in which PES are implemented.

This thesis goes in the same direction. Two main analyses were conducted: a qualitative analysis, conducted with the assistance of Hagendorn’s framework Institutions of Sustainability (IoS), and a quantitative analysis, in which a logistic regression model was developed. Substantial effort was employed to integrate these two approaches under a single research framework, in order to generate more consistent results and coherent interpretations.

This chapter starts by describing the selected cases. Case selection criteria with basic social, economic and ecological characteristics, and a description of PES projects in the three areas are presented. Afterwards, the IoS framework and its theoretical foundation are described, as well as the data sources used to generate the information subject to this analysis. It describes how qualitative and quantitative approaches are integrated at the methodological and practical level. Subsequently, variables selection and operationalization are explained, including the definition of the latent variables and the opportunity cost estimations. The chapter finalizes presenting quantitative data sources, survey and sample selection methods.

3.1. PES Cases: Extrema, ProdutorES and Oásis Apucarana

This sub-section introduces the three Brazilian PES-water cases analysed in this study. Reasons for the selection of these three specific cases are explained according to selection criteria. Elementary social, economic and ecological characteristics are described, along with the main characteristics of the PES projects. This basic description establishes the information background in which the analysis is later developed.

3.1.1. Cases selection criteria

A recent study published by the Brazilian Ministry of Environment identified eight PES-water project in Brazil already under implementation in 2010 (Gavaldão & Veiga, 2011). The authors considered as “under implementation” those projects that had contracted farmers receiving payments, where conservation or restoration activities were already being implemented, and land use changes were being monitored. Considering these eight projects, the potential area under some conservation or restoration activity totalled approximately 13 thousand hectares, 345 farmers were defined as ES providers, and roughly 22 million people were considered potential ES users (idem, p.135). Interviews conducted with the authors revealed that these numbers are already outdated, given the dynamic environment in which PES-water policies in Brazil are being developed. But due to the inexistence of more up-to-date published figures on the topic, they formed the basis for case selection in this thesis.

Five selection criteria are proposed:

i. Active projects: this criterion is similar to the “under implementation” classification used by Gavaldão and Veiga. Only projects with farmers already contracted and receiving payments for at least one year were selected;

ii. Active participation: a reasonable number of farmers participating in a given project would be required to constitute a representative sample size for the quantitative analysis. Therefore, restriction was imposed to selected those projects with at least 60 farmers participating;

iii. Different governance structures: in order to compare institutional performance and innovations – an important element of analysis in the IoS framework – different governance structures were preferred. The three selected cases differ in their governing bodies, coordinating organisations, sources of funding, monitoring systems, etc;

iv. Different farming regions: this criterion is important to allow higher variance in the independent variables used in the regression model;

v. Similar ecological objectives: although heterogeneous cases were selected, the ecological objectives unite three cases. Even though PES-water projects may have a very diverse set of specific ecological objectives; selected cases present the following as the most important ones: a) watershed protection, b) restoration of riparian vegetation, and c) increase in forest cover.

This five selection criteria reduced the research to a sample consisting of Extrema, ProdutorES and Apucarana cases, described in the Table 3:

Table 3. PES case selection criteria

illustration not visible in this excerpt

Source: author field data.

The selected PES cases are in the South and Southeast regions of Brazil, in the Atlantic Rainforest biome. Extrema and Apucarana are local projects implemented at municipal level. ProdutorES is a state-level programme run by the state of Espírito Santo, up till now implemented in five municipalities: Afonso Cláudio, Alfredo Chavez, Alto Rio Novo, Brejetuba and Mantenópolis. Figure 1 presents a map with the locations of the cases.

Figure 1. Extrema, ProdutorES and Apucarana locations

Abbildung in dieser Leseprobe nicht enthalten

Source: author field data.

3.1.2. Basic social, economic and ecological characteristics of study areas

In order to provide a contextual background to the understanding of the PES cases, some basic social, economic and ecological characteristics of the study areas are described in this sub-section. These include information on the history of occupation and settlements, recent economic dynamism in the rural areas, basic economic figures and farmers’ profile, and information on forest’s profile and regions’ hydrology.

3.1.2.1. Extrema

Extrema is a municipality located in the very south of Minas Gerais State, bordering São Paulo State. It is considered an old settlement for Brazilian standards, with early inhabitants dating from the XVIth century. The city is strategically located at the margins of BR-381, one of the most important highways of the country, which connects its biggest city, São Paulo, with Belo Horizonte, the capital of Minas Gerais State and an important regional centre. Due to its location, Extrema has recently attracted a number of industrial investments, which are installing their operations to benefit from its proximity with São Paulo and preferential tax regime offered by Minas Gerais State. These investments raised sharply city’s industrial production and Gross Domestic Product (GDP), placing Extrema on the top of all municipalities of the state of Minas Gerais concerning GDP per person, with R$ 47.4 thousand in 2008 (approx. EUR 20.6 thousand^[2] ), 196% higher than national average and even 46% higher than the city of São Paulo. Industrial investments also contribute significantly to city council’s budget, leading PES project leader Paulo Henrique Pereira to declare that the relatively wealth and comfortable city finances were important factors to the development of Extrema PES project. Nevertheless, Extrema’s economic growth has also attracted an important number of newcomers, which raised city’s population by 49% from 2000 to 2010 to a total of 28.5 thousand people, posing additional urban development problems to the already existing ones.

Rural population remained more stable, summing 2,572 people in 2010, with a relatively high rural density of approximately 85 habitants/km2 (IBGE, 2008). However, Extrema’s rural space is passing through a fast process of rural transformation. According to interviewees, small holder traditional agriculture, mostly based on extensive dairy livestock and low input crops for self-consumption, has been gradually abandoned due to several economic constrains. On the other hand, new actors have emerged – defined by the President of the local Farmer’s Union as the Paulistas, referring to urban people, usually from São Paulo city, that have been acquiring land either for leisure, retirement or rural tourism investments. This new emerging rural society was described by Graziano et al. as the Brazilian New Rural (2002), characterized by, among other factors, high shares of off-farm income on total household income and by a higher presence of non-agricultural purposes in rural spaces. Particularly in the case of Extrema, this process is markedly due to the already mentioned industrialization process, and its relative proximity to a large metropolitan zone, which makes Extrema particularly attractive as a weekend destination.

The municipality lies on the mountain range of Mantiqueira, which defines its hilly landscape, with elevations ranging from 950m in lowest rivers to more than 1700m on its highest peaks. Annual precipitation situates around 1400mm, distributed unevenly between the summer and rainy season and the more dry winter. Heavy rains are frequent during the summer. This tropical humid climate favoured the presence of a rich tropical Atlantic rainforest that dominated coastal areas of Brazil before arrival of European colonization. Today in the region less than 10% of this native vegetation cover is present, even though well preserved portions of rainforest can still be found in Extrema, particularly in its south mountainous border. Several streams and two rivers cross Extrema’s territory, Camanducaia and Jaguari, which later flow into the Cantareira system of reservoirs, an important water supplier to the metropolitan area of São Paulo (Pereira et al., 2010). The general ecological objective of the Extrema PES project is to stabilize the water flow of some important streams that originate in the municipality. In particular, the project seeks to reduce the difference in water supply between rainy and dry seasons and to diminish the risk of flash floods.

3.1.2.2. ProdutorES

The five municipalities benefited from the ProdutorES PES programme have similar economic profiles, even though some differences are prominent. The municipalities of Alfredo Chaves, Afonso Cláudio and Brejetuba have diversified local economies, in which agriculture and related services still play an important role. Particularly in Afonso Cláudio, rural tourism has been providing some dynamic economic opportunities. The municipalities of Alto Rio Novo and Mantenópolis are more isolated from the more economic vibrant centres of Espírito Santo State. Nevertheless economic profiles are still comparable to the other cities.

These cities have medium to high economic and social standards and a relatively high share of its populations living in rural areas, ranging from 37% to 71%, totalling almost 40 thousand rural habitants in the five municipalities. Some mountainous regions of Espírito Santo State – such as those from Alfredo Chaves, Afonso Cláudio and Brejetuba – were barely occupied until early XXth century, when a wave of mostly Italian and Germanic immigrants settled attracted by a state-led immigration policy. Initially, these immigrants suffered from remoteness, but managed to establish flourishing rural communities in which Italian and Germanic influences are still easily noticed on the accent and surnames (Neto, 2009). Arabica coffee in the higher zones, and Robusta coffee and banana in the lower zones dominate agricultural production, conducted mostly by small holder farmers.

In Alto Rio Novo and Mantenópolis, occupation also dates from early XXth century, when first settlers started to grow coffee in the region. This cultural practice is still of great economic importance for the five municipalities, even though more recently planted forests of eucalyptus and other fast-growing trees are substituting pastures and other agricultural areas. In all five municipalities, farmers usually share their agricultural area between their main cash crop – coffee, banana or eucalyptus – with vegetables and other food crops, normally for self-consumption.

[...]

---

^[1] For instance, according to Pagiola (2011) and Guedes and Seehusen (2011) and as explained in the section 1.2, in Brazil there are approximately 30-50 PES projects under implementation, while the OAS database presents only 8 projects. The three PES projects here studied are not in database.

^[2] All currency conversions from R$ to EUR use 2011 average exchange rate (R$ 1 ≈ EUR 0.44).