Diffusion of Information in Agriculture in Senegal

i

ACKNOWLEDGEMENTS

I would like to thank my supervisor, Prof. Dr. Hermann Waibel, who provided me this special opportunity, and was always ready to give advice, encouragement and support.

I am heavily indebted to Dr. Diemuth Pemsl, my tutor, for her patience, her never-tiring assistance and encouragement.

Gratitude goes also to the whole staff of Ceres Locustox, especially to Mr. Barka Dieng, the national coordinator of the IPPM program in Senegal and his assistant, Mme. Fall, for the warm welcome, and all the support during my stay in Senegal.

Thanks to Ndiaw Diop, my special friend, for many interesting and inspiring discussions and the recreative and enjoyable moments at his home, as well as to my friends, Drame Zale, Pape Baji, Modou Diop and Alioune Badara Diop, who did a very good job as translators and enumerators.

My special thanks goes to Ariane Kleiner for her valuable help and encouragement, and last but not least I would like to express my sincerest gratitude to my family for emotional and financial support.

  ii NA  

CONTENTS

LIST OF FIGURES ....................................................................................................................................V

LIST OF TABLES .....................................................................................................................................VI

LIST OF ABBREVIATIONS .....................................................................................................................VII

  ABSTRACT  VIII

1  BACKGROUND OF THE STUDY 1  

1.1  AGRICULTURAL EXTENSION FROM TRADITIONAL TOP-DOWN MODELS TO THE  

  PARTICIPATORY EXTENSION APPROACH  1

1.1.1  History and role of agricultural extension  1

1.1.2  The Transfer of Technology model  2

1.1.3  The Participatory Extension Approach  3

1.2  THE FARMER FIELD SCHOOL CONCEPT OF THE FAO  7

1.2.1  The origins  7

1.2.2  The concept of FFS  7

1.2.3  Advantages and benefits of IP-MFFS  8

1.2.4  Problems and criticism  10

1.2.5  IPP-MFFS in Senegal  12

1.3  OBJECTIVE OF THE STUDY  14

2  CONCEPTUAL FRAMEWORK 16  

2.1  DIFFUSION THEORY  16

2.1.1  Definition Diffusion and Adoption  16

2.1.2  Diffusion of innovations  17

2.1.2.1  Rate of adoption  17

2.1.2.2  Rate of awareness and innovation-decision period  17

2.1.2.3  Critical mass  18

2.1.3  Diffusion research  19

2.2  SOCIAL COGNITIVE THEORY  23

2.2.1  Basic assumption  23

2.2.2  Information and knowledge  24

2.2.3  What can informal interaction achieve  24

2.2.4  The value of information  25

  iii NA  

2.3  THEORY OF COGNITIVE DISSONANCE  26

2.3.1  General description  26

2.3.2  Social pressure  28

2.3.3  Heterogeneity of population  28

2.4  HYPOTHESES OF THE STUDY  30

2.4.1  FFS intensity and information diffusion  30

2.4.2  Information and stage of adoption  31

2.4.3  Transformation of FFS intensity into individual exposure  32

2.4.4  Specification of the survey hypothesis  32

3  DATA COLLECTION AND METHODOLOGY 35  

3.1  SAMPLING PROCEDURE  35

3.2  QUESTIONNAIRE  37

3.3  GENERAL DESCRIPTION OF THE STUDY SITE  38

3.4  METHODOLOGY OF ANALYSIS  40

3.4.1  Social Network Analysis  40

3.4.1.1  Respect and Advice Network  41

3.4.1.2  IPPM Talks Network  41

3.4.2  The logistic regression model  42

3.4.2.1  The specification of the logistic model  42

3.4.2.2  Estimation of the logistic function  44

3.4.2.3  General evaluation of the model  44

3.4.2.4  Evaluation of independent variables  46

4  RESULTS OF THE SURVEY 47  

4.1  SOCIO-ECONOMIC AND INSTITUIONAL CONDITIONS IN THE STUDY AREA  47

4.1.1  Demographic conditions  47

4.1.2  Social conditions  51

4.1.2.1  Respect and Advice Network  51

4.1.2.2  IPPM Talks Network  53

4.1.2.3  Summary of results  56

4.1.3  FFS intensity and individual exposure  56

4.1.3.1  Rationale of individual exposure  56

4.1.3.2  The intensity and time of exposure  57

  iv NA  

4.2  DIFFUSION OF INFORMATION ABOUT IPPM  60

4.2.1  The likelihood of receiving IPP-Mrelated information  60

4.2.2  Cognitive dissonance and search for information  64

4.2.3  The intensity of information reception  67

4.2.3.1  Quality of Information  68

4.2.3.2  Quantity of Information  68

4.2.3.3  Regression results  70

4.2.4  Summary  74

4.3  EFFECTS OF INFORMATION DIFFUSION ON ADOPTION BEHAVIOR  75

4.4  SUMMARY OF THE RESULTS  82

5  CONCLUSIONS AND RECOMMENDATIONS 84  

5.1  CONCLUSIONS DRAWN CONCERNING THE DIFFUSION OF FFS-ACQUIRED KNOWLEDGE  84

5.1.1  What vs How  84

5.1.2  Diffusion of IPPM knowledge in Senegal  85

5.2  QUALITATIVE DATA AND EXPERIENCES IN THE FIELD  87

5.2.1  Problems to the implementation of IPPM  87

5.2.1.1  Lack of water  87

5.2.1.2  Limited or no credit supply  88

5.2.1.3  Problems of commercialization and conservation  88

5.2.2  Incentives to participate  89

5.3  IMPLICATIONS OF SURVEY FINDINGS FOR THE DESIGN AND IMPLEMENTATION OF FFS-IPPM  

  PROJECTS  90

5.4  OUTLOOK  91

6  REFERENCES 93  

7  APPENDICES 99  

v

LIST OF FIGURES

  Figure 1: The Transfer of Technology model  3

  Figure 2: The participatory extension approach  4

  Figure 3: The participatory extension approach cycle  6

  Figure 4: Distribution of adopters and their categorization on the basis of innovativeness  17

  Figure 5: The rate of awareness-knowledge and the innovation-decision period  18

  Figure 6: The rate of adoption for usual and interactive innovations the impact of critical mass  19

  Figure 7: Variables Determining the Rate of Adoption  20

  Figure 8: Relations between the three classes of determinants in the reciprocal causation  23

  Figure 9: Knowledge environment and their interface  24

  Figure 10: Relation between magnitude of dissonance and active seeking of new information  27

  Figure 11: The increase in exposure and adoption  31

  Figure 12: The adoption process  32

  Figure 13: Schematical presentation of the theoretical design of the study  33

  Figure 14: Two villages as two points in time on the path of diffusion (rate of adoption)  35

  Figure 15: Schematic structure of the villages  36

  Figure 16: Region de Niayes (velvet) and the survey area  38

  Figure 17: Distribution of the logistic function  43

  Figure 18: Education distribution in the villages  48

  Figure 19: Education and gender  48

  Figure 20: Land tenure and gender in both villages  49

  Figure 21: Average farm size owned by men women  50

  Figure 22: Right of decision and gender  50

  Figure 23: Sociometric status for Gollam  52

  Figure 24: Sociometric status for Keur Abdou Ndoye  53

  Figure 25: Closeness index for Gollam  55

  Figure 26: Closeness index for Keur Abdou Ndoye  55

  Figure 27: Distribution of FFS farmers within the villages  57

  Figure 28: Frequency of time of exposure in years  59

  Figure 29: FFS intensity and rate of awareness  67

  Figure 30: Number of talks about IPPM in the last month  70

  Figure 31: Attitude of exposed farmers towards IPPM  75

  Figure 32: Constraints to the diffusion of IPPM knowledge  85

  Figure 33: Diffusion of IPPM to Non-FFS farmers in Senegal  86

  vi NA  

LIST OF TABLES

Table 1:  Comparison of transfer of technology and participatory extension  5

Table 2:  Examples of immediate and developmental impacts of IP-MFFS  9

Table 3:  Key characteristics of visited villages  35

Table 4:  Sample characteristics in Gollam and KAN  37

Table 5:  Land tenure issues  49

Table 6:  Social network characteristics of the Respect and Advice Network  51

Table 7:  Social network characteristics for IPPM Talks Network  54

Table 8:  The shares of FFS Exposed and Non-Exposed farmer groups by village  56

Table 9:  Variables of individual exposure  58

Table 10:  Cases of the dependent variable used for the analysis  60

Table 11:  Explanatory variables used for the analysis  61

Table 12:  Expected signs of the regression coefficients of the logistic model  63

Table 13:  Results of the regression  63

Table 14:  Initiator of the talks  65

Table 15:  Logistic regression results Dependent variable: Passive exposure  

  or active search for information  65

Table 16:  Quantity of information transfer  68

Table 17:  OLS-estimation results Dependent variable: Number of persons talked with  

  about IPPM  71

Table 18:  OLS-estimation results Dependent variable: Frequency of IPP-Mrelated talks  

  in the last month  73

Table 19:  OLS regression results The dependent variable is IPPM is a solution to  

  prevailing agricultural problems  76

Table 20:  Results of the binary logistic regression Dependent variable: wish to adopt  77

Table 22:  Diffusion of IPPM knowledge to Non-FFS farmers  86

vii

LIST OF ABBREVIATIONS

EIQ - Environmental Impact Quotient FAO - Food and Agriculture Organazation FFS - Farmer Field School FP - Farmer Practice IA - Impact Assessment IPM - Integrated Pest Management IPPM - Integrated Pest and Production Management IRRI - International Rice Research Institute KAN - Keur Abdou Ndoye LDC - Least Developed Country NGO - Non Governmental Organazation OLS - Ordinary Least Squares PE - Participatory Extension PEA - Participatory Extension Approach SNA - Social Network Analysis SPSS - Superior Performing Software System TOFT - Training of Farmer Trainers ToT - Transfer of Technology

viii

ABSTRACT

Using participatory training approaches such as farmer field schools (FFS) is perceived to be an appropriate technique to improve farmers’ knowledge of complex agro-ecological systems and Integrated Production and Pest Management (IPPM) technologies. Through intensive and exploratory learning farmers are trained to make competent and adequate decisions adapted to their local, specific environment. However, the FFS approach appears to be more costly than alternative less intensive approaches of knowledge transfer. Calculations of the costs per farmer trained in Indonesia and the Philippines amount to US$ 45 to 60, which are considered to be rather underestimates of the true costs. These high per capita expenditures are more justified, if knowledge can be disseminated through informal farmer-to-farmer interactions. Participatory extension approaches therefore rely on interpersonal channels and group mechanisms for diffusing greater awareness and facilitating learning among the group of untrained farmers. Although diffusion is not an explicit goal of the FFS approach, it is nevertheless a desired side- effect, which could invalidate the reproach of fiscal unsustainability.

However, several empirical studies found that knowledge generated by participatory extension training does not always sufficiently diffuse to non-participating farmers, because of the complexity of knowledge imparted in the course of a FFS. But even though there seems to be little diffusion of knowledge, farmers nevertheless exchange their experiences with each other, and thus, information about IPPM is being disseminated in the village. This diffusion of information can have a significant impact on adoption behavior and be an important factor for a successful introduction and establishment of an innovation like IPPM. Depending on the project placement strategy, diffusion of information can be boosted and promoted, or it can be hampered by creating unfavorable conditions.

This case study, conducted in the ‘Région de Niayes’ in Senegal in 2004, investigates the effects of training intensity on the diffusion of information. A total of 341 vegetable growers were interviewed in two villages (Gollam and Keur Abdou Ndoye) that had different shares of farmers trained in IPPM-FFS (Gollam: 3% - Keur Abdou Ndoye: 14%) but are similar in all other respects. The objective of the study was to identify the factors determining the intensity of information diffusion. A set of predominantly closed questions was used to generate data on demographic, farm-, IPPM-, and information-related issues, which are considered important to capture the diffusion processes. The data has been analysed using a logistic regression model as well as the ordinary least-square (OLS) -estimation model.

The results show that the proportion of trained farmers affects the dissemination of IPPM- related information. A higher share of FFS participants increases the individual exposure of non-participants. As a result, the likelihood of receiving information about IPPM is four times higher in Keur Abdou Ndoye than in Gollam. Consequently, the number of exposed farmers

ix

approaches 100% in Keur Abdou Ndoye. In addition, the quantity of information that is shared is higher. The individual exposure to IPPM has a statistically significant positive impact on information sharing. A second level of impact is the stage of adoption of FFS farmers and the observability of IPPM practices and benefits, which is a crucial variable for the intrinsic motivation of exposed farmers to adopt such technologies.

1

1 BACKGROUND OF THE STUDY

1.1 AGRICULTURAL EXTENSION – FROM TRADITIONAL TOP-DOWN

MODELS TO THE PARTICIPATORY EXTENSION APPROACH

1.1.1 History and role of agricultural extension

Poverty and hunger, food production and natural resource degradation are the great challenges in today’s world. The most affected and vulnerable group is the rural population of developing countries. More than 80% of the poor live in rural areas. In the face of a rising population in the LDCs 1 (2.4% - annual growth rate 2000-2002) 2 , improving agricultural productivity and farmers’ incomes are important issues in the fight against poverty, which are addressed by all involved governments and organizations of development assistance.

Agricultural extension is a tool, that can supply improved skills, information and ideas to people involved in the agricultural sector in order to develop an agriculture that will meet complex demand patterns, reduce poverty, and preserve ecological resources. 3

The term “extension” originates in the discussion of the Oxford and Cambridge universities about how to serve the educational needs of the growing population in the urban area in the 19 th century. The movement that emerged out of this discussion was designated “university

extension” 4 . The success of this work in Britain initiated similar activity elsewhere. Since the 19 th century and the potato famine attempts were developed in many European countries to impart useful knowledge to farmers through itinerant agricultural teachers or in other forms. 5 This significant social innovation and important force in agricultural change, has then been created and recreated, adapted and developed over the centuries. Particularly in the last decades extension operations “may well be the largest institutional development effort the world has ever known. Hundreds of thousands of technicians have been trained and hundreds of millions of farmers have had contact with and likely benefited from extension services.” 6 Today we understand agricultural extension as both, a system (institutional interpretation), and a set of functions (functional interpretation) performed by that system to induce voluntary change among rural people. It is largely publicly funded (80 percent of the world’s extension services), as most of the aspects of agricultural knowledge supply have the nature of public goods. 7 Worldwide agricultural extension employs more than 800.000 extension workers. Especially the developing country governments invested largely, as they expect an increased

1 Least Developed Country 2 The Least Developed Countries Report 2004, p.321 3 see Feder, Willett and Zijp (1998), p.1 4 Jones and Garforth (1997), p.1 5 see Jones and Garforth (1997), pp.1 and 5 6 Anderson and Feder (2003), p.21 7 see Fleischer, Waibel and Walter-Echols (2002), p.310

2

agricultural production induced by the information input provided by extension efforts. “Between 1959 and 1980, spending in real terms for extension grew more than six-fold in Latin America, tripled in Asia, and more than doubled in Africa.” 8

The goals of extension are transferring knowledge from researchers to farmers, advising farmers in their decision-making, educating farmers to be able to make similar decisions in the future, enabling them to clarify their own goals and possibilities and to fulfill them, and stimulating desirable agricultural developments. 9

1.1.2 The Transfer of Technology model

Until recently extension consisted mainly of farmers and communities being instructed what to do, often by institutions that had not carefully and thoroughly identified their local needs. The Transfer of Technology model (ToT), a rather “engineering” approach was the prevalent practice for developing and spreading of innovations. It bases on the assumption that gaps between the actual productivity of the farms and the potential productivity with better technology and know-how could be bridged if farmers had better access to certain inputs and used them according to a set of prescribed instructions. Extension would thus reduce the differential between potential and actual yields in farmers’ fields by accelerating technology transfer (reducing the technology gap) and helping farmers to become better farm managers (reducing the management gap) 10 . In this view farmers are often considered as the main constraint to development, as mis-managers of their resources. The role of the extension agent is to assist farmers in putting the ready-made technology into practice, thus running the danger that it may not be appropriate. 11

The results tended to be poor. The ToT-model based on a one-way communication that discouraged feedback of information. Researches worked independently of farmers and extension workers. The “innovations” were developed on research stations without considering the on-farm problems of the farmers. The dissemination of these innovations and technologies relied then on the extension workers. They were seen as technical agents. Social competence was not required as the socio-organisational issues were neglected or reduced to a technical level. In short, this linear conception viewed farmers, extensionists and researchers as three separate strata and the links between them have been weak or non-existent (Figure 1).

8 Feder, Willett and Zijp (1998), p.4 9 see Feder, Willett and Zijp (1998), p.3 10 see Anderson and Feder (2003), p.3 11 Hagmann, Chuma, Murwira and Connolly (1999), p3

3

Figure 1: The Transfer of Technology model

Source: Semana (2002), p. 3

The top-heavy ToT model led to highly bureaucratic structures with all their specific problems like lack of accountability 12 , fiscal unsustainability and poor interaction with other stakeholders 13 , and also implied low adoption rates, because farmers did not have any sense of ownership of the ideas imposed on them. 14 The methods were technology-driven, not farmer- driven; they were centrally uniform, not locally adaptive. 15

This top-down approach has been often criticized and a general change towards participatory approaches is sweeping through the development movement. The notion that “extension systems should be made more accountable to the clients and service delivery should be demand-pulled rather than supply-driven” 16 is more and more prevalent.

1.1.3 The Participatory Extension Approach

It is recognized that extension has a dual function. It not only does translate information from the researchers to farmers, it also has an important role in supporting researchers to tailor technology to the agro-ecological and resource conditions of farmers. At the heart of this change is the awareness that rural people themselves are the owners and shapers of their own development. The extension agent is no longer seen as the expert who has all the knowledge and technical solutions. The client’s own knowledge and ingenuity is seen as a major resource. Solutions to local problems are to be developed in partnership between research, extension agents and the rural population. This definition of the so-called participatory extension approach (PEA) implies three principles:

Participation: The process of technology identification, development and transfer must include the farmers in the locality being served.

Integration: The process of extension must involve researcher, extension agent and farmer in an integrative manner, using also local resources or personnel.

Practical relevance: Technology development and transfer must focus on actual and immediate problems of farmers. 17

12 see Feder, Willett and Zijp (1998), p.9 13 see Fleischer, Waibel and Walter-Echols (2002), p.309 14 see Hagmann, Chuma, Murwira and Connolly (1999), p.1 15 see Pontius, Dilts and Bartlett (2002), p.15 16 Fleischer, Waibel and Walter-Echols (2002), p.310 17 see Asiabaka (2001), p.2

4

Figure 2: The participatory extension approach

Source: Semana (2002), p. 5

The main failure of traditional extension models was the underestimation of the fact that outsiders are not able to truly determinate the “best practice” for specific local conditions. Farmers know their personal circumstances better and therefore are in a better position to decide about the management of their farms within the many environmental and social constraints they face. It is an illusion that outsiders can fully understand the totality of factors, which influence the behavior of local stakeholders. For successful innovation development and adoption, farmers need to experiment, adapt, evaluate and determine the practices most appropriate for their own situation. They need to become “experts in farming” 18 , mastering the ecological principles and developing solutions themselves. Agricultural extension therefore has to be participatory, considering all the different interwoven social, economical, cultural, political and ecological factors that determine the working conditions of farmers. Within this concept, the extension agent is seen as a “facilitator”, not as a “teacher” or “instructor”. Instead he plays the role of a mentor or educational companion. Both, facilitators and farmers become active participants in the educational process. The hierarchy between them is eliminated and a sense of community formed. Knowledge is created not transferred and is considered to be located in the community rather than in the individual. 19 Problems are identified and solutions are generated in collaboration between farmers and the facilitator. Table 1 summarizes the differences between the ToT and the PE approaches.

18 Pontius, Dilts and Bartlett (2002), p.16 19 see Whipple (1987), p.3

5

Table 1: Comparison of “transfer of technology” and “participatory extension”

Source: Hagmann et al. (1998), p.5

As far as the operational implementation is concerned, the participatory extension process can

be divided into four phases:

Phase 1: Social mobilisation: Facilitating the communities’ own analysis of their situation

Phase 2: Community-level action planning

Phase 3: Implementation and farmer experimentation

Phase 4: Monitoring the process through sharing experiences, ideas and self-evaluation

Figure 3 visualizes this interactive education process, as developed through experiences of

extension work in Zimbabwe. The self-evaluation at the end of the first cycle in the process

leads to the next cycle, which starts again with social mobilisation. PEA is a continuous process

of learning. 20 The potentials and pitfalls of participatory extension will be discussed in the next

chapter, in connection with the farmer field school approach.

20 see Hagmann, Chuma, Murwira and Connolly (1999), p.6

6

Figure 3: The participatory extension approach cycle

Source: Hagmann et al. (1998), p.15

7

1.2 THE FARMER FIELD SCHOOL CONCEPT OF THE FAO

1.2.1 The origins

Within the framework of participatory extension several methodologies or tools have been developed as for example group extension methods, farmer-to-farmer extensions, farmer field schools (FFS), master farmer training, diagnostic survey, farmer participatory research and rapid rural appraisal 21 .

The FFS approach emerged in the late 1980s and was first implemented by the FAO in Indonesia as an answer to the poor performance and even the failure of precedent extension work, especially during the “green revolution” in East Asia from the 1960s to the 1990s. The long-term effects of the traditional engineering extension concept had been disastrous. The standardized “production packages” were not adaptable to local conditions. Inadequate and abusive use of pesticides caused dramatic losses to national food production and also aggravation of the financial situation of the farmers 22 . The estimated yield loss in Indonesia in the year 1976-1977 for example, caused by the brown planthopper, was 364,500 tons of rice – “enough to feed three million people for an entire year.” 23 International Rice Research Institute (IRRI) researchers found in studies in the late 1980s that farmers had the capacity to learn, innovate and even outperform research stations in terms of average yield. If they only had the knowledge about new technologies, they could discriminate among the solutions offered to them by the research system, adapt the technologies to their particular environmental conditions, and provide supervision of inputs to ensure the appropriate application of the technology. 24 This acknowledgement of the indigenous potential of farmers led to a change of the existing extension systems.

1.2.2 The concept of FFS

The Farmer Field School is a group approach to agricultural technology development, focusing on adult, non-formal education through hands-on field discovery learning. 25 Originally designed as a way to introduce knowledge on Integrated Pest Management (IPM) to irrigated rice farmers in Asia, the FFS approach emphasises four principles: i. Grow a healthy crop ii. Conserve natural enemies of insect pests iii. Monitor the fields regularly iv. Become IPM experts through participation in FFS 26

21 see Hagmann, Chuma, Murwira and Connolly (1999), p.2 22 Pesticides can often be the cause for severe pest outbreaks, as they also diminish the population of natural enemies of certain pests.

23 Pontius, Dilts and Bartlett (2002), p.14 24 see Pontius, Dilts and Bartlett (2002), p.13 and 15 25 see Friis-Hansen, Maganga and Sokoni (2004), p.59 26 see Friis-Hansen, Maganga and Sokoni(2004), p.59

8

Operationally, the FFSs are organized around a series of 9-12 half-day long weekly sessions in a group of 20-25 farmers during a single crop season, focussing on biology, agronomic and management issues, where farmers conduct agro-ecosystem analysis, identify problems and then design, carry out, and interpret field experiments using IPM and Non-IPM comparisons. 27

1.2.3 Advantages and benefits of IPM-FFS

This concept of learning and education is highly responsive to local needs over a wide range of conditions, and with a wide range of crops. It helps farmers to acquire an understanding of important “system” concepts and relationships. In addition FFSs also include a significant focus on group and individual capacity building.

A key feature of the FFS approach is that it rectifies the problem of accountability, traditional extension models were so often confronted with. This aspect is addressed in two ways:

i. The facilitators are bound by a strict timetable of sessions within a predefined curriculum that can be easily looked-over by supervisors.

ii. The continuous interaction with a cohesive group of farmers creates a certain accountability to the group, in particular when the facilitator is a farmer-trainer, who is member of the same community.

These characteristics of FFS are expected to ensure the quality of knowledge provided to the farmers. 28 As mentioned above, the FFS approach was designed originally as a way to introduce knowledge on Integrated Pest Management. Although the farmer field school concept has been adopted and imitated by many official and non-governmental organizations, with different educational contents, the FFS programs conducted by the major player in this field, the FAO, are still designed to promote IPM.

The performance of the FFS (IPM) approach has been examined and evaluated by numerous Impact Assessment (IA) studies. Most of them concentrated on measuring immediate impacts like the effects on pesticide use and yield. The recent discussion however leads to a more complete assessment of the broad range of developmental impacts, including changes in the social and political domain. Table 2 resumes the possible impact dimensions of the farmer field schools.

27 see Owens and Simpson (2004), p.66 28 see Anderson and Feder (2003), p.20

9

Table 2: Examples of immediate and developmental impacts of IPM-FFS

Source: van der Berg (2004), Chapter 2.2

In a meta-study on Impact Assessment van der Berg draws this recapitulatory conclusion on the impact of IPM-FFS throughout the world: “The majority of studies measured the immediate impact of training through aggregated data, and reported substantial and consistent reductions in pesticide use attributable to the effect of training. In a number of cases, there was also a convincing increase in yield due to training. (…) A number of studies described broader, developmental impacts of training often using qualitative methods, and in some cases involving farmers in identifying and describing the impacts. Results demonstrated remarkable, widespread and lasting developmental impacts, which have been best documented for Indonesia. It was found that the Farmer Field School stimulated continued learning, and that it strengthened social and political skills, which apparently triggered a range of local activities, relationships and policies related to improved agro-ecosystem management.” 29 Beyond these impact effects, recent studies emphasize also the significant environmental and health benefits

29 van der Berg (2004), Chapter 5

10

of IPM-FFS, which is measured for example using the concept of the Environmental Impact Quotient (EIQ) 30

1.2.4 Problems and criticism

However, participatory extension methods such as farmer field schools face the same problem that has always dogged large extension systems – fiscal unsustainability. The FFS and other PE approaches can be relatively expensive (if compared to media campaigns for example) – both in time and in related training costs – because they require farmers’ hands-on participation in small, trainer-facilitated groups. Calculations of the costs per farmer trained in Indonesia and the Philippines amount to US$ 45 to 60, which are considered to be rather underestimates of the true costs. 31 Therefore diffusion effects play an important role. PEAs rely heavily on interpersonal channels and group mechanisms for diffusing greater awareness and facilitating learning among the group of untrained farmers. Although diffusion is not an explicit goal of participatory extension approaches like FFS, it is nevertheless a desired side-effect, which could invalidate the reproach of fiscal unsustainability. Knowledge of IPM, disseminated to non- trained farmers, can uplift the general agro-economical performance of small-scale farmers. Economically spoken, it’s the positive external effects of FFS that ensure a better efficiency of the training. If certain contents and practices could be communicated and adopted without official expenses, the benefit-cost ratio would increase. This farmer-to-farmer extension could raise the program coverage, allow FFS-acquired knowledge to spread faster and thereby make the FFS a more cost-effective approach to agricultural extension.

The FFS approach relies on two knowledge transmission principles, in order to reduce the fiscal burden and to accelerate the diffusion of IPM. First, the training of farmer trainers (TOFT). FFS graduates are encouraged to undertake a TOFT, thus becoming facilitators themselves, and subsequently to train other farmers. Through TOFT a large number of facilitators can be trained, who know their local conditions well and who can conduct FFS in their own or in neighboring villages. Second, in addition to this formal diffusion mechanism the transmission of knowledge also works through informal farmer-to-farmer communication.

Several studies have been conducted to identify the diffusion of FFS-acquired knowledge. The empirical evidence has brought some criticism to the FFS approach, especially from the World Bank. A scientific debate over the fiscal sustainability of FFS is now going on since 4 or 5 years, which was set off by a study by Agnes C. Rola and Serlie B. Jamias (2002) 32 , a second study by Gershon Feder, Rinku Murgai and Jaime Quizon (2002) 33 , and a repeated study by Feder et al. in

30 Walter-Echols (2004) 31 see Quizon, Feder and Murgai (2004), pp.51 and 55 32 “Do Farmer Field School Graduates Retain and Share What They Learn? An Investigation in Iloilo, Philippines” 33 “Sending Farmers Back to School – The Impact of Farmer Field Schools in Indonesia”

11

2004 34 . All three studies questioned the diffusion effects that before had been assumed to multiply the knowledge that had been imparted on the FFS farmers.

These studies found that knowledge gained by farmers in a FFS “does not diffuse in a significant way to other members of their villages” 35 , and “non-participants appear hardly to be affected” 36 . This again leads to the conclusion that “if the likelihood of farmer-to-farmer diffusion of FFS information is negligible, then the program will need to train directly large numbers of farmers, otherwise, it will have a very limited impact at the national level. But given the high costs associated with the intensive FFS training, the fiscal dimension becomes a serious obstacle as many countries would not be able to afford the large fiscal expenses, over a long period of time.” 37 An extrapolation of the training efforts of farmers who had undergone the TOFT in Indonesia shows furthermore that even the first diffusion effect “cannot be relied upon to maintain a significant training effort under the FFS approach.” 38 The reasons for this “failure” could be the complexity of knowledge, imparted in the course of an FFS, as the training aims at developing the analytical skills, critical thinking, and creativity of farmers and to help them learn to make better decisions. FFS graduates do not master a specific set of contents or “messages”, they rather master a “process of learning”. 39 For this reason knowledge does not diffuse well to other village members without the exploratory activities that are the key part of the FFS approach itself. 40

These findings were repeatedly challenged by other scholars 41 . Several letters have been sent to the authors of the mentioned study 42 . Workshops on FFS have taken up this issue and discussed and criticized the methodology and findings of the survey, sometimes in a rather polemic manner. Gallagher replies to Feder’s critique 43 : “The paper points out the obvious – that most extension services have no money to implement programs without external budgets. (…) Using their paper, one would also conclude that any public expenditure is too expensive. But governments still fund education, health and other services.” Other calculations seem to contradict or relativize the calculations by Feder et al. (2004) and Quizon et al. (2004). Owens and Simpson (2004) stated that the costs per farmer under the East African conditions vary between US$9 and 35, and in Ghana the estimated costs amount even to less than US$10 44 . It seems, however, that the view of the World Bank on the fiscal sustainability of the FFS has not changed. But even if researchers have reason to disagree with Feder’s conclusions concerning

34 “The Acquisition and Diffusion of Knowledge: The Case of Pest Management Training in Farmer Field Schools, Indonesia” 35 Feder, Murgai and Quizon (2004), p.238 36 Rola and Jamias (2002), p.9 37 Feder, Murgai and Quizon (2004), p.238 38 Quizon, Feder and Murgai (2004), pp.54 and 56 39 Quizon, Feder and Murgai (2004), p.50 40 see Rola and Jamias (2002), p.10 41 e.g. Waibel, Fleischer, Walter-Echols, Pemsl, Praneetvatakul or Gallagher 42 Feder, Murgai and Quizon (2002): “Sending Farmers Back to School…” 43 Gallagher (2002), p.3 44 see Simpson and Owens (2002), p.5

12

the fiscal sustainability of FFS, the World Bank studies have quite serious implications regarding the diffusion of knowledge through informal interaction.

1.2.5 IPPM-FFS in Senegal

In Africa, the FFS approach was first introduced by the FAO Global IPM Facility within the IPM-FFS program, in the year 1995 in Ghana. Since then 6.000 farmers and 400 extension agents have been trained in Ghana, covering a dozen different crop species. 45 Following the efforts in Ghana, a National IPM Program was established in Mali in 1998, which spread rapidly in the following years. At the same time similar efforts were initiated in Kenya and Zimbabwe. To date the Global IPM Facility has helped to start, or is currently working with FFS programs in over a dozen African countries, from Senegal to South Africa. Several of these have moved beyond the pilot stage and are expanding their activities.

In 1999, Senegal was selected as one of three pilot countries for a West African regional project on Integrated Production and Pest Management (IPPM). (The introduction of IPM FFS in Africa has shown that there are broader agronomic, management and production issues that have to be addressed by the facilitators. This has led the FAO to talk of IPPM rather than just IPM. IPPM goes beyond plant protection with links to irrigation and fertilizer information.) Set up by the Global IPM Facility for the Food and Agriculture Organization of the United Nations the project is carried out by the Foundation CERES-Locustox 46 . The first phase of the main program started in 2000 and continued until the end of 2004, after which the program was to be continued for a second phase. The focus of the program was on smallholder vegetable growers because of their increasing reluctance to continue with existing pesticide practices, attendant risks of residues of unauthorized pesticides, accidents and the high costs resulting therefrom. In Senegal, pesticide use is highest among vegetable and cotton growers. 47 Beyond that a great anxiety prevails among farmers because of lack of sufficient irrigation water and increasing infertility of soil. The FFS curriculum took up some lessons on compost, organic manure and prevention measures against fast evaporation. The pilot phase of farmer training through farmer field schools was a success. 205 FFS have been conducted, with 166 on vegetables. Until now some 3700 farmers have been trained, and a preliminary Impact Assessment study based on the FFS experimental plots has shown encouraging results like positive yield effects and a significant replacement of chemical pesticides by biological pesticides through FFS IPPM training 48 . At the end of the first phase, the project has been evaluated by FAO staff and has been approved as worthwhile to be continued. In addition, a large cross-country Impact Assessment Study is being conducted by the University of Hanover in collaboration with the national coordinators of Senegal, Mali and Burkina Faso, and

45 see Simpson and Owens (2002), p.2 46 CERES-Locustox is a training and research centre for environmental toxicology, mainly of pesticides in the Sahel, PO Box 3300, Dakar, Senegal 47 Diallo, Dieng and Everts (2004) 48 Pemsl (2004); and Diallo, Dieng and Everts (2004)

13

supported by the FAO, to examine the economic impact of FFS-IPPM training on the performance of small-scale rice and vegetable growers. For the second phase, which will continue for another four years, Cap Verde and Benin will join the project. 49

49 Settle (2005)

14

1.3 OBJECTIVE OF THE STUDY

This study analyzes the implications of different approaches to the design and the implementation of IPPM-FFS for the process of diffusion. Most important is the question, how the proportion of trained farmers in a village influences the dissemination of knowledge or information. Designing an FFS project several approaches can be used as far as project placement is concerned. The object is generally always the same: to reach as many farmers as possible in a certain region. But the way of implementing can differ. On the one hand, the decision makers can follow a “drop-strategy”, conducting one FFS per village, thus introducing IPPM to as many villages as possible, but receiving a very small share of adopters per village. On the other hand, they may concentrate on view villages and continually support them over a longer period, ensuring a high adoption rate and a profound and intensive training of farmers. Both approaches have different implications for the diffusion of knowledge.

Two interdependent aspects can be identified in the process of diffusion. First, the spread of information or knowledge about IPPM and second, the actual adoption of IPPM. Both issues are interrelated as with more and more farmers adopting IPPM the intensity (quality and quantity) of information exchange increases. Contrarily, with an increasing village-wide attention to, and intense communication about, IPPM, the social forces that influence adoption increase, so that IPPM becomes a socially accepted and valued practice. At a certain adoption rate (share of adopters in the village), this process of mutual effects becomes self-sustaining and pushes adoption forward to an equilibrium point, or even up to 100%. This adoption rate is called “the critical mass”.

In order to evaluate the different implementation approaches, this study tries to investigate the flow of IPPM-related information in two villages with different adoption rates - one below and one above the hypothetical critical mass of around 10%.

The results of this study may open up better ways to exploit the potential of FFS training. They can possibly be used to make better decisions with regard to project placement, achieving a higher acceptability and stage of adoption of IPPM, and thus a more efficient use of resources in the second phase of the project.

The next chapters are organised as follows:

Section 2.1 contains a concise overview of diffusion theory, mainly based on the classical work by Everett M. Rogers 50 . In section 2.3 an exact definition and scope of the study, the general hypothesis, and several questions derived from it, will be given, which shall serve as a guideline in the analysis of the survey data.

Chapter 3 will then give an overview of data collection and the methodology used for analysis, particularly the logistic regression model.

50 Rogers (2003): “Diffusion of Innovations”

15

The main part of this study, the results of the analysis, are presented in chapter 4, following the

outline of hypotheses of section 2.3.

In Chapter 5 finally some conclusions and recommendations are given.