Sustainability Impact Assessment of a Certification Scheme in the Indonesian Cocoa Industry: 2012 Pilot Survey Results

Transcription

1 Sustainability Impact Assessment of a Certification Scheme in the Indonesian Cocoa Industry: 2012 Pilot Survey Results Authors: Hiswaty Hafid, Jeff Neilson, Tula Mount and Fiona McKenzie University of Sydney, Australia Research funded by the Australian Centre for International Agricultural Research (ACIAR) This Discussion Paper was released in March 2013, with the intention that interested stakeholders can provide feedback to the authors to improve the assessment indicators and approach used for further research. For feedback or more information, please contact Jeffrey.neilson@sydney.edu.au

2 Table of Contents Executive Summary Introduction Methods Results of the pilot survey... 5 A. GENERAL INFORMATION... 6 B. FARM CHARACTERISTICS... 6 B.I. Vital Statistics... 6 C. ECONOMIC SUSTAINABILITY C.I. Cocoa production, prices and competitiveness C.II. Involvement in Producer Organisations C.III. Farmer perceptions of economic sustainability D. ENVIRONMENTAL SUSTAINABILITY D.I. Water D.II. Use of agrochemicals D.III. Soil D.IV. Attitudes towards the environment E. SOCIAL SUSTAINABILITY E.I. Basic human rights, worker rights, and equity E.II. Perception F. FARMER PERCEPTIONS OF THE CERTIFICATION PROGRAM Suitability of survey method Comparing indicators used with COSA indicators Survey effectiveness Reflections on the need for greater observation based results Tentative findings on the impacts of certification schemes Conclusion References Appendix A - ACIAR Cocoa Sustainability Survey

3 Executive Summary A challenge facing certification schemes is that their impact and effectiveness is variable or unclear something that all stakeholders agree needs to be overcome. This report outlines the results of a pilot survey that was undertaken to test the effectiveness of a set of farm-level indicators in assessing the role of market-based certification schemes in encouraging improved farm practices and the provision of farmer support services. The study approach reflects some elements of the Committee on Sustainability Assessment (COSA) methodology, but has key differences and was conducted independently of COSA. In conducting this relatively large-scale pilot, our objective was to identify improvements to develop a method more suited to the local producer context in Indonesia that could then be conducted in subsequent years ( ). The key methodology used to analyse impacts was a farm-level survey that used various indicators of sustainability. This involved trialling the survey with 158 cocoa farmers in West Sulawesi in July 2012, including target farmers who were already certified and a control group with similar characteristics but who were not certified. The challenge in undertaking a more localised assessment is to select indicators that fit with local sustainability issues. It was evident from the pilot that we partially achieved this, while some indicators were less useful. Low response rates were recorded for questions that relied upon memory recall. For example, questions about household income and certification premiums were difficult for farmers to respond to. Questions that lacked relevance also had a low response rate. For example, involvement in cooperatives, distance to markets and hiring of labour weren t necessarily relevant as few farmers were in cooperatives, most sold to local collectors not a distant market and few farmers hired labour. Detail cannot always be captured through quantitative questions. In measuring impact, we attempted to move beyond comparing linear changes of farm performance (yields, farm income, and profit), to also capturing intermediate (and often less tangible) outcomes that may indirectly support or lead to improved sustainability in the long-term. We tried to capture additional information through subjective questions that asked for the farmer s perception of certain factors. Additional improvements to the survey will be required. Observation-based assessments by the interviewer may also contribute to a better description of economic, environmental and social conditions. While results are preliminary and based on a set of draft indicators that will be further reviewed, we were able to draw some tentative conclusions about farmers experiences of the certification process. For example, farmers participating in certification schemes were overwhelmingly positive about the benefits of the program. They were particularly positive about the economic benefits and the provision of associated services, such as training and credit. Certification was also associated with significantly more active producer organisations, which were delivering several benefits to participating farmers (eg. labour sharing, access to credit, collective marketing). Other impacts are harder to disassociate from other influences. For example, certified farmers (even after taking into account extended drying) are receiving higher prices than non-certified 1

4 farmers, although it is difficult to ascertain whether this is due to the certification scheme specifically or the associated direct-purchasing program. Productivity levels appear to be significantly higher amongst certified farmers (based on farmer yield estimates), although it is difficult to dissociate the influence of certification from the influence of other programs (eg. the USAID Amarta program) that had been active in the target community. Longer term impacts that are only starting to materialise were also evident. For example, certification is associated with increased levels of farm-level record-keeping (sales, use of inputs), which may, in time, result in heightened levels of financial literacy and improved production efficiency as farmers become more aware of the impacts on profitability of undertaking certain practices. Participation in certification schemes may have unexpected impacts on the role of women in cocoa marketing, as men seem to be assuming a greater role in cocoa marketing through the new farmer organisation structures associated with certification. What we can definitely conclude is that this pilot assessment has reinforced the need for a more comprehensive methodological approach to assessing the impacts of interventions such as certification. We will therefore use the findings from this pilot activity to inform the methodological design of further studies planned for

5 1. Introduction Indonesia is an important producer of cocoa the third largest in the world after the Ivory Coast and Ghana. Indonesian cocoa is also a significant smallholder crop, with approximately 90% of cocoa grown by smallholders across about 1.5 million hectares (BPS, 2012). Within Indonesia, Sulawesi contributes approximately two thirds of total cocoa production. This production is concentrated across four provinces: South Sulawesi; Southeast Sulawesi; Central Sulawesi; and West Sulawesi (NAFED, 2010). Globally, value chain interventions such as certification schemes are becoming increasingly popular as an industry tool to meet sustainability commitments. In Indonesia, numerous interventions have already been introduced in the cocoa sector, including certification and verification schemes for sustainable production, such as Rainforest Alliance and Utz Certified. Certification schemes attempt to provide assurance to consumers about product origin, quality, safety, social or environmental attributes. In the process, they create a conduit for linking farmers with new markets, and an exchange of knowledge and development assistance along a value chain. In practice, certification schemes set farm management standards with which farmers and farmer groups must comply. While seeking to promote environmental sustainability and improve economic and social outcomes, certification is also being driven by a response to consumer demand for greater accountability, product traceability and sustainability. A challenge facing certification schemes is that their impact and effectiveness is variable or unclear, particularly the costs and benefits of compliance for smallholders (Giovannucci and Potts, 2008). This is not necessarily because benefits do not exist, but because the means to conduct a systematic assessment of impacts is still in development. Proponents of certification schemes and industry participants in the schemes recognise this, and are actively seeking solutions to ensuring that greater benefits flow to farmers and to the environment. As part of ACIAR project HORT/2010/011, Improving the sustainability of cocoa production in eastern Indonesia, we have prepared a trial set of survey indicators to assess the role of market-based certification schemes in encouraging improved farm practices and the provision of farmer support services. The main objective of this study was to assess the role of market based incentives and private sector certification schemes (eg. Rainforest Alliance, Utz certified) in facilitating knowledge transfer and shaping cocoa farmer behaviour (Keane, et al, 2010) as part of objective 4.2, HORT/2010/011 ACIAR project 1. The assessment relied on indicators of sustainability to be surveyed at the farm level. During July 2012, we conducted a pilot of the survey, with methods, results and conclusions outlined in the report below. It is our intention that this survey tool is refined and then scaled up for use at a number of locations throughout 2013, 2014, and ACIAR, Project proposal, Improving the sustainability of cocoa production in Eastern Indonesia through integrated pest, disease and soil management in an effective extension and policy environment, HORT/2010/011. 3

6 2. Methods The study approach has a number of similarities with an approach promoted internationally through the Committee on Sustainability Assessment (COSA). COSA is working towards the development and application of an internationally-recognized methodology and data gathering process to allow farmers and other stakeholders to more effectively measure and predict the outcomes, cost and benefits of the adoption of sustainability initiatives including certification (Giovannucci and Potts, 2008). Both our ACIAR pilot study and the COSA methodology apply an approach where farm-level indicators of sustainability are compared between a target group of certified farmers and a control group living nearby. In the design of this study, we liaised with COSA personnel and discussed our approach with them, and were influenced by their general approach towards the use of sustainability indicators. However, it should be noted that our method did not reflect the full use of the COSA methodology and was conducted independently of COSA. In order to conduct a short term pilot assessment, we modified the approach and emphasised the perceptions of cocoa farmers along with a more limited use of impact indicators. We chose to do this because very few studies of certification in the cocoa sector have been conducted anywhere in the world, and we decided it was necessary to undertake a phase of preliminary development and testing of an adapted survey. In conducting this relatively large-scale pilot, our objective was to identify improvements to develop a method more suited to the local producer context that could be conducted in subsequent years ( ). The data gathered in 2012 however, has also enabled us to provide an initial description of farmers experiences of the certification process. As described above, the key methodology used to analyse impacts was a farm-level survey that used various indicators of sustainability (the survey format can be found in Appendix A). This involved trialling the survey with 158 cocoa farmers in West Sulawesi in July The usefulness and effectiveness of these indicators is being reviewed and these preliminary results will help finalise indicators to be used from 2013 onwards. The draft indicators used in 2012, from which the following preliminary results were obtained, were prepared by the ACIAR team the authors of this report. Prior to conducting the survey, the indicators were shared and discussed with other organisations, including COSA and cocoa industry actors. The ACIAR team has continued discussions with COSA about closer collaboration in the future, and this report explicitly compares the ACIAR indicators with comparable questions within the COSA approach (Table 5) as a further step towards possible future collaboration. Certification programs in the Sulawesi cocoa industry are frequently introduced by the private sector, sometimes in collaboration with local Non-Government Organisations or development agencies. For the purposes of our survey, we worked together with an international cocoatrading company who has introduced a certification program in the Polewali-Mandar (Polman) District of West Sulawesi Province. Participating farmers were further supported by a local NGO, WASIAT, supported by USAID through their AMARTA program. In most areas where 4

7 certification schemes have been introduced, cocoa provides the main form of income, farmers are members of farmer groups, and farmers have participated in some degree of cocoa training in the past. We also targeted similar areas for our survey. Ideally, to avoid pre-existing selection bias, it is important to apply a randomization approach, randomly selected individuals from village populations who met the criteria. However, no database of individual farmer names existed. We therefore had to rely upon the lists of farmer groups. This list of farmer groups was randomized with the Microsoft Excel tool and farmer-group members were then selected randomly as respondents for this study. Two types of respondents were included: the control group and the target group. The target group meant farmers who are already certified. The control meant farmers who were not certified but who were similar to the target group in other criteria and could serve as comparable respondents. Introductory questions were used to ascertain whether the target and control groups were otherwise comparable, while subsequent questions were used to determine the possible effects of certification on the target group. Control groups are those who have not participated in the program of interest, but who are expected to have similar characteristics to those within the target group. Programs often target particular groups (for example those who have attended previous training sessions or who have good leadership) or areas (for example areas of high cocoa productivity or areas accessible by motorbike). Therefore the characteristics of participants may not always reflect the common characteristics of the broader village population. The presence of difference in criteria (called selection bias ) can affect a program s real impact. Consequently, this bias needs to be identified and minimized. Using a randomization method to select the group s respondent may help to address the issue. Before assigning random selection, identifying differences in variables in the population is crucial. Respondents can then be selected randomly based upon these variables. In this study, we identified that the certified groups had common criteria such as cocoa as a main source of income, previous participation in cocoa training and membership of a farmer group. Respondents from the control group were selected according to the basic criteria of having cocoa as their main source of income, although levels of training, farm size and wealth were found to be somewhat different between the target and control groups. At this stage, no statistical analyses had been performed to determine whether the control group was significantly different from the target group, although it is recognised that this will be essential for future surveys and analysis. 3. Results of the pilot survey It should be emphasised that the following results are preliminary and are based on a set of draft indicators that will be further reviewed. The primary purpose of this study was to trial a survey method, with a secondary purpose of analysing impacts from participation in a certification scheme. Results are presented here for initial discussions with relevant stakeholders only and should certainly not be used to judge the success, failure or otherwise of certification schemes. The survey format itself was organised into 6 sections: A. General information about the interview and respondent; B. Farm characteristics (includes exposure to 5

8 training and income sources); C. Economic Sustainability indicators; D. Environmental Sustainability indicators; E. Social Sustainability indicators; and F. Perceptions towards certification. The results for each section of the survey will now be discussed. A. GENERAL INFORMATION For administrative purposes, general information such as the name, age and village location of the participant was recorded, as well as survey date and time. This information is not included in this review. B. FARM CHARACTERISTICS B.I. Vital Statistics Information on household structure Information on farm and household characteristics was collected for both the control and target groups. This included data on mean family size, age of family members, percentage of gender composition, additional labour provided by family members, and family members who live elsewhere. This information is presented in Table 1. The primary purpose of these questions was to ascertain whether the baseline conditions of the target group were comparable to the control group. The basic characteristics of the control and target groups (age of respondents, family size, labour allocation to cocoa) were considered comparable, although women constituted a larger proportion of survey respondents within the target group (21%) than in the control group (only 7%), which could have affected responses. A slightly higher number of control households had family members working or living elsewhere, although this does not appear to be significant to the broader livelihood strategy of the household. Table 1. Household demography for the two respondent groups (target and control) Target Control Respondent Respondent Characteristics Respondent Respondent household household (n:76) (n:82) (N:351) (N:379) Family members (mean) n/a 4.6 n/a 4.6 Age (mean) Gender (Male) 78.9% 49.6% 92.7% 53.3% Work daily on farm 92.1% 34.2% 91.5% 38.3% Member lives elsewhere n/a 3.1% n/a 7.1% The official basic education system in Indonesia mandates nine years of schooling, consisting of 6 years in primary school (SD) and 3 years in junior secondary school (SMP). The education levels attained by farmers and their family members within both groups, aged between 17 and 30 years, were broadly comparable. It is assumed that all individuals aged 17 and over would have already had an opportunity to complete SMP, and those that haven t at that age are unlikely to subsequently complete the mandatory schooling. Education levels for those individuals aged over 30 are presented separately to identify generational change. These 6

9 results show a significant improvement in education levels within the younger generation for both groups. The target group is generally better educated than the control group for both generations. Figure 1. Education level of individuals in farm households Cocoa farm characteristics Table 2 presents information on farm condition across the two groups. Farmers often had more than one plot of cocoa in various sizes and tree densities. Our data is based on farmer estimates as it was not possible to accurately measure cocoa tree density per hectare. This would have required detailed information of tree numbers per. Farm condition varied between plots due to different levels of pest and disease infestation, tree density, soil health and farmers tendency to allocate their resources to the most productive sites. Based on total tree numbers for all plots per total farm size, cocoa farmers in Polman appear to have a relatively low density of cocoa trees, below 1000/ha, with the control group farms having a higher number of trees per hectare than target farms. The mean number of farm plots was similar between the groups, though the range varied from one to seven plots per farmer. This is expected to affect labour and time management in maintaining the farms. In addition, the average age of cocoa trees was relatively old - about 20 years - requiring more resources to sustain the farming business. Most farmers owned and cultivated their own land. 7

10 Table 2. Farm characteristics Characteristics Control (n:82) Target (n:76) Number of cocoa tree/ha (mean) Number of plot/farmer (mean) Average age of cocoa trees Total owned farm in hectare (mean) Producer owns unproductive farm 10% 32% Producer owns other crop farm 18% 43% Producer owns farms 98% 96% Share-cropping 2% 0% Farmer cultivates another s land 0% 4% Target group farmers also tended to own bigger farms, to cultivate cocoa less intensively, to have sufficient land to allocate to other crops, and to have more unproductive land. This is supported by the results presented in Figure 2. Target farmers were more likely to employ casual labour to work on their farm than control farmers, although the vast majority of farms exclusively relied on their own or family labour. Figure 2. Size of cocoa planting and total farm area. The general picture seems to be that that the control group is more dependent on cocoa cultivation than the target group, which may affect several sustainability indicator outcomes. This would indicate the target group is most likely wealthier than the control group. Such a conclusion is supported by the higher income estimates provided by the farmers (Table 3). 8

11 Table 3. Annual Income Estimates provided by farmers Average income (IDR) Median income (IDR) Control Group 4,962,492 3,080,000 Target Group 11,037,733 7,720,000 Unfortunately, this rather significant difference brings into question the comparability of the two groups and therefore the conclusions reached. The certification scheme may have been introduced to a group that was already more educated and wealthier than the surrounding community. Record Maintenance A high percentage of farmers were not keeping any records at all, especially for the control group. This certainly complicates the ability of a survey to obtain accurate information regarding costs and income from farmers as it based almost exclusively on notoriously unreliable recall data. The significantly higher percentage of certified farmers keeping records of crops sales and other records, such as inputs, (Figure 3) could also be interpreted as a positive impact resulting from the certification process. Figure 3. Farmer Record keeping Participation in training Farmers from both groups had participated in various cocoa-related training programs implemented by different organisations including private companies, development agencies and government over the years (Figure 4). In general, however, the target group had been exposed to substantially more private sector and international donor training, while the control group had received more government training and support. From the nature of the 9

12 survey question, however, it is not entirely clear whether the private sector / donor training was provided as a result of certification, or whether it preceded certification. For example, a local NGO, WASIAT, was engaged by one exporter to deliver training to farmers as part of the certification process. However, the same NGO was originally established as part of a USAID cocoa development program in the area. This substantial difference in the degree of exposure to donor and exporter training makes it very difficult to differentiate between impacts caused by the certification process compared to these other training-related interventions. Figure 4. Farmers who had participated in training by different organisations Figure 5 below shows how training focused on topics related to both on-farm and off-farm activities, marketing support, and the management of farm business and finance. It is clear from the responses that the broad topic of on-farm and off-farm practices (including Good Agricultural Practices and Integrated Pest Management, as well as post-harvest handling, such as fermentation and drying) was most common. This survey question should probably be revised to be more specific and to enable greater insights into the type of training provided. 10

13 Figure 5. Farmer training by topic according to provider (note: the length of each bar section represents the number of farmers who had recieved training on that topic by the identified provider) Only the WASIAT training (appearing under the exporter category), and implemented under the guidance of a certification organisation (Utz), included topics such as health, social and environmental issues. In this case, a master trainer from the certification organisation trained farmer trainers (who were members of the NGO WASIAT, as well as field facilitators from the export company, and from the government extension agency). As such, it was difficult to fully distinguish between training providers. These farmer trainers would then train representative farmers of each farmer group. In turn, the representative farmer would train other group members during regular group meetings and under farmer trainer supervision. The number of training days varied from one day to several weeks. The method of training was predominantly by theory and farmer field schools, with only a few smallholders having experienced a visit to another farm or a farmer field day. While the certified group was clearly exposed to more training, and a more diverse range of topics, this training was certainly not universally received amongst certified farmers. For example, less than a quarter of certified farmers had received non-production related training. Household Livelihoods Cocoa was the most important contributor to household income for farmers from both groups, followed by other crops and livestock. Farmers were asked to estimate the percentage of their income derived from cocoa and other sources. The average estimate of the contribution of cocoa-derived income from total household income was 83% for the target group and 79% for the control group. Figure 6 suggests that while most farmers obtained an income from more 11

14 than just cocoa, a far greater percentage of control farmers claimed to be 100% dependent on cocoa. Figure 6. Farmers claiming to be 100% dependent on cocoa income The most important secondary income source amongst both groups was tree crops (coconut, coffee, fruit trees, nuts etc). Target farmers had more varied income sources, and were much more likely to obtain an income from livestock than control group farmers (see Figure 7). Figure 7. Farmers obtaining an income from the different sources 12

15 Overall, target group farmers reported much higher total estimated incomes compared to the control group (see Figure 8). Figure 8. Estimated annual household income by income bracket C. ECONOMIC SUSTAINABILITY The economic sustainability of the cocoa sector is required to continue productive farming activities as well as to maintain quality of life. To develop indicators of economic sustainability, a range of parameters were used: production and post harvest activities; income and expenditure; quality and market access; and access to microfinance. C.I. Cocoa production, prices and competitiveness Production is a common indicator to determine improvements in the farming system. However, getting an answer to questions about a farmers production in the previous year was difficult. This is because farmers rarely keep records of their produce sales to middlemen or local traders (as shown in Figure 3 above). Nor is it standard practice for these buyers of cocoa to provide a receipt. This means that answers are reliant upon the memory recall of the farmer. Cocoa Production and yield Not all farmers are able to accurately estimate their cocoa production in the previous year (only 80% of the control group and 65% of the target group), although more were able estimate their cocoa harvest during the previous month (90% of the control group and 97% of the target group). The surveys were conducted in June through to August during the main harvest in Sulawesi. Target farmers reported much higher productivity during both time periods (Figure 9). 13

16 Figure 9. Farmer estimates of cocoa productivity (kg per ha of cocoa land) Post-harvest handling and marketing There was a very significant difference between the two groups in relation to post-harvest handling and marketing, suggesting that certification has been associated with a considerable transformation in this area. Figures 10 and 11 clearly show that most target farmers are drying their cocoa much longer (4-5 days) compared with most control farmers (1-2 days). Farmers receive significantly lower prices for wetter beans because of the high water content. Figure 12 shows that the certified target group is now overwhelmingly selling directly to the exporter rather than selling through more traditional local collector networks. However, target group farmers did also admit to selling some of their beans to local traders if they needed money quickly, or if the harvested volume was relatively small, or if they still had outstanding debts with the local collector. 14

17 Figure 10. Days drying of cocoa when sold and associated price (Control Group, n:82) Figure 11. Days drying of cocoa when sold and associated price (Target Group, n:76) 15

18 Figure 12. Marketing of cocoa beans by farmers Due to the lower moisture levels, target farmers received much higher prices than control group farmers. With each further day of drying, moisture content decreases and begin to approach the required 10% level. It is difficult to accurately estimate the moisture level from the survey based on number of days dried, as this will depend on a number of factors (drying yard materials, weather etc). Based on an assumption that cocoa dried for at least 5 days has reached the optimal moisture level (10%), and is therefore used as an index of 1, we have then adjusted the prices received for cocoa dried less than 5 days to take into account higher moisture levels and so obtain a more accurate estimate of prices received per content of cocoa 2. These adjusted prices are presented in Figure 13, and suggest that the prices received by target farmers are still significantly higher than in the control group. Interestingly, however the control group farmers who dried their beans properly for 5 days or more did actually receive a similar price to those in the target group. It seems likely that the direct trading relationship with the exporter has also contributed to the higher farm-gate price. This price data was collected during the period March July, 2012, and global price fluctuations might have influenced this data. According to data from the International Cocoa Organisation (ICCO, 2013), however, international process were relatively stable over the time period the survey was conducted. Monthly averages (in USD/ tonne) were: March (2,359); April (2,267); May (2,314); June (2,264), and July (2,350). The survey collected asked the date of the last sale, so it is possible to determine farm-gate prices as a percentage of prevailing world prices. 2 We have estimated that the price of cocoa sold after being dried for 4 days, should be divided by 0.95, for 3 days (0.9), for 2 days (0.8), and 1 day (0.6). This calculation is open to debate, but appears to be a best guestimate at this stage. 16

19 Figure 13. Farm-gate price of cocoa (adjusted to take into account moisture levels following drying) In addition to market fluctuations, bean quality, debt owed and absence of sales receipts, determining the influences on farm-gate price is further complicated because prices are subjectively determined by the collector. It is extremely difficult to differentiate the effects on farm-gate prices caused by variations in the marketing system from the effects of the certification process itself. Cocoa bean quality In order to access certified markets, good quality is required by the exporter, as well as good farm management practices. If farmers do not meet the bean quality standard, then the price is discounted. As explained above, discount prices can be applied because of excess moisture content, as well as for smaller bean size or the presence of waste and mould. Farmers were asked to rate the quality of their cocoa (Figure 14). Perhaps not surprisingly, most farmers in both control and target groups rated their bean quality as good for the various categories, although a significantly higher amount of control farmers answered don t know. This suggests that certification, or perhaps the altered marketing system, has been associated with an improvement in quality awareness amongst farmers. 17

20 Figure 14. Farmer assessment of cocoa bean quality Expenditure of cocoa income Farmers were asked how they used money from cocoa sales (answered were provided by 100% of respondents from both groups). Figure 15 Suggests that income was allocated primarily to education for each group, followed by daily needs, new house, motor bike, and new mobile phones. The purchasing of farm inputs was purportedly a low priority for both groups. Overall, the target group appeared to have a greater purchasing capacity than the control group, with more individuals spending cocoa income on a new house, ceremonies and furniture, although more control farmers bought a motorbike and a new mobile phone. Figure 15. Allocation of cocoa revenue to various activities (% of farmers who spent at least some of their cocoa income on each expense) Access to credit 38% of the control group and 57% of target group farmers confirmed they had obtained credit, and provided details of credit sources and amounts. Figure 16 suggests that the target group was more likely to access credit through a cooperative, while the control group accessed credit 18

21 through local collectors (tengkulak). The average amount of credit obtained within both groups from these sources was relatively low (less than 2 million Rupiah). Bank credit was also reported by around 10% of each of the groups, and was generally of much higher amounts. Figure 16. Farmer access to credit C.II. Involvement in Producer Organisations Benefits of participation Development programs are often implemented through farmer groups (kelompok tani) rather than directly with individual farmers. It was therefore anticipated that farmer groups are able to provide some sort of benefits to individuals. Farmers were asked if they were a member of a farmer group and, if so, what direct benefits they obtained from participation in the group. 100% of target farmers and 80% of control farmer claimed to be members of a farmers group. Farmers surveyed reported that benefits of farmer group participation include labour sharing, and the capacity to participate in the training and community meetings. Additional benefits were reported by the target group, who also participated in collective marketing and certification schemes (Figure 17). Target group farmers reported significantly greater benefits from participation in producer groups than control group farmers. 19

22 Figure 17. Benefits from being a member of a farmer group Role of women While farmer groups have become a vehicle for delivering development activities, equality in terms of an individual s right to group involvement is important the lack of which can indicate discrimination that may be embedded in cultural and social norms or religious practices. Figure 18 shows that male farmers of both groups were considered by respondents to be the most active in farmer groups, although more target group respondents (28%) felt that women were at least as active as men in the group. Figure 18. Farmer perception of which gender is more active in farmer groups In response to the question about whether there were woman leaders of any kind (association, head of village, school principle, etc) in the village, more respondents from the target group answered affirmatively (Figure 19). This could either demonstrate an improved awareness of 20

23 gender within the target group, or an actual greater number of women in leadership roles within the target group villages. Figure 19. Women in leadership roles C.III. Farmer perceptions of economic sustainability Farmers were asked to compare cocoa farming today to five years ago (Figure 20) and to provide their opinion on the future prospects of cocoa farming (Figure 21). The responses to both of these questions indicate that the target group is generally a lot more positive towards the increasing benefits associated with cocoa farming compared to the control group. It is possible that the recent introduction of the certification program, and its associated exporter linkages, have reinvigorated interest in cocoa farming amongst the target group. This outcome, of course, would be considered a substantial indicator of success for the corporate proponents of certification schemes, many of whom are concerned over long-term supply sustainability. Figure 20. Attitude towards cocoa farming today compared to 5 years ago 21

24 Figure 21. Perception of cocoa farming future prospects D. ENVIRONMENTAL SUSTAINABILITY D.I. Water Clean water is essential not only for drinking, but also for other household purposes. It is also used for agricultural activities. Water availability and access are therefore important indicators of environmental sustainability. In this survey, questions were asked about water sources, access and contamination. Water sources and access Groundwater from wells was the main source of drinking water for both groups, although more target group farmers were connected to a public water supply, again suggesting a greater affluence within this group (Figure 22). Figure 22. Sources of water Interestingly, in terms of access (Figure 23), a greater percentage of the target group had to travel further away to obtain water. In other words, the control group had more immediate 22

25 access to clean water, with only 2% with water more than 20 minutes away, compared to 17% of the target group. Figure 23. Accessibility of water source Preventing water contamination Figure 24 shows the different activities reported to have been undertaken by farmers in an effort to prevent water contamination. 56% of all control group farmers reported doing nothing to prevent water contamination, compared to only 18% of target group farmers. Overall, the target group respondents appear to have taken a greater number of efforts to prevent water contamination compared to the control group. While these claims were not verified, it suggests an enhanced awareness of the importance of preventing water contamination. Figure 24. Smallholder responses on activities to prevent water contamination 23

26 D.II. Use of agrochemicals Application of agrochemicals Two types of agrochemical inputs are commonly applied to cocoa trees: fertilizers and pesticides. In this survey, some form of pesticide or herbicide was used on almost every farm (96% of all farms). Previously (see Figure 15), farmers had said they allocated a small proportion of cocoa revenue to buying inputs. However, Figure 25 suggests a high percentage of farmers using synthetic fertilizers and pesticides. This may come back to an interpretation of revenue and the difference between ongoing farm costs and investments versus outright profit. Figure 25. Various brands of pesticides applied * Note: H= Herbicide; P= Pesticide/Insecticide; PA= Pesticide Adjuvant; F= Fungicide There were substantial differences in the types of pesticides used in the control and target group. For killing weeds, Gramoxone (active ingredient Paraquat) was used much more widely in the control group than the target group, which tended to prefer Supremo (Glyphosate). (Paraquat dichloride is an active ingredient found in Gramaxone that has been prohibited by some certification standard bodies.) Similarly, no use or highly reduced use of other pesticides such as Alika and Matador (both with active ingredient Lambda Cyhalothrin) were reported within the target group. Bento, Supremo and Vigor were both (relatively) commonly applied by both groups, possibly because these agrochemicals were distributed for free as part of the GERNAS program. Most farmers apply pesticides themselves, although the target group tended to hire more external contractors than the control group (Figure 26) again suggesting greater affluence. 24

27 Figure 26. Responsibility for spraying agrochemicals The need to restrict individuals from vulnerable groups (eg. pregnant woman and under age) from contact with chemicals through spraying was relatively well recognized by the members of both groups. In this survey, additional restrictions for old and sick people are accounted for in the category other (Figure 27). It is very difficult to ascertain the veracity of these claims. Figure 27. Vulnerable groups restricted from spraying agrochemicals Subjective responses to the question of whether farmers think they have been using more or less pesticides now compared to previously is reported in Figure 28. The target group generally reported using less pesticides now than 5 years ago, while the control group reported an increase use of pesticides. 25

28 Figure 28. Changes in pesticide usage compared to 5 years ago Almost all farmers from both groups reported having concerns over the negative effects of pesticides, although target farmers showed a heightened concern for environmental impacts (Figure 29). Figure 29. Perecieved negative impact of pesticides D.III. Soil More than 90% of cocoa farmers from both groups reported having sloping farm plots. However, Figure 30 below shows that more target farmer respondents claimed to have enacted soil control measures. 26

29 Figure 30. Smallholder responses on activities to prevent soil erosion Most farmers from both groups relied primarily on synthetic fertilizers to replenish soil nutrients (Figure 31), although more target farmers reported that they supplemented this with composting and manuring. Figure 31. Smallholder responses on activities to replenish soil nutrients D.IV. Attitudes towards the environment Interestingly, the perception of the importance of environmental issues was relatively similar for both groups (see Figure 32). One difference was that target group farmers were more concerned with solid waste/plastic management, while control group farmers were more concerned about chemical pollution - although they may not link this concern with their agrochemical practices. On-farm biodiversity seemed to be less important, while soil health and soil fertility was considered most important, followed by soil erosion and water quality in rivers. 27

30 Figure 32. Smallholder perception of the importance of environmental issues E. SOCIAL SUSTAINABILITY Social sustainability refers to the provision of basic needs, human rights and equity that contribute to creating a productive and humane working environment in the community. Four parameters were trialled here to assess social sustainability: 1) availability of basic public services; 2) basic human rights, worker rights and equity; 3) building smallholder capacity through training participation; and 4) smallholder perceptions. Availability of basic public services The affordability of proper basic public services is a key indicator of quality of life, particularly community health and education services. The most common accessible health service available around the village or hamlet were community health centres (PUSKESMAS) and paramedics (doctor or nurse), while public hospitals and private clinics were often available in nearby towns. Medical service costs were relatively affordable according to both groups, where the costs are either low and do not cause difficulty, or treatment is free (see Figure 33). 28

31 Figure 33. Opinion of basic medical service costs In terms of access to education, the distance of public schools from the place of residence was similar for both target and control groups (Figure 34). Figure 34. Distance of public schools from place of residence Despite this similarity in access to public schools, the target group spent more on education in the last year (2011) than the control group (Table 4), possibly suggesting a greater awareness of the value in education, or possibly greater disposable income. Table 4. Estimated expenditure on education in the last year Average (IDR) Median (IDR) Control Group 1,067, ,000 Target Group 1,904, ,000 29

32 Occupational health and safety is another component of social sustainability, with injuries and illness an important factor. In terms of on-farm injuries, the target group reported a total of 19 on-farm incidents in the previous year compared with the control having only 8. The inappropriate application of pesticides can potentially harm the health of smallholders. The use of protective clothing and equipment for the application of pesticides has therefore become a requirement of certification bodies. In measuring the use of protective measures, Figure 35 shows how more target farmers claimed to have used protective clothing and equipment than control farmers. Figure 35. Types of protective equipment claimed to have been worn by farmers when spraying E.I. Basic human rights, worker rights, and equity As an indicator of welfare, farmers were asked how many days they, or their family, went without rice in the previous year. No farmers reported going without rice for more than 10 days, and most (96% from both groups) reported going without rice for between 1 and 9 days in a year. Considering that few of the respondents actually grew enough rice to sell, and that most rice was purchased from the market, this suggests that few household across both groups were experiencing extreme poverty. Women on the farm Cocoa farming is commonly a family business where the on- and off-farm work is shared among the household members. Figures 36 and 37 show a strong gender division of labour for specific activities across both the control and target groups. While financial management and marketing is traditionally allocated to women, it is interesting that more men seem to have 30

33 become involved in this activity within the target group. It may be that the organisational requirements of collective marketing and certification have acted to increase the household role of men in this particular activity. It is unclear what implications this may have for the role of women in the division of on- and off-farm work. Figure 36. Household allocation of labour based on Gender (control group n: 82) Figure 37. Household allocation of labour based on gender (target group n: 76) E.II. Perception We assessed farmers perceptions on how their general quality of life had changed in the last 5 years and also how they felt environmental practices had changed (Figure 38). This second point was asked in terms of how their own environmental practices had changed, and then how they felt the environmental practices of others in their community had changed. There was little difference in terms of quality of life, with the control group having a marginally more positive attitude. It appears that the target group felt that their community s management of the environment had improved only marginally, although these results are inconclusive. 31

34 Figure 38. Smallholder perceptions regarding environment care and quality of life F. FARMER PERCEPTIONS OF THE CERTIFICATION PROGRAM The survey asked a number of questions specifically to those farmers participating in the certification program (the target group), particularly about their perceptions of change since implementation. There have been instances, elsewhere in Indonesia, where farmers were not aware that they had been included within a certification program. However, in this study, only 1 respondent (from 76 within the target group) claimed to be unaware that they were part of a certification scheme. Changes in farm practices All respondents claimed to have changes their on-farm practices in some way as a result of certification. They were then prompted about which particular farm practices had changed since becoming certified (Figure 39). Certified farmers claimed to have changed their practices most in relation to their use of safety equipment, increasing on-farm shade diversity and decreased chemical use. When asked if they had changed any other practices, a further five respondents claimed to have changed agronomic practices, such as pruning and sanitation, which should probably be included in future surveys. 32

35 Figure 39. Changes in farm practices since certification Changes in expenditure and costs Certified farmers were asked how they felt expenditure and labour allocation had changed since certification (Figure 40). Compliance with certification standards had contributed to a significant increase in time spent on training (acknowledged by 89% of farmers). A number of farmers (around 30%) reported spending more on hired labour, and the use of their labour, as a result of certification. Opinions seemed to be divided about whether certification had led to an increase or a decrease in expenditure on inputs, and whether it had led to an increase in paperwork and administration. This could perhaps be explained by the fact that each farmer group has an ICS (Internal Control System) team consisting of marketing and administration personnel who are responsible for sales and for managing group administration. The head of the farmer group is often responsible for both activities, and presumably these individuals are spending more time on administration, but this may not affect other individuals in the group. Figure 40. Changes in expenditure and time taken to comply with certification 33

36 Other Perceptions Farmers were asked about their perceptions of the impact of being involved in a certification scheme against a number of variables (Figure 41). In general, smallholder farmers had a positive response to certification, particularly in regard to economic aspects, including improved cocoa services, marketing, income, and better farm management respectively. Perceptions towards impacts on environmental conditions and environmental awareness were also positive, although less emphatically so ranging between somewhat agree and neutral. * Note: 1=strongly agree; 2=Somewhat agree; 3=Neutral; 4=Disagree; 5=Strongly disagree; 6=Doesn t know Figure 41. Farmer perception of participation in certification schemes 34