Participatory impact assessment of agricultural practices using the land use functions framework: case study from India

Abstract

What do different stakeholders think of the changing trends in agricultural practices and related policies? We answer this and related questions with respect to Karnataka, an Indian state showing signs of agrarian distress. Using the participatory impact assessment (PIA) method involving farmers, researchers and voluntary workers, we assess the impact of recent policy-driven farming practices. The land use functions (LUFs) framework, which resembles the ecosystem services framework, was adapted in the PIA to address multidimensional sustainability in agricultural landscapes. During the PIAs, participants ranked LUFs in the order of their perceived importance and projected the impact of different practice–policy scenarios on chosen indicators corresponding to each LUF. Three farming scenarios, namely organic, conventional (chemically intensive) and mixed input practices were assessed for their projected impacts on selected indicators of each LUF. The LUF ranking reveals that while stakeholder priorities vary, they remain contrasting to the common policy focus of profit and productivity maximisation. Farmers value family's health and water access the most and financial services the least as functions of their land. Indicator scoring in the PIA revealed that participants rated organic practices as the most beneficial, conventional scenario as detrimental and the now prevailing mixed inputs scenario as having little impact.

Keywords:

• participatory impact assessment
• agrarian distress
• small farms
• sustainability
• land use functions
• Indian agriculture

Introduction

Agriculture has always been an important source of livelihood in India, and today over 60% of its population depends on agriculture and allied sectors for their livelihood. In the 1960s low productivity, food shortages and rising food imports triggered the green revolution; a rapid, state-supported emphasis on production and consumption of synthetic inputs and hybrid varieties. The green revolution raised crop yields, but eventually resulted in problems such as land degradation, total dependence on external inputs (Evenson and Gollin 2003) and inter-farm and regional inequality (Sanyal 1983; Freebairn 1995). Although some productivity gains were sustained, the yield levels for most of the agricultural crops in India are far below yields in countries similar to India (Mahadevan 2003).

There has been a decline in public investment in agriculture, in rural economic growth (Dev 2004) and in the share of agriculture in gross domestic product (GDP), while the proportion of population dependent on agriculture for livelihood has remained unchanged (Suri 2006). Apart from problems with the weather, pests, storage and markets, farmers face risks posed by spurious seeds and pesticides, droughts and floods compounded by the absence of crop insurance schemes, poor extension services, lack of rural non-farm employment and health-care facilities (Dev 2006). Farmland is being increasingly fragmented; since 1970, the number of small and marginal farms (under 2 ha) has increased from 69% to 80% of all farms.

There are calls for the ‘unviable small farmers’ to sell or lease out their land (e.g. contract farming, see Gulati et al. 2008) and move to urban areas, allowing consolidation of holdings. This stance ignores the livelihood and cultural opportunities that the farm sector provides to a large percentage of the population. In addition, it seems to assume that farmers will be gainfully employed in the urban areas, although literature shows an increase in urban poverty levels and disparity (Sen and Himanshu 2004). Small farms in India also have higher productivity (per acre) than larger farms and a lower fertilizer imbalance index because of a more efficient use of inputs (Chand et al. 2011). Research has shown that small farmers are largely rational and respond to changes given the constraints and opportunities they face (Lutz 1998). Nevertheless, there has been an increase in farmer distress and suicides (for a detailed analysis, see Rao and Gopalappa 2004; Vasavi 2009). Despite the apparent need for attention for small farming, whether to support and sustain small-holder farming remains an open question in India, even academically.

Policy responses to the problems in agriculture have typically included subsidies for chemical fertilizers, loan waivers or relief packages leading to various social and ecological consequences. Industrial fertilizers subsidised by the central and the state governments have led to soil nutrient imbalances and consequent soil degradation (Chand and Pandey 2008; Kapila 2009). On the other hand, recent trends in provincial policies towards organic inputs appear to locally incentivise sustainable farming. In the context of such conflicting policies amidst agrarian distress, a holistic academic attention on the challenge of integrating food production with social and environmental issues is absent. The disconnect between farmers' needs, agricultural research and policies loom large (Pretty et al. 2010), although there is sporadic literature on the democratisation of agricultural research (Pimbert and Wakeford 2002; Pimbert 2004). Documented studies linking specific policy-driven practices and primary stakeholders are rare to find.

This article aims to explore how primary stakeholders' preferences and projections differ from those of policymakers. Using a systematically executed participatory method, we explore (1) which land use functions (LUFs) farmers and other stakeholders consider as important and (2) how different farming scenarios will impact these LUFs. The rest of this article consists of describing the study area (Section 2), a review of participatory approaches (Section 3), the approach adopted in this study (Section 4), results (Section 5), discussion (Section 6), ending with concluding remarks (Section 7).

Study area and the practice–policy context

This article focuses on the state of Karnataka in India, where roughly 56% of the population depends on agriculture for their livelihood (Directorate of Economics and Statistics 2006). Situated in south-western India, Karnataka registered an average annual economic growth of 7.4% in state GDP between 2000–2001 and 2010–2011. Despite its high economic growth, Karnataka's agriculture (mostly drought prone, only 28% of cultivable area is irrigated) remains plagued by declining soil fertility and ground water levels, increasing water salinity, soil erosion, loss of agrobiodiversity (Government of Karnataka 2003; State … 2011), stagnating yields (Rao and Gopalappa 2004), as well as indebtedness and marginalisation (Vasavi 1999; Deshpande and Prabhu 2005). Between 1999 and 2011, the contribution of agriculture to the state's GDP declined from 30% to 12%, and 28,937 farmers' suicides were recorded in the state (Report of the National Crime Records Bureau 1995–2011). These suicides were often driven by crop failure and inability to repay loans.

Agricultural practices currently prevailing in Karnataka can be categorised into farming with conventional, mixed and organic inputs. Conventional farming mainly involves the use of synthetic inputs popularised during the green revolution; mixed input farming is characterised by the use of both organic and chemical inputs, while organic farming, a resurgent practice in some areas, relies mostly on in situ (farm and the neighbourhood) generated inputs. This organic farming is mostly non-certified and targets the general open market without a price premium. Organic farming has been gaining attention in agricultural policy circles in the state with policies and schemes like the Karnataka State Policy on Organic Farming and the Karnataka State Organic Farming Mission, while huge subsidies to industrial fertilizers continue. In its 2010–2011 budget, Karnataka increased the allocation for organic farming to INR 1 billion, while INR 27 billion (INR 2107 per ha of gross cropped area) was spent on chemical fertilizers used in the state during 2007–2008.

During a workshop held in 2008, farmers, researchers and government officials discussed temporal changes in land use, cropping pattern and farmers' suicides in the state. This workshop was organised to help sharpen our approach and select districts to focus on. Five districts (Bijapur, Chitradurga, Chikballapur, Mysore and Udupi) that cover the major agroclimatic zones of the state (see Figure 1) were chosen. Three of these districts (Bijapur, Chitradurga and Chikballapur) are located in the dry zones of the state, one (Mysore) in the transitional zone and one (Udupi) in the coastal zone. Unlike the other districts chosen, Udupi is characterised by heavy rainfall and a relatively high human development index (HDI) (Table 1, Figure 1).

Table 1. Characteristics of districts studied

	Bijapur	Chitradurga	Chikballapur	Mysore	Udupi
Geographical area^a (ha)	1,053,471	770,702	404,501	676,382	356,446
Average size of operating land holding^a (2005–2006) (ha)	3.03	2.05	1.15	0.97	0.78
Net sown area^b (2008–2009) (%)	71	51	26	49	26
Net irrigated area^b (2008–2009) (%)	35	19	25	47	33
Fertilizers distributed^b (N + P + K, kg/acre of net sown area) (2009–2010)	92	72	221	229	69
Population density^c (2011) (per km^2)	207	197	298	437	304
Farmers' suicides^d (2003–2007)	408	1058	241	383	192
Notes: Chikballapur is a new district formed out from Kolar district in 2007; hence, these data are available only for Kolar.
Directorate of Economics and Statistics (2006 ^a, 2011 ^b); ^cCensus of India 2011; and ^dState Crime Records Bureau, Government of Karnataka.

Figure 1. Selected districts in the state of Karnataka, India, with information on the agroclimatic zones and HDI.

Frameworks for participatory impact assessments

Farmers' knowledge has been acknowledged as valuable for agricultural research (Dore et al. 2011). If scientific innovations were to help Indian farming in the long run, it would be best for farmers to provide information on what they prefer, what is feasible in their context and what conditions the technology needs to fulfil (Röling 2009). Growing public mistrust in the ability of government institutions and scientific expertise to cater to farmers' needs (Pimbert and Wakeford 2002) has generated interest in participatory methods to assess what can be done in this sector. Such methods can be employed as stand-alone tools to clarify the ground relevance of certain policy or scientific questions or can be used in conjunction with other methods, like multi-criteria analysis (e.g. Zendehdel et al. 2008; Purushothaman et al. 2012) or an interactive community forum (Becker et al. 2003). Our question on farmers' priorities and their perception of the impact of diverse practices needed a participatory method that can address questions on multiple dimensions and multiple stakes in agricultural practices. The role of such structured deliberative processes in constructing preferences has been instrumental in environmental policy research (Hermans et al. 2008).

We considered two frameworks suitable for the study: the ecosystem services (ES) framework (Millennium Ecosystem Assessment 2005) and the LUFs framework (Pérez-Soba et al. 2008). LUFs are defined as those goods and services that are produced through land in its interaction with the geophysical and sociocultural endowments. The concept of ES has been developed as a way to represent and value the dependence of human societies on natural ecosystems. Although comparable to the ES framework, the LUF framework is more adaptable to modified ecosystems in general and a participatory assessment in particular, as it enables the notion of multidimensional sustainability among a diverse set of indicators.

A comparative review of these frameworks by Schößer et al. (2010) concludes that the wide premise of the ES framework lies mostly within an ecological perspective, while LUFs can explain the link between land use change and sustainability issues. The ES framework has been generally used for monetary valuation of goods and services from ecosystems (de Groot et al. 2002; Tallis et al. 2008) that flow from a natural landscape to a dependent socio-economic system. The LUF frame facilitates exploring nature's services within a modified/production landscape for possible impacts and trade-offs among different economic, social and environmental dimensions. The LUF framework appears more effective in communicating and/or sharing outcomes with policymakers, especially in agriculture. Further, interactions with farmers based on the ES framework would have been difficult given the bundled nature of different ES (that makes a response on each service difficult) and diverse expectations from production land use. Many aspects that farmers consider as important services or benefits from their land, like providing employment options, access to a market economy, an asset base for financial capital and occupational preferences, may not be successfully addressed in the ES frame.

LUFs have earlier been employed in participatory assessment of agricultural problems (Zhen et al. 2009; König et al. 2010, 2012), using the framework for participatory impact assessment (FoPIA). FoPIA provides an inclusive process for the assessment of policy impacts at regional level. It follows a set of methods and procedures that harness the knowledge of national, regional and local stakeholders (Morris et al. 2011). In this approach, stakeholder participation is key in scenario impact assessment while enabling the social learning process among different stakeholder groups. Even though stakeholders' perceptions may be varied or even conflicting, this process leads to a critical engagement with their own position and facilitates consensual change in line with deliberative processes (Hermans et al. 2008). This can be used to generate wholesome knowledge about the impacts of proposed policies on sustainable land use. For example, König et al. (2010) applied FoPIA to assess the impacts of spatial zoning policies on rural–urban sustainability in Indonesia, and König et al. (2012) used this framework to assess a possible expansion of soil and water conservation measures on rural sustainability in Tunisia. In both cases, the exercise presented the regional problem and elicited new as well as implicit insights to support regional policy implementation.

We chose and adapted the LUF-based participatory approach to engage local stakeholders in the impact assessment process and to balance the assessment among social, economic and environmental dimensions.

Methodology

The participatory impact assessment (PIA) used in this study employs primary stakeholders to assess the ecological, economic and sociocultural sustainability of farming as impacted by different practices. LUFs, corresponding indicators, the range of LUF ranking and the scale of indicator scoring used in FoPIA were modified to suit the study context.

A set of 11 sociocultural, economic and ecological LUFs and associated indicators relevant to the study were defined (Table 2). Indicators were selected so as to represent the LUFs while being easily perceivable in the local context. An initial set of LUFs was adapted from Peréz-Soba et al. (2008) and Reidsma et al. (2011) and further elaborated to suit regional conditions through expert consultation, field observations and internal discussions. During PIA workshops this set of LUFs was presented to the group of participants and further specified and adapted based on deliberations. We made a presentation of these LUFs to the participants to facilitate interpretation of scenario impacts. Table 2 shows the framework of LUFs and indicators finally applied in the PIA.

Table 2. LUF and indicator FoPIA

Sustainability criteria	LUF	Indicator
Ecological	Water	Soil moisture
	Biodiversity	Agrobiodiversity
	Ecosystem processes	Soil organisms
Economic	Land-based production	Crop yield
	Infrastructure	Access to markets
	Economic production	Net income
	Financial services	No overdue loans
Sociocultural	Provision of work	Work satisfaction
	Family health	Visits to the doctor
	Food security	Food from farm
	Social capital	Collective activities

Unlike FoPIA workshops that confined to national stakeholders and policymakers, the workshops in this study primarily involved local stakeholders. Exploratory visits and interactions in the study districts helped us to prepare a list of about 140 invitees from four categories: farmers, voluntary sector or non-governmental organisation (NGO) representatives, researchers and officials of the Department of Agriculture. Out of 140 invitees, 97 (70%) accepted our invitation to participate in PIA workshops. Of these over 75% were farmers, and the rest comprised of an equal number of scientists and representatives of local NGOs and a very small representation of government officials. The proportion of invitees who attended the workshops was also highest among farmers (83% of those invited), followed by NGO representatives and researchers (over 70% of invitees each), while only 10% of the invited government officials (2 out of 20 invitees) attended the workshops.

PIA workshops for the study sites were conducted in vernacular (Kannada). They started with a discussion on the past and current situation of farming in Karnataka in general and the concerned study districts in particular. The research team was responsible for translation and presentation of information in Kannada, translation and analysis of responses, as also moderating the workshop, ensuring a smooth conduct avoiding domination of opinions while allowing a freewheeling discussion. Some participants without formal education required assistance in writing down their responses. The entire process of the PIA was explained to participants in three main steps: (1) scenario description, (2) assigning weights to LUFs and (3) indicator scoring for impact assessment.

Defining scenarios for assessment

The scenarios for impact assessment were envisaged after a time period of 5 years from the assessment year. The period of 5 years is the general time duration adopted for national policy planning and government tenure. ¹ The identified scenarios, following exploratory field interactions, literature and policy documents (Table 3), were presented in the PIA workshops.

Table 3. Scenarios for the PIA

Scenarios (farming with)	Practices
Organic inputs (S1)	• No use of chemical fertilizer and pesticides
	• Use of organic inputs
Mixed inputs (S2)	• Use of both organic and chemical inputs
Conventional inputs (S3)	• Use of chemical fertilizers and pesticides
	• Little or no use of organic inputs

S1 refers to farming with organic inputs using mostly in-farm or locally available resources. S3 involves the application of chemical fertilizers and pesticides popularised during the green revolution and S2 refers to the use of both organic and chemical fertilizers, a practice currently followed by many farmers, especially those trying to make a conversion to completely organic cultivation. Participants were quick to understand the three practice scenarios and the temporal implication of projecting them from the analysis year 2010 to 2015.

Assigning weights to LUFs

Participants were asked to prioritise LUFs by attaching weights to each of them. Each LUF was weighted separately on a range from 0 to 5; the same weight could be attributed to more than one LUF. The results of this exercise were presented to the participants and extreme weights were highlighted and discussed. This provided a chance for participants to modify, if convinced after deliberations on the reasoning behind attributed weightages. This step resembles the Delphi approach; the main difference is that in Delphi, experts do long-term forecasting of changes for which information (including information on who conducts the exercise) is scarce.

Impact assessment through indicator scoring

After prioritising the LUFs, participants discussed and modified a tentative list of indicators corresponding to each LUF, coming up with a set of perceivable and consensual measures. Although measurability was not an explicit criterion, only 1 (work satisfaction) out of 11 selected indicators was intangible.

As indicated earlier, the agenda of the PIA was to make participants assess the impact of practice scenarios on indicators in 5 years' time from the study year. Participants were asked to indicate how they thought the indicators would be affected in the three scenarios in 2015 and give them corresponding scores – ‘no change’ (0), ‘much worse’ (–2), ‘worse’ (–1), ‘better’ (1) or ‘much better’ (2). We used a range of scores with five points (between –2 and +2), as participants found a wider range difficult to work with. At the end of each session, results were presented to the participants and all scores, especially extreme scores, were highlighted and discussed to share the reasoning behind extreme scores.

Results

As described in Sections 4.1–4.3, we elicited and collated the most preferred LUFs and subsequently the expected impact of farming scenarios on selected indicators in PIA workshops. Preferences expressed by participants for different functions of their land are discussed in two sets: stakeholder-wise (Section 5.1) and site specific (Section 5.2). The impact of farming scenarios on selected indicators is presented in Section 5.3.

Preference for LUFs by different stakeholders

Preferences elicited from all participants from the study districts were pooled and classified into three groups of stakeholders: farmers, NGO representatives and researchers. ² Figure 2 depicts the average rank for each LUF attributed by these stakeholder groups across selected districts.

Figure 2. Mean weightage attributed to LUFs by participant groups.

Farmers and NGO representatives preferred family health and soil water content, while researchers gave food security and land-based production the highest importance. Financial services were given relatively low importance by all the three groups. Groups differed the most in terms of provision of work and biodiversity for which, compared to the other two groups, researchers attributed the lowest and highest preferences, respectively.

Site-specific preference for LUFs

In order to capture location-specific characteristics affecting LUF preferences, we segregated the weights between the districts studied. Table 4 presents the mean weights for each LUF attributed by all participants from each study district. Family health is given the highest importance in three districts and the second highest in one district. Udupi (the district with high HDI and good rainfall) is an exception with infrastructure and water as most preferred LUFs. Food security is scored as very important in Bijapur and Chikballapur, reflecting the subsistence need and remoteness of these sites. Provision of work emerges very important in Chitradurga and land-based production in Mysore.

Table 4. Mean weightage attributed to each LUF by participants in the study districts

Criteria	LUFs	Bijapur	Chitradurga	Chikballapur	Mysore	Udupi
Sociocultural	Provision of quality work	8.32	10.53	8.64	9.51	8.92
	Family health	11.10	10.91	10.05	10.56	9.39
	Food security	10.89	8.79	10.75	9.86	9.20
	Social capital	8.11	8.12	8.18	9.15	7.61
Economic	Land-based production	9.71	8.65	9.11	10.56	8.83
	Infrastructure	7.68	9.78	9.35	8.80	9.95
	Economic production	8.32	8.98	7.94	9.15	8.83
	Financial services	7.36	8.76	7.48	6.69	8.26
Environmental	Soil water content	9.82	8.94	9.58	9.86	9.95
	Biodiversity	8.32	7.57	8.88	8.45	9.48
	Ecosystem processes	10.35	8.97	10.05	7.39	9.58
Note: The four most important LUFs per district are highlighted in bold.

Impact assessment by indicator scoring

The impact of the three scenarios on selected LUFs was assessed by the participants through indicators. The indicator scores were classified into three participant categories. The pattern of assessed impacts across study districts was uniform, and hence we present only the difference in impacts revealed across stakeholder groups. 3–5 depict the average indicator scores in the S1, S2 and S3 scenarios, respectively, given by the participant groups.

Figure 3. Organic input use scenario (S1) impact scores on indicators assigned by participant groups.

Figure 4. Mixed input use scenario (S2) impact scores on indicators assigned by participant groups.

Figure 5. Conventional input use scenario (S3) impact scores on indicators assigned by participant groups.

The projected impact of the S1 scenario with organic inputs is positive on all LUF indicators (Figure 3). The highest positive impacts are projected by the NGO representatives, followed by farmers and researchers. Projected impacts on indicators do not diverge much between groups except for ‘Visits to the doctor’ (representing the LUF ‘family health’) for which farmers attributed lowest scores for S1.

In S2, with mixed inputs use, the impact on LUF indicators is small, although mostly positive (Figure 4). All groups assigned negative impacts for the highly valued LUF family health in S2.

In S3, with chemical inputs use, impacts are projected to be negative by all groups (Figure 5). The highest negative impacts are expressed by researchers, particularly for the three ecological indicators and for the indicator of family health.

Discussion

In the previous section, we summarised the differences and similarities across participant groups and study sites with respect to prioritised LUFs and expected impacts of different farming practices on indicators of these functions. The interpretation of results and methodological lessons are presented in the following subsections.

Interpretation of results

The three most important LUFs according to farmers were family health, soil water content and food security; for NGO workers these were family health, soil water content and quality of work, while for researchers these were food security, land-based production and biodiversity. These results imply how farmers, especially small holders (average holding size of participant farmers was 1.76 ha) preferred other than just financial gains from their land. This prioritisation should be considered in tandem with the rising food prices and volatility of agricultural prices (see FAO 2011 for analysis of food inflation and Sekhar 2004 for price volatility in Indian agriculture). In fact, profit and financial services (as also collective activities) emerged as low priority functions among the participant groups. It is interesting to note that these results are similar to what was found in the context of certified organic farming for premium price by Bhatta et al. (2009) using Strengths, Weakness, Opportunities and Threats (SWOT) analysis.

A notable divergence was found in the case of agrobiodiversity, which was relatively highly prioritised by researchers compared to farmers and NGO representatives. This was different from other contexts. For example, Ninan and Sathyapalan (2005) used revealed and stated preference methods and found that biodiversity was valued by locals. The results from the PIA imply that researchers link biodiversity and food security (see Chappell and La Valle 2011 for a detailed review of studies linking the two), unlike other stakeholder groups.

Looking closely at the results across the three stakeholder groups together, the LUFs that emerge as most important (family health, soil water content and food security) are also part of prioritised LUFs elsewhere in other studies (König et al. 2010; Morris et al. 2011). If we look at results across the study sites, the prioritised LUFs remain the same. However, especially in the relatively developed district (Udupi), ecological LUFs (ecosystem processes and biodiversity) along with provision for infrastructure also emerge as important.

Among selected indicators for impact assessment, the ecological indicators (soil moisture, agrobiodiversity and soil fertility) were perceived to be impacted most positively in S1 (with organic inputs) by the participant groups. Apart from these indicators, the positive impact on food security was also expected to be high implying that participants believe in the synergy between indicators in S1. Quantitative assessments in the same district also suggest this (Patil et al. Forthcoming 2012; Purushothaman et al. 2012). Although researchers in the PIA opined that organic agriculture may not be able to feed the burgeoning human population, there are studies showing that alternate farming methods can provide food security comparable to that of conventional agriculture (Chappell and LaValle 2011).

Participants emphasised better work satisfaction from organic practices, which resonates with the discussion in Rickson et al. (1999). Occupational preferences are often ignored in the development and popularisation of modern fertilizers and plant protection technology, although their negative impacts have been noted widely (Hart et al. 2004; Bandara et al. 2010). PIA participants considered reduced input costs and health benefits as major impacts from organic practices, which was also found in other parts of India (Panneerselvam et al. 2011).

Farming with mixed inputs (S2) was expected to result in almost no change from the current situation by all participant groups, although farmers thought that impacts of S2 could be positive for some indicators like crop yield, net income and collective activities. Researchers projected that this scenario had mainly negative impacts. The use and handling of pesticides is known to be unscientific (see Bag 2000; Devi 2010), although the used quantities are low compared to other countries (Gupta 2004). Farmers reported more nausea, headaches and skin rashes, and thus a negative impact on family health, if they continue working with chemicals. Hence, across the stakeholder groups, family health (indicator: visits to doctor) was considered to be most negatively affected in S3. According to the farmers, the ease in using chemicals was an attraction, although it came with a price.

Despite the earlier cited negative impacts of S3, all stakeholder groups noted the potential improvement in economic indicators in S3. However, these economic benefits were attributed low priority as LUFs across participant groups. Farmers reasoned that economic gains from S3 were at best short-lived and pinpointed the problem of pesticide resistance leading to increased usage of more toxic chemicals and further impacts akin to a technology treadmill (Kumar 2012).

Despite the expected positive impacts of organic agriculture compared to other practices, the results do not imply that farmers will definitely continue with (or convert to) organic practices. This is because of reasons like easy access to heavily subsidised chemical inputs, inadequate supply of organic inputs, as well as lack of price incentives and market for organic products.

Methodological reflections

Systematic organisation of the PIA workshop is important in order to get unbiased responses. Responses in the workshop can be biased if the group is small. Yet, a large group can prove chaotic, and so it is hard to pin down the optimal size within and between groups (Reed et al. 2009).

It is also likely that stakeholders from different backgrounds have opposing views, making the results skewed or inconclusive. It may be possible to shift the overall score simply by changing the sampling frame of participants, but to an extent this could be objectively managed by the organisers. The PIA workshops require careful preparation from the matters of location of the workshop to the range of weights and scores to be adopted.

Another methodological issue is the functional literacy of participants. Participants with minimal formal education will have difficulty in comprehending scenario development or matters of scale, unless the organisers have good communication skills and objective approaches. In this PIA, farmers were keen to give high weights to all LUFs. Initially, they were reluctant to prioritise one over the other. We also realised the need for good moderation by a neutral, firm and local researcher with good communication skills to contain peer pressure or influence by more vocal participants.

Conclusion

The disconnect between farmers' needs, agricultural research and policies loom large in India, despite sporadic literature on democratisation of agricultural research. Documented studies linking specific policy-driven practices and primary stakeholders are rare to find. Using a systematically executed participatory method, we explore (1) which LUFs farmers and other stakeholders consider as important and (2) how different farming scenarios will impact these LUFs.

We chose a LUF framework to design the PIA, so that many aspects that farmers consider as important functions (or services) from their land like providing employment options, access to a market economy, an asset base for financial capital and occupational preference may be addressed. A carefully designed and conducted PIA with its easily visualised and deliberated outputs can usefully engage farmers, NGO workers, agricultural researchers and government officials.

The study was characterised by the involvement of stakeholder groups in all stages of the assessment process: selecting policies, LUFs and indicators, prioritising LUFs and assessing and discussing impacts on indicators. Conflicting views among stakeholders turned out to be educational for all participants including us. While researchers prioritised food security and biodiversity, farmers and NGO representatives prioritised healthy living conditions and soil water content. Nevertheless, a converging conclusion of the exercise was that organic practices can sustain small farming as a livelihood option, even though participants had different priorities and perceptions on how exactly this would happen. This article points towards the socioecological dimensions of agriculture prioritised by farmers that are ignored by most policy assessments. Given the priorities elicited from the grassroots using the PIA, if sustaining small-scale agriculture is an objective, Indian agricultural policies should be addressing social and ecological concerns.

Acknowledgements

We gratefully acknowledge support from the EU-FP6 project ‘Land use policies and sustainable development in developing countries (LUPIS)’.

Notes

1. Thus, this could easily be related to policy assessment by the participants. Long-term scenarios for impacts did not sound reasonable to participant farmers, given the increasing uncertainties of market for agricultural commodities and climate, as also considering the increasing value of land especially in the peri-urban villages.

2. The two observations from government officials were merged into the closest (in terms of experience and interest of the participant) other categories.