The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, 46

70 A Study of Farmer’s Decision and Incentive Scheme to Reduce Highland Maize Farming: the Case of A Study of Farmer’s Decision and Incentive Scheme to Reduce Highland Maize Farming: the Case of Nan Province Sittidaj Pongkijvorasin and Khemarat Talerngsri A Study of Farmer’s Decision and Incentive Scheme to Reduce Highland Maize Farming: the Case of Nan Province1 Sittidaj Pongkijvorasin2 and Khemarat Talerngsri3

Abstract

During the past 10 years, there has been a considerable increase in highland maize farming in Nan. The increase in farming area causes an excessive forest encroachment and environmental problems, e.g., soil degradation, landslide, and pesticide contamination. One of the main reasons for this is the rapid expansion of maize farming area owing to the drastic increase in maize price, together with the government’s ineffective enforcement in forest conservation. Attempts made both at the national and local levels in the past to reduce highland maize farming have not shown much success. Maize farmers do not find it easy to leave their accustomed cultivation practice on highland in search of a more eco-friendly alternatives. Several reasons can be used to explain this sluggishness. For most highland farmers, once engaged in maize farming, they got locked into cycle of informal debt due to maize’s high production cost. Without sufficient and concrete incentives to induce change in farming behavior, it seems highly unlikely that the farmers would be able to reduce highland maize farming. The government policy to directly intervene or guarantee maize price, although intended to help relieve farmers’ endless problem of poverty, has indirectly fueled the maize farmers to expand their farming land and get locked into the vicious cycle of highland maize farming even deeper. Nonetheless, there are some areas that not only managed to resist the alluring market force of maize, but also found way to rehabilitate and conserve forests using local participation as the most vital mechanism. The Nam-Meed watershed has beautifully succeeded in the reforestation and reduced highland maize farming. The engagement of farmers in specifying and enforcing forest rule and regulations, coupled with appropriate role and timely supports by the officials of Nam-Meed Watershed Management Unit are believed to create profound impacts on the farmers’ attitudes toward conservation. In a rather different context, the Sop- Sai watershed has also managed to trigger a transformation from highland maize farming to reforestation. Since 2009, the area has received assistance in form of large-scale incentive scheme from “Pidthong Lungpra” foundation whose aims were to alleviate poverty problem, improving farmers’ livelihood and reduce deforestation in watershed area. This incentive

1 We would like to thank Thailand Research Fund for financial support; and Knowledge Network Institute of Thailand for facilitating the project. We are also highly indebted to Assoc.Prof.Somporn Isvilanonda for his valuable insights and useful suggestions; Bundit Chimchart, Nammeed Watershed Management Unit, for being a source of inspiration; and Wittawit Foojitnirun for his tireless assistance throughout the project. 2 Faculty of Economics, Chulalongkorn University, [email protected] 3 Faculty of Economics, Chulalongkorn University, [email protected]

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 47 mechanism has started to show significant results in changing farmers’ highland farming behavior, as they deserted maize farming to grow perennial economic trees. This research interviewed and surveyed data from 6 villages in Nan: 3 villages in Nam- Meed upper-watershed and 3 villages in Sop-Sai upper-watershed, so as to understand the fundamental mechanism and common factors which drove a sustainable natural transformation from maize farming to reforestation. The objectives of this research are to 1) demonstrate the possibility of a win-win solution in which farmers have higher incomes while forest is also restored; 2) analyze the effectiveness of various subsidy schemes; 3) explore important factors influencing farmers’ decision in reducing their highland maize farming; and 4) suggest plausible policy recommendations which lead to a reduction in highland maize farming. We found that the fundamental mechanism driving a sustainable transformation from highland maize farming to reforestation comprises 3 pillars; these are 1) adequate returns from the alternative practices; 2) a genuine love and attachment towards the forest and 3) effective enforcements of communal rules and regulations. These pillars essentially reinforce one another and are crucial factors for achieving the win-win solution in the long-run. According to surveyed data, alternative practices which can be used as a substitution for maize farming in highland can be either crop rotation in flatland, tree plantation in highland (economic forest), or reforestation in highland. With proper supports, these alternatives potentially generate higher income than maize farming in highland. In other words, farmers naturally quit growing maize in highland in search of higher income from a more forest-friendly alternative. To opt for any of the three alternatives or their mixture depends largely on the areas’ geographical and social conditions. While investigation of the natural transformation in Nam-Meed area highlighted the availability, allocation, and utilization of flatland as key driver of the change, such requirement does not always hold in the area where a rapid transformation was called for, like in Sop-Sai watershed. Subsidy schemes were used to supplement returns from new alternatives and to overcome limitations (e.g. geographical limitation) during transitional period. Using survey data from the area, we found that farmers of diverse socioeconomic backgrounds valued subsidy schemes differently. In order to successfully incentivize farmers’ behavior, the effective subsidy must take into account farmers’ socioeconomic characteristics and the area’s limitations. Our findings do not support a “one size fits all” subsidy approach. The first and foremost policy recommendation for the government is to send clear and strong signal of her commitment to solve the issue of deforestation in highland. Coherent directions of both agricultural policy and natural resource management are urgently needed. While the agricultural policy can aim to support alternatives that are contingent on ecological richness, the policy from a natural resource management standpoint cannot completely detach forests from local or indigenous people. Local community which conserves forests should be allowed not only with a balanced use of the forest, but also a communal right to regulate such use. In addition, collaborations from private sector in sending clear market signals are

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 48 considered vital in supplementing government’s efforts. Price mechanism as well as maize standards are examples of channels for appropriate signals. JEL Classification Codes: Q15, Q18, Q23, Q28 Keywords: Incentive scheme, maize farming, deforestation, highland, Thailand

Introduction

Deforestation of a colossal scale that has been happening in Nan, a mountainous province in the North of Thailand, over the past ten years has triggered alarm to not only Nan citizens but also environmentalists, academics, business sector and policymakers from all parts of the country. The unique characteristic of Nan as one of the largest upland watershed area of the country has made attempts to address the deforestation issue a matter of urgency. In the year 2003, Nan forest covered an area of 0.84 million hectares; that is 74% of Nan’s total land. Within a period of 9 years, the forest cover has decreased to 0.74 million hectares or 61% of Nan’s land. Interestingly, maize farming area had tripled from 46,934 hectares in the year 2006 to 136,355 hectares four years after. The latest statistics (year 2014) shows that the maize farming area in Nan now accounts for almost 10% of the country’s maize farming area. In terms of area, Nan, the province of very limited flatland, has become the second largest maize producer, next to Phetchabun (Office of Agricultural Economics, 2014). Attempts made both at the national and local levels in the past to reduce highland maize farming have not shown much success. Maize farmers do not find it easy to leave their accustomed cultivation practice on highland in search of a more eco-friendly alternatives. Several reasons can be used to explain this sluggishness. For most highland farmers, once engaged in maize farming, they got locked into cycle of informal debt due to maize’s high production cost. Without sufficient and concrete incentives to induce change in farming behavior, it seems highly unlikely that the farmers would be able to reduce highland maize farming. The government policy to directly intervene or guarantee maize price, although intended to help relieve farmers’ endless problem of poverty, has indirectly fueled the maize farmers to expand their farming land and get locked into the vicious cycle of highland maize farming even deeper (Talerngsri and Pongkijvorasin, 2012). With the government’s weak ability to enforce forest law and regulations, such large scale expansion of highland farming, hence forest clearing, seemed to be the most likely outcome (Pongkijvorasin and Talerngsri- Teerasuwannajak, 2014). Nonetheless, during the time of widespread deforestation in Nan, there are some areas that not only managed to resist the alluring market force of maize, but also found way to rehabilitate and conserve forests using local participation as the most vital mechanism. The Nammeed watershed in subdistrict Puea and Prabhutabat of Chiang Klang district has beautifully succeeded in the reforestation and reduced highland maize farming. The engagement of farmers in specifying and enforcing forest rule and regulations, coupled with appropriate role and timely supports by the officials of Nammeed Watershed Management Unit

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 49 are believed to create profound impacts on the farmers’ attitudes toward conservation. In a rather different context, the Sopsai watershed in Thawangpha district of Nan has also managed to trigger a transformation from highland maize farming to reforestation. Since 2009, the area has received assistance in form of large-scale incentive scheme from “Pidthong Lungpra” foundation whose aims were to alleviate poverty problem, improving farmers’ livelihood and reduce deforestation in watershed area. After 5 years, this incentive mechanism has started to show significant results in changing farmers’ highland farming behavior, as they deserted maize farming to grow perennial economic trees. What happened in these watershed areas has become a profound source of inspiration. This paper investigates these success cases and concentrates on the aspect of reduction in highland maize farming. In other words, we aim to identify significant determinants of such reduction. One of the main results reflects that while ownership of flatland plays compelling role for the natural transformation in Nammeed watershed, it casts no serious impact in the incentive-induced transformation case of Sopsai. In addition, we illustrate how different groups of farmers respond to various forms to incentives. The disparity in the farmer’s responsiveness to incentive measures sheds light on how to design effective incentive schemes that fit groups of farmers with diverse limitations. The paper is organized as follows. Section 2 elaborates on related literature. Section 3 lays down characteristics of the two sites and defines the scope of our survey study; the important statistics from the two areas are highlighted. The investigation of significant factors determining reduction of highland farming in the Nammeed and Sopsai watersheds is conducted in section 4. Section 5 focuses on factors influencing the natural transformation in Nammeed watershed. Section 6 explores the incentive schemes implemented in Sopsai area and investigates the induced transformation in detail. Lastly, the conclusions and policy implications are provided in section 7.

Related Literature

An approach used to address deforestation and other environmental problems in Thailand has long been geared towards command and control. Employing economic tools such as incentive scheme has rarely been the case in the country. Nonetheless, failures of the command and control approach in tackling environmental problem are mainly caused by its high cost of enforcement; the government does not have enough funding and personnel to track, monitor and enforce the laws effectively. People do not have enough incentive to cooperate in preserving environment. It is only lately that the government allows a few communities to be partners in taking care of national forests. Doing so helped reduce monitoring and enforcement cost. Yet, it has not been widely and officially accepted; hence not been used extensively. It is commonly acknowledged that conservation efforts by people themselves cannot come about unless people have enough money to live on. Hence, one of the goals of many agricultural related policies in Thailand has been to directly increase farmers’ income and improve their livelihood. These policies were launched with the hope that once the farmers are

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 50 content with their livelihood and feel secured about their potential income; their intentions to clear forests for farming would diminish. This is usually called a “win-win” outcome; where a society prospers economic-wise and environment is well-protected. It is believed that this outcome can be attained only when people have gained decent income levels (Shafik and Bandhyopadhyay, 1992; Cropper and Griffiths, 1994; Bhattarai and Hammig, 2001) or realized the environment’s true value, which results in supportive political institutions and good governance (Bhattarai and Hammig, 2001). Notable examples of developmental projects that aim to generate win-win outcomes are such as China’s Sloping Land Conversion Program (SLCP), the Forestry Bond Certificate in Advance (CAFa) project, the Forestry Development Fund (FDF) and the Programa de Pago de Servicios Ambientales (PSA) of Costa Rica. However, choosing appropriate means which can both increase farmers’ incomes and effectively generate conservation efforts is not easy. Various factors, such as characteristics of the problem, physical and geographical conditions, site and community characteristics, socioeconomic limitations and available fund, need to be taken into account when determining appropriate method or designing incentive schemes. Good incentive scheme must allow for sufficient monitoring and a certain degree of flexibility to handle unexpected interruptions (ABARE, 2001). A number of studies have looked into factors determining farmers’ efforts in reforestation, or their decisions to join reforestation projects, including the decision to adopt more eco-friendly agricultural practices (i.e. Dewees, 1992; Godoy, 1992; Thatcher et al., 1997; Nagendra, 2007). These factors can be grouped into five main categories: 1) household characteristics; 2) physical conditions such as soil quality; 3) institutional contexts such as land tenure; 4) economic/market conditions such as opportunity cost of reforestation and tangible benefit from forest; and 5) social conditions such as community leader and peer pressure. Farmers give various views regarding the significance of these factors. The decisions on which factors to be used in the studies and how to group these factors depend solely on the studies’ objectives and available data. However, most studies show that land ownership and rights to benefit from the trees grown are vital for the farmers to participate in the reforestation projects. Thatcher et al. (1997) study participations of small farmers in the CAFa and FDF reforestation projects of Costa Rica, which employed variety of economic tools such as subsidies and payment to eco-system service (PES) to induce farmers to change farming behavior and adopt a more eco-friendly agricultural practice, including reforestation. They find that the size of land owned, soil quality, household labour force and non-agricultural income are significant variables determining participation in the projects. The farmers wanted to register land with low productivity with the project, and viewed this option not only as just a way to make better earning, but also a good long term investment for their offspring. In addition, limited household agricultural labour force is also the reason why farmers are more willing to take part in the reforestation projects, the opportunity cost of sticking with agricultural occupation gets higher and higher as the farmers could find better source of income from non-agricultural sector. These results reaffirmed those found in earlier studies, for an example, Francis and Atta-Krah (1989). Furthermore, Huang et al. (2009) conduct a

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 51 comparative study of the incentive schemes implemented to preserve watershed areas in 15 sites all over Asia (i.e. China, Indonesia, the Philippines, Nepal, India and Vietnam). They focus on measures that have characteristics resembling PES1. The authors aim to understand what farmers truly want and elaborate on different forms of incentives used in these various sites. The incentives used may come in monetary form, provision of seeds, low-interest loans, land rights, a right to use and access reforested land, irrigation, infrastructure and trainings. However, due to the fact that there are usually large population residing in the watershed areas, and most of these lands are under state (government) control, it becomes difficult to establish real PES mechanism in these countries Land tenure has become a variable of interest in many studies. With data from 55 sites in Nepal, Nagendra (2007) attempts to determine factors with significant impacts on reforestation. Among many interesting variables such as initial condition of the forests, characteristics of communities that gain direct benefit from the forests, socioeconomic factors, the author pays particular attention to land ownership factor. She illustrates that forests under leasehold condition, whereby the forest on sloping land with very poor condition and difficult access was leased to some impoverished farmers, show significant positive change and highest increase in forest density, compared to forests under government control, and community forests. Furthermore, forests under governments control seem to be most problematic compared the other two types of land ownership. A study by Phanijchanont et al. (2011) shows similar results. Their study focuses on a Karen village at Ban Pala-U in Prachuapkhirikhan, a province in the south of Thailand, and studies people participation in the forest conservation project. They find that the size of land owned and type of land tenure have significant impacts on conservation behavior and determine participation in the project. One other approach of forest rehabilitation that has started to gain acceptance in developing countries is community forest. The studies of community forests in Thailand range from evaluation of forest benefits, management of community forests, to factors determining success of community forests or people participation in conserving the forests. These were mostly done qualitatively and have proposed different kinds of index for measuring success. These studies emphasize that the important determinants of people participation in caring for the community forests are people’s right to derive direct benefits from the community forest (e.g., Jailard, 2011; Tharasuk et al., 2013; Khunsan, 2001). One important reason is that the supplementary income from forest products helps mitigate overall income fluctuation, such that the farmers now have no need to migrate out of the area to find jobs during the non-farming season (Khunsan, 2001). On the contrary, Khanhirun (2003) studies incentive tools used in Khao-wong community forest management, Chaiyaphum. The author finds that a monetary incentive (such as monetary benefits from forest products) is not of high importance when people make decision to participate in the forest management. The most important determinants are people attitude towards forest conservation and their concerns over natural disaster. From

1 In practice, many PES-like schemes implemented in these countries share similar characteristics to government subsidy.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 52 the studies that focus on determining success factors of community forest in Thailand, most find that the main ingredients for success are positive attitudes towards forest conservation tasks and strong community leaders (Sornsri et al., 2004; Sothong, 2004). This paper focuses on farmers’ decision to reduce maize farming on highland, thus converting highland farm back to forests. We explore factors and conditions causing natural adjustment to reforestation in Nammeed watershed area, and compared to the case in Sopsai watershed where various forms of incentives were used to induce farmers to reduce maize farming on highland. We found that “the ownership of flatland”, as well as “age” are significant factors in determining the reduction in maize farming in the natural adjustment case, whereas level of education and age are important factors for the quick transformation under incentive scheme. The ownership of flatland no longer plays significant role when various incentive tools were employed to facilitate the transition. In the case of Sopsai watershed, we also look at how different group of farmers respond to different kinds of incentive. This helps shed light on the issue of choosing effective incentive forms so as to induce behavior change.

Characteristics of Nammeed and Sopsai Watershed

In order to study this, we conducted surveys in two watershed areas of Nan province, namely Nammeed and Sopsai watersheds. Nammeed watershed extends over 16 villages, covering the area of approximately 16,000 hectares. The change that happened over the last 15 year-period in this region represents a unique characteristic of gradual and natural transformation whereby farmers gradually left highland areas which were initially cleared for upland rice and maize farming, to concentrate on lowland cultivation. One important feature here is that there had been limited direct involvement from government sector at the initial phase. Assistances from the government officials of watershed management unit which came later, were mostly in forms of knowledge, infrastructures, farming tools, allocation of water for farmland and public utilities. A strong commitment of local communities in preserving forests together with effective rule and regulation implemented and monitored by local people have resulted in a rich forest cover of almost 12,800 hectares. Of a rather different nature, what has been happening in the Sopsai watershed reflects a more rapid change of upland usage. A massive clear out of forests in this area was due to concession since 1969. The forest condition got exacerbated by expansion of farming area especially maize farming, even after the government had implemented the policy to blockade forests all over the country in the year 1982. In the year 2009, the region suffered a traumatic impact of flash flood, which caused loss of lives and huge damage. A large-scale direct assistance and engagement from “Pidthong Lungpra” foundation, a development organization, to help alleviate poverty and secure better livelihood through various forms and degrees of direct subsidies since 2009 has led to a clear and structural change in upland and lowland farming in the region. A number of farmers have already changed from maize and upland rice farming to grow perennial plants and economic forests.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 53

For Nammeed watershed, we systematically chose three villages that differ in their natural geographical landscape; one village locates on highland with very limited flatland (Ban Nammeed), one located on slope and foothills with some decent amount of flatland (Ban Kwaeng), and another one with handsome amount of flatland and located by the foothill (Ban Denpattana). One common feature is that they all experienced period of extensive deforestation caused by maize and upland rice farming, but at present are successful in restoration and preservation of rich forests in their areas. The three representative villages from the Sopsai watershed are also varied in their geographical characteristics to a certain degree. Ban Huaythanu locates on slope and foothill, Ban Nampak locates on highland, but given good road condition, is not difficult to access. Ban Huaymuang locates on highland and is difficult to access to comparing to others. We gathered data through questionnaires with 107 farmers and in-depth interviews with community leaders, farmers and officials from these six villages. The main objective of questionnaire was to learn about factors that induce their changes in farming behavior. Additionally, in the case of Sopsai watershed; the farmers’ opinions regarding types of incentive scheme they received were also part of the questionnaire. Out of 107 observations, 69 farmers (63% of the sample) managed to reduce or quit highland maize farming. However, in Ban Denpatana, as large as 52% had never cultivated maize on highland, most people in the village have deserted highland maize farming. However, from the sample, Ban Nammeed and Ban Nampak are very successful in changing their highland farming behavior, 77% of the sample has reduced or quitted highland maize farming altogether (table 1). On average, the farmers in our sample managed to reduce highland maize farming by 1.92 hectares per household. The farmers who participated in the survey are heads of the household, with the average age of 50 years-old. Approximately 60% of them finished primary schools, and as high as 80% of the sample has got flatland for cultivation. The average size of land owned by these farmers in the past before the transformation took place was 5.12 hectares per household. The current income per year is $3,162 per household1 (see table 2).

Determinants of Success of the Transformation from Cultivation in Highland to Reforestation in Nammeed and Sopsai Watersheds

In this section, we identify significant determinants which affect farmers’ decision in reducing highland maize farming using OLS regression analysis. Given a vastly different nature of the transformations in the two regions, two separate regressions are used to determine important household characteristics that could become significant factors for the reduction of highland maize farming in each region. The common form of the regression is as follows;

Y= b0+ b1educ + b2age + b3noflatland + b4landp + b5ln(income) + e; where Y represents size of reduction of highland maize farming (hectare);

1 This paper uses the exchange rate of 30 baht/US$

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 54

educ represents highest level of education attained, with 0 if “did not finish primary school”, 1if “finish primary school”, 2 if “finish junior high school”, 3if “finish senior high school”, 4 if “finish vocational certificate”, 5 if “finish high vocational certificate”, 6if “finish bachelor degree”, and 7if “finish higher than bachelor degree”; age represents age of the household leader (year); noflatland represent the ownership of flatland, with 0 if the farmer has flatland and 1 if the farmer has no flatland; landp represents size of cultivating land in the past (hectare); and income represents income per year of a farmer ($/year). Results from table 3 shows that significant household characteristic variables that can be used to explain reduction in highland are age, lacking ownership of flatland and the size of cultivating land in the past (before the transformation). Farming on highland requires high degree of physical strength and stamina; as the farmers get older, they started to seek more age- suitable alternatives, and gradually retire from cultivating on highland. Lacking ownership of flatland implies lower opportunity of the farmers to seek other income-generating crops. Reducing farming on highland means loss of revenue generated from maize, hence, such action cannot be easily accomplished if the farmers fail to find suitable substitutes that generate sufficient compensation. Ownership of flatland has a profound impact on transformation in the case of Nammeed watershed; given other variables stay constant, the farmer with flatland has tendency to reduce highland farming more than the one with no flatland by ¾ hectare. As we consider data from the survey, it shows that most farmers who own a piece of flatland managed to reduce maize production, whereas only 28% of the farmers who do not own flatland managed to reduce highland farming. As for the size of cultivating land in the past, the associated coefficient implies that the farmer with large cultivating land in the past has a tendency to reduce highland farming than the one with small size land. It is not surprising as the farmers with large piece of land has more variety of agricultural options, hence, would not find reducing maize farming an unbearable loss. We obtain a rather different result for Sopsai watershed. In Sopsai case, variables that are significant in explaining the reduction of highland maize farming are education and the size of cultivating land in the past. The farmer with higher level of education tends to reduce highland maize farming more than the one with lower level of education1. This perhaps because the farmers who are equipped with higher level of education are more responsive to incentive schemes and are better at foreseeing benefits generated from the transformation and all incentive measures during the transition period. The variable capturing the size of land owned in the past is also significant in Sopsai watershed. Data from our survey clearly shows that all the farmers with large size of land owned in the past managed to reduce/quit high land farming,

1 Only formal educational system is captured in this regression model. The local wisdom, traditional agricultural knowledge and other practical knowledge which may be attributable to the transformation are not part of the regression study.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 55 whereas as high as 40% of the farmers who own moderate and small size of land in the past could not reduce highland farming.

Natural Transformation in the Nammeed Watershed

In this section, we look in detail at factors that may help explain the gradual and natural reduction of highland maize farming in Nammeed watershed. First of all, the household characteristics of Nammeed watershed are laid out, after that, some important features that define people in the area and their way of living are examined in turn. The data (table 2) shows that the average age of farmers in Nammeed watershed who participated in the survey is 52 years old. In the past, when deforestation was severe, each household’s farm area extended to 2.88 hectares on average. However, they succeeded in reducing cultivation on highland by 1.65 hectares per household. About 63% of these farmers finished primary school and 88% of the households managed to own at least one plot of flatland. The average annual income is $2,808 per household.

Ownership of Flatland

Results from in-depth interviews and simple statistical analysis all point out to the fact that owning a piece of flatland was vital for farmers’ decision to gradually leave highland cultivation. Table 4 shows how such decision varies according to whether or not the farmer owns flatland. A number of in-depth interviews with farmers gave a full picture of revenue and cost incurred from maize farming on highland as compared to rotational cropping on flatland. When taking all material and labor costs into account, farmers who grow maize on highland have average net income of $479 per hectare. However, in practice, almost all farmers still keep to the tradition of cooperative labor. When the labour costs are excluded from cost calculation, the maize grower receives average net income of $792 per hectare.1 Farmers in Nammeed watershed who managed to secure at least a piece of flatland typically engage in rotational cropping; this is very intensive in some villages such as Ban Denpatana. Each household may not own large piece of land, but suffice for a decent flow of income all year round. Farmers usually grow 3 crops per year on the same plot of land. As rainy season starts, the land is normally used to grow sticky rice, mainly for household consumption. After harvesting rice in September, farmers may grow tobacco or Chinese green mustard. Once the second crop is harvested, farmers would have a few months to grow and harvest the third crop such as low-land maize or peanuts before the rain comes in May. Flow of revenue arose from rotational cropping in Nammeed watershed has proved that the chosen plants grown on flatland give farmers better income compared to highland maize.

1 In this paper, net income is calculated using relatively high price of maize, at 8 bahts per kg.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 56

Table 5 shows that if farmers can keep up with cultivating 3 crops per year, they would be able to secure income 10 times larger than what they could have attained from highland maize farming. In other words, one hectare for flatland that is suitable for cultivation can generate income at least equal to that of 10 hectares highland maize farming. Not only that an ownership of flatland gives the farmer’s opportunity to seek higher income per hectare compared to highland maize farming1, the intensive agriculture on flatland give farmers no time to utilize their highland. They gradually desert cultivation on highland and leave the land idle. Farmers are becoming more and more willing to give up their highland ownership and support reforestation. In addition, these farmers realize that their way of living and income are dependent on their ability to utilize flatland, which in turns, is subject to availability of water. The irrigation system installed there would be meaningless if natural water from forest is scarce. The people there know that these all boil down to forest richness.

Size of Land Owned Before the Transformation Took Place

When we look at the size of land owned and cultivated before that transformation took place and how it links to farmer’s decision to reduce/quit highland maize farming, we find that farmers who owned moderate size (3.2 – 8 hectares) and large size (more than 8 hectares) of land in the past used to cultivate maize on highland. Almost all of them managed to reduce/quit highland maize growing. The story is a bit different for the farmers who owned small amount of land (less than 3.2 hectares). Although almost half of these farmers managed to reduce/quite maize farming in highland, but as high as 21.73% did not reduce their maize farm size. The fact that farmers that own large piece of land have better opportunity to try out different alternatives compared to those with already limited land could be a reason for this.

The Farmers’ Attitudes toward the Transformation in Nammeed Watershed

Our survey results show that the important reasons that made farmers in Nammeed watershed leave or reduce highland maize farming are 1) the revenue from maize was not worth it, it was relatively low considering the effort put in (48%); 2) farmers realized negative impacts of highland maize farming on environment (36%) and on their health and livelihood (33%)2. The farmers also pointed out that even though most farmers in the area realized the returns from maize farm in the past was not attractive, still a change to other better options could not come easily; there are a number of obstacles that prevent a quick transformation. The top two obstacles were the farmers’ concerns of the falling income during the change (48%) and lacking knowledge and know-how to seek better option or other career opportunity (27%). The groups of farmers that could not reduce that maize farm size on highland gave significant weights to these two obstacles.

1 It is important to note that full utilization of flatland can only come with sufficient irrigation development and appropriate market condition. 2 Detailed results of analysis in sections 5 and 6 are available from the authors upon request.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 57

As far as the necessary conditions or factors for the transformation are concerned, the farmers in Nammeed watershed emphasize the importance of community’s active participation in laying down local rules and regulations of forest use and access, including punishment (52%) and the engaging role of community leaders (41%). Furthermore, when asked about the intrinsic value transpires from the transformation to reforestation, most farmers feel proud of their actions in preserving forest that benefits society as a whole (61%). They also view that forest answers their needs on food security (42%) and stability of their income (36%).

Transformation under Incentive Scheme in Sopsai Watershed

The transformation from maize and upland rice farming to reforestation that is happenning in the Sopsai watershed has followed a different route. The starting point came from a community leader who wanted to tackle the problems of water management and natural disaster. The community joined “Pidthong Lungpra” program which has brought in various kinds of incentive tools to help trigger a change towards better livelihood. This external assistance started pouring in at the time when people have not yet fully understood positive impacts of reforestation. The poverty, debt, and sub-standard livelihood were the issues that people there most concerned with. However, Pidthong Lungpra project believes that unless the poverty problem is solved, the issue of deforestation cannot be addressed. One of the approaches used by the project is to use zoning system whereby forests are to be divided into three categories: 1) conserved forests whereby use or access is strictly prohibited, 2) community forests whereby the people in the community can benefits from forest products and woods to some extent; and 3) economic forests where the land owners grow perennial trees and benefit directly from their products. With this approach, forests would be well-preserved and people who take part in the forest conservation can benefit both directly in terms of forest products and indirectly in terms of water, environment and biodiversity from the protected forests. Table 2 shows that farmers in Sopsai watershed who participated in the survey and interviews age 48 years old on average. About 56% of the farmers finished primary school. In the past, farmers here had cultivated on land as large as 7.68 hectares per household, and they managed to reduce highland maize farming by 2.24 hectares per household on average. The average annual income is $3,564 per household. As we grouped farmers based on the size of land they owned in the past, and observe how their decision to reduce highland maize farming varies with the land size, we find that all farmers who owned large amount of land (more than 8 hectares) managed to reduce/quit their highland maize farming. However, the transformation was not easy for the farmers with moderate or limited land (less than 8 hectares); as high as 40% of the farmers in this group failed to reduce or quit highland farming. . One important objective of our survey study at this site is to look into detail of how farmers responded to different kinds of incentives. In other words, we look at how effective each type of incentive was in inducing a change in agricultural practice from maize farming to

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 58 eco-friendly alternative or reforestation. Hence, various incentive tools used in this area will be discussed in turn. Overall, the incentive schemes implemented by Pidthong Lungpra project in Sopsai area can be divided into three main categories which are: 1) Direct grant or subsidy to increase income directly so as to facilitate the change. Examples are subsidy for growing perennial trees, subsidy for patrolling forests. Farmers who undertake the forest patrolling task are required to spend a night in the forest and will be paid $13 each round. As for the subsidy for taking care of economic perennial trees, the farmers are paid $50 per hectares the first year after the trees are planted, and $40 per hectare per year for the next 4 consecutive years. This subsidy is considered vital especially during the transition period before the planted perennial trees bear fruits. 2 )Supporting means to reduce production or operational costs of taking up new alternative. These are seed funds, seedling/sapling funds. The main purpose of seedling/sapling fund is to help transform highland that was cleared for maize or upland rice farming to reforestation. In doing so, saplings of economic plants such as durian, cashew trees, banana, tamarind, longan, orange, rambutan, rattan and sugar palm are available for the farmers to choose1. The seedling/sapling fund project just started in year 2013 and will last for three years. In addition, the project helped set up various types of funds ranging from rice seed, farm plant seed, vegetable plant seed, seed for plants grown after rice to compost. These funds help reduce material costs of cultivation, and help end a typical problem of exploiting middlemen who used to play important role in agricultural culture of the area. Furthermore, Pidthong Lungpra also supports livestock farm as an possible alternative for farmers, in doing so, various livestock funds are set up such as pig fund, fish fund and fodder fund. 3) Improvement in infrastructures, capacity building, increasing productivity and knowledge transfers. These all aim to bring about better livelihood, higher productivity and to equip farmers with knowledge so as to be able to grab better options that are available to them. Examples are such as better water system, irrigation, public utilities, land improvement for cultivation, handicrafts and supplementary occupation. In terms of providing irrigation infrastructure to increase cultivation productivity, Pidthong Lungpra helped construct dykes, check dam, irrigation channels, water storage on mountain ridge, with villagers playing important role from construction to maintenance phase. Improvement of water system for everyday consumption was also one of the main supports of the project. Another important change to agricultural landscape is the introduction of rice terrace model and soil improvement in the area so as to increase productivity given limited flatland. Knowledge transfer and capacity building through training are elements to help bring about the transformation as well. These forms of incentive tool make the farmers understand

1 Saplings of trees to be grown in conserved forest and forest for community use are also available for villagers.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 59 what they are getting themselves into. Learning about the success case helps inspire the farmers and make them see the possibility of the transformation. Other supports from the foundation come in form of introducing supplementary jobs such as handicraft. From in-depth interviews, we were able to identify and group key incentive tools that might have significant influence over the transformation. The farmers were then asked to rank these tools according to how they value them. In other words, these tools were ranked based on how important they were in inducing the change in their farming behaviors. From a sample of 50 households in Sopsai watershed, we found that, the farmers give highest importance to irrigation infrastructure and water storage for agricultural purposes (71%), and 67% of the farmers found that assistance in the forms of seed/seedling and sapling funds are one of most crucial measures as these all help cut down production cost. About 57% of the farmers feel that they benefit from direct subsidy, i.e. compensation for forest patrol, and viewed that this form of subsidy is very important from operational point of view1. Support in terms of public utility such as pipeline system or water system for everyday use are ranked fourth in their importance. Other supporting measures such as increasing productivity through soil improvement, livestock funds, knowledge transfers/training and handicraft are considered not as crucial overall (figure 1). In order to understand how these incentive tools affect different groups of people, we divide farmers into groups based on various criteria, such as incomes, ownership of flatland, size of land owned at present, amount of debt, household labor force, land tenure and ability to reduce highland maize farming. Results from this investigation help shed light on the forms of incentives that would be suitable for farmers with different degree and aspects of limitations. In summary, the farmers in low income group are the one that give highest importance to direct subsidy such as compensation for forest patrol. The farmers who have flatland and those who own moderate size of land find that water for agricultural purposes is most valuable for them. The seed/seedling/sapling funds which aim to reduce production cost are most favored by the groups of farmers that own large size of land and those with low income. Although not many farmers found livestock fund very important, the farmers with large land size and high income do find this measure highly beneficial. This could be because they were in an advantageous position to explore new possibility compared to other groups of farmers. The farmers who have no flatland may not see benefit of the water supply for agricultural purposes but they give relatively high importance to water system for everyday consumption. When we look at the overall statistics, only 10% of the farmers value soil quality improvement as highly important, however, this measure is considered vital from the point of views of farmers that have limited land. Similarly, the knowledge transfer/capacity building and training may not hold much significance for farmers overall (just 12% find this measure valuable), but this form of assistance scores quite well in the groups of farmers who have no

1 The total percentage may exceed 100 because we asked each household to give top three measures that they found most important for the transformation.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 60 flatland/owned small amount of land and is ranked fourth in the group of farmers who managed to reduce or quit highland maize farming. The category that interests us most is when the farmers are grouped based on their decision to reduce/quit highland farming (figure 2). Both groups of farmers who managed to reduce highland farming and those that have not changed their practice view that irrigation system for agricultural purposes, seed/seedling/sapling funds and direct subsidy are three most valuable forms of supports. Nonetheless, the group of farmers who managed to leave or reduce maize farming on highland clearly gives high importance to water system for everyday consumption and capacity building, compared with another group. As they understand the imminent role of forest in securing constant and clean water supply for their consumption, they are more willing to give up part of their land for reforestation, or at least reduce forest clearing. The farmers’ attitudes toward the transformation in Sopsai watershed Data from the survey in Sopsai watershed indicates that the farmers who managed to reduce highland farming strongly viewed that low return from maize in relation to the effort put in was the main reason inducing the change (73% of the farmers in this group). The other important reasons were the negative effects maize farming imposed on their health and livelihood (24%) and supports and management from outside organization (24%). The main difference compared to what viewed by the farmers in Nammeed watershed is that the farmers in Nammeed rated impact on environment and natural disaster as one of the most important reasons for the reduction in maize farm, which is not the case in Sopsai. However, when asked about the obstacles impeding the change, the farmers in Sopsai area found that a concern over a falling income which was due to happen during the shift out of maize farming was the key obstacle (50%). Contrasting with Nammeed farmers, Sopsai farmers did not view that lacking knowledge about other potential options was an important obstacle. It could be that Pidthong Lungpra project succeeded in instilling the farmers here with necessary knowledge to support the transformation. As far as the necessary conditions or procedure leading to a complete transformation are concerned, unsurprisingly, the most important procedure viewed by the farmers in this area is assistance/support from outside organization (33%). They wanted to see steps taken to ensure that the new chosen option would provide higher income (20%). In addition, the compensation for taking care of forest (17%), including assistance via seed/saplings funds (17%) are relatively important to secure the transformation. Again, we find some interesting and contrasting pattern compared to what viewed by the Nammeed farmers. In Nammeed, the community leader and the involvement of community were viewed as most important procedure. When asked about what benefit they value most as the transformation to reforestation has taken place, the farmers who managed to reduce or stop highland maize farming indicate that food security (38%) income stability (34%) and stability in terms of asset to pass on to the younger generation (24%). Contrasting to the views of Nammeed farmers, the farmers here rated pride from forest conservation for public interest relatively low (11%).

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 61

Conclusion and Policy Implication

The situation of watershed areas in Nan does not look promising over these past ten years. The deforestation has been mainly caused by rapid expansion of maize farming on highland. The increase of maize price due to a steady rise in demand for animal feed has lured a large number of farmers to this cash crop. Once the farmers got locked into the vicious cycle of maize production, they found it difficult to leave. Notwithstanding the government’s attempts to curb forest clearing through command and control approach, the large scale deforestation in Nan seems to be a living proof of failure and non-functioning environmental and forest laws. The ineffectiveness of command and control approach is becoming more and more evident in many countries around the world (Poffenberger, 2001; Rao and Geisler, 1990; Ostrom, 1990; Hecht and Cockburn, 1989). However, amid the strong market force creating large demand for maize and failure of the command and control approach used by the government, there are a few watershed areas in Nan that have been able to transform themselves; leaving highland maize farming behind, and moving forward to reforestation. The Nammeed watershed in Nan becomes a success case of natural transformation with limited help from external organization and the government, whereas the rapid change that has been happening in Sopsai watershed is a showcase of what the external organization could do via incentive scheme so as to trigger and facilitate the transformation. However, what happened in both regions highlights the fact that a mechanism that is able to generate a “win-win” outcome is the only way forward to solve deforestation issue. Using data from household surveys of six villages and in-depth interviews with farmers and other stakeholders in the areas, this research aims to investigate significant determinants causing reduction of maize farming in highland. Through the use of OLS, we found that the farmers in Sopsai watershed (i.e. Ban Nampak, Ban Huay Muang, and Ban Huay Thanu) and those in the Nammeed watershed (i.e. Ban Kwaeng, Ban Denpatana and Ban Nammeed) show different patterns of behavioral change. For Nammeed watershed, the significant factors facilitating the reduction in maize farming are age, flatland ownership and size of land owned in the past. As the farmers get older, or physically weaker, they gradually retire their highland farming. The farmers who owned large plot of farming land was able to reduce to highland farming more than those who have limited land as more cultivating options are available to the former. The ownership of flatland variable is very interesting; we found that the farmers with some amount of flatland were able to reduce highland maize farming significantly. They were able to engage in alternative cultivating option to compensate for their loss from reduction of maize farming in highland. Furthermore, a simple statistical analysis shows that as high as 58.70% of the farmers with flatland who used to cultivate on highland managed to reduce or quit highland maize farming; whereas only 28.57% of the farmers with no flatland managed to reducing production on highland. For the Sopsai watershed, the result obtained from OLS model shows a different story. The farmers in this area received large-scale assistance from “Pidthong Lungpra” foundation.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 62

Various forms of incentive schemes carry very vital role here. We found that the significant factors determining the reduction in highland maize farming here are education level and size of land owned in the past. One explanation providing for the significance of education level is that, given other variables stay constant the farmers with higher education are more responsive to incentive schemes. The variable “size of land owned in the past” still plays important role for the farmers in this region. Statistics show that all the farmers who owned large size of land in the past managed to reduce/quit highland maize farming; whereas 40% of the farmers who own moderate or small piece of land could not manage to reduce highland farming. One other important point that could be drawn from these OLS results is that the “ownership of flatland” variable that had played important role for the transformation in Nammeed area carried no importance in the case of Sopsai. It could mean that the incentive schemes employed in Sopsai have already compensated for the necessity of owning a flatland. Compensations which came in various forms such as direct subsidy for forest patrol, taking care of economic forest and seeds/sapling funds are sufficient for the farmers such that reduction in maize farm did not hurt them much in terms of income. The education level which did not show significance in the case of Nammeed area, became important in the case of Sopsai as they need to weigh benefits from various supporting measures and adapt themselves to utilize assistances from external organization. Data from the site survey also reflect disparity in attitudes toward the transformation from maize farming to reforestation of the farmers in the two regions. Although farmers from both regions found that the concern over a falling income was the biggest obstacle to leave maize farming, the Nammeed farmers point out that lacking knowledge of better and more eco- friendly alternatives was also a main obstacle, while this factor is viewed as not significant in the case of Sopsai. This result highlights the success of “Pidthong Lungpra” foundation in Sopsai area especially in the aspect of knowledge transfer and trainings. Moreover, the Nammeed farmers view that 1) community participation in stipulating the rules and regulation for forest conservation 2) strong community leader and 3) rights to access and benefit fairly from forest they protected are most important elements for transformation success. With a rather different view, the farmers from Sopsai area point out the importance of 1) external assistance and 2) high returns from new alternatives as elements to secure the change. Our research helps shed light on what are essential ingredients for a transformation from maize farming to reforestation, especially from the farmers’ views. The government policy to help improve flatland productivity can help prevent farmers from expanding their farm land into inappropriate highland or watershed areas. However, in the case of where farmers do not have enough flatland to cultivate on, or in an area where deforestation has reached a critical level that the immediate assistance is required, the government or external organization have very important role to play. A well-designed incentive schemes which take into account differences in the farmers’ limitation, must be used in this case to induce appropriate responses and facilitate the transformation.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 63

References

Australian Bureau of Agricultural and Resource Economics (ABARE). 2001. “Alternative Policy Approaches to Natural Resource Management” Background Report to the Natural Resource Management Taskforce. ABARE. Canberra. Bhattarai, M., Hammig, M. 2001. “Institution and the environmental Kuznets curve for deforestation: a cross-country analysis for Latin America, Africa and Asia”. World Development 29(6), 995-1010. Cropper, M., Griffiths, C. (1994). “The interaction of population growth and environmental quality”. American Economic Review 84(2), 250-254. Dewees P.A. 1992. “Social and economic incentive for smallholders tree growing: a case study from Muranga district, Kenya” Community forest case study series no. 5, FAO, Rome, Italy Francis P.A. and Atta-Krah A.N. 1989. “Sociological and ecological factors in technology adoption: fodder trees in southeast Nigeria”, Experimental Agriculture 25, 1-10 Godoy R. 1992. “Determinants of smallholder commercial tree cultivation” World Development, 20, 713-725 Hecht, S. and Cockburn, A. 1989. The fate of the forest. Verso, London. Huang M., Upadhyaya S.K., Jindal R., and Kerr J. 2009. “Payments for Watershed Services in Asia: A Review of Current Initiatives”, Journal of Sustainable Forestry 28, Issue 3-5, 551-575 Jailard, J. 2011. “Local Participation in Community Forest Management: Ban-Kan-Tah, Khonkaen” Khonkaen University. (in Thai) Khanhiran, R. 2003. “Evaluation of the economic incentive of the Khao Wong community forest management cooperation” Research paper, Faculty of Graduate studies, Mahidol University. Khunsan, C. 2001. “The relationship between socio-economic factors and utilization of non- timber forest products of the communities around Mae Charim National Park, Changwat Nan” Thesis, Kasetsart Universtiy. (in Thai) Nagendra H. 2007. “Drivers of reforestation in human-dominated forests”, edited by Elinor Ostrom, PNAS, Vol. 104, no. 39 Office of Agricultural Economics. 2014. Statistics for agricultural products. Avialable online at http://www.oae.go.th/ewt_news.php?nid=13577 (last retrieved on August 18, 2014) Ostrom, E. 1990. Governing the commons. Cambridge University Press, New York. Phanijchanont, S., Vajirakachorn, S., Suwansujarit, S., and Suppadit, T. 2011. “Participation in Sufficiency Economy Project and Change in Environmental Concern and Conservation Behavior among Thai Karens in Pa La-U Village,” Journal of Environmental Management 7(1): 59-74. (in Thai) Poffenberger, M. 2001. Communities and forest management in Southeast Asia. IUCN, Switzerland. Pongkijvorasin, S. and Talerngsri-Teerasuwannajak, K. 2014. “Win-win solutions for reforestation and maize farming: A case study of Nan, Thailand” in A. Balisacan, U.

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 64

Chakravorty, and M. Ravago (eds.) Sustainable Economic Development: Resources, Environment and Institutions. Academic Press. Rao, K. and Geisler, C. 1990. “The Social Consequences of Protected Areas Development for Resident Populations” Society and Natural Resource 3 (1): 19-32. Shafik, N., Bandhyopadhyay, S. (1992). “Economic growth and environmental quality: time series and cross section evidence”. Working paper for the World Development Report 1992. The World Bank, Washington, DC. Sornsri, U., et al. 2004. “Value of Community Forest and Factors affecting Sustainable Forest Conservation in Ban-Rong-Kok, Pitsanulok” funded by Thailand Research Fund. (in Thai) Sothong, P. 2004. “Achievement of community forest management of Ban San Luang Samakkhi, Mae Pao, Phaya Mengrai, Changwat Chiangrai” Thesis, Chiangmai University. (in Thai) Talerngsri, K. and Pongkijvorasin, S. 2012. “Maize Farming and Inequality Mechanism in Rural Area: A study on supply chain of maize in Wiangsa District, Nan Province” funded by Thailand Research Fund. Available online at http://www.econ.chula.ac.th/research/project?topic=Poli_Topic (last retrieved on August 18, 2014) (in Thai) Tharasuk, W., Ratanachai, A., Chuadaungpui, P., and Masae A. 2013. “Factors Affecting Participation of Ban Tangsai Community in Mangrove Management in Tambon Lipang, Amphoe Palian, Changwat Trang” Paper presented in the 4th National and International Hatyai Conference. (in Thai) Thatcher T., Lee D.R. and Schelhas J. W. 1997. “Farmer participation in reforestation incentive programs in Costa Rica” Agroforestry Systems 35: 269-289

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 65

Table 1 Number of the farmers according to their farming behavior (figure in parentheses indicates percentage out of total number of household in the specified village) Reduce or quit No reduction in Total Has never grown Village highland maize highland maize observations highland maize farming farming 20 1 12 7 Ban Kwaeng (5) )60( (35) 13 11 1 Ban Denpattana 25 )52( (44) (4) 0 10 2 Ban Nammeed 12 (0) (77) (23) 2 17 3 Ban Nampak 22 (9) )77( (3) 0 11 5 Ban Huaythanu 16 (0) )49( (31) 0 8 4 Ban Huaymeung 12 (0) (67) (31) 16 69 22 Total 107 (15) (63) (20)

Table 2 Household characteristics of people in the two watershed areas Household Characteristics Nammeed Sopsai Total watershed watershed

Size of reduction in terms of land size (hectare) 1.65 2.24 1.92 Age of farmer (year) 52 48 50 Proportion of farmers whose highest educational 63 56 60 level is primary school (%) Proportion of farmers who own flatland (%) 88 70 80 Size of cultivated land in the past (hectare) 2.88 7.68 5.12 Income (US$/household/year) 2,808 3,564 3,162

Table 3 Comparison of regression results from Nammeed and Sopsai watersheds. Nammeed watershed Sopsai watershed Variable Coefficient P-Value Coefficient P-Value education -0.08 (0.690) 1.46 (0.038)** age 0.04 (0.011)*** 0.05 (0.263) No flatland -0.76 (0.026)** -0.16 (0.805) Size of cultivating 0.80 (0.000)*** 0.14 (0.100)* land in the past Income -0.62 (0.870) 4.66 (0.860) Note: figures in parentheses represent P-value with *** denoting 1% significance level, ** denoting 5% significance level and * denoting 10% significance level

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 66

Table 4 Decision to reduce/quit highland maize farming with flatland variable (percent) Never cultivate Reduce/quit No reduction in maize on highland highland highland cultivation cultivation Own flatland 28.26 58.70 13.04 Has no flatland 0.00 28.57 71.43

Table 5 Comparison of net income generated by different combination of rotational cropping and highland maize farming Proportion of Proportion of Farmer’s net Farmer’s net Combination of net income to net income to income a income b rotational cropping highland highland ($/hectare/year) ($/hectare /year) maizea maize b

Rice- tabacco- chilli 5,707 9,318 11.9 11.8 Rice- tabacco- lowland maize 5,686 8,797 11.9 11.1

Rice- tobacco- peanut 5,930 9,145 12.4 11.6 Rice- Chinese green mustard- Chilli 3,870 7,209 8.1 9.1 Rice- Chinese green mustard- lowland maize 3,849 6,689 8.0 8.4 Rice- Chinese green mustard- peanut 4,093 7,036 8.5 8.9

Average 4,856 8,032 10.1 10.1 Note: a represents the case if farmer’s own labor value is excluded; b represents a case if farmer’s own labor value is included

Table 6 Decision to reduce/quit highland maize farming and the size of land owned before the transformation (percent) Never cultivate Reduce or quit No reduction in maize on highland highland cultivation highland cultivation < 3.2 hectares 30.43 47.81 21.73 3.2 – 8 hectares 0.00 85.72 14.28 > 8 hectares 0.00 100.00 0.00

Table 7 Decision to reduce/quit highland maize farming and the size of land owned before the transformation for Sopsai watershed (percent) Never cultivate Reduce or quit No reduction in maize on highland highland cultivation highland cultivation < 3.2 hectares 20 40 40 3.2 – 8 hectares 0 60 40 > 8 hectares 0 100 0

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 67

Table 8 Ranking of incentive tools, viewed by different categories of farmers in Sopsai watershed (a proportion of farmers out of the total number of the farmers in that category) (1) (2) (3) (4) (5) (6) (7) (8) Total 71 67 57 27 12 10 8 2 Ownership of flatland Has flatland 83 67 58 19 8 8 8 3 No flatland 40 67 53 47 20 13 7 - Size of land owned Small size (< 3.2 hectares) 50 60 60 30 20 20 - - Moderate size (3.2 – 8 hectares) 87 57 48 26 4 4 4 - Large size (> 8 hectares) 61 83 67 28 17 11 17 6 Highland farming behaviour Reduce/quit highland farming 73 73 57 30 16 5 8 3 No reduction in highland farming 67 56 56 - - 22 11 - Farmers’ income Low income (< $1,667/year) 60 80 80 - - - - - Moderate income ($1,667- $3333/year) 71 63 50 38 17 13 4 4 High income (> $3,333/year) 69 63 63 31 13 13 13 - Size of debt No debt 68 55 45 23 5 9 0 0 Small debt (< $1,667) 67 87 53 27 27 13 20 7 Large debt (> $1,667) 79 64 79 36 7 14 7 0 Household agricultural labor force 2-1people 68 71 53 35 15 15 9 3 More than 2people 76 59 65 12 6 6 6 0 Land tenure Has land tenure 79 71 36 29 21 0 7 0 No land tenure 68 65 65 27 8 16 8 3 Note: (1) Irrigation for agriculture, Seedling/Sapling funds, Direct subsidy (wage), Water for household consumption, Knowledge, Soil improvement, Livestock subsidy, Handicrafts

The 9th ASAE International Conference: Transformation in agricultural and food economy in Asia 11-13 January 2017 Bangkok, Thailand 68

100 90 80 71 67 70 57 60 50 40 27 30 20 12 10 8 10 2 -

Figure 1 Rating the importance of incentive tools by the farmers in Sopsai watershed. 80 70 60 50 40 30 20 Reduce or quit 10 Not reduce -

Figure 2 Rating the importance of incentive tools, by the farmers in Sopsai watershed who could reduce highland maize farming and those that could not.