One size fits all? Contract farming among broiler producers in China

HUANG Ze-ying, XU Ying, ZENG Di, WANG Chen, WANG Ji-min

1 Institute of Agricultural Economics and Development, Chinese Academy of Agricultural Sciences, Beijing 100081, P.R.China

2 Centre for Global Food and Resources, University of Adelaide, Adelaide 5005, Australia

1. lntroduction

Agricultural contract can be generally defined as an agreement between producer and buyer which specifies the price of an agricultural commodity to be delivered at a certain point of time in the future (Harwoodet al. 1999). In the past decades, contract farming has observed increasing popularity in agrarian economies (Glover 1984; Nguyenet al. 2015). The expansion of contract farming is jointly triggered by increasing demand of agricultural products due to income and population growth, and increasing productivity given modern agricultural technologies in production, postharvest processing and distribution (Barrettet al. 2012).With potential benefits such as market assurance, input access for producers and risk mitigation for downstream contractors, contract farming is at the heart of agricultural value chain modernisation in developing countries (Glover 1990; Oya 2012; Martinez-Gomezet al. 2013).

As most of the poor in developing countries are small agricultural producers, the welfare impacts of contract farming on smallholders have attracted growing attention in search of policy implications (Nguyenet al. 2015). Existing literature, however, provides only inconsistent findings.Political economy analyses tend to agree that smallholders have weak bargaining powervis-à-vismonopsonistic contractors, especially if they have few alternatives of livelihood (Grosh 1994; Little 1994). For this reason,smallholders might have seen little welfare improvement from contract farming (Merryet al. 2004; Sivramkrishna and Jyotishi 2008), and likely become quasi-employees as they yield to the downstream contractor in production decision making (Reardon and Barrett 2000). Moreover, contractors may favour larger producers while poorer growers could be marginalised (Little and Watts 1994; Singh 2002; Dev and Rao 2005). Negative externalities may further occur(Welsh 2009). While these critics seem appealing, a growing number of studies from a purely empirical perspective actually suggest the opposite: smallholders in developing countries do benefit from contract farming in terms of income (Bolwig and Gibbonet al. 2009; Miyataet al. 2009;Bellemare 2012), return to capital (Simmonset al. 2005),food security (Bellemare and Novak 2015), and women’s employment (Raynolds 2002). These conflicting results can hardly assist policy decisions aiming to improve welfare through contract farming, and more evidence is needed to build up external validity.

Agricultural production in China has observed substantial changes in the past decades. The dietary structure has been shifting from plant-based to animal-based foods as a result of income growth (Fukase and Martin 2016), which is further magnified by population increase, yet the production behaviour in the poultry industry remains much unknown.In this article, we aim to narrow this knowledge gap through the investigation of contract farming participation of small broiler producers in China using a nationally representative survey. During the past decades, poultry industry in China has observed significant growth and has undergone rapid structural change from smallholder-dominated production to large-scale contract farming (Xie and Marchant 2015).Consequently, the poultry sector is no longer dominated by smallholders and many of them who used to keep broilers as a sideline activity have given up production (Ke and Han 2007). Remaining farmers generally specialise in broiler production (Xinet al. 2016). The welfare implications of broiler contract farming in China have not been formally analysed, though contract farming has grown rapidly in China (Guoet al. 2007), and empirical literature occasionally reports increased profitability for producers in neighbouring countries (Simmonset al. 2005; Narayanan 2014). The structural change and coexistence of small and large producers jointly provide a unique opportunity to analyse smallholder participation in contract farming and associated welfare results.

We approach this topic from both descriptive and quantitative perspectives. We first summarise the characteristics of surveyed households as well as the market prices and costs in their broiler production, the latter of which suggests that contract farming might not be optimal for small producers through cost-benefit comparisons. We then proceed with the estimation of a multinomial logistic regression model to reveal how farmers’ choices among individual farming and three types of contract farming are associated with their socioeconomic characteristics, where the possible endogeneity of broiler production is corrected for using a control function approach. It is found that farm size significantly and consistently explains the variation in these choices. We finally discuss the implications of ourfindings, and conclude our analysis specifically with strategic suggestions for small broiler producers.

2. Analytical framework

While the contradictory welfare results in literature may be puzzling, a closer look reveals that contract farming usually leads to welfare improvements where most producers are smallholders (Simmonset al. 2005; Bolwig and Gibbon 2009; Miyataet al. 2009; Bellemare 2012; Bellemare and Novak 2015). This is intuitive because, if the market is mostly supplied by smallholders who face similar production costs, contracts would only occur if the contracted price is high enough to incentivise most smallholders to participate.However, when large holders are also existent, there could be some room for arbitrage for the downstream contractor,and smallholders would therefore lose. This is likely the case when the contracted price is large enough to attract large holders with their economies of scale and thus lower per unit production costs, but is not sufficient to offset the higher production costs faced by smallholders. Hence, the smallholder welfare impacts of contract farming may vary with the supply-side market structure.

Fig. 1 depicts two different scenarios: with homogeneous producers (smallholders) and with heterogeneous producers (both small and large holders). In each graph,the upper block shows the cumulative percentage of total market production plotted against the cumulative percentage of the corresponding producers ranked by farm size from small to large, which is comparable to the Lorenz curve income distribution. The lower block,on the other hand, shows the average production cost among producers ranked in exactly the same manner.Therefore, with homogeneous producers, the cumulative production share curve is the 45-degree line and the average production cost is the same for all producers, as shown in Fig. 1-A. While with heterogeneous producers,the cumulative production share curve is below the 45-degree line as large holders produce disproportionally more than smallholders, and the average production cost is downward sloping given economies of scale that lowers the average cost when production increases. In both cases,for a certain among of production is to be delivered to the downstream contractor. With homogeneous producers,the contracted price is equal to the average production cost (a higher contracted price is not as profitable for the contractor while a lower contracted price generates a loss for farmers and thus may not be accepted). Producers ranked from b to c participate in contract farming in this case. With heterogeneous producers, however, only a small number of large producers participate in contract farming (from b to c), whose average production costs are below the contracted price, yet participation is not profitable for smaller producers ranked from a to b as their costs are higher than the contracted price. These smaller producers may therefore opt out of contract farming and look for alternative marketing opportunities to obtain higher profits.

Although the above literature has intensively explored the welfare implications of contract farming in a variety of countries with almost homogeneous smallholders, less is known about possible impacts with the coexistence of small and large holders, and how expected profitability could affect the participation of smallholders in contract farming. This information, however, is needed to assist policy designs aimed at improving smallholder welfare and ensuring reasonable allocation of benefits associated with contract farming. With the simultaneous presence of heterogeneous farmers, the broiler industry in China is an ideal opportunity to analyse the smallholder welfare in this regard.

Fig. 1 Contract farming with different supply-side market structures. A, contract farming with homogeneous producers (smallholders).B, contract farming with heterogeneous producers (small and large holders).

3. Data and method

The current study is facilitated by a recent national household survey of Chinese broiler producers implemented from June to August, 2015. Six provinces (autonomous region) are covered: Jilin, Hebei, Shandong, Hubei,Guangdong and Guangxi, which cover a wide spectrum of aggregate broiler production levels in the country and jointly represent 43% broiler meat production in China.The first three provinces are in northern China, while the latter three provinces (autonomous region) are in southern China. Stratified sampling strategy has been used for data collection. Within each stratum (province), four to five broiler-producing towns were randomly selected from a list of all broiler-producing towns provided by local agricultural experimental stations, and broiler producers were then randomly selected from producing household lists provided by the village committees. The survey was implemented through face-to-face interviews with farm household heads by the research team that consists of one professor and four doctoral students in the research field of poultry farm management. In our survey, 27 towns were selected with 15 farmers from each town, generating a sample frame of 405 broiler producers. Of these farmers, five refused to participate in the interviews, reducing the total sample to 400. However, in many cases the interviews were also interrupted mid-way through, or cancelled if the farmer was called away and an alternative time could not be identified. In the end, 373 questionnaires were conducted.367 questionnaires were finally deemed as complete with all information appropriately recorded, suggesting a valid response rate of 98.39%. These 367 producers thus serve as the full sample of our analysis.

Three types of contracts are existent in our data: 1) firm contract, 2) cooperative contract, and 3) hybrid contract.Firm contract (firm+producer) is the most popular contract type, which occurs directly between the downstream processor (firm) and broiler producer. Cooperative contracts(cooperative+producer) are less common yet still existent in many areas, with the downstream contractor being the agricultural marketing cooperative rather than the firm.Unlike the firm, an agricultural marketing cooperative is an entity jointly owned and operated by producers, which pools their individual resources to facilitate final marketing.Hybrid contracts (firm+cooperative+producer) have also been observed in the last decade among Chinese broiler producers due to the beliefs that individual farmers can barely negotiate with large downstream contractors but agricultural cooperatives can generally represent farmers’economic interests. In a hybrid contract, the agricultural cooperative negotiates and contracts with the downstreamfirm on behalf of individual farmers who collectively operate the cooperative, while the agreement between the cooperative and individual farmers are pre-arranged.Therefore, the hybrid contract consists of two contracts,one between the cooperative and the firm, and the other between individual farmers and the cooperative. We further differentiate cooperative contract from hybrid contract by the market role of the cooperative. In the cooperative contract,the cooperative takes full responsibility of final-stage broiler marketing and directly deals with consumers (e.g., individual consumers, restaurants, school cafeterias); while in the hybrid contract, the cooperative does not participate in finalstage broiler marketing but contracts with a downstream business for further broiler processing.

These contract types differ in strengths and weaknesses.While firm contract is perhaps the most straightforward type,it is often criticized as exploitive due to the weak bargaining power of individual producers (Grosh 1994; Little 1994).Cooperative contract, while voicing the needs of producers,can be inadvertently constrained by their lack of knowledge,skills, market information or appropriate business strategies.As a comparison, hybrid contract can help build a more stable relationship among producers and downstream agents as it both strengthens the collective bargaining power of producers and eases market information transmission back to producers given clear downstream needs, yet in reality it is more difficult to establish given the multiple parties involved and the difficulties of negotiation.

In our survey, all the contracts were written and were signed before broiler production. Focus group study suggests farmers generally believe written contracts are better than oral ones in terms of securing the contracted marketing channel and clarifying terms and conditions(thereby facilitating easier conflict resolution in case of need). To create a contract, the downstream contractor first offers the written contract that specifies the quality, price, and production details including the time of transaction. Focus group study suggests product quality is a major incentive to use contracts, which accords with earlier literature (e.g.,Guoet al. 2007). Then negotiation occurs and the contract is signed when terms and conditions are agreed upon, and so the contracted price can vary among producers. In each contract type, the downstream contractor partly shares some of the responsibilities of broiler breed provision, technology training, consulting, and disease control practices, which results in looser contract structures as compared to vertical integration commonly observed in horticultural production.However, the contract relationship is reasonably stable as most farmers contract with the same downstream entity over the years given that the number of downstream contractors are limited in most localities. The nonparticipants of broiler production contracts always sold their broilers directly to the market in our data.

Table 1 presents the summary statistics of the full samples and subsamples broken down by participation status of broiler contract farming and contract types.Household head characteristics, farm characteristics and broiler production practices are respectively presented in three panels. Among surveyed broiler producers, 59.4%(218 producers) participated in firm contract, whereas much fewer participated in either cooperative (46 producers)or hybrid (25 producers) contracts. The remaining 78 producers (21.3%) directly sold their broilers to the market without any contract engagement. While most household head and farm characteristics were very similar across these subsamples, firm contract participants did appear to be more likely female-headed and less likely Communist Party members, while hybrid contract participants were the most educated group with the lowest proportion of risk-averse producers. Nonparticipants of broiler contract farming looked rather alike, except that they possessed an above-average education level and there were fewer village cadres among them. Regarding farm characteristics, hybrid contract participants had the largest land area, which was about twice that of firm and cooperative contract participants and one third larger than that of nonparticipants. Moreover,nonparticipants were more likely to raise local broiler breeds and to be close to major markets. Most of these discrepancies, however, are small.

More interesting patterns can be seen from a simple cost-benefit analysis of broiler production among surveyed farmers. Nonparticipants had the smallest annual production,which is about half that of the contract participants, which accords with earlier observations that contract farming might favour larger producers (Little and Watts 1994; Singh 2002; Dev and Rao 2005). On the other hand, the annual production of hybrid contract participants was the largest.Moreover, although hybrid contract participants realised the smallest profit per broiler, their total broiler income was still the highest among all subgroups given their largest annual production.

A comparison among all subgroups further suggests that economies of scale are evident, as annual production and per broiler production cost rank in exactly reversed order.The average broiler weight at farm gate is the smallest for nonparticipants and their per broiler production cost is the highest, which could be partly due to the fact that there was no technological support from downstream contractors.However, the broiler prices they received directly from the market were substantially higher than those offered by contractors, and therefore the profit per broiler for nonparticipants (6.090 CNY or 0.968 USD in 2015) was averaged 327% that of all contract participants. As a result,even with the smallest annual production, the total broiler income of nonparticipants was not necessarily lower than(e.g., firm) contract participants. This is consistent with earlier findings (Verhaegen and Van Huylenbroeck 2001),and these observations imply that it might not be profitable for the former to participate in contract farming. To test this,we implement a simple counterfactual exercise by replacing the individual broiler market prices nonparticipants received with the province-level average contract price participants received. In that case, 74 out of 78 nonparticipants would have negative profits if they were participants, which confirms the above speculation.

The above observation accords with our hypothesis that contract farming could have opposite welfare results between smaller and larger producers. To investigate the role of farm size in broiler contract farming participation,there is a need to further implement quantitative analysis where multiple factors are simultaneously controlled for. It is towards such analysis that we now turn.

Table 1 Descriptive statistics of full samples and subsamples by contract type

4. Results

We employ multinomial logistic regression modelling to investigate the possible effect of farm size in broiler contract farming participation. Multinomial logistic regression is specifically useful in predicting the probabilities of mutually exclusive discrete choices with a set of independent variables (Greene 2012). In our case, the four discrete choices are firm contract participation, cooperative contract participation, hybrid contract participation and nonparticipation. Multinomial logistic regression is appropriate in our analysis because on one hand, a simple logistic regression with the binary outcome of participation/nonparticipation would not be capable to differentiate possible heterogeneity among choices over three contract types. On the other hand, contract participation decision and contract choice could not be separated as both were made before broiler production (or a nested logistic model would be more appropriate). Independent variables in our analysis include household head characteristics, farm characteristics and annual broiler production as described in Table 1. Broiler sales price, production cost and profit measures are not included as predictors because they were realised during or after the production process, and could not logically affect the participation decision or contract choice made before production. Neither included is broiler breed,which is usually provided by the downstream contractors if the producer participates in contract farming and is not a separate decision. However, annual production is included as it is the best available indicator of economies of scale, which might not change much during a single year.Finally, town dummies are included to capture any possible unobserved heterogeneity across localities.

Empirical estimation of the multinomial logistic regression relies on the assumption of independence of irrelevant alternatives. It states that the odds, or relative probability, of preferring one choice over another are independent from the presence or absence of other alternatives, which therefore can be considered as “irrelevant”. We proceed with formal Hausman-type tests to see whether this assumption holds since the violation of which could invalidate our results, andfind that it does.

現(xiàn)行國家評價體系中一級指標還有“保護措施”一項，用以評價保護措施的合理性、完善性和執(zhí)行力度。由于本文詞頻分析中出現(xiàn)的與保護管理相關的“管理、治理”二詞占比較低，故沒有將“保護措施”作為一級指標，而是在三級指標中出現(xiàn)。

A potential concern of direct maximum likelihood estimation of the multinomial logistic model regards the possible endogeneity of annual broiler production, a decision made by the producer, which could result in biased and inconsistent estimates. Endogeneity has been recognized in contract choices in agricultural production (Ackerberg and Botticini 2002), yet addressing endogeneity in a multinomial logistic regression is technically very difficult (Wooldridge 2015). In our case, we employ the two-step control function estimation procedure proposed by Wooldridge (2015).In the first step, we estimate simple linear regression using an excluded instrument and covariates against the endogenous variable of annual broiler production, and obtain the residuals. In the second step, we then estimate a multinomial logistic model using the residual as the control function along with covariates. Standard errors are finally obtained using bootstrapping.

The excluded instrument in our analysis is the area of broiler production facilities, which should intuitively affect annual broiler production but not affect broiler contract choice other than through annual production, thereby meeting the criteria of the instrumental variable. The pairwise correlation coefficient between the instrument and annual broiler production is 0.57, probably not as high as one may speculate, because 1) the different practices among producers lead to varying broiler densities during production,and 2) certain off-farm employment opportunities have prevented the broiler production facilities to reach their full capacities. In the first-step regression, the overallF-value is 41.76 (with aP-value of 0.000), suggesting the excluded instrument is reasonably strong.

In the multinomial logistic regression, nonparticipation is set as the default outcome. Moreover, we should focus on the average partial effects of covariates to ease the interpretation of results. Table 2 presents the average partial effects obtained by estimating the average structural function (Wooldridge 2015). The overall model specification is statistically significant at the 1% level, suggesting all covariates together successfully explain part of the outcome variations.

Among all covariates, annual production, which captures economies of scale, is statistically significant in all contract choices. This is in line with recent literature (Abebeet al.2016). Specifically, the probabilities of participating in firm,cooperative and hybrid contracts would be 0.010, 0.013 and 0.005 higher as annual production level increases by 1 000 broilers. While the average partial effects are small, the actual differences may not necessarily be. The interquartile range in annual production in our data is 30 000 broilers,which implies that the probabilities of participating in each type of contract can be 30, 39 and 15% higher for the median large producer (75% percentile of annual production–45 000 broilers) than for the median small producer (25%percentile of annual production –15 000 broilers). Thesefindings confirm our earlier observation that smallholders would be less likely to participate in broiler contract farming.

Several patterns of the associations between other covariates and contract choices are also worth discussion.Being a village cadre encourages firm contract participation,while being a Communist Party member discourages it. In China, village cadres are usually experienced,entrepreneurial and more capable farmers who take the lead in economic activities, while Party members tend to be conservative individuals who are more likely to follow instructions of the local government. As very few producers were both village cadres and Party members in our data,these contradictory effects could simply reflect the conflicting roles of these engagements in this foremost contract type.While education does not seem to stimulate contract farming in general, better educated producers are less likely to participate in cooperative contract, where the cooperative is run by farmers and represent their economic interests.Moreover, risk-averse producers are found more likely to participate in either firm or cooperative contract, which is also intuitive given the more secure marketing channel with contract farming, though such association is not seen with hybrid contract participation. Most other covariates do not seem to be correlated with contract choices.

Table 2 Multinomial logistic control function estimation of broiler contract choices (n=367)1）

5. Discussion

These findings further imply that it could be a rational choice for small producers to opt out of contract farming,which differs from the existing concern that they might be passively excluded from participation given that downstream contractors would prefer large producers (Little and Watts 1994; Singh 2002; Dev and Rao 2005). The underlying belief of the latter argument is that contract farming could potentially benefit all producers, and therefore nonparticipation would result in undesirable welfare loss for small producers,which is not the case in our study. Specifically, the poultry industry in China has been through rapid structural change where farmers increasingly specialise in broiler production,leading to the coexistence of small and large producers(Xie and Marchant 2015), which would largely explain the unprofitability of smaller producers in contract farming. This coexistence could be temporary and might fade over time given the progressive specialisation in broiler production and the fast increase of productivity among remaining broiler producers (Xinet al. 2016). However, our results are still of policy importance because, although broiler industry in China has been transitioning from smallholder-dominated production to large-scale contract farming, the ongoing process may take a long time and many small broiler producers are still expected to exist in the foreseeable future. Moreover, our data may also mirror the future market patterns of smallholder agriculture which is widely existent in agrarian economies and yet to be modernised. Hence,rather than contradicting recent literature that documents positive smallholder welfare impacts in those economies(e.g., Simmonset al. 2005; Bolwig and Gibbon 2009; Miyataet al. 2009; Bellemare 2012; Bellemare and Novak 2015),the current study complements it in terms of forecasting possible smallholder welfare scenarios that might occur as markets evolve.

As contract farming may benefit smallholders only if they provide most of market supply, a key question is about the future of smallholders given the increasing prevalence of contract farming with agricultural modernisation.Expectedly, not all small producers would continue broiler production, increase productivity or participate in contract farming. While the welfare consequences of quitting broiler production are not necessarily undesirable, especially when outside opportunities are more profitable, possible impacts on remaining small producers still demand more attention.Specifically, as producers are price-driven, it is necessary to investigate how evolving market prices affect smallholder welfare.

One interesting observation in our data is that the market prices received by nonparticipants of contract farming were much higher than the contracted prices for participants(Table 1). While middlemen’s profits might partly explain this difference, more can be revealed through a simple comparison of broiler breeds of nonparticipants. Among 76 nonparticipants, 33 produced broilers of improved breeds and 45 produced broilers of local breeds. The average price received by the former group (11.67 CNY, or 1.86 USD in 2015) is significantly lower than that of the latter group (19.28 CNY, or 3.07 USD in 2015,P-value through pairwiset-test: 0.003). This pattern, along with the fact that nonparticipants produce broilers of local breeds the most (Table 1), confirms earlier observation that Chinese consumers generally prefer “l(fā)ocal” chicken, which are free-range, slowly-growing with poor feed conversion, but with better meat quality in terms of flavour, tenderness,fat deposition, muscle fibre composition, as well as visual appearances (Yang and Jiang 2005). Therefore, consumers’willingness to pay for those “l(fā)ocal” chicken could be much higher. The implication of this consumer preference can be more profound when placed in the context regarding the future of small producers. A possible strategy is to meet these market needs by specialising in “l(fā)ocal” chicken or poultry products for which small producers possess a comparative advantage. This has already taken place in multiple food sectors. For example, organic food production has experienced an annual farmland increase of 8.9%between 2001 and 2011 to meet the increasing demand(Paull 2011). In the broiler industry, possible opportunities similarly exist for not only organic but also free-range chicken given their price privileges (Husaket al. 2008). Even in developed countries, the market opportunity for both organic and free range poultry products is yet to be fully developed(Castelliniet al. 2008). Specialising in these products could therefore be a possible long-term business strategy for small broiler producers given that contract farming is not profitable.

6. Conclusion

We investigate whether contract farming is a financially good option for small broiler producers using a recent nationally representative survey in China. In contrast with many recent studies, we find that contract farming does not benefit small producers in face of larger competitors, the latter of which assume lower production costs given their economies of scale. Descriptive and quantitative analyses further show that small producers were not disadvantageous in terms of socioeconomic characteristics. Certainly, low profits made them opt out of contract farming, given the much higher prices they received from directly selling to the market.Therefore, the current study complements recent empirical literature in extending the almost homogeneous supplyside market structure to cover the coexistence of small and large producers, and conclude that contract farming may not be able to effectively improve smallholder welfare in the latter case.

We further discuss that small broiler producers may consider alternative opportunities to continue operating in the poultry industry, such as specialising in organic and free-range chicken, which better utilises their comparative advantages. Full cost-benefit analysis of these strategies,however, is beyond the scope of this study. Data limitations further exclude us from investigating the smallholder welfare dynamics of contract farming over time, especially with market structure evolvement from homogeneous small producers to the coexistence of small and large producers.As a first study on broiler contract farming in China, ourfindings only speak to this specific industry and more research is needed to establish external validity. These remaining issues are open for further investigation that could build on the current analysis to better derive policy implications regarding the welfare impacts of contract farming.

Acknowledgements

This study is supported by the earmarked fund for the Agricultural Science and Technology Innovation Program,Chinese Academy of Agricultural Sciences (ASTIPIAED-2017) and the Modern Agro-industry Technology Research System, the Ministry of Agriculture of China(CARS-42-G24). We thank Prof. Liu Chunfang, Dr. E Yuzhou and Miss Zhou Rui from Chinese Academy of Agricultural Sciences for their assistance in survey data collection and the supports of staffs from broilers test stations in Hebei,Jilin, Shandong, Guangxi, Hubei and Guangdong.

Abebe G K, Bijman J, Royer A. 2016. Are middlemen facilitators or barriers to improve smallholders’ welfare in rural economies? Empirical evidence from Ethiopia.Journal of Rural Studies, 43, 203–213.

Ackerberg D A, Botticini M. 2002. Endogenous matching and the empirical determinants of contract form.Journal of Political Economy, 110, 564–591.

Barrett C B, Bachke M E, Bellemare M F, Michelson H C,Narayanan S, Walker T F. 2012. Smallholder participation in contract farming: Comparative evidence from five countries.World Development, 40, 715–730.

Bellemare M F. 2012. As you sow, so shall you reap: The welfare impacts of contract farming.World Development,40, 1418–1434.

Bellemare M F, Novak L. 2015. Contract farming and food security.American Journal of Agricultural Economics, 99,357–378.

Bolwig S, Gibbon P. 2009. The economics of smallholder organic contract farming in tropical Africa.World Development, 37,1094–1104.

Castellini C, Berri C, Le Bihan-Duval E, Martino G. 2008.Qualitative attributes and consumer perception of organic and free-range poultry meat.World’s Poultry Science Journal, 64, 500–512.

Dev S M, Rao N C. 2005. Food processing and contract farming in Andhra Pradesh: A small farmer perspective.Economic and Political Weekly, 40, 2705–2713.

Fukase E, Martin W. 2016. Who will feed China in the 21st century? Income growth and food demand and supply in China.Journal of Agricultural Economics, 67, 3–23.

Glover D. 1990. Contract farming and outgrower schemes in east and southern Africa.Journal of Agricultural Economics,41,303–315.

Glover D J. 1984. Contract farming and smallholder outgrower schemes in less developed countries.World Development,12, 1143–1157.

Greene W H. 2012.Econometric Analysis. 7th ed. Prentice Hall,Upper Saddle River. pp. 763–766.

Grosh B. 1994. Contract farming in Africa: An application of the new institutional economies.Journal of African Economics,3, 231–261.

Guo H, Jolly R W, Zhu J. 2007. Contract farming in China:Perspectives of farm households and agribusiness firms.Comparative Economic Studies, 49, 285–312.

Harwood J, Heifner R, Coble K, Perry J, Somwaru A. 1999.Managing Risk in Farming: Concepts, Research, and Analysis.Market and Trade Economics Division and Resource Economics Division, Economic Research Service,U.S. Department of Agriculture. Agricultural Economic Report No. 774.

Husak R L, Sebranek J G, Bregendahl K. 2008. A survey of commercially available broilers marketed as organic, freerange, and conventional broilers for cooked meat yields,meat composition, and relative value.Poultry Science,87, 2367–2376.

Ke B, Han Y. 2007. Poultry sector in China: Structural changes during the past decade and future trends. In: Thieme O,Pilling D, eds.,Poultry in the 21st Century: Avian Influenza and Beyond. Food and Agriculture Organization of the United Nations, Bangkok, Rome, Italy. pp. 1–33.

Little P. 1994. Contract farming and the development question.In: Little P, Watts M, eds.,Living Under Contract: Contract Farming and Agrarian Transformation in Sub-Saharan Africa. University of Wisconsin Press, Madison.

Little P D, Watts M J. 1994.Living Under Contract: Contract Farming and Agrarian Transformation in the Sub-Saharan Africa. University of Wisconsin Press, Madison.

Martinez-Gomez F, Aboites-Manrique G, Constance D H.2013. Neoliberal restructuring, neoregulation, and the Mexican poultry industry.Agriculture and Human Values,30, 495–510.

Merry F D, Sheikh P A, Mcgrath D G. 2004. The role of informal contracts in the growth of small cattle herds on thefloodplains of the Lower Amazon.Agriculture and Human Values, 21, 377–386.

Miyata S, Minot N, Hu D. 2009. Impact of contract farming on income: Linking small farmers, packers, and supermarkets in China.World Development, 37, 1781–1790.

Narayanan S. 2014. Profits from participation in high value agriculture: Evidence of heterogeneous benefits in contract farming schemes in Southern India.Food Policy, 44,142–157.

Nguyen A T, Dzator J, Nadolny A. 2015. Does contract farming improve productivity and income of farmers? A review of theory and evidence.Journal of Developing Areas, 49,531–538.

Oya C. 2012. Contract farming in Sub-Saharan Africa: A survey of approaches, debates and issues.Journal of Agrarian Change, 12, 1–33.

Paull J. 2011. The Uptake of organic agriculture: A decade of worldwide development.Journal of Social and Development Sciences, 2, 111–120.

Raynolds L T. 2002. Wages for wives: Renegotiating gender and production relations in contract farming in the Dominican Republic.World Development, 30, 783–798.

Reardon T, Barrett C B. 2000. Agroindustrialization, globalization,and international development: An overview of issues,patterns, and determinants.Agricultural Economics, 23,195–205.

Simmons P, Winters P, Patricka I. 2005. An analysis of contract farming in East Java, Bali, and Lombok, Indonesia.Agricultural Economics, 33, 513–525.

Singh S. 2002. Contracting out solutions: Political economy of contract farming in the Indian Punjab.World Development,30, 1621–1638.

Sivramkrishna S, Jyotishi A. 2008. Monopsonistic exploitation in contract farming: Articulating a strategy for grower cooperation.Journal of International Development, 20,280–296.

Verhaegen I, Van Huylenbroeck G. 2001. Costs and benefits for farmers participating in innovative marketing channels for quality food products.Journal of Rural Studies, 17, 443–456.

Welsh R. 2009. Farm and market structure, industrial regulation and rural community welfare: Conceptual and methodological issues.Agriculture and Human Values,26, 21–28.

Wooldridge J M. 2015. Control function methods in applied econometrics.Journal of Human Resources, 50, 420–445.

Xie C, Marchant M A. 2015. Supplying China’s growing appetite for poultry.International Food and Agribusiness Management Review, 18, 115–136.

Xin X, Zhang Y, Wang J, Nuetah J A. 2016. Effects of farm size on technical efficiency in China’s broiler sector: A stochastic meta-frontier approach.Canadian Journal of Agricultural Economics, 64, 493–516.

Yang N, Jiang R S. 2005. Recent advances in breeding for quality chickens.World’s Poultry Science Journal, 61,373–381.