ظهور درونی از بازارهای اعتباری: قرارداد در پاسخ به فناوری نوین در غنا

Access to credit is important for the productivity and overall welfare of farmers in developing countries. We present a theoretical framework which shows that a change in the mode of shipping (from air to sea) in the Ghanaian pineapple industry made it profitable for pineapple exporters to provide myopic farmers with both in-kind loans (to improve productivity) and cash loans (for consumption smoothing) despite being unable to monitor farmers or enforce repayment. The innovative theoretical result is that providing farmers with additional cash loans can enforce greater input use without compromising repayment. We provide evidence in the form of a case study documenting the dramatic rise of informal credit (through contract farming) after the switch to sea-freight between 1996 and 2001. Using this anecdote, we argue that credit arrangements can arise spontaneously, absent non-market interventions to meet market needs even in the absence of proper legal protections for creditors.

There is general agreement among researchers and policymakers that poor agricultural households in developing economies lack adequate access to credit. These poor households do not meet the traditional criteria for borrowing (especially given the long production cycle in agriculture) and hence are often perceived as bad credit risks by private financial institutions. This has a significant impact on these households for whom credit is vital. Loans allow smallholder farmers to invest enough to get productivity gains and it has been shown that alleviating capital constraints can have big effects.1 In addition, loans may be crucial for poor households to smooth consumption in emergencies and thus can affect nutrition, health and overall household welfare. In response to this need for credit, policymakers have attempted to intervene in markets. Governments have introduced state-owned development banks and while some have been successful,2 others have proved to be an inefficient source of credit,3 have been subject to political capture4 and have had distortionary effects.5 Microfinance institutions have been shown to have impacts but the impacts seem to be small on average with bigger benefits for only a small proportion of people.6 Ultimately, developing economies need a range of financial products and no single type of credit instrument can suffice. For example, microfinance loans are typically paid every week and are hence of limited use for capital investments in the agricultural cycle. Our contribution in this paper is primarily theoretical. We use anecdotal case study evidence from the pineapple industry in Ghana to document the advent of a new shipping technology and the concurrent emergence of an informal credit market via which farmers were able to acquire both loans for inputs and for consumption smoothing from pineapple exporters–credit contracts commonly referred to as contract farming.7 We then develop a model of informal contracting which explains why a credit market functioned after, but could not function before the technology change. Using this theoretical framework, we show that not only did the credit benefit the smallholder farmer, but the form of the loan additionally allowed the credit provider to better control the farmer's actions. Moreover, our model explains the nature of the offered credit contracts by showing that the provision of credit through multiple channels (cash and in-kind) can facilitate contracting and lead to higher profits for the loan providers. This sort of credit institution is common in developing economies. It is known that the quality of institutions determines economic progress and development.8 Hence, we feel that our results may be important for policy makers who seek to understand the factors which lead to the emergence of institutions where they are absent. In particular, we develop a model of informal credit through contract farming aimed at explaining both the reasons behind its emergence in our anecdote from Ghana and the form that the credit contracts took. Present in the model are three of the main issues which could lead to a breakdown of informal contracting. The first is that a farmer who receives credit may default and end up selling her output to a buyer other than the one with whom the farmer is contracting (side selling or extra-contractual marketing). By holding up the buyer, the farmer does not have to repay her loan obligation. The second is the moral hazard problem that stems from the buyer's inability to monitor the farmer's production process. There is no guarantee that the farmer will use the inputs provided on the crop as opposed to using them on a different crop or selling them for cash in the market. Lastly, in spite of the farmer's best efforts, the crop may end up being of poor quality in which case the exporter cannot recover his loan or, in other words, the farmer has limited liability. This implies that providing loans is always risky for the exporter. Since contracting is informal, the exporter must provide dynamic incentives to overcome these hurdles. Providing dynamic incentives is tricky as farmers in developing countries are often myopic.9 Firstly, the model uncovers how the difference in observable quality at the time of shipping altered the hold-up problem which affects informal contracting. We argue that this made contracting possible under sea freight when it wasn't under air freight. In the case of air freight, the quality of the fruit was observable to the farmer when she was deciding whether to default on the loan or not. Thus, if the fruit was of high quality, there were large gains to be had by holding-up the exporter and selling the fruit to someone else. By contrast, at the time of shipment by sea, the farmer could not foresee the end quality of the fruit once it arrived at Europe. Therefore, the expected gains from holding-up the exporter were lower and this made the prospect of default less attractive. We argue that for myopic farmers, this easing up of the hold-up problem was necessary in order for the exporters to provide them with credit without the fear of default. Secondly, the model shows that, under certain conditions, it is profitable for the exporter to offer credit not just in the form of inputs to production but also cash for consumption. Put differently, the exporter can benefit by providing the farmer with money in consumption emergencies, knowing that it would not be spent on production. This is a counterintuitive result as it seems that by providing the farmer with a cash loan for consumption, that the exporter is needlessly bearing additional risk from credit provision without receiving any benefits from improved production. Hence, such behavior might seem to be altruistic. However, we argue that this additional credit raises the value of the relationship for the farmer and makes her more reluctant to default. This in turn allows the buyer to provide higher amounts in-kind credit without fearing default thereby increasing profits. That the type of transfer may have critical implications in the functioning of informal contracts is an important implication of our model and, to our knowledge, is a result that hasn't been identified in previous theoretical work. We discuss how this result may be relevant in other settings in Section 6 where we provide concluding remarks. We provide anecdotal evidence from the field in the form of a case study10 from Ghana where in 1996, refrigerated sea freight became an option for pineapple exporters to ship their produce to European markets. The cost savings to the exporters of using sea freight as compared to the only other option, air shipping, was substantial. However, we argue from our qualitative interviews that this technological innovation meant that the quality of the pineapple at the time it was shipped was no longer observable. This is because the transit time to Europe took ten to twelve days by sea as opposed to a few hours by air. For the case of air freight, the quality of the pineapple at the time of shipping was essentially the quality at the time it was offered to the end customer (due to the short freight time) and so was observable for the intent and purposes of the exporter. However, for the case of sea shipping, due to the long transit time, the quality of the pineapple by the time it was received in Europe could no longer be predicted in advance by either the exporter or the farmer. We argue is that it was precisely this unobservable quality that both necessitated and facilitated the advent of credit arrangements. In response to this change in shipping technology, as the model predicts, credit arrangements emerged between the exporters and the pineapple growers in which farmers were offered cash loans (for consumption) and in-kind loans of high quality fertilizer (for production) in exchange for a contract on the fruit. This allowed the farmers to apply fertilizer in production which they otherwise could not afford — the application of high quality fertilizer increases the odds that the ripe fruit is of high quality. This case study is thus an instance where credit institutions tailored to particular markets seemed to emerge endogenously without government interventions in response to market needs.

In future work, we hope to test our theoretical framework empirically. The data required to test our model is quite intensive and unfortunately, such data does not exist for the pineapple industry in Ghana during the period we study. That said, it would be easy to construct an experiment around our framework which would allow for a testing of the model using a randomized control trial. More generally, we feel that empirical work aimed at uncovering the conditions under which contract farming can function successfully is an important topic for future research. There a few obvious ways in which the model can be generalized. We could allow the production function to be nonlinear and allow the application of inputs to affect not just quality but also quantity. If the probability of high quality fruit remains 0 when ki = 0, most of our results can still be obtained. The essential parameter determining the success of contracting would then by the marginal returns to investing in fertilizer when ki = 0. Similarly, utility from consumption can also be made nonlinear and once again the marginal utility of consumption at ci = 0 would be a critical parameter. In addition, we could explicitly model the shipping cost s of the fruit. This too would not affect our results as all this would do is make the effective price q–s in the export market. Our results show that contracting is facilitated by a high export market price. Thus the higher shipping cost under air freight is an additional reason why there was contracting in Ghana before the advent of sea freight. Lastly, we could endow the farmer with nonzero wealth. Once again, we could obtain our results if farmers are sufficiently poor. An important extension of the model would allow for the exporter to hold-up the farmer. Here we suggest one way in which the model can be adapted to account for this. Suppose, the exporter contracts with a positive measure of infinitesimal identical farmers each of whom draws her consumption state independently. In a symmetric equilibrium, the law of large numbers would ensure that the proportion of high quality fruit upon arrival of the shipment would be exactly γhϕkh⁎ + γlϕkl⁎. If the farmers were able to communicate among themselves, any profitable hold-up by the exporter would be detectable. Hence, a sufficiently patient exporter would not want to cheat the farmers as this would result in the loss of future supply. Such a model would be a reasonable approximation of our case where farmers are closely clustered smallholders and each exporter contracts with many farmers. That said, we consider the development of a general model of contract farming with two sided hold-up to be an important extension and we hope to do so in future work. While we use the case study from Ghana to motivate our theoretical framework, our model of contracting under sea freight could be used to explain other instances of successful contracting schemes. There is a literature consisting of specific case studies and a vast body of work that describes more generally when and where contract farming will be successful (see, for example, Grosh, 1994, Little, 1994, Minot, 2007, Poulton et al., 1998 and Watts, 1994). Overall these studies generally point to the following being important29: a. It works where there are large buyers (exporters, large scale processors and supermarkets), but with competition among the buyers. b. It does not work for commodities that are homogeneous, non-perishable and where quality is easily observable—the transaction costs here are low and spot markets (outside options) are therefore efficient; it only works for cases where spot markets are not efficient, i.e. where spot markets cannot convey information on aspects of quality that final consumers care about. c. It happens where crops have important quality variation and specialized inputs are needed to raise quality (sometimes inputs that are otherwise not easily available to farmers, see Goldsmith (1985) for an example), or for high value crops, or for highly perishable crops with technically difficult production.30 d. Contract farming cannot be sustained if there is easy leakage (farmers side selling)—strong repayment incentives are needed. e. A strong demand for the crop. Our model has points a–c built into it. Of particular importance is point b which is precisely the difference we highlight between the air and sea freight in Ghana. Of course, the primary goal of the model is to demonstrate how repayment incentives (point d) can be generated in the absence of formal legal enforcement. Lastly, Proposition 3 shows that contracting only emerges if the price for the crop is high (point e). High demand implies a high price q which we have argued can be sufficient for contracting to function even if farmers are myopic. Our model can also be used to analyze certain failures of contract farming in the developing world (Daddieh, 1994, Grosh, 1994, Jaffee, 1994 and Poulton et al., 1998). Most of these failures occurred due to disparities between the prices offered in the contract and those in the spot market. Recall that in our model the price offered to farmers upon default is endogenous—it is assumed that there is always another buyer willing to match the price offered in the contract. That said, it is easy to write down the model with an exogenously specified spot market price ps available to farmers who default. The only way this alters problem (★) is by changing the right side of the default constraint (Defi) which becomes View the MathML sourceϕkipi−ki+ci+μici+δ1−δ∑j∈hlγjϕkjpj−kj+cj+μjcj≥maxkc∈Bkiciϕkps+μic. Turn MathJax on It is possible to show that even when ϕq > 1 + αμh, if the spot market price ps > q − 1 / ϕ, then credit will not be offered to myopic farmers in stationary equilibrium. Put differently, if the spot market price were unexpectedly to jump above q− 1/ϕ, myopic farmers will default and contracting will break down. To conclude, economies in Sub-Saharan Africa are characterized by the lack of formal financial markets, in particular credit and insurance markets. Households thus have very poor access to these formal institutions and the ranges of products they offer. In addition, the governments in these economies do not play a strong role in improving financial access and there is a lack of public banks that could potentially play this role. As a result in these economies, informal institutions often endogenously arise to fill these gaps and it is important for policy makers to understand the reasons behind their emergence. We think models like the one presented in this paper can be applied to, for instance, the literature on informal risk sharing institutions (the seminal piece in this literature is Townsend, 1994; other important work is de Weerdt and Dercon, 2006, Fafchamps and Lund, 2003, Goldstein, 1999, Grimard, 1997, Ligon, 1998, Ligon, 2001, Morduch, 1999, Morduch, 2001 and Suri, 2011 among others). All these models of informal risk sharing depend on there being transfers between households that can maintain the informal contracts and thereby lead to consumption smoothing which results in welfare gains. These risk sharing institutions may not be fully efficient, for reasons such as moral hazard, limited liability, hidden income and transaction costs see (for example Jack and Suri, 2011, Kinnan, 2010, Ligon et al., 2000 and Ligon et al., 2002). However, there has been no work that we are aware of in this literature that examines whether the form of these transfers matter—they can be either in cash or in-kind (labor sharing for example) or both. The form of the transfer may have important ramifications both for static and dynamic incentives on both sides of these contracts. We see this as a natural next extension of our theoretical contributions in this paper. As an elaboration on this, consider Ligon et al. (2002) who present a model studying informal insurance arrangements under limited commitment. Here, people may join a risk-pooling network because welfare gains can be derived from their participation, yet they are tempted to default afterwards when they realize they have to make a positive contribution to the network. As in our model, default is avoided only if the long-term benefits exceed the short-term costs, that is, if promises of future reciprocation are perceived to be credible and sufficiently attractive. Ligon, Thomas and Worrall's model can fully explain the dynamic response of consumption to income, however, it is unable to explain the distribution of consumption across households. Thus, there seems to be scope to generalize their model potentially leading to a closer fit of the data. Our results suggest one such avenue. Mutual aid among households need not be strictly financial, it can also be in the form of time and labor that may be required during emergencies, illness or death.31 This can be incorporated into their model by allowing for household production which depends both on capital and labor where both are subject to shocks. The type of transfers also plays a crucial role in direct studies of transfers and remittances. Here, migrants need to overcome the moral hazard problem arising from the fact that they cannot observe or control how the money remitted is being spent. This moral hazard problem arises as members of the household need not have aligned preferences. In an interesting empirical paper, Ashraf et al. (2011) use a field experiment to study what happens to remittances when migrants sending money to their families can do so through multiple channels which gives them more control over how the money will be spent. They find that the additional control induces migrants to accumulate more in savings (i.e. they remit more). In our model, buyers provide both cash and in-kind credit for similar reasons. Thus, a potentially promising direction for future work is a theoretical analysis aiming to uncover the role played by the types of transfers in general risk sharing and investment relationships between individuals in developing economies. Table 1. Main Pineapple Exporters (> 80 tons/annum), 2000. Exporter Quantity Value ($) (000's) Value (cedis) (in) Interview Have Outgrowers Jei River 6342.5 2,544 12,931 No No Milani Impex 2907.5 1,314 7,065 Yes Yes Farmapine 2974.98 1.303 6,828 Yes Yes Koranco 3147.4 1,115 6,122 No - Georgefiek 2113.5 968 5,177 Yes Yes Prudent 2124.97 788 4,438 No - Unregistered 1368.4 596 2,657 No - Greenspan 1365.5 570 2,848 Yes Yes John Lawence Farms 734.2 334 1,872 No - Pioneer Quality Farms 852.7 326 1,766 Yes Yes Integral Ghana Ltd. 535.4 239 1,244 Yes Yes Burt & Baker Farms 500.9 220 1,227 Yes Yes Combined Farmers 397.96 179 937 No - Gannat Farms 372.2 152 923 Yes Yes Ultrafresh Farms 309.4 121 764 No - Bomart Farms 294.2 107 570 No - Mashaco Limited 200.1 88 454 Yes Yes Lartey Associates 144.8 82 316 No - Jesflam Farms 137.5 68 343 No - SOA Farms 143.8 64 410 Yes Yes Horizon Farms 153.1 62 381 No - Kalmoni Farms 151.3 48 292 No - Chartered Impex 87.8 47 312 Yes Yes Opintin Farms 103.8 40 217 No - KA Farms & Mechanical 83.6 39 184 No - Vilawoe Farms 99.99 37 163 No - Farmex Limited 97.0 31 141 No - Evelyn Farms 117.5 18 103 No - Average per Main Exporter 995.1 410.7 2167.3 Average per Main Exporter Interviewed 1095.9 481 2568.5 Industry Total 28511.6 11,853 62,674 Source: Ghana Export Proinotion Council (2001).