Rural institutions, agricultural development and pro-poor economic growth

James Roumasset
University of Hawaii

Conventional wisdom: invest and intervene

According to conventional wisdom, the ideal form of pro-poor economic development is through investment in agriculturally led growth.^[8] In the early stages of growth, greater production decreases food prices and shifts out the demand for labour. Inasmuch as poor households disproportionately consume food and earn a relatively large share of their income from labour, both mechanisms benefit the poor. Agricultural economists typically recommend a panoply of government interventions to go along with the investments in new technology and infrastructure, including price support and stabilization schemes, credit and input subsidies and crop insurance. The interventionist policy recommendations, however, are based on a variety of misconceptions and misinterpretations about farmers’ behaviour and rural institutions.

The interventionist doctrine for agriculture and rural development has remained remarkably resilient in the face of policy liberalization and globalization that took place in the 1980s and 1990s. Agricultural economists continue to justify regulations and subsidies of all kinds, presumably contributing to agriculture’s resistance to the liberalization and globalization of industry. Concurrently, donor support for agricultural development has waned. Three factors may account for this. First, the interventionist doctrine was at odds with prevailing "neo-liberal" attitudes. Second, there was growing dissatisfaction with the performance of many of the agricultural projects and programmes. Third, many observers concluded that low agricultural prices signaled success and that further efforts were unnecessary.

The following section reviews some of the intellectual failures contributing to the popularity of interventionism in agricultural development circles and provides specific examples of how faulty reasoning has led to policy failures in factor and output markets. The paper then shows how some of the very institutions and phenomena that have been used as evidence of inefficiency are in fact consistent with efficiency. The review so provided exemplifies a fundamental framework for policy analysis known as the "new institutional economics" (NIE).

Intellectual failures and challenges

In the 1950s, 1960s and 1970s, the economics of agricultural development called for a major role of government in providing essentials (incentives, transportation and marketing, new technology and access to inputs) and accelerators (extension, credit, irrigation, farmer cooperatives and development planning).^[9] Johnston and Mellor (1961) and Mellor (1966) emphasized the positive linkages between agricultural growth and economic development and continued to presume that a wide variety of government regulations and subsidies were appropriate to get agriculture moving. This legacy continued even into the 1980s as agricultural economists continued to argue for pushing the agricultural sector but were somewhat indiscriminant about the appropriate instruments for doing so. For example, a major collection of readings (Eicher and Statz 1984) failed to note the excess burden and dynamic costs of agricultural protection even as these became the focal point of industry and trade policies.

In the late 1980s and 1990s, new theories that took account of imperfect information augmented intellectual support for interventionism. The most general interventionist doctrine was based on the Greenwald-Stiglitz (1986) theorem which states that a competitive equilibrium is not constrained Pareto-optimal, i.e. is not on the feasible utility-frontier, whose limits are determined by feasible government actions as well as technology, factor endowments and consumer preferences. This theoretical result was interpreted to mean that government can always find a coercive intervention to increase economic efficiency over that achieved by voluntary contracting and competitive markets. Stiglitz (1993, 2002) himself has often used the institution of share tenancy to exemplify how economic organization can be in equilibrium but massively inefficient, asserting that a landlord’s share of one-half would have the same disincentive effects as a 50 percent income tax. In this new information economics, market failures are not limited to the usual cases of externalities, public goods and non-convexities, but also include the far more pervasive failures due to moral hazard, adverse selection and other information problems.^[10]

Similarly, de Janvry et al. (2001), while acknowledging the role that transaction costs play in rural organization, nonetheless concluded that "indirect sources of market failure need to be eliminated" including those that plague credit and insurance markets. This and the paper by de Janvry and Sadoulet (2000) have been misconstrued to mean that government should intervene in such markets with mandates and subsidies (see e.g. Weber et al. 2002). Some investments in agriculture, notably in agricultural research, were prematurely rejected in this view as mere "technofix".

These propositions are subject to Nirvana Fallacy (Demsetz 1969), however. The equilibrium concept in question is a straw man in two important respects. First, it doesn’t admit multilateral voluntary contracting. Second, it doesn’t admit private governance of moral hazard and other information problems, e.g. as described in Jensen (2000). Even if the Greenwald-Stiglitz theorem were generalized to allow for multiple distortions and even if some pervasive efficiency-improving interventions were found, the results would still suffer from "blackboard economics". (Note that "blackboard economics" should not be taken as a general condemnation of rigor but rather of equilibrium concepts that abstract from real world institutions and which internalize spillovers and mitigate information problems.)

In the following section, I review some of the intellectual failures in the context of agricultural and rural development. All of them result from misplaced exogeneity and a failure to provide a fundamental explanation for the phenomenon at issue.

Land and labour institutions

Asian agriculture displays the coexistence of disparate property, tenure and contractual institutions that connect labour to land. While most Asian agriculture is smallholder in character, with notable exceptions of growers of tree crops such as oil palm and coconuts, there has recently been an increase in the number of larger commercial farms, e.g. for the production of sugar in Indonesia (Fairhurst 2003). Instead of explaining the diversity, however, much of the economics of agricultural development have sought to identify which behaviours and institutions are inefficient, much as the old industrial economics regarded market structure as exogenous and proceeded to examine the conduct of different organizational forms and to evaluate the efficiency of their performance.

For example, most agricultural economists assert that smallholder agriculture is inherently more efficient than large-scale commercial farming because it economizes on hired labour.^[11] Utilizing family labour economizes on recruiting and supervision costs - the latter is said to be so because hired labour supposedly suffers in terms of both quality and an inherent tendency for effort shirking. These labour market imperfections supposedly result in the "productive superiority of family farms" (Deininger 2003, p. 84) and to the characterization of hired labour as inefficient (Otsuka 2002). Using the ICRISAT village data, Frisvold (1994) finds that family labour is indeed more productive than hired labour even before deducting the costs of supervision. The inefficiency of hired labour is also said to be at least partially responsible for the notorious inverse relationship between small and large farms, assuming that the latter are relatively more labour-dependent (Otsuka 2002; Deininger 2003). Similarly, Hayami (2003) finds that while plantation agriculture was an efficient institution for the exploitation of Western colonies in Asia, family farms have more recently "proved to be equally or more efficient producers of tropical export crops using family labour of low supervision costs, relative to plantations based on hired labour."

However, these studies fail to account for why labour is hired (and for which tasks) and for the incomplete substitutability of hired and family labour. They also fail to account for the role of land quality in crop choice and intensity of cultivation. It is not surprising, therefore, that one can find contradictory empirical results. Indeed, Benjamin (1992) finds that hired labour is significantly neither more nor less productive than family labour. This may simply be because there are both gains and losses involved. For example, hired labour facilitates specialization. In a prototypical farm where both family labour and hired labour are employed, rational choice implies that there will be a non-random division of tasks between the two types and that, at the margin, the difference in their productivities will be equal to the difference in opportunity costs.

Share tenancy is another institution that is commonly attacked for being inefficient. The literature has been unduly influenced by Stiglitz’s (1974) canonical model wherein sharecropping is viewed as a pairwise-efficient means of providing labour greater incentive to work (or not to shirk) relative to wage contracts but without the cost of risk-bearing that would be imposed under rent contracts. After reviewing the leading theories of share tenancy, Hayami and Otsuka (1993) conclude that the risk-aversion vs. moral hazard model indeed "justifies the existence of share tenancy in the theoretically most consistent manner..." As noted above, Stiglitz (1993, 2002) remains convinced that the Marshallian effort disincentive of share tenancy is socially inefficient. The inefficiency hypothesis has been further buttressed by econometric studies, most notably Shaban (1987) for the case of India. Jacoby and Mansuri (2002) report similar results for Pakistan. Bautista (1991) observes that share tenancy in the Philippines is less productive as well as inequitable.^[12]

As is the case with the literature on the inefficiency of large farms and hired labour, however, this conclusion is premature. First, the canonical model does not imply, as originally claimed (by Stiglitz, in 1974), that the optimal landlord’s share varies positively with the tenant’s degree of risk aversion, because risk aversion also blunts the tenant’s incentive to shirk. Second, the model is incapable of explaining the empirical or actual distributions of tenant shares, which cluster around of 50 percent, with a smaller cluster around two-thirds.^[13]

A more fundamental problem is that the canonical theory treats share tenancy as a mere labour contract and thereby misses its essence as a typically long-term contractual arrangement for bringing management together with land that facilitates the tenant’s learning-by-doing of production decisions (Reid 1976; Murrel 1983; Eswaran-Kotwal 1986; Roumasset 1995). Share tenants themselves hire
substantial amounts of labour, especially for the more arduous and routine tasks. Share contracting is a popular labour contract for specific tasks. Indeed, share tenants often hire casual workers on a share basis to do harvesting, weeding and transplanting. But receiving a share of the harvest does not make such workers tenants.

The persistent fallacy in all of the inefficiency arguments is one of misplaced exogeneity.

Trying to judge the inherent efficiency of particular institutions is tantamount to the old structure-conduct- performance paradigm whereby a market structure was taken to be exogenous, its conduct diagnosed and its resulting performance judged. For example, the conduct of monopoly is characterized as increasing price by lowering quantity below its competitive level and its performance is judged to be inefficient. As the above examples illustrate, this paradigm, while now defunct in industrial organization, is alive and well in development economics.

Even leaving the identification problem aside, the tenant’s compensation is not necessarily limited to his share of the harvest. Asian tenants often receive credit from their landlords at concessionary rates (often zero interest) and landlords help with tenant family’s needs and emergencies (Roumasset 1976; Sadoulet et al. 1997).

Credit and marketing

Credit and marketing institutions are similarly castigated as exploitative and inefficient. The stereotypical middleman charges excessive interest rates for credit and pays the farmer pitifully low prices. The following description (about Pakistan) is typical:

...owing to the involvement of many layers of middlemen between the growers and the consumers, every year the government has to intervene in the agriculture commodity markets to rescue the farmers from the clutches of the middleman by acting as a second buyer. (Badar 2002)

In Southeast Asia, such claims are often directed specifically at ethnic Chinese: "It is not unusual to hear ... that farmers or consumers are exploited by ...Chinese middlemen."^[14]

Hayami and Kawagoe (1993) have documented how, in general, "the stereotype has not held up under empirical tests", particularly so in Indonesia.^[15] They go on to document the nature of marketing operations in Western Java and Sumatra. The stylized marketing organization relies on village collectors (often themselves farmers), intervillage collectors, traders and processors. Because the village collectors have a low opportunity cost of time, traders readily adapt to the demands of marketing entrepreneurship, and because institutions and dynamic relationships are developed to provide quality control and mitigate the "holdup" problem, the marketing system tends to efficiency. The main obstacle to efficiency in this view is the tendency of governments, in the alleged quest to limit excesses of the ubiquitous middleman, to actually suppress entry and the natural evolution of appropriate institutions and entrepreneurship. Rather, government policy should be focused on increasing entry and fostering market integration through appropriate contractual and physical infrastructure as well as on providing market information and facilitating standards and grading.

In the current era of globalization, the efficiency of small-scale marketing systems may be in decline, however. In traditional marketing systems, production is indirectly coordinated with the marketplace only through successive layers of collection and distribution. Smoothing fluctuations in both demand and supply is done through inventories and through international trade. With the rise of supermarkets and "big box" discount stores, however, Reardon et al. (2003) have shown that retailers often interact directly with producers for delivery of goods, processed and packaged to specifications, at particular places and times. This confers competitive advantages to larger producers and partially displaces traditional marketing systems.

The Berkeley/World Bank group of applied economists acknowledges economies of scale in agricultural marketing but denies that these undermine their conclusion that large farms are inefficient, asserting that farmer associations can exploit large-scale marketing opportunities.^[16] This is a remarkable inconsistency. Small farms are said to be more efficient because they avoid the additional contracting costs associated with hired labour. But the proposition that small farms can simply overcome diseconomies through contracts blithely ignores the requisite contracting costs. Indeed farmer cooperatives are notorious for broken agreements and favouritism, both of which undermine the sustainability of group contracts.

The "evil middleman" syndrome has similarly led to widespread interventionism in credit markets, in particular the "directed credit" syndrome whereby interest rate ceilings are combined with concessionary lending to rural banks, which qualify for subsidies by targeting rural and agricultural clients. While this particular policy failure has been widely diagnosed, disagreement remains between intellectual supporters of interventionism and those who trust credit mobilization and allocation through competitive markets.

The "Ohio State School" asserts that the high rates in the informal sector are warranted by transaction costs and the risk of default. They note that the low interest rates mandated by government regulations direct credit primarily to larger commercial farmers and other borrowers with above average incomes (Meyer and Nagarajan 2000; Coleman 2002). Accordingly they advocate spontaneous institutions such as microcredit programmes and competitive market allocation of loanable funds.

The intellectual climate regarding credit policy may have swung too far towards the laissez faire extreme, however. For example, the donor consortium CGAP has come out with a set of "best practices" based on the "win-win" approach to rural credit.^[17] In this approach (also called the "new paradigm" by Meyer and Nagarajan [2000]), rural lending institutions should attain financial sustainability by eschewing government and donor assistance and charging rates commensurate with the full cost of the loan and high enough to successfully mobilize savings. As Morduch (2000) points out, this approach is mandated neither by logic nor empirical evidence. First the goals of financial sustainability and growth through profitability are not coincident with maximum impact on the poorest of the poor. Moreover, microfinance success stories have tended to "stretch accounting data in order to claim profitability" (Morduch [2000], p. 627). As a result, microfinance organizations have attempted to replicate apparent success stories, albeit with disappointing results. "Some donors believe that little more than 5 percent of all programmes today will be financially sustainable ever" (Morduch [2000], p. 618). Morduch (2000) argues instead that financial sustainability and programme expansion are consistent with some degree of subsidy. What is important is that the subsidy be a "hard budget constraint" and that sound prudential management be maintained, including selection and monitoring procedures that emphasize repayment. If these fundamentals are kept, the resulting diversity of programme designs will contribute to the evolution of successful approaches.

On the other hand, "second-best" interventionists argue that market institutions are inefficient due to problems of imperfect information (e.g. Stiglitz and Weiss 1981). Ray (1998) reviews both the theoretical literature and empirical studies from Asia and concludes that rural credit institutions display substantial inefficiency even after uncertainty and transaction costs are taken into account. Accordingly, Stiglitz and Uy (1996) argue for "mild" financial repression with both interest rate ceilings and policy discrimination across types of investors.

What is needed to progress from this impasse is a conceptual framework that is capable of evaluating the consequences of alternative credit policies. The theory must be able to explain the coexistence of formal and informal institutions for rural credit and other patterns that characterize the nature of credit institutions under a variety of policy umbrellas.

Among the many "market failures" and alleged justifications for government intervention, perhaps the most misunderstood concerns stabilization policy. One of the common justifications of a state trading enterprise to control domestic rice markets in Asia, for example, is that without government control, market prices would be unacceptably volatile. It is surprising how readily this justification is accepted without a compelling rationale that governments can and should control prices. It is also rather remarkable that while the intellectual climate regarding credit policy has largely swung to non-interventionism, the intellectual climate for stabilization (like that for intervention in land and labour markets) has proved to be more resilient.

The case for government stabilization of prices is weak at best. If the source of domestic price instability is international price variability, even costless stabilization would be welfare reducing. Consumers gain more from low prices than they lose from high prices. The reverse is true for producers. If domestic supply were the source of unstable domestic prices, price stabilization via a costless buffer stocking scheme would be welfare increasing, but of course no such free lunch exists. Feasible acquisition and release strategies are likely to be welfare reducing even when they work, due to the limited degree of stabilization and high costs. Moreover, empirical evidence suggests that attempts to stabilize grain prices do not succeed (Roumasset 2000, 2003b), and theoretical analysis shows that stabilization strategies involving buffer stocks tend to be destabilizing in the long run due to the probability that stocks, storage capacity or available budgets will eventually be exhausted (Wright and Williams 1990).

Policy failures: a synthesis

The intellectual failures reviewed above include market failure, behavioural failure and institutional failure. All of these result from misplaced exogeneity. A full understanding of policy failures, however, goes beyond diagnosing errors in economic reasoning. Political economy instructs us that bad policy results from rent-seeking as much as bad economics.^[18]

Directed credit programmes, for example, may have been justified by defunct economics but served as viable mechanisms for political patronage in many Asian countries. Subsidized interest rates resulted in excess demand for loans. Inasmuch as the programmes are "directed", there is room for rationing of loans to be done on the basis of various indicators of political loyalty instead of potential investment productivity. In an extensive review of several Asian economies, Meyer and Nagarajan (2000) characterize the predominant form of bank lending to the rural sector in the 1960s through the 1980s as targeted (e.g. to farmers), funded by governments and donors at subsidized rates and typified by negligible selection and monitoring procedures. Borrowers correctly perceived the programmes as entitlements, not obligations, and repayment rates were extremely low, with the exception of economies with "strong civil and professional traditions" such as Republic of Korea and Taiwan Province of China. As described above, this approach hampered the natural evolution of both the formal and informal sectors.

Similarly in Pakistan, the Agricultural Development Bank of Pakistan (ADBP), which provides most formal loans in rural areas, lends to large landowners much more than to small landowners (Faruqee and Khandker 2001). Large borrowers with lower marginal benefits use formal loans unproductively and have high rates of default. As a result, the ADBP’s operations impose a heavy burden on the government because of large subsidies required to sustain its operations every year. The Asian Development Bank is now recommending not subsidizing interest rates in rural finance operations (Asian Development Bank 2003).

Rural credit programmes in the formal sector have expanded substantially in most Asian countries, but it has been mostly short-term credit targeted to farmers. Because of subsidized rates and poor prudential practices, these programmes have not been financially sustainable. Rather, programmes are renewed, renamed and revived through additional tranches from international donors or the general funds of governments. The "band-aid" response of international donors in the 1980s was to seek to make small farmers more creditworthy by subsidizing ambitious programmes on formal land titling (e.g. in Northern Thailand). Feder et al. (1988) argue that simultaneously subsidizing the establishment of formal land titles and otherwise expanding formal lending improves welfare by funding agricultural investments with high present values that had formerly been rationed out of the credit market. Econometric support for such claims is not founded on any viable theoretical construct, however, and remains suspect. Models are needed that can rationalize the coexistence of formal and informal credit markets and that can be used to examine the consequences of subsidies, regulations and changes in property rights.

Another area of policy failure in Asia is land reform. In the Philippines, for example, land reform outlawed share tenancy. As a result, land reform beneficiaries hired permanent workers who were paid a fixed amount for the season. Hayami and Otsuka (1993) conclude that this has been an inferior substitute for share tenancy.

Another Philippine example concerns the failure to consider properly basing landlord compensation on land quality. By basing compensation on the principle that 25 percent of yield is a fair rent, reform confiscates value from owners of good and average farms but actually over-rewards owners of poor quality land (Roumasset and James 1979). As a result, friends and relatives of owners of poor quality land submit bogus claims that they have been working the land as tenants so that the landlord receives more than the land is worth (and landownership remains in the family).

There is a good reason, however, why politicians embrace the bad analysis supporting land reform. The implementation of land reform has always been very spotty. The administration in power can be very strict towards its enemies in the implementation of reform and very lax with its friends. Thus land reform becomes a potent political weapon.

Another persistent policy cockroach^[19] relates to the attempt to control agricultural prices through government parastatals, which are tasked with the impossible mission of maintaining high and stable producer prices as well as low and stable consumer prices (Roumasset 2000). Trying to distort and stabilize prices by prohibiting private trade and enabling parastatal monopolies has the undesirable effect of fragmenting markets and blunting incentives for farmers and the agribusiness sector. Attempts to control prices can decrease the welfare of consumers, producers or both. Moreover the inframarginal nature of price controls that results from limited parastatal resources implies that favourable inframarginal prices will be conferred on those who have gained a political advantage or whose political favour is curried by politicians and bureaucrats.

The antidote to so-called "blackboard economics" is methodological fundamentalism (Nozick 1975).

Economic cooperation in agriculture is more complex "than is imagined in your calculus, Horatio."^[20] The principle of comparative advantage implies that different characteristics of land and landowners will call for different intensities and composition of inputs and organizational forms with unlimited differences in architecture. Judging the relative efficiency of different organizational forms commits the most fundamental fallacy in economics - judging performance without understanding the nature and causes of the phenomenon of interest. Prescribing policy reforms based on the premise that politicians, bureaucrats and academics can socially engineer institutions superior to those shaped, tested and improved in the crucible of evolution is a recipe for government failure. The new institutional economics provides an alternative paradigm that encourages greater caution in tinkering with institutions that have evolved in the crucible of competition.

The new institutional economics of agricultural organization

The alternative to misplaced exogeneity involves characterizing the true nature and seeking the fundamental causes of behaviour and organizational differences. In a cross section of farms, for example, questions to ask include: Which type of land is allocated in large parcels, to which economic actors, and why? How has the composition between family and hired labour changed and why? Under what conditions do landlords choose to contract with tenants to manage their land?

The central decision-making model of development microeconomics is the farm-household model. A simple version is depicted in Figure 1, which shows the household labour supply schedule of a representative farm-household and three possible labour-demand schedules, depending on (quality-adjusted) farm size.

For D₁, the family exports its excess labour and the relevant shadow price of labour is w_s, the "selling wage" after deducting journey to work and other necessary expenses from the nominal wage. For D3, the farm-household imports hired labour, and the shadow wage is w_h, the hiring wage after including the employer’s agency cost, recruiting and supervision costs and the residual costs of labour shirking (see previous section). If labour demand intersects household supply in the intermediate range between w_h and w_s, the shadow wage rate is given by the household’s marginal opportunity cost of labour.^[21] Accordingly, the rational farm household can be said to be maximizing shadow profits, based on the shadow wage schedule,

w = w_s, L < L₁
w_h, L > L₂
S_L, L₁ < L < L₂

The profit maximization problem of the farm is only quasi-separable from the household utility maximization problem inasmuch as the labour supply schedule is not independent of farm income.

Similarly, the household-farm produces the (shadow) profit maximizing quantity of the agricultural commodity, where the shadow price is bounded by the buying price and the selling price and coincident with the household demand schedule in between. Again there is a limited source of non-separability inasmuch as household demand is dependent on farm income.

The "wedge model" - where transaction costs drive a wedge between prices - contrasts with the household-farm model of Lau et al. (1981) and Ahn et al. (1981) wherein household consumption is determined recursively, based on the profit-maximizing behaviour of the farm. Nonetheless a recursive algorithm can be employed to solve the wedge model, albeit by guessing household consumption and iterating until the guessed consumption level is consistent with both the household utility function and shadow-profit-maximizing farm income.

Figure 1. Quasi-separability of farm labour demand and farm household supply

However, the wedge model begs the question regarding determination of the unit transaction cost wedge. That is provided by agency theory.

Figure 2 illustrates agency theory in the context of alternative labour contracts. Piece rates are commonly used in situations where the product of labour is easily observable, for example, sizing and sharpening the cane stalks prior to planting and planting stalks at uniform spacing. These tasks are tantamount to intermediate products delivered to the farm operator, who pays according to quantity. This institution economizes on minimum agency cost (MAC), i.e. the minimum sum of monitoring cost and (quality) shirking cost. For tasks that are not amenable to ex post inspection, supervision is used to concurrently monitor the labour activity in question and workers are paid according to the time spent on an activity, not its result.

The four panels illustrate the comparative statics proposition that if tasks are sufficiently easy to monitor through ex post inspection then the corresponding agency cost at optimal monitoring will be lower than the agency cost under wage contracts. The opposite is true for tasks that are hard to monitor. For each task, the unit transaction cost is given by the least of the two minimum agency costs for the task in question.

The wedge model can be used to explain behaviour of the household-farm, the basic building block for theories of agricultural development. The agency cost model can be used to explain rural institutions. Both are essential for understanding the consequences of contemplated policy reforms.

New institutional economics (e.g. Roumasset 1978) also recognizes that different levels of analysis may be appropriate for the analysis of different problems. Models that recognize transaction costs such as the two above are classified as "second best".^[22] When the subject of inquiry is the terms of agricultural organization (e.g. tenants’ and harvesters’ share of production), the "first-best" model, which abstracts from transaction costs, has been found to be appropriate. In first-best analysis the terms of contracts are set such that factors receive their marginal products just as if there were competitive markets.^[23] "Third-best" analysis or political economy allows for multilateral opportunism in the pursuit of favourable government treatment by special interests (Dixit 1996). The following three sections rely primarily on the second-best level of analysis. A brief synthesis using all three levels of analysis is thereafter provided.

Figure 2. Specialization of contracts by task

Land, labour and the nature of the farm

Consider the evolution of hired labour. In Marxist view, the new rice and wheat technology that swept through Asia in the 1970s disenfranchised the peasantry and led to falling wages and increased unemployment. In the "induced innovation" view (Binswanger and Ruttan 1978; Ruttan 2003), the causation was just the reverse. Population pressure on limited land resources drove down wages thereby inducing land-saving technological change. In effect, this allowed "biological capital" (modern varieties and chemical inputs) and labour to substitute for land. The increased demand for labour had a positive effect on wages but was just not enough to offset the effect of population pressure (Hayami and Kikuchi 1982).

The induced-technological-change explanation just described is a first-best argument. However, not only did labour per hectare increase, its composition changed dramatically. In the ten years following the adoption of the new rice varieties in the Philippines, hired labour in weeding for example increased from less than 20 percent of total labour to more than 80 percent (Roumasset and Smith 1981). Figure 3 illustrates the use of the wedge model to explain this dramatic institutional change.

Figure 3. HYVs and the advent of labour markets

The graph represents a typical farm household in the province of Laguna and shows how four factors combined to increase hired labour dramatically. First, and most importantly, the intensification of production, ultimately caused by increasing land scarcity and accommodated by the new rice technology, increased the demand for labour per hectare. This is illustrated by the shift in the demand curve to the right. Second, increased farmer incomes resulted in increased schooling of farm children. This combined with the increased specialization among farm workers lowered the amount of farm-household labour per hectare. These higher opportunity costs and lower substitutability for skilled labour are illustrated by the shift in the labour supply curve to the le.. Third, the market wage went
down (from Wm0 to Wm1) as population growth, including in-migration, increased by more than enough to supply the increased labour demand. Fourthly, the transaction cost wedge between the market wage and the gross hiring wage decreased due to the advent of labour contractors and other new institutions of labour contracting (Roumasset and Uy 1987), illustrated by a downward shift in the gross hiring wage (from W_h0 to W_h1).

As hired labour increased, a menu of agricultural contracts emerged for "incentivizing" (i.e. providing incentives to) labour in different tasks. We have already discussed Figure 2, which shows how agency theory can be used to explain the tendency for piece rate contracts to be chosen when the task amounts to delivering an observable intermediate product. Statistical analysis of sugarcane contracts in the Philippines confirms this tendency (Roumasset and Uy 1980). For example, cane stalks are prepared for planting (uniformly sized and sharpened) and laid out for inspection. The farm operator simply inspects them for quality and uniformity. Next the stalks are planted, and the operator inspects for proper height and spacing.

Gama or Ilani, as practiced in the Philippines, is an institutional arrangement whereby the worker contracts to weed and harvest a specified parcel of land for typically one-sixth of the rice harvested for that parcel; ceblokan, practiced in Indonesia, typically requires transplanting in addition to harvesting and weeding for the same one-sixth share (Roumasset 1978; Hayami and Kikuchi 1982).^[24] These arrangements were preceded by hunusan in the Philippines and bawon in Indonesia wherein only harvesting was done for the share of the harvest, typically one-sixth. Before the new institutions of gama and bawon, the share was sometimes lowered to one-eighth (Roumasset 1978).

Why did the share settle at one-sixth with the amount of work increasing instead of the share of the harvest simply declining? Hayami (1998) suggests that another function of gama/ceblokan was to provide an explicit selection mechanism for choosing who would weed/harvest and to allocate a specific parcel to each group of workers. In addition to selection, this provides improved incentives over the hunusan/bawon systems that were open to anyone in the village. Under the old system, a kind of free-riding occurred wherein workers would harvest faster than efficiency warrants just to be able to harvest more. Having workers harvest the same plot that they weeded (and sometimes transplanted) provided additional incentives to weed/transplant with greater care. Thus, while first-best principles can explain either the falling harvesters share or the increased work required, second-best considerations are required to understand why one institution was favoured over the other.

Figure 4 provides a second-best efficiency explanation of the institution of share tenancy.

The larger the tenant’s share, the lesser is the agency cost of labour shirking (monitoring cost plus residual shirking costs). On the other hand, the greater the tenant’s share, the greater the tenant’s incentive to overuse (or undermaintain) land quality. Share tenancy with a tenant’s share of roughly one-half minimizes the agency cost. There is nothing inherently inefficient in the contract, just explicit recognition of the contracting costs inherent in specialization.

Inasmuch as the tenant is the farm manager, not a worker, it is futile to classify "forms of tenure" as share tenant, lessee and wage worker. Rather, we need to classify organizational forms by which ownership, management and labour are connected. Figure 5 illustrates a taxonomy of firms classified according to degree of specialization. Note that pure owner-operator and owner-manager are on opposite sides of the specialization spectrum even though the conventional taxonomy classifies them both as owner-operator. The pure owner-operator household provides all the management and all the labour - i.e. there is no hired labour. The owner-manager hires most of the labour and reserves for himself only those tasks that are bundled with managerial discretion, e.g. fertilization. Share tenancy is characterized by an intermediate amount of specialization - the tenant does most of the management, all discretionary tasks and some other tasks, e.g. land preparation.

Figure 4. An electric theory of share tenancy

Figure 5. A spectrum of agricultural firms

Evidence from the Philippines and Nepal confirms that specialization is driven by intensity of cultivation, which is driven in turn by favourable land quality, location and economic environment (Roumasset 1995). Intensification can also be driven by population pressure, demand growth and rising land values. Not only does intensification warrant more specialized agricultural firms, but the organization of hired labour itself becomes increasingly specialized. This is elaborated further in the succeeding sections.

The evolution of lending institutions

In applying the new institutional economics to credit markets, the first task at hand is to rationalize the coexistence of the informal and informal sectors. It is natural to assume that the formal sector specializes in enforcement through the formal sector (e.g. through legal foreclosure procedures) and that the informal sector specializes in more personalized mechanisms such as repeated transactions, reputation and idiosyncratic bonding devices.^[25] Formal institutions such as rural banks concentrate on production loans. The informal sector lends to relatively poorer households for both production and consumption purposes and at high unsubsidized rates.

The widespread policy of usury laws and subsidized rural banks in Asia has perverse effects on both the formal and informal sectors from the perspective of the model just described. The natural evolution of banks will be directly jeopardized by subsidizing banks that charge low interest and compete for the same customers as banks that rely on savings mobilization, charge borrowers higher rates and aim for financial sustainability. Inasmuch as mobile factors such as loanable funds and skilled labour are drawn from the unprotected sector to the protected sector, subsidizing the formal sector also stunts the growth of the informal sector instead of expanding it so as to widen access to commercial credit.^[26] But instead of letting these failed credit programmes die a natural death, donors have subsidized new programmes (e.g. formalizing land titles) and have justified new tranches of funds for directed lending, thereby inhibiting natural market development for an even longer period.

Moreover, the new loans are disproportionately given to those with previous dealings in the formal sector; these displace informal loans whose enforcement depends on personalized information and repeated interactions. Thus the interventions tend to shrink the informal sector and its high shadow price of credit and expand the formal sector, characterized by a low shadow price.

Instead of measures that artificially fragment credit markets and penalize the informal sector, what are needed are policies that deepen credit markets by building on existing institutions. At any given level of market development, shadow prices of credit differ across both market lenders and borrowers. Institutional development occurs when the benefits of arbitraging across different shadow prices is greater than the additional governance costs of the new institutions.

The nature of economic integration: transaction costs and specialization

In modern parlance, the classical engine of growth ala Adam Smith is "falling unit transaction costs", which facilitate ever-increasing transactions and specialization of economic organization (Yang and Ng 1993). This proposition emanated from the new institutional economics and was used to explain the role of labour specialization in agricultural development (Roumasset and Smith 1981). In Yang’s formalization (2003, ch.4), unit transaction costs are driven down by the endogenous emergence of middlemen, whose specialization is warranted by the extent of the (growing) potential market. For example, the institution of piece rates with teams (Roumasset and Uy 1980) economized on labour recruiting and supervision costs by relying on direct contracting between the farm operator and the team leader, who maintained a reputation for reliability.

Figure 6 provides a stylized evolutionary pattern of labour contracts. During Stage I, labour is provided by the farm household; exchange arrangements are then additionally made with residents in the same village. During Stage II, i.e. the next three rows of Figure 6, hired labour emerges. At first, labour is hired on a wage basis and workers are not differentiated with respect to task. As horizontal specialization increases, piece rate workers are hired for selected tasks (those which are relatively easy to monitor) and undifferentiated wage labour declines. The third phase of Stage II involves a further decrease in undifferentiated wage labour, a decline in individually hired piece workers and the advent of two new contracts. In "piece rate with team labour", the farm operator negotiates with a labour contractor who also serves as team leader and supervisor. The other new form (specialized wage labour) involves skilled labourers who specialize in particular tasks and get paid in wages. These new forms come to dominate the other forms of hired labour in Stage III. Piece rate with teams continues to replace individual piece rate contracts whereas specialized wage labour replaces undifferentiated wage labour and most of household labour.^[27]

Figure 6. Intensification and specialization

The explanation of the above dynamic pattern of labour contracts is similar to the agency theory explanation of the spectrum of agricultural firms (Figure 5). In both cases, the objective is to explain a spectrum of contracts ranked according to specialization. In the cross section case, the same preconditions for production intensity (e.g. land quality) also predispose a more specialized organizational form. As the profit-maximizing level of inputs increases, more production management is warranted, indicating an organizational form wherein the manager is rewarded with a larger share of the residual. That is, the agency costs associated with shirking of non-labour inputs increase moving towards better quality land, and these costs are best economized by supervising labour and incentivizing managers (Roumasset 1995).

As farm production intensifies, labour inputs increase, until the last stage wherein capital-labour substitution overcomes input intensification. Labour contracts are increasingly specialized, eventually with labour contracts made on a task-by-task basis. Thus intensification and specialization are coevolutionary. The diagram also helps to resolve the fundamental paradox that total transaction costs increase as economic development proceeds (North and Wallis 1982). In this scenario, unit costs of transportation and communication (unit transaction costs) tend to fall while improved institutions tend to lower agency costs (supervision plus residual shirking costs) per unit of labour hired, but because more labour is hired and because specialization increases the number of contracts (even if normalized by yield per hectare), transaction expenditures rise.

Note, however, that economic efficiency does not imply that shadow price differences across space and time disappear altogether. The efficiency condition is rather that such differences cannot exceed the cost of transport and storage, respectively (Kratz and Roumasset 2001). Econometric tests for market integration, using modern techniques of cointegration, appear to have failed to specify this integration hypothesis correctly. Moreover, shadow prices of inputs and outputs can vary across agents in the efficient solution according to the household wedge model discussed above.

Summary and implications for development policy

Policy failures result from a combination of bad economics and rent-seeking behaviour of politicians. Through "blackboard economics" - including misplaced exogeneity - analysts unwittingly assume inefficiency in order to conclude that market inefficiency exists while naively presuming that government actions will not be plagued by the very transaction costs that limit markets. When donors and politicians alike are in denial about their failures and throw more money at the very problems they have exacerbated, "band-aid" and "black-hole" cycles of ever greater public spending and worsening distortions are promulgated.

The key to avoiding misplaced exogeneity is to capture the essence of institutions and to provide fundamental explanations thereof. New institutional economics provides an explanatory framework with three levels of analysis. First-best analysis abstracts from transaction costs. Second-best analysis incorporates transaction costs. Third-best analysis incorporates the costs of political action and other elements of public choice. For example, the case for land-to-the-tiller reform, which is based on the inverse correlation between farm size and yield-per-hectare, can be refuted at both the first and second-best levels. On the first-best level, it can be shown that efficient organization of family farming requires that good quality land be organized in larger farms than poorer quality land (Roumasset and James 1979) thereby revealing the fallacy in the inverse-relationship-implies-inefficiency thesis. On the second-best level, the wedge model can be used to show that smaller farms face higher shadow prices of labour (Sah 1986) such that second-best efficiency implies the inverse relationship, thereby undermining the interventionist logic once again.

Third-best analysis is exemplified by the explanation of why agricultural protection increases with a country’s per-capita income (Balisacan and Roumasset 1987). In this arena of public choice, we must explicitly consider the costs and benefits of coalitional investment in political influence in order to get the appropriate comparative-statics results.

One source of confusion regarding the new institutional economics concerns the plethora of definitions of transaction costs. Transaction costs have been defined most broadly by Nobel Laureate Kenneth Arrow as costs of running the economic system and are the economic equivalent of friction in physical systems (Williamson 1985). Sublevels of transaction-like costs can also be distinguished. The first is unit transaction costs, e.g. the cost of one man-hour of supervision. Another is agency cost, e.g. the unit cost of supervision times the man-hours of supervision plus the residual shirking cost. (This concept was illustrated in the agency diagrams above.)

These distinctions make it possible to explain the essence of economic development as envisioned by Adam Smith in The Nature and Causes of the Wealth of Nations. As social capital (including infrastructure) increases, unit transaction costs fall, thereby facilitating greater specialization. In particular, the number of both final and intermediate goods increase as does the number of distinct labour tasks and opportunities for learning-by-doing. Thus economic specialization and integration are part of the same evolutionary process (Yang 2003). Total transaction costs, in the broadest sense, increase with efficient development, i.e. the income elasticity of transaction costs is greater than one. This means that economic organization gets more complex and market deepening proceeds, faster than unit transaction costs decline.

However, natural market deepening is impeded by market-distorting interventions including trade restrictions, price interventions, shipping and other regulations and failure to provide public infrastructure, including quality standards. Marketing regulations, such as those embodied by parastatals, exemplify how government policy can stagnate the natural evolutionary process and impede an industry instead of promote it. Economic integration can be enhanced by removing these policy distortions and by focusing on facilitating actions such as agricultural research and the provision of transportation and communication infrastructure.^[28]

The best stabilization programme would be to abolish parastatals that monopolize international trade in grains and eliminate government-imposed barriers to entry. This policy would not only render the industry competitive but also create a rapid-response capability to import in times of unexpectedly high domestic prices by removing the elaborate contracting, procurement, bidding and other administrative requirements that delay government purchases.

It may also be appropriate for governments to assure the maintenance of a small strategic reserve for emergency purposes. But a maximum size should be established for the strategic reserve based on the conceivable number of regions that can be in deficit at the same time, the availability of rice in the local market and the minimum delivery time of foreign-sourced grain. It is difficult to imagine how such considerations can justify stocks greater than 15 times the daily consumption rate.

By considering specialization and institutional choice as endogenous, we can understand two beneficial effects that are often overlooked. First, inasmuch as institutional change is induced by changing factor prices (Ruttan 1978, 2003) - e.g. falling wages relative to rents - it allows greater substitution of labour for land, thus partially ameliorating downward pressure on wages. Second, to the extent that institutional change facilitates specialization and the external economies associated therewith (Yang 2003), it may actually overcome the original downward pressure on wages (Roumasset and Van Assche 2003).^[29]

Conclusion: stop, push and facilitate

There has long been a tendency among economists and others to use statistical evidence and stylized facts to castigate behaviour and organization in developing countries as sources of inefficiency and inequity and to propose coercive mechanisms for reshaping the economy. These attempts illustrate that empirical analysis cannot be stronger than the underlying theory. Unless the theory accounts for the nature and causes of economic organization, econometric analysis can only deliver statistical patterns. It cannot be used as the basis of policy recommendations.

The assertion that government intervention can always improve efficiency is based on a straw man version of the market in which neither private governance nor multilateral agreements are allowed. Even if such circumscribed characterizations were accepted, the theory leads only to the claim that some kind efficiency-improving intervention exists. However, the nature of the theory and the available evidence make it infeasible to prescribe specific policy reforms or to determine their consequences (Besley 1994).

When a more fundamental approach is taken, we find substantial evidence that institutional change evolves in much the same way as would be warranted by efficiency. A healthy respect for institutional evolution leads us to the conclusion that governments should stop trying to engineer behaviour and organization. Rather the focus should be on facilitating economic cooperation through the provision of information, legal infrastructure and opportunities for multilateral cooperation. The prerequisites for cooperation will render the time-honored strategy of pushing agricultural development through investments in research and infrastructure even more effective, especially if modern principles of public administration are employed.^[30]

The first priority for policy reforms should be to roll back those regulations, excessive taxes and subsidies that inhibit the normal evolution of rural institutions and markets. Beyond this, reforms should be focused on increasing entry and fostering market integration through appropriate contractual and physical infrastructure. The benefits of such facilitation derive from equilibrium differences in shadow prices that prevail - e.g. due to communication and transport costs and limitations in the rule of law. This does not mean, however, that developing countries should imitate the modern institutions of high income economies, e.g. by spending vast sums on modern cadastral surveys and court proceedings in order to confer Western-style land titles before their benefits warrant their costs. Rather, appropriate rules regarding property and contracting should be allowed to evolve along with the increasing specialization and intensification of production.

The economics of rural organization with endogenous behaviour and organization is in its infancy. There is a promising body of theory featuring specialization as the central pillar of economic organization (e.g. Yang 2003) and a rich tapestry of rural institutions waiting to be described and explained. Much remains to be done.

References

Adams, D., Graham, D. & von Pischke, J. 1984. Undermining rural development with cheap credit. Boulder, United States of America, Westview Press.

Ahmed, I. 1981. Farm size and labour use: some alternative explanations. Oxford Bulletin of Economics and Statistics, 43: 73-88.

Ahn, C., Singh, L. & Squire, L. 1981. A model of an agricultural household in a multi-crop economy: the case of Korea. Review of Economics and Statistics, (63): 520-525.

Asian Development Bank. 2003. Sharing insights, lessons and experience in rural finance. In Paving the Way Forward for Rural Finance: An International Conference on Best Practices. 2-4 June. Washington, DC, United States of America.

Badar, H. 2002. Farm marketing and poverty reductions. (available at www.Pakissan.com)

Balisacan, A. & Roumasset, J. 1987. Public choice of economic policy: the growth of agricultural protection. Weltwirtschaftliches Archives.

Bauer, P. 1964. West African trade: a study of competition, oligopoly and monopoly in a changing economy. Cambridge, England, Cambridge University Press.

Bautista, R. 1991. Dynamics of rural development: analytical issues and policy perspectives. PIDS Working Paper No. 91-07. Manila, Philippine Institute for Development Studies.

Benjamin, D. 1992. Household composition, labor markets and labor demand: testing for separation in agricultural household models. Econometrica, 60: 287-322.

Berry, A. & Cline, W. 1979. Agrarian structure and productivity in developing countries. Baltimore, United States of America, Johns Hopkins University Press.

Besley, T. 1994. How do market failures justify interventions in credit markets? World Bank Research Observer, 9(1): 27-47. Washington, DC, World Bank.

Binswanger, H., Deininger, K. & Feder, G. 1995. Power, distortions, revolt and reform in agricultural land relations. In J. Behrman & T.N. Srinivasan, eds. Handbook of development economics Vol. II. Amsterdam,
New York and Oxford, Elsevier Science, North-Holland Publishing Co.

Binswanger, H. & Rozenweig, M.R. 1986. Behavioral and material determinants of production relations in agriculture. Journal of Development Studies, 22: 503-539.

Binswanger, H. & Ruttan, V. 1978. Induced innovation. Baltimore, United States of America, Johns Hopkins University Press.

Blinder, A. 1987. Hard heads, so. hearts: tough-minded economics for a just society. Massachusetts, United States of America, Addison-Wesley.

Bose, P. 1998. Formal-informal sector interaction in rural credit markets. Journal of Development Economics. 56: 256-80.

Boserup, E. 1965. The conditions of agricultural growth: the economics of agrarian change under population pressure. London, Allen & Unwin.

Coleman, B. 2002. Microfinance in Northeast Thailand: who benefits and how much? ERD Working Paper Series No. 9. Manila, Asian Development Bank.

De Janvry, A., Fafchamps, M. & Sadoulet, E. 1991. Peasant household behavior with missing markets: some paradoxes explained. Economic Journal, 101: 1400-17.

De Janvry, A., Gordillo, G., Platteau, J.P. & Sadoulet, E. 2001. Access to land, rural poverty and public action. Oxford, United Kingdom, Oxford University Press.

De Janvry, A. & Sadoulet, E. 2000. Poverty determinants in Latin America. Food Policy, 25: 389-409.

Deininger, K. 2003. Land policies for growth and poverty reduction. World Bank Policy Research Report. World Bank and Oxford University Press.

Demsetz, H. 1969. Information and efficiency: another viewpoint. Journal of Law and Economics, 12: 1-22.

Deweaver, M. & Roumasset, J. 2002. Risk aversion as effort incentive: A correction and prima facie test of the moral hazard theory of share tenancy. Economics Bulletin, 15(4): 1-16.

Dixit, A. 1996. The making of economic policy: a transaction-cost politics perspective. Massachusetts, United States of America, MIT Press.

Easterly, W. 2002. The elusive quest for growth. Massachusetts, United States of America, MIT Press.

Eastwood, R. & Lipton, M. 2000. Pro-poor growth and pro-growth poverty reduction: meaning, evidence and policy implications. Asian Development Review.

Eicher, C.K. &. Statz, J.M. 1984. Agricultural development in the third world. Baltimore, United States of America, Johns Hopkins University Press.

Eswaran, M. & Kotwal, A. 1986. Access to capital and agrarian production organization. Economic Journal, 96: 482-498.

Fairhurst, T. 2003. Global food security and the role of sustainable fertilization. IFA-FAO conference on commercial farms in Southeast Asia. 26-28 March. Rome, Italy. http://www.fertilizer.org/ifa/publicat/PDF/2003_rome_fairhurst.pdf

Faruqee, R. & Khandker, S. 2001. The impact of farm credit in Pakistan. Working Papers in Agriculture: Land, Commodity Prices, Markets No. 2653. Washington, DC, World Bank.

Feder, G., Ongchon, T., Chalamwong, Y. & Honglaradom, C. 1988. Land policies and farm productivity in Thailand. Baltimore, United States of America, Johns Hopkins University Press.

Frisvold, G.B. 1994. Does supervision matter? Some hypothesis tests using Indian farm-level data. Journal of Development Economics, 43(2): 217-38.

Greenwald, B. & Stiglitz, J. 1986. Externalities in economies with imperfect information. Quarterly Journal of Economics, 100: 229-64.

Hayami, Y. 1998. Norms and rationality in the evolution of economic systems: a view from Asian villages. The Japanese Economic Review, 49(1): 36-53.

Hayami, Y. 2000. An ecological and historical perspective on agricultural development in Southeast Asia. World Bank Policy Research Working Paper No. 2296. Washington, DC, World Bank.

Hayami, Y. 2003. Family farms and plantations in tropical development. FASID Discussion Paper No. 2003-001.

Hayami, Y. & Kawagoe, T. 1993. The agrarian origins of commerce and industry. New York, St. Martin’s Press.

Hayami, Y. & Kikuchi, M. 1982. Asian village at the crossroads: an economic approach to institutional change. Baltimore, United States of America, Johns Hopkins University Press.

Hayami, Y. & Otsuka, K. 1993. The economics of contract choice: an agrarian perspective. Oxford, Clarendon Press.

Hoff, K. & Stiglitz, J. 1998. Moneylenders and bankers: price-increasing subsidies in a monopolistically competitive market. Journal of Development Economics, 55: 485-518.

Jacoby, H. & Mansuri, G. 2002. Incomplete contracts and investment: a study of land tenancy in Pakistan. Washington, DC, World Bank. (available at www.williams.edu/Economics/neudc/papers/investment. pdf).

James, W. & Roumasset, J. 1984. Migration and the evolution of tenure contracts in newly settled regions. Journal of Development Economics, 14: 147-162.

Jensen, M. 2000. A theory of the firm: governance, residual claims and organizational forms. Massachusetts, United States of America, Harvard University Press.

Johnston, B. & Mellor, J. 1961. The role of agriculture in economic development. American Economic Review, 51: 566-93.

Jones, W.O. 1972. Marketing staple food crops in tropical Africa. Ithaca and London, Cornell University Press.

Jorgenson, D. 1961. The development of a dual economy. Economic Journal, 71(282): 309-34.

Kratz, F. & Roumasset, J. 2001. Trade diversion, excess storage and state trading enterprises: rice markets in the Philippines. University of Hawaii, Hawaii, United States of America.

Kydd, J. 2002. Agriculture and rural livelihoods: is globalization creating or blocking paths out of rural poverty? (available at http://www.odi.org.uk/agren/papers/agrenpaper_121.pdf).

Laffont, J.J. & Tirole, J. 1999. Competition in telecommunications. Massachusetts, United States of America, MIT Press.

Lau, L., Yotopoulos, P., Chou, E. & Lin, W. 1981. The microeconomics of distribution: a simulation of the farm economy. Journal of Policy Modeling, 3: 175-206.

Lele, U. 1971. Food grain marketing in India. New York, United States of America, Cornell University Press.

Mears, L.A. 1981. The new rice economy of Indonesia. Yogyakarta, Indonesia, Gadjah Mada University Press.

Mellor, J. 1966. The economics of agricultural development. New York, United States of America, Cornell University Press.

Mellor, J., ed. 1995. Agriculture on the Road to Industrialization. Baltimore, United States of America, Johns Hopkins University Press.

Mellor, J. & Johnston, B. 1984. The world food equation: interrelationships among development, employment and food consumption. Journal of Economic Literature, 22: 531-74.

Meyer, R. & Nagarajan, G. 2000. Rural financial markets in Asia: policies, paradigms and performance. New York, Oxford University Press.

Morduch, J. 2000. The microfinance schism. World Development, 28(4): 617-629.

Mosher, A. 1966. Getting agriculture moving. New York, Frederick A. Praeger.

Murrel, P. 1983. The economics of sharing: a transaction cost analysis of contractual choice in farming. Bell Journal of Economics, 14: 283-293.

Natsios, A. 2003. Notes from Natsios on foreign aid in the national interest: promoting freedom, security and opportunity. (available at http://www.usaid.gov).

North, D. 1987. Institutions, transaction costs and economic growth. Economic Inquiry, 25(3): 419-28.

North, D. & Wallis, J. 1982. American government expenditures: a historical perspective. American Economic Review, 72(2): 336-40.

Nozick, R. 1975. Anarchy, state and Utopia. New York, Basic Books.

Otsuka, K. 2002. Land markets. Bretton Woods Conference.

Ray, D. 1998. Development economics. New Jersey, United States of America, Princeton University Press.

Reardon, T., Timmer, C.P., Barrett, C. & Berdegu’e, J. 2003. The rise of supermarkets in Africa, Asia and Latin America. American Journal of Agricultural Economics, 85(5): 1140-46.

Reid, J.D. 1976. Sharecropping and agricultural uncertainty. Economic Development and Cultural Change, 24: 549-576.

Roumasset, J. 1979. Risk aversion, agricultural development and the indirect utility function. In J. Roumasset, J.M. Boussard & I.J. Singh, eds. Risk, uncertainty and agricultural development. Philippines, SEARCA/ADC.

Roumasset, J. 1976. Rice and risk: decision-making among low-income farmers in theory and practice. Amsterdam, North-Holland Publishing Co.

Roumasset, J. 1978. The new institutional economics and agricultural organization. Philippine Economic Journal, 37: 331-348.

Roumasset, J. 1983. Village at the crossroads - a review. Population and Development Review, 9: 169-171.

Roumasset, J. 1986. The welfare economics of rural credit and the benefits of land titles. Washington, DC, World Bank. (manuscript)

Roumasset, J. 1987. The public economics of irrigation management and cost recovery. Washington, DC, World Bank.(manuscript)

Roumasset, J. 1989. Decentralization and local public goods: getting the incentives right. Philippine Review of Business and Economics, 26(1): 1-13.

Roumasset, J. 1992a. Investing in agriculturally-led growth: the Philippine case. South East Asian Journal of Agricultural Economics, 1(1).

Roumasset, J. 1992b. The role of government in economic cooperation. In J. Roumasset & S. Barr, eds. The economics of cooperation: East Asian Development and the case for pro-market intervention. Boulder, United States of America, Westview Press.

Roumasset, J. 1995. The nature of the agricultural firm. Journal of Economic Behavior and Organization, 26: 171-177.

Roumasset, J. 2000. Black-hole security. Department of Economics Working Paper No. 00-5. University of Hawaii. (available at http://www.economics.hawaii.edu).

Roumasset, J. 2002. The microeconomics of agricultural development in the Philippines. Department of Economics Working Paper No. 02-10. Hawaii, United States of America, University of Hawaii. (available at http://www.economics.hawaii.edu).

Roumasset, J. 2003a. The role of agriculture in pro-poor economic development: a Philippine perspective. (mimeo)

Roumasset, J. 2003b. Agricultural parastatals and pro-poor economic growth. International Workshop on Agribusiness: From Parastatals to Private Trade - Why, When and How? 15-6 December. New Delhi, India.

Roumasset, J. & Barr S., eds. 1992. The Economics of Cooperation: East Asian Development and the Case for Pro-Market Intervention. Boulder, United States of America, Westview Press.

Roumasset, J. & Evenson, R. 1986. Markets, institutions and family size in rural Philippine households. Journal of Philippine Development, 23: 141-162.

Roumasset, J. & James, W. 1979. Explaining variations in share contracts: land quality, population pressure and technological change. Australian Journal of Agricultural Economics, 23(2): 116-27.

Roumasset, J. & Van Assche, A. 2003. Malthus to Solow: a new classical approach to induced innovation. SABE Behavior Economics Conference Paper. Lake Tahoe, NV, United States of America.

Roumasset, J., Setboonsarng, S., Wickramasinghe, U., Estudillo, J. & Evenson, R. 1995. Institutional reform and the informal sector. Working paper. University of Maryland, Maryland, United States of America.

Roumasset, J. & Smith, J. 1981. Population, technological change and the evolution of labor markets. Population and Development Review, 7(3): 401-19.

Roumasset, J. & Uy, M. 1980. Piece rates, time rates and teams: explaining patterns in the employment relation. Journal of Economic Behavior and Organization, 1(4): 343-360.

Roumasset, J. & Uy, M. 1987. Agency costs and the agricultural firm. Land Economics, 63(3): 290-302.

Ruttan, V. 1978. Induced institutional change. In J. Binswanger & V. Ruttan. Induced Innovation. Baltimore, United States of America, Johns Hopkins University Press.

Ruttan, V. 2003. Social science knowledge and economic development - an institutional design perspective. Chicago, United States of America, University of Michigan Press.

Sadoulet, E. and de Janvry, A. 1995. Quantitative development policy analysis. Baltimore, United States of America, Johns Hopkins University Press.

Sadoulet, E., de Janvry, A. & Fukui, S. 1997. The meaning of kinship in sharecropping contracts. American Journal of Agricultural Economics. 79(2): 394-406.

Sah, R.K. 1986. Size, supervision and patterns of labor transactions. Journal of Philippine Development, 13(1, 2). (also available at http://dirp4.pids.gov.ph/ris/pjd/pidsjpd86-labortrans.pdf).

Setboonsarng, S. 1991. The evolution of rural labor institutions in a rice-based economy: a simultaneous equation model with selectivity bias. University of Hawaii, Hawaii, United States of America. (Ph.D. dissertation)

Siamwalla, A., Pinthong, C., Paopongsakorn, N., Satsanguan, P., Nettayarak, P., Mingmaneenakin, W. & Tubpun, Y. 1990. The Thai rural credit system: public subsidies, private information and segmented markets. World Bank Economic Review, 4(3): 271-95.

Shaban, R. 1987. Testing between competing models of share tenancy. Journal of Political Economy, 95(5): 893-920.

Stiglitz, J. 1974. Incentives and risk sharing in sharecropping. Review of Economic Studies, 41(2): 219-56.

Stiglitz, J. 1993. Whither socialism? Massachusetts, United States of America, MIT Press.

Stiglitz, J. 2002. Globalization and it’s discontents. New York, Norton.

Stiglitz, J. & Weiss, A. 1981. Credit rationing in markets with imperfect information. American Economic Review, 71(3): 393-410.

Stiglitz, J. & Uy, M. 1996. Financial markets, public policy and the East Asian miracle. World Bank Research Observer, 11(2): 249-276.

Timmer, C.P. 1987. The corn economy of Indonesia. New York, Cornell University Press.

Unnevehr, L.S. 1984. Marketing and price formation. In F. Falcon, W. Jones, S. Pearson, et al., eds. The Cassava Economy of Java. pp. 136-62. California, Stanford University Press.

Weber, C., Franklin, D., Diaz, E., Bathnik, D., Zuvekas, C. & Southgate D. 2002. Rural prosperity in the Latin American and Caribbean Region. USAID/LAC White Paper. Washington, DC.

Williamson, O. 1985. The economic institutions of capitalism: firms, markets and relational contracting. New York, Free Press.

Wright, B. & Williams, J. 1990. The behavior of markets for storable commodities. Presented at the Annual Conference of the Australian Agricultural Economic Society. Brisbane, Australia.

Yang, X. 2002. Development Economics: Inframarginal vs. Marginal Analysis. Massachusetts, United States of America, Blackwell.

Yang, X. & Ng, Y.K. 1993. Specialization and economic organization: a new classical microeconomic framework. Amsterdam, North-Holland.