Access to land through rental markets: a (counter-) evolution in the World Bank’s land policy?

U. Pica Ciamarra
Dipartimento di Economia Pubblica, Università degli Studi di Roma "La Sapienza"
and FAO Consultant, Pro-Poor Livestock Policy Initiative (PPLPI), Rome, Italy

The 2003 World Bank policy research report Land policies for growth and poverty reduction reviews theoretical analyses and practical experiences of the economics and the political economy of land access, and presents some general guidelines for formulating policy advice. It stresses that secure property rights to land can greatly increase the welfare of the poor and that access to land can be provided through a heterogeneous variety of policies, both state-led and market-driven. In particular, it argues that in developing countries, long-term secure tenancy contracts can provide ample opportunities for securing efficient resources allocation and benefiting the poor, and that governments should remove all unjustified restrictions in the operation of land rental markets. In line with the World Bank report, this article argues that the available evidence suggests, on the one hand, that long-term tenancy contracts are not necessarily essential for first-best investment incentives and, on the other hand, that rental markets are hardly ever pro-poor. Policy-makers should be cautious about supporting liberalization of land rental markets and the argument that long-term tenure is good for the poor; in the resulting liberalized land lease markets, equity and efficiency could be a costly trade-off.

INTRODUCTION

In the last decade land reallocation policies have returned in full force to the policy agendas of the international community and individual governments, and the World Bank has been in the forefront of reviewing, updating and interpreting the evolution of practical experiences and economic theory of land policies in developing and transition countries. The 2003 World Bank Policy Research Report Land policies for growth and poverty reduction can thus be regarded as a milestone in the World Bank’s land policy, and it is likely to have some policy influence in the near future: "this report aims to summarize available insights on land policies in a form that is relatively accessible; to present general recommendations; and to illustrate how these could be translated into specific, real world situations"^[1] (World Bank, 2003, p. 1). Furthermore, the report builds on the outputs of four regional workshops in Africa, Asia, Eastern Europe and Latin America and an electronic discussion that, as emphasized in the foreword, "provide a strong basis for using the report as an input into the development of strategies and activities at the country level"^[2] (World Bank, 2003, p. xi).

The report revisits the theoretical and empirical economic literature on land property rights and land reallocation policies (e.g. Bardhan, 1989; Deininger, 2001; Hayami and Otsuka, 1993) and provides a comprehensive review of the role of markets (chapter three) and the state (chapter four) in increasing access to land for rural people, enhancing productive capacity and reducing poverty levels. The report critically analyses land rental markets, which have often been disregarded by academic scholars and policy-makers alike, despite the fact that they can provide ample opportunities for securing an efficient allocation of scarce resources, promoting economic growth, and benefiting the poor (Deininger et al., 2003). From a political economy viewpoint, the report points to two preconditions for the effectiveness of land lease markets: secure property rights and secure and long-term tenancy contracts. The ensuing policy guidelines are that pro-poor governments should establish and enforce secure land property rights, and remove all unjustified restrictions on the functioning of land rental markets.

This article provides a non-technical overview of the conceptual and empirical premises of land rental markets and argues that, though the policy directives of the World Bank report appear reliable and persuasive, the available evidence suggests that: (i) long-term and secure tenancy contracts are not necessarily essential for efficient resource allocation; and (ii) although land rental markets are efficiency-enhancing, whether they are also equity-enhancing and pro-poor remains controversial. The ultimate aim is not to argue that the World Bank report is wrong - it is not - but rather to draw attention to a number of supposedly clear-cut statements in the report that are in fact contentious.

THE WORLD BANK REPORT IN A HISTORICAL PERSPECTIVE

State-led land redistributive policies were high on the political agenda in the 1950s (in Asia and the Near East) and 1960s (in Latin America), especially in countries with high land property concentration, great social and economic inequality, abject rural poverty and widespread landlessness. Land reforms were largely political and aimed at mitigating rural riots and protests so as to perpetuate the ruling government. It is hardly surprising, therefore, that the economic performance of reforms has been unsatisfactory. De Janvry (1981) maintains that, in the 1960s, Latin American land reforms became instruments for promoting technological change in the non-reformed sector, rather than improving the livelihoods of rural dwellers. Otsuka (1993) contends that, in Asia, land redistribution did not have any significant impact on rural poverty, and El Ghonemy (2002) asserts that African land reforms destroyed functional systems of collective land property rights in semi-arid regions.

Following the disappointing experiences of the 1960s, land reforms largely disappeared from the political agenda. In the 1970s, agricultural development was primarily associated with the technology of the Green Revolution, with a central role given to international institutions and the state. In the 1980s, in the attempt to "get the prices right", the role of governments was reduced and the economies liberalized under stabilization and structural adjustment policies. De Janvry, Murgay and Sadoulet (2001) define this period as one of retrogression in rural development, but at the same time one of setting the context for new approaches to rural development in terms of market incentives and new roles for civil society. Since the 1990s, in fact, the rural development agenda has been characterized by the market/state/civil society trilogy, and the issue of access to land has returned in full force to political debates (Borras, 2003; Deininger, 1999).

In particular, the World Bank suggests that, under some circumstances, government-directed one-time land redistribution could be substituted by non-coercive mechanisms of adjustment in access to land. Reference is made to market-assisted, market-driven, market-friendly or negotiated land reforms to indicate strategies that are intended to facilitate land market transfers. First, macroeconomic policies are to shift outward the supply of land by removing distortions that encourage people to accumulate land for reasons other than farming. Second, the demand for land is activated through allowing potential beneficiaries, assisted by the community and local government, to receive a combination of grants and loans from the public and private sectors, which they use to negotiate the purchase of land from willing sellers and to set up viable farms.

Pilot projects on market-assisted land reform have been implemented in Brazil, Colombia and South Africa. The evidence is very mixed so far, and common opinion is that results have fallen short of expectations (see Borras, 2003; Lyne, Zille and Graham, 2000; Sauer, 2001; Sender and Johnston, 2004). The World Bank has also not yet reached clear-cut conclusions about the effectiveness of market-assisted programmes of land reallocation. First, the Bank has commissioned a manual for monitoring and evaluating market-assisted land reform that aims to describe: "the methods, measures, and data needs to investigate land reform projects’ effect on income, agricultural productivity, and market activation; as well as the aggregate costs and benefits of the project". Second, the 2003 report does not comprehensively analyse market-assisted land reform, maintaining that it is simply one among several instruments to promote efficient resource allocations in rural areas. Finally, the report highlights how rental markets, which have traditionally been underemphasized compared with land sales markets, can also be an effective mechanism for land reallocation in developing countries:

Traditionally, much of the discussion on land markets has focused on the permanent transfer of ownership through sales. However, similar benefits can accrue from often informal transactions in land rental markets that are widespread across the world and that are less likely to be affected by, or can adjust more easily to, the market imperfections that are pervasive in rural areas of developing countries.

(World Bank, 2003, p. 79)

Rental markets (a) allow flexibility in adjusting the land area used with low transaction costs; (b) require only a limited capital outlay, thereby leaving some liquidity available for productive investments rather than locking it all up in land; (c) facilitate easy reallocation of land toward more efficient users than the current owners [...]; (d) provide a stepping stone toward land ownership by the landless; and (e) help overcome, through sharecropping contracts, market failures in labor, insurance, credit, management, and supervision, thereby potentially helping secure the competitiveness of participants.

(World Bank, 2003, p. 85)

The overall conclusions are that: (i) tenure security is a precondition for effective rental markets that benefit the poor; (ii) long-term secure tenancy contracts could enhance the efficient allocation of scarce resources and also secure access to land for the rural poor; (iii) governments should eliminate all unjustified restrictions to the operation of land markets (see Box 1 for details of interpreting the report).

These conclusions are centred on neoclassical economic theory, which has reached clear-cut results in the case of the one-period or static contract, i.e. when the contract is signed at the beginning of a crop season and payment is made at the end. While one-period pure-wage contracts provide no incentive to farmers to put in effort (their compensation schedule is independent of outcome), fixed-rent tenancy contracts entail the first-best allocation of resources as farmers are the ultimate beneficiaries of their effort (the rent is fixed and independent of production). When agents are risk-averse, an appropriate solution could be a sharecropping contract, which also enables the landlord and the tenant to pool non-contractible inputs and resources so as to bypass factor market imperfections (e.g. inadequate access to credit) (see Bardhan, 1989; Hayami and Otsuka, 1993; Salanié, 1997).

It is significant that the report maintains that tenancy contracts (either fixed-rent or sharecropping) are efficiency-enhancing and that governments should eliminate all unjustified restrictions on the operation of land markets. But the report also argues that tenancy contracts are only efficient when they are long-term, and that rental markets are likely to be pro-poor, i.e. they are also equity-enhancing. However, it does not thoroughly elaborate on the optimal lengths of tenancy contracts; surprisingly, very few studies have theoretically analysed and collected data on the length of tenancy contracts in developing countries (see Bandiera, 2003; Banerjee and Ghatak, 2004; Jacoby and Mansuri, 2002; Roy and Serfes, 2000). On the other hand, whether or not rental markets are pro-poor is essentially an empirical question, the answer to which depends on local idiosyncratic conditions.

SHORT-TERM VERSUS LONG-TERM TENANCY CONTRACTS

Static contract theory is primarily concerned with optimal contract design and its implications for the immediate outcome, and neglects the relationships between the parameters of the contract and the incentives they provide for future effort or existing long-term investments. It follows that it can be appropriate for formulating policy as long as one assumes that those who deviate from the agreement can be penalized adequately, so that a one-period contract can be replicated indefinitely. The penalty will be endogenous if it originates in the non-replication of the contract; it will be exogenous if it is imposed by a third supervising party (the court). In general, however, the existence of substitute jobs or contract opportunities is likely to limit the amount of the endogenous penalty. Furthermore, even if courts exist, litigation is often prohibitively costly and the bargaining power is tilted in favour of large landowners. It follows that time and intertemporal choice give the parties an additional degree of freedom when contracting and dealing with market imperfections. As a consequence, the outcome of a one-shot game, or one-period contract, and an infinitely repeated game, or multi-period contract, might be different. The conventional and intuitive view is that, when compared with one-period agreements, long-term and secure tenancy contracts provide the tenant with incentives for maintenance and long-term production investments.

Investment incentives obviously depend on the length of the contract. Intuitively, if tenancy contracts are short-term, long-term non-contractible (non-observable) land-specific investments, such as effort for plant maintenance, will typically be underprovided by the tenant. A one-period tenant will invest only up to the point where the marginal return in the current period (i.e. the period of the contract) equals the shadow price, and this inefficiency will be common to both share and fixed-rent tenants. Worse still, in the case of one-period contracts the tenant could maximize his/her utility by disregarding soil depletion and other damage to the land that will negatively affect its future productivity (Hayami and Otsuka, 1993; Roy and Serfes, 2000).

Long-term tenancy contracts could be a solution to the under-provision of effort. In particular, they are a valuable incentive mechanism when non-observable effort has a long-term investment-like component, such as tree maintenance and careful application of chemicals, that ensures the agent a stake in future output. Longer duration of contracts should therefore be associated with greater non-contractible investment, either "investment effort" (Banerjee and Ghatak, 2004) or "maintenance effort" (Bandiera, 2003). Furthermore, in a multiperiod setting, future output includes information about the agent’s current non-contractible effort level, since the effect of random factors on production is filtered out over the longer term, and thus a penalty can be imposed if the expected and actual outputs differ significantly.

Compared with short-term contracts, long-term contracts carry two further advantages. First, they involve lower transaction costs because they have to be agreed upon less frequently. It follows that in industrialized countries and more developed areas of developing countries, where the opportunity cost of economic agents is high, long-term contracts are likely to be more pervasive. Second, long-term contracts can be used to smooth consumption and lessen the overall risk shouldered by the agent when credit/ insurance markets are imperfect or incomplete (see Bandiera, 2003).

However, long-term tenancy agreements also entail costs, both for landowners and tenants. Landowners forsake the use of eviction threats, which could otherwise be used to elicit both present and future effort. On the one hand, a tenant operating under a short-term contract and subject to an eviction threat could exercise the first-best effort today in order to increase output and his/her reputation as someone with whom to contract in the future; on the other hand, eviction threats can also provide incentives for long-term investment, as the tenant increases output in the next period and hence lowers the probability of eviction^[3] (Bardhan, 1984). Banerjee and Ghatak (2004) have recently modelled a dynamic game with non-contractible investments to examine the role of eviction threats as a possible contractual instrument. They conclude as follows: "In this paper we have analyzed a very simple model to demonstrate the possibility that eviction threats can increase unobservable investment effort. This suggests the conventional (and eminently intuitive) argument that security of tenure is good for investment should be applied with caution" (p. 16).

Griffin, Khan and Ickowitz (2002) elaborate further on this issue and argue first that both tenants’ and landowners’ investment incentives will be unaltered if contracts are short-term. In particular, provided that tenants can take their capital with them when their contract is terminated, their incentive to invest will be unimpaired; and as long as short-term contracts allow landowners to increase rents when investment results in an increase in total returns, their incentive to invest will be unaffected. An implication is that tenants will invest in movable assets (e.g. livestock) while landowners will invest in land improvement and fixed assets (e.g. fences).

Second, long-term tenancy agreements prevent landowners not only from cultivating the land directly for the duration of the contract, but also from adjusting the terms of the contract to suit changes in the exogenous environment. For instance, short-term tenure enables landowners to adjust the parameters of the contract in order to harvest the full benefits of a new technology or to make profits on new market and institutional opportunities. A bleak implication is that if landowners are unable to modify the terms of a contract because tenancy agreements are long-term, they will have fewer incentives to invest in their land; this will have negative consequences for agricultural growth (Griffin, Khan and Ickowitz, 2002).

Finally, long-term tenancy contracts may also be costly for tenants, for example when the investment is non-contractible and the landlord is unable to commit fully to retaining the tenant for the duration of the contract at a given price. This is a common situation in rural areas of developing countries where the bargaining power is typically tilted in favour of large landowners. The tenants, aware of the impending opportunism on the part of the landlord, fear that they will not recoup all the fruits of their effort and may undersupply land-specific investment, even if the contract provides strong incentives (i.e. a long-term fixed-rent contract). In an empirical investigation in rural Pakistan, Roy and Serfes (2000) found that a landlord’s specific unobservable parameter (e.g. reputation), which is uncorrelated with assets, accounted for at least half of the variation in tenancy durations in their sample. Sharma and Drèze (1996) report that in Palanpur, Uttar Pradesh, the quality most frequently sought in their partner by both landlords and tenants is "honesty".

Undeniably, when extending the single-period agrarian contract model beyond one cropping season, one can cast doubts on the position expressed in the World Bank report that "Short-term contracts will provide only limited incentives for users to undertake land-related investment" (p. xxxvi); and that "Unless secure long-term contracts are available, the incentive for either tenants [...] or for landlords [...] to make investments in land may be severely limited" (p. 116). In effect, the report does not elaborate sufficiently on long-term tenancy contracts. However, the existing economic literature suggests that long-term tenancy contracts are not necessarily superior to short-term ones; moreover, even though available data on contract length are very mixed, it appears that the majority of agrarian contracts are for one year or one cropping season (Bardhan, 1984; Lastarria-Cornhiel and Melmed-Sanjak, 1999; Newbery, 1977). In fact, why should landlords offer long-term tenure if short-term tenure is not necessarily detrimental to investment incentives?

THE SCANT INCIDENCE AND EXCLUSIONARY ASPECT OF LAND RENTAL MARKETS

Economic literature has thoroughly analysed the circumstances under which state-led land reform and land sales markets can promote land reallocation favouring the poor (e.g. Deininger, 1999). The World Bank report also considers land rental markets as an option for land redistribution: "land sales markets are likely to be much less active than land rental markets virtually everywhere in the world because of higher transaction costs, difficulties in accessing long-term capital to finance land purchases, and insecurity about future economic development that would significantly affect land prices" (p. 120). It is uncontroversial that imperfections in rural markets may prevent land sales from leading to rapid equalization of land holding patterns. According to the report, however, land rental markets suffer less from these disadvantages and could thus contribute to increasing equity and efficiency. Indeed, the extent to which rental markets are able to equalize the distribution of operational holdings and be pro-poor is very much an empirical question.

Table 1 presents statistics on the distribution of farms and farmland by land tenure status for major regions of the world. Even though these are subject to several limitations and conceal deep structural differences between regions and countries of the world, they are the only data set available for macrorepresentation.

First, Africa and Asia present almost an equal distribution of agricultural area, with a concentration index of the agricultural area of 0.53 and 0.56 respectively. The European concentration index is in line with those of Asia and Africa, while the distribution of agricultural land is more skewed in North and South America, with concentrations of 0.75 and 0.82 respectively.

TABLE 1
Distribution of farms and farmland by land tenure status and by major regions, as reported in the 1990 World Census of Agriculture

Source: elaborated from FAO (1997).

n.a. = not available.

Reporting countries given below; data for the distribution of farms and farmland reported only from those countries marked with an asterisk.

Africa - Burkina Faso, Democratic Republic of the Congo, Djibouti, Egypt, Ethiopia, Guinea, Guinea-Bissau, Lesotho, Libyan Arab

Jamahiriya, Malawi, Namibia, Réunion, Uganda.

Asia - Cyprus, India,* Indonesia, Iran (Islamic Republic of), Israel, Japan, Myanmar, Nepal,* Pakistan,* Philippines,* Republic of Korea, Thailand,* Turkey,* Viet Nam.

Europe - Austria,* Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy,* Luxembourg, Netherlands, Norway, Portugal, Slovenia, Spain,* Switzerland, United Kingdom.

North America - Canada, United States,* British Virgin Islands.*

South America - Argentina, Brazil, Colombia,* French Guiana, Paraguay,* Peru.

The census coverage is not complete for all countries, as nomadic livestock, modern plantations and holdings below a certain size were sometimes excluded. Percentages, therefore, may not always add up to 100.

Second, the table suggests that owner cultivation is the most pervasive form of tenure in all regions, excluding Africa for which data are not available, both in terms of farms and farmland, with percentages over 60 percent worldwide, and higher than 90 percent in Asia. Rental markets appear to be developed in Europe and North America. Multiple tenure forms, in fact, mainly refer to ownership-cum-tenancy arrangements, where large farms use rental markets to increase the size of their operated areas and realize the full potential of technical indivisibilities.

Swinnen (2002), quoted in the World Bank report, states that 71 percent of farmland is rented in Belgium, 48 percent in the Netherlands and 47 percent in France. Sadoulet, Murgai and de Janvry (1998) report that the percentage of farms with rented-in land is 75 percent in the United States of America and Belgium, 47 percent in Spain, and 35 percent in Italy. In contrast, rental contracts in Africa, Asia and Latin America are de facto nonexistent, with percentages lower than 5 percent both in terms of number of farms and farmland.^[4] Individual country studies often contrast with this evidence. In Asia estimates are that 15 percent of rural households rent in land in India, 14 percent in Thailand and 40 percent in Bangladesh (Lastarria-Cornhiel and Melmed-Sanjak, 1999). Estimations for Brazil and Paraguay indicate that 17.7 and 9.2 percent of holdings are rented in respectively (de Janvry and Sadoulet, 2002).

This suggests that land rental markets are particularly thin in developing countries, supporting the hypothesis of the World Bank report that there is room for expanding tenancy contracts. However, the table neither explains why rental markets are atrophied, nor does it indicate whether effective rental markets are equity-enhancing. The question remains: although rental markets are thin, are they able to shift land from large owners to the landless and small producers?

To our knowledge, Lastarria-Cornhiel and Melmed-Sanjak (1999) present one of the most comprehensive reviews of land tenancy contracts in the developing world. Even though the authors do not specifically focus on the characteristics of the tenants (i.e. whether or not they are poor), their analysis suggests the existence of a variety of different situations that point to heterogeneous patterns of land lease transactions, either pro-poor or not. Tenants could be small owner-cultivators who adjust their holdings through contracting, landless poor labourers who obtain access to a plot of land, or large landowners or agribusiness firms that lease lands from small owners (reverse tenancy). In general, however, the dominant pattern is of small-to-small lease transactions, and leasing contracts tend to equalize land operational size where the opportunity cost of agricultural activities is growing and migrants rent out their land.

A review of some of the few country studies available provides deeper insights into the characteristics of tenants. Sharma and Drèze (1996) analysed the nature of sharecropping contracts in an Indian village in Uttar Pradesh; Deininger, Zegarra and Lavadenz (2003) examined the performance of land markets in Nicaragua following a structural adjustment programme; Sadoulet, Murgai and de Janvry (1998) investigated the functioning of land rental markets both in a Tunisian village and in four Pakistani rural districts; and De Silva (2000) studied tenancy contracts among rice farmers in Sri Lanka.

In India, very few landless households lease in land, as the ownership of land - even a very small plot - appears to be an indispensable precondition for leasing land. In contrast, in Pakistan and Tunisia large owners rent out land while small owners and the landless rent in land, denoting a process of equalization of the operational size, especially when the opportunity cost of farming is high (see also Kung, 2002, for China). In Nicaragua, even though a limited amount of land is actually exchanged on the rental market, 80 percent of those renting in land in the sample were completely landless.

Equalization of land operational size through rental markets, however, does not necessarily benefit the poor. In Uttar Pradesh, the wealthy landowners are those who usually lease in land, so as to realize the potential of indivisible productive assets, such as threshers and pumping sets. In Tunisia, there is a positive correlation between the endowment positions of the landlords and the tenants, which is particularly evident for the more intensive crops. In the richest districts of Pakistan landlords look for tenants who can match their capital contributions, and access to land for the rural poor is difficult to obtain. In Nicaragua, if both livestock and coffee producers lease in and lease out lands, there is a higher propensity for renting land out among producers of traditional and less-profitable staples, pointing to some exclusionary aspects of land rental markets. Ruben and Masset (2003) even show that there are powerful pressures on poor Nicaraguan farmers towards distress sales, and argue that the country is going through a process of peasant differentiation. In Sri Lanka, only landless and land-scarce households that are endowed with resources similar to the landlords’ are able to lease in land as fixed-rent tenants. To sum up, it appears that either there is a general complementarity between the landlords’ and tenants’ asset endowments, or the wealthy landowners are more likely to lease in land, as in Uttar Pradesh and Nicaragua. In other words, land rental markets often bypass the poor.

Finally, an interesting lesson can be drawn from West Bengal. The widely known 1977 West Bengal Land Reform Act specified strict conditions under which landlords could evict tenants and launched a registration campaign (Operation Barga) in which over 1.4 million sharecroppers signed on with the public authority, earning tenure security and a series of rights in the land and credit markets (Ramachandran and Ramakumar, 2000). This reform has had positive effects on agricultural productivity through an increased supply of non-contractible inputs (Banerjee, Gertler and Ghatak, 2002). However, the 1992 National Sample Survey found that only 30.6 percent of tenants were on record and that, in the case of the landless, the percentage was as low as 16 percent, denoting remnants of feudal agrarian relations (fear of being evicted) (Shankar, 1999). Thus, the benefits of Operation Barga appear to have gone to the relatively better-off tenants; in fact, the majority of efficient farms, irrespective of tenure status, belong to the medium-size farm class (Chattopadhyay and Sengupta, 2001).

The above review leads to the following conclusions. Rental markets in developing countries appear to be rather inactive and frequently exclusionary and segmented. In the case of incomplete contracting, wealthy individuals are more likely to be offered either fixed-rent or sharecropping contracts, while small producers and the poor landless are bypassed by any willing-landlord willing-tenant exchange, although exceptions exist. Therefore, removing all unjustified restrictions on the operation of land rental markets, as suggested by the World Bank report, does not necessarily benefit the poor. As is emphasized in the literature on poverty traps, low physical and human capital may leave landless rural dwellers no alternative but to emigrate or to look for low-paid agricultural jobs, which keep them at the bottom of the earning ladder (Carter and Olinto, 2003). It follows that those efficiency-enhancing policies that remove restrictions on rental markets are not automatically equity-enhancing, and may benefit the poor only through (long-term) trickle-down effects.

CONCLUSIONS

The World Bank report considers land rental markets as a potential efficiency and equity-enhancing mechanism for reallocating land among poor rural dwellers. It argues that: (i) tenure security is a precondition for rental markets to work; (ii) long-term tenancy contracts could enhance the efficient allocation of scarce resources and secure access to land also for the rural poor; and (iii) governments should eliminate all unjustified restrictions to the operations of land markets.

This article has shown that the available evidence suggests that tenure security and long-term land tenancy contracts are not necessarily the first-best arrangements. In particular, through briefly comparing static and dynamic contract theory, it has been found that long-term tenancy contracts present both advantages and disadvantages when compared with short-term contracts and that, in several circumstances, landlords may wish to offer tenants short-term rather than long-term contracts.

Moreover, this article has questioned the assumption that rental markets are both efficiency- and equity-enhancing. A review of the empirical evidence, namely aggregate data for world regions and the results of individual country studies, indicates that rental markets are particularly thin in developing countries and in several cases tend to be exclusionary, in that the poorest households are likely to be bypassed in a segmented land rental market. The case, therefore, is not about tenancy per se, but about the quality of tenancy and the role of public actions in enabling rental markets to function for the poor. What are these public actions? Are these policies less demanding than market-driven and state-led programmes of land reallocation?

Much research is needed in at least four areas. First, economic literature on the lengths of tenancy contracts is still very scarce and consensus has not been reached in several respects. Second, the political economy of rental markets is a new area in economic literature and raises the question of how to activate land rental markets. Third, even though imperfections in the factors market do exist and may justify land tenancy agreements, it is widely acknowledged that small farmers have a far better record than large farmers in credit repayments. From this perspective, large farmers acting as credit suppliers to tenants through interlinked contracts could well constrain the development of rural credit markets and encourage the persistence of a non-egalitarian agrarian economy. Fourth, an equal distribution of land property rights may encourage some public actions that would not be supported by an equalization of land access through rental markets (e.g. a school site); in other words, some distribution of land property rights (rather than land-use rights) can enhance the efficiency of public resources allocation, while others can preclude some efficiency-enhancing collective actions. It follows that options for the full transfer of property rights are not to be completely disregarded: in a polarized agrarian economy, getting prices right, institutions right and promoting macrostability may not be enough to include the poor in the process of economic development (Carter and Barham, 1996). Policy-makers, therefore, should be highly cautious when applying the argument that long-term tenure security is good for the poor and supporting the liberalization of land rental markets.

ACKNOWLEDGEMENTS

The author would like to thank Marcello Gorgoni, Piero Conforti and Ludovica Mancini for their insightful comments on an earlier draft paper. The opinions expressed in this paper are the author’s and do not necessarily reflect views and policies of FAO and FAO/PPLPI.

REFERENCES

Bandiera, O. 2003. Contract duration and investment incentives: evidence from land tenancy agreements. London, London School of Economics and Political Science. (mimeo)

Banerjee, A.V. & Ghatak, M. 2004. Eviction threats and investment incentives. J. Dev. Econ., 74(2): 469-488.

Banerjee, A.V., Gertler, P.J. & Ghatak, M. 2002. Tenancy reform in West Bengal. J. Polit. Economy, 110(2): 239-280.

Bardhan, P.K. 1984. Land, labor and rural poverty: essays in development economics. New York, Columbia University Press and New Delhi, Oxford University Press.

Bardhan, P.K., ed. 1989. The economic theory of agrarian institutions. Oxford, UK, Clarendon Press.

Binswanger, H.P., Deininger, K. & Feder, G. 1995. Power, distortions, revolt and reform in agricultural land relations. In J.R. Behrman & T.N. Srinivasan, eds. Handbook of development economics, Vol. III, pp. 2659-2772. Amsterdam, Elsevier.

Borras, S.M. Jr. 2003. Questioning market-led agrarian reform: experiences from Brazil, Colombia and South Africa. J. Agrarian Change, 3(3): 367-394.

Carter, M.R. & Olinto, P. 2003. Getting institutions "right" for whom? Credit constraints and the impact of property rights on the quantity and composition of investment. Am. J. Agric. Econ., 85(1): 173-186.

Carter, M.R. & Barham, B.L. 1996. Level playing fields and laissez faire: post-liberal development strategy in inegalitarian agrarian economies. World Devel., 24(7): 1133-1149.

Chattopadhyay, M. & Sengupta, A. 2001. Tenancy inefficiency: a study based on West Bengal agriculture. Economic and Political Weekly, 36(5-6).

Deininger, K. 1999. Making negotiated land reform work: initial experience from Colombia, Brazil and South Africa. World Devel., 27(4): 651-672.

Deininger, K. 2001. Land markets and land reform. Paper presented at the International Conference on Access to Land: Innovative Agrarian Reforms for Sustainability and Poverty Reduction, 19-23 March 2001. Bonn, Germany (available at www.dse.de/zel/access-to-land/dateien/Deininger.pdf).

Deininger, K. & Binswanger, H.P. 1999. The evolution of the World Bank’s land policy: principles, experience, and future challenges. World Bank Research Observer, 14(2): 247-276.

Deininger, K., Zegarra, E. & Lavadenz, I. 2003. Determinants and impacts of rural land market activity: evidence from Nicaragua. World Devel., 31(8): 1385-1404.

Deininger, K., Feder, G., Gordillo de Anda, G. & Munro-Faure, P. 2003. Land policy to facilitate growth and poverty reduction. Land Reform, Land Settlement and Cooperatives, 2003/3: 5-18.

de Janvry, A. 1981. The agrarian question and reformism in Latin America. Baltimore, MD, USA and London, The Johns Hopkins University Press.

de Janvry, A. & Sadoulet, E. 2002. Access to land for the rural poor: how to keep it open and effective for poverty reduction? Revista Econômica, 4(2): 253-277.

de Janvry, A., Murgay, R. & Sadoulet, E. 2001. Rural development and rural policy. In B. Gardner & G. Rausser, eds. Handbook of agricultural economics, Vol. 2A. Amsterdam, Elsevier.

De Silva, S. 2000. Skills, partnership and tenancy in Sri Lankan rice farms. Discussion Paper No. 819. New Haven, CT, USA, Economic Growth Center, Yale University.

El Ghonemy, M.R. 2002. The land market approach to rural development. In V.K. Ramachandran & M. Swaminathan, eds. Agrarian studies - essays on agrarian relations in less-developed countries. New Delhi, Tulika Books.

FAO. 1997. Report on the 1990 world census of agriculture. FAO Statistical Development Series No. 9. Rome.

Griffin, K., Khan, A.R. & Ickowitz, A. 2002. Poverty and the distribution of land. J. Agrarian Change, 2(3): 279-330.

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

Jacoby, H.G. & Mansuri, G. 2002. Incomplete contracts and investment: a study of land tenancy in Pakistan. Washington, DC, World Bank. (mimeo)

Kung, J.K. 2002. Off-farm labor markets and the emergence of land rental markets in rural China. J. Compar. Econ., 30(2): 395-414.

Lastarria-Cornhiel, S. & Melmed-Sanjak, J. 1999. Land tenancy in Asia, Africa and Latin America: a look at the past and a view to the future. Working Paper No. 27. Madison, USA, Land Tenure Center, University of Wisconsin. (mimeo)

Lyne, M., Zille, P. & Graham, D. 2000. Financing the market-based redistribution of land to disadvantaged farmers and farm workers in South Africa: recent performance of the Land Reform Credit Facility. Madison, USA, Land Tenure Center, University of Wisconsin. (mimeo)

Newbery, D.M.G. 1977. Risk-sharing, sharecropping, and uncertain labor markets. Rev. Econ. Stud., 44: 585-594.

Otsuka, K. 1993. Land tenure and rural poverty. In M.G. Quibria, ed. Rural poverty in Asia. Priority issues and policy options. Manila, Asian Development Bank.

Paasch, A. 2003. The failure of market assisted land reforms and some necessary consequences. Comments on the World Bank’s Policy Research Report. Bonn, Germany, FoodFirst Information and Action Network. (mimeo)

Ramachandran, V.K. & Ramakumar, R. 2000. Agrarian reforms and rural development policies in India. A note. Paper presented at the International Conference on Agrarian Reform and Rural Development, 5-8 December 2000. Tagaytay City, Philippines.

Roy, J. & Serfes, K. 2000. Strategic lengths of tenancy contracts. Paper presented at the International Symposium on Development Policies for the New Millennium, 13-16 July 2000. Mumbai, India, Indira Gandhi Institute of Development Research.

Ruben, R. & Masset, E. 2003. Land markets, risk and distress sales in Nicaragua: the impact of income shocks on rural differentiation. J. Agrarian Change, 3(4): 481-499.

Sadoulet, E., Murgai, R. & de Janvry, A. 1998. Access to land via rental markets. Paper presented at the FAO/WIDER Seminar on Access to Land, Rural Poverty and Public Action, April 1998. Santiago, Chile.

Salanié, B. 1997. The economics of contracts. A primer. Cambridge, MA, USA, The MIT Press.

Sauer, S. 2001. A proposta de "reforma agrária de mercado" do Banco Mundial no Brasil. Brasilia, Senado Federal. (mimeo)

Sender, J. & Johnston, D. 2004. Searching for a weapon of mass production in rural Africa: unconvincing arguments for land reform. J. Agrarian Change, 4(1-2): 142-164.

Shankar, K. 1999. Tenancy reform: rhetoric and reality. Economic and Political Weekly, 20- 26 November.

Sharma, N. & Drèze, J. 1996. Sharecropping in a north Indian village. J. Devel. Studies, 33(1):1-39.

Swinnen, J.F.M. 2002. Political reforms, rural crises, and land tenure: a historic analysis of land leasing and tenure reform in Western Europe. Leuven, Belgium, Policy Research Group, Leuven University, Belgium. (mimeo)

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

U. PICA CIAMARRA

Accès aux terres et loyers fonciers: évolution (ou régression) de la politique foncière de la Banque mondiale?

Le rapport sur les recherches en matière de politiques, publié en 2003 par la Banque mondiale et intitulé «Land policies for growth and poverty reduction» passe en revue les analyses théoriques et les expériences concrètes réalisées, sous l’angle de l’économie et des politiques économiques concernant l’accès à la terre, et présente des orientations générales pour la formulation d’avis en matière de politiques. Il démontre que les droits de propriété foncière garantis peuvent considérablement améliorer le bien-être des plus démunis et que l’accès aux terres peut être assuré grâce à un ensemble hétérogène de politiques, qu’elles soient gouvernementales ou qu’elles obéissent aux lois du marché. En particulier, il est montré que, dans les pays en développement, les baux à ferme garantis à long terme peuvent assurer une allocation rationnelle des ressources et servir les intérêts des pauvres, et que les gouvernements devraient lever les restrictions arbitraires entravant les opérations sur les marchés des loyers fonciers. Dans le droit fil du rapport de la Banque mondiale, cet article montre que selon les renseignements disponibles, les contrats de location à long terme ne sont pas nécessairement indispensables pour mettre en place des incitations optimales à l’investissement, et que les marchés des loyers fonciers sont rarement axés sur les pauvres. Les décideurs devraient procéder avec circonspection lorsqu’ils préconisent la libéralisation des marchés des loyers fonciers et lorsqu’ils avancent que les baux à long terme sont bénéfiques aux pauvres, car sur les marchés des loyers fonciers libéralisés qui seraient ainsi créés, avoirs et efficacité pourraient représenter un compromis coûteux.

Acceso a la tierra mediante los mercados de arrendamiento: ¿una (contra)evolución en la política agraria del Banco Mundial?

El informe de investigación sobre Políticas agrarias para el crecimiento y la reducción de la pobreza del Banco Mundial, publicado en 2003, examina análisis teóricos y experiencias prácticas de la economía y la economía política del acceso a la tierra, y presenta algunas orientaciones generales para prestar asesoramiento sobre políticas. Hace hincapié en que el bienestar de los pobres puede aumentar en gran medida gracias a los derechos seguros de propiedad de la tierra, y en que una heterogénea variedad de políticas, tanto estatales como impulsadas por el mercado, pueden facilitar el acceso a la tierra. En particular, se argumenta que, en los países en desarrollo, los contratos seguros de tenencia de la tierra de larga duración pueden brindar amplias oportunidades para garantizar una asignación eficiente de los recursos y beneficiar a los pobres, y que los gobiernos deberían suprimir todas las restricciones injustificadas al funcionamiento de los mercados de arrendamiento de tierras. En consonancia con el informe del Banco Mundial, en el artículo se demuestra, por un lado, que los contratos de tenencia de larga duración prueban que no tienen por qué ser esenciales para incentivar más la inversión inicialmente y, por otra parte, que los mercados de arrendamiento casi nunca son favorables a los pobres. Los encargados de formular políticas deberían ser prudentes con respecto al apoyo a la liberalización de los mercados de arrendamiento de tierras y al argumento de que la tenencia de larga duración es buena para los pobres; en los mercados liberalizados de arrendamiento de tierras resultantes el costo de la equidad y la eficacia podría ser elevado.