不留遗憾地进行专业化分工: 在工业化经济体中的土地转让权、农业生产率和投资

___e_ o: POLICY RESEARCH WORKING PAPER 2202 Specialization without In China, where collectives own farmland but farmers Regret may hold 'use rights" to the land, a case can be made for Transfer Rights, Agricultural a property rights system with incomplete security of tenure Productivity, and Investment but with strong transfer in an Industrializing Economy rights, which permit "specialization without regret" MicbaelRP. Ca?ter -so farmers can recoup the Yang Yao value of an investment even if they exit farming The World Bank Development Research Group Rural Development October 1999  POLICY RESEARCH WORKING PAPER 2202 Summary findings A number of studies have examined the effects of secure increased if the household population increases, and how tenure on agricultural investment and productivity. frequent average land adjustments are under the Carter and Yao also study the importance of rights to household responsibility system. household residual income and land use being Analyzing panel data for a sample of farm households, transferable. Carter and Yao study the "investment regret mitigation Contemporary China - where industrialization has effect," which results when greater transfer rights make spread rapidly, if unevenly - is a good place to study the households more willing to invest because they are less economic effects of transfer rights as well as likely to regret such investments when they can recoup conventional security of tenure. Village collectives the investment value even if they exit farming. formally own land in China, so there can be no Carter and Yao find that transfer rights may be individual land sales, but farmers are sometimes entitled especially important in an industrializing economy. A to sell their rights to use the land allocated to them under property rights system with incomplete security of tenure the household responsibility system. but with strong transfer rights that permit "specialization Whether a household has secure tenure depends on without regret" - so farmers can recoup the value of an whether its landholding will be reduced if the household investment even if they exit farming - may have much population declines, whether the landholding will be to recommend it. This paper - a product of Rural Development, Developnment Research Group - is part of a larger effort in the group to study the determinants and impact of property rights systems and land tenure regimes in the process of development. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Maria Fernandez, room MC3 -542, telephone 202-473-3766, fax 202-522-1151, Internet address mfernandez2@ worldbank.org. Policy Research Working Papers are also posted on the Web at http:!/wTww.worldbank.org/html/dec/Publications/ Workpapers/home.html. The authors may be contacted at carter@caae.wisc.edu or yyao@ccer.pku.edu.cn. October 1999. (50 pages) The Policy Research Working Paper Series dissemninates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the Wtorld Bank, its Executive Directors, or the countries they represent. Produced by the Policy Research Dissemination Center  SPECIALIZATION WITHOUT REGRET: TRANSFER RiGHTS, AGRICULTURAL PRODUCTIVITY AND INVESTMENT IN AN INDUSTRIALIZING ECoNoMY* Michael R. Carter Department of Agricultural and Applied Economics University of Wisconsin-Madison Madison, WI 53706 USA and Yang Yao The China Center for Econornic Research Beijing University Beijing, China 100871 * The authors thank the World Bank and the Ford Foundation for their support of this research. Gershon Feder commented extensively on an earlier version of this work. We him, Klaus Deininger, Hanan Jacoby, Pedro Olinto and seminar participants at the World Bank, the University of Wisconsin and the .1998 American Agricultural Economics Association Annual Meetings for their comments. Any remaining errors and all views expressed in this paper remain the responsibility of the authors and should not be attributed to employing or funding institutions.   Specialization without Regret: Transfer Rights, Agricultural Productivity and Investment in an Industrializing Economy Among those mutable factors that constrain economic performnance in low income and transitional economies, property rights over land rank high by most calculations.' With a few important exceptions, the literature that has tried to quantify the economic effects of property rights has concentrated on the security with which individuals hold rights over land. The goal of this paper is to broaden the focus of these earlier efforts to include transfer rights, meaning an individual's liberty to rent or sell the use, income and other rights that he or she holds. Theoretically this paper shows that transfer rights permit specialization in off-farm activities economy industrializes and its population begins to specialize in non-agricultural activities. Transfer rights should also facilitate (shadow) factor price equalization across households and resolve problems of allocative inefficiency that are likely to become more severe under the pressure of industrialization. This paper's econometric analysis of household level panel data from China largely corroborates these theoretical propositions concerning transfer rights. With its finding that transfer rights seem to matter more than tenure security, the empirical analysis also suggests some novel directions for future property rights reform in China. The liberalization of transfer rights has emerged as a contested and yet poorly understood policy issue in a number of countries, both in those where land is or until recently has been held under customary tenure arrangements, and in those where earlier redistributive reforms reshaped land ownership structure. This paper considers transfer rights in the context I See for example the discussion in Alston, Libecap and Schneider (1996) and the many references contained therein. 1  of China and its particular history of reforms that generated a highly egalitarian distribution of land.2 China's household responsibility system reform of the late 1970s and early 1980s replaced collective production teams with household-based production units, and assigned use and residual income rights to individual households. Transfer rights and tenure security were not fixed nationally at the time of these reforms, and they have been subject to local determination and subsequent evolution (Liu et al. 1998). A number of analysts have credited the hybrid property rights system created by the household responsibility system reforms for the rapid and sustained agricultural growth that took place in China in the early 1980s. While the exact weight of the reform in this growth spell can be disputed,3 there is little doubt that the reforms unambiguously sharpened individual work incentives. Moreover, for an economy that was labor intensive and where most rural people worked full-time in agriculture, the reforms should also have attained approximate allocative efficiency without any factor market transactions given that land was allocated to families in rough proportion to their labor endowments (Burgess 1998). The limited transfer rights and weak tenure security that emerged in many areas may have been relatively unimportant in the initial post-reform period. However, the subsequent and rapid growth in industrial employment and wages has undercut the conditions that may have rendered transfer rights unimportant.4 With the 2 Liberalization of transfer rights could be argued to work differently in less egalitarian economies (e.g., Mexcio or Nicaragua) because large wealth differentials create potentially radically different patterns of capital access that spill over and influence the function of the land market (eg., see Carter and Salgado, forthcoming). 3 Lin (1992) and MacMillian, Whalley, and Zhu (1989) argue that the property rights reform was central, while Putterman (1992) warns that their analysis overstates the importance of property rights reform relative to price reform. 4 Carter and Yao (1998) for exanple show that land-labor endowment distribution has become more disperse over time. 2  slowdown in agricultural growth after mid-1980s, the wisdom of the hybrid, village-based land tenure system has come under both academic and policy scrutiny. Some argue that restricted transfer rights and frequent insecurity-inducing land redistribution and small landholdings rooted in the village-based land tenure system have become the major hindrance to the improvement of agricultural productivity. Others argue that whatever its costs, the current land tenure system provides important benefits by functioning as a mechanism of rural social insurance (Dong, 1996 and Kung 1994). While the work presented here does not address the social insurance benefits of the current system, it does try to identify the more narrowly construed investment and productivity benefits that might attend a further liberalization of property rights. To this end, Section 1 below develops a two period model that identifies three kinds of effects that further property rights reform in China may have: 1. A Security-Induced Investment Demand Effect that results when households perceive a reduction in the likelihood that land in which they might sink, attached, long-lived investment will be reallocated to other households5; 2. An Investment Regret Mitigation Effect, that results when greater transfer rights make households more willing to make investments because they are less likely to regret sinking investment in the land because they become able to recoup the value of the investment even if they should exit farming. 3. A Factor Price Equalization Effect that results when increased transfer rights facilitate the equalization of returns to land, labor and short-lived capital between household farming units. The first of these three effects has been studied extensively in the literature (see the summary 5We omit the collateral effect because land is not allowed to be used as loan collateral in current China. Besides, using land as a collateral is a doubtful punishment for loan default in an economy where land per capita ranges from only one thirtieth to one tenth of a hectare, as our survey revealed (Table 1). 3  by Feder and Akihiko, 1996).6 Distinguishing it from the latter two effects, which result from liberalized transfer rights, faces a number of challenges, including the fact that transfer rights and tenure security often move together. Fortunately, the theory developed in Section 1 below suggests a number of quite specific tests that permit us to distinguish security from regret and factor price equalization effects. Many of these tests are variations on what have become known as separabilty tests that ask whether shadow prices and factor choices in production independent of household endowments and consumption choices. Interestingly, the modeling done here suggest that much conventional separability testing has been mis- specified in the sense that non-separability, if it holds at all, applies differently across different market participation regimes and that a single regression model that pools observations across regimes is inappropriate. Section 2 develops an econometric implementation of the theoretical model. While offering a powerful, if statistically conservative, control for latent variables, the panel data methods put forward force reliance upon simulated maximum likelihood (SML) methods given multiple land market participation regimes and the presence of censoring in the investment data. Using data collected in 1988 and 1993 on. a sample of 400 rice-producing households spread across 40 villages in 2 Chinese provinces, Section 3 then presents the SML results. While the models and methods demand a lot of the data, both investment regret and factor price equalizations effects prove significant. Interestingly, the tenure security effects 6 For China, recent studies by Rozelle, Li and Brandt (1998), Carter and Yao (1998), and Jacoby Li, and Rozelle (1998) find some evidence of tenure security effects, standing in contrast to the earlier study by Feder, Lau, Lin, and Luo (1992) who found that the link between fanners' perception of tenure security and farm investment was very weak if positive at all. The investment regret effect is similar to what Besley identifies as the tenns of trade effect, while the factor price equalization effect has parallels to Gavian and Fafchamps (1996) tests of allocative inefficiency. 4  appear insignificant. As discussed in the concluding Section 4, these results provide insights into promising directions for further property rights reform. Section 1 A Model of the Impact of Tenure Security and Transfer Rights on Allocative Efficiency and Investment Incentives This section develops a two period model that identifies the avenues through which transfer rights and tenure security affect agricultural productivity and investment. Constrained by the degree of tenure security and land transfer rights that they hold, households in the first period make investment decisions and allocate variable factors of production between on- and off-farm uses, taking into account the expected impact of these decisions on their future wellbeing. Prior to the second period, two village-level shocks are realized, a land redistribution shock and an off-farm employment growth shock. Households then allocate variable factors of production for second period production. In order to highlight the factor price equalization effect of land transfer rights, this section first analyzes this model for the special case in which there is no prospect that off-farm TVE employment will expand in the second period. We then consider the more general case of employment growth shocks and show the additional impacts generated by the resulting investment regret effect. 1.1 The Model Households enjoy endowments of labor time, L, and contract land, IC. The exact bundle of rights held over contract land is variable and is detailed below. Each period, households divide their endowment of family labor between on-farm agricultural uses (Lf, home or sideline enterprises (L'), and wage employment in township and village enterprises 5  (L). Income in these activities is generated as follows: Ai F(K, Tf , Lf ) g(Ls); and, wLw, where A, is a household-specific technical efficiency (or agricultural comparative advantage) term, K is capital, Tf is farmed or operated land F is a conventional constant returns to scale production function with positive cross partials (e.g., FLK >0).7 We assume that returns to labor allocated to sideline diminish (g' >0, g"< 0), and that the TVE wages exceed the marginal agricultural labor product (Yao 1998) such that the time allocated by a household to these activities is bound by the constraint Lw < Law Capital in the agricultural production function is long-lived, and attached to the land. Examples of such capital include terraces, fencing, and fertile soil. In first period, there is no inherited capital stock, and all capital is purchased by a price of rK, and capital goods are spread out evenly across a farm's cultivated area (both own and rented land). To sidestep issues concerning investment disincentives on rented land, we assume that landlords pay a per- mu rebate of kjvk , where k1 is the installed capital stock per-mu, and vk is the rebate rate that 7 Because in our particular empirical application, households' endowments of productive resources are small enough that labor hiring is trivial (except for peak season agricultural tasks), we ignore hired labor, effectively assuming that households are optimally in the regime where no labor hiring is desirable. Generalizing the model to consider hired labor could take one of two directions. The first would be to assume that labor effort-measured in efficiency units-can simply be hired at a fixed agricultural wage rate. A second, and more realistic approach, would be to assume that households face what Bowles (1985) calls a "labor extraction" problem, meaning that without supervision or other incentives, hired laborers supply very little efficiency labor per-unit time. Under the latter specification, the model changes little if supervisory costs increase with farm size as the equilibrium shadow wage will increase with farm size, as in the present model. 6  that individuals expect to equal to their expected marginal second period use value of capital.8 Households hold full current use and residual income rights to their contract land endowment. There are, however, two other important dimensions of the property rights to contract land. First, as discussed above, in some areas of rural China, households may not rent-out their land, while in other areas land renting is allowed subject to various conditions that range from shifting the incidence of the land tax and other collective duties, to obtaining the acquiescence and approval of local authorities. In still other areas, land may be freely rented with no restrictions or regulation (see Liu et al., 1998). The severity of restrictions on private transfers of land rights, M, is defined such that higher values indicate less encumbered (more freely transferable) land rights. We assume that the restrictions on land transactions create a per-mu cost, co(M) (with co' <O) for those that rent-out land; and, c(AM) (with c" <0) for those that rent-in.9 The net returns from renting out R