Research on the relation between security of land tenure and agricultural productivity has not been extensive; more results are available on the relation between farm size and productivity. The empirical evidence available regarding the role of tenure all has confirmed that the relationship between tenure security and measures of productivity is positive. Path-breaking work in this regard has been carried out by Gershon Feder and associates in Thailand. Feder and Tongroj Onchan analyzed data for more than 500 farms in three provinces of Thailand, and they concluded that land titling gave rise to greater on-farm investment in two of the provinces and to the acquisition of more farmland in the third.46 Feder and Noronha cite research in Costa Rica and Brazil which showed a positive relation between ownership security and the amount of investment per unit of land.47 Intensity of land use also has been shown to respond positively to tenure security for the case of Jamaica.48 Atwood has cited research demonstrating a link between tenure security and farm income in the case of Kenya (D. A. Atwood, 1990, p. 659). Binswanger et al. (1995) also cite the results of a less formal study in Ecuador which supports the relationship between titling and farm income levels.
Feder and colleagues have found a clear, positive relationship between land titling and the availability of credit to the farmer. They present evidence from India, Thailand and South Korea which shows that both formal and informal institutions are more willing to lend to farmers when land is used as collateral.49 In addition, they analyze data from Lop-Buri Province, Thailand,
46. Gershon Feder and Tongroj Onchan, 'Land Ownership Security and Capital Formation in Rural Thailand', Discussion Paper ARU 50 (revised version), Research Unit, Agriculture and Rural Development Department, The World Bank, Washington, DC, USA, February 1986.
49. Reprinted from Agricultural Economics, 2, G. Feder, T. Onchan and Tejaswi Raparla, 'Collateral, guarantees and rural credit in developing countries: evidence from Asia', pp. 234-236, Copyright (1988), with permission from Elsevier.
under two different econometric models, concluding that (p. 243) 'in the area studied, the pledging of land collateral increases the amount of institutional credit offered by 43% ... or 55% ... as compared to a loan without a security'.
Feder and Noronha mention that a study by Seligson in Costa Rica50 showed that titling increased the share of farms obtaining credit from 18% to 31.7%, and a study in Jamaica51 showed that titling enabled almost half of the recipients to increase their borrowing.52
In an analysis of land titling experiences in Nicaragua, Deininger and Juan Sebastian Chamorro found that a program of conferring registered (legally binding) titles on small farms increased land values by an average of 30% and increased the propensity to invest in the land by 8-9 %.53
For the relationship between farm size and productivity, perhaps the most thorough empirical investigation was carried out by Rasmus Heltberg for Pakistan. With farm-level panel data, he analyzed the effect of farm size on value added and land returns per hectare, controlling for differences in soil quality, market imperfections in labor and credit, and other variables. He obtains 'highly significant' statistical results that indicate a U-shaped relationship between farm size and each of the two measures of productivity, i.e. value added and land returns per hectare. However, in a range of farm sizes that account for 90 % of all farms and 65% of farmland, his results show an inverse relationship. He hypothesizes (as other authors have done) that the main explanation for the inverse relationship is that hired labor is an imperfect substitute for family labor. In his own words, his main conclusions are as follows:
Three lines of criticism have been raised in the literature against the IR hypothesis [inverse relationship between farm size and productivity]: (a) that the empirical evidence is flawed due to omitted variable bias, (b) that the relationship may no longer hold after the Green Revolution, and (c) that a consistent explanation for the inverse relationship is missing.
With respect to the first criticism, the article has presented strong evidence to support the presence of an inverse size-output relationship even when soil and other heterogeneity is controlled for. .. . With regard to the second point of critique, it was found that small farms are still significantly more productive than big farms, also in irrigated and relatively developed areas of Pakistan. With respect to the third point... a set of reasonable hypotheses about labor, land, credit and insurance market failures was set up to account for systematic size-output and size-profit relationships. .. . the market imperfections framework conforms well with the data.54
Heltberg finds imperfections in the market for credit as well as farm labor. There also is a U-shaped relationship between credit used per hectare and farm size, with the credit variable declining over a considerable range of farm sizes. When credit per hectare begins to increase with farm size, he conjectures that it may contribute to the greater productivity of very large farms over medium-sized farms.
Another careful empirical study of the farm size-productivity relationship was conducted by Fidele Byiringiro and Thomas Reardon for Rwanda. Their principal results were consistent
50. M. A. Seligson, 'Agrarian reform in Costa Rica: the impact of the titles security program', Inter-American Economic Affairs, 35(4), Spring 1982.
51. Inter-American Development Bank, 'Jamaica Land Titling Project Feasibility Report', Inter-American Development Bank, Washington, DC, USA, 1986.
52. Feder and Noronha, 1987, pp. 144-145.
53. Klaus Deininger and Juan Sebastian Chamorro, 'Investment and income effects of land regularization: The case of Nicaragua', The World Bank and the University of Wisconsin at Madison, Washington, DC and Madison, WI, USA, mimeo, January 2002.
54. R. Heltberg, 'Rural market imperfections and the farm size-productivity relationship: evidence from Pakistan', World Development, 26(10), 1998, pp. 1823-1824.
with those of Heltberg, including the finding of a U-shaped relationship over a wider range of farm sizes. Their conclusions are as follows:
We explored: (1) whether the smaller farms have greater average and marginal land productivity than the larger farms, and whether the smaller farms are less allocatively efficient; and (2) whether . .. soil erosion reduces, and soil conservation investments increase, land productivity. Both queries were answered with a strong affirmative. Moreover, the inverse relationship is not mitigated by the smaller farms' being more eroded, despite their farming more intensively (with less fallow). In fact, smaller farms are not more eroded than larger farms. Moreover, the inverse relationship is not mitigated by larger farms' using more nonlabor inputs or by their putting more of their land under cash perennials. In fact, larger farms do not do more of either compared to smaller farms.
... we find the marginal value product of land on smaller farms to be well above the rental price of land, implying factor use inefficiency and constraints to land access. By contrast, the marginal value product of labor on smaller farms was well below the market wage. This implies a 'bottling up' of labor on smaller farms, and constraints to access to labor market opportunities. . . .55
In addition to these econometric studies, tabulations of data from agricultural censuses and surveys in a number of countries show markedly higher productivity per hectare on small farms than on large farms. Binswanger, Deininger and Feder cited early findings along these lines by Albert Berry and William Cline which show that the value of output per hectare is 5.6 times larger on small farms than on the largest ones in Northeast Brazil, 2.7 larger in Pakistan's Punjab, and 1.5 times larger in the Muda irrigation scheme of Malaysia.56 Similarly, for Honduras census data for the mid-1970s showed that farming income per hectare fell from 584 lempiras for farms of 0-2 hectares to 215 lempiras for farms of 10-20 hectares.57 Comparable findings have been made for a number of other countries. The inverse relationship between farm size and productivity per hectare in developing countries is now widely accepted as generally correct, although exceptions always may be found.
Was this article helpful?