9512.net
甜梦文库
当前位置:首页 >> 经管营销 >>

Impacts of land redistribution on land management and productivity


land degradation & development
Land Degrad. Develop. 12: 555±568 (2001) DOI: 10.1002/ldr.473

IMPACTS OF LAND REDISTRIBUTION ON LAND MANAGEMENT AND PRODUCTIVITY IN THE ETHIOPIAN HIGHLANDS
S. BENIN1* AND J. PENDER2
2 1 International Livestock Research Institute, Addis Ababa, Ethiopia International Food Policy Research Institute, Washington, DC, USA

Received 6 September 2001; Accepted 27 September 2001

ABSTRACT The increasing problem of landlessness in Ethiopia has put pressure on regional governments to redistribute land. In 1997 and 1998, a major land redistribution was undertaken in the Amhara Region, reducing landlessness where implemented. While the impacts of such redistributions have been hotly debated, little empirical evidence exists concerning the actual impacts of redistribution. We nd that land redistribution in the Amhara Region has had a positive impact on land productivity, by increasing access to land for farmers who are more interested or able to use purchased inputs such as fertilizer and herbicides. Our results, however, do not show much effect of the recent land redistribution or expectations of future redistribution on land improvement and management. Thus, to the extent that investments in land improvement are necessary for conservation purposes, it appears that policy change to stop land redistributions is unlikely to have a substantial impact on reducing land degradation. Credit and extension programmes and improving land rental markets, however, present better strategies for improving land management in this region of Ethiopia. Copyright # 2001 John Wiley & Sons, Ltd.
key words: Ethiopian highlands; land redistribution; land management and productivity

INTRODUCTION Reducing environmental degradation is a major challenge in the Ethiopian highlands. Land degradation, especially soil erosion [averaging 42 t ha1 yr1 on cultivated lands (Hurni, 1988)], low and declining soil fertility, soil moisture stress and deforestation are critical problems contributing to low agricultural productivity (which is reected in cereal yields averaging less than 1 t ha1), poverty and food insecurity in these areas (Fistum et al., 1999; Lakew et al., 2000). Several government policies and strategies have been implemented, both in the past and present, to solve these problems. One of the major ones was the redistributive land reform in 1974, when the military government (/ Derg) / took power and ended all forms of tenancy, nationalized all rural lands, and redistributed land to the tillers. Individual land rights were restricted (i.e. land sales and leasing, using land as collateral, and hiring labour were prohibited) to ensure that the tillers remained the beneciaries of agricultural production. Land redistribution by the government was then established as the only means to improve access to land and reduce landlessness. Since the fall of the military government in 1991, the new Ethiopian government has allowed land leasing and hiring labour, but the prohibition on land sales has continued, codied in the new constitution. The increasing problem of landlessness has put pressure on regional governments to redistribute land frequently. For example, in 1997 and 1998 the Amhara regional government implemented a major land redistribution. While the impacts of such land redistributions have been hotly debated in Ethiopia and elsewhere, little empirical evidence has been available concerning the actual impacts of land redistribution.
Correspondence to: Dr S. Benin, International Livestock Research Institute, PO Box 5689, Addis Ababa, Ethiopia. E-mail: s.benin@cgiar.org Contract/grant sponsor: The Norwegian Ministry of Foreign Affairs. Copyright # 2001 John Wiley & Sons, Ltd.

556

S. BENIN AND J. PENDER

Although redistributive land reform is sometimes utilized as a deliberate policy instrument to improve agricultural productivity and reduce poverty,1 it is argued that land redistribution erodes tenure security and that farmers will not undertake land-improving investments because they may not be able to claim fully the returns on their investment. Thus, to the extent that investments in land are required for conservative purposes and to increase productivity, land redistribution will further promote land degradation and reduce farm output. Currently, the Amhara regional government is considering a land policy that will end land redistributions. A critical question therefore is whether abolishing land redistribution will improve investments in land that will in turn reduce land degradation and increase productivity. In this paper, we present evidence and policy implications of the impacts on land management and productivity of land redistributions in the Amhara Region of Ethiopia. We also investigate other factors (including agricultural potential, market access, population density, credit and extension programmes and education) that may affect land management and agricultural productivity. The study is based on analysis of a community-level survey conducted in 98 villages ( gots ) in the highland / / areas (above 1500 m.a.s.l.) of the Amhara Region in 2000. A stratied random sample of 49 peasant associations (PAs, usually consisting of three to ve villages) and two villages randomly selected from each PA were selected from highland areas of the region. Using district (// wereda) level secondary data, the stratication was based upon / indicators of agricultural potential (whether the wereda is drought-prone or non-drought-prone/higher rainfall, as // / classied by the Ethiopian Disaster Prevention and Preparedness Committee), market access (access or no access to an all-weather road) and population density (1994 rural population density greater than or less than 100 persons km2). Two additional strata were dened for PAs where an irrigation project is present (in drought-prone vs. higher rainfall areas), resulting in a total of 10 strata. Five PAs were then randomly selected from each stratum (except the irrigated drought-prone stratum, in which there were only four PAs), for a total of 49 PAs and 98 villages.Weredas predominantly (more than 50 per cent of total area) below 1500 m.a.s.l. were excluded from the /// sample frame. Information were collected at both PA and village level using group interviews with about ten respondents from each PA and village, selected to represent different genders, ages, occupations, and in the PA-level survey, different villages. Information collected includes changes in land investments, land management practices, input use and agricultural production since 1991 (the year when the current government replaced the former Marxist government). The data were supplemented by secondary information on population from the 1994 population census, geo-referenced maps of the boundaries of each sample PA, and geographic attributes, including altitude and climate. LAND REDISTRIBUTION Land redistribution has been carried out in many developing countries, often as part of land reform in the wake of social and political revolution. Sometimes, however, land redistribution is utilized as a deliberate policy instrument to capture the efciency benets of the family farm, decrease urban food prices and reduce poverty (Prosterman and Riedinger, 1987). The 1974 redistributive land reform in Ethiopia shared many similar attributes (e.g. prohibiting farmland sales and other transfers and hiring of labour to ensure that the actual farmers remained beneciaries of the land) to Land-to-the-Tiller programmes implemented in other countries (e.g. Philippines in 1972). However, in the case of Ethiopia, redistribution of farmland was undertaken regularly (in many cases as often as every one to two years during the / regime; Fistum et al., 1999) to reduce landlessness and to equalize Derg / land holding and quality. Thus, those with larger elds had part of their land taken away and given to others that did not have land or who had smaller elds. Local administrations, known as peasant associations, were set up and charged with the responsibility for land redistribution.

1 This policy stems from many case studies that show an inverse relation between farm size and productivity (Berry and Cline, 1979; Prosterman and Riedinger, 1987). The underlying argument is that by giving land to the actual tillers, the high transaction and monitoring costs associated with tenancy and using hired labour are eliminated.

Copyright # 2001 John Wiley & Sons, Ltd.

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

IMPACTS OF LAND REDISTRIBUTION

557

After the fall of the / in 1991, the new government constitutionalized state ownership of all rural lands. Derg / However, the new constitution, drawn in 1994, allows temporary leases and guarantees the rights of peasants of free access to land and the right to improvements they make on land including the right to bequeath, transfer, remove or claim compensation for such improvements when the right to use the land expires. In principle, farmers now have the right to use the land indenitely, lease it out temporarily to other farmers, and transfer it only to their children. However, they still cannot sell or mortgage their lands. Although the constitution has resolved some issues, it seems to create other ambiguities and does not address some important issues (Fistum et al., 1999). For example, given land scarcity, it is not clear how farmers' rights of free access to land can be assured in practice, and how much land they are entitled to.2 Regional governments have been charged to resolve those issues and there have been signicant differences across the regions with respect to development of a regional land policy and redistribution of land. For example, in the Tigray region, land redistribution was stopped in 1991, and the policy of no future redistribution was made ofcial by a new land policy in 1997. In the Oromiya region, too, there has not been a redistribution for more than 10 years (Bezuayehu et al., 2000), although the regional government has not made any ofcial statement about abandoning it. In the Amhara Region, however, land redistribution has been very common, with a recent and major one undertaken in 1997 and 1998. Although there is no regional land policy per se, administration and use of land in the region have been guided by the provisions made in the national constitution. In 2000, the regional government passed a land policy document that will determine the administration and use of rural land in the region. The document is yet to be made public or proclaimed, however. The provisions in that document are similar to those provided in the national constitution, including: the right of peasants to free and indenite use of land, transfer to dependents, consolidate holdings and rent out; right to use, sell, exchange or transfer the wealth cultivated on their land; but not the right to sell or exchange the land. Other important issues such as registration of the timing and limitations of renting out land, maximum land holding and plot size to be used for rain-fed and irrigated agriculture have been relegated to by-laws that will be decided in the future. On the issue of land redistribution, the draft document states: So long as giving a land free to farmers is maintained, land redistribution shall not be effective unless otherwise the land division does not affect the productive capacity required by the community and unless decided by law (ANRSC, 2000: section 3, article 10). Although the document is yet to be proclaimed, the above statement suggests that land redistribution in the future is not completely ruled out. Examining the incidence of land redistribution in the Amhara Region, the survey conducted in the region shows that every community has experienced at least one redistribution since 1974, and nearly half have had a land redistribution since 1991, mainly in the recent redistribution in 1997 and 1998. The average number of land redistributions is three, with one village experiencing as many as 14 since 1974. About four-fths of the communities expect a redistribution in the future, most within the next few years. Informal discussions with some of the farmers in the region revealed that although they do not fully support land redistribution, it is seen as a necessary tool by which landless farmers gain access to farmland, especially since land sales are prohibited and other transfers are restricted. In the next section, we examine the conceptual framework for analysing the impacts of land redistribution on land investments, land management, input use and productivity. CONCEPTUAL FRAMEWORK AND HYPOTHESES The conceptual framework and hypotheses about how land redistribution may inuence land-improving investments, land management, input use and productivity draw from the literature on property rights and
2

In principle, peasants are entitled to farmland when they reach the age of 18, and the maximum allowable holding is set at 10 ha.

Copyright # 2001 John Wiley & Sons, Ltd.

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

558

S. BENIN AND J. PENDER

investment incentives (Barrows and Roth, 1990; Migot-Adholla et al., 1991; Feder and Feeny, 1993; Place and Hazell, 1993; Besley, 1995; Gavian and Fafchamps, 1996; Quisumbing et al., 1999; Pender and Kerr, 1999; Place and Swallow, 2000). Although land redistributions cause tenure insecurity, they may have mixed impacts on farmers' land management and productivity through short- and long-term effects of redistribution. On one hand, expectations of land redistribution may undermine farmers' incentives to invest in land improvements and soil fertility, since farmers' abilities to reap the benets of such investments are undermined (Feder and Feeny, 1993). On the other hand, redistribution may improve access to land of households that have relative surpluses of other important factors of production, such as labour, oxen or cash to purchase inputs, particularly in the context of prohibited land sales and restricted lease markets as exist in Ethiopia. Therefore, land redistribution may increase input intensity, which may in turn increase productivity. Furthermore, the threat of redistribution may encourage farmers to invest if investments reduce the perceived likelihood of losing access to a given piece of land (Snyder, 1996; Quisumbing et al., 1999). Thus, land redistribution may either increase or decrease investments in land improvement, the intensity of land management, use of purchased inputs and productivity. ECONOMETRIC APPROACH AND RESULTS Econometric analysis was used to investigate the effects of recent (1997 and 1998) land redistribution and expectations of future redistribution on: (1) farmers' land investments (construction of stone terraces, soil bunds, check dams, drainage ditches, waterways and irrigation canals, and planting of trees and live barriers) since 1991; (2) farmers' land management practices (use of burning to prepare elds, traditional fallow, improved fallow, contour ploughing, crop rotation, mulching, manuring, composting, ploughing in crop residues, soil burning and minimum tillage) in 1999; (3) farmers' use of purchased inputs (fertilizer, pesticides, herbicides and improved seeds) in 1999; and (4) crop yields (barley, wheat and teff) in 1999. Summary statistics of these variables are presented in Table I. With the exception of drainage ditches, which seem to be part of normal cultural practices, the other investments since 1991 have been undertaken by less than 50 per cent of the households. Most of the management practices have not changed since 1991, except for fallowing and soil burning, which have declined by more than 50 per cent. Use of fertilizers, herbicides and improved seeds have increased by more than 50 per cent, with the use of improved seeds almost doubling. Although yields of local varieties of barley, teff and wheat are less than 1 t ha1, they have declined between 1991 and 1999, while yields of improved varieties of wheat has increased by about 40 per cent. The analysis also investigated other factors that may affect these responses and outcomes, including indicators of agricultural potential, access to markets, population pressure, credit and extension programmes and adult literacy. Agricultural potential, access to markets and population pressure are hypothesized to be especially important in determining comparative advantages (Pender et al., 1999). We estimate the econometric model given by: yv a b1 x1 b2 x2 czv ev v v 1

where yv is the proportion of farmers in village v that have invested in land conservation and improvement since 1991, undertaken land management practices or used purchased inputs in 1999 or yv is the average crop yield in village v in 1999; x1 is a dummy variable equal to one if there was a land redistribution in village v in 1997 or 1998; v x2 is a dummy variable equal to one if land redistribution is expected in village v in the future; zv is a vector of v observed factors that affect y; and ev are unobserved factors that affect y. Table I also shows a description of the explanatory variables, and their means and standard errors. About 45 per cent of the villages have had a land redistribution in 1997 or 1998 and slightly more than 80 per cent expect a redistribution in the future. Agricultural potential is measured by average annual rainfall (with a mean of 1218 mm), altitude (2182 m.a.s.l.), proportion of area irrigated in 1999 (0 02 per cent) and proportion of farmland with good soil (38 per cent). Access to markets is measured by distance to the wereda town (37 km). The other // / factors are population pressure, which is measured by the number of households per hectare in 1991 (35) and proportion of landless households in 1999 (17 per cent), proportion of households obtaining credit and associated

Copyright # 2001 John Wiley & Sons, Ltd.

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

IMPACTS OF LAND REDISTRIBUTION

559

Table I. Summary statistics of data: number of observations (N), means, and standard errors Variable Dependent variables Proportion of households investing since 1991: Stone terrace Soil bund Check dam Drainage ditch Irrigation canal Waterway Tree planting Live barriers Proportion of households practising: Burning to clear eld Traditional fallow Improved fallow Contour ploughing Mulching Manuring Composting Ploughing in crop residues Soil burning Minimum tillage Proportion of households using: Fertilizer Pesticides Herbicides Improved seed Average cereal yield (100 kg ha1) Local varieties of barley Local varieties of teff Local varieties of wheat Improved varieties of wheat Explanatory variables If land redistribution since 1991 If land redistribution in 1997 or 1998 If expect land redistribution in the future Average annual rainfall (1000 mm) Altitude (1000 metres above sea level (m.a.s.l.)) Proportion of area irrigated Proportion of land with good soil Distance (100 km) to wereda town // / 100 households ha1 Proportion of landless households Proportion of households receiving credit and associated extension from: Bureau of Agriculture (BoA) Amhara Credit and Savings Institution (ACSI) Other formal sources Proportion of adult literates N 1991 1999

94 94 94 94 93 94 94 94 94 92 94 94 94 94 94 94 94 94 93 94 94 93 72 76 56 41 98 98 98 98 98 98 98 98 98 98 98 98 98 98

NA NA NA NA NA NA NA NA 0 365 0 165 0 075 0 918 0 068 0 335 0 021 0 478 0 051 0 260 0 246 0 164 0 040 0 018 9 872 7 547 8 129 7 967 (0 075) (0 043) (0 033) (0 043) (0 035) (0 061) (0 015) (0 077) (0 032) (0 057) (0 056) (0 050) (0 034) (0 010) (1 509) (1 127) (0 748) (1 019)

0 387 0 157 0 273 0 776 0 113 0 549 0 096 0 440 0 336 0 050 0 039 0 934 0 069 0 369 0 056 0 459 0 005 0 231 0 519 0 191 0 142 0 341 7 427 6 582 6 302 12 55 0 488 0 447 0 814 1 218 2 182 0 002 0 380 0 374 0 453 0 169 0 254 0 089 0 189 0 533

(0 061) (0 048) (0 048) (0 049) (0 049) (0 081) (0 035) (0 069) (0 079) (0 020) (0 027) (0 033) (0 036) (0 060) (0 029) (0 079) (0 005) (0 054) (0 061) (0 046) (0 046) (0 063) (1 432) (0 710) (1 069) (1 895) (0 067) (0 057) (0 057) (0 031) (0 081) (0 001) (0 045) (0 057) (0 063) (0 026) (0 060) (0 030) (0 072) (0 035)

NA NA NA 1 218 (0 031) 2 182 (0 081) 0 001 (0 001) 0 533 (0 039) 0 374 (0 057) 0 351 (0 054) 0 206 (0 033) 0 055 (0 030) NA 0 001 (0 001) 0 388 (0 035)

Notes: Standard errors are in brackets. Sample means and standard errors are adjusted for stratication, weighting and clustering of sample. NA Not available.

Copyright # 2001 John Wiley & Sons, Ltd.

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

560

S. BENIN AND J. PENDER

extension in 1999 from the Bureau of Agriculture (BoA; 25 per cent), Amhara Credit and Savings Institution (ACSI; 9 per cent), and other formal sources (19 per cent), and proportion of adult literates in 1999 (53 per cent). An econometric problem to address in this model is that land investment and management and input-use dependent variables are censored, since they are based on proportions data. For example, if the proportion of households using fertilizer was either 0 or 1, the dependent variable was left or right censored, respectively. We therefore used a maximum likelihood censored regression model (or `two-limit tobit model') to estimate the model specied in Equation (1) for these three types of dependent variables, taking into account both left and right censoring. Ordinary least squares was used to estimate the model for cereal yields. There is also a conceptual problem with the model as specied in Equation (1) in the sense that it may bias the estimate of the impact of recent land redistribution in 1997 and 1998. That is, Equation (1) does not control for dynamic processes such as changes in agricultural potential, market access, population pressure, access to credit and extension, and education that may be important in inuencing investment in land management and input use, which may in turn affect agricultural productivity. This is because these dynamic processes inuence the awareness, availability, risks, costs, and benets associated with investment in land improvement and input use. Therefore, we also estimate the rst difference model given in Equation (2):3 yv a2 a1 bxv c1 z1 z1 c2 c2 z2 ev2 ev1 v2 v1 2 1 v 2

where yv represents the dependent variable in village v, xv is a dummy variable equal to one if there has been land redistribution in village v since 1991, z1 is a vector of observed time-varying factors affecting y, z2 is a vector of vt v observed xed factors affecting y, and evt are unobservable factors affecting y. The observed xed factors, z2 , v will have an impact only if the marginal effect of such factors has changed over time. The dependent variables in this model are: (1) proportion of farmers undertaking land investments since 1991; (2) change between 1999 and 1991 in proportion of farmers adopting land management practices; and (3) change between 1999 and 1991 in proportion of farmers using of purchased inputs.4 The time-varying factors are changes between 1999 and 1991 in the proportions of area irrigated, farmland with good soil, landless households, and households obtaining credit and extension from BoA, ACSI5 and other formal sources, and households per hectare and adult literacy. The xed factors are average annual rainfall, altitude, and distance to the wereda town. // / The problem of censoring associated with the dependent variables, as discussed with the previous model, also applies here and so maximum likelihood censored regression is used to estimate Equation (2). Another econometric problem to address in Equation (2) is that the time-varying explanatory variables may be endogenous. Population growth and changes in participation in credit and extension programs, proportion of area irrigated, proportion of farmland with good soil, and adult literacy may respond to or be affected by changing opportunities in agriculture and changes in land management and productivity. We therefore tested for exogeneity of those potentially endogenous explanatory variables using a Hausman test (Hausman, 1978; Greene, 1993).6 We failed to reject exogeneity of those explanatory variables in the regressions, except the ones for changes in proportion of households undertaking drainage ditches and using herbicides. Nevertheless, we report the robustness of the signicant coefcients to using predicted values of those potentially endogenous variables.
The rst difference model eliminates unobservable xed factors as a source of omitted variable bias. The change in average cereal yields between 1999 and 1991 could not be estimated due to insufcient number of observations, as cereals were not grown in both years in some of the villages. 5 ACSI started operating in the Amhara Region in 1995 and so the proportion of households participating in 1999 represents the change. 6 The instrumental variables used to predict the potentially endogenous explanatory variables, in addition to the exogenous variables in the regressions, include the values of each those endogenous variables in 1991, walking time to nearest bus station in 1991 and change since 1991, walking time to the nearest grain mill in 1991 and change since 1991, walking time to the nearest primary school in 1991 and change since 1991, walking time to the nearest all weather road in 1991 and change since 1991. The instruments predicted most of the potentially endogenous variables fairly well: R2 0 87 for change in proportion of landless households, 0 65 for change in proportion of households obtaining credit from BoA, 0 61 for change in household density, 0 55 for change in proportion of households obtaining credit from other formal sources, 0 40 for proportion of households obtaining credit from ACSI, 0 32 for change in proportion of farmlands with good soil, 0 32 for change in adult literacy and 0 22 for change in proportion of area irrigated.
4 3

Copyright # 2001 John Wiley & Sons, Ltd.

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

IMPACTS OF LAND REDISTRIBUTION

561

RESULTS The results of both model estimations are corrected for sample stratication and clustering. We present only those results in which the overall estimated model was statistically signicant at the 10 per cent level of signicance. Tables II and III show results of the cross-section model given in Equation 1, while Tables IV and V show results of the rst-difference model given in Equation 2. Investments in land improvement We nd that land redistribution and expectations of future land redistribution have a statistically insignicant and usually small association with land investments since 1991 (Tables II and IV). Of the other explanatory variables, we nd that increased use of credit and extension from other formal sources is associated with higher and increasing investments in drainage ditches. Increased use of credit from the ACSI, however, has mixed impacts: it is associated with lower investments in irrigation canals, but increasing investments in waterways. Increase in the proportion of poor soils is associated with more investments in stone terraces, drainage ditches and waterways. This is probably because the returns to soil and water conservation are higher on more degraded soils (Herweg, 1993). As expected, increase in the proportion of area irrigated increases investment in irrigation canals and waterways. Poor market access (increase in distance to the wereda town), however, is associated with lower investment in // / irrigation canals and waterways. This is probably because the protability (through increased output) of investment in irrigation is reduced by the high transaction costs associated with distant markets. More densely populated areas are associated with higher investments in live barriers, but lower investments in drainage ditches. In addition, increase in proportion of landless households is associated with lower and declining investments in drainage ditches. Landless households also tend to be renters and sharecroppers and so they may not have incentive to invest in land improvement, especially given the short leases (usually one season long) as exist in most of Ethiopia. Land management practices With the exception of traditional fallow, we did not nd any statistically signicant association between land redistribution or expectations of future redistribution and the proportion of farmers undertaking land management practices. The reduction in the proportion of farmers practising traditional fallow where there has been a recent redistribution (Table II), suggests that farmers may be farming more intensively to compensate for the reduction in plot sizes (usually less than 1 ha per household), as farmers cannot afford to fallow part of their already small elds. Of the other explanatory variables, we nd that the practice of traditional fallow increases with altitude and declines with rainfall (Table II), while the practice of burning to clear elds increases with altitude (Table IV). Higher altitude areas tend to be more degraded and may require more fallow to replenish lost nutrients, due to more burning and/or lower use of fertility-improving technologies. Participation in credit and extension programmes has mixed impacts. Higher participation in credit from the ACSI is associated with lower practice of traditional fallow, due probably to the ACSI offering credit in the form of fertilizer, use of which reduces the need for fallowing. Higher participation in credit from other formal sources, on the other hand, is associated with higher practice of traditional fallow. This is probably because there are several non-governmental organizations in the region providing credit for small enterprises that generate non-farm income, which may allow more farmers to fallow their elds. We also nd that increase in participation in credit and extension from the BoA and ACSI is associated with more practice of burning to clear elds. With increase in use of credit in the form of fertilizers, more farmers can afford to burn their elds as a quicker and cheaper method of clearing, and then utilize the fertilizers to replenish the nutrients lost through burning. Higher incidence of landless households is associated with higher and increasing practice of traditional fallow, but associated with declining practice of crop rotation. The landless also tend to be renters and sharecroppers who may mine the soil, requiring more fallow to replenish lost nutrients. Furthermore, with short leases of one season long, there is little incentive and opportunity for practising crop rotation. Increase in adult literacy is associated
Copyright # 2001 John Wiley & Sons, Ltd. LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

562

Table II. Regression results of the proportion of households undertaking land investments since 1991 and practising traditional fallow in 1999 in the highlands of Amhara Region Stone terraces 0 3663 0 0229 0 9759 0 0354 0 7652 1 0533*** 0 0583 0 3170 0 2309 0 3681 0 0665 0 0103 0 4609 2 1043*** 2 72** 52 25 17 94 2 23** 18 16 60 94 0 5509 0 1035 3 7158*** 1 6538 1 9234 0 0032 2 4381* 0 1544 0 6853 0 5932 3 10*** 22 59 12 93 0 5487 0 7261 2 3301 0 6575 5 5610 2 1236** 0 6087 0 7547* 3 2133*** 0 2926 0 2968 0 2218 0 1258 55 7670*** 0 1202 0 5548** 0 0201 0 0233 Drainage ditches Irrigation canals Live barriers 0 0539 0 3382 1 2157 0 3907* 0 4559 0 2729 0 3099 0 6860*** 0 5857 0 0484 0 3363 0 3404 0 0106 0 8000 2 61** 37 24 33 94 Traditional fallow 4 3456*** 0 4956 2 8608** 0 7329*** 6 0512 0 9224 0 3012 0 2683 2 3019*** 2 4718 6 7829* 1 2285** 0 5564 0 3394 4 04*** 10 78 4 92

Copyright # 2001 John Wiley & Sons, Ltd.

Explanatory variable

If land redistribution in 1997 or 1998 If expect land redistribution in the future Annual rainfall (1000 mm) Altitude (1000 m.a.s.l) Proportion of area irrigated in 1999 Proportion of land with good soil in 1999 Distance (100 km) to wereda town // / 100 households ha1 in 1991 Proportion of landless households in 1999 Proportion of households in 1999 receiving credit and associated extension from: Bureau of Agriculture (BoA) Amhara Credit and Savings Institution (ACSI) Other formal sources Proportion of adult literates in 1999 Intercept

S. BENIN AND J. PENDER

F Uncensored observations Left-censored observations Right-censored observations Total observations

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

Notes: Coefcients and standard errors are adjusted for stratication, weighting and clustering of sample. *Statistically signicant at the 10% level; **statistically signicant at the 5% level; ***statistically signicant at the 1% level.

Table III. Regression results of the proportion of households using purchased inputs in 1999 and average cereal yields (100 kg ha1) in 1999 in the highlands of Amhara Region Fertilizers 0 4990*** 0 0002 0 5901*** 0 0696 5 7691 0 1119 0 0909 0 1579* 0 5076* 0 6183*** 0 7438*** 0 6435*** 0 7679*** 0 3404 9 76*** 61 13 19 93 30 51 13 94 53 25 15 93 4 06*** 6 49*** 0 4751* 0 7939 1 4565 0 9323*** 0 4146 0 0348 0 9043*** 0 7266*** 0 0721 0 1489 3 1056* 4 4021 3 0146* 0 5081 8 8989* 30 82*** 0 73 70 6 1616*** 0 6375 2 6232 0 2544 4 6839 16 66** 0 77 53 0 2292 0 1277 0 2470 0 1004 1 3798** 0 5067 0 6035** 0 0854 24 4110* 6 9681 0 3406 0 3256 0 5889 0 1754* 0 0280 0 2295*** 2 2316*** 0 1906 4 2782* 4 0881*** 13 8790*** 1 2004 166 5900 5 6830*** 0 0899 8 0390*** 3 1299 3 6424*** 0 8567 8 0569 0 6495 358 3500 4 3948* 0 0997 1 5155 3 1865 Pesticides Improved seeds Local barley Local wheat Local teff Improved wheat

Copyright # 2001 John Wiley & Sons, Ltd.

Explanatory variable

3 3351*** 5 9116*** 1 2362 12 1290*** 0 0414 36 8720*** 0 7090 3 0108 10 3370 134 6400 1 9026 10 1780* 0 2162 1 5547 3 1070*** 7 0553*** 1 0942 7 0550 3 4812*** 1 1109 3 4949*** 1 2662 3 4329 15 58*** 0 68 74 1 4537 3 7549 1 6295 6 1071 11 4210 15 94*** 0 77 41

If land redistribution in 1997 or 1998 If expect land redistribution in the future Annual rainfall (1000 mm) Altitude (1000 m.a.s.l) Proportion of area irrigated in 1999 Proportion of land with good soil in 1999 Distance (100 km) to wereda town // / 100 households ha1 in 1991 Proportion of landless households in 1999 Proportion of households in 1999 receiving credit and associated extension from: Bureau of Agriculture (BoA) Amhara Credit and Savings Institution (ACSI) Other formal sources Proportion of adult literates in 1999 Intercept

IMPACTS OF LAND REDISTRIBUTION

F R2 Uncensored observations Left-censored observations Right-censored observations Total observations

Notes: Coefcients and standard errors are adjusted for stratication, weighting and clustering of sample. *Statistically signicant at the 10% level; **statistically signicant at the 5% level; ***statistically signicant at the 1% level.

563

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

564

Table IV. Regression results of changes between 1999 and 1991 in the proportion of households undertaking land investments and land management practices in the highlands of Amhara Region Drainage ditches 0 0588 2 3969 0 4107 307 8300*** 2 0329 1 2034 0 2866 2 2026* 0 1853 0 8221 4 2091** 3 2162 1 8479 4 51*** 16 15 55 86 1 90* 18 34 34 86 0 9965 4 4849*** 1 3322 1 1446 5 5691 0 0981***R 0 0963* 0 0290 0 1342 0 0463 3 59*** 86 0 0 86 0 3139 0 6498 2 1741**R 277 4300*R 6 3835*** 2 4818**R 3 9424 0 3353 0 0079 0 0832 0 0353* 7 9136 0 1033 0 0039 0 0956 0 1668 Waterways Burning to clear eld Traditional fallow 0 0029 0 1048 0 0325 1 8864 0 1154 0 0591 0 5133 0 5879***R 0 1115 0 1832 0 0573 0 4488* 0 1091 2 44** 83 2 0 85 Crop rotation 0 0027 0 0249 0 0038 2 8411 0 1519 0 0339 0 2413 0 2125* 0 0218 0 0072 0 0207 0 1253 0 0151 1 92* 86 0 0 86

Copyright # 2001 John Wiley & Sons, Ltd.

Explanatory variable

S. BENIN AND J. PENDER

If land redistribution since 1991 Annual rainfall (1000 mm) Altitude (1000 m.a.s.l) Change in proportion of area irrigated Change in proportion of land with good soil Distance (100 km) to wereda town // / Change in 100 households ha1 Change in proportion of landless households Change in proportion of households receiving credit and associated extension from: Bureau of Agriculture (BoA) Amhara Credit and Savings Institution (ACSI) Other formal sources Change in proportion of adult literates Intercept

F Uncensored observations Left-censored observations Right-censored observations Total observations

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

Notes: Change in explanatory variable refers to difference between 1999 and 1991 levels. Coefcients and standard errors area adjusted for stratication, weighting and clustering of sample. *Statistically signicant at the 10% level; **statistically signicant at the 5% level; ***statistically signicant at the 1% level. Rmeans coefcient of same sign and signicant at 10% level when predicted values used for changes in proportions of area irrigated, farmlands with good soil, landless households, households obtaining credit and extension from ACSI, BoA, and other formal sources, adult literates and change in household density.

IMPACTS OF LAND REDISTRIBUTION

565

Table V. Regression results of the changes between 1999 and 1991 in the proportion of households using purchased inputs in the highlands of Amhara Region Explanatory variable If land redistribution since 1991 Annual rainfall (1000 mm) Altitude (1000 m.a.s.l) Change in proportion of area irrigated Change in proportion of land with good soil Distance (100 km) to wereda town // / Change in 100 households ha1 Change in proportion of landless households Change in proportion of households receiving credit and associated extension from: Bureau of Agriculture (BoA) Amhara Credit and Savings Institution (ACSI) Other formal sources Change in proportion of adult literates Intercept F Uncensored observations Left-censored observations Right-censored observations Total observations Fertilizers 0 2763* 1 0309***R 0 0848 32 8710* 0 1892 0 0606 0 6766 0 1098 Pesticides 0 1173 0 2863 0 0461 28 8540***R 0 0049 0 0389 0 5353 0 0556 Herbicides Improved seeds

0 1991***R 0 0177 0 0962 0 1641 0 0378 0 0930 1 2785 65 5570*** 0 0774 0 3957 0 0994 0 1315 0 2492 0 7160** 0 2296** 0 1902 0 5291***R 0 1963 0 9303***R 0 1690 0 2166 70 68*** 75 0 10 85

0 1519 0 0754 0 0818 0 0677 0 0739 0 2667*R R 0 3055*** 0 2323*** 0 0787* 0 2989 0 1148 0 1998 1 0639***R 0 5059*R 0 0871 3 62*** 82 0 3 85 2 28** 82 3 1 86 5 01*** 86 0 0 70

Notes: Change in explanatory variable refers to difference between 1999 and 1991 levels. Coefcients and standard errors are adjusted for stratication, weighting and clustering of sample. *Statistically signicant at the 10% level; **statistically signicant at the 5% level; ***statistically signicant at the 1% level. Rmeans coefcient of same sign and signicant at 10% level when predicted values used for changes in proportions of area irrigated, farmlands with good soil, landless households, households obtaining credit and extension from ACSI, BoA, and other formal sources, adult literates, and change in household density.

with declining practice of traditional fallow. It may be that better educated people are more likely to employ other fertility-improving technologies (i.e. fertilizer) and engage in more intensive agriculture. This hypothesis is discussed further in the next subsection relating to use of purchased inputs. Use of purchased inputs In contrast to the limited impacts on land investment and management practices, increased use of fertilizers in 1999 is strongly associated with recent land redistribution in 1997 or 1998 (Table III). It might be hypothesized that this is because the credit and extension programmes focused more attention on areas where the land redistribution occurred. However, we nd a negative and relatively small correlation between the presence of extension and credit programmes and where land redistribution has occurred. Thus, it appears that land redistribution has contributed to greater input intensity by increasing access to land among households with greater proclivity or ability to use inputs. This may be because the younger farmers who gain access through redistribution are more educated or have access to off-farm sources of income with which to nance input purchases. Increased use of fertilizers is also consistent with our nding of a reduction in the practice of fallow where there has been recent land redistribution. With more intensive and continuous cropping, there is the need to maintain (or improve) fertility through fertilizers and other fertility-improving practices. We do not nd any statistically signicant association between expectations of future land redistribution and use of purchased inputs. We also nd that land redistribution since 1991 has a statistically signicant positive impact on change in proportion of households using fertilizers and herbicides (Table IV). This nding reinforces the discussion above. Of the other explanatory variables, we nd that higher participation (and growth) in credit and extension programmes have statistically signicant positive association with use of most of the purchased inputs. The only
Copyright # 2001 John Wiley & Sons, Ltd. LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

566

S. BENIN AND J. PENDER

exception is the use of herbicides, which declines with increased participation in credit and extension from other formal sources, although the coefcient measuring this impact is not robust (Table V). As expected, increase in area irrigated is associated with higher use of pesticides, and increasing use of fertilizers, pesticides, and improved seeds. Higher altitude areas are associated with lower use of pesticides, probably because of the lower incidence of pests at higher altitudes. Rainfall and population density have mixed impacts on the use of purchased inputs. While higher rainfall is associated with higher use of pesticides, it is associated with lower and declining use of fertilizers. The latter may be due to the active promotion by the government of fertilizers in the lower rainfall areas under the previous extension system within the last decade. More densely populated areas are associated with higher use of fertilizers and improved seeds (Table III), but population growth is associated with a decline in the use of improved seeds, although the coefcient measuring this impact is not robust (Table V). More densely populated areas also tend to have smaller plots and, thus, farmers here may need to farm more intensively. On the other hand, land±labour ratios may be low enough to induce more labour-intensive, modern cultural practices (Boserup, 1965). We also nd that areas with higher proportion of landless households are associated with lower use of fertilizers, but higher use of pesticides. As mentioned earlier, landless households also tend to be renters and sharecroppers and so they may have little incentive (given short leases) to use fertilizers, which may have long-term fertility improving effects of which they may not be able to reap the benets. Cereal yields Consistent with the positive impacts of land redistribution on fertilizer use, we also nd that yields of barley, wheat and teff in 1999 are higher in communities where there has been a recent redistribution (Table III). Yields of local varieties of these crops are about 400 kg ha1 higher and yields of improved wheat are about 600 kg ha1 higher where there has been a recent redistribution. We also nd that expectations of future redistribution have a positive association with yields of the local cereal varieties (although it is statistically signicant in the case of barley only), but a negative association with yields of improved varieties of wheat. The positive association may indicate soil mining on those plots that farmers anticipate will be redistributed in the future. On the other hand, it may indicate that farmers increase their productivity if they believe that higher yields will prevent their plots from being redistributed. Further research is needed to test these hypotheses. The negative association between expectations of future redistribution and improved wheat yield is not apparent. Therefore, while the impact of redistribution on the yields of the local varieties is uniformly positive, that on the improved varieties is not and depends on expectations about the future.7 Of the other explanatory variables, we nd that areas with higher rainfall, better soils and higher participation in credit and extension programmes have higher yields in one or more of the cereals. While more densely populated areas are associated with higher yields of local varieties of teff, they are associated with lower yields of improved varieties of wheat. The model for the change in average cereal yields between 1999 and 1991 could not be estimated due to insufcient number of observations, as the cereals were not grown in both years for some of the villages. CONCLUSIONS AND IMPLICATIONS Overall, these results suggest that the recent land redistribution in Amhara Region has had a positive impact on land productivity, at least in the short term, by increasing access to land of farmers who are more interested or able to use purchased inputs such as fertilizers. This does not mean that land redistribution must be continued and used as a tool to improve access to land, as the longer term impacts of such redistributions depend upon how these may affect farmers' perceptions of tenure security and incentives to invest in land improvement. Nevertheless, it is difcult to continue to use redistribution as a tool to address landlessness because of the very small size of farm
7 We estimated a probit model of expecting land redistribution in future as a function of recent redistribution (with a positive and signicant coefcient), number of redistributions in the past (negative), population density (positive), proportion of landless households (positive and signicant), and proportion of households that have their land registered (negative).

Copyright # 2001 John Wiley & Sons, Ltd.

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

IMPACTS OF LAND REDISTRIBUTION

567

holdings in the Ethiopian highlands. Improving access to farmland through other means such as development of land rental markets may present a more sustainable and only strategy to improving agricultural productivity, especially since land sales are prohibited and other transfers restricted. For the most part, however, our results do not show much effect of land redistribution or expectations of future redistribution on land improvement. In addition, almost all farmers expect future redistributions. Although the regional government is considering a policy that may end land redistributions, we nd that respondents' expectations of future redistributions are also signicantly affected by landlessness, which also tend to negatively affect use of several of the land-improving technologies analysed here. Thus, to the extent that investments in land improvement are necessary for conservation purposes, it appears that as long as landlessness is prevalent, the intended policy change to end redistributions is unlikely to have a substantial impact on reducing land degradation in this region of Ethiopia. However, since landless households tend to be renters and sharecroppers who have little incentive in investing in rented land, improving land rental markets, especially to encourage longer leases, will have a positive impact on land management by reducing soil mining and by increasing use of fertilizers and crop rotation. Furthermore, given that other factors such as participation in credit and extension programmes were more important determinants of land investment and management, these factors may present better strategies for reducing land degradation in the Ethiopian highlands. For example, improving credit for non-farm income generating activities will allow more farmers to fallow their (marginal) lands or provide them with cash to purchase and use fertilizers. acknowledgement The authors wish to acknowledge the Norwegian Ministry of Foreign Affairs for funding this research. references
Amhara National Regional State Council. 2000. The administration and use of rural land in Amhara Region. Draft policy document. ANRSC: Bahir Dar. Barrows R, Roth M. 1990. Land tenure and investment in African Agriculture: theory and practice. Journal of Modern African Studies 28(2): 265±297. Berry RA, Cline WR. 1979. Agrarian Structure and Productivity in Developing Countries. John Hopkins University Press: Baltimore, MD. Besley T. 1995. Property rights and investment incentives: theory and evidence from Ghana. Journal of Political Economy 103(5): 913±937. Bezuayehu T, Gezahen A, Yigezu A, Paulos D, Jabbar MA. 2000. Nature and causes of land degradation in the Oromiya region: a review of literature. Paper presented at the workshop on Policies for Sustainable Land Management in the Highlands of Ethiopia, May 22±23, 2000, International Livestock Research Institute: Addis Ababa. Boserup E. 1965. The Conditions of Agricultural Growth: The Economics of Agrarian Change under Population Pressure. Aldine: New York. Feder G, Feeny D. 1993. The theory of land tenure and property rights. In The Economics of Rural Organization: Theory, Practice and Policy, Hoff K, Braverman A, Stiglitz JE (eds). The World Bank: Washington, DC. Fistum H, Pender J, Nega G. 1999. Land degradation in the highlands of Tigray and strategies for sustainable land management. Socioeconomic and Policy Research Working Paper No. 25. The International Livestock Research Institute: Addis Ababa. Gavian S, Fafchamps M. 1996. Land tenure and allocative efciency in Niger. American Journal of Agricultural Economics 78(2): 460±471. Greene WH. 1993. Econometric Analysis. Macmillan: New York. Hausman J. 1978. Specication tests in econometrics. Econometrica 46: 1251±1271. Herweg K. 1993. Problems of acceptance and adoption of soil conservation in Ethiopia. Topics in Applied Resource Management 3: 391±411. Hurni H. 1988. Degradation and conservation of resources in the Ethiopian highlands. Mountain Research and Development 8: 123±130. Lakew D, Menale K, Benin S, Pender J. 2000. Land degradation and strategies for sustainable development in the Ethiopian highlands: Amhara region. Socioeconomic and Policy Research Working Paper No. 32. The International Livestock Research Institute: Nairobi. Migot-Adholla SE, Hazell P, Blarel B, Place F. 1991. Indigenous land rights systems in sub-Saharan Africa: a constraint on productivity? World Bank Economic Review 5(1): 155±175. Pender J, Place F, Ehui S. 1999. Strategies for sustainable agricultural development in the east African highlands. Environment and Production Technology Department Discussion Paper No. 41. International Food Policy Research Institute: Washington, DC. Pender J, Kerr J. 1999. The effect of land sales restrictions: evidence from south India. Agricultural Economics 21: 279±294. Place F, Hazell P. 1993. Productivity effects of indigenous land tenure systems in sub-Saharan Africa. American Journal of Agricultural Economics 74(3): 360±373.

Copyright # 2001 John Wiley & Sons, Ltd.

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)

568

S. BENIN AND J. PENDER

Place F, Swallow B. 2000. Assessing the relationships between property rights and technology adoption in smallholder agriculture: a review of issues and empirical methods. CAPRi Working Paper No. 2. International Food Policy Research Institute: Washington, DC. Prosterman RL, Riedinger JM. 1987. Land Reform and Democratic Development. John Hopkins University Press: Baltimore, MD. Quisumbing AR, Payongayong E, Aidoo JB, Otsuka K. 1999. Women's land rights in the transition to individualized ownership: implications for the management of tree resources in Western Ghana. Food Consumption and Nutrition Department Discussion Paper No. 58. International Food Policy Research Institute: Washington, DC. Snyder KA. 1996. Agrarian change and land use strategies among Iraqw farmers in Northern Tanzania. Human Ecology 24(3): 315±340.

Copyright # 2001 John Wiley & Sons, Ltd.

LAND DEGRADATION & DEVELOPMENT, 12: 555±568 (2001)



更多相关文章:
Biodiversity impacts ecosystem productivity as much as ....pdf
Biodiversity impacts ecosystem productivity as much as resources, disturbance, or herbivory_生物学_自然科学_专业资料。Biodiversity impacts ecosystem productivity ...
...of Technology Change on Durum Wheat Productivity: A Cross ....pdf
Impact Assessment of Technology Change on Durum Wheat Productivity: A Cross Sectional An_电子/电路_工程科技_专业资料。Journal ofAgricultural Science and ...
...Impacts on Vegetation Distribution and Productivity_论文_....pdf
Climate Variability in Niger: Potential Impacts on Vegetation Distribution and Productivity_环境科学/食品科学_工程科技_专业资料。JunlfninnaSic n niergB12 )95...
Software reuse and its impact on Productivity, Quality and ....pdf
Software reuse and its impact on Productivity, Quality and Time-to-market_... which allows the project management to assess how well the project is ...
土地资源管理专业英语词汇.doc
The major effect of land use on land cover ...Management Bureau of State Land Supervision 市国土...productivity land suitability land efficiency land ...
Multiple Impacts of Land-UseCoverChange.pdf
Chapter 4 Multiple Impacts of Land-Use/Cover ... productivity increases through land-use ...redistribution is restricted by the geographical ...
...Management Practices on Turnover, Productivity, ....pdf
AMJ1995 The Impact of Human Resource Management Practices on Turnover, Productivity, and CORPORATE f_经管营销_专业资料。AMJ 年度最佳论文...
The Impacts of Governance and Education on Agricultural ....pdf
Management Conference The Impacts of Governance and...Governance affects agricultural productivity through ...(land): It is estimated by the arable land ...
...development policies and sustainable land use in....pdf
(Lps/year) Determinants of land management ...effect on crop productivity or per capita income,...Land redistribution programs seeking to increase ...
Chapter 13 Human Impacts on Natural Systems.ppt
Climate prediction (Maldives - 80% of land are less is not an exact ...? Impact on developing countries is high but uncertain. Global productivity ...
...of Management Practices on Retail Store Productivity.doc
Do any models currently exist, that model the impact on management practices on productivity in the retail sector? 3. What modelling approaches are ...
...of Climate and Environment Changes in China (II) Impacts, ....pdf
2.3 Impacts on land degradation Because of the ...The future trend of land salinization shows that:...productivity and distribution of agriculture, ...
Land use and cover in ecological economics.pdf
HUMAN IMPACTS ON LAND... 23页 免费 Land Use...productivity can be defined in terms of land: ...redistributions in the age-class structure of the...
...and Aggregate Industry Productivity.pdf
The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity_经管营销_专业资料。HO理论The Impact of Trade on Intra-Industry Real...
SUSTAINABLE LAND USE CONVERGENCE IN.doc
(2) maintaining resource productivity, and (3) ...319-327 need for integrated resource management is...impacts of alternative land use policies on ...
Container terminal planning - improving system productivity ....pdf
productivity to service larger container vessels__...The impacts of larger vessel size on terminal ...No longer does terminal management have to ...
...Effects of Code Coupling on Team Dynamics and Productivity....pdf
into the Effects of Code Coupling on Team Dynamics and Productivity_专业资料...Keywords Information Hiding, Dependency Management, Coupling, Large Scale ...
...impact on development and productivity of pepper....pdf
Biochar impact on development and productivity of... S (eds) Biochar for environmental management. ...greenhouse gases through emissions from land-use ...
...Studies of the Impact of Broadband on Productivity_图文_....ppt
Designing and Assessing Studies of the Impact of Broadband on Productivity_经济学_高等教育_教育专区。Designing and Assessing Studies of the Impact of ...
Automated Driving Impacts on the Vehicle Archite....pdf
Automated Driving Impacts on the Vehicle ... parking management, integration of the consumer ...productivity; time in transit becomes more ...
更多相关标签:

All rights reserved Powered by 甜梦文库 9512.net

copyright ©right 2010-2021。
甜梦文库内容来自网络,如有侵犯请联系客服。zhit325@126.com|网站地图