Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Agrarian reform and small-farmer welfare: Evidence from four Mexican communities

Billie R. DeWalt, Kathleen M. DeWalt, José Carlos Escudero, and David Barkin


Most scholars would agree that adequate nutrition for normal growth and development is one of the most basic of all human needs, and that it depends on several critical conditions being met. First and foremost, of course, adequate amounts of food must be available to individuals and families. For much of human history, food was obtained primarily by the gathering, producing, and harvesting of foodstuffs from a local resource base. Self-provisioning depended on people's having access to the natural resources from which foods could be hunted and gathered, or to land on which to produce the food required.

Without access to such natural resources, people can obtain adequate amounts of food if they have economic resources or purchasing power. For most individuals this means having employment from which they can earn sufficient income to buy sufficient food.

While governments can and often do provide subsidized food to those for whom sufficient food is not available, this alternative is less desirable than providing people with the means to generate their own subsistence needs. Thus, to meet nutritional and other needs, development efforts should have the goal of providing people with access either to natural resources or to remunerative employment opportunities.

One of the most popular strategies that has been advocated in Latin América for providing the poor with the resources with which to satisfy their basic needs is land reform. Land reform achieved official recognition in the Charter of the Alliance for Progress, which was signed in Punta del Este in 1961 [1, 2]. A major world conference in 1979 concluded: "Equitable distribution and efficient use of land . . . are indispensable for rural development, for the mobilization of human resources, and for increased production for the alleviation of poverty" [3]. While land reform has long been advocated by socialists and liberals, it has recently been advocated as a part of conservative development strategies [4].

The Latin American country with the longest experience with agrarian reform is Mexico. After its bloody revolution, substantial agrarian reform was effected, with almost 50 per cent of the land eventually coming under the control of ejidos, communities created to receive redistributed lands. Most of this land was redistributed during the presidency of Lazaro Cardenas (1900-1936), who saw the ejido as the basis on which a modern Mexico could be built. As one of his officials put it in 1935:

We have dreamt of a Mexico of ejidos and small industrial communities, electrified, with sanitation, in which goods will be produced for the purpose of satisfying the needs of the people, in which machinery will be employed to relieve man from heavy toil, and not for so-called overproduction. [5]

About this same time, Simpson wrote a thoughtful book examining the problems and prospects of the system that asked whether the ejido might be Mexico's "way out" [6].

Four decades later there is substantial debate in Mexico about whether the ejido system has outlived its usefulness. Although our research was not designed to address this issue directly, we wanted to determine whether the system has been beneficial in terms of improving the welfare of poor rural communities. This article reports the results of our investigations into how access to natural resources (especially land), alternative income-producing opportunities employment), and general health conditions are related to the nutritional status of infants and infant mortality in ejidos in four regions of Mexico. This research was begun in 1984 by investigators from the University of Kentucky and the Universidad Autonoma Metropolitana-Xochimilco as a part of the International Sorghum/Millet Collaborative Research Support Program (INTSORMIL). The main purpose was to document the effect on rural Mexican communities of a change in crops from maize to sorghum. Sorghum, which was virtually unknown in the country until the late 1950s, had become the second largest crop planted by 1984 [7, 8].

The communities

The four communities chosen for study are located in the states of Michoacán, Morelos, San Luis Potosí, and Tamaulipas. They were selected for several reasons. First, they are ejido communities in very different ecological regions, yet sorghum cultivation has become quite important in each, and other aspects of agricultural modernization such as credit, mechanization, and so on are prominent. Second, we felt that it was important to study regions that are less well known in the community-study literature in Mexico. All are mestizo communities that are quite representative of the majority culture in the country. Ejidos like these control almost 50 per cent of the land and thus are important components of the agricultural system.

The total population of the communities ranged from a low of 610 in the community in Tamaulipas to over 2,800 in the community in Morelos. The ejidos connected to these communities received their land at different periods of time: in the late 1920s in San Luis Potosí, during the presidency of Lázaro Cárdenas in the 1930s in Morelos, and not until the 1960s in Michoacán and Tamaulipas. Extensive cultivation of sorghum in these communities was begun in the late 1960s or early 1970s. Agricultural mechanization also dates from about this same time; tractors are now available in all four communities to carry out some of the agricultural labour.

Economic characteristics of the communities

The first objective of our research was to determine the access to economic resources of the people in the four communities. We were especially interested in (1) access to land, (2) access to other employment opportunities, (3) the cash income resulting from these activities, and (4) migration patterns that would indicate whether the local people perceived these communities as economically viable.

All of these communities were beneficiaries of the agrarian reform that resulted from the Mexican Revolution, so that there is at least some access to land. Table 1 shows the data on land access among the households sampled. Despite agrarian reform, in all the communities except the one in San Luis Potosí approximately one-third of the households do not have access to land. Landholdings are scarcest in the community in Morelos, where about 44 per cent of the households have access to less than three hectares. In Tamaulipas, in contrast, the 71 per cent of families with access to ejido land each have about 20 hectares, 15 of which are irrigated. Thus, although many households are able to do some farming, land distribution is substantially unequal within and among these different communities.

As we indicated in an earlier report [9] access to land is only one element in the ability to use agricultural resources to provide for one's family. Mexican agrarian history since 1940 has been characterized by a relative lack of support for the ejido and smallfarm sector in terms of credit, technical assistance, infrastructure, and research [10].

One indicator of how successful these communities' agricultural systems are is the portion of income that is derived from agricultural pursuits (table 2). Households depend on a wide range of sources to satisfy their economic needs. Grain crops such as sorghum and maize, which are extensively grown in each of the communities, provide the major part of the reported income only in the community in Tamaulipas. The community in Morelos is least dependent on selling grain crops; sales of livestock provide the most income, with substantial contributions also coming from sales of fruits and vegetables and from wage labour.

TABLE 1. Distribution of land among the sampled farmers in the four communities

Number of hectares controlled Michoacán San Luis
Morelos Tamaulipa
No. % No. % No. % No. %
None (non ejidatarios) 25 30.1 6 10.0 32 33.0 22 29.3
<3 1 1.2 4 6.7 11 11.3 0 0
3.1-5 17 20.5 9 15.0 17 17.5 0 0
5.1-7.9 24 28.9 32 53.3 14 14.4 0 0
8-12 14 16.9 9 15.0 19 20.0 0 0
>12.1 3 3.6 0 0 4 4.1 53 70.7
Total 83 60 97 75

Wage labour is least important in the community in San Luis Potosí; here remittances from people working legally and illegally in the United States are quite important in maintaining the community economically. The data on mean cash income for the households indicate that the communities in San Luis Potosí and Michoacán are quite poor, with estimated annual incomes of only US$617 and US$806 respectively in 1984. In contrast, in the communities in both Morelos and Tamaulipas the households have mean cash incomes of over US$1,800 per year.

TABLE 2. Reported sources of cash income

  Percentage of income
Michoacán (N = 83) San Luis Potosí (N = 60) Morelos (N = 97) Tamaulipas (N = 75)
Grain crops 34.2 33.3 17.9 51.3
Fruits and vegetables 6.3 0 12.2 4.6
Livestock 12.0 14.9 30.0 4.9
Wage labour 24.4 9.5 23.1 19.5
Equipment rental 9.6 0.2 0.3 3.8
Remittances 0. 5 10.5 5.1 1.7
Other 10.9 31.5 11.5 14.5
Cash income (pesos)  
Total 10,702,640 5,918,550 29,037,426 23,801,649
Mean per household 128,947 98,642 299,355 317,355
(US$806) (US$617) (US$1,871) (US$1,983)

These are average figures, and it should be emphasized that they mask considerable variation: many households have cash incomes much smaller than the average figures, and a few have cash incomes substantially higher.

Other indicators were used to establish the relative success of agriculture in these communities. In a direct question about whether the household head earned the greater part of his or her income from agriculture, only in Michoacán and Tamaulipas did more than 50 per cent indicate that they did; in San Luis Potosí and Morelos this figure was less than 20 per cent. In all the communities, more than 60 per cent of the household heads indicated that they had worked outside the community at some point; more than 95 per cent in San Luis Potosí had temporarily migrated from the community to work. Over 80 per cent of the household heads in this community had at one time worked in the United States; the percentages in the other communities were smaller.

The relative success of the agricultural economy may also be gauged by comparing the residences of the children over 15 years old. Agrarian reform may have benefited the older generation by providing access to land, but what has happened to the children and grandchildren of these individuals? A substantial percentage (ranging from 37 to 56 per cent) of these younger persons have found it necessary to leave their community. The community in Tamaulipas, where ejidatarios have access to the largest amount of land, has been able to retain the largest percentage of its children. In San Luis Potosí not only have a majority of the children over the age of 15 left the community, but at the time of the interview almost one-third of them were living in the United States, most as illegal aliens. This percentage was smaller for the other communities. The attractiveness of the United States as a place to derive income is indicated by the fact that 43 per cent of the households in the San Luis Potosí community and 23 per cent in Michoacán had children in that country. More children from the four communities had migrated there in search of work than had gone to any Mexican city.

Thus, in terms of the economic welfare of these communities, the picture is somewhat mixed. Agrarian reform provided access to land, but this has not led to economic security. Only in the Tamaulipas and Michoacán communities does agriculture provide households with a major part of their income, but even there the picture is quite mixed. The one in Tamaulipas seems to be a thriving agricultural community, but the one in Michoacán is very poor. Although agriculture is important in the latter, it seems that this is true only because the people have not found better alternative employment opportunities.

In all the communities, the average cash income for the household is below the gross national product per capita of Mexico as a whole. The communities in Michoacán and San Luis Potosí are especially poor, with cash incomes substantially less than those in Morelos and Tamaulipas. Poverty in the former communities has led many of these individuals, in both the older and the younger generations, to migrate temporarily or permanently in search of better opportunities.

Nutritional status and infant mortality rates

We used several different measurements of health. The first was the nutritional status of children under five years of age. This was designed to give us some indication of the current nutritional adequacy in the different communities.

Estimates of nutritional status were based on anthropometric measurements of children 60 months of age and under. Length was measured in centimetres, using an infantometer for children unable to stand unaided. For children able to stand, height was measured using a measuring board with an embedded metal metre tape and a sliding headboard. Weight was measured to the nearest 100 9 using a dial-faced spring scale (ITAC). Children's weight for age, height for age, and weight for height were calculated as percentages of the standard NCHS median values ( see table 3).

Weight less than 90 per cent of standard for age suggested some degree of undernutrition at some point in children's lives. The percentage of children in this category ranged from about 27 per cent in the communities in Morelos and Tamaulipas and over 38 per cent in San Luis Potosí to a high of 56 per cent in Michoacán. Stunting (height for age less than 95 per cent of standard), an indicator of chronic undernutrition, characterized 32 per cent of the children in Morelos and approximately 50 per cent in the other three communities. The percentage of children whose current weight for height was under 90 per cent of standard, an indicator of current nutritional status, was much smaller. It ranged from 5.6 per cent in Tamaulipas to 12.5 per cent in San Luis Potosí. The overall conclusion is that between 40 and 50 per cent of all the children in these communities appeared to have experienced nutritional stress at some time in their first five years of life. Acute undernutrition (as measured by low weight for current height) is less of a problem.

TABLE 3. Comparison of nutritional indicators for children under five years of age, 1984

  Michoacán (N = 80) San Luis Potosí
(N = 56)
Morelos (N = 70) Tamaulipas (N = 53)
Average % of children Average % of children Average % of children Average % of children
Height for age 94.8   95.6   97.6   94.7  
> 95% 49.0 50.0 68.0 53.0
< 95%   51.0   50.0   32.0   47.0
Weight for age 91.1   92.7   99.5   96.1  
> 110% 6.4 5.3 18.0 9.4
90-110% 37.2 56.1 53.5 64.2
75-90% 50.0 33.4 21.8 24.5
c 75%   6.4   5.3   5.7   1.9
Weight for height 101.4   101.6   104.1   106.1  
> 110% 11.5 16.0 27.0 28.3
90-110% 82.0 71.5 63.0 66.1
85-90% 2.6 5.4 5.0 5.6
<85%   3.8   7.1   5.0   0.0

Reference values are NCHS median values for the appropriate age, sex, and measurement categories.

These data are indicative of the nutritional situation in Mexico as a whole. A recent historical review of food consumption between 1950 and 1984 indicated that undernutrition has remained one of the country's gravest public health problems. In 1984 it was estimated that between 40 and 50 per cent of the population were undernourished [11]. Thus, in spite of agrarian reform and agricultural modernization of recent years, nutritional problems in rural communities like these have not been solved. The resources to which these families have access are insufficient to provide a diet that will lead to normal growth rates of a substantial percentage of children.

Infant mortality rates were estimated by collecting a history of all births among females in the households sampled. By aggregating responses on number of births, children still living, and children who had died, we were able to assemble a historical pattern. This could be related to significant events, such as land redistribution, that have affected the communities.

We converted the reports of the number of children born, the number surviving, and at what age and in what year children died into rates of infant and child mortality by decade. Table 4 shows that the women in San Luis Potosí reported the smallest proportion of deaths to the total number of children born, while those in Michoacán reported the largest proportion.

Infant mortality for all the communities was substantially above 100 per 1,000 live births in the 1940s and 1950s. During that period, only in the community in Morelos was the rate below that for Mexico as a whole.

TABLE 4. Mothers reporting deaths of children in the four communities

Number of deaths Michoacàn San Luis Potosí Morelos Tamaulipas
No. % No. % No. % No. %
0 52 58 49 65 60 61 46 62
1 14 16 13 17 27 28 11 15
2 13 14 4 5 6 6 10 14
3 7 8 5 7 5 5 4 5
4 3 3 1 1 0 0 2 3
5 1 1 3 4 0 0 0 0
6 0 0 0 0 0 0 1 1
Deaths as % of all children born 12.4 8.6 9.3 10.8

Infant mortality dropped substantially in the community in Michoacán during the 1960s. It is not unreasonable to suggest that this came about as a result of the improved economic circumstances that would have accompanied the creation of the ejido in 1962. Many of those who received land were actually poor, landless individuals who had been living around Lake Chapala. They were given land and resettled in the Apatzingán region in 1962. Although it is somewhat speculative to draw such a conclusion on the basis of relatively small samples, infant mortality since the 1960s has actually increased in the community in Michoacán. The ejido grants of three hectares may have been an improvement in these individuals' life circumstances, but it has not been sufficient to lead to long-term economic improvement, especially with the population growing. The infant mortality rates, which had dropped below the average for Mexico as a whole in the 1960s, were higher than those for the country in the 1970s and 1980s.

Additional indication of the beneficial effects of land redistribution on small-farmer welfare comes from an examination of the data on infant mortality for the community in Tamaulipas. Although the ejido was created in 1964, the land could not be cultivated profitably until the opening of the irrigation system in 1972. The marked decline in infant mortality in the 1970s and 1980s roughly coincided with this period of time. Many of the people who received land in Tamaulipas migrated to the area from the state of San Luis Potosí during the late 1950s and early 1960s.

The community in Morelos began with lower infant mortality rates than the other communities, probably because of its larger size and closer proximity to Mexico City. Its health-care system was more developed for these reasons. Rates were quite low in the 1950s, although we are inclined to believe that they were probably not as low as the 12 per 1,000 indicated by our data. During the 1960s and 1970s they were a bit above the national average. In 1978, wells were drilled that provided irrigation to the land of many of these ejidatarios in Morelos. Infant mortality did drop, but again we cannot be sure of the linkage.

Perhaps the most surprising result concerning infant mortality was the very low rate in recent years for the community in San Luis Potosí. This cannot be related to any improvement in access to land or to irrigation. We are much more inclined to believe that the decline relates to improvements in the medical services and transportation system in the region. During the last several years, nurses have visited the community once a week to attend to those needing medical attention. In addition, this community is located approximately 15 kilometres from a relatively large town in which there are several medical doctors and government clinics. Transportation into town is readily available. The two factors seem to have resulted in a dramatic improvement in infant mortality.

These data for the four ejidos are consistent with more macro-level data for Mexico. In an analysis of census data it was found that states in which ejidatarios comprise a higher proportion of the labour force had substantially lower than average child mortality rates [12].


The data presented here can be related to current debates in Mexico regarding the potential of ejidal agriculture to provide a viable living in rural areas. There is much discussion about whether the ejido system is an institution that has outlived its usefulness. Some scholars have jumped into the fray by suggesting that ejidos are inefficient and should be absorbed into partnerships with the private sector [13, 14]. Others point out that since the 1940s the system has systematically been discriminated against in Mexico's agricultural development. They note that the ejidos are productive, and socially and economically efficient, given the infrastructural support, agricultural research, access to credit, and other resources to which they have access [15].

We do not wish to address this debate here. What our data do indicate, however, is that access to land -especially good-quality irrigated land -and other resources necessary for agriculture are associated with general improvement in the economics, nutrition, and health of rural communities. Of the four communities studied here, the one with the best access to land and resources (in amounts usually available only to commercial farmers) had the best overall economic situation. Farmers were productive and were making good use of their resources. The level of migration was the lowest of any of the communities, and nutritional status and infant mortality seemed to be improving. This was also the case in the community in Morelos, where access to land and other employment opportunities was quite good.

The communities in Michoacán and in San Luis Potosí, on the other hand, are much like many other Mexican villages and towns that are in crisis. The several hectares of rain-fed land available to ejidatarios there do not provide a decent living. Land resources are poor, economic access is poor, and the people are poor. Infant mortality is high, especially in Michoacán, and the nutritional status of children in both communities is poor. Migration has been a solution for many individuals, especially for the children. Illegal migration to the United States is a major factor in allowing these communities to continue to exist, albeit in marginal circumstances [16, 17].

The next step in our analysis will be to look at the intracommunity variation in the various welfare indicators we have used. We expect that these will show more directly the relationship between access to economic resources and the cost in terms of health and nutrition when such access is blocked.


The authors are grateful to the International Sorghum/Millet Collaborative Research Support Program for financial assistance in making this research possible. We also appreciate the field research assistance of Oscar San Germán Alanis, Rosalia González, and Mark Skolnik in Michoacán; Marco Antonio Callejas Fuentes, Brenda McIntyre, and Uriel Sanchez Modesto in Morelos; Valentin Niembro Dominguez, Hugo Ortiz Martinez, and Gloria Quintero Boucheli in San Luis Potosí; and Jose Colli Misset, Pedro Vargas González, and Bernal Gellida Esquinca in Tamaulipas.


1. Harrison P. Inside the third world: the anatomy of poverty. New York: Penguin Books, 1983.

2. de Janvry A. The role of land reform in economic development: policies and politics. In: Eicher C, Staatz J. eds.Agricultural development in the third world. Baltimore: Johns Hopkins University Press, 1984.

3. Food and Agriculture Organization (FAO). Report: world conference on agrarian reform and rural development. Rome: FAO, 1979.

4. Rabkin R. A conservative case for land reform. Natl Rev 1985;37:36-40.

5. Beteta R. Economic aspects of the six year plan. Economic and social programs of Mexico (Mexico, 1935), quoted in Hewitt de Alcantara C. Modernizing Mexican agriculture. Geneva: United Nations Research Institute for Social Development, 1976.

6. Simpson EN. The ejido: Mexico's way out. Chapel Hill, NC, USA: University of North Carolina Press, 1937.

7. Barkin D, DeWalt BR. La crisis alimentaria mexicana y el sorgo. Problemas del desarrollo 1985;61:65-85.

8. DeWalt BR. Mexico's second green revolution: food for feed. Mexican Studies/Estudios Mexicanos 1985; 1: 29-60.

9. DeWalt BR, Barkin D. El sorgo y la crisis alimentaria mexicana. In: Paul CL, DeWalt BR, eds. El sorgo en sistemas de producción en America Latina. Mexico: CIMMYT and INTSORMIL, 1985.

10. Sanderson SW. Land reform in Mexico: 1910-1980. New York: Academic Press, 1984.

11. Reig N. Las tendencies alimentarias a largo plazo en Mexico: 1950-1984. Problemas de Desarrollo 1985;6:9-64.

12. Wood R. Land tenure and basic needs satisfaction in Mexico. (Unpublished paper) Deland, Fla, USA: Economics Department, Stetson University, 1987.

13. Yates PL. Mexico's agricultural dilemma. Tucson, Ariz, USA: University of Arizona Press, 1981.

14. Solis L. Mexican economic policy in the post-war period: the view of Mexican economists. Am Econ Rev 1973; 61 (3): 11.

15. Barchfield J. Land tenure and social productivity in Mexico. Madison, Wis, USA: Land Tenure Center, University of Wisconsin, 1979 (LTC 121 ).

16. Camara F. Kemper RV, eds. Migration across frontiers: Mexico and the United States. Albany, NY, USA: Latin American Anthropology Group, State University of New York, 1979.

17. Cornelius W. Mexican migration to the United States: the view from rural sending communities. Cambridge, Mass, USA: Center for International Studies, Massachusetts Institute of Technology, 1976.

Contents - Previous - Next