The efficiency of sharecropping: evidence from the postbellum south.

• Author: Garrett, Martin A.

1. Introduction

   Views on sharecropping have been and remain controversial. Following Adam Smith, virtually all classical economists considered sharecropping inefficient. Their argument was epitomized by Alfred Marshall (1920) and became the well-known Marshallian theory of sharecropping. (1) According to the Marshallian theory, sharecropping leads to Pareto-inefficient allocation of labor because sharecroppers are paid only a percentage, rather than all, of their marginal product of labor and thus would rationally reduce their work effort. Although it has been contradicted by other theoretical arguments (Johnson 1950; Cheung 1969; Reid 1973, 1977; Roumasset 1973), the Marshallian theory dominated the literature on sharecropping for several decades. (2)

   Throughout time and across countries, however, sharecropping has been and still is an important method of agricultural production. For example, between 1880 and 1910 in the United States, owner-operated farms, as a proportion of all farms, decreased from 64 to 50% in the South and from 74 to 63% in the nation. (3) Meanwhile, rented and sharecropped farms rose sharply. In 1880, 12% of farms in the South were rented and 24% were sharecropped; the corresponding percentages were 8% rented and 18% sharecropped for the entire United States. In 1910, rented farms increased to 15% in the South and 11% in the nation, and sharecropped farms rose to 35% in the South and 26% in the nation. Interestingly, criticisms of sharecropping in the American South emerged immediately in early years of the postbellum period, although they are somewhat difficult to understand because the percentage of sharecroppers in the South increased only about 2% between the antebellum period and 1880 (Reid 1973, p. 111).

   The expansion of sharecropping in the postbellum South may be attributed to the economic situation and institutional arrangement of that time. After the Civil War, southern farmers faced a major problem in trying to acquire capital. Money was scarce at the end of the war, and banks and credit institutions had never existed in the vast agrarian interior of the South. Ownership required a significant amount of capital. Even fixed-rent farming required renters to provide all inputs, including draft animals, except land. In the absence of such needed capital, farming could become extremely difficult. (4) Even if the capital became available, many freedmen from the Civil War might not have the expertise necessary to farm by themselves. In addition, President Johnson's wholesale amnesty of ex-confederates completely wiped out the freedmen's hope for 40 acres of land and a mule from the U.S. government. Consequently, farmers were prompted to look for better working arrangements, such as sharecropping.

   The incompleteness of the labor market may also explain the expansion of sharecropping in the postbellum South. Following emancipation, there was severe disenchantment with work gang labor, and wage labor was simply ill suited for agricultural production. Agricultural production was spread over space, which made landlords' supervision complex and costly. Because it was highly seasonal and could not be postponed, the seasonal sequences in which fanning must be done demanded dependable workers throughout the growing and harvesting seasons. Falling cotton prices reduced the planters' and freedmen's earnings. Yet expectations of high cotton prices in the future created competition among planters to entice labor with higher wages. As a result, labor often migrated in midseason, and planters spent an inordinate amount of time securing labor, especially for the harvest season.

   The following insight from Parker (1980) provided us with perhaps the most compelling argument for the disenchantment of work gang and wage labor. As Parker (1980, p. 1025) pointed out, the main obstacle to a wage economy in Southern commercial agriculture in the 1860s was the same one that created slavery in the first place and gave strength to peasants' holdings and family-sized farms in free societies. In simple fact, a free labor force on a farm cannot be readily motivated, supervised, or kept honest, reliable, and dependable by direct controls on a large scale on the model of a factory. Consequently, as early as 1867, a majority of planters and freedmen were ready to adopt alternative crop-producing arrangements. Fixed rents and sharecropping became natural choices because they had long existed throughout the world. In analyzing the origins of Southern sharecropping, Shlomowitz (1979) suggested that the transition to sharecropping was, in many cases, affected by an intermediate stage--the squad system--a nd this process might have taken at least a decade and a half to work itself Out. According to Shlomowitz (1979, p. 572), the final transition to sharecropping "... probably lies in the increased effectiveness of the incentives implicit in the share arrangement, more closely matching effort and reward at the individual family level...."

   Beginning with Johnson (1950) and Cheung (1969), who led the way to argue for sharecropping, Reid (1973, 1977) and Alston and Higgs (1982), among others, have argued that sharecropped farms yield as much output as rented or owner-cultivated farms in the American South. In addressing share tenancy as a logical market response, Reid (1973, 1977) provided excellent discussions of the economic situation and the advantages of sharecropping in the postbellum South. Reid (1973) argued that "the central feature of a sharecropping contract is the continuing interest of both landlord and tenant in the efficiency of agricultural production" (p. 126) and that "evidence and theory combined imply that the rise of tenancy should have increased southern agricultural productivity, ceteris paribus" (p. 197). Reid (1977, p. 406) further maintained that "... when complementary factors need to be coordinated in ways difficult or costly to specify in advance or to supervise at the time, the theory of tenure choice predicts that sh arecropping will be preferred" and that the "continuing incentive to cooperate likely explains much of the popularity of sharecropping." Alston and Higgs (1982) also contended that the efficiency of sharecropping could result from the combination of a landlord's supervision and tenants' labor.

   Yet empirical analysis of the relative efficiency of sharecropping compared with other forms of tenancy in the postbellum South is virtually nonexistent. Although Reid (1973, 1977) and Alston and Higgs (1982) argued that the rise of sharecropping in the American South could have resulted from the advantages of sharecropping over other forms of agricultural production, they did not provide such empirical evidence. While empirical investigation was absent in Reid's work, Alston and Higgs (1982) did use farm data from 22 Georgia plantations to test the impact of a set of "predetermined" variables on the choice of contractual mix. They found that the value of land, improved acres, work stock, cotton acreage, and landlords' supervision did have statistically significant impacts on the choice of contractual mix in Southern agriculture. But they did not investigate the efficiency of sharecropping relative to other forms of tenancy. In a more recent study, Brinkley (1997, p. 130) provided some empirical work and foun d "... sharecropping being more productive than farm ownership ... and being less productive at the margin as indicated by the negative impact on output of its interaction with capital." He concluded that "the net effect of sharecropping appears to be close to zero with greater output in some counties and smaller output in others." While Brinkley's study provides us with some knowledge about the role of sharecropping in the postbellum South, the role of sharecropping was not the focus of his study. Hence, he did not attempt to explain why landlords and farmers chose sharecropping if the net effect of sharecropping on output was zero.

   Given the prevailing Marshallian view that sharecropping is inefficient, the empirical work in this article provides significant insight into the efficiency of sharecropping in the postbellum South. Our hypothesis is that, given the economic situation and institutional arrangement after the Civil War, sharecropping in the postbellum South was desirable because sharecropped farms could be more productive than owner-operated farms and perhaps more productive than rented farms. Although it is difficult to prove why sharecropping was more efficient, our explanation for the hypothesis is that sharecropping should be viewed as an incentive to work rather than a tax that produced a disincentive to work. Because highly disaggregated data at the farm level are not currently available, we use county data from the postbellum South to explore whether the output elasticity was higher under sharecropping than under other forms of agricultural production. We focus our empirical inquiry directly on the productivity of sharec ropping and compare it with that of ownership and tenants.

   Our empirical results show that the output elasticity of sharecroppers is indeed significantly higher than that of owners on all occasions and that of tenants on some occasions. Most important, the results are very robust. Therefore, our study helps explain why sharecropping rose sharply in the postbellum south and is complementary to the existing arguments in the literature. It opens a new avenue for researchers in various disciplines to understand sharecropping better and to examine more carefully this interesting form of production.

2. The Empirical Model and the Data

   To test the hypothesis that sharecropping was desirable in the postbellum South because sharecropped farms could be more productive than owner-operated farms and perhaps more productive than rented farms, we may estimate separately the production function for owners, sharecroppers, and tenants and then compare the...