According to the hypothesis of an inverse relationship between farm size and land productivity, farm size can be an important determinant of success forthe land reform farms in South Africa. The analysis in this thesis provides empirical evidence in the debate on whether the land reform should equalize the land distributions or keep the large-scale sector intact. A strong inverse relationship is found between farm size and land productivity indicating that small-scale beneficiaries in the land reform are more efficient than large-scalebeneficiaries. Through indirect testing the result is attributed to intensive labour use and more extensive land use on small farms. The comparative advantage of the small-scale beneficiaries provides a strong argument to redistribute smaller holdings. This would increase output of land reform farms and probably absorb more labour, which will be a crucial contribution to alleviation of rural poverty and the success of land reform farms.
This thesis will first present the theory and discuss previous literature of the inverse relationship. Then I will use data on farms controlled by beneficiaries from a cross sectional data set from the Quality of Life survey (QoL) of 2005.Using Stata 10.0, the analysis starts with regressing farm size on land yield,defined as the value of output per land unit, which is the classical approach to test if there exists an inverse relationship between farm size and productivity.
The theory of the inverse relationship postulates that the use of cheaper familylabour on small farms, is the source of the often observed relationship.The empirical analysis shows that there is indeed a negative relationship between farm size and land productivity among the bene ciaries of the land reform. This indicates that small farms are more efficient and that labourmarket imperfections are dominant. Critics of the above analysis assert that the observation of an inverse relationship is caused by unobserved land quality differences (Benjamin, 1995). However, when controlling for land quality and other variables that may bias the results the inverse relationship remains intact. To test if labour market imperfections are the reason for the results the analysis is redone only for large farms. If the results are caused by familylabour use on smaller farms the IR should become less significant since the entire range of large farms are to a larger degree dependent on hired labour (Bhalla, 1979). The results show that the relationship between farm size and land productivity is less significant for large farms, indicating that the intensive labour use on small farms is an important determinant of the observed inverse relationship. According to this analysis subdivision of landreform projects will increase output, more of the available land will be used and the projects will create more jobs in the agricultural sector. A relevant policy implication is that smaller land holdings should be made available on the market, and that subdivision of farms can indeed be a criteria for success in the South African land reform.