Variations in inequality of landholdings in Indonesia, 1983

Abstract

The objectives of this study are to investigate the extent to which Gini coefficients of landholdings varies across Kabupaten (Regencies) both in Java and in the Other Islands and to examine determinants of size inequality of holdings by Kabupaten (Regency). The primary source of data used in this study was the 1983 Agricultural Census, conducted by the Central Bureau of Statistics, Jakarta. Other,secondary, data which related to the study were the 1903, 1963, 1973 Agricultural Censuses and the 1920 and 1980 Population Censuses. But, the many limitations and weaknesses present in these data must be taken into consideration when interpreting the results. In 1983 agrarian structure in Indonesia showed some important changes particularly regarding the degree of inequality of landholdings, agricultural households, size distribution of farms and the proportion of land rented-in and rented-out. These may be due to the increased in populatlon pressure on land and no land reforms. In the analysis of variations in inequality of landholdings, the Gini coefficients for every Kabupaten in Indonesia have been calculated. It was found that between 1973 and 1983 in Indonesia landholdings have been unequally distributed and there has been only a slight decline in the degree of inequality. The degree of inequality of landholdings varies considerably between regions in Indonesia. The highest degree of inequality was in Kalimantan and Other Islands (0.5417 and 0.5407 respectively). However, compared with other countries such as Spain, Argentina, Peru, Venezuela, Columbia, Uruguay, India, Pakistan and Turkey, for which data available, the Gini coefficients of landholdings concentration for all places (Regencies, Provinces and Islands) in Indonesia was "moderate". The hypotheses tested quantitatively, in relation to inequality of landholdings, were that variations in inequality of landholdings, indicated by Gini coefficients, was associated with certain factors. These factors were the proportion of irrigated land to total land, average holding size, the proportion of land in owned holdings, the proportion of labour force in agriculture and the population growth rate. These hypotheses were tested using simple and multiple regression analysis (OLS), emphasising the degree of association (correlation), but not establishing causal relationship between variables. The simple regression analysis confirms that at the confidence level of 5 per cent variations in inequality of landholdings was associated with the proportion of irrigated land. A significant relationship was found in regions which are characterized by a high proportion of irrigated land such as Java and Sumatra. For Sumatra, a higher percentage of irrigated land is negatively correlated to variations in inequality of landholdings, but for Java (excluding West Java) it was positively correlated. The relationship between variations in inequality of landholdings and average holdings size was significant only at the 10 per cent level. Regions which confirm this hypothesis were Eastern Indonesia, Sulawesi and Indonesia as a whole. However, the relationship was negatively correlated. In other words, the higher average holding size, the lower will be the degree of inequality of landholdings. This suggests that increasing pressure on land (proxied by average holding size) leads to a concentration of land in fewer hands, which seems to be more true in these regions than elsewhere. Variations in inequality of landholdings also is associated with the proportion of land owned holdings and proportion of labour force in agriculture. It appears that at the 5 and 10 per cent level the relationship between variations in inequality of landholdings and the proportion of land in owned holdings was negatively correlated for regions in Indonesia as a whole. East Java, Java, Java and Bali, Sulawesi, Eastern Indonesia and Kalimantan. However, for groups of provinces D.I Aceh, North and West Sumatra it was positively correlated. Therefore, for the former regions it could be argued that tenancy is not promoting equality of landholdings, but it is in the latter regions. The test also confirms the hypothesis that the proportion of the labour force in agriculture affected variations in inequality of landholdings, except for West Java and group of provinces D.T Aceh, North and West Sumatra. But, these two variables are negativeJy correlated. This suggests that the higher proportion of labour force in agriculture, the lower will be the degree of inequality of landholdings. On the other hand, perhaps in some regions inequality of holdings is forcing labour force out of agriculture. Therefore, it could be argued that causality between proportion of labour force in agriculture and the Gini coefficient can be a two-way relationship. Further, using multiple regression analysis (logarithmic functional form) we observed that there is no region in Indonesia where all the explanatory variables have a significant impact on the degree of inequality of landholdings. Only in East Java, were variations in inequality of landholdings significantly associated with the proportion of irrigated land, the proportion of land in owned holdings and the proportion of the labour force in agriculture. Finally, within Java the population growth rate has an effect on changes in the degree of inequality of landholdings. The only region which was significant in this respect was East Java. In this area, the more rapid population growth rate, the higher will be the degree of inequality of landholdings. Therefore, it is suggested that policies to reduce this problem should be implemented here.