An analysis of changes in Indonesia's income distribution, 1996-2008

This thesis examines the evolution of income distribution in Indonesia during the period 1996-2008, a period characterized by the region’s worst financial crisis in 1997 as well as by economic and institutional reforms. It investigates the changes in income distribution as well as in inequality and poverty, what determining factors have the greatest impact on the changes, and the parts of the distribution influenced greatest by the factors. It draws relevant policy implications from the analysis as well. In this study, Bayesian analysis is utilized to empirically investigate the changes in Indonesia’s income distribution from the perspective of the complete distribution. Two three-parameter functions, namely, the Singh-Maddala and the Dagum distributions are proposed to model the income distribution. A model comparison confirmed that the Dagum function is the best-fitting parametric density to model the expenditure distribution for both urban and rural regions. The performance of the univariate Dagum density is then improved through the use of a mixture of Dagum densities. Next, the mixture of the Dagum densities is employed to further reveal changes in income levels, the modes and inequality and poverty measures over the observed period. To determine the potential causes that have driven the changes in the expenditure distribution as well as in the inequality and poverty, the DiNardo et al. (1996) decomposition framework is adopted. Bayesian estimation is then applied to the counterfactual densities, assuming that the densities follow the Dagum distribution. The study found that the change in the income distribution improved the welfare of a large number of individuals, however the inequality among them also increased. The findings of the decomposition analysis suggest the changes are mainly associated with gradual improvement in the levels of the formal education of the population, which acted substantially to reduce poverty. Concurrently, the same factor is also part of the cause of rising inequality over the years. The reason for this is that access to better education is still dominated by the non-poor income group, pushing their income to the upper-middle range of the distribution and thus increasing inequality. On the other hand, occupational status and age demonstrated a neutral effect on changes in poverty and inequality. In addition to adhering to policy that aims to achieve higher economic growth and liberalization, the government also needs to ensure that equal participation among the population in such growth, and thus equal access to income across regions, are realised. This can be achieved through policies that strongly regulate equal access to all levels of education and training services regardless of an individual’s background. The development of the education policy itself needs to emphasize the employability of graduates based on skills that are sought-after in the labour market, besides the quality and efficiency of education and training services. To increase the population’s productivity and participation in the economic growth and liberalization process, policies that highly stimulate rural and agricultural development and favourably promote equal employment opportunities across gender and age are also essential. The strategy also needs to be complemented by a social safety net program to ensure rapid response to vulnerable groups during difficult times