Abstract.
The objective of this paper is to present an inequality decomposition method, the two-stage nested Theil decomposition method, which is an extension of the ordinary one-stage Theil decomposition method. The method is analogous to a two-stage nested design in the analysis of variance (ANOVA). It considers the three-level hierarchical structure of a country: region-province-district, and decomposes overall regional inequality, as measured by Theil indices based on district-level mean incomes, into three components: the between-region, between-province, and within-province inequality components. The within-province component is a weighted-average of within-province income inequalities for each province, while the between-province component is a weighted-average of between-province income inequalities within each region. The method uses a district as the underlying regional unit to measure regional income inequality, rather than a province, and thus can analyze the contribution of within-province inequalities as well as between-province and between-region inequalities to the overall regional income inequality in a coherent framework. This paper applies the two-stage nested Theil decomposition method to district-level income and population data in China and Indonesia and explores factors determining regional income inequality in China and Indonesia.
Similar content being viewed by others
Author information
Authors and Affiliations
Additional information
Received: October 2000/Accepted: June 2002
The author is grateful to the Foundation of International Education (Kokusai Kyoiku Zaidan) and Japan Society for the Promotion of Science (Grant-in-Aid for Scientific Research No. 12630073) for their financial support, and to anonymous referees for helpful comments.
Rights and permissions
About this article
Cite this article
Akita, T. Decomposing regional income inequality in China and Indonesia using two-stage nested Theil decomposition method. Ann Reg Sci 37, 55–77 (2003). https://doi.org/10.1007/s001680200107
Issue Date:
DOI: https://doi.org/10.1007/s001680200107