TY - JOUR
T1 - Decomposing changes in inequality
T2 - Evidence from Myanmar
AU - Aung, Lwin Lwin
AU - Warr, Peter
N1 - Publisher Copyright:
© 2021 John Wiley & Sons Ltd
PY - 2021/8
Y1 - 2021/8
N2 - We study the determinants of changes in the distribution of household expenditures in Myanmar, using a large, nationwide panel household survey data set covering 2005 and 2010, the only panel household survey data currently existing for Myanmar. There was no statistically significant change in mean real consumption expenditures between these years, but the distribution of that consumption across households became significantly less unequal. The sources of this large decline in measured inequality have not been systematically investigated. The paper attempts that investigation by using the above data and applying regression-based inequality decomposition methods developed by Fields and its subsequent extension by Yun. The results using Fields’ approach indicate that region-specific variables, occupational changes, and education were the main factors contributing to the narrowing of expenditure inequality but that road development expenditures had the opposite effect. On the methodology of decomposition, we argue first that the Yun extension of Fields’ approach promises enhanced policy relevance, by distinguishing between changes in household characteristics and changes in their impact on expenditures. Second, we show that the decomposition described by Yun entails arbitrary sequencing of the distributions being compared, substantially influencing the decomposition obtained. We explain the underlying reason for this problem and demonstrate a simple solution.
AB - We study the determinants of changes in the distribution of household expenditures in Myanmar, using a large, nationwide panel household survey data set covering 2005 and 2010, the only panel household survey data currently existing for Myanmar. There was no statistically significant change in mean real consumption expenditures between these years, but the distribution of that consumption across households became significantly less unequal. The sources of this large decline in measured inequality have not been systematically investigated. The paper attempts that investigation by using the above data and applying regression-based inequality decomposition methods developed by Fields and its subsequent extension by Yun. The results using Fields’ approach indicate that region-specific variables, occupational changes, and education were the main factors contributing to the narrowing of expenditure inequality but that road development expenditures had the opposite effect. On the methodology of decomposition, we argue first that the Yun extension of Fields’ approach promises enhanced policy relevance, by distinguishing between changes in household characteristics and changes in their impact on expenditures. Second, we show that the decomposition described by Yun entails arbitrary sequencing of the distributions being compared, substantially influencing the decomposition obtained. We explain the underlying reason for this problem and demonstrate a simple solution.
KW - Myanmar
KW - expenditure inequality
KW - panel data
KW - regression-based decompositions
UR - http://www.scopus.com/inward/record.url?scp=85111926825&partnerID=8YFLogxK
U2 - 10.1111/rode.12781
DO - 10.1111/rode.12781
M3 - Article
SN - 1363-6669
VL - 25
SP - 1172
EP - 1196
JO - Review of Development Economics
JF - Review of Development Economics
IS - 3
ER -