*While earnings inequality remained virtually unchanged in urban India between 2004-05 and 2011-12, it declined sharply in rural India over this period. This column finds that although the change in the distribution of education among paid workers had an inequality-increasing effect, there was a net decline in rural inequality because returns to increased levels of education improved more for low-earning workers than high-earning ones.*

**Earnings inequality in India**

**Figure 1. Real weekly earnings, by percentile, 2004-05 and 2011-12**

^{1}for 2004-05 and 2011-12. At each percentile, earnings were higher in 2011-12 than in 2004-05. The gap between the two curves reveals that the increase in earnings was, in absolute terms, greater for higher percentiles. For instance, real weekly earnings increased by Rs. 99 at the first decile, Rs. 194 at the median, and Rs. 307 at the ninth decile. However, as seen in Figure 2, the percentage increase in earnings was greater at the lower end of the distribution. For instance, earnings increased by 91% at the first decile, 74% at the median, and by 44% at the ninth decile. Thus, earnings inequality - defined in relative rather than absolute terms - declined over the seven-year period.

**Figure 2. Change in log of real weekly earnings, by percentile, 2004-05 to 2011-12**

^{2}. The Gini coefficient of real weekly earnings fell from 0.462 to 0.396. This is in sharp contrast to the picture in urban India where earnings inequality remained virtually unchanged over the period: The Gini coefficient of real weekly earnings in urban India was 0.506 in 2004-05 and 0.499 in 2011-12.

**Decomposition of the change in earnings**

^{3}as observed in 2004-05, with the rates of return (essentially how the labour market rewards males versus females, and how it rewards illiterates, high schoolers and college graduates) as observed in 2011-12. Consequently, the difference between the 2004-05 distribution and the hypothetical distribution gives us the structure effect, that is, the part arising due to changes in the rates of return keeping the distribution of characteristics fixed at 2004-05 levels; while the difference between the hypothetical distribution and the 2011-12 distribution gives the composition effect, that is, the part due to changes in the distribution of worker characteristics keeping the rates of return fixed at 2011-12 rates.

**Figure 3. Aggregate decomposition of earnings**

**Conclusions and policy implications**

*The research on which this column is based received funding from the European Community’s Seventh Framework Programme (FP7/2007–2013) under grant agreement No. 290752. The views expressed are of the authors and do not reflect those of Statistics Canada or other institutions that the authors are affiliated to.*

*Notes:*

- A percentile (or a centile) is a measure used in statistics indicating the value below which a given percentage of observations in a group of observations fall. For example, the 20
^{th}percentile is the value (or score) below which 20% of the observations may be found. - Gini coefficient is the most commonly used measure of inequality. The coefficient varies between 0 (reflecting complete equality) and 1 (reflecting complete inequality).
- The complete set of worker characteristics that we consider are gender, marital status, religion, caste, education, industry, occupation and state of residence.

**Further Reading**

- Khanna, Shantanu, Deepti Goel and René Morissette (2016), “Decomposition Analysis of Earnings Inequality in Rural India: 2004—2012”,
*IZA Journal of Labor & Development*; 5: 18. - Kotwal, Ashok, Bharat Ramaswami and Wilima Wadhwa (2011), “Economic Liberalization and Indian Economic Growth: What´s the Evidence?”,
*Journal of Economic Literature*, 49(4):1152-1199. Available here.