Income Inequality in Côte d'Ivoire: 1985-2014

Data on income/consumption distributions in Sub-Saharan Africa have been mainly used to study the welfare of the poorest. Yet, the rapid growth experienced by several countries in the last decades has drawn the attention towards higher earnings groups in the income/consumption distributions. However, due to under-reporting and non-response, surveys often fail to accurately measure the income of the wealthiest. Little is known about the size of such biases as it requires to have access to more reliable sources of information. In this paper we confront the 2014-2015 household survey with first-hand income tax files in the case of Côte d’Ivoire, 2014. We first identify, within the survey, a sub-sample corresponding to the one for which we have fiscal data. Comparing the earning distribution of this sub-sample with the one extrapolated from the fiscal data, we are able to measure the magnitude and the distribution of the bias among top earners in the survey. We then use this estimation to adjust the pre-tax and pre-transfer income distribution of the entire survey sample and thus recover corrected nationally representative inequality statistics. Our results show that the 2014-2015 survey significantly underestimates income inequalities. After our correction, the top 1 % share increases from 11.57 % to 17.15 %, the top 10 % share from 40.34 % to 48.28 %, and the Gini coefficient from 0.53 to 0.59. We compare our estimates with more commonly used consumption inequality measures and discuss the potential sources of differences. Making the assumption that the bias is constant over time for a given level of income, we also extend our correction to previous surveys. After correction, top 1 % shares increase by 5-6 percentage points, top 10 % shares by 7-8 percentage points and Gini coefficients increase by 6 points, making Côte d’Ivoire’s inequality levels comparable to that of the US.