Beyond tax-survey combination: inequality and the blurry household-firm border

Inequality evidence based on surveys, tax records, or their combination often result in divergent trends, fueling the distributional debate in Latin America. Beyond the strengths and weaknesses of these sources and their combination, tax-survey data face two shortcomings: they are unable to account for aggregate household or national income, and they are affected by firm owners’ decisions about the distribution of profits, changing which incomes researchers can actually observe. We combine social security data, household surveys and matched personal and firm tax records, which allows us to accurately account for all income sources, particularly capital incomes at the firm and individual level. Based on these unique data, we assess inequality trends in Uruguay, showing that increasing profit-distribution by firms pushes tax-survey top shares upwards, but that this trend is offset when undistributed profits are accounted for. These results call for caution when using tax-survey data without considering changes in profit-distribution.