Between communism and capitalism: long-term inequality in Poland, 1892–2015

Abstract

We construct the first consistent series on the long-term distribution of income in Poland by combining tax, household survey and national accounts data. We document a U-shaped evolution of inequalities from the end of the nineteenth century until today: (1) inequality was high before WWII; (2) abruptly fell after the introduction of communism in 1947 and stagnated at low levels during the whole communist period; (3) experienced a sharp rise with the return to capitalism in 1989. We find that official survey-based measures strongly under-estimate the rise in inequality since 1989. Our results highlight the prominent role of capital income in driving the U-shaped evolution of top income shares. The unique inequality history of Poland speaks to the central role of institutions and policies in shaping inequality in the long run.

Wealth Inequality in Interwar Poland

“…The level of wealth inequality in interwar Poland was much higher than today. A comparison of our estimates of wealth inequality with estimates of income inequality in the late 1920s calculated by Bukowski and Novokmet (2021) shows that wealth inequality was higher than income inequality. This is a common situation.…”

“…We have also to keep in mind that large state-owned enterprises were exempt from taxation. As the number of legal persons is small, similarly to Bukowski and Novokmet (2021) we think that their inclusion in has limited impact on our estimates. To assess the impact of the inclusion of legal persons in wealth tax statistics we perform a robustness exercise, in which we estimate the value of wealth held by legal persons based on the capitalization of income tax paid by legal persons.…”

“…Estimates of income inequality in interwar Poland are available in the literature. Bukowski and Novokmet (2021) estimate long-run income inequality using income tax data combined with survey data. According to their results in 1925, the top one percentile had 10.5% of total income, while the income share of the top 0.01% equaled 0.9%.…”

“…We estimate wealth inequality using Pareto interpolation techniques. This is a common approach in the case in which tax data is expressed in brackets (Kuzents, 1955;Piketty, 2001;Atkinson, 2007;Atkinson andPiketty, 2007, Cowell, 2011;Wroński, 2019;Bukowski and Novokmet, 2021). In our case, we have information on the number of taxpayers and total wealth in each tax bracket (si;si+1).…”

“…Statistical sources do not contain direct information on the number of households in interwar Poland. To estimate the size of our control population we follow Bukowski and Novokmet (2021). We use outcomes of the 1921 Census presented in the Statistical Yearbook 1926 (GUS, 1927).…”