Inequality in pre-income survey times: a methodological proposal

We propose different alternatives of inequality estimation for economies with a big agricultural sector where land is a decisive factor in income generation and where we do not have enough information about personal earnings. To this end, we use the Uruguayan case to test our methodology, because: (i) Uruguay's economy has the characteristics described above; (ii) we have some information available about incomes and economically active population in agriculture and (iii) we can also contrast our series with previous estimates. We propose six analytical exercises where Gini indexes are calculated, and as reference we choose the estimation that better adjusts to some theoretical and empirical conditions. Finally, we check the historical accuracy of the series by looking at explanatory variables of income distribution and the shape of the Inequality possibility frontier. Our results are consistent with the economic and social events of the period (1870-1912) and with previous estimates which reveal worsening trends in income distribution. Our annual data allow capturing the dynamics of the process where breaks in the series are observed and improvements and declines alternate in the evolution of income distribution.