The Statistical Mechanics of Income in Peripheral Capitalism: Peru, 2004-2022

This paper analyzes the evolution of household income in Peru over the past two decades using the maximum entropy method and an expectation-maximization (EM) algorithm to estimate a mixture of density components that reflect various aspects of the political economy. The findings suggest that Peru’s income distribution aligns with a combination of exponential and Pareto distributions, similar to patterns observed in advanced capitalist economies. For the period 2004-2022, the estimates indicate that 10% of the income distribution is best described by an exponential density, 8% by a log-normal distribution, and 82% by a Pareto distribution. The chapter also confirms the persistence of these exponential-log-normal-Pareto patterns in Peru and other Latin American economies. In response to criticisms regarding the method’s handling of homoplutia, the chapter presents descriptive statistics showing that variations in homoplutia across countries correlate with profitability trends, providing evidence for the relatively low significance of this phenomenon in Peru’s household income distribution. Lastly, the estimation of the Lorenz curve for the Peruvian income distribution suggests that the finite mixture model, along with the exponential density, performs better in capturing the dynamics of income distribution.