Determinants of income among rural self-employed women: a longitudinal analysis

Abstract

The objective of this study was to analyze the factors associated with the annualized real earnings of rural self-employed women in Peru during 2020–2024. The study employed a quantitative approach with a non-experimental, longitudinal design and a correlational scope. Data from the National Household Survey (ENAHO), modules 200 and 500, were used, based on an unbalanced panel of 2,978 person-years corresponding to 1,177 women. We estimated clustered ordinary least squares models with random effects, fixed effects, and correlated random effects using the Mundlak approach, with robust standard errors clustered by person. The results showed that the absence of registration with the National Superintendency of Customs and Tax Administration (SUNAT) was negatively and significantly associated with annualized real earnings in the fixed-effects model (β = -0.482; p = 0.011), equivalent to a difference of 38.2%. In the clustered models, age, educational attainment, and internet use showed positive associations with earnings; however, these associations were not statistically significant in the fixed-effects model. Bank account ownership also did not show a robust association. We conclude that a lack of registration with SUNAT is associated with lower earnings. This result does not establish causality and should be interpreted with caution, given that only 36 women changed their registration status and that single observations were excluded. The findings suggest promoting comprehensive strategies for formalization, connectivity, training, and financial inclusion linked to market opportunities for rural women.