An analysis on the relationship between income distribution and human development with panel data and machine learning

The aim of this study is to examine the effects of socioeconomic indicators such as human development,\rurbanization and female employment rate on income distribution in MIST countries, with both econometric\rand machine learning methods, and to predict the performance of these variables in income distribution\restimation.. Although there are many studies on growth in the economics literature, the income distribution\rhas not found the place it should be. To fill this void, the intensity of studies between growth and income\rdistribution has increased recently. For this purpose, we employed two separate analyzes using the panel data\rmethod and the Support Vector Regression method, which is one of the machine learning methods, by reaching\rthe Gini coefficients, HDI, urbanization, and female labor force participation data of the MIST countries for\rthe years 1990-2019. As a result, all determinants in the random-effects model have a statistically significant\reffect on income inequality. In the model, while HDI and urbanization are significant at 5%, female labor force\rparticipation is significant at 0.1%. Signs of all explanatory variables are negative, and therefore they have a\rdecreasing effect on income inequality. As HDI, urbanization, and female labor force participation values of\rMIST countries have improved since 1990, the Gini coefficients have improved. In other words, the income\rdistribution of these countries has become fairer. Unlike other studies in the literature, the Support Vector\rRegression model was used in this study, and SVR produced a more suitable model for predicting income\rinequality.