Replication Data for: How can democratization reduce the income inequality?: Global panel data analysis based on random forest method

The mechanism of the relationship between democratization and inequality has been one of the main focuses of political and economic research, albeit with no consensus. The presence of missing values, the voluminous social, economic, environmental measures across countries and the discretion needed to select proxies/measures, leading to difficulties in conducting multinational panel data analysis. This study utilizes one of the machine-learning techniques, the random forecast model, to interpolate missing values and select candidates of control variables based on importance ranking, which greatly reduces the variable selection bias and increases the continuity of data in time series. We then constructed unbalance panel data analysis on data from 151 countries (1993-2017). Results from different models come to a consensus that the promotion of democracy can significantly reduce income inequality, but this effect has declined slightly, both in terms of wealth and in the long run. However, such mechanism only exists in non-colonial nations and presidential states.