The Dynamics of Wealth & Income Inequality: Assessment of Temporal Relationships and Influence of GDP & Interest Rates
GitHub Repo for semester project at HU Berlin using a VAR model to explain wealth and income inequality developments in the U.S.
- Economic inequality is high in the United States, both when assessing income and wealth inequality. The top 10% receives 45.5% of the total income (Chancel et al., 2022). In contrast, the bottom 50% earns just 13.3%. The economic divide is even more apparent when examining wealth inequality, where the top 10% owns above 70% of the nation’s wealth, whereas the bottom half owns 1.5% of the total wealth.
- While much work has been done on the descriptive measurement of inequality and also its consequences, little research has modeled indicators of economic inequality using time series analyses. Only recently, research has began to study the effects of macroeconomic variables on the trajectories of wealth and income inequality (e.g., Berisha & Meszaros, 2020).
- Assess the reciprocal dynamics of wealth and income inequality in the U.S. based on a new dataset by Kuhn et al. (2020) on U.S. households in the postwar period.
- Study the effects of interest rates and GDP on both inequality indicators to partially explain the trajectories.
- Use a Vector Autoregressive Model (VAR) to model the reciprocal relationships between the inequality measures and a VARX model that allows the use of exogenous variables.
- Follow a five-step process to derive results, which focuses on explaining relationships between variables and testing hypotheses (not prediction performance):
- A VAR(6)-model showed the best model fit according to AIC and SC criteria. The resulting models are presented below:
- Also, a VARX(3,3)-model was used including interest rates (endogenous) and GDP (exogenous) based on a Granger causality test:
- Surprisingly, wealth inequality did not granger cause income inequality nor vice versa. GDP reduced the residual error covariance matrix compared to a pure VAR model and interest rates showed a reciprocal relationship with income inequality and an effect on wealth inequality in some lags.
- Install R (4.2.1) and RStudio
- Download the macroeconomic dataset from the GitHub Repo and the SCF+ data from Kuhn et al. (2017) from the Google Drive: https://drive.google.com/drive/folders/1fDQmMi-66BoRTC8Ob9RjARZuvg1Yr2md?usp=sharing
- Change the data loading path to your local working directory
- Run the script and read the accompagnying written report