Answers are grounded in this paper and cite their sources.
In a perfect democracy, social characteristics of people should not affect their political participation. If they do, the voice of people is skewed, and certain groups are underheard and underrepresented. I state that the extent to which social characteristics affect political participation defines political participation inequality. In this paper, I develop a new Political Voice Inequality Index (PVI-index) to measure political participation inequality. Via simulations, I show that the adjusted R-squared coefficient is a good metric for the PVI-index because it is a valid, comparable, multidimensional, simple, generalizable, and robust measure, and it outperforms the alternative measures such as b-coefficients, (unadjusted) R-squared, MAE, MSE, and RMSE, and other measures proposed in the literature. To validate the index, I analyze about 1.5 million respondents from high quality cross-national longitudinal surveys: ANES, CSES, ESS, ISSP, and WVS – spanning 1980 to 2022, comprising 700 country-years. I measure PVI-index as the adjusted R squared coefficient by regressing political participation on seven basic social characteristics: age, sex, education, employment, urban residence, marital status, and religiosity, – within each country-year. Machine learning modeling based on about 180 country-level indicators showed that the best predictors of political participation inequality were the form of government (presidential vs mixed or parliamentary systems), the number of changes in government per year, the share of women in parliament, the Index of Participation of the Democracy Barometer, and other variables. The PVI-index can be updated yearly for all countries with available data, becoming an important indicator of the health of democracy across the world.
No takes yet. Share an insight, caveat, or question.
Vardan Barsegyan (2023) studied this question.