Political Variables in Cross‐country Growth Analysis
这篇论文梳理了跨国增长回归中常用的五类政治变量（民主、政府稳定、政治暴力、政策波动、主观政治感知），并比较了它们的解释力，发现民主指标效果最差，政策波动和主观感知效果最好。
The recent boom in cross‐country growth analysis has renewed the empirical interest in political sources of economic growth. A large number of studies have tested political variables in growth regressions and a very heterogenous set of variables has been proposed to measure economically relevant differences in the political systems of countries. This paper distinguishes five categories of relevant political variables: democracy, government stability, political violence, policy volatility, and subjective perception of politics. For each of these categories, the specification, testing method, and results of the most relevant studies are presented and are critically discussed. The paper concludes that measures of democracy are least successful and measures of policy volatility and subjective perception of politics most successful as explanatory variables in cross‐country growth regressions.