Trust and Growth in the 1990s - A Robustness Analysis

We conduct an extensive robustness analysis of the relationship between trust and growth for a later time period (the 1990s) and with a bigger sample (63 countries) than previous studies. In addition to robustness tests that focus on model uncertainty, we use Least Trimmed Squares, a robust estimation technique, to identify outliers and investigate how they affect the results. We find that the trust-growth relationship is less robust with respect to empirical specification and to countries in the sample than previously claimed, and that outliers affect the results. Nevertheless trust seems quite important compared with many other growth-regression variables.