Question

A researcher plans to study the causal effect of a strong legal system on the number of scandals in a country, using data from a random sample of coun- tries in Asia. The researcher plans to regress the number of scandals on how strong a legal system is in the countries (an indicator variable taking the value 1 or 0, based on expert opinion).

A. Do you think this regression suffers from omitted variable bias? Explain why. Which variables would you add to the regression?
B. Using the expression for omitted variable, assess whether the regression will likely over- or underestimate the effect of a strong legal system on the number of scandals in a country. That is, do you think that Ể > B, or ŝi < Bı?

Answer 1

Yes, the regression suffers from omitted variable bias. The researcher should include additional variables in the regression model. Omitted variable bias will likely underestimate the effect of a strong legal system on the number of scandals in a country.

Step-by-step explanation:

Yes, the regression suffers from omitted variable bias. Omitted variable bias occurs when there are important variables that are not included in the regression model and are correlated with both the dependent and independent variable. In this case, there are likely other variables that may influence the number of scandals in a country, such as government transparency, economic stability, and cultural norms. To address this bias, the researcher should include these additional variables in the regression model.

Using the expression for omitted variable bias, it is likely that the regression will underestimate the effect of a strong legal system on the number of scandals in a country (ŝi < Bi). Omitted variable bias tends to bias the coefficient of the included variable towards zero. Since a strong legal system can potentially reduce the number of scandals, not accounting for other influential variables may lead to an underestimation of this effect.