Tuesday, November 25, 2014

URSP Student Amy Handlan Builds a Panel Data Set from Annual Data and Running Fixed Effect Panel Regressions

Hello, my name is Amy Handlan. I am an economics major, math minor at George Mason University. This is my second semester with URSP researching the relationship between government type and financial instability. Last semester I found that there is a significant, negative quadratic relationship between Polity score, an index of government ranging from autocratic to democratic, and frequency of financial crisis, a measure of financial instability. I found this relationship in a cross-section of 70 countries from 1970 to 2006. However, cross-section regressions do not account for variations over time or for country time-invariant traits.

Accordingly, this semester I have been working on building a panel data set from the annual data and running fixed effect panel regressions. Once accounting for country characteristics and time, the coefficients became positive. This implies that there could be a positive relationship between changes in government and frequency of financial crises. While getting significant coefficients in my econometrics is exciting, my biggest challenge this semester has been working with the theories behind my empirical relationships.
           
Therefore, this semester I have been focusing on developing my review of related literature and looking for possible channels of causation. This week I have started reading

Charles Calomiris and Stephen Haber’s Fragile by Design: The Political Origins of Banking Crises and Scarce Credit. The focus of their book is that banking systems are infused with politics because “the property-rights system underpinning banking systems is an outcome of political deal making” (13). Accordingly, when political structures are different or change, there will be an effect on the banking system and thus financial markets. This book is helping me build a foundation for developing the theoretical channel through which different government types could create different frequencies of financial crises.