83rd International Atlantic Economic Conference

March 22 - 25, 2017 | Berlin, Germany

Red fed, blue fed: Why the Federal Open Market Committee votes the way it does

Friday, 24 March 2017: 15:50
Paul Pieper, Ph.D. , Economics, University of Illinois at Chicago, Chicago, IL
This paper examines the effect of the political culture of the twelve Federal Reserve Districts on the voting behavior of their presidents in the Federal Open Market Committee (FOMC). The Federal Reserve districts are categorized into “red” and “blue” districts based on the percentage of the presidential vote received by the Democratic candidate in 2008. Since all but one district crosses state lines, this calculation is made on the county level. There are several reasons why district bank presidents will reflect the political views of their districts. First, the presidents must be from the district that they serve. Second, they have frequent interactions with business and community leaders from their district and will to some extent reflect the political views of their constituents. Third, the presidents are reviewed by their banks' Board of Directors. It has long been argued that liberal policymakers favor a lower unemployment target than conservative policymakers while the latter group is more sensitive to concerns about inflation. It follows that bank presidents representing conservative (red) districts will prefer a more hawkish inflation policy and will vote more frequently to raise interest rates than presidents from blue districts.

This hypothesis is tested using FOMC voting records from 2000 to 2016. The relationship is estimated using a mutinomial model, where the dependent variable can take three values: vote with the chair, dissent in favor of higher interest rates, and dissent in favor of lower interest rates. Control variables include the inflation rate, the unemployment rates, and a dummy variable for the appointing Federal Reserve chairman. We find that the political culture variable has its hypothesized sign and is highly statistically significant.