Is the Federal Reserve a typical bureaucracy?

Some analysts have considered whether, in a public choice framework, the Federal Reserve behaves as a typical bureaucracy. In this regard, Shughart and Tollison hypothesized that growth in the monetary base partly reflects the financing of growth in the number of staff employed by the Federal Reserve. They found evidence of correlation and one-way causality in support of their hypothesis. The authors reexamine the Shughart-Tollison empirical findings, and conclude that when one adjusts the variables to be sure that the residuals are white noise, there is no statistically significant, contemporaneous correlation between Federal Reserve employment and the monetary base. Also, an apparently significant correlation between current growth of the monetary base and lagged growth in Federal Reserve employment disappears when a few initial observations are deleted from the sample. The same shortening of the sample period also affects their causality results. Increases in Federal Reserve employment no longer cause increases in the monetary base; in fact, there is no evidence of causality in either direction, based on standard Granger causality tests. Finally, when an error-correction model is used, the authors find evidence of uni-directional causality in the other direction, from the monetary base to Federal Reserve employment.