Bureaucratic Behavior Modeled by Reduced-Rank Regression: The Case of Expenditures from the Soviet State Budget
Working Paper 96-1
A bureaucratic hierarchy compresses information flowing upward and disaggregates directives passing downward, thus making itself amenable to modeling by reduced-rank regression. Applying this technique to the Soviet government, we find that about eight latent variables are required to characterize the mapping of predetermined variables into a vector of budgetary expenditures. The latent variables can be interpreted as targets for expenditures in high-priority sectors.
JEL classification: C5, H5, P35
Key words: bureaucratic behavior, reduced-rank regression, Soviet economy, public finance.