Three essays: Military base closures and federal spending

This work of three essays investigates government decisions. Two essays examine military base closure policies and the third essay studies the determinants of federal (defense and nondefense) spending.; The first essay develops a positive analysis of military base closure policy following an exogenous shock that increases national security. The analysis finds that, when labor resources cannot move costlessly from the military sector to the civilian sector, the optimal policy when commitments are possible is not incentive compatible in a discretionary regime. In the discretionary regime, base closure policy involves excessive employment in the military sector--a positive bias in protection of the military.; The second essay develops an empirical model to explore congressional influence and institutional changes on military base closing decisions. The model is applied to data of base closings announced from 1961 to 1995. The findings extend the limited empirical analysis of congressional influence on base close decisions in two ways: first by examining additional explanatory factors not previously studied and second by taking into consideration the timing of base closure decision. Using a failure time estimation procedure, it is found that both descriptive characteristics and base representation on congressional subcommittees are useful in determining which bases will remain open longer.; The final essay develops empirical models to investigate the role of constituent interests, political institutions, and vote trading behavior in decisions of both defense spending and nondefense spending across states. Using panel data from 1981 to 1992 in two-stage least square regressions, the study finds that both defense spending and nondefense spending depend on the concentration of interest groups and the organization structure of the spending decisions. Empirical evidence also suggest that vote trading behavior occurs in defense spending decisions; however, the evidence does not support such behavior in nondefense spending decisions.