EMPIRICAL DETERMINATION OF THE GROWTH—
MILITARY EXPENDITURES RELATIONSHIP
We begin our empirical analysis with a simple comparison of trends for each of the “great power” countries from approximately 1870 to 1939. Simple graphs show how such historical events as the RussoJapanese War or the Versailles Treaty influenced movements in military expenditures and economic output. They also provide an initial test of the universality of the military expenditures—growth relationship without being subject to the data requirements of a more formal statistical analysis.
To control for some of the other factors that may influence domestic resource allocation toward the military, however, statistical analysis is a useful tool. In particular, a system of regressions explicitly allows for the possibility that increases in one country's military expenditures, or in the size of its armed forces, might influence another country's decision on how much to spend. We stress, however, that our analysis here asks simply whether the relationships illustrated in the graphs appear robust when other variables are factored in. Our focus is on the sign rather than the magnitude of particular parameters because data limitations—including significant numbers of missing observations—preclude more precise estimation.
Our data consist of annual measures of national output, military expenditures, military personnel, and government expenditures for each country, plus price deflators and currency exchange rates for making cross-national comparisons. Measures of real output growth are calculated by taking the log difference of real output. An index