ABSTRACT

Modeling the influence and interaction of key variables in national political economies has attracted increasing attention in the past few years. With the expansion of interest has come a proliferation of questions, contexts in which to ask them, and, inevitably, conclusions that are not always compatible. Do military regimes allocate resources differently than civilian regimes? Do parties of the left spend differently when they are in power than parties of the right? Is the political business cycle a persistent feature of electoral systems or simply an intermittently on-off phenomenon? How do political decision-makers respond to political-economic crises? How do political economies respond to political decision-makers who create domestic crises through over-or underspending? The questions are numerous and the answers that are produced will always depend to some extent on how the questions are posed. For example, a cross-sectional analysis of 20, 40, 80, or 160 political economies is likely to yield a different statistical outcome each time the number of cases is changed. A longitudinal analysis is also likely to produce different answers depending on how long the time series are. Considerations such as these underscore the need to be as explicit as possible at the outset about what sort of investigation is being undertaken.