Political Budget Cycle

Here's an interesting paper on the political budget cycle. Page 22 of the paper (page 24 of the pdf) tells most of the story-- that the political budget cycle exists, and that while only Latin America hit statistical significance with their estimates, the budgetary pattern around elections still holds in the OECD countries as well. It would be interesting to see how the democracy variable plays in but I don't see it in any of the regression tables; the abstract is below:

While existing cross-country studies on political budget cycles rely on annual data, we build a panel with quarterly and monthly data from Latin American and OECD countries over the 1980-2005 period. Disaggregated data allow to center the electoral year more precisely, and show the effects are concentrated in a three-quarter window around elections. Cycles are statistically significant only in Latin America, but the pattern is similar to OECD countries: the budget surplus/GDP ratio falls in the election period and rises in the post-election period. In line with the logic of rational opportunistic manipulation, these effects cancel out.