Regression results for fiscally conservative behaviour

Conclusions to Part 1

Did the debate around the Stimulus funds led by U.S. governors carry more bark than its bite? Here we find that while in FY 2009, spending changes at the state level were indeed linked to partisanship of the state governor, GDP per capita and education, whether a governor opposed the Obama stimulus package was not a significant predictor of changes in spending. Only in FY 2010, post budget crisis, do we find opposition to the stimulus to be the single greatest predictor of changes in spending, with all other variables, including governor partisanship being insignificant. Neither governor’s partisanship nor opposition to stimulus affect net changes in taxes of countercyclical or fiscally conservative behaviour significantly for FY 2010. In fact, none of the dependent variables are significant at a confidence level of 5% for multivariate regressions explaining net changes in taxes, countercyclical or fiscally conservative behaviour in FY 2010. In regards to the ARRA stimulus, while it was intended to help states behave counter-cyclically, we estimate that only in nine states did the funds actually swing the spending change from being negative to positive. Only three states actually behave counter-cyclically in FY-2010, and among them we find Alaska, where Governor Palin was one of the most vocal opponents of the funds. These findings lead us to conclude that while spending changes in FY 2009 followed a pattern that could be explained with economic data and by governor partisanship, spending changes in FY 2010 were more erratic, with the only significant explanatory variable being opposition to Stimulus. And while opposition to stimulus explained spending changes in FY 2010, the fact that it did not explain them in FY 2009 and does not explain tax changes, fiscally conservative behaviour or counter-cyclical behaviour implies that lowering spending in states where governors opposed stimulus is perhaps more of a consequence of having spoken out against the stimulus (in order to uphold the dialogue), rather than a profound difference in beliefs. Changes in spending may well be the most visible kind of fiscal behaviour, and thus most influenced by partisanship. This paper also provides evidence to the effect that partisanship does not necessarily impact taxation, fiscally conservative or counter-cyclical budget behaviour in the states. For example, all three states with positive counter-cyclical behaviour are Republican. The Obama stimulus, also criticized as it was speculated to increase government spending, is actually estimated to have been more used to decrease spending cuts than it was to actually increase total spending, which it only did, and minimally, in a small number of cases. A look at the fiscal behaviour of Canadian provinces is attached in an Appendix.

PART 2: US-Canada comparison: two different crises

Date: 2016-04-22; view: 581