We exploit differences in the stringency of balanced budget rules across US states to estimate the effect of fiscal policy cyclicality on GSP growth. While most states have passed laws restricting deficits, the nature and strictness of these laws vary greatly.  States with more stringent balanced budget restrictions run more pro-cyclical fiscal policy.  We use the diversity in these laws as an instrument for the cyclicality of state government spending.  We find modest evidence that a more counter-cyclical fiscal policy increases a state's average growth per capita.  In fact, our point estimates suggest that a state could increase its annual growth rate by 0.4% by relaxing the "ex-post" balanced budget restriction.  This estimated effect is statistically significant at the 10% level, but loses its significance when we control for the initial debt to GSP ratio. 

Objectives:

Using US state-level data from 1977 – 2005, our goal is to analyze whether the cyclicality of fiscal policy affects long-run growth.  If it does, we would like to determine which components of public spending are most effective. 

Data / Methods:

We have disaggregated US state-level data on the expenditures of state and local governments.  The components of total expenditures that we analyze in our paper are current operations and capital outlays.  This dataset ranges from 1977 – 2005.  We also have information on the balanced budget rules in each state.  The stringency of these rules is summarized by an index created by the Advisory Commission on Intergovernmental Relations.  Finally, we have data on state GDP for these years, as well as data on political characteristics and debt levels in each state.               

Since the cyclicality of fiscal policy is a choice variable of each state government, we run an instrumental variables analysis using the stringency of the balanced budget rules as an instrument for cyclicality.  Our IV regressions are all cross-sectional tests.  In our first stage, we regress cyclicality on the balanced budget rules.  Using this information, we then regress average state growth rates on cyclicality in our second stage. 

Results:

We draw two main conclusions from our analysis.  First, states with the strictest balanced budget rule implement policy that is significantly more pro-cyclical than states without the restriction.  The same result holds for states that have a higher ACIR index value.  It seems, though, that lenient balanced budget rules do not constrain a state government's ability to run counter-cyclical policy. 

Second, we find that a more counter-cyclical fiscal policy has a modest, yet positive impact on growth.  The estimated coefficient suggests that a state can increase its growth rate by approximately 0.4% per year if it relaxes the ex-post balanced budget restriction.  This result, while significant at the 10% level in our basic specifications, loses its statistical significance after controlling for each state's initial debt to GSP ratio in 1977.   These results are then compared to the existing literature on fiscal policy.