Abstract
We consider whether there is statistical evidence for a causal relationship between federal government expenditures and growth in real GDP in the United States, using available data going back to 1791. After studying the time-series properties of these variables for stationarity and cointegration, we investigate Granger causality in detail in the context of a Vector Error Correction Model. While we find causal evidence that faster GDP growth leads to faster growth in government spending, we find no evidence supporting the common assertion that a larger government sector leads to slower economic growth.
Highlights
Is a larger government bad for growth? Much of the current U.S debate over economic stimulus versus debt reduction assumes that public spending dampens economic growth in the long run, and tax cuts are a more effective policy
A common methodological pitfall in the literature that tries to uncover the causal links between government size and long-term economic growth is that these studies regularly conduct Granger causality tests outside the cointegration framework, though Jones and Joulfain [19], Ghali [20] and Islam [17] are among the exceptions
Stationarity and cointegration are the first steps of analysis, because if the variables turn out to be non-stationary in their levels and are cointegrated, the Vector Error Correction (VEC) model is the appropriate analytical tool to use
Summary
Much of the current U.S debate over economic stimulus versus debt reduction assumes that public spending dampens economic growth in the long run, and tax cuts are a more effective policy Both the theoretical effects of government size and the empirical record are much more mixed than the US political dialogue asserts. A common methodological pitfall in the literature that tries to uncover the causal links between government size and long-term economic growth is that these studies regularly conduct Granger causality tests outside the cointegration framework, though Jones and Joulfain [19], Ghali [20] and Islam [17] are among the exceptions. We focus on federal government outlays, not government revenues—which are more clearly driven in the short-run by changes in GDP—or the financing through debt In the latter case, the fiscal and time-series properties were thoroughly explored by Kremers [23], using data from 1920-1985. We conclude with a brief summary and suggestions for further research
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
