Cyclicality of U.S. Discretionary Fiscal Policy (Job Market Paper)
The literature on fiscal policy conduct has failed to yield a consensus on even the most basic aspects of policy, such as whether and how it responds to the business cycle. Conflicting results may stem in part from model uncertainty, particularly uncertainty about which covariates belong in the underlying model of fiscal policy. I estimate the response of U.S. federal discretionary policy to different business cycle measures using a Bayesian framework that explicitly accounts for model uncertainty. I find that policy is countercyclical, responding primarily to the change in the unemployment rate, and that taxes make up a larger portion of the response than spending. Distinguishing between expansions and recessions makes it clear that countercyclical policy is limited to recessions; during expansions, in contrast, policymakers are unlikely to respond to economic conditions. Finally, I find no evidence of a structural break in business cycle responses, nor of substantive differences between intended policy and actual policy outcomes.
Regime Switching in U.S. Discretionary Fiscal Policy Conduct
There is some evidence to suggest that discretionary fiscal policymakers vary policy responses according to the state of the business cycle and debt levels. However, these findings rely on strong assumptions about the structure of the underlying model. I use Bayesian model comparison techniques to compare linear and threshold models that differ with respect to the included covariates and, for threshold models, the threshold variable and value that govern regime switching. I consider twenty-nine possible threshold variables related to the business cycle, debt, monetary policy, inflation, and political environment. My results provide strong evidence that discretionary fiscal policy is regime dependent. Among the threshold variables that I consider, the change in nonfarm payrolls and the change in the federal funds rate are the most probable. Threshold value posteriors for both of these threshold variables split observations into regimes consistent with periods of "normal" and "bad" economic performance that roughly coincide with recession dates.
Direct and Indirect Monetary Policy Responses to Tax and Spending Shocks
The effectiveness of fiscal policy depends critically on the central bank's response. In fact, a central bank that engages in monetary offset exactly counteracts the macroeconomic effects of fiscal shocks, rendering them an ineffective means of stabilizing the economy. In this paper I investigate empirically whether the Federal Reserve offsets fiscal policy through the estimation of a ``direct response" to tax and spending shocks that measures the federal funds rate response to a shock different from what is justified by the shock's predicted effects on the economy. I find evidence of non-offsetting behavior as well as asymmetries in the way the Federal Open Market Committee (FOMC) responds to different types of shocks. Specifically, the FOMC reinforces the anticipated effects of spending shocks, for example responding to an increase in spending by decreasing the federal funds rate, and impedes the anticipated effects of tax shocks, for example responding to a decrease in taxes by increasing the federal funds rate. The direct response to expansionary shocks is larger and more persistent than the response to contractionary shocks, as is the response to local spending relative to federal spending. My results also indicate that the direct response to tax shocks has declined in recent years while the response to spending is largely unchanged.
Works in Progress
Cyclicality of U.S. Subnational Discretionary Fiscal Policy
Subnational fiscal policy operates under markedly different conditions from federal policy. Most notably, unlike the federal government all states (except Vermont) are subject to balanced budget requirements, which limit their ability to run deficits. As a result, subnational fiscal policy is likely to be less countercyclical and more responsive to debt than federal policy. There is evidence to suggest that this was true during the Great Recession, when seemingly large stimulus by the U.S. federal government was accompanied by contractionary policy at the subnational level. This paper aims to estimate the response of subnational fiscal policy to business cycle measures using Bayesian techniques that incorporate uncertainty about the underlying model of fiscal policy.