What Accounts for the Changes in U.S. Fiscal Policy Transmission?

Article excerpt

ONE OF THE MOST PROMINENT ISSUES IN macroeconomics concerns the effects of an increase in government spending. The topic takes center stage in the policy debate and has received great attention in the theoretical literature at least since Keynes' General Theory. Recently, empirical research on this issue has flourished, as well. In a seminal study based on vector autoregressions (VARs) for a long postwar sample, Blanchard and Perotti (2002) provide evidence indicating a positive response of consumption and output to a one-time fiscal shock. Specifically, the authors analyze U.S. time series data from 1960 to 1997 and find a spending multiplier for consumption between one-third and one. Similar findings are also reported by Fatas and Mihov (2001) and Gali, Lopez-Salido, and Valles (2007). Other more recent empirical studies, however, suggest that the transmission of fiscal policy shocks may have actually changed around the early 1980s. Indeed, both Perotti (2005) and Mihov (2003) provide fresh VAR-based evidence for the United States showing a substantial reduction in the expansionary effects of spending shocks after 1980.

What accounts for the changes in the transmission of U.S. fiscal policy over time? The fact that the aforementioned studies point to a break around 1980 suggests several interesting hypotheses. First, it is widely accepted that the conduct of U.S. monetary policy differed substantially before and after the 1980s. This change may have affected more generally the transmission of shocks in the economy. A second hypothesis draws on the observation that fiscal policy itself has changed. Perotti (2005), for example, reports that a typical shock to government spending displays much less persistence in the more recent period. A third explanation stresses the role of private consumption behavior by pointing out the possible consequences arising from increased asset market participation. In fact, retail financial markets were subject to significant restrictions until the late 1970s. Bilbiie and Straub (2006) argue that these restrictions may have effectively prevented a large fraction of households from smoothing consumption in the desired way. Shut out from asset markets, such households would tend to exhibit an extreme version of "Keynesian" consumption behavior, where current consumption perfectly tracks current income, as has been suggested in recent papers by Gali, Lopez-Salido, and Valles (2007) and Bilbiie and Straub (2004). To the extent that this explains the strong crowding-in effects of government spending documented for the 1960s and 1970s, one may conjecture that the change in fiscal transmission around 1980 is critically related to the financial liberalization occurring at the same time. (1) Specifically, deregulation and financial innovation may have widened private access to asset markets, reducing the number of households who fail to smooth their consumption profiles (in response to government spending shocks). Thus, the evolution of household finance competes with changes in monetary and fiscal policies as candidate explanations for the observed decline in the effects of U.S. government spending.

The objective of the present paper is to evaluate the relative importance of these different causes. A better understanding of how and why fiscal transmission changed during the early 1980s seems valuable in its own right but also with respect to the more general changes in business cycle behavior that have come to be called the "Great Moderation." Kim and Nelson (1999) and McConnell and Perez-Quiros (2000) were the first to highlight a marked decrease in the volatility of economic activity since the mid-1980s. Several subsequent papers, including Stock and Watson (2004) and Ahmed, Levin, and Wilson (2004), have attempted to explain the sources of this phenomenon, examining some of the same aspects that we focus on, notably, changes in macroeconomic policies and the behavior of the private sector. …