Poster Paper: The Effects of Fiscal Stimulus On Crime

Saturday, November 9, 2013
West End Ballroom A (Washington Marriott)

*Names in bold indicate Presenter

Filippos Petroulakis, University of Maryland, College Park
Throughout the 2000s, crime had been on a steady slow decline since its last peak in the wake of the dot com recession. Yet this trend dramatically accelerated during the Great Recession, during which the rate of change of crime fell precipitously and across the board in the United States after 2008. The growth of property crime rate fell from an average of around -2% a year from 2002 to 2008, to -5.3% in 2009 and another -3.3% in 2010. These numbers not been seen for a decade, since the unprecedented drop in crime the 1990s.

Property crime rate fell a median of 6.5% in 44 of the 49 continental states (including the District of Columbia) in 2009, and in a further 36 in 2010, following a leveling in 2008. This decline came at a time of devastating economic decline and huge spikes in joblessness. Past experience concerning the relationship between crime and macroeconomic conditions indicates that this is rather uncommon: the correlation between the unemployment rate and the property crime rate is almost 0.60 from 1960 to 2007, and falls to only 0.21 once we include 2008-2010.

This moderation in the falling trend in 2008, the first year of the recession, and the steep decline thereafter coincides with the response of the federal government to the crisis, and poses the natural question of whether government action assisted in this decline. I investigate the impact the American Recovery and Reinvestment Act (ARRA), more commonly known as the Obama stimulus, had on crime. The hypothesis is straightforward: ARRA directed federal funds to boost social support and employment in areas most hit by the recession, and by doing so prevented crime from rising, and even expedited its decline.

I use county level crime data from the FBI, and county level unemployment and retail wage data (an commonly used proxy for the wages of low skilled men) from BLS, and construct a short panel of approximately 850 counties. Using a fixed effects specification, I argue that stimulus had a modest but statistically significant negative impact on crime. Controlling for unemployment, retail wages, poverty, arrest, and other controls, a 10% rise in total per capita stimulus outlays resulted in a reduction in property crime rate of approximately 1.5%. When federal outlays are expressed as a fraction of county GDP lagged one period, a 10% rise in outlays over lagged county GDP leads to a reduction in crime of approximately 0.5%. I conduct several different checks and conclude that my results are robust. To my knowledge, this is the first paper to have analyzed the effects of government stimulus on crime.