Economics Letters 118 (2013) 297–299 Contents lists available at SciVerse ScienceDirect Economics Letters journal homepage: www.elsevier.com/locate/ecolet School shootings and private school enrollment Rahi Abouk, Scott Adams ∗ Department of Economics, University of Wisconsin-Milwaukee, Milwaukee, WI 53201, United States article info Article history: Received 12 July 2012 Received in revised form 6 November 2012 Accepted 11 November 2012 Available online 24 November 2012 JEL classification: I20 Keywords: School shooting School enrollment Media effects abstract We find that school shootings are followed by a 10%–12% increase in private high school enrollment. The effects are most pronounced following shootings in nonurban areas, which is consistent with their more intense media coverage. © 2012 Published by Elsevier B.V. 1. Introduction School shootings, through excessive media coverage, create panic (Jemphrey and Berrington, 2000; Muschert, 2007) and generate a perception of a public health problem that exceeds its actual danger (Burns and Crawford, 1999). According to a post- Columbine survey, two-thirds of Americans believed that it is at least somewhat likely that a similar shooting could happen in their area (Saad, 1999). Such widespread fear may prompt reactions from the public that surpass the scope of these isolated events. We offer evidence of one such reaction to school shootings by showing a significant increase in private school enrollments the school year following a shooting. At the same time, public school enrollments decline. 1 We confirm that these reactions are likely media-driven. Since socioeconomic factors drive the nature of how school violence is covered by the media, with shootings in urban and predominantly minority schools receiving more limited coverage (see, e.g., Menifield et al., 2001), we would expect stronger effects to follow shootings in nonurban settings. This is supported by our data. We also find that the impact of school shootings is temporary, with a post-shooting effect only observed in the Fall following the shooting. No effects are found in subsequent years, confirming ∗ Corresponding author. Tel.: +1 414 229 4212; fax: +1 414 229 3860. E-mail address: sjadams@uwm.edu (S. Adams). 1 The only study of the effect of school shootings by economists that we are aware of showed that test scores for high school students fell following school shootings in Finland. The results were linked to post-traumatic stress syndrome (Poutvaara and Ropponen, 2010). that these enrollment decisions are likely heat-of-the-moment in nature. These results suggest school shootings serve as a viable source of exogenous variation in private school enrollments for future studies wishing to employ a natural experiment approach for assessing the economic implications of private school enrollment. 2. Basic empirical approach and data To assess the impact of school shootings on enrollments, we estimate weighted least squares regressions on state-level data, summarized by: lprivate it = α i + γ t + δ 1 SS it + δ ′ 2 X it + ε it (1) and lpublic it = α i + γ t + β 1 SS it + β ′ 2 X it + ε it . (2) These regressions are weighted by population, which allows for less weight to be placed on smaller states that have higher variability in enrollment. The variables lprivate and lpublic are logs of private and public school enrollment in state i in year t for the school year beginning in the Fall. State (α) and year (γ ) fixed effects are included. The variable SS is a dummy variable indicating that a shooting occurred in the previous academic year. We also separate the SS variable into urban shootings and nonurban shootings. Specifically, we define urban shootings as those in places with more than 100,000 people. 2 2 There were two cases (Antioch, CA and El Cajon, CA) that were close to the 100,000 cutoff but circumstances of the shootings clearly guided the classifications into urban and nonurban, respectively. 0165-1765/$ – see front matter © 2012 Published by Elsevier B.V. doi:10.1016/j.econlet.2012.11.009