![Sample our Law journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5942/TOC-Subject-Sample-2021-Law.jpg"})
Abstract
Prior research has examined the effects of social welfare on levels of crime in the U.S. The emphasis has been on how governmental support in the form of social spending may influence crime levels because it serves as a mechanism to lessen the impact of adverse financial circumstances that may stimulate criminal behavior. This study builds on prior research by assessing the impact of another mechanism that affects the distribution of economic resources: the structure of the tax system. Specifically, we incorporate the “Suits Index,” a novel measure of the progressivity of the tax system. Our analysis examines the effects of changes in the Suits Index on changes in robbery and burglary rates using data for the 50 U.S. states between 1995 and 2017. Results indicate that movement towards greater progressivity in the tax structure was associated with decreases in robbery and burglary rates, net of changes in social spending.
Keywords:
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 Data points were available for all measures under scrutiny with the latest year of 2017. The exceptions were alcohol consumption, social spending, and the Suits Index. Specifically, data on alcohol consumption and social spending were available until 2016 only. Accordingly, for 2017, we used information from 2016 for these three measures. The 2017 estimate for the Suits Index was interpolated from 2015 & 2018 data, as explained below.
2 Information necessary to compute the Suits Index was not available for 2017. Given that data on this measure were available for 2015 and 2018 (the latest available year prior to 2018), we used the Suits data from 2015 and 2018 to estimate the 2017 value (Davis et al., Citation2015; Wiehe et al., Citation2018).
3 We are very grateful to O’Brien for providing data on the Suits Index for years 2002, 2005, and 2007 along with formulas necessary to calculate more recent estimates.
4 Substantively similar results were obtained when only state direct expenditures on welfare per capita or total state direct expenditures per capita were included in the alternative model specifications. While previous work used a wide range of welfare measures, as discussed earlier, there is no other program with data available at the national level that has been consistently implemented from 1995 to after 2010. The effect of such programs could potentially vary over time, while some of the other programs either emerged in the mid-2000s or finished by the mid-2000s (e.g. AFDC/TANF). In recent years, the state-level programs took over; however, the reporting on those programs is inconsistent and incomplete, precluding their use as controls even if interacted with time. Accordingly, we incorporated food stamp recipient data in the present analysis, which has been consistently collected by the CPS during the period under investigation.
5 The data on the percentage of children in female-headed households were not collected for 1995 and 2002. Accordingly, we used linear interpolation to impute values.
6 Another approach to control for time is to include a time trend in the analysis. Such an approach removes the dynamics of crime rates in each state. The benefit of a model with fixed effects for year is its nonparametric form, which means that a fixed effects model does not specify the precise temporal dependence of the omitted variable bias (Greenberg, Citation2014). Given the objectives of our research, we proceeded with fixed effects for time as a suitable strategy. In additional analyses, we controlled for time with time trend included in the model (instead of fixed effects for time). The results were substantially the same as those presented in the manuscript (results available upon request).
7 We also computed Variance Inflation Factors (VIF). The highest value is 3.2 for immigrant concentration in both models in Table 4. This is well within an acceptable range given the sample size and other features of our data (see O’Brien, Citation2007).
8 Phillips, (Citation2006) offered the illustration of how region might be implicated in such omitted variable bias. Strictly speaking, region is a time-invariant measure. However, it may be a proxy measure of time-varying factors related to region that are excluded from the within-unit models and may thus contribute to different between- and within-unit associations with crime rates.
Additional information
Notes on contributors
Sylwia J. Piatkowska
Sylwia J. Piatkowska is an assistant professor in the College of Criminology and Criminal Justice at Florida State University. Her areas of interest include hate crime, suicide and suicidal behavior, immigration and crime, and both international and comparative criminology.
Steven F. Messner
Steven F. Messner is Distinguished Teaching Professor of Sociology at the University at Albany, State University of New York. His research focuses on social institutions and crime, understanding spatial and temporal patterns of crime, and crime and social control in China.
Colin Gruner
Colin Gruner received his PhD from the University at Albany, SUNY. His research focuses on neighborhood correlates of crime and the relations between welfare policy and crime rates. The opinions, findings, and conclusions expressed in this publication are those of the author and not their employer.
Eric P. Baumer
Eric P. Baumer is a professor of sociology and criminology at Pennsylvania State University. His research explores demographic, temporal, and spatial patterns of illicit activity; the mobilization of law; and the application of criminal justice sanctions.