Article Abstract Online ISSN: 1099-176X    Print ISSN: 1091-4358 The Journal of Mental Health Policy and Economics Volume 7, Issue 4, 2004. Pages: 191-205 Published Online: 13 Dec 2004 Copyright © 2004 ICMPE.

* Correspondence to: Joseph J. Sabia, 22 Webcowet Rd, Arlington, MA 02474, USA
E-mail: Jjs10@cornell.edu

Source of Funding: Department of Policy Analysis & Management, Cornell University, Ithaca, NY 14853-4401, USA

Abstract

A recent high-profile murder case and the passage of the Unborn Victims of Violence Act have focused policymakers’ attention on domestic violence against pregnant women. The well-established link between men’s alcohol consumption and spousal abuse has led some to suggest that more stringent alcohol regulations could ameliorate domestic violence.  I examine this hypothesis using two waves of the Fragile Families and Child Well-Being Study. I find that while men’s alcohol consumption is positively associated with domestic violence against new mothers, there is little evidence that this effect is large in magnitude.  Bivariate probit, two-stage least squares, and fixed effects estimates reflect that unobservables correlated with both drinking and domestic abuse may be more important in explaining the drinking-abuse relationship.  Moreover, I find that higher liquor taxes and state restrictions on alcohol supply are not effective policies to ameliorate domestic violence.

Background: A recent high-profile murder case and the passage of the Unborn Victims of Violence Act have focused policymakers' attention on domestic violence against pregnant women and new mothers. The link between men's alcohol consumption and spousal abuse has led some to suggest that more stringent alcohol regulations could ameliorate domestic violence. Aim of Study: (i) To examine the correlation between men's alcohol consumption and domestic violence against new mothers and test how sensitive the correlation is to assumptions about unobserved heterogeneity, (ii) To test whether higher liquor taxes and more stringent alcohol control regulations are associated with a lower incidence of domestic abuse. Methods: Using the Fragile Families and Child Wellbeing Study, I estimate ordinary least squares, bivariate probit, two-stage least squares, and fixed effects models to test the relationship between alcohol consumption and domestic violence. Results: My findings suggest that while there is a strong positive association between men's alcohol consumption and the commission of domestic violence against new mothers, this correlation is highly sensitive to assumptions about unobservables. There is little evidence that higher liquor taxes or more stringent alcohol regulations will significantly reduce domestic violence. Discussion and Limitations: The empirical results suggest evidence for an ‘unobserved bum hypothesis’ That is, unobservable characteristics of the father may be correlated with both the likelihood that he abuses pregnant women (or new mothers) and that he drinks heavily. While the findings of this paper cannot rule out the possibility that men's alcohol consumption has some effect on domestic violence, there is little evidence to suggest that the impact is large in magnitude. Moreover, there is little evidence that higher liquor taxes or stricter alcohol supply regulations reduce the incidence of domestic abuse. However, greater policy heterogeneity across states and over time would be beneficial in further exploring this issue. Implications for Health Policy: Alcohol regulations, such as higher liquor taxes, are rather ineffective policies at reducing domestic violence against pregnant women and new mothers. Moreover, because policies that regulate alcohol availability or tax its consumption will harm non-violent drinkers, such policies may also be target inefficient. Rather than using non-credible, indirect mechanisms such as alcohol regulation, increasing criminal penalties for harming pregnant women or their unborn children may be a more direct method reducing domestic violence. Implications for Further Research: Future research that attempts to estimate the impact of alcohol consumption on the likelihood of domestic violence must carefully consider how unobservables may impact both drinking and abuse. Future work should utilize a longer panel dataset with greater within-state alcohol policy variation to allow for more robust tests of the impact of alcohol regulations on the prevalence of domestic violence.

Received 25 August 2004; accepted 3 November, 2004

Copyright © 2004 ICMPE