Family violence is an increasingly recognized problem in the U.S. and other countries, with substantial physical, psychological, and societal costs. Although overall rates of family violence are difficult to measure, estimates for the U.S. range from 1 to 3 million annual incidents of violence against an intimate partner (U.S. Department of Justice, 1998). Yet despite several decades of research and analysis, there remains considerable debate about the appropriate conceptual model of family violence and about the efficacy of interventions that are derived from alternative theoretical perspectives. We propose to advance the study of family violence by assembling and analyzing a comprehensive database of domestic violence incidents from the National Incidence Based Reporting System (NIBRS). NIBRS includes all incidents reported to police -- not just those that result in an arrest -- in participating jurisdictions across the country. We will use these data to conduct a comprehensive descriptive analysis of the 2.5 million family violence incidents reported from 1994 to 2004. Our findings will be an important complement to existing studies that are mainly based on retrospective reports of victimization. Our preliminary analysis indicates that NIBRS data imply a higher incidence rate of intimate partner violence against women compared to other data sources, and a much smaller trend reduction in the risk of family violence over the 1995-2004 period. These data will allow us to fill many important gaps in knowledge about the situational determinants of family violence and to re-assess conclusions that have been drawn from other studies. We will also use the NIBRS data to study the effect of unexpected emotional cues on family violence. This analysis will provide one of the first rigorous tests of alternative micro-level causal models of violence. Specifically, we will combine the NIBRS data with information on the dates, betting odds, and final outcomes for NFL (professional football) games played during our sample period. We will then evaluate the impact of unexpected game outcomes (e.g., upset wins and losses) on rates of family violence. Standard rational choice models of family violence -- including models in which violence is valued directly by perpetrators, and models in which violence plays an instrumental role in family bargaining -- predict that unexpected emotional cues will have no impact on family violence. In contrast, behavioral decision-making models that allow emotional triggers or visceral factors to affect preferences suggest that rates of family violence will be affected by emotional cues. A preliminary analysis of NIBRS data for the state of Colorado reveals an important positive effect of upset wins and losses. In particular, we find a 37 percent increase in family assaults on days in which the local professional football team suffers an unexpected loss. Our study of family violence has a strong relevance for public health. Violence within the family results in over 2 million injuries and 1,700 deaths in the U.S. every year (Centers for Disease Control and Prevention, 2003, and Department of Health and Human Services, 2005).
Our study of family violence has a strong relevance for public health. Violence within the family results in over 2 million injuries and 1,700 deaths in the U.S. every year (Centers for Disease Control and Prevention, 2003, and Department of Health and Human Services, 2005).
| Card, David; Dahl, Gordon B (2011) Family violence and football: the effect of unexpected emotional cues on violent behavior. Q J Econ 126:103-43 |
