Domestic Violence Rises 10 Percent After NFL Upsets

By Gordon Dahl, Ph.D.

Violence by men against members of their own families is one of the most common yet perplexing forms of criminal behavior. While there are several explanations for family violence, one of the leading theories is that family violence arises unintentionally when emotional situations escalate out of control. This interpretation suggests an important role for emotional cues in precipitating violence. According to recent theories by economists and psychologists, the strongest emotional cues should come from events that are both unexpected and upsetting.

[Photo: Gordon Dahl, Ph.D.]

Football games are emotionally laden events of widespread interest, typically garnering 25 percent or more of a local television viewing audience. Could the disappointment of an unexpected loss raise the risk that football fans react inappropriately and abuse their spouses or partners? In a recent paper co-authored with Professor David Card at UC Berkeley, we set out to answer this question.

We analyze the effects of the emotional cues associated with wins and losses by local professional football teams, using police reports of family violence during the regular season of the National Football League (NFL). We hypothesize the risk of violence was affected by the “gain-loss utility” associated with game outcomes around a rationally expected reference point. In layman’s terms, the gain-loss theory says two things: (i) an individual’s happiness depends not on actual outcomes, but on outcomes judged relative to expectations, and (ii) unexpected disappointments have a larger impact compared to pleasant surprises.

Gain-Loss Utility: Emotions and Expectations

To illustrate this theory, think about what happens when you get a 3 percent raise when you expected 5 percent. If you’re like most people, you’re disappointed and in a bad mood after hearing the news. In contrast, now think about what happens when you get a 3 percent raise you were expecting. You probably aren’t especially happy or sad, since the raise was completely anticipated. This example demonstrates the first prediction of the gain-loss theory: Your expectation (or “reference point”) plays an important role in modulating your emotions. Now take the scenario one step further. What happens when you get a 5 percent raise but only expected 3 percent? The theory predicts your happiness increases due to the unexpected good news, but not by as a large an amount as your happiness decreases when you receive unexpected bad news.


It has proven difficult to test this theory of “gain-loss utility” outside of laboratory experiments. Even inside the lab, it is difficult to create emotionally charged expectations and then crush them (or exceed them) to see how violent behavior changes. So, to test the theory of gain-loss utility in the context of domestic violence, we take advantage of what might be called a “natural experiment.” We use the emotional cues created by unexpected wins and losses of NFL football broadcasts and analyze how police reports of domestic violence change in areas near a local NFL team.

A key feature of our natural experiment is that there is a well-organized betting market that allows us to infer the expected outcome of each game. We use the point spread from these betting markets to construct the reference point for gain-loss utility. We compare the pregame betting odds to the game results of regular-season games between 1995 and 2006 for six NFL teams: the Carolina Panthers, Detroit Lions, New England Patriots, Denver Broncos, Kansas City Chiefs and Tennessee Titans. This information is matched to records collected from 763 jurisdictions in the relevant states from the National Incident-Based Reporting System (NIBRS), a database of local police reports.


We find that police reports of male-on-female intimate partner violence rise 10 percent in areas where the local NFL team lost a game they were favored to win by four or more points (i.e., after an upset loss). This spike is relative to weeks the local NFL team did not have a game, and the result controls for both the pregame point spread and the size of the local television viewing audience. The increase in violence after an upset loss are concentrated in a narrow time frame at the end of the game, as one might expect if the spike in violence is due to a transitory emotional shock. Consistent with reference-point behavior, losses when the game was expected to be close do not have a significant effect on family violence. Upset wins (i.e., victories when the home team was expected to lose) also have no significant impact on the rate of violence, suggesting an important asymmetry in the reaction to unanticipated losses and gains.

Interestingly, the pattern is most pronounced for games that are more emotionally charged. Upset losses to a traditional rival result in a 20 percent increase in police reports compared to an 8 percent increase for upset losses to a nonrival. Violence also increases by a larger amount when the team is still in playoff contention or has a particularly frustrating performance (i.e., a large number of sacks/turnovers or an excessive number of penalties).

A related question is whether individuals update their expectations (i.e., their reference points) slowly or quickly. If people update their expectations quickly, then there is more opportunity for a series of small surprises and less opportunity for large surprises. We find no evidence for reference-point updating based on the halftime score, even though the halftime score is more predictive of the final game outcome than the pregame betting odds. Apparently, in this setting at least, people are slow to update their expectations when presented with new information.

Managing Expectations

Our findings suggest that emotional cues based on the outcomes of professional football games exert a relatively strong effect on the occurrence of family violence. The estimated impact of an upset loss is comparable to the effect of a hot day, or about one-third as large as the jump in violence on a major holiday like the Fourth of July. The magnitude of the cueing effect attributable to an upset loss is especially large, considering that only a fraction of the population are serious football fans.

From a policy perspective, our paper suggests that better awareness and management of expectations could help to reduce violence within families. Managing expectations could also be very useful when it comes time for your next salary raise!

The complete paper was published in The Quarterly Journal of Economics.