Academic journal article The American Journal of Economics and Sociology

From the Arena into the Streets: Hockey Violence, Economic Incentives and Public Policy

Academic journal article The American Journal of Economics and Sociology

From the Arena into the Streets: Hockey Violence, Economic Incentives and Public Policy

Article excerpt



Over the last two decades a large multi-disciplinary literature has emerged concerned with the causes and effects of, and the policy initiatives directed towards rectifying violence in professional team sport.(1) Although the contribution of economics to this literature has been modest, recent research on the National Hockey League (NHL) has focused on treating violence as a "goods characteristic," an attribute of the product deliberately fostered by teams to generate revenue in their drive to maximize profits.(2)

Specifically, one "goods characteristic" hypothesis is that there is a direct positive relationship between violence and game attendance (reflecting spectator preference for hockey as a "blood sport"), and, hence, a positive relationship between violence and revenue. If this hypothesis is correct, the two standard methods for controlling hockey violence - allowing the league to practice self regulation, and/or prosecuting in the courts the immediate perpetrators, the players, - are unlikely to succeed. Self regulation will not be effective if there is an incentive to promote violence, and target the team's "agents", the players, occasionally in the courts will not produce the desired effect as long as team incentives for violence exist and teams go unchallenged by the law.(3) But, if neither self regulation nor the judicial system are effective constraints on violence, what is the alternative? The answer is not self evident.

Accordingly, the object of this paper is to consider the policy dilemmas associated with controlling violence in the NHL in light of the potential existence of team incentives to promote violence. Specifically, we empirically test for the hypothesized positive violence - game attendance relationship. If the hypothesis is validated its implications for the use of alternative policy instruments to control violence will be considered.

A demand model of NHL game attendance will focus on the violence-attendance relationship. The data set is basically game by game data for the 1989/90 season, the last period for which all data necessary to estimate the model were available. The major conclusions are: a positive violence-attendance relationship exists, although it is principally an American, not a Canadian, phenomenon; and, since self regulation and the courts in their present forms do not appear to be effective instruments of violence control, consideration should be given to putting a more direct mechanism in place.

The remainder of the paper is divided into three parts: II. the model is specified, estimated and the results analyzed; III. control alternatives in light of II. are discussed; and IV. conclusions are drawn.


The Model and the Results

The empirical model of team behavior adopted here consists of a simultaneous system defined by the following two reduced form equations:

ATT = [[Alpha].sub.0] + [[Alpha].sub.1] CAN + [[Alpha].sub.2] POP + [[Alpha].sub.3] INC + [[Alpha].sub.4] HTRL + [[Alpha].sub.5] VTRL + [[Alpha].sub.6] UT3 + [[Alpha].sub.7] UT3B3 + [[Alpha].sub.8] PLD4Q + [[Alpha].sub.9] DAY + [[Alpha].sub.10] VIOL + [e.sub.1] [1]

p = [[Beta].sub.0] + [[Beta].sub.1] CAN + [[Beta].sub.2] POP + [[Beta].sub.3] INC + [e.sub.2] [2]

The endogenous variables are attendance (ATT) and price (p). The exogenous variables are either location specific representing games played in Canada (CAN), population (POP) and income (INC); or team specific, representing, success of home and visiting teams respectively, (HTRL and VTRL), high and low outcome uncertainty (UT3 and UT3B3), play-off drive (PLD4Q), games played on the weekend (DAY) and violence (VIOL). The error terms are [e.sub.1] and [e.sub.2]. The variable definitions are set out in Table 1 and the following points should be emphasized about the specification, content and estimation of the model.

The reduced form model is derived from a recursive structural form in which attendance, as in equation [1], is dependent on price and price is determined by equation [2]. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.