Atl Econ J (2011) 39:303311 DOI 10.

1007/s11293-011-9275-1

Determinants of Attendance in the Quebec Major Junior Hockey League: Role of Winning, Scoring, and Fighting
Rodney J. Paul & Andrew P. Weinbach

Published online: 29 May 2011 # International Atlantic Economic Society 2011

Abstract Attendance in the Quebec Major Junior Hockey League was studied for the 2009 to 2010 season. This junior league, part of the Canadian Hockey League system, serves as a development league for teenage players who attempt to parlay their participation in this league into a professional career. Fan demand for this level of hockey is found to be sensitive to the success of the home team and to exhibit normal consumer responses to weekday and monthly effects with weekends being more popular and attendance increasing throughout the season toward the playoffs. On-ice factors such as scoring, a proxy for excitement, and fighting are not shown to have a significant effect on attendance. Keywords Attendance . Sports . Hockey . Demand . Violence

Introduction The economic study of sports, once thought to be confined to antitrust law and labor contracts, took on a much different context after the introduction of Sportometrics by Goff and Tollison (1990). Economists now routinely study sports as a laboratory of general economic behavior by investigating the actions of players, coaches, referees, general managers, and team owners. Topics such as individual and managerial incentives, the organizational behavior, income distribution, competitive structure, and the efficient markets hypothesis were studied with sport-specific and general economic insights and conclusions obtained. An additional group which can also be examined through Sportometrics-style studies is consumers of sports; the fans themselves. Estimating consumer demand for sports is important to the overall
R. J. Paul (*) Department of Finance, St. Bonaventure University, Allegany, NY, USA e-mail: rpaul@sbu.edu A. P. Weinbach Coastal Carolina University, Conway, SC, USA

304

R.J. Paul, A.P. Weinbach

business of sports and to the individual teams and leagues involved. Major and minor league sports have been investigated to attempt to determine what factors attract fans to spend their disposable income on sports-related entertainment. Attendance in the sport of hockey has not been studied as often as other sports, particularly baseball, where economists have always had a keen interest. Attendance studies of hockey have mostly focused on the effects of fighting and rule changes. These studies of hockey include Jones (1984), Jones et al. (1993), Jones et al. (1996) and Paul (2003) in relation to the highest professional level of hockey, the National Hockey League (NHL). Recently, more general studies of hockey attendance have begun to surface including studies by Hong (2009) and Rascher et al. (2009). Hong (2009) focuses on the different between casual and avid fans and the factors which attract them to hockey games. Rascher et al. (2009) estimated the impact of the NHL lockout on minor league hockey and the National Basketball Association (NBA). This study expands the investigation of the determinants of hockey attendance to major junior hockey. The Canadian Hockey League (CHL) consists of three separate leagues: the Ontario Hockey League (OHL), the Western Hockey League (WHL), and the Quebec Major Junior Hockey League (QMJHL). These leagues consist of many of the top teenagers in Canada and imports from around the world. These leagues directly compete for talent with National Collegiate Athletic Association (NCAA) collegiate hockey in the United States, with the majority of the top players typically choosing the junior hockey route over collegiate hockey (with some notable exceptions). Major junior hockey is closer in stature to the NHL and its minor leagues as the CHL leagues allow fighting within their contests and generally are considered to have a more physical style of game than the U.S. college system. Due to its similarities with the NHL, coupled with the fact that these players are only teenagers, not full adult professionals, allows for a study of fan preference for hockey attributes such as winning, scoring, and fighting in this setting. This study specifically studies the Quebec Major Junior Hockey League for the 2009 to 2010 season. Each team in the QMJHL played 68 regular season games. Data on attendance, team records, scoring, and fighting were taken from box scores available on the official league website, www.lhjmq.qc.ca. Our study attempts to determine how sensitive fans of major junior hockey in Quebec (and surrounding areas) are to the key game attributes of hockey. This information will help shed light on if hockey fans have universal preferences across leagues or if there are some attributes which fans of junior hockey, specifically junior hockey in Quebec, find more or less favorable than fans of other levels of hockey or in other areas of the country.

Regression Model for 2009 to 2010 QMJHL Attendance Game-by-game attendance for each contest was gathered from box scores on the QMJHL website, www.lhjmq.qc.ca. For each game, the final score, the number of fights, and basic information such as the month of the year and the day of the week was taken from the league schedule and individual box scores. The win percentage variable included in the regression model is calculated as a percentage of the total number of points possible within the QMJHL hockey structure. Each win earns a

Quebec Major Junior Hockey League

305

team two points, each overtime loss or shootout loss earns a team one point, and a regulation loss gives a team zero points for that game. The win percentage variable is calculated on a running basis by dividing the number of points earned by the total number of points possible (two times the number of games played). The scoring variable, total goals per game, is calculated as a running average throughout the season (for any individual game it is the total number of goals per game for that team going into the current days game). The total goals per game is calculated by adding the goals for per game average to the goals against per game average for each team. This variable serves as a proxy for excitement within the game as higher-scoring games are expected to be more exciting than lower-scoring games. Including only goals for per game, for each team, was considered, but goal scoring is highly correlated with winning, as teams need to score goals to win games. By taking the sum of goals for and goals against, for each team, it differentiates between teams which play high-scoring games from those that play lower-scoring games, while not creating the multicollinearity problems that the inclusion of goals for average itself creates. For the number of fights, this information was taken from the penalty section of the box score, where a single fight consisted of two players (one on each team) receiving a five-minute major penalty for fighting. In rare cases where one player was given a fighting major and no corresponding player on the opposing team was given a major, a fight was also included in the data calculation. As mentioned in the introduction, the key variables of interest we wish to study are the effects of winning, scoring, and fighting at the Major Junior Hockey level in the Quebec League. With this in mind, the data is arranged as a pool, with each home team having a fixed effect within the pool relating to that city. Although we considered using explanatory variables to attempt to explain differences between cities, including variables such as income per capita, population, arena size, etc., the explanatory power of the regression was very weak. Using the pooled approach allows for the attributes of each of the QMJHL cities to be captured by the dummies for the home team in each game in this attendance model. While it would be quite useful to know the optimal size and income per capita of cities for major junior hockey, this information might be better understood by incorporating a study over a longer time frame than a single season. With our information from box scores for one season, such as fights and scoring, the pooled approach seemed to best allow the evaluation of these individual factors while allowing for the fixed effects to capture individual home team attributes. The regression model is specified using dummy variables for the months of the year of the season and the days of the week of the individual games. The monthly dummies attempt to capture fan interest as it changes throughout the season, with the likely effect being later months in the season (February and March) being more popular due to the push for the playoffs. September is the omitted dummy variable for the months of the season. Days of the week also have likely predictable effects with it being highly probable that weekend days are more popular than weekday games due to the opportunity cost of fans time during the work/school week. Thursday is the omitted dummy variable for the day of the week category. Individual dummy variables are also included for the road team in the hockey game. If certain road teams, likely the best teams, attract larger crowds when they are

306

R.J. Paul, A.P. Weinbach

the visiting team, then the individual dummy variables for the road team should capture this effect. Like the dummies included in the fixed effects of the home team, the excluded dummy category is the team from Acadie-Bathurst (first alphabetically), with all other results for the dummies compared to this team. Summary statistics for the non-binary variables are presented in Table 1. The regression results for the 2009 to 2010 QMJHL season attendance model is presented in Table 2. The results in relation to all variables are discussed following the results, with special interest in the key sports/hockey effects including winning, scoring, and fighting. To begin with the fixed-effects dummy variables, it can be seen that each individual home team has a statistically significant and different coefficient, compared to the omitted team dummy for Acadie-Bathurst. In relation to the road team dummies, the only significant results were found in relation to the best teams in the league for the 2009 to 2010 season. Statistically significant and positive effects on attendance were found for Quebec (additional 552 fans on the average) and Saint John (additional 406 fans on the average) when they were the visiting teams. Quebec and Saint John each won their respective divisions (Eastern and Atlantic). Given that the teams in the central and western divisions were not found to have a significant effect on attendance on the road, perhaps fans of teams in the eastern region of the QMJHL are more sensitive to the best road teams than fans of teams outside the eastern region. The month and day of the week dummies revealed expected results. QMJHL games became more popular as the season progressed, with significantly positive effects on attendance found in January, February, and March (compared to the omitted month, September, at the start of the season). Through these months, attendance increased for these hockey games, reaching its average peak during the month of March during the push for the playoffs. In relation to the days of the week, the middle of the week games were found to be the least popular, likely due to the opportunity cost of fans time during the work week. Weekend games were found to be popular, along, surprisingly, with games on Monday. Monday games were the least frequent during the season, with only five games played on this day of the week. Therefore, these results could be due to Mondays actually being popular game nights or due to the quality of the five games played on Monday during the season. In relation to the key variables of interest, the on-ice performance factors, fans of QMJHL hockey appear to respond to winning teams. The win percentage variable was found to have a positive and significant effect on attendance. Home teams which had higher win percentages attracted a greater number of fans. Fans at the junior league level appear to respond to the best teams, as is expected across many sports.
Table 1 Summary statisticsQMJHL hockey game results and statistics Attendance Mean Standard deviation Median 3386.93 2451.55 2846.50 Total goals per game average 6.86 0.73 6.86 Fights per game average 0.69 0.27 0.67

Quebec Major Junior Hockey League

307

Table 2 2009 to 2010 QMJHL hockey league attendance. Dependent variable: Attendance (per game) Independent variable Intercept October November December January February March Win Percentaget-1 Home team Baie-Comeau Cape Breton Chicoutimi Drummondville Gatineau Halifax Lewiston Moncton Montreal PEI Quebec Rimouski Rouyn-Noranda Saint John Shawinigan Val-dOr Victoriaville R-squared Adjusted R-squared 363.3364*** (2.8068) 1854.694*** (13.2330) 1742.149*** (15.9907) 1179.143*** (9.2080) 1102.999*** (9.6828) 4204.026*** (19.3931) 365.7436*** (2.9478) 3025.151*** (15.1458) 1308.055*** (8.8207) 459.3279*** (3.8268) 10587*** (40.8226) 1996.558*** (164.5872) 504.7993*** (4.2529) 2501.639*** (12.2525) 1777.371*** (17.2255) 408.5756*** (2.7401) 934.2417*** (7.8565) 0.9422 0.9372 Coefficient (T-Statistic) Independent variable Coefficient (T-Statistic)

735.8459** (2.1408) 21.2557 (0.2138) 89.8647 (0.9114) 244.0732** (2.1942) 248.6385** (2.4659) 441.1153*** (3.6591) 633.3499*** (5.1172) 524.4194** (2.4801)

Sunday Monday Tuesday Wednesday Friday Saturday Total goals per game averaget-1 Fights per game averaget-1 Visiting team Baie-Comeau Cape Breton Chicoutimi Drummondville Gatineau Halifax Lewiston Moncton Montreal PEI Quebec Rimouski Rouyn-Noranda Saint John Shawinigan Val-dOr Victoriaville Durbin-Watson Stat

389.7312*** (3.8460) 686.0159** (2.2928) 217.2258 (1.5351) 130.8287 (1.1566) 528.0551*** (5.2785) 603.9432*** (5.5066) 29.1921 (0.8254) 118.6830 (0.8653)

61.7351 (0.4021) 191.7763 (1.1307) 161.9510 (1.1463) 161.5152 (1.0102) 84.2258 (0.5707) 123.0253 (0.7059) 18.3313 (0.1193) 280.8926 (1.4100) 24.3774 (0.1862) 122.3703 (0.8126) 551.8083 (4.3034) 172.2907 (1.1354) 69.3463 (0.4887) 406.4225 (2.4574) 12.9913 (0.0854) 2.3498 (0.0167) 45.0450 (0.3371) 1.9584

On the other hand, teams which played higher scoring games or were involved in more fights did not attract a greater number of fans. Fans do not appear to come out in greater numbers when teams play in higher scoring games. Also, fighting does not appear to attract more fans in this league. This is in contrast with studies of the NHL, which have shown positive attendance effects due to fighting in hockey games (Jones 1984; Jones et al. 1993; Jones et al. 1996 and Paul 2003). Fans of junior hockey may not have the same preferences for fighting in hockey games as fans of the NHL, possibly due to the players in the junior league being teenagers, rather than adults. Fans may not wish to see younger developing players participate in fights in the same manner as grown men in the professional leagues. Other possible

308

R.J. Paul, A.P. Weinbach

explanations for this result include cultural differences as fans in Quebec may not enjoy seeing fighting within a hockey game as fans from other regions of Canada and the United States. Although there could be cultural differences in relation to fan likes and dislikes of fighting across Canada (and the world), there is some evidence that at least some fans in Quebec do like fighting in hockey. A Facebook page exists which is titled Keep Fighting in the QMJHL that receives ample attention. During the mid-1990s, some predominant NHL fighters of that era came from the QMJHL, where they often fought. This group includes Donald Brashear, Sandy McCarthy, Gino Odjick, and Georges Laraque. Therefore, although many believe the junior hockey game in QMJHL to not be as physical or have much fighting, there still appears to be quite a history of fighting in this leagues games. Some of the perception on fighting in the QMJHL may have changed in 2008 as tougher rules were introduced relating to bench-clearing brawls in games. These rules changes stemmed from a major event involving legal charges arising from a particularly bloody brawl between Quebec and Chicoutimi. Jonathan Roy, son of legendary NHL goalie Patrick Roy, had a savage attack on Bobby Nadeau (the opposing goalie in the game), which received considerable media attention in Canada, on SportsCenter in the U.S., and on computers around the world through frequent YouTube viewings. Rule changes aimed to diminish fighting and brawls included players who are deemed the aggressor in fights can be suspended up to 15 games at the discretion of a league disciplinary committee, players who fight for a second time during the same pause in a game will now receive a five-game suspension, players who instigate fights during pre-game warm-ups will receive a ten-game suspension, a player fighting a goaltender will receive a five-game suspension, and a goaltender who is considered an aggressor will receive an automatic ten-game suspension. The role of coaches in orchestrating fights also has come under more scrutiny as coaches may also receive suspensions with organizations fined when this occurs.1 In addition, players who receive a second fighting major in a game are ejected (unless the other player in the fight is deemed the instigator), and automatic suspensions exist for fights in the last five minutes of a game when the score differential is two goals or more. An automatic one-game suspension is given to a QMJHL player after four misconducts or match penalties. This is slightly different from the NHL, where a third instigator penalty in a season yields a two-game suspension, which increases with further instigator penalties later in the season (fourth instigator is a four-game suspension, fifth instigator is a six-game suspension, etc.) Due to the increased scrutiny on fighting and violence in the QMJHL, coupled with the stiffer penalties related to fighting, the regression results in Table 2 could still represent some short-term reaction to the memorable brawl a couple of years before this data was gathered. Fans may still wish to see fighting, although many may have been disgusted with the violence and media circus surrounding the Jonathan Roy events. Therefore, long-run preferences for fighting could be different than those shown in this sample. For the hockey year of 2009 to 2010 in the
1

The summary of these rules changes in the QMJHL was taken from an article in The Canadian Press from September 10, 2008 entitled QMJHL beefs up fighting penalties.

Quebec Major Junior Hockey League

309

QMJHL, however, fans did not appear to attend games in greater numbers for teams which fought more often.

What Attracts Fans to QMJHL Games The regression results from Table 2 above may be surprising to hockey fans and sports fans in general with respect to fan attendance. Hockey seems to rely heavily upon exciting goals, physical play, and memorable fights to attract an audience in a sportsentertainment market which offers many easily available substitutes, both in person and on television. With previous academic studies showing a preference for fighting, the question of why QMJHL fans do not appear to be affected by fighting or scoring demands further investigation. One likely reason that fans attend QMJHL games without being overly sensitive to the amount of scoring or fighting is due to the inherent cultural link of hockey to small-town (and big-city as well) Canada. Although we are not aware of many socioeconomic academic studies that have attempted to account for these underlying tastes, various books on Canadian culture shed some insight into the love of playing and watching hockey throughout Canada. Perhaps of more importance to the researcher delving into the world of junior hockey is that the town rink and its junior team is often the lifeblood that feeds and sustains the town. Throughout his classic hockey book The Game (Dryden 1983) and directly addressed in his follow-up book Home Game (Dryden and MacGregor 1989), legendary Montreal Canadiens Goalie Ken Dryden notes the importance of hockey to small-town Canada. In his books, he gives illustrations and quotes local citizens about how cultural activity flows through the local hockey rink during the long winters. He notes how towns suffer population losses or outright cease to exist soon after a local rink closes. When thinking of hockey in Canada, the introduction of Drydens Home Game seems to sum-up the underlying love and importance of playing hockey and following the local team in the typical Canadian town. He wrote, Hockey is a part of life in Canada. Thousands play it, millions follow it, and millions more surely do their best to ignore it altogether. But if they do, their disregard must be purposeful, done in conscious escape, for hockeys evidences are everywhereon television and radio, in newspapers, in playgrounds and offices, on the streets, in sights and sounds, in the feeling of the season. In Canada, hockey is one of winters expectations. He goes on to explain, Hockey is part sport and recreation, part entertainment, part business, part community-builder, social connector, and fantasy-maker.Others in newspapers, television, radio, and film also seem to echo this sentiment about hockey in Canadian life, including direct stories of hockey in small local areas in William Boyds Hockey Towns: Stories of Small Town Hockey in Canada (Boyd 1988). This underlying love and dedication to the game which often fuels sold-out local arenas, not just for playoff games, but everyday regular season games as well, is not only limited in scope to hockey in Canada, but is similar to the love of high school basketball seen in Indiana, the fervor for high school football in Texas and throughout the Southeast, and the love of football (soccer) in so many places around

310

R.J. Paul, A.P. Weinbach

the world. These underlying tastes and preferences, which may be quite difficult to explain and model, put fans in the seats at these arenas and stadiums. It is important to note, however, as this study of the QMJHL shows, that this love of the game does not translate directly into the same number of fans for each game. The simple laws of economics are still at play; as opportunity cost, scarcity, supply, and demand are evident at many levels. Fans still respond to the opportunity cost of their time as weekends are more popular days at the gate as opposed to weekends. Interest in games intensifies as the season progresses toward the playoffs. Also, as evidenced by the home team win percentage variable and the visiting team dummy variables, these junior hockey fans prefer watching successful teams play hockey as opposed to poorly-performing teams. More fans appear to become interested enough to buy tickets to a game as a team wins more often. Fans appear to stay at home when a team performs poorly. These fans may not wish to see losing hockey because they prefer to see their team win, but it could also show that these fans endure a painful experience when their team loses. In either case, some of the local audience appears to stop attending games when the home team is not successful.

Conclusion The study of QMJHL attendance for the 2009 to 2010 season revealed that this junior hockey league mimics many other professional and minor leagues in terms of its significant determinants. Attendance is highest on the weekends for these hockey games and attendance generally increases toward the end of the season during the rush for the playoffs and playoff positioning. Individual team dummy variables for the road team revealed that the best teams during the regular season attracted the largest crowds in opposing arenas. In relation to on-ice hockey-related factors, the QMJHL had some similarities and differences to the NHL and other hockey leagues. Fans of this junior league did respond to winning teams as the win percentage (calculated as points as a percentage of total possible points) of the home team was shown to have a positive and significant effect on per-game attendance. Apart from valuing winning teams, however, fans of the QMJHL did not seem to be influenced by teams which played higher scoring games (proxied by the sum of the goals-for average and goals-against average of the home teamto avoid multicollinearity problems of including only goals-for average due to the relationship between scoring and winning a hockey game) or had more fights. Both the scoring and fighting independent variables were not found to have significant effects on attendance. While fighting was found to increase attendance at NHL games in previous research, this did not appear to be an important factor in the QMJHL. This could be due to the younger ages of these players (teenagers) who fans may not wish to see fight in these games, or it could have something to do with regional preferences as fans in Quebec may not have the same preferences for fighting as other areas of Canada and the U.S. Overall, it appears from the results of the Quebec Major Junior Hockey League, that attendance in junior leagues in Canada (and possibly other places) is similar to other sports in that fans care about winning teams, enjoy attending games during the

Quebec Major Junior Hockey League

311

push for the playoffs, and attend more games when the opportunity cost of their time is lower (weekends). In relation to scoring and fighting, future studies of other leagues may help in the understanding if preferences for these variables are a functional of regional demand or other factors.

References
Boyd, W. (1998). Hockey towns: Stories of small town hockey in Canada. Toronto: Doubleday. Dryden, K. (1983). The game. Hoboken: Wiley. Dryden, K., & MacGregor, R. (1989). Home game. Toronto: McClelland & Stewart Inc. Goff, B. L., & Tollison, R. D. (1990) Sportometrics. Texas A&M University Economics Series; no. 11. Texas A&M University Press, College Station, TX. Hong, J. (2009). A comparison of motivational factors affecting attendance between avid and casual fans at minor hockey games. International Journal of Sport Management and Marketing, 5(12), 115131. Jones, J. (1984). Winners, losers, and hosers: demand and survival in the National Hockey League. Atlantic Economic Journal, 12(3), 5463. Jones, J., Ferguson, D., & Stewart, K. (1993). Blood sports and cherry pie: some economics of violence in the National Hockey League. American Journal of Economics and Sociology, 52(1), 87101. Jones, J., Stewart, K., & Sunderman, R. (1996). From the arena into the streets: hockey violence, economic incentives, and public policy. American Journal of Economics and Sociology, 55(2), 231 249. Paul, R. (2003). Variations in NHL attendance: the effect of violence, scoring, and regional rivalries. American Journal of Economics and Sociology, 62, 345364. Rascher, D., Brown, M., Nagel, M., & McEvoy, C. (2009). Where did National Hockey League fans go during the 20042005 lockout? An analysis of economic competition between leagues. International Journal of Sport Management and Marketing, 5(12), 183195.