Determinants of Russian Football Club Brands.

Coates, Dennis ; Naidenova, Iuliia ; Parshakov, Petr 等

Introduction

It has become standard for managers of football clubs to treat their clubs as brands (Gladden, 2002). Doing so "enhances and contributes to brand loyalty among consumers" (Ross, Russell, & Bang, 2008, p. 322). As Gladden and Funk (2001) state, "Brand loyalty ensures a more stable following even when the core product's performance falters (i.e., the team has a losing season)" (p. 68). This explanation of the value of brand loyalty to a club is inherently a dynamic one; teams can only have a losing season over time. Moreover, a stable following suggests a very practical objective, to ensure that demand for the team's core product does not falter at the same time team performance does. Brand is one of several types of intangible capital that are important determinants of a company's ability to create value. Focusing on four different industries, Megna and Mueller (1991) hypothesize that profit rates differ among firms in a given sector because of a failure to measure assets properly, particularly failure to account for intangible assets from research and development, and advertising expenditures. McGrattan and Prescott (2014) introduce intangible capital, including brand equity, into real business cycle theory and use microevidence to show that intangible investments are large and also highly correlated with tangible investments. Clearly, then, investing in and managing this capital is an economic decision of firms.

There is little sport economics research focused on capital investment by sport clubs, whether on tangible or intangible capital, like brand; sport marketing research has generally moved away from the use of objective, market-based measures of brand, such as attendance or revenues, in favor of subjective consumer-based measures like brand awareness and brand recognition (Bauer et al., 2008; Bauer, Sauer, & Exler, 2005; Ross, Russell, & Bang, 2008), or fan identification with the club (Hoegele, Schmidt, & Torgler, 2014a, 2014b). The current paper proposes an alternative measure of brand equity or brand strength of sports clubs that is derived from observed consumer decisions, attendance at matches, rather than on survey questions. After obtaining the measure of brand strength, the analysis focuses on determinants of this equity and how it varies over time.

A deeper understanding of brand management in football and other sports clubs is important for the club ownership because efficient brand management enables the football club to maximize or maintain its financial value through periods of both on-field success and on-field failure. Richelieu and Lessard (2014) consider how formerly successful clubs can rebuild their brand after a period of relatively poor performance on the pitch, suggesting club attributes and choices to enhance their brand. We take the Richelieu and Lessard (2014) discussion as focused on brand management for any sport club, not simply those attempting to rebuild brand. If they are correct in their specification, then we should find support for their proposed club characteristics in our analysis, and clubs can use those ideas to improve their brand management. Football club stakeholders may be interested in the determinants of their club's brand strength because of how it affects economic returns; Millward (2013) points out the influence of brand strength on a club's profits, and Couvelaere and Richelieu (2005) link brand to a club's commercial viability. The analysis here does not and cannot address whether brand generates economic returns.

For analysts and academicians, a deeper understanding of sports club brands is valuable both for improved explanations of club behavior and for the lessons it may or may not have for brand development in other industries, particularly in the evaluation of the dynamics of brand. Sports clubs provide a unique opportunity to study the dynamics of brand equity because of the array of high-quality, market-based information that is readily available year after year. Mizik (2014) studies brand equity in a short time series using financial market data to study the dynamics of brand in multiple industries. The analysis here also uses multiple years of data to address the dynamics of brand.

Our analysis focuses on the football clubs of the Russian Premier League (RPL). This focus has both advantages and disadvantages. One disadvantage is that the league is relatively young, limiting the ability to extend the data set into the past. An advantage of the focus is that clubs of the Russian Premier League are little known outside of Russia and have certainly not made concerted efforts to build an international fan base the way clubs of Western European leagues have. Consequently, our measure of intangible brand capital is particularly relevant for clubs of the RPL for whom own-generated revenues are primarily domestically based. Additionally, improved brand management may be especially important for most Russian football clubs because they are financially inefficient. Operating with little transparency and soft budget constraints--most receive government support--these clubs have little incentive to effectively manage resources, including their brand. Indeed, these characteristics may apply to many companies in transition countries or in countries with weak protection of property rights and the rule of law. Understanding the relevant factors for brand management may allow these clubs to use their resources more wisely, both in terms of putting a better, more competitive team on the pitch and in terms of providing value for the club ownership and the fans. For example, the league has varied its policy on the use of foreign players over time, often very close to the start of a season. Changing the number of foreign players a team can have on the pitch affects relevant factors of brand management hypothesized in the marketing literature (Richelieu & Lessard, 2014). These changes complicate brand management and affect how the clubs structure their rosters, and hence their team wage bills.

Here brand strength, or brand equity, is estimated as the power of a visiting club to boost the attendance of the home team. In this way, brand can be thought of as a determinant of demand. Pawlowski and Anders (2012) have suggested that attendance at football matches in which the home team is expected to lose may be driven by the brand following of the visiting club. Berri and Schmidt (2006) examined factors influencing road attendance in the National Basketball Association (NBA), finding that the strength of the opponent team affects game attendance. Coates and Humphreys (2010, 2012) estimated the impact of visiting teams on National Football League (NFL) and National Hockey League (NHL) attendance, attributing this boost in part to the fans of the visiting team that might live in or travel to the city where the game is held. The visitor effects were stronger for the NFL than the NHL. Coates and Humphreys did not extend their analysis to determinants of these visiting team effects. In this paper we do such an analysis for the clubs of the Russian Premier League. Our key findings are that a history of club success builds brand, but that this brand can deteriorate with a period of reduced success. In other words, for professional sports franchises, the best marketing is winning.

Literature Review

As previously mentioned, sport marketing research has moved away from objective observable measures of brand to measures based on survey responses. This movement has its antecedents in the work of Aaker and Keller (1990), Aaker (1991), Keller (1993), which was extended to the sport context in Ross (2006), and Ross, James, and Vargas (2006). For example, according to Keller (2008), ''A brand is something that has actually created a certain amount of awareness, reputation, prominence, and so on in the market place'' (p. 2). Shank (1999) provides a more specific definition of brand in sports: ''a name, design, symbol, or any combination that a sports organization uses to help differentiate its product from the competition'' (p. 239). It seems that there is not much difference between these two definitions. However, a sport brand has an important feature: a brand is determined only with intangible characteristics whereas brands of other products have plenty of other tangible parameters (e.g, quality; Cordiner, 2001). Previous studies (Bauer et al., 2008; Bauer, Sauer, & Exler, 2005; Bauer, Sauer, & Schmitt, 2005; Hoegele, Schmidt, & Torgler, 2014a, 2014b) have shown that a brand is a major determinant of fans' loyalty and team identification, which are relevant because loyalty and identification help sports clubs maintain demand for their product even when that product--performance on the field--varies in quality.

There are different classifications of factors that drive football clubs' brands. Richelieu and Lessard (2014) surveyed management at 19 successful European football clubs from which they identified eight catalyzing factors for reviving the brand of a once successful club. These factors are also reasonably identified as factors for building and maintaining the brand intangible capital. These eight factors are: on-field performance, authenticity/integrity, strategic management, internationalization, market research, regional anchor, youth development, and acquisition of foreign players. Gladden and Funk (2002) suggest a Team Association Model of brand that has three distinct themes, each with multiple dimensions. These themes and their associated dimensions are: Attribute (success, head coach, star player, management, stadium, logo design, product delivery, and tradition), Benefit (identification, nostalgia, pride in place, escape, and peer group acceptance), and Attitude (importance, knowledge, and affect). Ross et al. (2006) suggest another metric of brand in sports: non player personnel, team success, team history, stadium community, team play characteristics, brand mark, commitment, organizational attributes, concessions, social interaction, and rivalry. Lee et al. (2012) shows that "satisfaction with the stadium experience, in turn, had a positive, direct impact on intention to revisit" (p. 490). Clearly, these different authors suggest similar determinants of brand. Taking as given that brand is a source of stability in demand, the analysis here attempts to use these dimensions of brand as a guide to find its determinants for the clubs of the Russian Premier League.

There is a large literature estimating demand for attendance at sporting events, none of which explicitly discusses brand as a determinant of attendance. As our measure of brand intangible capital derives from attendance, it is important to review this literature. Rottenberg (1956) was the first to describe the demand for sport. His discussion included the usual determinants of demand, prices, income, and preferences. Rottenberg raised the issue of the quality of the play of the teams both in an absolute sense and relative to one another. He hypothesized that fans would be more interested in seeing games in which the teams were equally matched. This has come to be known as the uncertainty of outcome hypothesis. Neale (1964) added to this notion, positing that churning of the standings, with clubs regularly moving up and down the standings, would make for greater fan interest, and hence attendance. Borland and MacDonald (2003) and Garcia and Rodriguez (2002) review the theory and empirical issues of demand for attendance.

The evidence from demand estimates is mixed on all counts. For example, estimates of price elasticity range from highly inelastic to unitary elasticity to fairly price elastic. A consensus on price inelasticity has developed, giving rise to studies seeking to explain why--despite the essential monopoly structure of the sport attendance market--estimates do not indicate price elastic demand (Coates & Humphreys, 2007; Fort, 2004; Krautmann & Berri, 2007). A difficulty with estimating price elasticity is that the price faced by the marginal attendee is impossible to determine. Consequently, researchers are left with imprecise proxies of price. A second issue is that ticket price is theoretically endogenous, so researchers must find valid instruments to identify the price or to simply estimate reduced form models that omit price. Finally, if the research uses annual attendance rather than game attendance, the price variable has no variation within a year. The upshot of this is that in panel estimation, either price or year fixed effects can be used but not both.

Income elasticity estimates are also problematic. Average income in a community may be a poor proxy for the income of the marginal attendee. Moreover, when using game attendance there will be no variation in income from game to game for a given team. Estimates of the income elasticity are generally small, positive, and not statistically significant.

By far the issue grabbing the most attention is uncertainty of outcome. Studies of uncertainty of outcome (UOH) and attendance use either individual game attendance (Borland, 1987; Coates & Humphreys, 2010, 2012; Coates, Humphreys, & Zhou, 2014; Forrest & Simmons, 2002) or season-long attendance (Lee & Fort, 2005; Mills & Fort, 2014; Schmidt & Berri, 2004). For game attendance, Coates, Humphreys, and Zhou (2014) note that models that use measures of home team win probability to measure uncertainty strongly reject the UOH. They develop a model of attendance demand that builds on reference dependent preferences and loss aversion. Their formal model has two parameters that reflect game uncertainty, measured by the probability of the home team winning the game. Depending on the signs of these two parameters, the data indicates support for the uncertainty of outcome or loss aversion and reference dependent preferences.

Methodology

The first stage of the analysis is to assume that the coefficient on the visiting team dummy in an attendance regression captures the influence of brand, brand loyalty, brand strength, or intangible brand capital. The model of attendance is a linear reduced form regression so that price can be omitted. We include in the attendance regression those factors that vary from game to game. The maintained assumption is that intangible brand capital does not vary by game. Formally, a reduced form attendance equation is specified as:

[mathematical expression not reproducible]

where attendance is number of people who attended the game; temperature (in degrees Celsius) and precipitation (equal to 1 if there were any kinds of precipitation and 0 otherwise) characterize weather conditions on the day of the game; distance is the number of kilometers between the city of the visiting team and the city of the home team; goalsh and goalsv are the average number of goals per game scored in previous games of the season by the home team and visiting team, respectively; goalsallowedh and goalsallowedv are the average number of goals allowed by the home and away teams, respectively, before the game; stadiumcapacity is the total capacity of the stadium where the game was held; nhs is a dummy variable that is equal to 1 if the game is not played in the usual home stadium of the home team and 0 otherwise; derby is a dummy variable equal to 1 if the match is between two teams from the same city and 0 otherwise; and [epsilon] is the error term. The model also includes season-specific home and visiting team dummies represented by feh and fev, respectively.

Understanding the fixed effects is important for the analysis. The home team fixed effect captures both the popularity of the team within its city and the popularity of football more generally within the city. It is assumed that a football club's brand cannot change significantly during one season but it can differ from season to season. We interpret the visiting team dummy variable coefficients in this equation as indicators of a football club's brand. Note that this is different from brand discussions in marketing. First, it is entirely objective, based on observed data; the typical measures of brand are based on name recognition or other survey information and do not link directly to actual decisions of consumers. Second, the visiting team fixed effect captures the recognition of the club outside its home area, something very much like the marketing approach that uses product recognition to measure brand.

In the second step, the determinants of the football clubs' brands are considered. Ideally, the analysis would include data on advertising and promotional expenditures and other direct investments in fan loyalty. Such information is not available. Our choice of determinants is based on the results of Richelieu and Lessard (2014). However, since the idea of this study is to elaborate an objective measure of sports club brand and its determinants, we do not use survey-based determinants. For this reason we include five of eight factors, suggested by Richelieu and Lessard (2014): on-field performance, acquisition of foreign players, youth development, strategic management, and internationalization. It is impossible to find objective measures or non-survey-based indicators of authenticity/integrity in market research, whether the club is a regional anchor or not. For the five measures included, some, for example number of years in the top division, are weak proxies for the concept of interest, in this case strategic management.

On-field performance is captured in multiple ways: history of championships is the number of national championships prior to the current season, place is the team's place in the national championship table in the previous season (the first place is for the winner), and since last measures the number of seasons since the team last won the league championship. Champions league and Europa League are binary variables equal to 1 if a team plays in the Champions League and Europa League, respectively, in the season and reflects both on-field performance and internationalization. Acquisition of foreign players is captured by the percentage of the team comprised of foreign nationals, foreign players. Having a head coach that is a foreign national also captures the acquisition of foreign players aspect of brand. Of course, foreign players and foreign coaches also contribute to internationalization of the club. Richelieu and Lessard (2014) describe youth development as the "(a)bility to make stars of almost unknown young players through the Clubs' football academies, soccer camps, etc" (p. 290). To capture this aspect of brand (re)invigoration, the analysis includes the average age of the players on the team. Richelieu and Lessard (2014) describe strategic management of the brand as comprised of many factors, such as the ability to attract a large number of diverse stakeholders and to connect emotionally with fans. This also includes a respect for the team's history and its roots. As a very imperfect proxy for this strategic management of the brand, the analysis includes the number of years that the club has been in the top division of Russian football. Having the best Russian players is also important for the club's brand both because of the quota system on foreign players, identified with foreign players, and because players from the Russian national soccer team are well known in Russia. This is captured by the National team players variable. The strength of a club's players is captured by average transfer value of players during a season. We also believe that change of the head coach (coachchange) can contribute to club's brand because change of the head coach suggests a change of the team's strategy and possible improvement of team performance. Additionally, we included a set of season dummies because of the change in the league calendar.

Thus in the second step we estimate the following equation:

[mathematical expression not reproducible]

where the [beta]s are parameters to be estimated and e is the error term.

An issue arises in the analysis because of promotion and relegation. Several clubs have spells in the premier league that last only one season or are interrupted during our time period. Of course, it is difficult to imagine that such clubs have a significant brand, as measured by attendance increases associated with them being the visiting team. To address this, the analysis does not include visiting team year dummy variables for these clubs but rather uses them as the omitted category, that is, as the constant term in the first-stage equation. It is possible that a prominent club with a significant following is relegated at the end of the first season and that club's visiting team effect from that season would reflect its brand strength. No such clubs exist in the data. Likewise, all of the clubs that have interrupted spells in the Russian Premier League are from smaller cities with little history in the top divisions of Russian football either during the Soviet period or since. For the second-stage equation, the dependent variable is defined in four distinct ways. First, the dependent variable is simply the annual visiting team coefficients taken directly from the regression output. Of course, each of these coefficients has its own variance, so the model would be heteroskedastic. In the first instance, the estimation uses heteroskedasticity robust standard errors. In the second specification, each coefficient is first normalized by its standard error, so each has a variance of 1. In the third specification, each visiting team fixed effect [delta]k is added to the first-stage equation constant term [alpha]0. The model is then estimated with robust standard errors. Finally, we use the sum of the visiting team fixed effect and the constant term as in the third specification, but in this case we divide the sum by the standard error of the sum, [mathematical expression not reproducible], to correct for the induced heteroscedasticity.

Data

The focus here is on Russian football clubs that played in the Russian Football Championship during the 2010-14 seasons. In the Russian Football Championship there are 16 teams that play each other twice, once at home and once away, for a total of 30 matches. Three points are awarded for a win, one for a draw, and none for a loss. Teams that capture first and second place qualify for the UEFA Champions League group stage. The third-place team qualifies for the Champions League's second qualifying round. The fourth- and fifth-place teams qualify for the UEFA Europa League. The bottom two teams are relegated to the lower division.

The Russian Football Championship (or Russian Premier League) was organized in 1992 and traditionally the season ran in the summer, from March to November. In 2010 the Russian Football Union decided to change the calendar beginning with the 2012-13 season, and now the league season follows the autumn-to-spring model common in Western Europe. To make the change gradual, there was a transitional season that began in early 2011 and continued until summer of 2012. During the transitional season, 16 Premier League teams played each other twice over the course of the 2011 calendar year; then they were split into two groups of eight, and the teams played the other teams in their group twice more for a total of 44 games (30 in 2011 and 14 in 2012; UEFA.com, 2010).

The characteristics of game attendance and weather conditions are presented in Table 1. Attendance data is collected from the league's official website (http://www.championat.com) and weather data from a Russian-based weather website (https://nuipogoda.ru). We excluded all games that were postponed or played behind closed doors (i.e., a game in which the league has banned fans as a punishment for bad fan behavior). Average attendance of Russian Premier League games is about 12,000 people with the minimum of 1,500 and the maximum of 68,000 people. Stadium capacities range from 10,600 to 84,745, though only one stadium is actually larger than 36,000. Luzhniki Stadium, capacity 84,745, was the national stadium of the Soviet Union and the primary venue for the 1980 Olympic Summer Games. Very few games are sellouts, about 7% of our sample, even when a sellout is defined as 95% of the stadium capacity. Only 12 matches have actual attendance equal to stadium capacity. For this reason, first-stage regressions are simple OLS rather than Tobit or hurdle model estimations. The temperature varies greatly: from -14 to +38 Celsius. During one third of game days precipitation occurred. The largest distance between home team and away team cities is more than 4,000 kilometers.

Tables 2 and 3 provide information about the clubs in our analysis. CSKA, Dinamo, Lokomotiv, and Spartak are from Moscow, and Zenit is from St. Petersburg. CSKA was the "Red Army" team during the Soviet era; Dinamo was associated with the police; and Lokomotiv was associated with the railway. Spartak was created by the trade unions and was therefore identified as the "peoples' team." Clubs in the Russian Premier League may be privately owned, as is Spartak today, or held by corporations. Zenit, for example, is owned by Gazprom, the petroleum and natural gas company.

Results

The presentation of results begins with a stripped down model that does not include any of the club dummy variables. These results are presented in Table 4, Model 1. The findings are largely as one would expect. For example, characteristics of the facility, the weather, and derbies all have the expected signs; stadium capacity, warmer temperatures, no precipitation, and derbies are each associated with greater attendance. Home and visitor goals per game scored prior to this match both are positively associated with attendance, while home and visiting goals allowed per game are negatively related to attendance. It appears Russian football fans want to see high scoring games, but do not want to see their team give up lots of goals. Games not played in the home team's usual stadium and distance between cities of the two teams are not statistically significant. Model 2 in Table 4 includes dummy variables for home and visiting teams for each game. These dummies are allowed to vary by season. The signs and significance of the coefficients change a bit in the model with home and visiting team dummies compared to the model without them.

Table 4, Model 2 provides the estimates on the variables discussed in the preceding paragraph when the model includes the annual home and visiting team dummy variables. With these annual team effects included, the visitor and home team goals scored per game and precipitation become insignificant. Stadium capacity, temperature, and derbies remain significant, but distance becomes significant. Importantly, the annual home and visiting team dummy variables are jointly significant (F(118, 933) = 12.34). The measure of brand in this analysis is the team-season specific visitor effect on the home team attendance obtained from Model 2 and reported in Appendix A. These coefficients are the dependent variable in Model 3 of Table 5. Model 4 uses these coefficients relative to their standard error instead. In other words, the dependent variable of Model 4 is the t-ratio for each of the team-season specific visitor effects. Model 5 adds the team-season coefficient to the estimated constant from Model 2, and Model 6 uses this sum divided by its standard error as the dependent variable. The constant from Model 2 incorporates the influence of the omitted visiting team effect, Ural in the 2013 season.

Each of the three different measures of brand strength produce similar results. In each of Models 3-6, the number of championships won by a team raises the attendance at matches that team plays on the road. The length of time since the last championship does not reduce this brand strength and, indeed, the coefficient estimate is positive. This result is contrary to expectation. One possible explanation is that we do not take into account achievements of the clubs during the Soviet period. Only one of the teams--Dinamo Moskow--won a championship in the Soviet Top League. Its last championship was in 1976. Including this information in our data, the coefficient on the natural logarithm since last championship turns to negative but it does not become statistically significant.

Two of the four models indicate that having Russian national team players increases brand strength. Interestingly, appearances in the Champions League, in the Europe League, and where the team ranked in the table at the end of the previous season are all statistically insignificantly different from 0. Foreign players, average age of the team, coach change, and foreign coach also have no effect on team brand. The quality of the players measured by the average transfer value is positive and statistically significant.

The 2012-13 transition season dummy is significant and negative, indicating that the brands of all clubs were lower in this season. However, besides the transition from the summer schedule to the fall-winter schedule, the league also introduced a change in the quota system for foreign players. We try to separate these two influences by interacting the season dummy and foreign players share. If the brand effect was due to the change in the quota, this interaction term would be statistically significant. However, the interaction term is insignificant, which means the negative effect of the 2012-13 season is explained by the transition to the new schedule.

The results indicate that brand following rises with the number of championships a club has won. To put this in context, consider that the coefficient on the number of championships won in Model 3 indicates an additional championship and implies an additional 548 fans will attend matches when that club is the visitor. Considering that average attendance in the sample is 12,000 to 13,000, these extra 548 fans are about 4.3% of the total attendance. For Spartak with nine championships since 1992, the impact on attendance when it is the visitor is a quite substantial 4,932 increase in fans. Of course, one natural question is whether the relationship is linear or if the marginal impact of an additional championship declines as the number of championships rise. Estimating the model with championships and championships squared suggest the linear form is appropriate. Interestingly, our results show that the impact of championships on brand strength does not depend upon how recently they were won. Apparently, winning a championship provides a boost to brand that does not depreciate over time. Another way to raise the brand is to increase the quality of players. Increasing the transfer value by one standard deviation gives 1,117 more additional fans, which is 9% of the average attendance. Interestingly, buying an additional Russian national team player increases the brand by 3%, which is slightly lower than winning a championship.

In order to address the issue of robustness of our results we have experimented with methods to deal with the large number of zero championship wins and the large outlier of Spartak with nine championship wins including a dummy variable for Spartak, a dummy variable separating out teams that have won championships from those that have not, and interactions between these dummies and years since the last championship. Estimation results are available upon request. Results on number of championships, transfer values, and Russian national team players are robust to these alternatives.

Conclusion

We find empirical evidence that the game attendance of Russian football clubs is affected by the intangible brand capital of the opponent. Interestingly, the strength of the visiting club's brand does not change significantly over seasons. In other words, the results imply that a club's brand does not depreciate after winning a championship. Teams with great history, such as Spartak Moscow, still attract many fans to the stadium even if at the current moment the team is not so successful. Possibly the depreciation of a club's brand takes more time and Russian Premier League has not enough history to estimate it. On the other hand, an issue we do not address in the analysis is success below winning the championship. For example, while Spartak has not won the championship since 2001, it has placed second five times, third once, and fourth twice. The failure to find a deterioration of brand over time may be that our analysis is insufficiently reflective of the type of poor results that would produce a waning of followers. Support for this notion is the evidence of Dinamo Moscow, whose last championship in Russian football was in 1976. Inclusion of that information in the time since last championship produced a negative coefficient reflecting depreciation. For Spartak, it is possible that the club had not fallen on such hard times to induce large depreciation in its brand during the run of our data. However, since the 2011-12 season when the club placed second, it has finished fourth once, sixth twice, and fifth once. Perhaps if our data had extended to these last two seasons, our result for time since the last championship would be different. And, if Spartak continues to finish no higher than fourth for a few more seasons, that may also produce the depreciation of brand we expected to find.

Our results suggest that a club can develop its brand. Winning championships frequently strengthens the brand but is quite difficult. However, club management can opt for a squad made up of more Russian team players or simply buy better players. The effect of buying additional national team players is lower, but comparable to winning a championship. An increase in the club's player budget generates growth in the brand. However, the opposite is also true. A decrease in a team's player budget leads to the rapid destruction of the club's brand. This conclusion corresponds to the situation with Anzhi soccer club, which was purchased by Russian billionaire Suleiman Kerimov in 2011. The club bought star player Samuel Eto'o and hired top soccer coach Guus Hiddink. However, in 2013 the club lost the financial support of Kerimov, sold its top players, changed the head coach, and moved from the top to the bottom of the league (RT Question More, 2013).

Our findings in this research are subject to at least three limitations. First, results obtained in the framework of this project may not be transferable to other leagues because of the unique features of the Russian Premier League. Second, the period of analysis is perhaps too short to study the dynamics of brands. Third, other factors that influence brand, such as promotional and marketing efforts, are not included. Future studies should conduct a similar study of brand determinants for a number of football leagues. Such research can provide developing leagues with a list of best practices of brand management in developed leagues. For developed leagues, in turn, such a study can reveal key drivers of brands in emerging leagues that may help those leagues to enter new markets.

References

Aaker, D. (1991). Managing brand equity. New York, NY: Free Press.

Aaker, D., & Keller, K. (1990). Consumer evaluations of brand extensions. Journal of Marketing, 54, 27-41.

Bauer, H., Sauer, N., & Exler, S. (2005). The loyalty of German soccer fans: Does a team's brand image matter? International Journal of Sports Marketing & Sponsorship, 7, 14-22.

Bauer, H., Sauer, N., & Exler, S. (2008). Brand image and fan loyalty in professional team sport: A refined model and empirical assessment. Journal of Sport Management, 22, 205-226.

Bauer, H., Stokburger-Sauer, N., & Schmitt, (2005). Customer-based brand equity in the team sport industry: Operationalization and impact on the economic success of sport teams. European Journal of Marketing, 39, 496-513.

Berri, D. J., & Schmidt, M. B. (2006). On the road with the National Basketball Association's superstar externality. Journal of Sports Economics, 7, 347-358.

Borland, J., & MacDonald, R. (2003). Demand for sport. Oxford Review of Economic Policy, 19, 478-502.

Coates, D., & Humphreys, B. R. (2007). Ticket prices, concessions and attendance at professional sporting events. International Journal of Sport Finance, 2, 161-170.

Coates, D., & Humphreys, B. R. (2012). Game attendance and outcome uncertainty in the National Hockey League. Journal of Sports Economics, 13, 364-377.

Coates, D., & Humphreys, B. R. (2010). Week to week attendance and competitive balance in the National Football League. International Journal of Sport Finance, 5, 239-252.

Coates, D., Humphreys, B. R., & Zhou, L. (2014). Reference-dependent preferences, loss aversion, and live game attendance. Economic Inquiry, 52, 959-973.

Cordiner, R. (2001). Sport wakes up to the value of brands. Sports Marketing, 75, 13.

Couvelaere, V., & Richelieu, A. (2005). Brand strategy in professional sports: The case of French soccer teams. European Sport Management Quarterly, 5, 23-46.

Forrest, D., & Simmons, R. (2002). Outcome uncertainty and attendance demand in sport: The case of English soccer. Journal of the Royal Statistical Society: Series D (The Statistician), 51, 229-241.

Gladden, J. M., & Funk, D. C. (2002). Developing an understanding of brand associations in team sport: Empirical evidence from consumers of professional sport. Journal of Sport Management, 16, 54-81.

Hoegele, D., Schmidt, S. L., & Torgler, B. (2014a). Superstars and disloyal fans: Factors influencing attraction to competition. Sport, Business and Management: An International Journal, 4, 298-316.

Hoegele, D., Schmidt, S. L., & Torgler, B. (2014b). Superstars as drivers of organizational identification: Empirical findings from professional soccer. Psychology and Marketing, 31, 736-757.

Keller, K. (2008). Strategic brand management: Building, measuring and managing brand equity. Upper Saddle River, NJ: Prentice Hall.

Keller, K. L. (1993). Conceptualizing, measuring and managing customer-based brand equity. Journal of Marketing, 57, 1-22.

Krautmann, A. C., & Berri, D. J. (2007). Can we find it at the concessions? Understanding price elasticity in professional sports. Journal of Sports Economics, 8, 183-191.

Lee, S., Lee, H., Seo, W., & Green, C. (2012). A new approach to stadium experience: The dynamics of the sensoryscape, social interaction, and sense of home. Journal of Sport Management, 6, 490-505.

Lee, Y. H., & Fort, R. (2005). Structural change in MLB competitive balance: The depression, team location, and integration. Economic Inquiry, 43, 158-169.

McGrattan, E. R., & Prescott, E. C. (2014). A reassessment of real business cycle theory. American Economic Review: Papers & Proceedings, 104, 177-182.

Megna, P., & Mueller, D. C. (1991). Profit rates and intangible capital. The Review of Economics and Statistics, 73, 632-642.

Mills, B., & Fort, R. (2014). League-level attendance and outcome uncertainty in US pro sports leagues. Economic Inquiry, 52, 205-218.

Millward, P. (2013). New football directors in the twenty-first century: Profit and revenue in the English Premier League's transnational age. Leisure Studies, 32, 399-414.

Mizik, N. (2014). Assessing the total financial performance impact of brand equity with limited time-series data. Journal of Marketing Research, 51, 691-706.

Neale, W. (1964). The peculiar economics of professional sports. Quarterly Journal of Economics, 78, 1-14.

Pawlowski, T., & Anders, C. (2012). Stadium attendance in German professional football--The (un) importance of uncertainty of outcome reconsidered. Applied Economics Letters, 19, 1553-1556.

Richelieu, A., & Lessard, S. (2014). Long gone the glory days. Sport, Business and Management: An International Journal, 4, 284-297.

Ross, S. (2006). A conceptual framework for understanding spectator-based brand equity. Journal of Sport Management, 20, 22-38.

Ross, S. D., Russell, K. C., & Bang, H. (2008). An empirical assessment of spectator-based brand equity. Journal of Sport Management, 22, 322-337.

Ross, S., James, J., & Vargas, P. (2006). Development of a scale to measure brand associations in professional sport. Journal of Sport Management, 20, 260-279.

Rottenberg, S. (1956). The baseball players' labor market. Journal of Political Economy, 64, 242-258.

Schofield, J. A. (1983). Performance and attendance at professional team sports. Journal of Sports Behaviour, 6, 196-206.

Shank, M. (1999). Sports marketing: A strategic perspective (2nd ed.). Upper Saddle River, NJ: Prentice Hall.

Russian league switches to new calendar. (2010). Retrieved from http://www.uefa.com/member-associations/association=rus/news/newsid=1527870.html?cid=PULSE_TW

Eto'over? Billionaire Kerimov sells Russia's richest football club Anzhi top players, RT Question More. (2013). Retrieved from https://www.rt.com/news/russian-football-anzhi-sell-134/

Authors' Note

This study comprises research findings from the Intangible-driven dynamics in economics and finance project carried out within the International Laboratory of Intangible-Driven Economy (ID Lab) of the National Research University Higher School of Economics' Basic Research Program in 2017. Appendix A. Visiting Team Year Effects Team Season Effect t-stat P>t Krylia Sovetov 2010 1532.868 1.00 0.315 Alania 2010 1891.354 1.25 0.213 Amkar 2010 1894.975 1.22 0.224 Saturn 2010 1986.963 1.10 0.273 Spartak Nalchik 2010 2000.053 1.39 0.164 Terek 2010 2006.751 1.39 0.164 Sibir 2010 2055.411 1.29 0.196 Tom 2010 2331.674 1.74 0.082 Rubin 2010 2361.345 1.39 0.165 Lokomotiv 2010 3090.082 2.05 0.040 Spartak 2010 4878.494 3.63 0.000 Dinamo 2010 5276.259 2.31 0.021 Zenit 2010 5978.716 3.64 0.000 Anzhi 2010 622.7952 0.41 0.685 CSKA 2010 6501.436 3.15 0.002 Rostov 2010 803.6664 0.55 0.581 Rubin 2011/12 1523.966 1.13 0.260 Lokomotiv 2011/12 1547.894 1.07 0.284 Volga 2011/12 1596.333 1.21 0.228 Krylia_Sovetov 2011/12 1798.517 1.34 0.181 Kuban 2011/12 1802.13 1.28 0.200 Tom 2011/12 2027.365 1.62 0.105 Krasnodar 2011/12 35.60914 0.03 0.977 Rostov 2011/12 447.7662 0.36 0.717 Amkar 2011/12 488.9131 0.39 0.694 Spartak Nalchik 2011/12 -5.743966 -0.01 0.996 Anzhi 2011/12 5683.562 3.89 0.000 Zenit 2011/12 8035.25 4.60 0.000 CSKA 2011/12 8410.644 4.11 0.000 Spartak 2011/12 8483.158 4.50 0.000 Dinamo 2011/12 854.5931 0.54 0.591 Terek 2011/12 939.6296 0.80 0.422 Lokomotiv 2012/2013 1053.313 0.58 0.565 Krasnodar 2012/2013 -1385.47 -0.88 0.382 Amkar 2012/2013 -1461.411 -1.06 0.290 Mordovia 2012/2013 -1680.098 -1.12 0.264 Rubin 2012/2013 2264.601 1.27 0.203 Volga 2012/2013 -2418.766 -1.72 0.086 Anzhi 2012/2013 4395.782 2.64 0.008 Dinamo 2012/2013 469.4985 0.25 0.803 Zenit 2012/2013 4769.925 2.62 0.009 Krylia Sovetov 2012/2013 630.3338 0.43 0.670 Terek 2012/2013 -661.6142 -0.46 0.642 CSKA 2012/2013 6863.917 2.88 0.004 Rostov 2012/2013 762.9267 0.48 0.628 Alania 2012/2013 768.1143 0.55 0.584 Kuban 2012/2013 786.2146 0.51 0.613 Spartak 2012/2013 8278.462 2.76 0.006 Rostov 2013/2014 1364.642 1.40 0.161 Lokomotiv 2013/2014 1806.75 1.35 0.176 Amkar 2013/2014 21.52744 0.02 0.985 Anzhi 2013/2014 2179.545 2.64 0.008 Kuban 2013/2014 2210.576 2.31 0.021 Rubin 2013/2014 2408.065 1.88 0.061 Krylia Sovetov 2013/2014 -417.111 -0.44 0.660 Tom 2013/2014 494.8916 0.67 0.502 CSKA 2013/2014 5339.591 4.11 0.000 Volga 2013/2014 572.4839 0.69 0.492 Terek 2013/2014 -601.9295 -0.67 0.504 Dinamo 2013/2014 675.7112 0.56 0.578 Spartak 2013/2014 7096.194 4.35 0.000 Krasnodar 2013/2014 787.0927 0.96 0.337 Zenit 2013/2014 8236.977 6.44 0.000 Omitted visiting team-year is Ural 2013.

Dennis Coates (1), Iuliia Naidenova (2), and Petr Parshakov (2)

(1) UMBC

(2) National Research University Higher School of Economics, Russia

Dennis Coates is a professor of economics at UMBC and a leading international researcher at the International Laboratory of Intangible-Driven Economy at the National Research University Higher School of Economics. His research focuses on topics in sport economics and public economics.

Iuliia Naidenova is a researcher and lecturer at the International Laboratory of Intangible-Driven Economy at the National Research Higher School of Economics. Her research interests include corporate finance, intellectual capital, and sport economics.

Petr Parshakov is a researcher at the International Laboratory of Intangible-Driven Economy at the National Research Higher School of Economics. His research interests include sport economics and empirical corporate finance Table 1. Game Attendance in Russian Football Championship and its Determinants Season Number Average Average Average Average of games attendance temp. precipitation stadium capacity 2010 239 12,187 15.16 0.46 23,940 2011-12 352 12,903 13.67 0.28 28,954 2012-13 236 13,166 13.95 0.30 29,992 2013-14 234 11,717 12.30 0.30 24,385 Season Average Average distance transfer value 2010 1,762 60.96 2011-12 1,517 49.36 2012-13 1,217 67.70 2013-14 1,561 73.46 Table 2. Game Attendance in Russian Football Championship and its Determinants Team Number Average Number Share of Average of games attendance of times games foreign participation with age of players in Champion's foreign players league coach Alania 60 14,753.3 0 0,00 22.9 Amkar 134 10,002.3 0 0,27 26.9 Anzhi 134 13,978.8 0 0,38 24.9 CSKA 134 11,792.5 2 0,00 23.8 Dinamo 134 8,383.1 0 0,64 25.5 Krasnodar 104 10,245.2 0 0,86 26.0 Krylia 134 13,641.6 0 0,00 24.5 Sovetov Kuban 104 18,856.1 0 0,80 25.8 Lokomotiv 134 13,686.4 0 0,66 25.3 Mordovia 30 7,022.8 0 0,37 25.7 Rostov 134 10,106.9 0 0,83 26.0 Rubin 134 11,947.2 2 0,25 25.0 Saturn 30 6,380.0 0 0,00 25.4 Sibir 30 9,933.3 0 0,97 25.4 Spartak 134 18,142.3 2 0,14 23.9 Spartak 74 6,603.4 0 0,00 20.7 Nalchik Terek 134 15,034.6 0 0,07 25.3 Tom 104 8,675.0 0 0,00 26.2 Ural 30 13,188.9 0 0,00 27.3 Volga 104 7,308.7 0 0,37 26.6 Zenit 134 19,140.9 4 0,98 26.1 Team Share of Average Average place in transfer previous value season Alania 30.8 20.0 22.54 Amkar 46.9 12.1 19.42 Anzhi 27.2 9.7 88.58 CSKA 41.9 2.7 134.31 Dinamo 36.6 6.6 98.66 Krasnodar 42.9 13.9 36.49 Krylia 29.5 12.0 21.47 Sovetov Kuban 38.0 12.2 35.16 Lokomotiv 39.9 6.2 93.75 Mordovia 28.1 20.0 17.95 Rostov 47.2 11.5 29.46 Rubin 44.6 3.9 98.53 Saturn 0.0 7.0 1.39 Sibir 25.5 20.0 17.31 Spartak 45.4 3.1 98.79 Spartak 8.3 8.0 5.97 Nalchik Terek 46.4 10.6 26.70 Tom 24.0 11.8 18.81 Ural 24.0 20.0 14.95 Volga 28.4 16.3 13.91 Zenit 40.6 1.7 167.74 Table 3. Clubs' Characteristics at 2014 Team Champions in St. Dev. Alania 1 15 Amkar 0 10 Anzhi 0 7 CSKA 4 22 Dinamo 0 22 Krasnodar 0 3 Krylia Sovetov 0 22 Kuban 0 6 Lokomotiv 2 22 Mordovia 0 1 Rostov 0 20 Rubin 2 11 Saturn 0 12 Sibir 0 1 Spartak 9 22 Spartak Nalchik 0 6 Terek 0 7 Tom 0 8 Ural 0 6 Volga 0 3 Zenit 3 20 Table 4. Determinants of Attendance Model 1 Pooled stadium capacity 0.1287 (***) (0.0177) temperature 98.5424 (***) (19.3746) precipitation -1,229.9148 (***) (366.9204) distance between cities 0.1924 (0.1565) derby 7,185.8069 (***) (1,379.0475) home team average goals scored 1,425.7702 (***) prior to this match (404.1446) visiting team average goals scored 2,634.2210 (***) prior to this match (364.7955) home team average goals allowed -1,916.3406 (***) prior to this match (395.0727) visiting team average goals allowed -2,920.6999 (***) prior to this match (359.9019) game not at home stadium -2,022.8271 (1,369.7121) Home team effects Visiting team effects Constant 8,404.7845 (***) (976.3356) Observations 1,061 R-squared 0.3550 Model 2 Fixed effects stadium capacity 0.1259 (***) (0.0313) temperature 78.3550 (***) (16.0873) precipitation -490.1135 (297.7887) distance between cities -0.3663 (**) (0.1701) derby 7,954.9309 (***) (1,283.7306) home team average goals scored 1,570.6792 (***) prior to this match (527.8201) visiting team average goals scored 14.0692 prior to this match (482.7963) home team average goals allowed -645.6055 prior to this match (532.0650) visiting team average goals allowed -1,529.9503 (***) prior to this match (471.1912) game not at home stadium Home team effects Included Visiting team effects Included Constant 7,157.2588 (***) (1,611.8028) Observations 1,061 R-squared 0.6745 Robust standard errors in parentheses (***) p<0.01, (**) p<0.05, (*) p<0.1 The omitted category includes all home clubs with only one season in the Premier League or whose time in the Premier League is interrupted, and Ural as a visitor in 2013. Table 5. Determinants of Brand Quality Model 3 Model 4 Model 5 Model 2 Visiting Visiting constant team effect team t-stat + visiting team effect Champions League 1,002.1062 0.5500 981.4256 (976.9315) (0.6196) (962.1998) Europa League 818.3260 0.4701 802.9494 (553.8093) (0.3513) (543.4268) Number of 547.9852 (***) 0.2459 (***) 546.6115 (***) championships won (129.0354) (0.0818) (127.5786) Natural log of 723.4342 0.3070 722.0643 years since last championship (454.6195) (0.2884) (450.0222) Place in previous 17.2170 0.0006 14.3362 season (54.1615) (0.0344) (51.8615) Average age of 75.5357 0.0761 70.5522 players (184.1559) (0.1168) (180.7843) Average transfer 21.1920 (**) 0.0154 (***) 21.1559 (**) value of players (8.3360) (0.0053) (8.2508) Share of foreign -27.6426 -0.0226 (*) -26.5192 players (20.7208) (0.0131) (19.8143) Number of Russian 378.0035 (**) 0.1116 377.2752 (**) national team players (186.2330) (0.1181) (184.3367) Years in premier -4.0045 -0.0069 -4.2149 league (35.9059) (0.0228) (35.5326) Foreign coach -363.2050 -0.1679 -346.3486 (606.1595) (0.3845) (594.7790) Coach change -600.9614 -0.3915 -593.2816 (526.6325) (0.3340) (520.0956) Season 2011/12 387.7717 0.2461 384.9972 (607.0563) (0.3850) (600.8366) Season 2012/13 -1,134.5516 (*) -0.9514 (**) -1,149.5819 (*) (622.4433) (0.3948) (612.1035) Season 2013/14 -761.1856 0.0672 -795.5028 (780.3103) (0.4949) (755.2770) Constant -2,558.7703 -1.5331 4,724.6741 (4,803.2659) (3.0466) (4,717.6295) Observations 63 63 64 R-squared 0.7698 0.7302 0.7722 Model 6 Model 2 constant + visiting team effect relative to the standard error Champions League 0.4940 (0.5740) Europa League 0.2634 (0.3254) Number of 0.1647 (**) championships won (0.0758) Natural log of 0.1066 years since last championship (0.2671) Place in previous -0.0106 season (0.0318) Average age of -0.0061 players (0.1082) Average transfer 0.0122 (**) value of players (0.0049) Share of foreign -0.0074 players (0.0122) Number of Russian -0.0188 national team players (0.1094) Years in premier -0.0146 league (0.0211) Foreign coach -0.1306 (0.3561) Coach change -0.5159 (0.3094) Season 2011/12 0.5774 (0.3567) Season 2012/13 -1.0874 (***) (0.3657) Season 2013/14 -0.3401 (0.4584) Constant 5.4938 (*) (2.8220) Observations 63 R-squared 0.6986 Standard errors in parentheses (***) p<0.01, (**) p<0.05, (*) p<0.1 Season 2010 is the omitted category