Fireworks saturation and attendance in minor league baseball.
Paul, Rodney J. ; Weinbach, Andrew P.
In recent years, the number of fireworks shows that Minor League
Baseball teams offer during the season has drastically increased.
Fireworks, which at one time were largely used as a promotion on the 4th
of July, Memorial Day, and/or Labor Day, now frequently appear on the
season schedule for various teams. Some teams have fireworks every
Friday. Others have fireworks every weekend. Some teams seem to have
fireworks nearly every home game of the season.
Due to the increased frequency in the use of fireworks as a game
promotion, the problem facing teams is determining the point at which
fans grow tired of seeing fireworks. How many times can a team use
fireworks before the team sees a decline in marginal attendance? Is it
possible to have too many fireworks shows during the summer? All of
these questions are ultimately testable through observation and study of
game attendance during the baseball season.
Using a regression model with individual game attendance as the
dependent variable, the marginal impact of additional fireworks shows is
investigated using data from baseball teams at the AAA and AA levels.
Using data from five leagues, two AAA and three AA, the impact of city
demographics, team performance, particular days and months, weather, and
other promotions are controlled for, and the impact of having fireworks
and their relative frequencies is tested for its impact on game
attendance.
To account for a possible saturation effect with fireworks, where
fans are not as interested in attending games with fireworks as their
frequency increases, fireworks are included in the model in three ways.
First, fireworks are included as a simple dummy variable. Second, the
frequency of fireworks as a linear sequence (1 for the first time, 2 for
the second time, etc.) is also included in the model. Third, to account
for possible nonlinear saturation effects in respect to fireworks
frequency, the square of fireworks is also included in the model.
If the market for fireworks at the Minor League Baseball level
becomes saturated, either the fireworks frequency term or the fireworks
frequency squared term (or both) should have negative and significant
effects. If these saturation effects are found, the empirical question
to be answered is at what point do additional fireworks shows become
detrimental to of ticket sales and the revenues of the baseball team? Or
is it possible that fans have an insatiable appetite for fireworks and
the expected saturation effects that we would anticipate with increased
usage never materialize? Answering these questions is the goal of this
research.
This paper is organized as follows. The next section gives an
overview of the literature on major and Minor League Baseball
attendance. The third section describes the data used in the study for
the five minor leagues we investigated. The fourth section presents the
empirical findings of the regression model and explains the results. The
final section presents the conclusions.
Literature Review
Economists have estimated demand for baseball game attendance
through a variety of model specifications. Some of the independent
variables that have been studied in the literature include population,
per-capita income, star players, and recent success (Noll, 1974);
televised games, quality of the team, and availability of substitutes
(Demmert, 1973); expected probabilities of winning a championship
(Whitney, 1988); salary structure (Richards & Guell, 1998); turnover
in team rosters (Kahane & Shmanske, 1997); impacts of interleague
play (Butler, 2002; Paul, et al., 2004); and new stadium effects (Coates
& Humphreys, 2005; Depken, 2006). More complete reviews of the
literature on the factors affecting baseball attendance have been
conducted by Schofield (1983), McDonald and Rascher (2000), and Villar
and Guerrero (2009).
Attendance in Minor League Baseball has also been studied in recent
years, both at the annual and game-to-game levels. Some studies found
virtually no link between team performance and attendance (e.g.,
Siegfried & Eisenberg, 1980; Gifis & Sommers, 2006). Gitter and
Rhoads (2010) studied all teams and leagues at the AAA, AA, and A levels
of play from 1992 to 2006 and found that team performance (win
percentage) had a positive and significant effect on attendance. In
addition, Gitter and Rhoads (2010) found that Minor League Baseball and
Major League Baseball are substitutes and that the closest Major League
team's performance only affects Minor League attendance when the
teams are affiliated. Gitter and Rhoads (2011) also did a similar study
at the AAA and AA levels, focusing on the impact of highly touted
prospects (future stars). They found that prospects only affected
attendance at the AAA level and this was only when the future stars were
ranked in the top five. Although statistically significant effects of
these elite prospects were found on attendance, the authors note that
the actual increase in the number of fans attending these games was
small.
Gifis and Sommers (2006) focused on the impact of specific
promotions on attendance for the 2004 Trenton Thunder (AA). They found
that fireworks, guest appearances, and bobblehead giveaways all had a
positive and significant effect on attendance, with fireworks having the
biggest impact. In addition, they found that weekend days, summer days,
and day games also led to increases in attendance over other games.
Cebula, Toma, and Carmichael (2009) studied the Carolina League
(A-Advanced) for a single season (2006) and, in 2013, Cebula (2013)
examined the same league for five consecutive seasons (1997-2001). Both
studies focused on the determinants of fan attendance, including the
study of game promotions, events, and giveaways. Cebula, Toma, and
Carmichael (2009) found that fireworks played a key role in increasing
attendance at Carolina League games (as did per capita income of the
city, runs scored, and Friday and Saturday games), while Monday games,
rainy weather, and team errors had a negative impact on fans in the
seats. Using a fixed-effects pooled sample of five seasons in the
Carolina League, Cebula (2013) found that specific promotions, including
fireworks, merchandise giveaways, food and drink specials, beer
specials, and family/group nights, all had a positive and significant
impact on attendance. In addition to these promotional games, Cebula
(2013) found that population, per capita income, and Friday and Saturday
games also were key drivers of game attendance.
Anthony, et al., (2011) studied three leagues in the southeastern
United States, the Southern (AA), Florida State (A-Advanced), and South
Atlantic (A) leagues, and found similar fan preferences in relation to
attendance in these cities and leagues. Fireworks were shown to have a
large and consistent impact on attendance across the three leagues.
Concerts, family nights, and athlete appearances were also shown to
increase attendance. In addition, weekend games, sunny days, and
population were shown to have positive and significant effects, while
higher per capita income was shown to decrease attendance. Similarities
between leagues were found for the leagues of the Pacific Northwest
(Pioneer and Northwest Leagues) in Paul and Weinbach (2011).
Although a study of Major League Baseball, McDonald and Rascher
(2000) researched the impact of both promotions and the frequency of
their use. Using a data set consisting of 19 teams in Major League
Baseball for the 1996 season, with the inclusion of a large array of
independent variables related to team performance, weather, and
city-specific variables, McDonald and Rascher (2000) found that
promotions as a whole increased attendance in baseball by 14%. They also
found that as the use of promotions increased, there was a watering down
effect as the impact of the promotions subsided. They discovered,
however, that the loss in attendance due to the watering down effect was
offset by the additional attendance gained by having the promotion.
Summary of Data and Fireworks Frequencies by Team and League
This section describes the data used in this study. All data,
unless otherwise noted, were obtained from www.minorleaguebaseball.com
and its individual team websites. Game data were gathered from box
scores of individual games. The promotional information, including
fireworks, was obtained from the promotional schedule listed on each
team's website. Although Minor League Baseball now has a
standardized format for team and league web pages, it does not appear
mandatory. Two teams, Akron of the Eastern League (AA) and Montgomery of
the Southern League (AA), do not conform to normal website standards,
leaving their game data unavailable. We attempted to contact the teams
to gain this information, but were not able to obtain the full necessary
data from these two teams. Therefore, they were excluded from the
analysis. The data listed below and the regression results that follow
include all teams at the AAA level and every team except two at the AA
level. All of the individual game data is from the 2011 season. In any
data set encompassing one season there could be unique attributes to
that year that would leave the results more specific than general. Given
that we have five leagues from different areas of the country and the
individual game information from each game played within the leagues, we
are less concerned with having one season of data than we would be in
other circumstances.
The following tables contain summary information on some of the key
variables related to Minor League Baseball attendance and its
determinants. Table 1 contains summary statistics, by league, for
attendance, population, per-capita income, and game-time temperature.
For each variable, the mean, median, and standard deviation are noted.
Table 2 contains a summary of frequencies of each weather category
(taken from the game box score) and promotions used by teams during the
2011 Minor League Baseball season.
More specifically and directly related to the main focus of this
study, the next set of tables shows the frequency of fireworks
promotions for each team in the five leagues studied. Table 3 notes
fireworks frequency for AAA baseball (International League and Pacific
Coast League) and Table 4 notes fireworks frequency for AA baseball
(Eastern, Southern, and Texas leagues).
At the AAA level, each team had fireworks multiple times during the
2011 season. In the International League, four teams (Columbus,
Pawtucket, Syracuse, and Norfolk) had single-digit fireworks offerings,
with Columbus having the fewest shows at 5. Toledo topped the list with
26 fireworks dates during the season. Rochester also offered more than
20 dates where fireworks displays happened after the game.
In the Pacific Coast League, Tucson, and Las Vegas had the least
frequent fireworks displays with 4 and 5 dates, respectively. None of
the Pacific Coast League teams reached 20 fireworks displays during the
2011 season. Nashville and Omaha each offered 17 fireworks shows, while
Reno, Round Rock, and Tacoma had the next most-frequent fireworks
offerings at 15 shows each.
In the Eastern League, the number of times fireworks were used as a
promotion ranged from a low of 5 times (Portland) to 30 or more (New
Hampshire and Reading). Teams in the Southern League mostly had 12-18
fireworks shows, but Huntsville and Mobile did not list any fireworks
dates for the 2011 season. In the Texas league, Frisco and NW Arkansas
led the league with sixteen fireworks dates, while Arkansas offered the
fewest dates at three.
Overall, there is a great deal of variation across teams in their
use of fireworks throughout the 2011 baseball season. In some markets,
20 or 30 fireworks shows were offered, while others offered only a few
(if any) dates where post-game festivities included fireworks. This
variation should allow for an investigation into differences in fan
interest in fireworks across cities, leagues, and levels and may
potentially reveal where the market for post-game fireworks in Minor
League Baseball becomes saturated.
Regression Model of Minor League Baseball Attendance
To investigate the role of fireworks and the marginal impact of
repeated fireworks displays throughout the season, a simple regression
model was specified with per-game attendance for each team as the
dependent variable. The attendance figure is announced attendance taken
from the box scores of the games in the five minor leagues studied.
Given that ticket prices are set once at the beginning of the season
(with discounts in ticket prices listed as a promotional dummy in the
model), we assume that each team sets ticket price with the expectations
of maximizing profits given local economic conditions. Therefore, we use
a reduced-form model to estimate the marginal impacts of the independent
variables described below. The regression model is as follows:
[Attendance.sub.i] = [[alpha].sub.0] + [[beta].sub.1]
([Population.sub.i]) + [[beta].sub.2] ([income per capita.sub.i]) +
[beta](days of week) + [beta](months) + [beta](home win percentage) +
[beta](road win percentage) + [beta](distance) + [beta](temperature) +
[beta](weather categories) + [beta](promotions) + [beta](fireworks) +
[beta](fireworks number) + [beta](fireworks number (2)) +
[[epsilon].sub.1] (1)
The independent variables consist of a wide variety of factors that
typically are thought to have an impact on baseball attendance. For ease
of explanation, the independent variables are grouped by category. Each
category represents a specific facet of the likely determinants of
attendance for Minor League Baseball. We break the individual
determinants into the categories of city demographics, weekdays, months,
team performance, weather factors, nonfireworks promotions, and
fireworks/fireworks frequency. We will discuss each independent
variable, in order, by category below.
To account for the local city demographics, we include fixed-effect
factors for the population and income per capita of the host city. For
teams who take the name of a region, as opposed to a city, we use the
local city where the team stadium is located. All data on population and
income per capita are taken from www.city-data.com. If larger cities
draw more fans to the game, the population should have a positive and
significant effect on attendance. Possible exceptions to this are in
very large cities, where Minor League Baseball may give way to other
entertainment possibilities in the area. If Minor League Baseball is a
normal good, the income per capita variable should also have a positive
and significant effect on attendance. If Minor League Baseball is an
inferior good, however, the sign will be reversed. In wealthier cities,
it is quite possible that citizens may prefer higher-priced options for
sports and other entertainment, thereby lowering attendance for the
local Minor League Baseball club.
The second and third categories of independent variable are the
days of the week and the months of the season. The implications of these
variables are rather straightforward. Due to the opportunity cost of
time, we would expect that weekend days will attract larger crowds than
weekdays. The months of the year capture the partial impacts of weather
(also included directly below), the overlap of other sports seasons,
school year effects (ending in late spring, starting in late
August/early September), and potential excitement (or disappointment) in
races to win the division and make the playoffs. Wednesday and June are
the excluded day of the week and month of the season dummies.
Team performance is proxied by the win percentage of the team. Win
percentage is calculated as a moving average throughout the season and
is updated for each game on the team schedule. If fans care about
winning, the win percentage should have a positive and significant
effect on attendance. It is possible, however (especially at the minor
league level), that fans may care more about the overall game day
experience than home team wins.
The visiting team win percentage is also included as an independent
variable in the regression analysis. If fans prefer seeing better
opposing teams, this variable should have a positive and significant
effect on attendance. If fans prefer weaker-quality opposing teams,
making it more likely the home team will win the game, the impact of
this variable could be negative. In addition to the quality of the
opposing team, the distance from the opposing team city to the home city
is included in the model to account for possibly rivalry effects and for
fans of the visiting team traveling to see their team in an opposing
park.
The weather for each game is noted in the daily box scores
throughout the season, allowing for an inclusion of these variables in
the attendance regression model. Temperature at game time is included as
an independent variable. In general, warmer temperatures are preferred
to colder temperatures for outdoor sporting events, although in hot
climates, the temperature could hurt attendance due to the
unpleasantness of being outside in the heat for a few hours for a
baseball game. The square of temperature was also tried within the
regression model, but did not provide statistically significant insights
beyond the inclusion of simple temperature (and did not meaningfully
change results).
In addition to the temperature, different categories of weather are
also included in the game box scores. The different categories of
weather seen during the 2011 Minor League Baseball season in the five
leagues studied consisted of partly cloudy, cloudy, clear, sunny, rainy,
and drizzling. In general, poor weather days (rainy and drizzling) are
expected to have a negative impact on game attendance. Sunny and clear
days are likely to be the most popular days. Partly cloudy days are the
excluded categorical dummy variable.
In addition to fireworks, Minor League Baseball teams use many
other forms of promotions throughout the season. To account for these
different promotions, we decided to group promotions by common
categories seen across Minor League Baseball schedules. We ultimately
decided upon categories of opening day, merchandise giveaways, group
nights, free or reduced-price food, free or reduced-price beer, discount
tickets, dogs to the park, and concerts (in addition to the fireworks
categories listed below). If fans enjoy these promotions, the impact of
these promotions should be positive. It is possible, however, due to
sponsorship dollars gained by having a corporate sponsor of a
promotional item or event, that promotions may be worthwhile to the team
even though they may not generate much fan interest at the gate.
The main focus of this paper is on a special category of promotions
at Minor League Baseball games: fireworks. As described in the
introduction, the number of fireworks shows has been increasing over
time, leading to the question driving this research about possible
saturation in the market for fireworks. Previous tables on fireworks
frequency have shown that the vast majority of teams use fireworks, with
many teams using fireworks double-digit times throughout the season, and
some teams using fireworks over thirty times in a season.
To account for fireworks and fireworks frequency throughout the
season, we decided on a simple combination of three variables. The first
variable, fireworks, is a dummy variable that takes a value of one if
there is a post-game fireworks show on that day. The second variable
related to fireworks, fireworks frequency, notes the number of times
that fireworks have been used that season by that particular team. The
first time fireworks are used is noted as one and then it increases by
one each additional time fireworks are displayed throughout the season.
The third variable accounts for nonlinearity in this relationship and is
the fireworks frequency squared. The model was first tried with only
fireworks and fireworks frequency and was expanded to fireworks,
fireworks frequency, and fireworks frequency squared. Ultimately, we
believe the specification of all three variables in the regression model
best accounts for the relationship between fan interest (attendance) and
fireworks shows. (1)
If fans grow tired of seeing fireworks shows during the season, the
marginal attendance increase associated with each additional fireworks
show should decline. Ultimately, this is likely to take the form of a
statistically significant fireworks frequency squared term, which will
ultimately offset part (or all) of the general impact of fireworks.
Alternatively, it could be a simple linear relationship, where the
number of fireworks has a negative and significant impact on attendance.
If the market for fireworks is not saturated, however, the impact of the
frequency variables should not have a meaningful impact on marginal
attendance, especially at the current levels of fireworks frequency seen
in 2011 Minor League Baseball.
Results of the attendance model for all five leagues are presented
in Table 5 below. (2) For ease of exposition, the results are shown over
multiple tables. Table 5-A notes the results for the intercept, city
demographics, weekday, and monthly dummy variables. Table 5-B describes
the results for temperature, weather category, and nonfireworks
promotions. Table 5-C presents the results for the fireworks-related
variables. A description of the results follows each individual table.
Newey-West HAC standard errors and covariances are used to account for
heteroskedasticity and autocorrelation problems identified when
initially running the regressions. *-notation following the coefficient
denotes statistical significance with * being significant at the 10%
level, ** being significant at the 5% level, and *** being significant
at the 1% level.
In three of the five leagues studied (International, Eastern, and
Southern), population had a positive and significant effect on game
attendance. In the same three leagues, Minor League Baseball was found
to be an inferior good, as income per capita negatively affected game
attendance. This is likely due to the availability of entertainment
options in these cities other than Minor League Baseball. In the Texas
League, however, per-capita income was found to have a positive and
significant effect on attendance.
The days of the week revealed consistent and expected results
across leagues. Weekend days, in addition to Thursday, were shown to
have a positive and significant impact on attendance compared to the
omitted categorical dummy variable, Wednesday. Wednesday was the least
attended day of the week in both AAA leagues and in the Eastern League.
Mondays and Tuesdays were also quite poorly attended in the Southern
League, with Mondays being slightly worse than Wednesdays in the Texas
League.
There were also similarities across leagues in the
month-of-the-season effects. Early-season games tended to suffer when it
came to performance at the gate, with many fewer (1900+ in the
International League in April) fans choosing to attend games in the
presummer months. In terms of statistically significant results, only
games in August and September showed a positive and significant impact
compared to June in the Pacific Coast League, while negative and
significant results were shown in August in the Southern and Texas
Leagues.
Win percentage was found to have a positive and significant effect
on attendance at the AAA level, but did not have a statistically
significant impact on attendance at the AA level. Visiting team win
percentage was only found to have a significant effect on attendance in
the Texas League, where it was found to have a negative impact on
attendance. This result shows that fans of the Texas League prefer to
see games where it is more likely their home team win the game. Distance
from the opposing city to the home team was not found to have a
significant effect in any of the leagues studied.
Temperature was mostly found to be insignificant, except for the
Texas league, where the often oppressive heat in the summer was found to
negatively affect attendance. The categorical weather conditions listed
in the box score of games were found to affect attendance in a few
categories and leagues. Cloudy days in the Eastern League (AA) were
found to have a negative and significant effect on attendance with over
650 fewer fans attending games on these days. Days with rain (both rainy
and drizzling) were found to have negative and significant effects on
attendance. Rain was found to have a significant negative impact on
attendance in the Pacific Coast and Eastern Leagues, while the category
of drizzling was found to negatively affect the Eastern and Texas
Leagues. Clear days were also found to have a statistically significant
(10% level) and negative effect in the Pacific Coast League, perhaps due
to the availability of other outdoor entertainment activities on clear
days in cities in California and other states and regions in this
league.
For the nonfireworks promotions, opening day and merchandise
giveaways were found to have a positive and significant impact on
attendance across all five leagues. Group nights showed mixed results
across leagues, while reduced-price or free food was found to
significantly increase attendance in the Texas League. Beer-related
promotions were shown to have a negative and significant effect on
attendance in both the Eastern and Southern Leagues. This negative
impact is likely due to externalities associated with free or
reduced-price beer that cause some portion of the fan base (likely
families with children) to avoid these games. Discount ticket nights did
not have a positive impact on attendance at the AAA and AA levels and
actually had a negative and significant effect on attendance in the
International League. This result may be capturing the fact that general
managers schedule these discount ticket promotions for games with low
projected attendance. On these discount ticket days, fans were not very
responsive to lowering the price of already inexpensive tickets.
Post-game concerts were found to be a success in the International
League, Eastern League, and Southern League, each attracting over 1,000
additional fans per game (with the International League games attracting
nearly 2,000 more fans).
The dummy variable for having a fireworks show was found to have a
positive and significant effect on attendance across all five leagues
studied. The Southern League exhibited the smallest impact from
fireworks, adding 892 fans for fireworks promotions. The Pacific Coast
League was found to have the largest fireworks effect, with over 1,800
additional fans attending games with fireworks. The impact of fireworks
was between 1,000 and 1,600 fans across the three other leagues.
[FIGURE 1 OMITTED]
Through the use of the fireworks number and fireworks number
squared terms, the impact of possible saturation of fireworks in the
market for Minor League Baseball was investigated. In AA baseball,
statistically significant results of either term were not found. In the
International League, the fireworks number variable was found to have a
positive and significant impact on attendance, implying that more fans
attended games with increased exposure to fireworks promotions. Although
the fireworks number squared variable was found to be negative for the
International League, it was not found to be statistically significant.
The Pacific Coast League was the only league where the fireworks number
squared variable was found to have a negative and significant effect,
implying some sort of fireworks saturation effect within this market.
To more clearly illustrate the impact of the fireworks number and
fireworks number squared variables for the Pacific Coast League, where
there appears to be evidence of saturation in the marketplace, and for
the other leagues where fireworks saturation does not appear to be a
problem, the following figure shows the relationship of the fireworks,
fireworks number, and fireworks number squared variables plotted against
the number of fireworks shows. The vertical axis is the marginal impact
on attendance of fireworks shows [(fireworks dummy value) + (fireworks
number coefficient x the number of fireworks) + (fireworks number
squared coefficient x number of fireworks squared) ], while the
horizontal axis represents the number of fireworks shows in a season.
Each league is separately plotted on Figure 1. Although all five leagues
are presented in the figure, the only statistically significant results
lie in the Pacific Coast League. The other leagues are shown for
comparative purposes.
As can be seen in Figure 1, plotting the relationship of fireworks
frequency and marginal attendance reveals concave relationships in four
of the five leagues, while the Eastern League actually takes a convex
shape over a reasonable range of fireworks shows. The only situation
where the relationship drops negative across the range of fireworks
shows is in the Pacific Coast League. The point where the curve crosses
the zero point on the vertical axis in the Pacific Coast League is
between 18 and 19 events. Interestingly, the highest frequency of
fireworks seen in the Pacific Coast League was 17 times (in Omaha and
Nashville), implying that teams in these leagues implicitly understand
how frequently they can utilize fireworks on their schedule.
Mapping out the relationship estimated from the regression model
reveals that in four of the five leagues, fans do not appear to grow
overly tired of fireworks shows. In the Eastern League, fireworks shows
actually showed an increase in attendance with additional shows. In the
International and Southern Leagues, the drop-off in attendance with
additional fireworks shows begins in the teens. While in the Texas
League, attendance reaches its peak on the 21st fireworks show, with
only a slight decline in attendance seen after that point.
Discussion and Conclusions
As Minor League Baseball teams have increased the frequency of
fireworks shows from only a few times a year on holidays (4th of July,
Memorial Day, Labor Day) to ten, twenty, or even thirty-plus times a
year, the possibility of saturating the market for fireworks exists. If
fans grow tired of seeing fireworks shows over the course of the season,
it is possible that diminishing marginal returns set in as the number of
fireworks shows increases. This could lead to a reduced impact on
attendance with each additional fireworks show added to the promotional
schedule and presents the possibility that too many fireworks shows
could actually lead to a negative marginal impact on attendance for
games.
To test for possible saturation effects in the market for Minor
League Baseball fireworks shows, we studied five leagues at the top
tiers of the minor league system. We gathered data on per-game
attendance, city demographics, weekday and month of game, weather
conditions, nonfireworks promotions, and fireworks shows for the two AAA
leagues (highest level in Minor League Baseball) and the three AA
leagues (next highest level) for the 2011 season.
Within the framework of a regression model, with per-game
attendance as the dependent variable, we controlled for other factors
and examined the impact of fireworks in three ways. First, a dummy
variable was included for having a fireworks show on game night. Second,
a fireworks frequency variable was created that noted the number of
times fireworks shows had been used by that particular team during the
season, starting with one for the first time and increasing by one for
each additional show throughout the season for the team. Third, to
account for the possibility of a nonlinear relationship and to
illustrate the possible saturation effects within this market, the
square of the fireworks frequency variable was also included in the
model.
Overall, fireworks had a significant and sizeable positive impact
on attendance across the five leagues studied. The results of fireworks
having some of the largest, if not the largest, increases in attendance
as promotional nights for Minor League Baseball teams are similar to the
findings of other papers that have investigated the use of fireworks at
minor league parks (e.g., Gifis & Sommers, 2006; Cebula, Toma, &
Carmichael, 2007; Cebula, 2013).
In terms of fireworks frequency, despite some teams using fireworks
more than 20 or 30 times a season, increased use of fireworks shows over
the course of the seasons did not reveal much of an impact across the
five leagues studied. In 4 of the 5 leagues, the relative decline in
marginal attendance associated with each additional fireworks show was
rather small. In 3 of the 5 leagues studied, International (AAA),
Southern (AA), and Texas (AA), marginal increases in attendance were
seen with each additional fireworks show well into the teens in terms of
frequency over the course of the season. The Eastern League (AA)
actually showed a convex relationship, where teams with a very high
number of fireworks shows showed increases in attendance with additional
fireworks shows late in the season. The only league where a large
saturation effect was noticed was with the Pacific Coast League (AAA),
where the marginal impact on attendance at original fireworks shows was
shown to have zero marginal impact around the 18th show of the season.
The fact that no team in the Pacific Coast League had more than 17
fireworks shows in the 2011 season suggests an implicit understanding of
this relationship on the part of team management.
Overall, increased frequency of fireworks shows throughout the
season does not appear to present too many difficulties for Minor League
Baseball teams as it relates to saturation within the marketplace. It
appears that across the AAA and AA levels of Minor League Baseball teams
can increase the frequency of their fireworks shows well into the teens,
with the distinct possibility of financial advantages to the team still
possible with even 20, 30, or more fireworks shows during the season.
The range of observable frequencies of fireworks shows that in AAA
and AA baseball fans do not appear to grow tired of fireworks shows.
Increases in the number of fireworks shows over time have not seemed to
reach a point where fans treat fireworks shows as ordinary. Fireworks
nights are some of the most popular, if not the most popular, nights on
teams' promotional schedules and increasing their frequency appears
to have been quite successful for the teams.
Perhaps the result that the market for fireworks has not become
saturated in the minds of consumers of Minor League Baseball should not
be surprising. Places like Disney World, Six Flags theme parks, and
other amusement businesses tend to have fireworks nearly every night
they are open. Although consumers may attend these parks for many days
over the course of a vacation, many attendees are likely to watch the
fireworks shows multiple times and still gain enjoyment from their
display. Minor league baseball teams, through their own experimenting
with their promotional schedule, and following the lead of other
entertainment ventures (such as Disney World) have increased the
frequency of their fireworks shows without detrimental marginal effects.
In fact, most teams and leagues studied here find little decline over
the course of the season when adding additional fireworks shows,
effectively offering fans more of what they desire (post-game fireworks)
and increasing their ticket sales and revenues along the way.
References
Anthony, T., Kahn, T., Madison, B., Paul, R., & Weinbach, A.
(2011). Similarities in fan preferences for minor-league baseball across
the American southeast. Journal of Economics and Finance. In press,
online first, December, 2011.
Butler, M. (2002). Interleague play and baseball attendance.
Journal of Sports Economics, 3, 320-334.
Coates, D., & Humphreys, B. (2005). Novelty effects of new
facilities on attendance at professional sports events. Contemporary
Economic Policy, 23, 436-455.
Demmert, H. G. (1973). The economics of professional team sports.
Lexington, MA: Lexington Books.
Depken, C. A. II (2006). The impact of new stadiums on professional
baseball team finances. Public Finance and Management, 6, 436-474.
Gifis, L., & Sommers, P. (2006). Promotions and attendance in
Minor League Baseball. Atlantic Economic Journal, 34, 513-514.
Gitter, S. R., & Rhoads, T. A. (2010). If you win they will
come: Fans care about winning in Minor League Baseball. Journal of
Sports Economics, 11, 614-628.
Kahane, L., & Shmanske, S. (1997). Team roster turnover and
attendance in Major League Baseball. Applied Economics, 29, 425-431.
McDonald, M., & Rascher D. (2000). Does bat day make cents? The
effect of promotions on demand for Major League Baseball. Journal of
Sport Management, 14, 8-27.
Noll, R. (1974). Attendance and price setting. In R. Noll (Ed.),
Government and the sports business (115-158). Washington, DC: The
Brookings Institute.
Paul, R. J., & Weinbach, A. P. (2011). Minor League Baseball
attendance in the Pacific Northwest. Ekonomika a Manageament (Economics
and Management), 2011.
Paul, R. J., Weinbach, A. P., & Melvin, P. (2004). The Yankee
effect: The impact of interleague play and the unbalanced schedule on
Major League Baseball attendance. New York Economic Review, 35, 3-15.
Richards, D. G., & Guell, R. C. (1988). Baseball success and
the structure of salaries. Applied Economics Letters, 5, 291-296.
Schofield, J. A. (1983). Performance and attendance at professional
team sports. Journal of Sport Behaviour, 6, 196-206.
Siegfried, J., & Eisenberg, J. (1980). The demand for Minor
League Baseball. Atlantic Economic Journal, 8, 56-69.
Villar, J. G., & Guerrero, P. R. (2009). Sports attendance: A
survey of the literature 1973-2007. Rivista di Diritto ed Economia dello
Sport, 5, 111-151.
Whitney, J.D. (1988). Winning games versus winning championships:
The economics of fan interest and team performance. Economic Inquiry,
26, 703-724.
Rodney J. Paul [1] and Andrew P. Weinbach [2]
[1] Syracuse University
[2] Coastal Carolina University
Endnotes
(1) We also ran the regression with a long series of individual
dummy variables for the number of times fireworks were used by each team
in the league. Results were similar to the results described in the
paper. These results are available upon request from the authors.
(2) R-squared for each regression is as follows: International
League (0.4310), Pacific Coast League (0.2987), Eastern League (0.4826),
Southern League (0.3976), and Texas League (0.4465).
Rodney J. Paul is a professor in the Department of Sport Management
at Syracuse University where he specializes in the economics and finance
of sport, macroeconomics, and international economics. He has presented
at conferences nationally and internationally, and his work has appeared
in a number of journals and book chapters on sport economics and
business.
Andrew P. Weinbach is an associate professor in the E. Craig Wall
Sr. College of Business Administration at Coastal Carolina University.
His research interests include applied microeconomics, sports economics
(patterns of consumer interest in live sporting events, including fan
attendance, television ratings, and betting participation), industrial
organization, and financial economics.
Table 1: Summary Statistics by League
International Pacific Coast Eastern
League League League
(AAA) (AAA) (AA)
Attendance mean 7,015 5,988 4,975
Attendance median 6,830 5,561 4,915
Attendance std. dev. 2,873 2,617 2,038
Population mean 333,365 407,372 84,312
Population median 242,803 416,467 73,206
Population std. dev. 250,772 170,117 43,687
Income per cap mean 22,677 24,237 22,462
Income per cap median 21,472 24,596 19,493
Income per cap std. dev 5,310 2,649 7,141
Temperature mean 75 79 74
Temperature median 78 82 77
Temperature std. dev. 12 15 11
Southern Texas
League League
(AA) (AA)
Attendance mean 3,152 5,162
Attendance median 2,877 4,888
Attendance std. dev. 1,685 1,784
Population mean 216,681 340,384
Population median 178,874 193,524
Population std. dev. 223,849 395,183
Income per cap mean 22,118 25,359
Income per cap median 21,957 26,072
Income per cap std. dev 3,537 6,563
Temperature mean 84 89
Temperature median 86 91
Temperature std. dev. 10 12
Table 2: Weather Category and Promotions Frequencies by League
International Pacific Coast Eastern
League League League
(AAA) (AA) (AA)
Partly cloudy 358 310 280
Cloudy 273 177 176
Clear 236 422 235
Sunny 72 71 75
Rainy 16 16 10
Drizzling 11 12 12
Overcast 46 61 0
Opening day 16 16 11
Merchandise 186 185 193
Fireworks 191 192 201
Group 104 487 214
Food 126 206 66
Beer 31 87 76
Discount tickets 161 118 206
Dogs to park 15 14 21
Concerts 17 24 29
# observations 1012 1069 788
Southern Texas
League League
(AA) (AA)
Partly cloudy 255 191
Cloudy 129 52
Clear 202 271
Sunny 46 28
Rainy 6 0
Drizzling 1 3
Overcast 0 23
Opening day 9 8
Merchandise 64 131
Fireworks 114 95
Group 195 107
Food 81 76
Beer 65 39
Discount tickets 94 114
Dogs to park 10 8
Concerts 22 11
# observations 639 568
Table 3: Fireworks Frequencies by Team for 2011 MiLB Season--AAA
International Fireworks Pacific Coast Fireworks
League frequency League frequency
Buffalo 18 Albuquerque 10
Durham 18 Colorado Springs 9
Charlotte 13 Iowa 14
Columbus 5 Fresno 14
Gwinnett 11 Las Vegas 5
Indianapolis 12 Memphis 9
Lehigh Valley 16 Nashville 17
Louisville 15 New Orleans 12
Norfolk 9 Oklahoma City 14
Pawtucket 6 Omaha 17
Rochester 21 Reno 15
Scranton/WB 12 Round Rock 15
Syracuse 8 Sacramento 13
Toledo 26 Salt Lake City 9
Tacoma 15
Tucson 4
Table 4: Fireworks Frequencies by Team for 2011 MiLB Season--AA
Eastern Fireworks
League frequency
Altoona 19
Binghamton 20
Bowie 20
Erie 9
Harrisburg 12
New Briton 10
New Hampshire 30
Portland 5
Reading 32
Rochester 22
Southern Fireworks
League frequency
Birmingham 18
Carolina 13
Chattanooga 15
Huntsville 0
Jacksonville 14
Jackson 14
Mississippi 12
Mobile 0
Tennessee 15
Texas Fireworks
League frequency
Arkansas 3
Corpus Christi 13
Frisco 16
Midland 11
NW Arkansas 16
San Antonio 11
Springfield 12
Tulsa 13
Table 5-A Minor League Baseball Attendance Model--Demographics,
Days, Months
International Pacific Coast Eastern
League League League
(AAA) (AAA) (AA)
Intercept 7309.781 *** 5813.703 *** 4676.587 ***
Population 0.0017 *** 0.00009 0.0167 ***
Income -0.1714 *** -0.4661 -0.0399 ***
Sunday 596.9434** 1007.590 *** 665.6975 ***
Monday 259.8857 119.7745 129.1654
Tuesday 571.8509 ** 80.3541 49.5631
Thursday 750.8177 *** 439.2510 * 373.1304 *
Friday 921.6393 *** 880.8712 *** 757.9834 ***
Saturday 1529.164 *** 1696.877 *** 1149.432 ***
April -1905.774 *** -1601.886 *** -1721.560 ***
May -963.4507 *** -408.5644 -621.6619 ***
July 282.0655 363.0001 136.3867
August 381.9272 574.1225 * 110.9559
September 912.6473 852.2727 * -25.5043
Southern Texas
League League
(AA) (AA)
Intercept 2680.270 ** 6826.677 ***
Population 0.0015 *** 0.0001
Income -0.5219 * 0.1182 ***
Sunday 47.2498 483.2296 *
Monday -45.3585 -39.9049
Tuesday -121.69 58.0932
Thursday 1024.437 *** 517.2146 **
Friday 362.7559 911.8899 ***
Saturday 1199.863 *** 1545.316 ***
April -102.8820 -1127.162 ***
May 224.8959 -447.3246 *
July -6.9600 -167.8915
August -454.6245 * -444.3399 **
September -766.8636 -36.7136
Table 5-B Minor League Baseball Attendance Model--Win Percentage,
Weather, Promos (Nonfireworks)
International Pacific Coast Eastern
League League League
(AAA) (AAA) (AA)
Win % 2296.446 ** 3052.709 *** -644.7778
Opp. win % 267.1338 -964.3709 272.3190
Distance 0.6996 -0.1532 0.2149
Temperature 10.3481 -8.8193 -7.8039
Cloudy 16.4735 -93.4428 -668.0490 ***
Clear -117.2755 -329.6222 * 26.7392
Sunny 125.4362 -120.7051 86.3536
Rainy -328.9915 -1748.342 *** -1433.407 ***
Drizzling -530.1078 -891.4772 -1041.193 **
Overcast 7.2399 -381.1136
Opening day 3647.609 *** 2077.092 *** 2060.581 ***
Merchandise 715.1980 *** 470.4060 ** 331.4087 ***
Group -638.9716 ** 469.8240 *** 581.6479 ***
Food 51.8970 -6.3889 -164.9732
Beer 223.9100 320.9583 -552.2472 **
Disc. tickets -1003.258 *** 90.5262 -68.5331
Concerts 1931.957 *** 637.1199 1006.334 ***
Southern Texas
League League
(AA) (AA)
Win % -580.7215 -1569.816
Opp. win % 92.5554 -2105.585 **
Distance 0.3671 -0.2720
Temperature 8.6974 -37.7640 ***
Cloudy 1.9648 -28.9287
Clear 133.4777 71.1647
Sunny 426.3894 244.8204
Rainy 260.5367
Drizzling -1014.316 -1098.715 **
Overcast -3.2166 ***
Opening day 1656.475 *** 977.1821 **
Merchandise 451.0689 ** 296.5533*
Group 111.3207 -285.1888 *
Food 90.3159 758.1683 ***
Beer -1067.228 *** 181.6710
Disc. tickets -52.7640 84.0881
Concerts 1142.575 *** 157.3840
Table 5-C Minor League Baseball Attendance Model--Fireworks
International Pacific Coast Eastern
League League League
(AAA) (AAA) (AA)
Fireworks 1019.253 ** 1837.794 *** 1590.823 ***
Fireworks number 180.7091 ** 246.1657 -83.2247
Fireworks number (2) -5.6839 -19.0648 ** 3.4058
Southern Texas
League League
(AA) (AA)
Fireworks 891.8914 * 1125.768 **
Fireworks number 162.0472 64.0257
Fireworks number (2) -7.0278 -1.4968
