Citation
Location location location

Material Information

Title:
Location location location an econometric study of the placement of professional sports teams
Creator:
Rebideau, Jacqueline
Publication Date:
Language:
English
Physical Description:
42 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Sports franchises -- Location -- United States ( lcsh )
Sports franchises -- Location ( fast )
United States ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 41-42).
General Note:
Department of Economics
Statement of Responsibility:
by Jacqueline Rebideau.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
49664852 ( OCLC )
ocm49664852
Classification:
LD1190.L53 2001m .R42 ( lcc )

Full Text
LOCATION LOCATION LOCATION
AN ECONOMETRIC STUDY OF THE PLACEMENT OF PROFESSIONAL
SPORTS TEAMS
by
Jacqueline Rebideau
A.B., Occidental College, 1992
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
Economics
2001


This thesis for the Master of Arts
degree by
Jacqueline Rebideau
has been approved
by
Brian J. Duncan
Steven G. Medema
s<
S2Z
Date


Rebideau, Jacqueline (M.A., Economics)
Location Location Location: An Econometric Study of the Placement of
Professional Sports Teams
Thesis directed by Assistant Professor Brian J. Duncan
Between 1970 and 2000, 21 professional sports teams relocated and 39
expansion franchises were introduced. The purpose of this study is to
evaluate the characteristics that identify team locations; an analysis to
determine if certain economic and demographic conditions heighten the
appeal of some locations over others. In this paper, I first attempt to identify
these characteristics then use 1980 and 1990 data to evaluate the importance
of each. Next, using 2000 data, I use the most accurate model to predict
future locations of professional sports teams. The evaluation of my three
models indicate that only total population, the distance between teams in the
league, and the number of teams from other sports located in the city are
important determinants of team location.
This abstract accurately represents the conte*" t
recommend its publication.
ABSTRACT
Signed
Brian J. Duncan
in


CONTENTS
Tables......................................................................v
Chapter
1. Introduction............................................................1
2. Background Issues.......................................................3
2.1 Why Cities Seek Teams...................................................3
2.2 Where Teams Locate......................................................5
3. Theoretical Framework...................................................7
3.1 Local Effects...........................................................7
3.2 League Effects...................................................... 10
4. Data...................................................................12
4.1 Dependent Variable.....................................................12
4.2 Independent Variables.................................................15
5. Discussion of Results..................................................23
5.1 Model 1: Past Empirical Work...........................................23
5.2 Model 2: Expanded Model................................................27
5.3 Model 3: Predicting Future Team Locations..............................30
6. Conclusion.............................................................33
Appendix...................................................................35
Bibliography...............................................................41
IV


TABLES
Table
4.1 Variables and Definitions.............................................13
4.2 Percent of the 56 Sample MSAs Hosting Teams..........................14
4.3 Independent Variable Means for 1980..................................16
4.4 Independent Variable Means for 1990..................................17
4.5 1980 Cities Below Median Population..................................18
4.6 1990 Cities Below Median Population..................................18
4.7 Number of Teams in Each Region.......................................20
4.8 Average Distance Between Teams.......................................21
5.1 Model 1 OLS Regression Results........................................24
5.2 Expansion/Relocation Sites Between 1981 and 2000......................26
5.3 Model 2 OLS Regression Results........................................28
5.4 Expansion/Relocation Sites Between 1981 and 2000.....................29
5.5 Model 3 OLS Regression Results.......................................31
5.6 Team Placement (2001/02 2010/11 seasons)...........................32
v


1. Introduction
As professional sports have grown into American big business, they have also
become the focus of much research and analysis. The market structure of the industry
and the debate over public funding of stadiums have evolved into a myriad of studies.
The monopolistic structure of professional sports resulting in limited supply, excess
demand, and inflated franchise prices though, has not deterred communities from
reaching deep into their pockets to obtain a team. In fact, each time a new franchise is
introduced or an existing team is exploring relocation, many suitors emerge. Yet,
even with all the interest in professional sports, little research has addressed what
characteristics, if any, affect the placement of these teams.
Between 1970 and 2000, 39 expansion franchises were introduced: 5 in the
National Football League (NFL), 6 in Major League Baseball (MLB), 12 in the
National Basketball Association (NBA), and 16 in the National Hockey League
(NHL). This resulted in league totals of 31, 30, 29, and 30 teams during the 2000/01
season, respectively. Over this time, too, the cost associated with obtaining a new
franchise increased substantially. Beginning play in 1998, the new owners of
baseballs Arizona Diamondbacks and Tampa Bay Devil Rays had to each pay the
league $130 million. Likewise, the four new teams introduced by the NHL between
1998 and 2000 required an expansion fee of $80 million each. These compare to the
1


$7 million expansion price for the Seattle Mariners (MLB) and Toronto Blue Jays
(MLB) in 1977 and $6 million spent on the hockeys Buffalo Sabres and Vancouver
Canucks in 1970.
This 30 year period also witnessed the movement of many existing teams. In
fact, 21 teams relocated: 6 in the NFL, 2 in MLB, 7 in the NBA, and 6 in the NHL.
Although many factors may contribute to a team's relocation, poor financial success
and low fan support top the list. Yet even after some of these cities lose one
professional sports team they are able to acquire another within a few years. For
example, hockey returned to Minneapolis and Denver, while a football team again
plays in Cleveland and will be in Houston again in the 2002 season. These
communities, therefore, must contain adequate resources to support teams. Since
these cities obviously were viable locations, the question remains: what factors led to
the original loss and reacquisition of a professional sports team?
The purpose of this study is to evaluate the characteristics that identify team
locations; an analysis to determine if certain economic and demographic conditions
heighten the appeal of some locations over others. In this paper, I first attempt to
identify these factors and use 1980 and 1990 data to evaluate the importance of them.
Then, using 2000 data, I use the most accurate model to predict future locations of
professional sports teams.
2


2. Background Issues
2.1 Why Cities Seek Teams
Cities enthusiastically pursue professional sports teams. Although they offer
cities various benefits, scientific research has shown that professional sports teams are
not necessarily wise financial investments for a city. In fact, economists (Baade,
1987; Noll and Zimbalist, 1997; Baade and Sanderson, 1997) have shown that the
actual social economic growth associated with professional sports is close to zero.
This is not to say that there are no economic effects realized by hosting a team, rather
the overall impacts net near zero. For instance, new employment opportunities are
introduced though most are unskilled, seasonal jobs. New restaurants and bars open
in the vicinity of the stadium or arena. But this simply results in a transfer of wealth
from one area of the city to another rather than an introduction of additional monies
into the local economy. In addition to these financial gains, the community also
enjoys utility gain from hosting a local team. Despite the lack of obvious financial
gain, the continuing desire to obtain a team allows the leagues to maintain their
leverage over cities, especially when coupled with the strategic introduction of
franchise teams. Because expansion is a relatively slow process, it enhances the
intercity competition for teams. And the most blatant form of this competition has
become the public stadium subsidy.
3


During the 2000/01 season, professional teams played in 97 different playing
facilities. Of these, 47 were less than ten years old and nearly all had at least partial
public contribution. Many of the other 50 facilities are in the process of being
replaced. In fact, stadium construction had become such an important requisite for
entry into the sports industry that in 1990 the Fort Lauderdale community chose to
construct a $138 million publicly funded baseball stadium simply to increase its
chances of luring the Chicago White Sox. After attempts to recruit the Seattle
Mariners and the San Francisco Giants also proved unsuccessful, Tropicana Field
(nee Sun Coast Dome) became the home of the expansion franchise Tampa Bay Devil
Rays. A $70 million renovation readied the dome for their inaugural season in 1998.
On the other hand, Los Angeles appears to be the only city not willing to
commit public dollars to these ventures. Both Dodger Stadium and the $375 million
Staples Arena were both privately funded. The city also saw the 32nd NFL franchise
slip through its fingers after a lack of public funding impeded a suitable stadium
proposal. Houston, on the other hand, ensured a $310 million facility ($195 million
from taxpayers) and secured the 2002 expansion team.
The pressures for public subsidies, the introduction of jobs, and the creation of
dining establishments and bars are no doubt similar for all team locations. This paper
therefore will not address these issues, rather this study will focus on community
factors that may influence the placements of professional sports team.
4


2.2 Where Teams Locate
The apparent prerequisites for an expansion franchise would appear simple: a
city should have a population greater than 900,000 and should not be within a
reasonable drive from another major league baseball franchise. (Sheehan,
1996)
Little research, though, has examined whether population and distance, as well
as additional variables, deem important in the process of selecting future sports team
locations. The following studies, though, addressed this issue to varying extents.
Noll (1974) focused on demographic and economic factors as they influenced
the demand for baseball games. Understanding that greater audience for games
results in greater profits, owners and leagues strive to maximize attendance by
hosting teams in markets that yield the largest demand. Four factors that were
discussed in the study did indeed relate to the characteristics of the city. They
included: population, income, sports competition, and percent of population that is
black. The other six variables pertained to either the team profile or the playing
facility.
A study by Jozsa and Guthrie (1999) examined factors that play a role in the
relocation and expansion placement of football, baseball, and basketball teams. The
structure of this study was such that it reviewed data on teams after they had been in
new locations rather than before they were placed. Similar to the study by Noll this
too included population and income, but also included discussed population growth
was also an important factor.
5


Bruggink and Zamparelli (1999) published the first study that explored the use
of econometric models to predict future locations of professional baseball teams.
Their study attempted to identify factors that make one location more appealing than
others for current and prospective MLB owners. Their final model included:
population, population growth, income, the number of other sports teams located in
the city, the number of Fortune 500 companies, and the distance to the closest team.
A previous model had also included the number of households with televisions but
was later excluded due to its correlation with population. The authors then used these
data to predict and defend future placements of baseball teams. Unfortunately, no
new MLB placement has occurred since the release of this study so validation of their
predictions cannot yet be done.
Many of the economic and demographic characteristics explored in these three
earlier studies will also be incorporated in this study, as well as examining additional
factors.
6


3. Theoretical Framework
Franchise relocation is a natural adjustment to market conditions, and no
person or institution is better equipped than an owner to make such decisions.
After all, owners want to maximize their profits. To do this, they must
respond to market conditions. (Jozsa & Guthrie, 1999)
While the individual team owners seek to maximize their own profits and
increase the value their franchises, the league acts as a single entity and seeks these
same goals but in the aggregate. The structure of sports leagues is such that all
proposed expansion franchise locations and existing team relocations are subject to
majority league approval.1 Therefore a site is likely to be approved only if it is in the
best interest of all the league. To do so, one must measure both (a) the size and
composition of the local market and (b) the intraleague and interleague effects.
3.1 Local Effects
In recent years, sports leagues have seen great increases in revenue from game
attendance. During the 1999/00 season, the average price of an individual game
ticket was $45.63 for the NFL, $16.65 for MLB, $48.37 for the NBA, and $45.70 for
the NHL. This is roughly a 100 percent increase in the last decade alone. Although
the importance of attendance varies by sport dependent upon the number of games
per season and the stadium/arena capacity a teams ability to fill the stands has
The four sports leagues require a 75 percent approval on all proposed sites. In the MLB, though, the league in which the team
plays (National League/American League) must approve at 75 percent while the other league approves at 50 percent.
7


become increasingly lucrative. Combining the rise in game prices, with the fact that
Irani (1997) showed that a 10 percent increase in population resulted in a 1.2 percent
increase in attendance, emphasizes the importance of evaluating potential market
demand.
Three measures are used as a proxy for the potential professional sports
audience: (1) total population, (2) population density, and (3) population growth.
Total population and population density indicate current demand with the former
illustrating the entire consumer base and therefore acts as a proxy for consumption.
The latter demonstrates a fan's willingness to support a particular team. That is,
consumers in the Rocky Mountain area would be more willing to drive further
distances to attend a sporting even than, say, those in the densely populated
northeastern states. Population growth, on the other hand, estimates the future market
for this team by showing if the consumer base is increasing, decreasing, or remaining
near constant.
Revenue from game attendance is strongly determined by the segment of the
audience that fills the luxury suites. As a by-product of the rampant stadium
construction craze, luxury boxes have proven an irreplaceable means of revenue.
Although the first boxes were introduced in 1965 with the Houston Astrodome, their
increasing popularity corresponded with the increasing number of new stadiums in
the 1980s and 1990s. Each season, local businesses spend millions of dollars for
8


these suites. The number of Fortune 500 companies headquartered in each city,
would act as an estimate of this segment of the audience.
In addition to the size of the audience, certain demographic characteristics can
influence demand. Although the prediction of consumer behavior is far from an exact
science, market composition could bias the demand for professional sports.
First, consumers generally allocate a certain amount of their income to spend
on recreational activities. This money, for example, can be spent on seeing a favorite
sports team play, but then it is no longer available to attend other events. As game
tickets approach $50 each, combined with parking and concession expenses, a family
of four is likely to spend up to $300 for a single game attendance. That's a number of
forgone movies or an evening of dinner and a concert. Noll (1974) claimed that game
attendees were blue collar workers which resulted in a negative correlation between
income and attendance. The escalating price of game tickets, though, has likely
caused sports to become a luxury good shifting patronage to middle class fans with
higher disposable income.
Second, it has been suggested (Noll, 1974) that the racial composition of
players may influence fan preference. For example, the MLB has, by far, the greatest
representation of hispanic players. The 25 percent representation compared to less
than one percent representation in the other sports.2 Likewise, black Americans are
2 All statistics from the Northeastern Universitys Center of the Study of Sport in Society for 1997/98 and 1998 seasons.
9


more highly represented in basketball and football. The NBA consisted of 77 percent
black players while 65 percent of professional football players were black. MLB,
though, had only a 15 percent representation and the NHL could count the number of
its black players on two hands. Unfortunately, in practice, the correlation between
player and attendee race is uncertain. The recent financial burden of game attendance
may outweigh the natural tendency of fans.
Third, weather may influence attendance. Noll (1974) stated that weather
certainly plays a role in baseball game attendance arguing that fans in good weather
areas have the option of doing other outdoor activities. Therefore, he contended that
the fewer sunny days a city experiences the greater game attendance. On the other
hand, the construction of numerous domed stadiums has allowed fans to attend games
in comfort even in more inclement weather.
3.2 League Effects
It is in the best interests of all owners in a league to provide competitive
performances and rivalries throughout the divisions and conferences, thereby
maximizing attendance and league profits. (Jozsa & Guthrie, 1999)
In order to analyze the geographical competition, three variables were
included in this study: the region of the United States in which the city exists, the
distance from the city to the next nearest team in that league, and the number of teams
from other sports located in the city.
Although the northeastern and midwestem regions of the United States were
10


the original homes of professional sports teams, more cities in the south and west now
host teams than the midwest and northeast. Of the 39 cities hosting teams during
2000/01, 7 were located in the northeast, 10 in the midwest, 12 in the south, and 10 in
the west. Today, then, would the leagues favor the two original regions for
prospective locations because they have experienced a longer history with them? Or
does the novelty of the south and west continue to look more favorable?
If a city already hosts a professional team, is it more attractive as a potential
site for other sports? Or does it simply reflect that the same demographic and
economic conditions that attract one team ensure appeal to another? Noll (1974) and
Bruggink and Zamparelli (1999) argue the former, stating that if a city already hosts
one team this indicates that it is a "sports town." That is, since the location has the
ability to support one professional team, its consumer base is also able to support
another.
On the other hand, it could be argued that additional teams results in potential
entertainment competition. Since the season of professional sports partially overlap,
this increases the amount of competition. And limited entertainment budgets have
caused consumers to carefully choose their activities.
11


4. Data
The 12 sets of data used in this analysis captured the 4 professional sports
over 3 time periods. The 1980 and 1990 data allowed for the validation of predicted
locations versus the actual placement of professional sports teams while the 2000 data
were used to predict future locations. Statistics for each sport and for each year were
similarly collected. Since each sport was included as a separate data set, expressions
like nearest team and between teams refer to those teams that are within the same
sport unless otherwise noted. Also note that due to the difficulty in acquiring certain
types of data, only U.S. statistics have been included. That is, Canadian cities are not
included as possible expansion/relocation sites. Table 4.1 defines these variables.
4.1 Dependent Variable
The geographical areas to be used in this study include the most populous
metropolitan statistical areas (MS As) as defined by the U. S. Census Bureau for 1998.
By definition of the U.S. Office of Management and Budget, an MSA consists of a
core area containing a large population nucleus, together with adjacent communities
having a high degree of social and economic integration with that core.3 The MSAs
in each data set were then assigned a 'O' or a T.' The former if the area hosted no
3 For the purposes of this paper, though, the terms MSA and city have been used interchangeably.
12


Table 4.1-Variables and Definitions
Variable uelimtion
TEAM = 1 if the metropolitan areag hosted at least one professional team; 0 otherwise
POP3 = total population of the metropolitan areag
GROWTH3 = percentage of population growth of the metropolitan areag
DENSITY3 = population density (per square kilometer) of the metropolitan areag
INCOME^ = average per capita income of the metropolitan areag
COMPANY0 = number of Fortune 500 companies located in the metropolitan areag
SOUTH^ = 1 if metropolitan areag is located in the southern region of the U.S.; 0 otherwise
MIDWEST^ = 1 if metropolitan areag is located in the midwestem region of the U.S.; 0 otherwise
WEST^ = 1 if metropolitan areag is located in the western region of the U.S.; 0 otherwise
BLACKe = percent of total population that is black
HISPANICe = percent of total population that is hispanic
JULYf = inches of average July rainfall received in the metropolitan areag
SNOWf = inches of average annual snowfall received in the metropolitan areag
ALLSPORT = number of teams from the other three sports leagues located in metropolitan areag
NFLSPORT = number of NFL teams located in the metropolitan areag (excluded if dependent
variable is NFL)
MLBSPORT = number of MLB teams located in the metropolitan areag (excluded if dependent
variable is MLB)
NBASPORT = number of NBA teams located in the metropolitan areag (excluded if dependent
variable is NBA)
NHLSPORT = number of NHL teams located in the metropolitan areag (excluded if dependent
variable is NHL)
DISTANCE = miles to closest metropolitan areag hosting a team of this sport
The values for total population are in millions. The data for total population and population growth rates were obtained
from the U.S. Census Bureau for the years 1980, 1990, and 1998. Values for 1980 growth were calculated using 1970 and
1980 total population data; 1990 with 1980 and 1990, and 2000 estimates with 1990 and 1998. Population densities were
calculated. They were computed utilizing the total population values and the land area (in square kilometres) of each MSA
as reported in 1992 County and City Data Book Extra.
bThe values for average per capita income were that of the 1979,1989, and 1998 U.S. Census Bureau.
cThe number of Fortune 500 companies located in each MSAs was obtained by reviewing the lists published in the May
05, 1980; April 23 1990; and April 17, 2000 editions of Fortune magazine.
dDummy variables were used for MSAs located in the midwestem, southern, and western regions of the United States.
"The data for the percentage of black and hispanic populations were obtained by the 1992 County and City Data Book
Extra, with included the entire MSA, and the 1997 U.S. Census Bureau, which only included the nucleus city.
rAU weather statistics were obtained by the U.S. National Oceanic and Atmospheric Administration (for July data) and the
Climate Diagnostic Center (for snowfall). Because not all of the individual areas in New Jersey were listed, the data of
nearest cities were used as estimates.
Metropolitan area is defined here as a core area containing a large population nucleus, together with adjacent communities
having a high degree of socal and economic integration with that core. (U.S. Census Bureau, 1998)
13


team in that league, the latter if one or more teams of that sport were located there.
This variable, TEAM, is the dependent variable in each of the models.
To determine the eligible sites, I restrict my sample to the 56 MS As with the
greatest population. This ensures that at least half of the dependent variables in each
of the 12 regressions has a value equal to 0.4 Note that contained within these 56
MS As were actually 59 areas. Recognizing the close proximity of three pairs of
them, I decided to combined and included as a single unit: Los Angeles with Orange
County, the District of Columbia with Baltimore, and Dallas with Fort Worth. In the
Appendix, Tables Al, A2, and A3 illustrate the location of teams during the 1980/81,
1990/91, and 2000/01 season, respectively.
Summarizing the data in these tables, the percent of cities that hosted teams
(i.e., dependent variable equal to 1) are listed in Table 4.2.
Table 4.2 Percent of the 56 Sample MSAs Hosting Teams
1980/81 Season 1990/91 Season 2000/01 Season
NFL 45% NFL 46% NFL 50%
MLB 36% MLB 36% MLB 43%
NBA 41% NBA 46% NBA 46%
NHL 23% NHL 23% NHL 39%
There is no difference between the 1980/81 and 1990/91 MLB and NHL percentages.
This is a result of no expansion teams being introduced during those decades, though
a relocation occurred in hockey but none in MLB. Both leagues, though, introduced
4 The NFL held teams in 28 cities during the 2000/01 season. This was the most of any data set.
14


expansion team between 1990 and 2000 and therefore there are increases in Table 4.2.
The NFL also introduced no expansion teams between 1980 and 1990, but the
interregional relocation of the Baltimore Colts to Indianapolis resulted in the NFL
increase from 45 percent to 46 percent in Table 4.2. Expanions teams then increased
this total to 50 percent in 2000/01. The NBA was the only league to introduce
introduce expansion during the 1980 and 1990 periods, but experienced no activity
between 1990 and 2000.
4.2 Independent Variables
In order to validate whether certain demographic and economic conditions
created more appealing markets for owners, data on a number of independent
variables were collected and analyzed. Tables 4.3 and 4.4 include the means of these
variables for 1980 and 1990, respectively.
15


Table 4.3 Independent Variable Means for 1980
1980 MEANS ALL MSA WITH NFL W/O NFL WITH MLB W/O MLB WITH NBA W/O NBA WITH NHL W/O NHL
POP3 1.98 3.09 1.08 3.50 1.13 3.04 1.25 4.00 1.37
GROWTH3 13.70% 8.30% 18.00% 6.50% 17.70% 10.80% 15.70% 1.90% 17.30%
DENSITY3 327 445 232 494 234 479 222 627 237
INCOME0 $13,282 $13,732 $12,920 $14,000 $12,883 $13,895 $12,855 $13,788 $13,130
COMPANY' 9 17 1 20 3 17 4 21 6
SOUTH" 20 7 13 4 16 5 15 1 19
MIDWEST" 11 8 3 8 3 6 5 4 7
WEST3 12 5 7 4 8 8 4 2 10
BLACK8 24.40% 30.70% 19.20% 31.30% 20.50% 25.60% 23.50% 31.90% 22.10%
HISPANIC8 9.30% 9.80% 8.90% 7.60% 10.20% 12.20% 7.30% 9.30% 9.30%
JULY' 3.31 3.41 3.22 3.09 3.43 2.69 3.74 3.18 3.34
SNOW' 21.85 24.98 19.32 23.59 20.88 22.47 21.41 39.66 16.46
ALLSPORT - 2.16 0.23 2.50 0.42 2.13 0.52 2.69 0.93
NFLSPORT 0.5 - - 1.15 0.14 0.90 0.24 1.08 0.33
MLBSPORT 0.43 0.96 0.00 - - 0.87 0.12 1.00 0.26
NBASPORT 0.41 0.68 0.19 0.80 0.19 - - 0.62 0.35
NHLSPORT 0.25 0.52 0.32 0.55 0.83 0.39 0.15 - -
DISTANCE (NFL) 211 270 163 - - - - - -
DISTANCE (MLB) 253 - - 258 250 - - - -
DISTANCE (NBA) 249 - - - - 268 235 - -
DISTANCE (NHL) 422 - - - - - - 382 434
The values for total population are in millions. The data for total population and population growth rates were obtained from the U.S. Census Bureau for the years 1980,
1990, and 1998. Values for 1980 growth were calculated using 1970 and 1980 total population data; 1990 with 1980 and 1990, and 2000 estimates with 1990 and 1998.
Population densities were calculated. They were computed utilizing the total population values and the land area (in square kilometres) of each MSA as reported in 1992
County and City Data Book Extra.
The values for average per capita income were that of the 1979, 1989, and 1998 U.S. Census Bureau.
The number of Fortune 500 companies located in each MSAs was obtained by reviewing the lists published in the May 05, 1980; April 23 1990; and April 17, 2000
editions of Fortune magazine.
dDummy variables were used for MSAs located in the midwestem, southern, and western regions of the United States.
The data for the percentage of black and hispanic populations were obtained by the 1992 County and City Data Book Extra, with included the entire MSA, and the 1997
U.S. Census Bureau, which only included the nucleus city.
rAll weather statistics were obtained by the U.S. National Oceanic and Atmospheric Administration (for July data) and the Climate Diagnostic Center (for snowfall).
Because not all of the individual areas in New Jersey were listed, the data of nearest cities were used as estimates.


Table 4.4 Independent Variab
e Means for 1990
1990 MEANS ALL MSA WITH NFL W/O NFL WITH MLB W/O MLB WITH NBA W/O NBA WITH NHL W/O NHL
POP3 2.23 3.31 1.29 3.85 1.33 3.16 1.42 4.23 1.62
GROWTH3 16.70% 12.40% 20.40% 11.20% 19.70% 16.10% 17.20% 4.30% 20.40%
DENSITY3 357 463 266 529 262 479 252 734 243
INCOME0 $15,919 $15,996 $15,852 $16,557 $15,564 $16,214 $15,663 $17,047 $15,578
COMPANY0 4 7 2 9 1 7 2 9 3
SOUTH3 20 7 13 4 16 8 12 1 19
MIDWEST0 11 8 3 8 3 6 5 4 7
WEST3 12 6 6 4 8 8 4 1 11
BLACK6 13.50% 14.60% 12.50% 14.90% 12.70% 13.20% 13.70% 14.50% 13.20%
HISPANIC6 9.10% 10.20% 8.20% 8.60% 9.30% 12.20% 6.40% 8.50% 9.30%
JULY' 3.31 3.34 3.28 3.09 3.43 2.99 3.58 3.38 3.28
SNOW' 21.85 24.23 19.78 23.59 20.88 21.57 22.08 37.06 17.25
ALLSPORT - 2.08 0.37 2.40 0.58 1.92 0.53 2.69 1.02
NFLSPORT 0.5 - - 1.05 0.19 0.81 0.23 0.92 0.37
MLBSPORT 0.43 0.88 0.03 - - 0.73 0.17 1.00 0.26
NBASPORT 0.41 0.77 0.23 0.80 0.31 - - 0.77 0.40
NHLSPORT 0.25 0.42 0.10 0.55 0.83 0.38 0.13 - -
DISTANCE (NFL) 212 268 164 - - - - - -
DISTANCE (MLB) 253 - - 260 250 - - - -
DISTANCE (NBA) 189 - - - - 232 151 - -
DISTANCE (NHL) 424 - - - - - - 311 458
The values for total population are in millions. The data for total population and population growth rates were obtained from the U-S. Census Bureau for the
years 1980, 1990, and 1998. Values for 1980 growth were calculated using 1970 and 1980 total population data; 1990 with 1980 and 1990, and 2000 estimates
with 1990 and 1998. Population densities were calculated. They were computed utilizing the total population values and the land area (in square kilometres) of
each MSA as reported in 1992 County and City Data Book Extra.
bThe values for average per capita income were that of the 1979, 1989,and 1998 U.S. Census Bureau.
The number of Fortune 500 companies located in each MSAs was obtained by reviewing the lists published in the May 05, 1980; April 23 1990; and April 17,
2000 editions of Fortune magazine.
dDummy variables were used for MSAs located in the midwestem, southern, and western regions of the United States.
The data for the percentage of black and hispanic populations were obtained by the 1992 County and City Data Book Extra, with included the entire MSA, and
the 1997 U.S. Census Bureau, which only included the nucleus city.
fAll weather statistics were obtained by the U.S. National Oceanic and Atmospheric Administration (for July data) and the Climate Diagnostic Center (for
snowfall). Because not all of the individual areas in New Jersey were listed, the data of nearest cities were used as estimates.


Population. First, the median population of the 56 MSAs was 1.32 million for the
1980 data and 1.47 million for 1990. Tables 4.5 and 4.6 indicate the percent of teams
in each sport located in below the median population of these 56 MSAs.
Table 4.5 -1980 Cities Below Median Population
# of MSAs hosting teams in 1980 Percent below median population
NFL 25 8%
MLB 19 0%
NBA 23 22%
NHL 13 15%
Table 4.6 -1990 Cities Below Median Population
# of MSAs hosting teams in 1990 Percent below median population
NFL 26 15%
MLB 20 5%
NBA 26 31%
NHL 13 23%
For example, in 1980, of the 25 NFL teams, only 8 percent were located in cities with
populations less than 1.32 million, and 92 percent were in larger cities. This pattern
holds true for all sports suggesting that population may be an important determinant
of team location.
As expected, all sports experienced an increase in the number of teams located
in below median metropolitan areas from 1980 to 1990 as smaller cities acquire
teams. Baseball though continued to have little representation in those markets
growing only from zero to five percent.
On the other hand, the mean population was 1.98 million for the 1980 sample
18


and 2.23 million for 1990. (New York and Los Angeles Orange County cause the
large discrepancies between the mean and median values.) When only the MSAs that
hosted teams were analyzed, the mean populations greatly increased to 3.32 million
and 3.53 million, respectively. One would expect MLB to have the highest means
due to its concentration in large markets, while the NBA is expected to have the
lowest due their greatest presence in smaller cities. In actuality, the NHL had the
highest total population means in both years. The fact that there are fewer hockey
teams, combined with their presence in the largest markets, could cause this
distortion. As expected, the NBA did indeed have the lowest mean populations. Its
presence in locations like Sacramento and Salt Lake City contribute to this. In all
leagues, though, the cities with teams average at least three times more populated than
those without.
Second, the average growth rate for all 56 MSAs was 13.7 percent in 1980,
with ten cities experiencing negative growth, and 16.6 percent in 1990, with seven
being negative. Six MSAs experienced negative growth during both periods:
Buffalo-Niagara Falls, Cleveland-Lorain-Elyria, Detroit, Pittsburgh, Newark, and
Bergen-Passaic.
Third, the mean population densities were 320 people per square kilometer in
1980 and 357 in 1990. New York and Riverside-San Bemadino held the most and
least dense populations, respectively, in each year, with the former being as much as
19


127 times more dense than the latter. Cities with teams were consistently much more
densely populated than those without.
Income. The mean per capita income for these years were $13,282 and
$15,919, respectively. In 1980, the cities with MLB teams averaged the highest and
the cities with NFL teams had the lowest. In 1990, the NHL took over the high spot
while the NFL maintained the lowest. Across the board, though, those cities with
teams had higher per capita income than those without.
Regions. Of the 56 MSAs, 23 percent (13) were located in the northeast, 20
percent (11) in the midwest, 36 percent (20) in the south, and 21 percent (12) in the
west. Table 4.7 shows the location of teams by region.
Table 4.7 Number of Teams in Each Rei gion
1980 NFL 1990 NFL 1980 MLB 1990 MLB 1980 NBA 1990 NBA 1980 NHL 1990 NHL
Total U.S. teams 28 28 24 24 23 27 14 14
Number in East 8 7 8 8 4 5 7 8
Number in Midwest 8 8 8 8 6 6 4 4
Number in South 7 7 4 4 5 8 1 1
Number in West 5 6 4 4 8 8 2 1
Because there was expansion only in the NBA, regional changes in the other sports
were due solely to relocations. In reference to the changes between regions and
between years, the NFL saw a relocation from Baltimore to Indianapolis and the Saint
Louis Cardinals moved to Phoenix. The MLB saw no activity while NBA
experienced the most with two relocations and four expansion franchises. And only
one hockey team moved: the Denver Rockies became the New Jersey Devils.
20


Weather. Not surprisingly, the NHL cities averaged the most snowfall in both
years. The NFL, though, had the rainiest city in 1980 while the NHL had the rainiest
in 1990. During the decade, the NFL teams moved into slightly drier climates; the
NBA and NHL both moved into slightly rainier, but with less snow. Again,
remember the MLB saw no changes.
Teams. An important measure of competition in this study was the number of
other teams being hosted by the MSAs. Overall, when a team was located in an area,
that area was likely to possess at least one other. For instance, when the hockey cities
are reviewed, they, on average, host a total 2.69 teams. It remains unclear whether
one sport attracted another, or if the same conditions that made a site appealing for the
first sports team were the same that made it appealing for the second or third.
Distance. The distance variable has been defined as the number of miles
between this city and the next closest city hosting a team of this same sport. The
average distances for each league have been summarized in Table 4.8.
Table 4.8 Average Distance Between Teams
1980 (miles) 1990 (miles)
NFL 211 212
MLB 253 253
NBA 249 189
NHL 422 424
Bearing in mind that basketball was the only sport to experience a great deal
of movement during this time the average distance factors for the NBA were the only
21


ones to see a drastic change in average distance.
Due to differences in data collection, no comparison can be made of
companies and race between the two years.
22


5. Discussion of Results
Three models are estimated for each sport in 1980 and 1990. The first model
allows a comparison to prior research, while the latter two models expand the first to
evaluate importance of additional factors in the placement of professional sports
teams. Although no previously published work had included a complete analysis of
both expansions and relocations for all four sports, studies by Bruggink and
Zamparelli (1999), Jozsa and Guthrie (1999), and Noll (1974) helped create the
foundation for the empirical work of Model 1. Model 2 then expands upon this first
model by including additional demographic and economic variables that could
potentially affect placement. Last, after evaluating the results of the first two models,
the selection of variables for the third model was made. The 2000 data sets were
applied to this final model for forecasting purposes.
5.1 Model 1: Past Empirical Work
Six of the independent variables discussed in the previous DATA section are
included in this model. The equation estimated in this model is identical to that of
Bruggink and Zamparelli (1999):
23


TEAMy = a+ b, POPy + ^GROWTHu + ^INCOMEy + ft,COMPANYy + A.j/lLLSPORTy (1)
+ ^DISTANCE,] + e0
where: i = city
j = {NFL, MLB, NBA, NHL}
a = a constant
bk = parameters
ey = error term
The coefficients for equation (1) were then estimated as a linear probability model
using eight data sets: the NFL, MLB, NBA, and NHL for 1980 and 1990. The results
are presented in Table 5.1.
Table 5.1 Model 1 OLS Regression Results
1980 NFL 1980 MLB 1980 NBA 1980 NHL
INTERCEPT 0.073 (.456) -0.276 (.402) -0.769 (.562) 0.020 (.435)
POP -0.010 (.060) 0.025 (.048) 0.085 (.071) 0.184** (.049)
GROWTH -0.157 (-341) -0.149 (.351) -0.033 (-491) -0.757** (.386)
INCOME -0.000 (.035) 0.024 (.031) 0.063 (.043) 0.001 (.034)
COMPANY -0.005 (.005) -0.002 (.004) 0.002 (.006) -0.006 (.005)
SPORT 0.320* (.081) 0.283** (.064) 0.058 (.085) -0.013 (.051)
DISTANCE 0.557* (.327) -0.061 (.258) 0.391 (.359) 0.033 (.139)
* Significant at the 10 percent level
** Significant at the 5 percent level
Standard errors are in parentheses
1990 NFL 1990 MLB 1990 NBA 1990 NHL
INTERCEPT 0.362 (.373) -0.134 (.329) -0.378 (.447) 0.061 (.307)
POP 0.035 (.059) 0.055 (.052) 0.066 (.067) 0.154** (.046)
GROWTH -0.353 (.333) -0.216 (.329) 0.256 (.394) -0.780** (.296)
INCOME -0.018 (.022) 0.010 (.020) 0.021 (.026) 0.004 (.019)
COMPANY -0.004 (.011) 0.009 (.010) 0.007 (.013) -0.008 (.009)
SPORT 0.203** (.084) 0.158** (.072) 0.066 (.093) -0.006 (.054)
DISTANCE 0.707** (.355) 0.083 (.284) 0.001** (.000) -0.143 (-137)
The model yield some expected, yet also some unexpected results. First, the
results for both years are quite similar. Although not always precisely estimated,
teams are typically more likely to locate in larger cities than smaller, but population
growth rates were negatively correlated with location. Therefore it appears that most
teams are located in larger, more stable cities rather than recently developed MSAs.
24


Second, the uncertainly of the role of per capita income continued with this
model. None of the coefficients were significant. Third, the NFL and MLB both
depend on the number of other teams in the community. In fact, nearly every U.S.
city that have MLB teams also hosts football: of the 20 cities hosting MLB teams in
each 1980 and 1990, 20 and 19, respectively, also had football. This may suggest that
it is the presence of another team, in addition to the measured characteristics of the
city, that promote locations of football and baseball teams.5 Last, the distance
between teams, when significant was always positive.
The next step was to test the validity of this model. To do so, the above
coefficients were applied to each MSA. This yielded the likelihood that a city that
currently did not have a team in the respective sport would obtain a future team either
through expansion or relocation. A summary of the results is in Table 5.2.
Disecting the first line of the Table 5.2, for example, shows that 33 of the 56
sample MS As cities used in equation (1) did not host a basketball team during the
1980/81 season. Minneapolis, being one of the 33, did, however, obtain a NBA team
during the decade of 1981 and 1990. The NBA factor predicted by equation (1) for
Minneapolis was .65, which of all 33 factors was closest to 1.00, determining that
Minneapolis had the greater the chance of obtaining a NBA team during this period of
time. (Note, that if any MSA received a negative factor, then the model suggested
5 When equation (1) was rerun without SPORT: POP was significant for all sports; GROWTH was significant for MLB and the
25


that if (a) the MSA already had at least one team then it had a chance of losing it or
(b) it didnt already have a team it is very unlikely to get one.)
Table 5.2 also includes a ranking of the MS As. A ranking of 1 indicates
that equation (1) predicted this city as the one most suited for a team in the league
listed. The rank, therefore, compare the factors of the actual team locations to those
Table 5.2 Expansion/Relocation Sites Between 1981 and 2000
Decade New Location League Available Cities* Model 1 Factor Rank**
1981-1990 Minneapolis NBA 33 0.65 1
1991-2000 Saint Louis NFL 30 0.71 1
1981-1990 Phoenix Mesa NFL 31 0.44 2
1991-2000 Denver MLB 36 0.47 2
1991-2000 Miami MLB 36 0.45 3
1981-1990 Indianapolis NFL 31 0.36 4
1981-1990 Miami NBA 33 0.50 4
1991-2000 Charlotte NFL 30 0.44 4
1991-2000 Phoenix Mesa MLB 36 0.40 5
1981-1990 Bergen Passaic NHL 43 0.25 7
1991-2000 Tampa Saint Petersburg MLB 36 0.27 7
1991-2000 Dallas NHL 43 0.25 7
1991-2000 Columbus NHL 43 0.20 13
1991-2000 Jacksonville NFL 30 0.23 15
1991-2000 San Jose NHL 43 0.19 15
1991-2000 Nashville NFL 30 0.21 17
1991-2000 Atlanta NHL 43 0.16 17
1981-1990 Orlando NBA 33 0.21 18
1981-1990 Sacramento NBA 33 0.17 21
1991-2000 Denver NHL 43 0.12 24
1991-2000 Nashville NHL 43 0.10 25
1981-1990 Charlotte NBA 33 0.14 26
1991-2000 Miami NHL 43 0.09 26
1991-2000 Phoenix Mesa NHL 43 0.08 28
1991-2000 Tampa Saint Petersburg NHL 43 0.07 29
1991-2000 Raleigh Durham NHL 43 -0.01 34
*A city that does not currently host a team of that sport, i.e., teairij equals 0.
Rank minus 1 is the number of available cities that received higher factors.
NHL; and INCOME was significant for MLB and the NBA.
26


of all the available cities. Of the 26 new locations listed in the table, 12 were in the
top 10 expected sites and only 1 had a negative factor.
5.2 Model 2: Expanded Model
This model included many of the additional variables discussed in the Data
section. This second equation became:
TEAMjj = a,+ AyPOPy + A2GROWTHiJ + ^DENSITY* + ^INCOMEy + ^COMPANYy (2)
+ AfSOUTHy + ^MIDWESTy + *sWESTy + 6,BLACKy + */flHISPANICy + 6JULYy
+ 6,,SNOWy + 6yjNFLSPORTy + AMLBSPORTy + 6;jNBASPORTy+ 6/(SNHLSPORT,j +
£/7DISTANCEy + eu
where: i = city
j = {NFL, MLB, NBA, NHL}
a = a constant
bk = parameters
ey : error term
As with equation (1), this equation was used as a validity tool so the 1980 and 1990
data sets were again applied. The results are included in Table 5.3.
The six characteristics belonging to both this equation and the first equation
showed minor variations. Total population and distance continued to show positive
relationships, and the dependency between the NFL and MLB became more vivid.
The additional variables added a number of significant results. The three regions
appeared negative in nearly every instance. This showed the emphasis of teams,
especially hockey, in the eastern region of the United States.6
The race effects are somewhat surprising. First, there is no significant
27


relationship between the proportion of the population that is black and team location
in any sport. Second, cities with a higher proportion of hispanics are more likely to
have an NFL team in 1980 and less likely to have a baseball team.
Table 5.3 Model 2 OLS Regression Results
1980 NFL 1980 MLB 1980 NBA 1980 NHL
INTERCEPT -0.268 (.536) 0.645 (.446) 0.469 (.776) 0.716 (-451)
POP -0.079 (.079) 0.100* (.052) 0.221 (.105) 0.270 (.049)
GROWTH 0.464 (-475) 0.106 (.384) 0.071 (.661) 0.083 (.402)
DENSITY -0.062 (.104) -0.111 (.083) -0.161 (.142) -0.208** (.083)
INCOME -0.006 (.035) -0.003 (.029) 0.017 (.050) -0.012 (.030)
COMPANY -0.006 (.005) 0.001 (.004) 0.005 (.007) -0.006 (.004)
SOUTH -0.132 (.229) -0.352** (.178) -0.501 (.306) -0.578** (.164)
MIDWEST . -0.057 (.157) 0.164 (.122) -0.074 (.219) -0.246* (.126)
WEST -0.057 (.270) -0.546** (.209) -0.610 (.370) -0.692** (.192)
BLACK 0.005 (.005) -0.001 (.004) -0.002 (.006) -0.004 (.004)
HISP 0.007* (.004) -0.009 (.003) 0.001 (.006) -0.005 (.004)
JULY 0:066 (.043) -0.068* (.036) -0.092 (.066) -0.030 (.038)
SNOW 0.002 (.003) -0.006 (.002) -0.002 (.004) -0.000 (.002)
NFLSPORT - 0.439 (.095) 0.186 (.195) 0.210* (.113)
MLBSPORT 0.766** (.165) -0.099 (-287) -0.243 (.161)
NBASPORT 0.074 (.122) 0.109 (.096) -0.196 (.093)
NHLSPORT 0.273 (.189) -0.191 (.148) -0.613 (.253)
DISTANCE 0.722** (.356) 0.115 (.241) 0.556 (.424) 0.098 (-135)
* Significant at the 10 percent level
** Significant at the 5 percent level
Standard errors in parentheses
1990 NFL 1990 MLB 1990 NBA 1990 NHL
INTERCEPT 0.210 (.568) -0.248 (.404) -0.724 (.740) 0.443 (.512)
POP 0.076 (.068) 0.082* (.047) 0.062 (.087) 0.179" (.057)
GROWTH -0.237 (.445) -0.257 (-327) -0.184 (.574) -0.255 (.411)
DENSITY -0.428 (.266) -0.624 (.192) 0.000 (.000) 0.035 (.242)
INCOME -0.018 (.027) 0.043 (.019) 0.044 (.036) -0.021 (.025)
COMPANY 0.000 (.015) 0.030 (.010) 0.012 (.020) -0.017 (.0140)
SOUTH -0.384* (.226) -0.067 (.177) 0.330 (.294) -0.281 (.225)
MIDWEST -0.138 (.187) 0.320** (.123) 0.262 (.237) -0.202 (.165)
WEST -0.142 (.249) -0.227 (.190) 0.115 (.318) -0.296 (.228)
BLACK -0.001 (.009) 0.001 (.007) -0.010 (.012) -0.007 (.009)
HISP 0.006 (.006) -0.005 (.004) 0.007 (.008) -0.003 (.006)
JULY 0.104 (.050) -0.068* (.038) -0.021 (.068) 0.037 (.050)
SNOW 0.000 (.003) -0.004 (.002) 0.000 (.004) 0.002 (.003)
NFLSPORT 0.291 (.111) 0.218 (.206) -0.139 (-142)
MLBSPORT 0.452" (.186) -0.100 (.251) 0.138 (.178)
NBASPORT 0.143 (123) -0.024 (.090) - 0.044 (.114)
NHLSPORT -0.095 (.171) 0.147 (.124) 0.039 (.220)
DISTANCE 0.670* (.398) 0.289 (.253) 0.001* (.001) -0.047 (.186)
Lastly, the two weather variables indicate that the presence of an MLB team 6
6 This, unfortunately, appears to contradict the negatively valued population density factors. The cities in the east are more
28


and precipatation are inversely related, while the NFL and both variables were
positively correlated. This latter result confirms Nolls (1974) argument that
consumers in nicer climates have more outdoor entertainment opportunities. The
NHL and NBA, on the other hand, yielded mixed results. But since these two sports
are played inside arenas, they are not as strongly dependent on weather conditions.
Table 5.4 Expansion/Relocation Sites Between 1981 and 2000
Decade New Location League Available Cities* Model 2 Factor Rank**
1991-2000 Saint Louis NFL 30 0.81 1
1991-2000 Denver MLB 36 0.51 2
1991-2000 Phoenix Mesa MLB 36 0.43 3
1981-1990 Miami NBA 33 0.58 4
1991-2000 Jacksonville NFL 30 0.30 6
1991-2000 Tampa Saint Petersburg MLB 36 0.21 6
1991-2000 Denver NHL 43 0.27 7
1991-2000 Charlotte NFL 30 0.23 8
1981-1990 Minneapolis NBA 33 0.37 11
1981-1990 Sacramento NBA 33 0.33 13
1991-2000 Dallas NHL 43 0.17 14
1981-1990 Phoenix Mesa NFL 31 0.11 16
1991-2000 Nashville NFL 30 0.12 17
1991-2000 Atlanta NHL 43 0.14 17
1981-1990 Indianapolis NFL 31 0.09 18
1991-2000 Tampa Saint Petersburg NHL 43 0.11 19
1981-1990 Charlotte NBA 33 0.08 22
1991-2000 Miami MLB 36 0.02 23
1991-2000 Nashville NHL 43 0.04 24
1981-1990 Bergen Passaic NHL 43 -0.01 29
1981-1990 Orlando NBA 33 -0.13 32
1991-2000 Phoenix Mesa NHL 43 -0.04 32
1991-2000 Columbus NHL 43 -0.05 33
1991-2000 Miami NHL 43 -0.07 35
1991-2000 Raleigh Durham NHL 43 -0.10 39
1991-2000 San Jose NHL 43 -0.16 41
*A city that does not currently host a team of that sport, i.e., teauii equals 0.
**Rank minus 1 is the number of available cities that received higher factors.
dense than most.
29


In order to test the validity of this model, the selected locations during these decades
were again evaluated. Table 5.4 summarizes the results. This time, 8 of the 26 actual
locations were listed as one of the top 10 expected sites and 7 had negative factors.
Therefore, although variables in equation (2) showed greater overall significance, this
model did not predict future locations as well as equation (1).
5.3 Model 3: Predicting Future Team Locations
After reviewing the results of the two previous models, it became apparent
that many of the variables do not play as significant a role in the placement of
professional sports teams as originally thought. Or at least this cannot be shown
empirically. Therefore the third model simply became a slight variation of the first
one.
TEAMjj = a + b, POPy + />,GROWTH;j + ft jNCOME, + ft^COMPANYy + *JNFLSPORTjj (3)
+ ^MLBSPORTy + ^NBASPORTy + A,NHLSPORTy + ^DISTANCEy + etj
where: i = city
j = {NFL, MLB, NBA, NHL}
a = a constant
bk = parameters
ey = error term
This equation is simply equation (1) with the ALLSPORT variable listed separately
as NFLSPORT, MLBSPORT, NBASPORT, and NHLSPORT.7 Table 5.5 reports the
results.
In this model, total population was clearly significant and distance remained postively
30


correlated. (Note that football and total population were inversely related which
contradicts all research assumptions. Though this effect could be swayed by the
strong relationship between MLB and NFL teams.)7 8
As seen in much of the two previous models: population growth continued to be
negative, income was positive for all but football, and the Fortune 500 companies
were a mixed bag. The strong relationship between the NFL and MLB teams
prevailed, while the NBA and NHL teams were negatively related due to their near
identical seasons.
Table 5.5 Model 3 OLS Regression Results
2000 NFL 2000 MLB 2000 NBA 2000 NHL
INTERCEPT 0.567 -0.120 -0.347 -0.383
(.371) (.314) (.445) (.422)
POP -0.120** 0.073* . 0.098* 0.124**
(-052) (.038) (.059) (.056)
GROWTH -0.438 -0.328 -0.006 -0.220
(.362) (.309) (.433) (.413)
INCOME -0.008 0.001 0.010 0.012
(.012) (-010) (.014) (.013)
COMPANY 0.000 0.001 0.003 -0.010
(.011) (.009) (.013) (.013)
NFLSPORT - 0.360** 0.099 0.182
(.091) (.140) (.139)
MLBSPORT 0.636** - 0.076 -0.014
(.156) (.199) (.195)
NBASPORT -0.011 0.099 - -0.073
(.121) (.103) (.143)
NHLSPORT 0.184 0.039 -0.069 -
(.117) (.100) (.137)
DISTANCE 0.521 0.581* 0.965** 0.540
(.351) (321) (470) (.340)
* Significant at the 10 percent level
** Significant at the 5 percent level
Standard errors in parentheses
Using the results from equation (3), the cities most likely to become the next
7 For validation purposes, the 1980 and 1990 data were applied to this model. As expected, they yielded near identical results.
8 When equation (3) was rerun without the SPORT variables, POP was significant for MLB, NBA, and NHL and GROWTH
was significant for the NFL and MLB.
31


locations for professional sports teams are listed in Table 5.6. This table lists the top
five predicted sites for expansion franchises or relocating teams. The total number of
available cities for each league were: 28 for the NFL, 32 for MLB, 30 for the NBA,
and 34 for the NHL.
By definition, a dependent variable with a value of 1 referred to cities that host
one or more teams. Therefore, the factor of 1.07 for Los Angeles Orange County
simply states that this city is suited to receive at least one future football team.
Interestingly, too, is that the top two NFL picks are cities that previously hosted teams
and were the two in competition for the 32nd NFL expansion franchise.
Table 5.6 Team Placement (2001/02 2010/11 seasons)
Only MSAs with no team during the 2000/01 season included
Top Five Predicted MSA for Each Professional Sport____________
NFL Los Angeles Orange County (1.07) NBA Kansas City (.73)
Houston (.54) Saint Louis (.57)
Salt Lake City (.42) San Diego (.52)
Columbus (.38) Memphis (.50)
Rochester (.37) New Orleans (.48)
MLB New Orleans (.55) NHL Seattle Bellevue Everett (.81)
Charlotte (.54) San Francisco Oakland (.76)
Buffalo Niagara Falls (.50) New Orleans (.53)
Indianapolis (.50) San Diego (.47)
Nashville (.47) Portland Vancouver (.44)
The number in parentheses illustrates the likelihood of obtaining a team. The closer the value is to 1
the greater the chance.
Another city of interest is New Orleans. Although it already hosts a football
team, it is deemed a good candidate for all other leagues by appearing as one of the
top five candidates for the MLB, NBA, and NHL.
32


6. Conclusion
In an effort to identify which economic and demographic characteristics are
important in the placement of professional sports teams, I evaluated 18 independent
variables within 3 OLS models for each sport. The first two models estimated the
effect of various factors important in team placement during the 1980s and 1990s.
The predicted values were then compared to actual team placements. The third model
included current data and was used to predict future placement of teams.
While this study is not only the first of its kind to validate such predictions, it
is the only to include all four professional sports (football, baseball, basketball, and
hockey) and the first to examine both expansion franchises and relocating teams.
Of the 18 city characteristics, I concluded that most do not play as important a
role in the placement of teams. In fact, only total population, the distance between
teams in the league, and the number of teams from other leagues are important
determinants. This provides some insight into the team placement process. That is,
the decision for all team locations is subject to league approval. Because financial
success and competitive performances motivate team owners, the league looks to
place teams in markets that will maximize these objectives. These markets will,
therefore, emphasize total population, distance between teams, and other sports teams.
33


Although many of the city characteristics I evaluated did not yield significant
results, other factors could play an important role in team placement. For example, a
communitys willingness to supply public subsidies for facility construction and
player/team performance could greatly affect team revenue. Likewise, the structure
of the sport season could impact revenue. Baseball, for example, has over 80 regular
season home games each year while football only has 8. Therefore, applying different
models to different sports may prove valuable.
34


Appendix
Table A1 Locations of 1980/81 Teams
Metropolitan Statistical Areas* NFL Team?** MLB Team?** NBA Team?** NHL Team?**
ATLANTA Yes Yes Yes
AUSTIN SAN MARCOS
BERGEN-PASSAIC Yes
BOSTON Yes Yes Yes Yes
BUFFALO-NIAGARA FALLS Yes Yes
CHARLOTTE
CHICAGO Yes Yes Yes Yes
CINCINNATI Yes Yes
CLEVELAND-LORAIN-ELYRIA Yes Yes Yes
COLUMBUS
DALLAS-FORT WORTH Yes Yes Yes
DENVER Yes Yes Yes
DETROIT Yes Yes Yes Yes
FORT LAUDERDALE
GRAND RAPIDS-MUSKEGON-HOLLAND
GREENSBORO-WINSTON-SALEM
HARTFORD Yes
HOUSTON Yes Yes Yes
INDIANAPOLIS Yes
JACKSONVILLE
KANSAS CITY Yes Yes Yes
LAS VEGAS
LOS ANGELES-ORANGE COUNTY Yes Yes Yes Yes
MEMPHIS
MIAMI Yes
MIDDLESEX-SOMERSET-HUNTERDON
MILWAUKEE-WAUKESHA Yes Yes Yes
MINNEAPOLIS Yes Yes Yes
MONMOUTH-OCEAN CITY
NASHVILLE
NASSAU-SUFFOLK
NEW ORLEANS Yes
NEW YORK Yes Yes Yes Yes
NEWARK
OKLAHOMA CITY
35


ORLANDO
PHILADELPHIA Yes Yes Yes Yes
PHOENIX-MESA Yes
PITTSBURGH Yes Yes Yes
PORTLAND-VANCOUVER Yes
PROVIDENCE
RALEIGH-DURHAM-CHAPEL HILL
RIVERSIDE-SAN BERNADINO
ROCHESTER
SACRAMENTO
SAINT LOUIS Yes Yes Yes
SALT LAKE CITY Yes
SAN ANTONIO Yes
SAN DIEGO Yes Yes Yes
SAN FRANCISCO-OAKLAND Yes Yes Yes
SAN JOSE
SEATTLE-BELLEVUE-EVERETT Yes Yes Yes
SOUTHERN VIRGINIA
TAMPA-SAINT PETERSBURG Yes
WASHINGTON-BALTIMORE Yes Yes Yes Yes
WEST PALM BEACH-BOCA RATON
*AII 56 MSAs used in this study are listed in this matrix.
**Left blank if MSA hosts no team.
36


Table A2 Locations of 1990/91 Teams
Metropolitan Statistical Areas* NFL Team?** MLB Team?** NBA Team?** NHL Team?**
ATLANTA Yes Yes Yes
AUSTIN SAN MARCOS
BERGEN-PASSAIC Yes Yes
BOSTON Yes Yes Yes Yes
BUFFALO-NIAGARA FALLS Yes Yes
CHARLOTTE Yes
CHICAGO Yes Yes Yes Yes
CINCINNATI Yes Yes
CLEVELAND-LORAIN-ELYRIA Yes Yes Yes
COLUMBUS
DALLAS-FORT WORTH Yes Yes Yes
DENVER Yes Yes
DETROIT Yes Yes Yes Yes
FORT LAUDERDALE
GRAND RAPIDS-MUSKEGON-HOLLAND
GREENSBORO-WINSTON-SALEM
HARTFORD Yes
HOUSTON Yes Yes Yes
INDIANAPOLIS Yes Yes
JACKSONVILLE
KANSAS CITY Yes Yes
LAS VEGAS
LOS ANGELES-ORANGE COUNTY Yes Yes Yes Yes
MEMPHIS
MIAMI Yes Yes
MIDDLESEX-SOMERSET-HUNTERDON
MILWAUKEE-WAUKESHA Yes Yes Yes
MINNEAPOLIS Yes Yes Yes Yes
MONMOUTH-OCEAN CITY
NASHVILLE
NASSAU-SUFFOLK
NEW ORLEANS Yes
NEW YORK Yes Yes Yes Yes
NEWARK
OKLAHOMA CITY
ORLANDO Yes
37


PHILADELPHIA Yes Yes Yes Yes
PHOENIX-MESA Yes Yes
PITTSBURGH Yes Yes Yes
PORTLAND-VANCOUVER Yes
PROVIDENCE
RALEIGH-DURHAM-CHAPEL HILL
RIVERSIDE-SAN BERNADINO
ROCHESTER
SACRAMENTO Yes
SAINT LOUIS Yes Yes
SALT LAKE CITY Yes
SAN ANTONIO Yes
SAN DIEGO Yes Yes
SAN FRANCISCO-OAKLAND Yes Yes Yes
SAN JOSE
SEATTLE-BELLEVUE-EVERETT Yes Yes Yes
SOUTHERN VIRGINIA
TAMPA-SAINT PETERSBURG Yes
WASHINGTON-BALTIMORE Yes Yes Yes Yes
WEST PALM BEACH-BOCA RATON
*AII 56 MSAs used in this study are listed in this matrix.
**Left blank if MSA hosts no team.
38


Table A3 Locations of 2000/01 Teams
Metropolitan Statistical Areas* NFL Team?** MLB Team?** NBA Team?** NHL Team?**
ATLANTA Yes Yes Yes Yes
AUSTIN SAN MARCOS
BERGEN-PASSAIC Yes Yes
BOSTON Yes Yes Yes Yes
B U FFALO-NI AG ARA FALLS Yes Yes
CHARLOTTE Yes Yes
CHICAGO Yes Yes Yes Yes
CINCINNATI Yes Yes
CLEVELAND-LORAIN-ELYRIA Yes Yes Yes
COLUMBUS Yes
DALLAS-FORT WORTH Yes Yes Yes Yes
DENVER Yes Yes Yes Yes
DETROIT Yes Yes Yes Yes
FORT LAUDERDALE
GRAND RAPIDS-MUSKEGON-HOLLAND
GREENSBORO-WINSTON-SALEM
HARTFORD
HOUSTON Yes Yes
INDIANAPOLIS Yes Yes
JACKSONVILLE Yes
KANSAS CITY Yes Yes
LAS VEGAS
LOS ANGELES-ORANGE COUNTY Yes Yes Yes
MEMPHIS
MIAMI Yes Yes Yes Yes
MIDDLESEX-SOMERSET-HUNTERDON
MILWAUKEE-WAUKESHA Yes Yes Yes
MINNEAPOLIS Yes Yes Yes Yes
MONMOUTH-OCEAN CITY
NASHVILLE Yes Yes
NASSAU-SUFFOLK
NEW ORLEANS Yes
NEW YORK Yes Yes Yes Yes
NEWARK
OKLAHOMA CITY
ORLANDO Yes
PHILADELPHIA Yes Yes Yes Yes
PHOENIX-MESA Yes Yes Yes Yes
PITTSBURGH Yes Yes Yes
PORTLAND-VANCOUVER Yes
39


PROVIDENCE
RALEIGH-DURHAM-CHAPEL HILL Yes
RIVERSIDE-SAN BERNADINO
ROCHESTER
SACRAMENTO Yes
SAINT LOUIS Yes Yes Yes
SALT LAKE CITY Yes
SAN ANTONIO Yes
SAN DIEGO Yes Yes
SAN FRANCISCO-OAKLAND Yes Yes Yes
SAN JOSE Yes
SEATTLE-BELLEVUE-EVERETT Yes Yes Yes
SOUTHERN VIRGINIA
TAMPA-SAINT PETERSBURG Yes Yes Yes
WASHINGTON-BALTIMORE Yes Yes Yes Yes
WEST PALM BEACH-BOCA RATON
*AII 56 MSAs used in this study are listed in this matrix.
40


Bibliography
Baade, R. A. (1987). Is there an economic rationale for subsidizing sports
stadiums? [On-line], www.heartland.org/studies/ports/baadel.htm
Baade, R.A. & Sanderson, A.R. (1997). The employment effect of teams and
sports facilities. In R.G. Noll & A. Zimbalist (Eds.), Sports, Jobs, and Taxes: The
Economic Impact of Sports and Stadiums (pp. 92 118). Washington, DC:
Brookings Institution Press.
Bruggink, T.H. & Zamparelli, J.M. (1999). Emerging markets in baseball:
An econometric model for predicting the expansion teams new cities. In J. Fizel, E.
Gustafson, & L. Hadley (Eds.), Sports Economics: Current Research (pp. 49 59).
Westport, CT: Praeger Publishing.
Irani, D. (1997). Public subsidies to stadiums: Do the costs outweigh the
benefits? Public Finance Review, 25(2), 238-253.
Jozsa, F.P., Jr. & Guthrie, J.J., Jr. (1999). Relocating teams and expanding
leagues in professional sports: How the major leagues respond to market conditions.
Westport, CT: Quorum Books.
Lapchick, R.E. (1999). Sport in society. [On-line].
www.sportinsociety.org/rgrc98.html
McManus, J. (Ed.). (2000, April 23). The 500 largest industrial corporations.
Fortune. 346-364.
Noll, R.G. (Ed.). (1974). Government and the sports business. Washington,
DC: The Brookings Institution.
Noll, R.G. & Zimbalist, A. (1997). The economic impact of sports teams and
facilities. In R.G. Noll & A. Zimbalist (Eds.), Sports, Jobs, and Taxes: The
Economic Impact of Sports Teams and Stadiums (pp. 55 91). Washington, DC:
Brookings Institution Press.
41


Pearlstine, N. (Ed.)- (1980, May 05). The 500 largest industrial corporations.
Fortune. 276 294.
Pearlstine, N. (Ed.). (2000, April 17). The 500 largest industrial corporations.
Fortune. FI -F19.
Sheehan, R. G. (1996). Keeping score: The economics of big-time sports.
South Bend, IN: Diamond Communications.
42