RÉSUMÉ
RÉALISÉ ÉLECTRONIQUEMENT
(pour un aperçu
de la recherche). - Comme l'étude n'a pas été encore été
officiellement publié, il m'est impossible de la rendre disponible
dans sa version complète. A.D.
Roughly 250 U.S. counties have
legalized casino gambling within their borders. Sixty of these
counties have established commercial casino operations, with the
remainder supporting tribal casinos. Past research has provided
mixed results regarding the impact of these casinos on market and
non-market outcomes. The goal of this research study is to estimate
the impact of casinos on two of these outcome variables --
individual and business bankruptcy rates -- over the decade of the
1990s. The study matches each casino county with a non-casino county
according to U.S. Census region, household income, population and
population density. Using simple descriptive statistics and
regression analysis, the study estimates the impacts of casinos on
bankruptcy rates. Our regression analysis on matched-pair counties
indicates that those counties that legalized casino gambling during
the 1990s experienced a cumulative growth rate in individual
bankruptcies that was more than double the growth rate for
corresponding non-casino counties. However, the cumulative rate of
change in business bankruptcy rates in the casino counties was, on
average, 35.4 percent lower than the applicable rate for the
non-casino counties.
NOTE: All estimates, opinions, and views expressed in this paper are
those of the authors alone and not those of the Congressional Budget
Office.
Twenty-five years ago, legalized gambling was confined to Nevada,
Atlantic City, New Jersey, a few racetracks, and two or three state
lotteries.
Since then, the U.S. has added almost 400 commercial casinos and 248
tribal casinos to the gambling landscape.
Casino revenues have grown along with the number of casinos.
In recent years, commercial (non-tribal) casinos have increased
adjusted gross revenues (AGR) from $19.7 billion in 1999 to $26.5
billion in 2002, or 10.0 percent per year.
Per thousand dollars of GDP, commercial casino AGR grew from $2.13
in 1999 to $2.63 in 2002.
Tribal casinos have also experienced significant revenue growth.
In 2001, tribal casinos in 28 states pulled in an estimated $12.7
billion in AGR, reflecting a growth rate of approximately 14 percent
per year from the $7.5 billion in AGR reported for 1997.
Figure 1 and Figure 2 profile commercial and tribal casino AGR.
This expansion has impacted the social costs of gambling, including
bankruptcy.
Much of the research examining the social costs has focused on the
problem or pathological gambler.
All seem to agree that pathological gamblers "engage in destructive
behaviors: they commit crimes, they run up large debts, they damage
relationships with family and friends, and they kill themselves.
With the increased availability of gambling and new gambling
technologies, pathological gambling has the potential to become even
more widespread."
During the rapid expansion of casino gambling during the 1990s,
personal bankruptcies expanded at comparably high rates.
On the other hand, during this same period of time, business
bankruptcies declined from 63,365 in 1990 to 37,183 in 1999.
Despite these seemingly contradictory relationships, many
politicians, sociologists and economists fault casinos for a large
share of the growth in U.S. bankruptcies.
However, other economic and demographic factors were also changing
during this period, making the assignment of cause for rising
bankruptcies impossible to isolate without a more in-depth analysis.
In the subsequent analysis, we use multivariate regression to
disentangle contributors to higher bankruptcy rates, specifically
focusing on the casinos.
Other researchers have also undertaken this same task.
For example, Barron, Staten and Wilshusen (2000) (hereafter referred
to as BSW) conclude that casinos had positive and statistically
significant impacts on personal bankruptcy rates in the casino
county and its geographic neighbors.
However, these researchers concluded that the increase in personal
bankruptcies attributable to casinos was only 8 percent, and that
other demographic and economic factors were much more important in
explaining the rapid growth in personal bankruptcies in the 1990s.
In the subsequent analysis, we expand on the BSW study by adding two
factors not considered by them.
First, we examine business bankruptcies; second, we add tribal
casinos, which were excluded from their analysis, to our assessment.
Our data set also excludes some data that the BSW study included.
The BSW study included counties adjacent to those hosting casinos --
what they term "collar counties" -- in their analysis, based on the
assumption that a higher incidence of pathological gambling behavior
was expected within a 50 mile radius of a casino facility.
However, our study focuses only on the casino counties.
Bankruptcy filings in collar counties may well include residents who
live more than 50 miles from a casino, who thus are not particularly
influenced by casino activity.
Our more limited focus may be viewed as providing a more
conservative measure of the bankruptcy impact of casinos, as it
reduces the possibility that those with more attenuated geographical
proximity to the casino operations may erroneously be attributed to
casino-related causation.
Large-scale Indian casino gambling is barely a decade old.
Its origins trace back to 1987, when the U. S. Supreme Court issued
its decision in California v. Cabazon Band of Mission Indians.
The Court held that the state of California had no authority to
apply its regulatory statutes to gambling activities conducted on
Indian reservations.
Tribal sovereignty was subordinate to the Federal government, and
state power to regulate was thus dependent on congressional
authorization.
which essentially recognized the right of Indian tribes to regulate
gambling and gaming facilities on their reservations as long as the
states in which they were located had some form of legalized
gambling.
Figure 3 shows tribal casinos by state.
Oklahoma had the largest number of tribal casinos at 50 followed by
California at 44 and Washington at 23.
After a brief respite, America returned to casinos in 2002.
Following more than a decade of explosive growth, the tragic events
of September 11th reduced air travel to spots such as Las Vegas.
But casinos responded with increased marketing to locals, and the
U.S. gaming industry (both casino and non-casino) posted a five
percent increase in revenues to an estimated $64 billion for 2002.
While all forms of gambling have grown, casino gambling has
experienced robust growth in recent years.
Since Nevada legalized casino gaming in 1931, an additional ten
states have legalized commercial casinos.
New Jersey legalized casino gaming in 1976, and its first casino
opened in 1978.
However, eight of the eleven states with commercial casinos began
casino construction in the 1990s, thus introducing new features into
their local economic and social structures.
According to Christiansen of Capital Advisors LLC, Americans today
pay out more on gambling than they spend on movie tickets, theme
parks, spectator sports, and video games combined.
Moreover, Merrill Lynch estimated that Americans lose a comparable
amount each year in illegal betting.
Figure 4 shows the number of commercial casinos by state.
As indicated, eleven states had a total of 432 commercial casinos in
2004.
Nevada had the most casinos at 249, while Michigan had the fewest at
3.
Table 1 compares commercial and tribal casinos.
Commercial casinos generate, on average, more revenue than tribal
casinos.
Furthermore, the effective tax rate is much higher for commercial
casinos and the growth rate of AGR has been much lower for
commercial casinos.
A study by the National Opinion Research Center at the University of
Chicago found that pathological gamblers generate 15 percent of the
industry's gross revenues and that each pathological gambler costs
society around $10,550 over his/her lifetime.
In its 1999 report, The National Gambling Impact Study Commission
singled out convenience gambling as providing fewer economic
benefits and greater social costs than other forms of gambling.
In order to evaluate the significance of bankruptcy data considered
in this study, a basic overview of federal bankruptcy laws will
prove helpful.
Federal bankruptcy laws serve two important purposes: providing a "fresh
start" for debtors by granting relief from burdensome financial
obligations, and providing a means for creditors to obtain payment
to the extent possible.
Generally speaking, Chapter 7 of the Bankruptcy Code provides for a
liquidation process, while Chapters 11, 12, and 13 provide
procedures for reorganization and rehabilitation of debtors.
A debtor commences bankruptcy by filing a petition that constitutes
an order for relief under the applicable chapter of the Bankruptcy
Code for which the debtor is eligible.
Individuals who are employees, as well as individuals who are sole
proprietors of businesses, are eligible.
Thus, a portion of Chapter 7 filings may reflect adverse financial
experiences with business activities, as well as financial
difficulties rooted in gambling activity.
Confirmation of this plan provides relief for the debtor by changing
the nature and extent of the debtor's financial obligations, which
may result in some creditors not being paid, or receiving lower
payments at a later time than reflected in their original bargain.
Other more detailed chapters also exist in the Bankruptcy Code to
address special types of debtors.
Chapter 9 provides special rules for municipalities.
Chapter 12 provides special procedures for family farmers with
regular income.
Given the limited applicability of Chapter 12 and the extremely
limited scope of Chapter 9, those bankruptcy filings are not
considered in this study.
Federal district courts have original and exclusive jurisdiction
over bankruptcy cases.
so in this sense the bankruptcy court is a unit of the federal
district court.
Bankruptcy petitions are subject to venue rules that affect the
proper geographical location for filing.
in which there is pending a case under Title 11 [i.e., the
bankruptcy title] concerning such person's affiliate, general
partner, or partnership."....
For purposes of this statute, "domicile" and "residence" may be in
different venues.
"Domicile" and "residence" are not used as synonyms in § 1408.
The term "domicile" is defined generally as residence in fact along
with the intent to remain there or to return when absent.
Once established, a domicile continues until a new one is acquired.
Although any United States citizen residing in the United States
always has a domicile in some state, a person may only have one
domicile at a time.
In contrast, a person may have several residences at the same time.
A person can change residences at will but a domicile, once
established, remains until a new domicile is established.
Actual residence is not necessary to preserve a domicile once a
domicile has been acquired.
Residence, when used in a sense other than domicile, is one of the
most nebulous terms in the legal dictionary and can have many
different meanings depending on the context in which it is used.
Residence is less inclusive than domicile, importing merely having
an abode at a particular place which may be one of any number of
such places at which one is, at least from time to time, physically
present.
The venue rules make it possible for a debtor to have a residence in
one jurisdiction, but to file a bankruptcy petition in another
jurisdiction that the debtor considers his domicile.
Alternatively, a debtor might file in still another venue in which
the debtor has a "principal place of business."
Employment in a particular location is not a sufficient basis for
venue under the "principal place of business" category.
However, an entrepreneur who owns a business, even one that is
presently inactive, may file in the location of that business, even
if it differs from his residence or domicile.
These venue rules show that filing may occur in a location that
differs from one's residence.
Although one might raise that fact as a basis for discounting the
validity of any correlation between casinos and bankruptcy filings,
the fact remains that the venue for filing will nevertheless be the
same as the residence and domicile in the vast majority of cases.
Given the stringent requirements for domicile and the inability to
use a place of employment to file in lieu of residence, the typical
employee debtor is likely to file in the same geographical area in
which he lives.
Even if the debtor is an entrepreneur with a business located in a
venue that might differ from his home, the business is nevertheless
subject to the effects of casinos, and the individual may also be
personally affected.
Thus, the foundation for correlation shown in the data is
substantial, though particular cases may well have exceptional facts
that limit the actual scope of any effect from casinos in that
jurisdiction.
The bankruptcy data used in this study was obtained from the
Administrative Office of the United States Bankruptcy Courts.
Report F-5A contains data by county for filings by business and
non-business debtors under each bankruptcy chapter.
Court administrators use data in this report to evaluate where
demand is greatest for bankruptcy court services.
The county-by-county presentation of this data is also useful for
the purpose of evaluating whether any correlation exists between the
availability of casino gambling in that county and the number of
bankruptcy petitions filed in that county.
Table 2 compares the bankruptcy experiences of counties with
commercial casinos, tribal casinos, and no casinos.
When aggregate bankruptcy data are considered, the tribal casino
counties experienced the highest growth rate; commercial casino
counties had a lower aggregate growth rate than both non-casino and
tribal casino counties.
However, it should be noted that this aggregate data includes
counties with preexisting casino operations.
Thus, it is hazardous to generalize about the discrete effects of
casino gambling on these figures alone.
The rest of Table 2 focuses on median bankruptcy data over this same
period.
The median individual bankruptcies per thousand of population data
show similar, though not identical, changes to the aggregate data
over the period.
However, median firm bankruptcies per thousand of population
decreased over this same period for all three categories.
Table 3 also compares counties with commercial casinos, tribal
casinos, and no casinos based on characteristics other than
bankruptcy rates.
In general, counties with commercial casinos had a larger
population, higher density of population, lower poverty rates,
higher bankruptcy rates, and higher family income than either
counties with no casino or counties with a tribal casino.
As presented, tribal casinos were located in counties that began the
decade in more financial distress in terms of higher poverty rates,
higher unemployment rates and slightly lower average family income.
Table 4 compares statistical results for counties with and without
casinos.
In this case, each casino county is matched with a non-casino county
according to population, population density, census region and
family income.
As presented, there were no dramatic differences in these factors
between casino counties and their non-casino counterparts.
In order to investigate differences in a multi-variate framework, we
next apply regression Equations (1) and (2) to the matched data.
Equations (1) and (2) estimate individual bankruptcies and business
bankruptcies, respectively, against factors hypothesized to affect
financial distress.
The two dependent variables, (IndBnk and (BusBnk, represent the
change from 1990 to 1999 in the individual and business bankruptcy
rates.
A description of each variable is contained in Table 5.
As presented, the addition of casinos in a county during the 1990s
had a negative and statistically significant impact on business
bankruptcy rates and a positive and statistically significant impact
on individual bankruptcy rates.
Using parameter estimates from Table 6 provides estimates of the
impact of individual and business bankruptcy rates.
Adding a casino in the 1990s increased the county personal
bankruptcy rate by 100 percent, but reduced business bankruptcy
rates by 35.4 percent on average.
Table 7 presents the estimated impact of factors, including the
addition of a casino on three types of bankruptcy filings.
As shown, the addition of a casino in the county had a positive
impact on Chapter 7 and 13 bankruptcy filings.
On the other hand, the addition of a casino in the county had a
negative impact on Chapter 11 filings.
However, there are some constraints associated with this methodology
that should be noted.
First, we examined data in two discrete years: 1990 and 1999.
Although this methodology may accurately measure the impact of the
addition of a casino operation during the interim years, it is
possible that such an introduction could have had discrete effects
in a given year that were not significant in 1999.
Second, we have omitted some potential variables that could affect
bankruptcy.
For example, it is theoretically possible that casino counties had
populations with higher debt loads than non-casino counties.
Our study did not include debt load data.
However, we find no evidence that such a disparity did, in fact,
exist.
Third, adding a casino could also be endogenous, to the extent that
a county that legalized casino gambling may have already experienced
increased bankruptcy rates before the casino was opened.
The differences in results between these chapters of the Bankruptcy
Code may be explainable by reference to the effects of casino
gambling in the particular locale.
Though individual employees may have an impact on their operational
effectiveness, the ability to diffuse that impact over a number of
employees would tend to prevent the compulsive behavior of one or a
few employees from impacting the financial wellbeing of the entire
enterprise.
Moreover, in many jurisdictions the business community surrounding a
casino may well experience beneficial economic impacts from the
inputs and services that casino operations require.
Summary
Results from applying regression
analysis to U.S. bankruptcy data for 1990 and 1999 indicate that
counties that legalized casinos during the period suffered
individual bankruptcy rates more than 100 percent higher than
casinos that remained “casinoless.” On the other hand, the casino
counties experienced business bankruptcy rates that were 35.4
percent less than their matching counties without casinos. Casino
counties were much more likely to experience Chapter 7 and Chapter
13 bankruptcies, but less likely to experience Chapter 11
bankruptcies
GRAPHIQUES
ET FIGURES DE LA RECHERCHE ICI
Résumé par Copernic Summarizer
