Content uploaded by Bram Lancee
Author content
All content in this area was uploaded by Bram Lancee on Aug 24, 2015
Content may be subject to copyright.
This article was downloaded by: [Wissenschaftszentrum Berlin]
On: 15 January 2014, At: 07:27
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Sociological Spectrum: Mid-South
Sociological Association
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/usls20
Income Inequality and Gambling: A Panel
Study in the United States (1980–1997)
Thijs Bol a , Bram Lancee b & Sander Steijn a
a University of Amsterdam, Amsterdam Center of Inequality Studies ,
Amsterdam , The Netherlands
b Utrecht University , Utrecht , The Netherlands
Published online: 13 Jan 2014.
To cite this article: Thijs Bol , Bram Lancee & Sander Steijn (2014) Income Inequality and Gambling:
A Panel Study in the United States (1980–1997), Sociological Spectrum: Mid-South Sociological
Association, 34:1, 61-75, DOI: 10.1080/02732173.2014.857196
To link to this article: http://dx.doi.org/10.1080/02732173.2014.857196
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
Income Inequality and Gambling: A Panel Study in the
United States (1980–1997)
Thijs Bol
University of Amsterdam, Amsterdam Center of Inequality Studies,
Amsterdam, The Netherlands
Bram Lancee
Utrecht University, Utrecht, The Netherlands
Sander Steijn
University of Amsterdam, Amsterdam Center of Inequality Studies,
Amsterdam, The Netherlands
While there are many studies that examine the consequences of increasing income inequality, its effects
on gambling behavior have not yet been studied. In this article, we argue that income inequality
increases the average expenditure on gambling. Using longitudinal state-level data for the United States
(1980–1997), we estimate fixed-effects models to analyze two types of gambling expenditure:
pari-mutuel betting and lottery spending. Our findings show a positive effect of increasing income
inequality on lottery expenditure. For pari-mutuel betting, the result is not linear, as for higher levels
of income inequality, the positive effect decreases, suggesting that the effect flattens out when the
increase in income inequality is highest. We argue that there are three reasons why we find a positive
effect of income inequality of gambling expenditure: increasing mobility aspirations, availability of
resources in the upper part of the distribution, and status anxiety in the lower part of the distribution.
The Lottery, with its weekly pay-out of enormous prizes, was the one public event to which the proles
paid serious attention. It was probable that there were some millions of proles for whom the Lottery
was the principal if not the only reason for remaining alive. It was their delight, their folly, their
anodyne, their intellectual stimulant. Where the Lottery was concerned, even people who could barely
read and write seemed capable of intricate calculations and staggering feats of memory.
George Orwell, 1984
INTRODUCTION
Research on the impact of inequality is rapidly expanding. Rising income inequality has been
linked to a wide range of undesirable societal outcomes, including behavior that can be seen
Address correspondence to Sander Steijn, Amsterdam Institute for Social Science Research, Kloveniersburgwal 48,
1012 CX Amsterdam, The Netherlands. E-mail: s.r.steijn@uva.nl
Sociological Spectrum, 34: 61–75, 2014
Copyright #Taylor & Francis Group, LLC
ISSN: 0273-2173 print/1521-0707 online
DOI: 10.1080/02732173.2014.857196
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
as risk taking, such as, for example, smoking or alcohol usage (e.g., Wilkinson and Pickett
2009). Surprisingly, one of the most widespread risk-taking behaviors, gambling, has not been
studied in relation with increasing income inequality. In this article, we argue that rising income
inequality serves as an explanation for increased risk-taking behavior and, more specifically, for
gambling expenditure.
The majority of research on gambling behavior in the social sciences has predominantly
focused on problematic (and pathological) gambling (e.g., Johansson et al. 2009; McNeilly
and Burke 2002; Volberg and Wray 2007). Many researchers seek explanations for gambling
in the psyche by, for example, studying to what extent people gamble out of excitement (e.g.,
Anderson and Brown 1984) and to what extent their decisions in gambling are cognitive (Sharpe
and Tarrier 1993; Rogers 1998).
However, gambling behavior is also a more general social phenomenon that has been largely
neglected as such in social scientific research (see also Volberg and Wray 2007). While previous
work analyzes which social groups are more likely to gamble (e.g., McEffery 1994; Sproston,
Erens, and Orford 2000; Beckert and Lutter 2009, 2012), contextual explanations for these
differences are often not considered. Overall, gambling behavior is scarcely studied in relation
to macro conditions (but see Freund and Morris [2005, 2006], who argue that income inequality
is increased by gambling due to income concentration). A more sociological perspective
on gambling behavior that seeks for its societal explanations can help to better understand
how gambling comes about.
We argue that income inequality is likely to increase gambling expenditure through three
mechanisms. Firstly, higher income inequality makes upward social mobility for individuals
in the lower part of the income distribution more desirable. Consequently, as a strategy to get
ahead in society, people gamble more. Secondly, in more unequal societies, the upper part of
the distribution has more resources to gamble. Put differently, under conditions of high
inequality, resources are more concentrated in the upper part of the distribution. As a result,
people in the top of the distribution can spend more and will do so in the form of gambling
because it confirms their status as ‘winners’ in society. Thirdly, inequality leads to more anxiety
and stress amongst the poor (Wilkinson and Pickett 2009), which increases gambling behavior
as relief in the lower part of the distribution. Based on these three arguments, we hypothesize
that income inequality increases gambling expenditure.
We test this hypothesis empirically by analyzing longitudinal state-level data containing both
income inequality and gambling expenditure in the United States for 36 states in the period of
1980–1997. We relate income inequality at t-1 in each state to the per capita expenditure for two
types of gambling: lottery playing and pari-mutuel betting.
1
We contribute to the existing litera-
ture by proposing and testing a broader, more sociological account of gambling, which adds to
the rapidly expanding literature on the consequences of income inequality (e.g,. Andersen and
Fetner 2008; Huisman and Oldehinkel 2009; Uslaner and Brown 2005; Wilkinson and Pickett
2009). The empirical contribution of this study is the use of longitudinal data to examine
consequences of inequality. Studies that examine the effect of inequality often rely on cross-
sectional data, making reversed causality and bias due to unobserved heterogeneity more likely.
1
Pari-mutuel betting is a betting system that places all bets in a prize pool without a priori providing odds to the con-
testants. This prize pool is eventually shared between all correct bets, proportional to their contribution to the prize pool.
A notable example of this system is the betting that occurs on greyhound racing.
62 T. BOL ET AL.
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
The longitudinal models that we estimate support the hypothesis that increasing income
inequality is positively related to gambling expenditure. This finding is in line with the more
general claim, advocated by Wilson and Pickett (2009), that income inequality enhances risk-
taking behavior.
THEORY AND HYPOTHESIS
Gambling as Risk-Taking Behavior
Despite the stereotypical stigma that is often placed on gambling, it is a widespread phenom-
enon. For example, in the United States, 82%of adults report to gamble at least once a year
(Welte et al. 2002). Several studies show that the vast majority of these gamblers are not
categorized as problematic: the prevalence rate for excessive gambling (both problematic and
pathological) is around 3%(for an overview see Stucki and Rihs-Middel 2007). The high par-
ticipation rates in gambling activities make it an interesting research object for sociologists.
Moreover, gambling is a clear example of a fundamental interest of sociologists: risk taking
(Frey and Eadington 1984:121; Cosgrave 2008). The concept of risk has often been studied
(Beck 1992; Giddens 1991), and scholars increasingly point to risk taking as desirable behavior.
Lyng (1990), for example, describes the increase in interest in voluntary risk-taking behavior
with the concept of edgework: behavior that involves a ‘‘threat to one’s physical or mental
well-being or one’s sense of an ordered existence’’ (Lyng 1990:857). The goal of this risk-taking
behavior is the excitement of taking the risk. The stereotypical example of edgework is extreme
sports, where failure results in death. Gambling, however, is also behavior where an in essence
unnecessary risk is taken: to wager large amounts of money is voluntary risk taking as a form of
leisure.
Gambling has been associated with risk-taking behavior before. Clotfelter and Cook (1989)
argue that gambling can be seen as a good proxy for risk-taking behavior. Brown, and colleagues
(2010) study the United Kingdom and connect gambling expenditure to financial risk taking
such as loans, hire purchase agreements, and credit club payments. There are, however, also
good reasons to see gambling not so much as risk-taking behavior. Buying a lottery ticket,
for example, is a relatively low investment, thereby having a relatively low risk (e.g., McEffery
1994). Still, we argue that this does not make gambling risk-free behavior: not gambling is
always more risk-averse than gambling, and the risk increases with the amount (proportional
of income) that is wagered.
2
In this article, we distinguish two types of risk-taking behavior: risk taking as a form of
leisure and risk taking as a means to social mobility. Layton and Worthington (1999) find that
individuals with higher incomes spend more on gambling, thereby supporting the claim that
there is indeed more gambling at the top of the income distribution. Gambling by people in
the top of the distribution is therefore more likely to be risk-taking behavior as leisure and as
a status symbol: because high-income individuals are already in the top of the distribution, they
cannot rise much more.
2
In our empirical study we focus on the expenditure on gambling, which validates gambling as a risk-taking process.
The higher the amount that is wagered (net of the wealth of a state), the higher the risk.
INCOME INEQUALITY AND GAMBLING 63
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
At the bottom part of the income distribution, gambling can be motivated both by relief or
escapism (risk taking as a form of leisure), as well as the prospect of upward social mobility (risk
taking as a means to social mobility). People gamble as a way of potentially improving their
societal position, for example by winning the lottery (Rogers 1998). However, in doing so they
must be willing to accept different levels of risks. For example, according to Young (2010), the
risk society produces a corollary, a demand for risk in the form of gambling. To get ahead in
society individuals increasingly are taking risks.
Previous work using individual level data finds that lower-class individuals indeed have
a higher probability to engage in gambling (Brown et al. 2010:23, Lang and Omori 2009;
Beckert and Lutter 2012). Layton and Worthington (1999) confirm these findings by showing
that in Australia blue-collar workers gamble more than white-collar workers. Besides the
frequency of gambling, there is also evidence that gambling expenditure is unevenly distributed
across the income distribution, low incomes spending relatively more than high incomes. In
a study of lottery play in the United States, Clottfelter and colleagues (1999) show not only that
participation in the lottery is frequent under lower income groups (48.5%of the people earning
less than $10,000 per year and 46.7%of the people earning between $10,000 and $25,000
a year), but also that spending on the lottery is relatively high among those groups (respectively
$597 and $569 per year). This means that the lower income groups outspend the higher income
groups both in relative and in absolute terms.
However, such high gambling expenditure by the lower income population comes at a cost.
Kearney (2005) shows that after the introduction of a state lottery, lower income people signifi-
cantly cut down expenditure on food (2.8%) and on home mortgage, rent, and other bills (5.8%)
to compensate for their gambling expenditure. Lang and Omori (2009) show that the least
wealthy are more likely to lose a higher proportion of their income than wealthier households.
While there is previous work on who gambles, and how much different groups spent on gam-
bling, we know relatively little about how the shape of the income distribution affects gambling.
Income Inequality and Gambling
Research that analyzes the consequences of inequality is growing steadily. Outcomes such as
crime, bad health, low social trust, and low levels of civic participation are seen as being depen-
dent on the income inequality in a society (Uslaner and Brown 2005; Kawachi et al. 1997; Daly,
Wilson, and Vasdev 2001; Huisman and Oldehinkel 2009; Kelly 2000;Wilkinson and Pickett
2009; Rothstein and Uslaner 2005; Solt 2008; Lancee and Van de Werfhorst 2012). While these
studies connect income inequality to risk-taking behavior, Neckerman and Torche (2007) argue
that individuals are more likely to be risk averse in contexts with a high level of income
inequality, as the cost of failure is higher. However, no empirical evidence exists to support
the statement. We discuss three mechanisms by which income inequality is likely to increase
gambling expenditure, and with that, implicitly, increases risk-taking behavior.
Firstly, rising income inequality makes upward social mobility more beneficial. Wilkinson
and Pickett state that ‘‘the scale of income differences has a powerful effect on how we relate
to each other’’ (2009:5). They argue that, with higher inequality, there are more status differ-
ences between individuals which results in larger status gaps. These gaps trigger status compe-
tition. According to Wilson and Daly (1985), especially men take risks in order to achieve social
64 T. BOL ET AL.
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
reputation or status. Already in 1951, Bloch (1951:215) argued that the ‘‘chance element in
human life is particularly exploited in those societies where status is largely competitive.’’ Stu-
dies indeed show that individuals are more likely to gamble in contexts where they perceive their
own income as low (Haisley et al. 2008). Along this line of reasoning, as growing inequality
fuels aspirations for upward mobility, inequality is likely to trigger gambling behavior. Using
data from the United States, Xu and Garand (2010) show that especially low-income individuals
residing in states with high income inequality are more likely to perceive large increases in
income inequality.
Mobility aspirations based on material motivations especially hold for people in the bottom
half of the income distribution. As Clotfelter and Cook (1989) argue, gambling is a way out for
those that cannot afford otherwise. They state that the ‘‘opportunity to purchase a chance to win
500 dollars for one dollar seems attractive to those that do not have other ways of obtaining that
much spending money’’ (Clotfelter and Cook 1989:71). On the basis of a Swedish study of
soccer betting, Frey and Eadington (1984:114) argue that ‘‘the propensity to gamble is not
necessarily determined by social class characteristics, but rather by the extent to which upward
mobility aspirations are held.’’ Under conditions of high inequality, the distance to the top is
larger, hence the potential gain of upward mobility is higher.
3
We therefore expect that since
rising inequality increases the potential gains of gambling, it will also increase the incidence
of gambling as a ways of achieving upward mobility, specifically in the lower and middle part
of the distribution. As a consequence, we expect gambling expenditure to rise.
A second mechanism that explains how rising income inequality increases gambling behavior
is the changing composition of the income distribution. In more unequal societies, the top of the
distribution by definition has more resources to gamble and thus expenditure is likely to be
higher. Indeed, gambling expenditure is found to increase with income (Welte et al. 2002).
The reason that people gamble more is that for individuals in the middle and upper part of
the income distribution, gambling is a leisure activity and serves as a status symbol confirming
and emphasizing people’s position on the societal ladder. Put differently, as a result of status
competition, people in the upper part of the distribution claim their position as ‘winners’ in
society by engaging in gambling activities. In more unequal societies, we therefore expect that
people in the upper part of the distribution gamble more because of leisure or as a status symbol.
The third argument that stipulates a positive relationship between income inequality and
gambling expenditure is increasing envy and stress in the lower part of the distribution (Eitle
and Taylor 2011; Wiggins et al. 2010). The ‘‘frustration’’ argument was first introduced by
Devereux, who argued that gambling is a ‘‘safety valve through which the repressed wishes
crowd for escape’’ (1980 [1949]:781). As Devereux argues, particularly people in the bottom
part of the income distribution tend to be frustrated. For them, higher inequality results in
increasing levels of frustration, since those at the bottom are lagging behind and simultaneously
see that another part of the population grows ever richer (Wilkinson & Pickett 2009). Already in
1984, Frey and Eadington (1984:110) conclude: ‘‘In sum, gambling is a mechanism of accom-
modation utilized by society to channel protestations resulting from frustration with the basic
economic and ethical systems in a socially approved- that is, functionally imperative-manner.’’
3
Note that we do not argue that gambling is the only strategy to achieve upward mobility. There are many ways
through which this can be achieved. We argue that upward mobility aspirations is one of the reasons that people gamble,
even under lower levels of income inequality.
INCOME INEQUALITY AND GAMBLING 65
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
More recently, Beckert and Lutter provided micro-level evidence for the relation between
frustration and gambling behavior. Using micro data from Germany they find that status
frustration is one of the main factors contributing to an individual’s gambling (lottery) expenditure
(2012:13).
In a situation of economic stratification the lower part of the income distribution
feels degraded and desires the resources they lack even more (Blau and Blau 1982). Uslaner
and Brown (2005; Rothstein and Uslaner 2005), examine the relation between inequality and
people’s civic engagement. They argue that under conditions of high inequality, people will
perceive that their views are not represented in the political system. As a consequence, people
may feel powerless and they will opt out of civic engagement. These feelings of frustration may
also manifest itself in increasing gambling behavior as a form of escapism (Cohen 2001).
The three mechanisms that we identify are not expected to explain the expenditure of lottery
playing and pari-mutuel betting equally (King 1985). The argument of mobility aspiration is
likely to apply mainly for lottery playing, as a small wager can give high returns. With only
a small investment, people have the chance to take a large step on the income ladder. In contrast,
the status and leisure mechanism is likely to apply mostly to pari-mutuel betting, because it is
a more visible form of gambling. The anxiety and frustration argument can apply to expenditure
on both types of gambling. When individuals feel detached from society, they can engage in all
forms of risk-taking behavior.
Unfortunately, empirically, we are not able to distinguish between the three mechanisms
because we do not have data on gambling expenditure in different parts of the distribution.
However, all three mechanisms predict a positive effect of income inequality on gambling
expenditure. Our general hypothesis to be tested is therefore that with an increase in income
inequality the average gambling expenditure increases as well.
DATA AND MEASUREMENT
We test our hypothesis using data from the Gambling Impact and Behavior Study (GIBS) carried
out in the years 1980–1997 by the National Opinion Research Center at the University of
Chicago, for the National Gambling Impact Study Commission (see NORC, 1999). In this
survey, information is gathered on per capita gambling expenditure for four types of gambling:
bingo, lottery, pari-mutuel betting, and casino. In the GIBS survey, data for per capita lottery
expenditure are gathered at the state level. Pari-mutuel expenditure per capita is, however,
gathered at the level of census designated places (CDPs). A place is defined by the U.S. Census
Bureau as ‘‘densely settled concentrations of population that are identifiable by name, but are
not legally incorporated places.’’ Out of a population of 32,000 CDPs 105 were selected by
using a simple stratified sampling method. The focus in our study is not on the level of CDPs,
but on the level of states. For pari-mutuel spending we therefore calculate the state-level per
capita spending as an average of all CPDs that are surveyed in that state.
We do not analyze the data of casino expenditure as casino gambling is highly conditioned
on the availability of casinos nearby. Unsurprisingly, a casino industry survey shows that in 1992
(about the midpoint of the GIBS survey) Nevada and Atlantic City, where casinos are legal,
account for the large chunk of casino visits. Out of about 50 million total visits to the casino,
more than 40 million took place in those two states. As legal restrictions to casinos influence
66 T. BOL ET AL.
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
the availability of casinos, it hence does not make sense to explain gambling behavior on the
basis of income inequality in states. We also exclude bingo expenditure because the majority of
observations in the GIBS survey are extrapolated instead of observed.
Measures
Data for per capita pari-mutuel expenditure were harmonized on the basis of various receipt
data sources. Per capita lottery expenditure is estimated on the basis of lottery sales and prize
data for the time period 1980 until 1992 and based on state lottery agency data for the remaining
period of 1992 to 1997.
4
We measure income inequality with the Gini-coefficient. The Gini-coefficient has a theore-
tical range from zero, indicating that all households have an equal share of income, to 100,
indicating that one household receives all income. The Gini-coefficient is an attractive measure
of income inequality, because it calculates overall inequality: it captures the income distances of
‘everybody to everyone.’ We use data from the University of Texas Inequality Project (2011).
5
The state level Gini-coefficient is available for all states and years. While our hypothesis
predicts a positive relation between income inequality and gambling, we do not presume
that this effect is linear. To account for possible non-linearity we therefore add a squared term.
We control for general wealth in each state by including yearly measures of per capita Gross
State Product (GSP). For 1977 to 1990, we use the GSP measures by Beemiller and Dunbar
(1993). For the remaining years, we rely on Beemiller and Woodruff (2003). To obtain the
GSP per capita, we divide it by the state population size (derived from the U.S. Census Bureau).
Descriptive statistics on all variables are displayed in Table 1. Furthermore, in models that
we do not show we controlled our results for the introduction of state lotteries (cf. Lutter
2011). The inclusion of this variable did not alter our results.
Method of Estimation
We estimate fixed effects regression models. The fixed effects (FE) model estimates an intercept
for each state. The FE model uses each variable’s difference from its within-state mean and hence
can estimate only coefficients that have within-state variation (Rabe-Hesketh and Skrondal
2008). The advantage of the FE model is that it controls for all differences between states,
thereby eliminating time-constant unobserved heterogeneity (Halaby 2004). Due to the FE
set-up, we focus on over-time variation within states and therefore do not need covariates to con-
trol for enduring differences between states. To better account for potential reversed causality,
we include our measure of inequality lagged one year.
6
Finally, we add year dummies, to control
for a general time trend.
4
We included a dummy variable to account for potential differences that arise out of the different coding of lottery
expenditure (a ‘‘post 1992’’ dummy). The inclusion of this dummy (tables not shown, available from corresponding
author upon request) did not alter our results. For more information on the measurement of the dependent variables
see the codebook of the Gambling Impact and Behaviour Study (NORC, 1999:13-15).
5
See Galbraith and Hale (2008) for the composition of this measure.
6
We also estimated models without and with longer time-lags. The results are consistent with the ones reported in the
next section.
INCOME INEQUALITY AND GAMBLING 67
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
RESULTS
Figure 1 provides an overview of the temporal development of pari-mutuel and lotto expenditure
(standardized to the range 0–1) along with the development of income inequality by state. Both
forms of gambling show an increase in expenditure in almost all states, although there are
differences. In Maine, for example, all two types of gambling increase at the same pace, while
in Connecticut the differences between lottery and pari-mutuel betting are large. Overall, income
TABLE 1
Descriptive Statistics
Observations Mean SD Min. Max.
Per capita lottery expenditure 334 43.47 28.49 2.72 145.42
Gini coefficient (t-1) 334 0.39 0.02 0.33 0.46
Gini squared (t-1) 334 0.15 0.02 0.11 0.21
GSP per capita (1,000 $) 334 21.25 5.69 8.91 37.37
Per capita pari-mutuel expenditure 514 12.18 11.17 0.03 49.46
Gini (t-1) 514 0.40 0.02 0.33 0.47
Gini squared (t-1) 514 0.16 0.02 0.11 0.22
GSP per capita (1,000 $) 514 20.25 6.20 8.70 40.24
Source: GIBS Survey 1980–1997, University of Texas Inequality Project (2011).
FIGURE 1 Over-time development of inequality and gambling expenditure, by U.S. states (color figure available
online).
68 T. BOL ET AL.
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
inequality, as well as gambling expenditure per capita, has risen between 1977 and 1996 in all
states in our sample. Hence, this descriptive picture, at first inspection, supports our hypothesis.
Using fixed effects models, where years are nested in states, we estimate the effect of income
inequality on gambling expenditure. Table 2 shows the results when predicting per capita
pari-mutuel expenditure. Model 1 only includes income inequality (at t-1), which is not statisti-
cally significant. If we, however, account for non-linearity, by adding a squared term, we find a
significant positive effect of income inequality, and a significant negative effect of squared
income inequality. Substantively, this implies that higher income inequality leads to more spend-
ing on pari-mutuel betting, but this positive effect is flattened down for higher values of income
inequality. The coefficients remain statistically significant once we control for GSP per capita
(which increased in almost all states in each year). Furthermore, even in the most restrictive
model, where we control for the general over-time increase in pari-mutuel betting by adding year
dummies, the effects remain significant. In the final model the explained variance is low,
although this might be caused because the measure is aggregated from lower level units to states.
Figure 2 graphically displays these findings. In this figure, for each state the predicted values
of pari-mutuel expenditure, based on the final model, are plotted against levels of income
inequality. For most cases, high values of income inequality are associated with high pari-mutuel
spending. However, for the highest values of income inequality, the positive effect flattens down
and even becomes negative for a few cases.
There are two reasons that might explain why the effect of income inequality is indeed
non-linear. First, there is only a certain amount of gambling that people in the upper part of
the distribution can do. In highly unequal societies the gambling expenditure is therefore
unlikely to rise even more. Furthermore, one could argue that for the wealthiest, other forms
of behavior can provide the status symbolism and leisure that is generated by gambling.
Investing in the stock market could take over the role of gambling (both in terms of status sym-
bolism as well as leisure). Secondly, when a society is very unequal, the bottom income group
TABLE 2
Fixed-Effects Regression Predicting Per Capita Pari-mutuel Expenditure in U.S. States (1980–1997)
Model 1 Model 2 Model 3 Model 4
b SE b SE b SE b SE
Gini-coefficient 5.156 (12.724) 775.917!!! (224.513) 790.496!! (233.175) 620.676!(271.666)
Gini-coefficient
squared
"975.176!!! (283.609) "987.6562!! (288.835) "799.291!(326.572)
GSP per capita
(1,000 $)
".012 (.051) ".057 (.123)
Year dummies yes
Constant 10.202!(4.999) "141.520!! (44.401) "145.072 (46.955) "106.279 (56.075)
nObservations 514 514 514 514
nStates 33 33 33 33
Within R
2
.000 .024 .025 .044
Overall R
2
.003 .007 .009 .036
Source: GIBS Survey 1980–1997, University of Texas Inequality Project (2011).
!p<.05; !!p<.01; !!!p<.001, two-tailed tests.
INCOME INEQUALITY AND GAMBLING 69
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
might feel so marginalized that even their hopes of upward social mobility disappear. The
disillusion of the gap might cause a cut-off point where inequality no longer affects gambling
expenditure.
Table 3 presents the estimates for per capita lottery expenditure. In the case of lottery
expenditure, we found no evidence for non-linearity in the effect of income inequality. The
first model shows a positive effect of income inequality (at t-1), indicating that in states where
inequality increased, expenditure on lottery also increased. This effect remains significant once
FIGURE 2 Predicted values of per capita pari-mutuel expenditure for each U.S. state (color figure available online).
TABLE 3
Fixed-Effects Regression Predicting Per Capita Lottery Expenditure in U.S. States (1980–1997)
Model 1 Model 2 Model 3
b SE b SE b SE
Gini-coefficient 1834.053!!! (78.235) 434.528!!! (110.257) 427.441!!! (143.355)
GSP per capita (1,000 $) 3.095!!! (.205) 5.086 (.592)
Year dummies yes
Constant "659.670!!! (30.056) "259.465!!! (57.614) "246.598 (55.243)
nObservations 334 334 334
nStates 26 26 26
Within R
2
.642 .794 .825
Overall R
2
.118 .438 .490
Source: GIBS Survey 1980–1997, University of Texas Inequality Project (2011).
!p<.05; !!p<.01; !!!p<.001, two-tailed tests.
70 T. BOL ET AL.
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
we control for wealth by adding GSP per capita. Furthermore, the effect persists even when
we control for a general time trend and increase of gambling, by adding year dummies. The
explained variance of these models is relatively high. The final model explains about 83%of
the observed variance in lottery spending within states. In Figure 3, the predicted values are
presented for each state. The figure clearly shows that, both within and between states, higher
income inequality corresponds with higher per capita lottery expenditure.
In sum, our findings support the theoretical prediction that income inequality positively
affects gambling expenditure, although this effect is not linear for expenditure on pari-mutuel
betting. Due to our fixed effects design, we control for all time-constant differences between
states, eliminating the problem of bias due to (time-constant) unobserved heterogeneity. Further-
more we predict gambling expenditure with income inequality in the previous year, making
reversed causality less likely.
CONCLUSION
In this study, we analyzed the effect of growing inequality on gambling expenditure in the
United States with longitudinal data. We argued that inequality affects gambling behavior for
three reasons. First, growing inequality is likely to increase spending on gambling activities
because an increase in income inequality makes upward social mobility for the lower part of
the income distribution more desirable. Gambling will therefore be used more often as a means
to get ahead in society. Secondly, in more unequal societies the upper part of the distribution has
more resources to gamble and will do so because of leisure and as a status symbol. Thirdly,
inequality leads to more anxiety and stress, which in turn increases gambling as a relief in
the lower part of the distribution.
FIGURE 3 Predicted values of per capita lottery expenditure for each U.S. state (color figure available online).
INCOME INEQUALITY AND GAMBLING 71
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
We found empirical support for this hypothesis. Analyzing two different dependent vari-
ables, our findings indicate that increasing income inequality is positively related to gambling
expenditure in U.S. states. For pari-mutuel betting, the effect is, however, not linear. The
effect of inequality most likely flattens because of a natural maximum to gambling expenditure,
as well as the unavailability of resources in the lowest part of the income distribution.
The results have to be seen in the light of some serious limitations. First of all, no (longitudi-
nal) individual level data was available that could be matched to contextual information. It was
therefore not possible to empirically examine the effect of inequality for people in different parts
of the income distribution. Based on the current findings, we therefore know that inequality
increases gambling expenditure, but we do not know who are the people that spend more. This
implies that we could not empirically differentiate the theoretical arguments that we identified.
Second, because casino gambling is highly regionally concentrated in the United States, it was
not possible to include it. Further research could focus on replicating the current findings in different
and cross-national contexts, and for different forms of gambling, like casino-based gambling.
Finally, since our sample ranged from 1980 to 1997, internet gambling was of no influence to
our results. As this recent technological development has made gambling more easily accessible
for a large group, further research has to take it into account. The rise of internet gambling could
of course have large effects on the relation between income inequality and gambling. It is highly
accessible, and the potential revenues of internet gambling are relatively high. We are therefore
likely to underestimate the relationship between income inequality and gambling expenditure:
whereas income inequality increased over the past decade, we are likely to underestimate the
overall gambling expenditure due to the absence of data on internet gambling.
Although we focus on gambling behavior, the results fit within the broader research agenda on
the consequences of (increasing) income inequality. Specifically, gambling behavior can be seen
as a proxy for being risk prone. To our knowledge, this is one of the first studies that examines
a relation between income inequality and legal risk-taking behavior in a longitudinal perspective
(but see Freund and Morris 2005, 2006). The findings suggest that differences in incomes
might affect the degree of risk that people are willing to take. As such, the results may have
implications for research on a number of societal outcomes that are related to risk taking as well
as social mobility, such as (but not exclusively) entrepreneurship, investing, and crime.
ACKNOWLEDGMENTS
All authors contributed equally and are listed alphabetically. The authors would like to thank
the members of the International Political Economy and Stratification (IPECS) and Institutions,
Inequalities and Lifecourses (IIL) groups at the University of Amsterdam, and the reviewers of
Sociological Spectrum for their excellent comments.
AUTHOR NOTES
Thijs Bol is Assistant Professor at the University of Amsterdam, where he also obtained his PhD.
His academic interests include inequalities in education and the labor market, social stratifi-
cation, and occupations. He has published articles in Comparative Education Review,Social
Science Research, and Research in Social Stratification and Mobility.
72 T. BOL ET AL.
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
Bram Lancee is Assistant Professor of Sociology at Utrecht University. He obtained his PhD
at the European University Institute in Florence. He held (visiting) positions at Social Science
Research Center Berlin (WZB), Oxford University, and the University of Amsterdam. His
academic interests include social capital and social participation, ethnic inequality on the labor
market, and attitudes toward immigration. Parts of his work are published in International
Migration Review, Ethnic & Racial Studies, Journal of Ethnic and Migration Studies, Political
Behavior, European Sociological Review, and Social Science Research.
Sander Steijn is a PhD candidate at the sociology department of the University of Amsterdam.
As a political sociologist, he is primarily interested in (public opinion on) the consequences of
income inequality and redistribution of incomes by government.
REFERENCES
Andersen, Robert and Tina Fetner. 2008. ‘‘Economic Inequality and Intolerance: Attitudes Toward Homosexuality in 35
Democracies.’’ American Journal of Political Science 52:942–958.
Anderson, G. and R. I. F. Brown. 1984. ‘‘Real and Laboratory Gambling, Sensation-Seeking and Arousal.’’ British
Journal of Psychology 75:401–410.
Beck, Ulrich. 1992. The Risk Society: Towards A New Modernity. London: Sage.
Beckert, Jens and Mark Lutter. 2009. ‘‘The Inequality of Fair Play: Lottery Gambling and Social Stratification in
Germany.’’ European Sociological Review 25:475–488.
Beckert, Jens and Mark Lutter. 2012. ‘‘Why the Poor Play the Lottery: Sociological Approaches to Explaining
Class-based Lottery Play.’’ Sociology doi:10.1177/0038038512457854.
Beemiller, Richard M. and Clifford H. Woodruff. 2000. ‘‘Gross State Product by Industry, 1977–98.’’ Survey of Current
Business October:69–90.
Beemiller, Richard M. and A. E. Dunbar. 1993. ‘‘Gross State Product, 1977–90.’’ Survey of Current Business
December:28–49.
Blau, Judith R. and Peter Blau. 1982. ‘‘The Cost of Inequality: Metropolitan Structure and Violent Crime.’’ American
Sociological Review 47:114–129.
Bloch, Herbert A. 1951. ‘‘The Sociology of Gambling.’’ American Journal of Sociology 57:215–221.
Brown, Sarah, Andy Dickerson, Jolian McHardy, and Karl Taylor. 2010. ‘‘Gambling and the Use of Credit: An
Individual and Household Level Analysis.’’ IZA DP No. 4808.
Clotfelter, Charles T., Philip J. Cook, Julie A. Edell, and Marian Moore. 1999. ‘‘State Lotteries at the Turn of the
Century. Report to the National Gambling Impact Study Commission.’’ Working paper, Duke University.
Clotfelter, Charles and Philip J. Cook. 1989. Selling Hope: State Lotteries in America. Cambridge: Harvard University
Press.
Cohen, Lloyd R. 2001. ‘‘The Lure of the Lottery.’’ Wake Forest Law Review 36:705–747.
Cosgrave, Jim F. 2008. ‘‘Goffman Revisited: Action and Character in the Era of Legalized Gambling.’’ International
Journal of Criminology and Sociological Theory 1:80–96.
Daly, Martin, Margo Wilson, and Shawn Vasdev. 2001. ‘‘Income Inequality and Homicide Rates in Canada and the
United States.’’ Canadian Journal of Criminology April:219–236.
Devereux, Edward C. Jr. 1980 [1949]. Gambling and the Social Structure. A Sociological Study of Lotteries and Horse
Racing in Contemporary America. New York: Arno Press.
Eitle, David and John Taylor. 2011. ‘‘General Strain Theory, BIS=BAS Levels, and Gambling Behavior.’’ Deviant
Behavior 32:1–37.
Freund, Elizabeth A. and Irwin L. Morris. 2005. ‘‘The Lottery and Income Inequality in the States.’’ Social Science
Quarterly 86:996–1012.
Freund, Elizabeth A. and Irwin L. Morris. 2006. ‘‘Income Inequality and Gambling in the United States.’’ Policy Studies
Journal 34:265–276.
Frey, James H. and William R. Eadington. 1984. ‘‘Gambling: Views from the Social Sciences.’’ Annals of the American
Academy of Political and Social Science 474:107–121.
INCOME INEQUALITY AND GAMBLING 73
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
Giddens, Anthony. 1991. Modernity and Self-Identity. Palo Alto, CA: Stanford University Press.
Galbraith, James K. and J. Travis Hale. 2008. ‘‘State Income Inequality and Presidential Election Turnout and
Outcomes.’’ Social Science Quarterly 89:887–901.
Halaby, Charles N. 2004. ‘‘Panel Models in Sociological Research: Theory into Practice.’’ Annual Review of Sociology
30:507.
Haisley, Emily, Romel Mostafa, and George Loewenstein. 2008. ‘‘Subjective Relative Income and Lottery Ticket
Purchases.’’ Journal of Behavorial Decision Making 21:283–295.
Huisman, M. and A. J. Oldehinkel. 2009. ‘‘Income Inequality, Social Capital and Self-Inflicted Injury and
Violence-Related Mortality.’’ Journal of Epidemiology and Community Health 6:31–37.
Johansson, A., J. E. Grant, S. W. Kim, B. Odlaug, and K. G. Go¨testam. 2008. ‘‘Risk Factors for Problematic Gambling:
A Critical Literature Review.’’ Journal of Gambling Studies 25(1):67–92.
Kawachi, Ichiro, Bruce P. Kennedy, Kimberly Lochner, and Deborah Prothrow-Stith. 1997. ‘‘Social Capital, Income
Inequality, and Mortality.’’ American Journal of Public Health 87:1491–1498.
Kearney, Melissa S. 2005. ‘‘State Lotteries and Consumer Behavior.’’ Journal of Public Economics 89:2269–2299.
Kelly, Morgan. 2000. ‘‘Inequality and Crime.’’ Review of Economics and Statistics 82:530–539.
King, Kim. 1985. ‘‘Gambling - Three Forms and Three Explanations.’’ Sociological Focus 18:235–248.
Lancee, Bram and Herman G. van de Werfhorst. 2012. ‘‘Income Inequality and Participation: A Comparison of 24
European Countries.’’ Social Science Research 41:1166–1178.
Lang, K. Brandon and Megumi Omori. 2009. ‘‘Can Demographic Variables Predict Lottery and Pari-mutuel Losses?
An Empirical Investigation.’’ Journal of Gambling Studies 25:171–183.
Layton, Allan and Andrew Worthington. 1999. ‘‘The Impact of Socio-Economic Factors on Gambling Expenditure.’’
International Journal of Social Economics 26:430–440.
Lutter, Mark. 2011. ‘‘The Adoption of Lotteries in the United States, 1964–2007: A Model of Conditional and
Time-Dynamical Diffusion.’’ MPIfG Discussion Paper 11=4. Cologne: Max Planck Institute for the Study of
Societies.
Lyng, Stephen. 1990. ‘‘Edgework: A Social Psychological Analysis of Voluntary Risk Taking.’’ American Journal of
Sociology 95:851–888.
McEffery, Edward J. 1994. ‘‘Why People Play Lotteries and Why It Matters.’’ Wisconsin Law Review 1994:71.
McNeilly, Dennis and William Burke. 2002. ‘‘Disposable Time and Disposable Income: Problem Casino Gambling
Behavior in Older Adults.’’ Journal of Clinical Geropsychology 8:75–85.
Neckerman, Kathryn M. and Florencia Torche. 2007. ‘‘Inequality: Causes and Consequences.’’ Annual Review of
Sociology 33:335–357.
NORC. 1999. Codebook for the Gambling Impact and Behavior Study. Ann Arbor: University of Michigan.
Rabe-Hesketh, Sophia and Anders Skrondal. 2008. Mulilevel and Longitudinal Modeling Using Stata. 2nd ed. College
Station, TX: Stata Press.
Rothstein, Bo and Eric M. Uslaner. 2005. ‘‘All for All: Equality, Corruption, and Social Trust.’’ World Politics 58:
41–72.
Rogers, Paul. 1998. ‘‘The Cognitive Psychology of Lottery Gambling: A Theoretical Review.’’ Journal of Gambling
Studies 14:111–134.
Sharpe, Louise and Nicholas Tarrier. 1993. ‘‘Towards a Cognitive-Behavioural Theory of Problem Gambling.’’ The
British Journal of Psychiatry 162:407–412.
Solt, Frederick. 2008. ‘‘Economic Inequality and Democratic Political Engagement.’’ American Journal of Political
Science 52:48–60.
Sproston, Kerry, Bob Erens, and Jim Orford. 2000. Gambling Behaviour in Britain: Results from the British Gambling
Prevalence Survey. London: National Center for Social Research.
Stucki, Stephanie and Margret Rihs-Middel. 2007. ‘‘Prevalence of Adult Problem and Pathological Gambling Between
2000 and 2005: An Update.’’ Journal of Gambling Studies 23:245–257.
University of Texas Inequality Project. 2011. ‘‘A Panel of Estimated Family Income Inequality, by State and Year,
1969–2004.’’ Retrieved June 15, 2011 from http://utip.gov.utexas.edu/data.html.
Uslaner, Eric M. and Mitchell Brown. 2005. ‘‘Inequality, Trust, and Civic Engagement.’’ American Politics Research
33:868–894.
Volberg, Rachel A. and Matt Wray. 2007. ‘‘Legal Gambling and Problem Gambling as Mechanisms of Social
Domination? Some Considerations for Future Research.’’ American Behavioral Scientist 51:56–85.
74 T. BOL ET AL.
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
Welte, John W., Grace M. Barnes, William F. Wieczorak, Marie-Cecile Tidwell, and John Parker. 2002. ‘‘Gambling
Participation in the U.S. - Results from a National Survey.’’ Journal of Gambling Studies 18:313–327.
Wiggins, Lyna, Lia Nower, Raymond Sanchez Mayers, and N. Andrew Peterson. 2010. ‘‘A Geospatial Statistical
Analysis of the Density of Lottery Outlets within Ethnically Concentrated Neighborhoods.’’ Journal of Community
Psychology 38:486–496.
Wilkinson, Richard G. and Kate Pickett. 2009. The Spirit Level. Why More Equal Societies Almost Always Do Better.
London: Allen Lane.
Wilson, Margo and Martin Daly. 1985. ‘‘Competitiveness, Risk-Taking, and Violence: The Young Male Syndrome.’’
Ethology and Sociobiology 6:59–73.
Xu, Ping and James C. Garand. 2010. ‘‘Economic Context and Americans’ Perceptions of Income Inequality.’’ Social
Science Quarterly 91:1220–1241.
Young, Martin. 2010. ‘‘Gambling, Capitalism and the State.’’ Journal of Consumer Culture 10:254–273.
INCOME INEQUALITY AND GAMBLING 75
Downloaded by [Wissenschaftszentrum Berlin] at 07:27 15 January 2014
