“Cultures of Inequality”: Ethnicity, Immigration, Social Welfare, and Imprisonment

The authors discuss the shift from classic culture of poverty arguments to more contemporary uses of cultural variables in explaining criminal justice practices in Western industrialized countries. The authors use “cultures of inequality” to refer to the increasing taste or tolerance for inequality in the general population across nations. They also elaborate a potential link between perceived threat of others and growing tastes for inequality, thereby extending the classic threat hypothesis. Using country-level data and data from the World Values Survey, the authors find that countries with higher than average tastes for inequality also have higher income inequality, more population heterogeneity, and higher percentages of others in prison. However, people in these countries do not necessarily have more hostile attitudes toward others . The United States shares several characteristics with other Western countries but appears to be driving the difference in the mean taste for inequality between countries with low and high imprisonment of others .

"CULTURES OF INEQUALITY" 135 on welfare reform while he was in the White House, said that Murray's analyses of the problems created by the welfare system-greater poverty and more exaggerated consequent social problems-were right on target, although he noted that Murray's solutions were too extreme (http://www.newsweek.com/id/125104/ page/1). Banfield and Murray, as well as Clinton and other officials, acknowledge that social structural conditions also contribute to crime, addiction, family disruption, and teen pregnancy but emphasize that the poor hold values that are the major source of their problems and life circumstances. In contrast, many social scientists argue that social structural conditions such as joblessness, inadequate schools, and racial discrimination lead to these problems (cf. Massey and Denton 1993;Sampson and Bean 2006;Wilson 1987). They also generally acknowledge that sometimes persistent disadvantage may lead to the emergence of norms and values that exacerbate the problematic lives of the poor and the communities they inhabit.
In this article, we offer a counter to the traditional culture of poverty argument: a culture of inequality thesis. A number of social problems are produced by persistent poverty, which exists not because of perverted values among the poor, but rather because of values in the larger society that are accepting of social inequality. Where this culture of inequality exists, many people believe that it is acceptable that substantial inequalities be allowed to persist. That is, cultures of inequality exist where the populace has a high "taste for inequality." In these contexts, the view prevails that government is not responsible for ameliorating the causes and products of social and economic inequality. This culture of inequality reaches its highest form among those modern-day social Darwinists who believe that the problems of the poor, the unemployed, and the uneducated are due to their own failures (Banfield 1968). This view holds that government intervention to help these groups will only lead to further dependency. Such efforts, according to this argument, only waste the money of hardworking taxpayers (Murray 1984).
A number of social problems are produced by persistent poverty, which exists not because of perverted values among the poor, but rather because of values in the larger society that are accepting of social inequality.
In our conceptualization, an expressed "taste for inequality" is the observable manifestation of the presence of a culture of inequality. The former is the 136 THE ANNALS OF THE AMERICAN ACADEMY operationalization of the latter. Here we will develop this thesis and use it as a framework for thinking about shifts in imprisonment and social welfare policies. We are especially interested in comparing countries that are more homogeneous with those that are more heterogeneous. However, we also want to tease out and compare countries with an already heterogeneous population base that are experiencing more immigration with countries in the rest of our sample. Large or growing numbers of "others" in a country are viewed as increasing threat (Blalock 1967;Jacobs and Kleban 2003), along with producing a growing taste for inequality. Where that taste becomes broadly accepted among the citizenry, a culture of inequality is likely to emerge. When marked by such a culture of inequality, nations are likely to invest less in social welfare and to respond more punitively to crime. Of particular interest is how the United States, which has had unprecedented increases in imprisonment as well as cutbacks in welfare programs, compares to other nations.

Cultures of Inequality
Except among segments of the general public and some politicians, the old notion that a persistent subculture of poverty is the cause of social problems has been largely discredited (Wilson 1987). Instead, most social scientists recognize that under some conditions, structured inequalities lead to the emergence of norms, values, and behavior patterns that make life difficult for poor communities (Anderson 1999;Miller 2008). Notably, Sampson and Bean (2006) contrast older notions of the existence of a subculture of poverty with contemporary analyses that stress that the social isolation of the poor exacerbates their precarious social positions and produces counterproductive behaviors in their communities.
Scholars are increasingly turning to culture as part of the broader explanation of variation in national imprisonment rates (Jacobs and Kleban 2003;Sutton 2004). Unable to explain patterns solely using indicators of policies or political leadership, they speculate that societies that are more individualistic, or willing to accept inequality, are also more likely to embrace heavy reliance on imprisonment for social control.
Consistent with these themes, we argue that cultures of inequality-where many of the politically powerful or the electorate have a high taste for inequalitydetermine societies' response to social problems. Where the taste for inequality is low, illness is frequently addressed with widespread public health care. Joblessness and unemployment bring about state support for welfare benefits. In contrast, a high taste for inequality leads to a minimalist welfare state and punitive criminal justice practices. People with high tastes for inequality see the source of the problems within disadvantaged individuals, assuming that these people are in dire straits through their own fault and arguing that transgressions should not be rewarded with health care, welfare, or a criminal justice system that does anything other than punish.
During the post-Great Society years, and especially since the election of Ronald Reagan, the United States has experienced widespread, popular acceptance of culture of inequality values. The result has been dramatic cutbacks in welfare, under the guise of "welfare reform," 1 and the ascension of an educational philosophy that assumes that every child in every school has the capacity to learn unless they, their parents, or their teachers fail to put forth sufficient effort. 2 As a result of this increase in the collective tastes for inequality, racial injustice is essentially reduced to a historical fact with little or no bearing on the contemporary life chances of people of color (Sowell 1981). Thus, affirmative action is now defined as discrimination against the privileged (Bonilla-Silva, Lewis, and Embrick 2004). Most crime is viewed as a consequence of rational choices by people unable or unwilling to defer gratification. In turn, society responds harshly to punish offenders and to deter the "not-yet-detected" (Stafford and Warr 1993). From 1973 to 1997, incarceration numbers in the United States increased fivefold (Caplow and Simon 1999). We suggest that this increase is explained in part by a growing taste for inequality as well as politicians' willingness to run for office on platforms that pander to the view that substantial inequality is acceptable and even just.
In addition to a higher incarceration rate and a much less developed welfare state apparatus compared to other Western countries, the United States has historically had a racially and ethnically heterogeneous population. And this heterogeneity is only increasing as a result of immigration. However, changes on the horizon in other Western countries may make them look more similar to the profile of the United States. The European Union has pressured member states to reduce government spending and limit the welfare state in the name of efficiency (Eichengreen et al. 1998). There is a perception that crime is a growing problem in Western Europe, and in some countries the number of people confined in prisons has risen. Canada and Western Europe are also becoming increasingly heterogeneous, which poses the question of whether their welfare generosity will continue unabated. Although the population in some European countries has been substantially diverse for decades, recent media accounts suggest that heterogeneity is growing. (Many immigrants have easy entry to Western European countries because they are citizens by virtue of birth in former colonies.) Thus, Spain has had increased immigration from Latin America (see Harter 2007, BBC News), and France's Muslim and African immigrants have increased (see Astier 2005, BBC News). Some evidence links the growth of heterogeneity to increases in ethnic conflict. For example, the United Kingdom has faced a nationalistic-nativist movement; there have been ethnic/religious stresses in the Netherlands (see Hardy 2006, BBC News); and in Germany and Italy, the presence of immigrant/guest workers has created tensions (see Broomby 2001, BBC News; Lewis 2008, BBC News).

Cultures of Inequality and Threat
The threat hypothesis posits that as a minority population grows, it is more likely to be perceived as a threat by the majority (Blalock 1967). Acts of hostility aimed at immigrants in the United States, Spain, and the United Kingdom are examples of anti-immigrant sentiments and actions that can be explained by the 138 THE ANNALS OF THE AMERICAN ACADEMY perception that "those people" threaten the well-being and the jobs of "deserving natives." Regarding imprisonment, Jacobs and his colleagues (Jacobs, Carmichael, and Kent 2005;Jacobs and Kleban 2003) argue that the threat that others are perceived to represent can help to explain increased sentencing severity. We contend that the threatening presence of immigrants, or others defined as outsiders-including racial and ethnic minorities who have been kept at the margins of societies-increases a society's taste for inequality, which in turn reduces support for welfare benefits and increases punitive criminal justice practices.
Of course, some argue that antiother sentiment exists because others are an actual criminal threat. Individuals explained anti-Latin American immigrant riots in Spain by saying that they were reacting to imported gang activity (see Harter 2007, BBC News). The French foreign minister justified punitive police actions in Paris suburbs on the grounds that the rioters were merely "thugs," who just happened to be North African and Muslim "immigrants." Notably, many participants have been residents of France for more than a generation but still are not considered to be French (Haddad and Balz 2006). In the United States, the disproportionate imprisonment of people of color is thought by many to simply be a function of higher crime rates among minorities. It is hard to dispute that higher levels of criminal involvement explain, in part, disproportionate incarceration of others in the United States. But a substantial literature (e.g., Blumstein 1982;Bridges and Crutchfield 1988) has addressed this question, and the weight of the results indicates that a significant amount of the racial disproportionality in American prisons cannot be accounted for solely by differential criminal involvement. Crime is part of the explanation, but we believe that punishment practices are also a consequence of cultures of inequality.
Data and Methods Tonry (1997) describes the problems inherent in cross-national comparative work. He argues that explaining cross-national racial and ethnic disparities, immigration policies, and other country characteristics can be difficult since the meaning of race, ethnicity, and immigrant alien varies across nations. The historical experience of immigrants to different countries also varies, as do the sentiments of host populations regarding others. Despite the problems involved, solid comparative work in criminology is possible and useful when comparable cross-national data are available, or when data can be used in a substantively comparable fashion. For example, if we use race to measure, in part, the percentage of others in U.S. prisons, while using the percentage of aliens as a measure in the Netherlands, the results may be expected to be fairly comparable. Indeed, using these measures, we find that the United States and Holland have roughly similar percentage of others in prison.

Individual attitudes and the World Values Survey (WVS).
Here, we use attitudinal data from the WVS, which is administered independently across countries. The four published waves are (1) 1981 and 1984, (2) 1989 and 1993, (3) 1994 and 1999, and (4) 1999 and 2004 (see www.worldvaluessurvey.org). The benefits of using the WVS are twofold: (1) it makes a comparison of public attitudes across different countries possible and (2) because the data are at the individual level, it provides large samples, which allows for a comparison of attitudes within and between countries.
We use principal component factor analysis to create scales that measure our three variables of interest: (1) taste for inequality, (2) jobs and immigrants, and (3) attitudes about others. Taste for Inequality is captured with three survey items: "should incomes be made more equal" versus "should income differences serve as incentives"; attitudes about the government's responsibility to ensure that everyone is provided for; and belief that "competition stimulates people to work hard and develop new ideas" versus "competition brings out the worst in people." Two scales are used to measure threat. Jobs and Immigrants uses "should immigrants be allowed in only when jobs are available" and "when jobs are scarce, should employers give priority to nationals." Attitudes about Others combine items that indicate whether respondents mind having immigrants, Muslims, or people of a different race as neighbors. For each of the three scales, higher values indicate greater taste for inequality, less favorable views of immigrants, and less favorable views of others, respectively.
Country-level data. We consider several national-level measures: the prison population, total immigration, homicide rate per one hundred thousand population (as a measure of crime), 3 and the GINI coefficient of income inequality. Sources for these measures are noted in Table 1. Our data pertain to fifteen WVS country samples of individual respondents. We pooled waves 3 and 4 of the survey to increase sample size in situations where countries were lacking data on questionnaire items of interest.
We are interested in comparing public attitudes and preferences, indicating cultures of inequality and threat, across countries grouped according to structural characteristics. For instance, we are interested in whether more heterogeneous countries differ on certain attitudes from those that are more homogeneous. Analysis of variance (ANOVA) is well suited to test the significance of the differences of group means across variables by comparing the size of the variance between samples with the size of the variance within samples (Hofstede 2001). We use one-way ANOVA, 4 which is similar to performing a t-test of the difference of means. A larger F-statistic means that there is more difference in the means across samples compared to the variability within each sample.

Results
These results should be viewed as a demonstration of the utility of the culture of inequality concept and not as a test of theory. Our assessment, which links tastes for inequality, attitudes about others, and the incarceration of others, addresses four expectations. First, we expect that countries with more others will have a higher 140 THE ANNALS OF THE AMERICAN ACADEMY taste for inequality and, in accordance with the threat hypothesis, will have less favorable opinions of immigrants. Second, we believe that it is too simplistic to assume that homogeneous countries differ from more heterogeneous countries on these attitudes. Rather, countries that already have a heterogeneous population and still experience high levels of immigration will have higher tastes for inequality and will also have more negative attitudes toward others. This expectation is based on the contention that heterogeneous nations already have a base level of perceived threat. Third, the consequences of cultures of inequality include more negative attitudes toward others and harsher punishment of others. Jacobs and his colleagues (Jacobs and Kleban 2003;Jacobs, Carmichael, and Kent 2005) argue that enhanced law enforcement tends to accompany threat. Finally, we suspect that the United States may be different (i.e., a case of "American exceptionalism") from the other Western countries in our sample due to its historically high levels of heterogeneity and racial conflict, coupled with high inequality. As a result, the United States should have considerably higher imprisonment of others compared to other countries. Before describing the results of the ANOVA, we present descriptions of the country-level data. We cannot conduct multivariate analyses because of our small sample of countries (i.e., fifteen), but the descriptive statistics suggest relationships between inequality, number of others in the population, and imprisonment. Table 1 includes key values for the fifteen nations. Comparisons across places reveal that although the United States has the highest imprisonment of others (59.7 percent), the Netherlands is a close runner-up at 55.6 percent. (Notably, in the United States, 25.2 percent of the total population is composed of others; while in the Netherlands, the comparable figure is just 10.1 percent.) Table 2 presents comparisons of mean homicide rates, percentage of national populations imprisoned, and proportion of the prison population that is others for countries that are high (above average) and low (below average) on the proportion of the prison population that is others, and the level of observed inequality (GINI coefficient). In nations with larger populations of others and with greater inequality, there are more homicides, people in prison, and others in prison. These patterns are in line with theoretical predictions linking "threatening" populations, income inequality, and crime and punishment. Obviously we cannot conclude from these data that others are actually the ones committing the murders, so we must settle for the ambiguous conclusion that these findings are consistent with studies conducted in the United States of racial disparities in imprisonment (e.g., Bridges and Crutchfield 1988).
The WVS data allow us to examine the expressed attitudes of people within and across these fifteen Western industrialized nations. We divided the sample into countries with high and low income inequality. Table 3 compares means for taste for inequality, jobs and immigrants, and attitudes about others between nations with above-and below-average populations of others. Nations with a higher proportion of others have a significantly higher taste for inequality compared to countries with a low proportion of others in that they are less supportive of welfare policies and of government efforts to correct economic inequality. Contrary to our expectations and the threat hypothesis, nations with more others have more positive attitudes about jobs and immigration, as indicated by the fact that they have less of a problem admitting others when jobs are available. This may reflect respect for the need for imported workers, or greater contact that has led to increased tolerance. However, while residents of high-others countries welcome immigrant labor, they are less inclined to be supportive of equalizing welfare policies. More heterogeneous nations also have more positive attitudes toward others.
As Table 3 shows, it is not more homogeneous nations where respondents are more tolerant of inequality. To examine the role of population heterogeneity, in one analysis (data not shown), we compared means for our key variables relative to increases in countries' immigration levels. This analysis did not yield signs of increased taste for inequality, or changes in the views of others, or jobs and immi gration. It may be that increased heterogeneity has not yet had an opportunity to appreciably shift 142 THE ANNALS OF THE AMERICAN ACADEMY attitudes. Also, some nations are experiencing substantial in-migration on top of considerable existing heterogeneity. The United States is a case in point. 5 To consider how increases in others combine with existing heterogeneity to influence the culture of inequality, we compare countries with a large base population of others to the rest of the sample. These comparisons are presented in Table 4. Already diverse countries, which received large numbers of in-migrants, have higher tastes for inequality. However, such nations do not differ from others on attitudes about immigration and jobs. Attitudes about others are more negative in homogeneous countries than in increasingly heterogeneous nations, a finding that is not consistent with the classic threat hypothesis. However, these countries do have small numbers of others, and perhaps this is sufficient to engender negative attitudes but insufficient to allow for meaningful contact that might lead to familiarity and comfort.
Countries with higher imprisonment of others also, on average, have higher tastes for inequality but are also more open to immigration if jobs are available (see Table 5). At the same time, however, attitudes about others do not vary across country groupings based on high and low imprisonment of others. In general, these results support the culture of inequality thesis, while also demonstrating that attitudes about others may not, on their own, adequately explain welfare policies or imprisonment.  [I]t appears that the United States is similar to its country peers in most comparisons, but it is driving the difference in the mean taste for inequality when comparing countries with low and high imprisonment of others.
This preliminary investigation demonstrates greater support for the concept of cultures of inequality than it does for the classic threat hypothesis. Although threat need not directly translate into negative attitudes about others, it may increase tolerance for inequality. For example, the United States has a considerably higher taste for inequality (see Table 6) but has slightly more positive attitudes about others 6 compared to other countries. We suggest that the United States might be a relatively unusual case driving our results. When the United States is excluded from the analysis (data not shown), the differences of means for taste for inequality is attenuated when comparing countries that are high and low on percentage others, but that difference remains statistically significant. No differences in means emerge for the other scales when the United States is excluded. The mean taste for inequality is not significantly different between countries with low and high others in prison when the United States is excluded from the sample. In brief, it appears that the United States is similar to its country peers in most comparisons, but it is driving the difference in the mean taste for inequality when comparing countries with low and high imprisonment of others.

Discussion
Rather than defining "cultures of inequality" in terms of subcultural norms and values, as was done with early culture of poverty arguments, we argue that acceptance or tolerance of inequality are beliefs held by the larger population. In turn, increased tolerance of inequality results in harsher punitive measures. At the same time, the threat hypothesis claims that where there is a perception that others pose a threat, harsher punishments also result.
The notion of a culture of inequality is heuristic, and our analyses suggest that it is a fruitful perspective for understanding some American public policies within a crossnational comparison. In general, we find support for the culture of inequality thesis. Attitudes about others, whether based on race, ethnicity, religion or nationality, are not enough to explain the treatment of others in Western societies. Ill treatment appears to require an acceptance of social and economic inequality.
In addition to our cross-national comparisons, we set out to examine whether the United States is in fact different from other Western industrialized countries on tolerance of inequality, attitudes about others, and immigrants and jobs. In other words, we asked whether the characteristics of the United States are indicative of "American exceptionalism." On one hand, the United States has a higher taste for inequality than the other fourteen countries in our sample. This appears to be especially meaningful in terms of imprisonment of others. The United States also has the highest number of others in prison. 7 As Figure 1 illustrates, by 1992, imprisonment of African Americans in the United States surpassed the imprisonment of whites, and the gap between the two continued to grow throughout the 1990s and into the new century. By the beginning of the 1990s, Americans' attitudes toward government involvement in welfare also became less favorable.
At the same time as sentiments toward federal involvement in welfare became less favorable, public opinion polls in the late 1980s and early and mid-1990s  pattern of sample sizes and variances while assuming that the means are equal. We found that the p-values of our simulations are very similar to the original results. 5. So too are Austria, Germany, Great Britain, Ireland, and Sweden. 6. As Lapinski et al. (1997) demonstrate, Americans tend to have favorable opinions of immigrants on a personal and cultural level, though they may simultaneously believe that they drain social services.
7. The Netherlands is a very close runner-up.

FIGURE 1 BLACK AND WHITE IMPRISONMENT AND PUBLIC SUPPORT FOR FEDERAL INVOLVEMENT IN WELFARE
SOURCE: Bureau of Justice Statistics and Stimson (2004, 84).