Social Capital,Racial Context,and Incarcerations in the American States

Abstract

This article examines the differential effects of social capital on policy equity in state outcomes. Specifically, it explores the relationship between social capital and incarceration rates in the American states paying particular attention to racial disparities in incarceration rates. Building on work by Hero, I present a theoretical explanation and empirical support for how social capital operates differently under different racial contexts. I argue that social capital enhances social empathy in homogeneous contexts and social controls in diverse contexts. Using state-level longitudinal data on the contiguous states, I find that social capital is positively associated with incarcerations, but only for African Americans. Furthermore, the effects of social capital appear to be conditional on racial context where this relationship is stronger as minority group size increases.

Keywords

social capital racial diversity inequality criminal justice

Introduction

The boom in incarcerations in the United States, which began in the early 1970s, has concerned academics and policy makers alike. Incarcerations continued to rise long after crime rates fell precipitously in the 1990s. Between 1972 and 2001, the federal and state prison incarceration rate went from 93 to 470 inmates per 100,000 population (Pattillo, Weiman, and Western 2004). Equally alarming is the racial disparity regarding who is incarcerated. While a majority of inmates are white (nearly 60%), blacks are significantly overrepresented in the prison population making up over 37% of the total prison population while comprising only 13% of the total U.S. population. According to Western (2006), black men are seven times more likely to be incarcerated than white men of a similar age. Even after accounting for age and educational attainment (high school dropouts vs. high school completion), black males are still 4.8 to 5.3 times more likely to spend time in prison.

This article examines the determinants of mass incarcerations and the racial disparity of incarcerations in the American states. While there are large literatures within political science, criminology, and sociology that explore how social, demographic, and economic factors as well as public policies contribute to incarcerations, this article is specifically interested in the role that social capital and racial context play in explaining incarceration rates across the states. Building on past work on social capital and inequality (e.g., Hero 2003; 2007), I argue that social capital can enhance social empathy and trust or it can strengthen social controls (including the penal system). I hypothesize that social capital will be more likely to enhance social control mechanisms (rather than empathy) as racial diversity and racial threat increase, which would result in increased racial disparities in incarceration rates. Using state-level panel data, I find support for the hypothesis that the effect of social capital on incarcerations is conditional on racial context and that social capital exacerbates the racial gap under certain conditions.

Social Capital and Criminal Justice

Social capital is considered a social glue that holds communities together and can act as a social lubricant that helps society function more smoothly (Anderson and Jack 2010; Grootaert 1998). Within political science, the most ubiquitous definition of social capital comes from Putnam (2000), where he defines social capital as “connections among individuals” or “social networks and the norms of reciprocity and trustworthiness that arise from them” (Putnam 2000, 19). Social capital includes both formal and informal networks, all of which make individuals and communities more connected. As a result, society functions more smoothly. Putnam argues that higher levels of social capital are associated with greater social trust, political participation, and citizen cooperation resulting in better social and policy outcomes in the aggregate.

While social capital has often been thought of as an individual-level characteristic that accrues within individuals much like economic or human capital, Putnam (2000) argues that it also has importance at the aggregate level. Community-level factors, such as social capital, have an effect on individual-level behavior and outcomes. He argues that people are “profoundly motivated not merely by their own choices and circumstances, but also by the choices and circumstances of their neighbors” (Putnam 2000, 312). He continues, “My fate depends not only on whether I study, stay off drugs, go to church, but also on whether my neighbors do these things” (Putnam 2000, 312). The idea is that social capital—or the lack thereof—can reinforce and cultivate patterns and behaviors within individuals that have collective, aggregate-level effects. Indeed, much of the empirical work on social capital within political science has been at the aggregate level.

There is a significant literature on the relationship between social context, including social capital, and crime.¹ Shaw and McKay (1942) contend that informal social controls and organization can work to prevent criminal activity. A simple example would be a neighborhood watch group where neighbors monitor each other’s properties. These informal social controls and organization are more likely to exist and thrive in communities with higher levels of social capital—a logical conclusion given how social capital has been conceptualized and operationalized. For example, Putnam (2000) argues that among the core components of social capital are civic participation, volunteerism, social trust, and engagement in public affairs. All of these factors should conceptually increase participation in developing and maintaining a successful neighborhood watch program within a community. Alternatively, we would not expect to see the development of such a program in areas that lacked these traits.

Bursik and Grasmick (1993) have added to Shaw and McKay’s classical social disorganization theory by considering the interaction between informal and formal processes of control. In the systemic model of crime, formal “public control”—for example, law enforcement—is bolstered by informal processes including community involvement and efforts to secure public services for the community. Rosenfeld, Messner, and Baumer (2001, 287) make the case that social capital plays a role in this process, and hence “areas with extensive civic engagement are better able to secure adequate policing and other resources relevant to the ‘public control’ of crime.” Therefore, using a social disorganization perspective, communities with higher levels of social capital should have lower crime rates either due to social capital’s reinforcement of informal social controls and organization via social trust and public engagement or because these communities are better able to secure resources for formal public control.²

Empirical work also supports a connection between social capital and crime. Sampson and Groves (1989) found higher levels of social organization and participation were related to lower crime rates in Great Britain. Rosenfeld, Messner, and Baumer (2001) find that social capital is negatively correlated to homicide rates in a study of 99 counties in the United States. Kennedy et al. (1998) also found that social capital (and inequality) was a significant predictor of violent crime rates in U.S. states. Similarly, other studies have found a negative relationship between social capital and both violent and property crimes in the American states (e.g., Hawes, Rocha, and Meier 2013; Kawachi, Kennedy, and Wilkinson 1999).³

There is less work, however, examining the relationship between social capital and incarcerations, and particularly disparities in incarceration rates. Social capital can affect incarcerations and the criminal justice system via both informal and formal mechanisms. On its face, if social capital enhances social trust and empathy, we may expect high social capital areas to have lower incarcerations because citizens (e.g., jurors, prosecutors) will be more empathetic and/or trusting of each other and of the accused. Higher empathy may make citizens more likely to support lower sentences or rehabilitation programs rather than harsh sentences. This can operate directly through how citizens behave and the role they play in the criminal justice process (how they report crime, participate on juries) or more indirectly through the development and adoption of public policy (whether they elect “law and order” candidates, support or oppose harsh crime legislation). Social trust and empathy should make citizens more likely to place themselves in their neighbors’ shoes, give others the benefit of the doubt, and support second chances. Increased social trust and empathy could result in lower incarceration rates in the aggregate, even after controlling for crime rates. Generally, this could be stated as a social trust/empathy hypothesis:

Social Empathy Thesis (Hypothesis 1): Higher levels of social capital will be negatively related to incarceration rates.

That said, it is also possible that empathy and the norms of reciprocity will only apply to those who follow these norms. Coleman (1988, S119) describes social capital as “obligations and expectations, which depend on trustworthiness of the social environment . . . and norms accompanied by sanctions.”. This suggests that higher levels of social capital will be associated with both having these norms and with being more likely to impose sanctions on those who break them. Those who break the norms (i.e., criminals) may be treated with less compassion than they would in community with weaker norms (i.e., lower social capital). That is, social empathy may only extend to those who play by the rules—rules that are more clearly established and adhered to in high social capital communities. Areas with low social capital are less likely to have strong social norms and as such may likely be less punitive to those who break them.

An alternative mechanism by which social capital may increase incarceration rates is via the effectiveness and efficiency of social controls. Social capital may increase police efficiency due to citizen cooperation and, in turn, increase the likelihood of solving and prosecuting crimes. Indeed, to the extent that social capital can enhance informal controls (e.g., neighborhood watch) and formal controls (e.g., increase funding and support for police departments), social capital may improve the effectiveness of police work, increasing case clearance rates and potential conviction rates. This could result in higher incarceration overall rates (controlling for crime rates). Collectively, we can think of this as a hypothesis where social capital and sanctions enhance social controls in society. Put formally,

Social Control Thesis (Hypothesis 2): Higher levels of social capital will be positively related to incarceration rates.

The Role of Inequality and Race

In terms of racial disparities in incarcerations, the relationship is less clear. Putnam (2000, 294) has argued that social capital works to promote racial equality. As he writes in Bowling Alone, “inequality and social solidarity are deeply incompatible.” However, Putnam (2000, 358) also acknowledges that “social capital, particularly social capital that bonds us with others like us, often reinforces social stratification.” Putnam (2007) builds on this with the concept of “Constrict Theory,” which suggests that racial and ethnic diversity may decrease both in-group and out-group solidarity, or bonding versus bridging social capital. Diversity, he finds, can trigger social isolation where “people living in ethnically diverse settings appear to ‘hunker down’” (Putnam 2007, 149). This effect, however, appears to be particularly strong for social trust, which is related to the social empathy hypothesis. If social capital reinforces bonding social capital (in-group) and stratification, then we might expect the black–white disparity in incarcerations to be higher in areas with higher levels of social capital. That is, perhaps the social trust and empathy that social capital engenders are conditional on racial context.

A common theme in much of the work on the causes of crime is that inequality—particularly economic inequality—is at the heart of this process. Indeed, social disorganization theory, capital disinvestment, and strain theory all have aspects of inequality at their core. For example, in strain theory, absolute levels of deprivation are less important than relative deprivation; that is, the level of deprivation compared with a reference group is more consequential than deprivation in an absolute sense (Burton and Dunaway 1994). Others, such as Blau and Blau (1982) and Sampson and Wilson (1995), argue that inequality—whether across racial groups or in terms of geographical concentrations of poverty—is a central cause of social disorganization.

The empirical work on the relationship between social capital, inequality, and incarcerations is less clear. Hero (2007), for example, has linked social capital to a number of measures of inequality in the United States and found a positive relationship. That is, social capital appears to be related to an increase in inequalities between Anglos and minority groups. This was also found to be the case in a study by Hawes and Rocha (2011) that examined inequality between whites and blacks across several policy areas including incarcerations.

One possible reason for this discrepancy in the relationship between social capital and crime and inequality in incarcerations is that the determinants of crime may be different from the determinants of incarcerations. The theoretical link between social capital and crime is fairly straightforward; however, incarcerations are more complex. Incarceration rates are obviously a function of actual crime rates; hence, it should also be related to social capital (assuming social capital is a determinant of crime). However, there are numerous other factors, including bureaucratic discretion and citizen participation (either in reporting crime or as jury members). If social capital and trust vary across and between racial and socioeconomic groups, decisions that are crucial in detecting, prosecuting, and convicting criminals may result in disparities in enforcement outcomes.

This suggests that social capital may reduce crime rates while increasing inequality in incarceration rates between racial groups. This proposition differs from the claims Putnam (2000) has made, in which he argues social capital should promote equality due to reinforced norms of reciprocity, inclusiveness, and a community-oriented perspective. Putnam (2000, p. 356) states, “far from being incompatible, liberty and fraternity are mutually supportive . . . the most tolerant communities in America are precisely the places with the greatest civic involvement.” The result, he contends, is that minorities will be better off in high social capital communities. Social capital, however, could result in discrimination toward minorities—whether intentional or not. This is in line with Hero’s (1998; 2003; Hero and Tolbert 1996) work in which he argues social capital or the “civic republicanism” tradition ignores ascriptive hierarchy, which is closely linked to racial/ethnic segregation and bifurcation within states. That is, social capital tends to be highly correlated with racial homogeneity. If social trust and norms of reciprocity are linked to race, these norms may not apply equally to racial and ethnic minorities. Building on Putnam’s Constrict Theory, I agree that social capital is enhanced in racially homogeneous contexts and diminished as racial diversity increases (see also Hero 2007). Theoretically, as diversity increases, we will see reductions in bridging social capital (out-of-group bonds) while bonding social capital (in-group) and stratification may be increased. This article argues that social capital itself operates differently under different racial contexts. I extend Putnam’s Constrict Theory by arguing that as racial/ethnic diversity increases, not only does social capital generally decrease (as Putnam argues), but that it enhances social controls, both formal and informal (e.g., neighborhood watch programs) at the expense of social empathy.

Take, for example, a hypothetical neighborhood watch program in a high social capital, racially homogeneous (white) neighborhood.⁴ Due to a high sense of community and civic engagement, neighbors vigilantly monitor the neighborhood and faithfully report “suspicious” activity. In such a scenario, a Hispanic or African American man driving through the neighborhood may be at an increased risk of being reported to police as “suspicious” than his white counterpart. This is both due to racial bifurcation and segregation and increased social capital. In a highly diverse community, it would not be strange to see an individual of a different race driving through the neighborhood. In addition, if social capital were low, there may not be an effective neighborhood watch program and the “suspicious” activity may go unreported. Thus, it is the combination of racial homogeneity and high social capital that results in an increased likelihood of a report being filed. An active and engaged neighborhood watch program could be thought of as more “effective” policing (i.e., more arrests and incarcerations), which in itself is a good thing. However, if social capital is also highly correlated with racial homogeneity, minorities who do live in these areas may be more likely to be targeted and profiled; hence, we may see higher disparities across racial/ethnic lines.

This is compatible with literature on the intergroup conflict thesis and specifically the racial threat hypothesis. This research posits that as the size of a minority group increases within a population, prejudicial attitudes of whites will also increase. This effect is due to a perceived threat that the minority group poses to scarce economic resources (e.g., jobs, spending priorities) and social privilege (Oliver and Wong 2003). Many empirical studies have found support for the racial threat thesis (Fossett and Kiecolt 1989; Giles and Buckner 1993; Glaeser 1994; Rocha and Espino 2009). With respect to social capital development, if the racial threat hypothesis were correct, we would expect that higher levels of minorities and/or diversity would create an environment of conflict as and negative racial attitudes toward minority groups. This conflict and hostility would have a negative effect on social trust and cohesion, which could, in turn, affect how social capital operates in society.

This could, in part, explain the massive disparity in incarceration rates between whites and blacks. The racial threat hypothesis suggests that social capital may enhance social trust in some contexts and social control in others. Indeed, Putnam (2007) finds that ethnic homogeneity enhances social capital and social trust. Thus, while overall levels of social capital⁵ are generally lower in more ethnically diverse communities (Hero 2007; Putnam 2007), I argue that social capital itself will tend to take different forms in diverse versus homogeneous communities. As racial diversity and minority group size decrease, social capital will work to enhance social trust. However, as racial diversity and minority group size increase, social capital will work to enhance social controls. The empirical implication of this is a conditional hypothesis:

Social Capital Conditional on Racial Context (Hypothesis 3): The effect of social capital on black incarceration rates relative to white rates is conditional on the racial context in the state. The social empathy hypothesis is more likely when racial homogeneity is high while the Social Control hypothesis is more likely as the racial context becomes more diverse.

Data and Method

The empirical analysis examines the relationship between social capital and incarcerations at the state level from 1986 to 2009.⁶ It examines both absolute incarceration rates and relative rates between whites and blacks. The empirical analysis employs a pooled cross-sectional time-series design to account for both the cross-sectional and temporal dynamics of these relationships. All dependent variables were tested for panel stationarity and were found to be panel stationary.⁷ The incarceration rate models use Seemingly Unrelated Regression (SUR) models, and the ratio models utilize herteroscedastic panel-corrected standard errors (Beck and Katz 1995) to control for heterogeneity across the states.

Dependent Variables

Putnam (2000) argues that minorities will generally be better off in high social capital areas because social capital is expected to produce more desirable policy outcomes (e.g., Putnam 2000). However, others contend that minorities may be worse off relative to whites in high social capital states (e.g., Hero 2003). It is possible that both arguments are correct. That is, minorities may be at a higher relative risk of being incarcerated compared with whites, but overall incarcerations may be lower—including those of minorities. Thus, we may see “better,” albeit more unequal, policy outcomes for minorities. To examine this possibility, both absolute and race-specific relative measures are needed.

State-level incarceration rates from 1986 to 2009 were obtained from the Bureau of Justice Statistics’ National Prisoners Statistics study.⁸ These data include incarceration rates of whites and blacks in state prisons (White/Black Inmates) and are measured as the number of white and black state prisoners per 100,000 of their respective state population. During this time period, the average incarceration rate for whites was 207 inmates per 100,000 white population with a maximum rate of 547 (Arkansas in 2007). The average black incarceration rate was significantly higher at 1,508 inmates per 100,000 black population, with a high of 9,545 inmates per 100,000 black population (South Dakota in 1992). In this same year, the white incarceration rate was only 167 per 100,000 white population in South Dakota.

This disparity highlights the high level of inequality that exists in criminal justice policy throughout the states. To capture this disparity, the final dependent variable is the incarceration odds ratio for black and white prisoners. It is simply the ratio of the black and white incarceration rate (black rate/white rate). A value of 1 represents parity, that is, whites and blacks have the same likelihood of being incarcerated. Alternatively, values higher than 1 indicate that blacks have higher odds of incarceration. The average black–white incarceration ratio during this time period is 8.32; hence, blacks were over eight times more likely to be incarcerated than whites. South Dakota and Iowa had the highest black–white ratios of 57.3 and 33.4, respectively. From 1986 to 2009, there were 24 states that had black–white incarceration ratios of 10 or greater (totaling 287 state-years). No state had a black–white ratio of 1 or less, that is, an equal odds ratio.⁹

Social Capital

Social capital can be measured in a variety of ways, and there is some debate as to what is the best way to measure it. As the debate in the literature regarding the relationship between social capital and inequality has largely centered around Putnam’s arguments, I use a measure of social capital that is in line with Putnam’s conceptualization of social capital. This conceptualization of social capital has many limitations. Most notably, it arguably does not explicitly capture social capital for minority groups because social capital can take very different forms in African American and Latino communities (e.g., Harris-Perry 2004). This is a key criticism of Putnam’s conceptualization and operationalization of social capital, that is, it can be characterized as middle-class, Anglo social capital. However, this is precisely the type of social capital that the theoretical arguments above are interested in and would be potentially sensitive to racial threat; thus, it is appropriate for the purposes of this analysis. An important caveat is that the findings and implications do not apply to other conceptualizations of social capital (e.g., black-centric or Latino-centric social capital) that are not adequately conceptualized or measured in Putnam’s work.

Putnam (2000) argues there are five components to social capital, three behavioral (community organizational life, public engagement, and volunteerism) and two attitudinal (social trust and informal sociability). For this article, social capital is measured using an index created by Hawes, Rocha, and Meier (2013) using factor analysis of 22 items that capture the behavioral components of social capital. This index primarily relies on state-level data from a market research firm (MediaMark, Inc.) that conducts large annual surveys that include items related to organizational membership, volunteerism, and civic engagement. They conduct factor analysis of 22 items that measure (1) associational membership, (2) engagement in public affairs, and (3) volunteerism. Associational membership includes the percent of population who belonged to one of the following: a fraternal order, a religious club, a civic club, a veteran club, a body of local government, or a country club. Engagement in public affairs was captured with 10 variables that measured the percent of the population that engaged in the following activities: (1) voted in a federal, state, or local election; (2) wrote to an elected official about a matter of public business; (3) wrote to an editor of a magazine or newspaper; (4) wrote or telephoned a radio or television station; (5) wrote something that has been published; (6) addressed a public meeting; (7) visited an elected official to express a point of view; (8) actively worked for a political party or candidate; (9) engaged in fund-raising; and (10) overall voter turnout in national elections. Finally, six items were used to measure volunteerism: (1) nonprofits per capita, (2) a generosity index, (3) percent who contributed to public television, (4) average amount of contributions, (5) percent engaged in nonpolitical volunteer work, and (6) percent active in a local civic issue.

Factor analysis of these items was used to produce a single composite index (eigenvalue = 7.57).¹⁰ Hawes, Rocha, and Meier (2013) validate this measure and find it is highly correlated with Putnam’s state-level index as well as many other variables the literature suggests social capital should be correlated with. This measure has been used in other work, including Hawes and Rocha (2011), to examine the relationship between social capital and inequality in racial policy outcomes across a range of policy areas.

Racial Diversity

Hero (1998; 2003; 2007) convincingly argues social diversity (or ascriptive hierarchy) can be equated with minority diversity. Furthermore, he argues, in considering the effects social capital has public policy outcomes for minorities, one must consider racial/ethnic diversity. Hero (1998; Hero and Tolbert 1996) argues that state politics are influenced by the level of racial/ethnic diversity within a state. Indeed, Hero (1998) finds that minorities are often better off, relative to whites, in states with high levels of minority diversity. Hawes and Rocha (2011) also find that state-level diversity is an important determinant of policy equity across several policy areas. Perhaps even more importantly, Hero (2007) suggests that the effect social capital has on state policy outcomes may be contingent on racial diversity. Thus, following Hero, I include racial diversity as a central theoretical variable in explaining crime and criminal justice policy outcomes. To do this, I use a Blau dissimilarity index to measure diversity at the state level.¹¹

While Hero has found that racial diversity is an important predictor in explaining state policy outcomes, the racial threat literature contends that minority group size—particularly black group size—is important in explaining incarcerations. Smith (2004), for example, finds that the percent of a state’s population that is black is a significant predictor of state incarceration rates. Furthermore, Yates and Fording (2005) argue that the percent of the population that is black moderates the relationship between conservative politics and black incarcerations. Thus, in addition to racial diversity, I also consider the effect of the percent of the population that is black on crime and incarcerations.¹²

Crime Rates

Perhaps the most obvious determinant of incarcerations is the actual occurrence of crime. However, a cursory glance at data suggests that the unprecedented increase in incarcerations cannot simply be explained by occurrence in crime, at least not in recent years. While crime rates increased significantly between 1960 and the early 1990s, they have been on the decline ever since. Indeed, the crime rate in 2005 was roughly equal to the crime rate in the early 1970s when we began to see the spike in incarcerations (Western 2006). Crime rates began to rise significantly in the mid-1960s and incarceration rates followed suit about 10 years later. Yet, crime rates peaked in the early 1990s and have seen a precipitous decline ever since. Nearly two decades later, however, incarceration rates have not reversed their trend. In fact, incarcerations have continually increased since the 1980s even if we consider incarceration rates relative to crime rates. Incarcerations relative to crime rates significantly decline through the 1960s and were relatively stable during the 1970s. However, beginning in the 1980s, there was an increase in the incarceration rate relative to crime rates which has accelerated since the 1990s. For example, in the mid-1970s, there were approximately 0.03 state inmates per crimes committed. This ratio steadily increased, and by 2009, the ratio was 1.25, over 40 times greater than the ratio in the 1970s. While the causes of the prison buildup are numerous and varied (see Useem and Piehl 2008), it does not appear to be simply a product of an increase in criminal activity (see Smith 2004; Yates and Fording 2005).

That said, I do include two measures of crime in the models. Every year, the Federal Bureau of Investigation (FBI) publishes the Uniform Crime Report (UCR)—a collection crime statistics from state and local law enforcement agencies. The UCR breaks down crime into either violent crime or property crime and provides the total number of crime rates in these categories per 100,000 population. These crime rates serve as two state-level measures for total crime in each state.¹³ These data are available from 1986 to 2009, the full time span of this study.

Control Variables

The models also control for a number of political, economic, and demographic factors. Table 1 presents the descriptive statistics and sources for all the variables used in the models. The models control for four state-level political variables: the percent of females in state legislature, the percent of African Americans in the state legislature, the percent of Democrats in state legislature, and Berry et al.’s (2013) measure of government ideology. We expect that states controlled by more males, fewer African Americans, and more conservative Republican lawmakers will be more likely to advocate “get tough on crime” legislation and allocate more state resources and support to law enforcement and prisons, which could in turn have an effect on both crime and incarceration rates.

Table 1.

Summary Statistics.

Variables	M	SD	Source
Dependent variables
Total inmates per 100,000 population	350.730	164.080	BJS National Prison Statistics (ICPSR Study No. 34981)
White inmates per 100,000 population	202.329	88.762	BJS National Prison Statistics (ICPSR Study No. 34981)
Black inmates per 100,000 population	1,503.607	628.849	BJS National Prison Statistics (ICPSR Study No. 34981)
Black–white incarceration ratio	8.264	4.005	BJS National Prison Statistics (ICPSR Study No. 34981)
Independent variables
Social capital index	0.131	0.973	Hawes, Rocha, and Meier (2013)
Racial diversity (Blau index)	0.317	0.158	U.S. Census Bureau Population Estimates
Percent black population	10.388	9.471	U.S. Census Bureau Population Estimates
Total violent crimes per 100,000 population	468.600	235.566	FBI Uniform Crime Report
Total property crimes per 100,000 population	3,930.259	1,116.649	FBI Uniform Crime Report
Government ideology	50.179	25.758	Berry et al. (2013)
Percent Democrats in state legislature	53.711	15.778	Indiana State University—Klarner Politics
Percent females in state legislature	20.346	7.842	Center for American Women and Politics
Percent African Americans in state legislature	7.068	6.603	National Conference of State Legislatures/Joint Center
Three-strikes law (yes/no)	0.374	0.484	National Conference of State Legislatures
Trial court clearance rate	86.897	26.634	State Court Statistics (ICPSR Study No. 9266)
Drug arrests/population	0.005	0.005	FBI Uniform Crime Report (Sorens, Muedini, Ruger 2008)
GSP per capita (in 2007 $1,000 USD)	30.459	10.358	U.S. Department of Commerce, Bureau of Economic Analysis
Unemployment rate	5.271	1.652	U.S. Department of Commerce, Bureau of Economic Analysis
Poverty rate	12.683	3.739	U.S. Census Bureau, Current Population Survey
Poverty inequality (black–white ratio)	2.825	0.697	U.S. Census Bureau, Current Population Survey
Percent with college degree	23.912	5.222	U.S. Census Bureau, Current Population Survey
Education inequality (black–white ratio)	0.636	0.191	U.S. Census Bureau, Current Population Survey
Divorces per 1,000 population	4.447	1.375	National Center for Health Statistics
Voting-ineligible felons per 100,000	1,199.759	801.093	George Mason University, United States Election Project

Note. A moving average (when possible) or the most recent available years were used to replace missing years for the independent variables. BJS = Bureau of Justice Statistics; ICPSR = Inter-university Consortium for Political and Social Research; FBI = Federal Bureau of Investigation; GSP = gross state product.

I also include three variables related to state criminal justice policy. Three-strikes law is a dichotomous variable that indicates whether the state has adopted a policy with enhanced sentences for habitual offenders. In 2009, there were 27 states with some form of three strikes laws, compared with only two states in 1990. The vast majority of the adoptions of three strikes laws occurred in the mid-1990s with 23 adoptions in 1994 and 1995 alone.¹⁴ I expect states with three strikes laws to have higher incarceration rates because they are required to incarcerate repeat offenders (in some cases for life). Trial court clearance rate captures the percent of cases brought to trial courts that are cleared in a given year.¹⁵ Arguably, states with higher clearance rates will have higher incarcerations as they are able to prosecute and move a higher percent of cases through the court system. Finally, I control for drug enforcement rates. Drug arrests/population measures the total number of drug-related arrests (age 18 or older) per capita. States that target and heavily enforce drug use are expected to have higher incarceration rates. I also expect all three of these policy variables to have a disproportionately large effect on blacks relative to whites. Therefore, these policies are expected to increase overall incarcerations and especially African American incarceration rates.

State demographic controls include gross state product (GSP) per capita to control for state-level economic conditions that could affect incarceration rates. GSP is measured in 2007 constant dollars and is also adjusted for across state cost-of-living differences.¹⁶ In addition, the models include the percent of the population that are living in poverty and the percent with a college degree. There are also two measures that capture the level of racial inequality in poverty and educational attainment: the black–white odds ratios for poverty and college degrees, respectively. Higher values on these measures indicate that blacks have a higher rate of poverty or educational attainment within the state than whites, respectively. The models also control for the divorce rate because it has been linked to higher levels of crime, particularly homicide rates, which could lead to higher incarcerations (Land, McCall, and Cohen 1990). Finally, the number of voting-ineligible felons per 100,000 voting-eligible persons is included as a control. As many felons face a probationary period (e.g., parole) where even minor offenses (drug or firearm possession) can result in incarceration, states with higher felon populations may have higher incarceration rates. Furthermore, states with a large number of felons may witness higher crime rates and may affect citizens’ and policy makers’ attitudes toward criminal justice policy resulting in harsher and longer sentencing, which may cumulatively increase the prison population over time. Table 1 presents the descriptive statistics for the variables.

Findings

Table 2 presents the results for the models examining the relationship between social capital and incarcerations in state prisons. The first column presents total incarceration rates, followed by white and black rates, respectively. As the dependent variables in the first three models are related to each other—that is, it is highly likely that factors that affect white rates will also affect black rates—I estimate these models using SUR models.¹⁷ While other estimators produce similar results, the SUR models are more efficient given that the errors are likely correlated across the models.¹⁸ Furthermore, this estimation is consistent with past research examining these dependent variables (e.g., Yates and Fording 2005).

Table 2.

Social Capital and Prison Incarceration Rate.

	Prison inmates per 100,000 population			ln(odds ratio)
	Total	White	Black	Black/white
Social capital	−0.076	−8.963***	59.535***	0.062***
Social capital	(2.250)	(1.754)	(14.356)	(0.013)
Racial diversity	410.888***	−32.231**	687.365***	0.624***
Racial diversity	(20.691)	(16.151)	(131.731)	(0.212)
ln(% black)	32.856***	16.570***	−95.869***	−0.090***
ln(% black)	(3.827)	(2.999)	(24.338)	(0.019)
Total violent crime rate	−0.015	−−0.010	0.049	0.00003
Total violent crime rate	(0.013)	(0.010)	(0.084)	(0.0001)
Total property crime rate	0.002	0.020***	0.076***	−0.00001
Total property crime rate	(0.003)	(0.002)	(0.016)	(0.00001)
Government ideology	−0.195***	−0.237***	−0.316	0.0004
Government ideology	(0.070)	(0.055)	(0.445)	(0.0004)
% Democrats in legislature	−2.282***	−1.471***	−6.342***	0.001*
% Democrats in legislature	(0.140)	(0.109)	(0.890)	(0.0008)
% females in legislature	−0.326	−0.088	10.905***	0.001
% females in legislature	(0.295)	(0.231)	(1.871)	(0.002)
% black legislators	4.796***	0.682*	0.377	−0.020***
% black legislators	(0.449)	(0.351)	(2.855)	(0.003)
Three-strikes law	−17.074***	−22.163***	−2.702	0.149***
Three-strikes law	(3.598)	(2.817)	(22.855)	(0.019)
Trial court clearance rate	−0.072	−0.198***	0.216	0.001***
Trial court clearance rate	(0.062)	(0.049)	(0.392)	(0.0002)
Drug arrests/population	−525.619*	563.270**	−8,629***	−6.868***
Drug arrests/population	(317.465)	(248.456)	(2,016)	(1.050)
State GSP per capita	6.344***	5.464***	21.084***	−0.003
State GSP per capita	(0.312)	(0.245)	(1.978)	(0.002)
Unemployment rate	−5.800***	−2.934***	−26.054***	−0.002
Unemployment rate	(0.964)	(0.751)	(6.149)	(0.006)
Poverty rate	3.616***	3.096***	4.132	−0.019***
Poverty rate	(0.611)	(0.478)	(3.885)	(0.004)
Black–white poverty ratio	−12.875***	−36.264***	155.336***	0.310***
Black–white poverty ratio	(2.748)	(2.170)	(17.404)	(0.014)
% college degrees	−6.906***	−2.648***	−11.186***	−0.003
% college degrees	(0.621)	(0.488)	(3.935)	(0.004)
Black–white college ratio	194.500***	143.770***	112.455	−0.673***
Black–white college ratio	(14.772)	(11.726)	(93.716)	(0.098)
Divorces per 1,000 population	19.116***	20.263***	68.873***	−0.057***
Divorces per 1,000 population	(1.409)	(1.107)	(8.942)	(0.007)
Felons per 100,000	0.058***	0.032***	0.225***	−0.00004***
Felons per 100,000	(0.002)	(0.002)	(0.015)	(0.00001)
Constant	—	—	—	2.210***
Constant	—	—	—	(0.165)
Observations	1,143	1,143	1,143	1,143
R ²	—	—	—	.563
Number of states	48	48	48	48

Note. Models 1 to 3—Seemingly Unrelated Regression (XTSUR). Model 4—Panel Corrected Standard Errors (PCSE) Model. GSP = gross state product.

Standard errors in parentheses. *p < .1. **p < .05. ***p < .01.

Most notable is that social capital does not appear to be statistically related to total incarceration rates, but it is a strong predictor of race-specific rates, albeit in opposite directions. Social capital is negatively related to total and white prison incarcerations; however, the opposite is true for black incarceration rates: there is a strong positive relationship between social capital and black incarcerations. A one standard deviation increase in social capital is associated with about nine fewer white incarcerations (per 100,000 white population) and nearly 60 more black incarcerations (per 100,000 black population). This finding is very robust and is not sensitive to model specification.¹⁹ This suggests that the effects of social capital are race specific and operates differently for whites than blacks—in this case, to the detriment of the latter.

Another way to think about how social capital has differential effects of white versus black citizens with respect to incarcerations is to model the odds ratios of black/white incarceration rates (see Hawes and Rocha 2011; Hero 2003). The odds ratio captures the relative likelihood of incarceration for blacks compared with whites. Higher values indicate that blacks are more likely to be incarcerated relative to the white incarceration rate. The final column in Table 2 presents the relationship between social capital and the relative likelihood of incarceration for blacks and whites.²⁰ Here we see that social capital is associated with a higher incarceration rate for blacks relative to whites. The effect of a one-point increase in social capital is a 6.2% increase in the logged prison ratio. Given that the average (median) black–white incarceration logged odds ratio in 2009 was 1.97, a 6% increase is not trivial. To put this in perspective, if we used the 2009 average white prison incarceration rate as a baseline (about 255 inmates per 100,000 white population), a logged black–white ratio of 1.97 (i.e., the median state ratio) means there would be about 2,060 black inmates per 100,000 black population. The model suggests that a one-point increase in social capital would translate into an additional 224 black inmates per 100,000 population, holding the white inmate rate constant.²¹ This would translate into an additional 780 black inmates in an average-sized state in 2009.²² Obviously, these figures would be significantly higher for states that have higher baseline incarceration rates.

Racial diversity is also associated with an increase in the black–white odds ratio. A significant increase in racial diversity (0.16) is associated with about 9.9% increase in the black–white incarceration logged odds ratio. This effect on incarcerations is larger than that of social capital.²³ Interestingly, the percent of the population that is black has a negative relationship on the black–white odds ratio. The models in Table 2 suggest that overall incarceration rates are positively related with the percent of a state’s black population, but that the black incarceration rate goes down—both in absolute numbers and relative to whites—as the percent of the black population increases. This relationship is not sensitive to whether or not racial diversity is included in the model. Past work (e.g., Smith 2004) has found a positive relationship between black population size and overall incarceration rates. These findings suggest, however, that black incarceration rates will be lower in states with a larger black presence, all else being equal.

The relationship between social capital and black–white incarcerations presented in Table 2 is only the short-term impacts of social capital. Table 3 presents autoregressive error correction models that capture the longer term relationship between social capital and incarceration rates. Here we include the lag of social capital, the change in social capital, and a disequilibrium term (DV_t−1 − Social Capital_t−1) that captures how quickly the short-term effects dissipate (Zhu 2013). Here we see that the effects of social capital appear to be long-lasting. That is, the disequilibrium term is moderate in size, suggesting that the effects of a change in social capital do not dissipate immediately or quickly. The long-run impact can be calculated as 1 − (β₂ / γ), where γ is the coefficient from the disequilibrium term (i.e., DV_t−1 − Social Capital_t−1). The long-run effect of a one-time, one-unit increase in social capital is about an 8.3% increase in the logged odds ratio. It is worth noting that a one-point increase in social capital is a very substantial change and very unlikely to occur from one year to the next. Indeed, only seven states witnessed such a one-year change in social capital from 1986 to 2009. That said, if we consider changes in social capital over time, even a short time period, a one-point shift in social capital is not unreasonable. From 1986 to 2009, there were 151 cases that witnessed at least a one-point change in social capital in a five-year period and 72 cases within a three-year period.

Table 3.

Long-Term Effect of Social Capital on Black–White Incarceration Ratios.

	Error correction models
	DV: Δ ln(black–white prison ratio)
	β	(SE)
ΔSocial capital	0.013	(0.012)
Social capital_(t−1)	−0.144***	(0.020)
DV_(t−1) − Social capital_(t−1)	−0.157***	(0.023)
Racial diversity	0.182**	(0.086)
ln(% black)	−0.007	(0.012)
Total violent crime rate	0.00003	(0.00003)
Total property crime rate	−3.59e-06	(6.55e-06)
Government ideology	−0.00003	(0.0002)
% Democrats in legislature	0.0004	(0.0005)
% females in legislature	0.0003	(0.0008)
% black legislators	−0.005***	(0.001)
Three-strikes law	0.025**	(0.011)
Trial court clearance rate	0.0002	(0.0001)
Drug arrests/population	−0.788	(0.569)
State GSP per capita	−0.002***	(0.0007)
Unemployment	−0.003	(0.003)
Poverty rate	−0.003	(0.002)
Black–white poverty ratio	0.046***	(0.011)
% college degrees	−0.0004	(0.002)
Black–white college degree	−0.054	(0.065)
Divorce per 1,000 population	−0.011***	(0.003)
Felons per 100,000	−0.00001*	(0.00000)
Constant	0.343***	(0.100)
Long-run effects 1 − (β2 / γ)	0.083
Observations	1,141
R ²	.096
Number of states	48

Note. Panel corrected standard errors in parentheses. DV = dependent variable; GSP = gross state product.

p < .1. **p < .05. ***p < .01.

The preceding analysis presents strong support for the hypothesis that social capital is related to incarcerations (particularly Hypothesis 2). This could be interpreted in a number of ways. It could be that states with high social capital also tend to have lower tolerance for lawbreakers, or that they have more stringent laws (hence more opportunities for them to be broken), or that they are generally more punitive. Alternatively, it may be that high social capital states have more effective policing, perhaps as a result of higher citizen and community participation. However, this explanation does not explain why the effect of social capital only applies to black incarcerations rather than all incarcerations. If high social capital states were simply more punitive or more efficient at policing, we would expect higher incarcerations across the board, not only for African Americans.

Social Capital and Racial Diversity

As argued above, it is possible that social capital operates differently based on racial context. Indeed, this is a central part of Rodney Hero’s thesis: racial diversity conditions the effect of social capital. In the context of criminal justice, this may operate via several mechanisms. As discussed earlier, social capital can take the form of social trust whereby individuals develop norms of reciprocity and empathy. Alternatively, social capital can be linked to increased social controls—for example, neighborhood watch—where citizens aid law enforcement, thus reducing crime and potentially increasing the apprehension of criminals (and hence incarcerations).

Is it possible that social capital operates differently based on the racial context as Hypothesis 3 posits? The finding that social capital increases incarcerations for blacks but not whites suggests that there are differential effects of social capital by race. However, it is also possible that the effect depends on the broader racial context. If so, we might expect that in homogeneous (white) states, social capital takes the form of social trust rather than social control. Thus, we would not expect social capital to have as great a “crime-fighting” effect because individuals give each other the benefit of the doubt rather than acting as neighborhood watch patrols. As racial diversity increases, however, we would expect social capital to enhance social controls. Similarly, the relationship between social capital and incarceration rates may be contingent on the percent of the population that is black. Past work has found that the effects of state-level political variables on the black–white incarceration disparity are conditioned on the size of the black population (Yates and Fording 2005).

Table 4 presents the results from two interactive models that include a multiplicative term between social capital and racial diversity and percent of the population that is black, respectively. Here, we are theoretically interested in how racial diversity and black group size moderate the effect that social capital has on incarcerations. Figures 1 and 2 present this conditional relationship by plotting the marginal effects²⁴ of social capital on the dependent variable (black–white incarceration odds ratio) for different values of racial diversity and percent black population, respectively. The results suggest that in both cases, the effect social capital is conditional on racial context. In homogeneous states, social capital is not associated with increased incarcerations for blacks relative to whites. In fact, the effect of social capital is negative and even statistically significant when the percent of black population is below about 1% (Figure 2). However, as racial diversity and percent black population size increase, respectively, the effect of social capital on the black–white incarceration odds ratio also increases and becomes statistically significant. In Figure 1, we see that once diversity reaches about 0.12 (12th percentile), social capital is associated with a statistically significant increase in the black–white ratio. Thus, it is only in the most homogeneous states that social capital does not increase the disparity in incarcerations. Similarly, in states with moderate-sized black populations (above 3.75% or the 35th percentile), social capital has a positive effect on the black–white incarceration disparity. However, in states with small black populations (less than 1.25% or the 19th percentile), social capital is associated with lower black–white incarceration ratios. The average effect of social capital in states with significant black populations (above 7% or the 50th percentile) is nearly double the effect reported in the ratio model presented in Table 2.²⁵

Table 4.

Interactive Models: Social Capital, Racial Diversity, and Black Population.

	DV: ln(black–white prison odds ratio)
Social capital	0.018	−0.057**
Social capital	(0.031)	(0.024)
Racial diversity index	0.620***	0.557***
Racial diversity index	(0.218)	(0.209)
ln(% black)	−0.101***	−0.143***
ln(% black)	(0.022)	(0.022)
Social Capital × Racial Diversity	0.161	—
Social Capital × Racial Diversity	(0.106)	—
Social Capital × ln(% Black)	—	0.065***
Social Capital × ln(% Black)	—	(0.010)
Total violent crime rate	0.00002	−0.00003
Total violent crime rate	(0.00008)	(0.00008)
Total property crime rate	−0.00001	−0.000001
Total property crime rate	(0.00001)	(0.00001)
Government ideology	0.0004	0.0007*
Government ideology	(0.0004)	(0.0004)
% Democrats in legislature	0.001*	0.001
% Democrats in legislature	(0.0008)	(0.0008)
% females in legislature	0.001	0.0003
% females in legislature	(0.002)	(0.002)
% black legislators	−0.018***	−0.011***
% black legislators	(0.003)	(0.003)
Three-strikes law	0.149***	0.141***
Three-strikes law	(0.019)	(0.018)
Trial court clearance rate	0.0005**	0.0004*
Trial court clearance rate	(0.0002)	(0.0002)
Drug arrests/population	−6.851***	−6.695***
Drug arrests/population	(1.031)	(1.017)
State GSP per capita	−0.004*	−0.006***
State GSP per capita	(0.002)	(0.002)
Unemployment	−0.003	−0.005
Unemployment	(0.006)	(0.006)
Poverty rate	−0.018***	−0.017***
Poverty rate	(0.004)	(0.004)
Black–white poverty ratio	0.310***	0.305***
Black–white poverty ratio	(0.015)	(0.015)
% college degrees	−0.003	−0.0009
% college degrees	(0.004)	(0.004)
Black–white college degree	−0.681***	−0.660***
Black–white college degree	(0.098)	(0.096)
Divorce per 1,000 population	−0.058***	−0.055***
Divorce per 1,000 population	(0.007)	(0.006)
Felons per 100,000	−0.00003**	−0.00002*
Felons per 100,000	(0.00001)	(0.00001)
Constant	2.258***	2.339***
Constant	(0.163)	(0.164)
Observations	1,143	1,143
R ²	.565	.586
Number of states	48	48

Note. Panel corrected standard errors in parentheses. DV = dependent variables; GSP = gross state product.

p < .1. **p < .05. ***p < .01.

Figure 1.

Marginal effect of social capital on black–white odds ratio conditional on racial diversity.

Figure 2.

Marginal effect of social capital on black–white ratio conditional on percent black population.

This finding provides strong support for Hypothesis 3 and is consistent with the thesis that social capital may transform from social trust to social control as racial diversity/minority population increases. If racial diversity is the trigger for social capital taking on a social control structure (rather than social trust), then it is no surprise that the consequence is higher incarcerations for minorities. It is the racial context that shapes how social capital is used. States with higher levels of diversity are more likely to adopt systems and institutions of control when social capital is present.

Conclusion

Building on work in political science, sociology, and criminology, this article presents analysis examining the relationship between social capital, crime, and incarcerations across the U.S. states. Using longitudinal state-level data, this article finds that social capital is associated with higher incarceration rates, but only for African Americans. In addition, the relationship between social capital and incarcerations seems to be conditional on racial context. The effect of social capital on incarceration inequality is strongest in racially diverse states and those with significant African American populations. Social capital appears to exasperate the disparity in black–white incarcerations, but only in racially diverse states. This supports the notion that social capital strengthens racially targeted social controls as the racial context becomes more diverse. This suggests social capital is triggered, perhaps, by the same considerations that create racial threat.

There are several caveats concerning the analysis and results. Further work is needed to examine the within-state temporal dynamics as well as to further examine the interaction between social capital and diversity. Does this relationship apply to other racial/ethnic minorities? Are there threshold effects? Questions of reverse causation and reciprocal effects should also be addressed. Furthermore, future work should examine the policy-making side of criminal justice policy to explore the causal mechanism linking social capital to higher incarcerations, particularly for minorities. This may follow different processes at the local and state level. Jail-level incarcerations may reflect discriminatory practices in both citizens reporting crime and police discretion in making arrests. State prison population figures, however, are more likely a reflection of state laws, prosecutorial discretion, parole hearings, and attitudes of judges and juries. Future work should examine incarcerations at multiple levels. As discussed earlier, if social capital, trust, and empathy are transferred more easily within racial groups (as opposed to across racial groups), “norms of reciprocity” may disproportionately benefit white defendants in the decisions made by law enforcement, prosecutors, judges, and juries. That is, inequality does not inherently entail proactive discrimination against minorities; it can also occur from favorable treatment toward members of the majority. Future work—both theoretical and empirical—should explore this relationship.

As noted earlier, these findings only apply to the form of social capital articulated and operationalized by Putnam (2000). This is arguably Anglo, middle-class social capital rather than social capital maintained and developed in diverse communities and populations. These findings do not speak to the effect of other forms of social capital—particularly social capital developed in black communities—on incarcerations. In addition, it does not directly capture attitudinal aspects to social capital, including social trust. Future work should explore these aspects. Furthermore, it would be incorrect to conclude that minority communities do not benefit from the social connections and networks they develop. Rather, this analysis examines how the social capital conceptualized by Putnam affects incarcerations. Criticisms of his work on conceptualization of social capital (particularly in his work from Bowling Alone) also apply to this analysis. However, it is exactly this conceptualization of social capital this article is interested in testing given the numerous claims associated with it. Future research should also further explore the relationship between bonding and bridging social capital and how it relates to social trust versus social controls. The findings here suggest that bonding social capital may enhance social controls for member of the out-group. If in-group/out-group cohesion is based on racial criteria, then it may increase racial disparities. However, this is not directly tested because the measure of social capital used cannot distinguish between bonding and bridging social capital. Future work is needed to develop new measures of social capital that can make these distinctions.²⁶

Social capital offers promise to be catalyst for improving our communities and society. It offers to make us all better neighbors by reinforcing norms of reciprocity, by applying the Golden Rule: to “do to others as you would have them do to you.”²⁷ It holds the promise that if we strive to increase social cohesion and social capital, we will see a multitude of benefits to society. However, the reality appears to be more complex and, like many other areas of social life, dependent on context. The findings in the analysis suggest that racial context plays an important role in how social capital operates. This supports Hero’s (2007) contention that the benefits of social capital are not equally felt by all members of society. Minority groups, in particular, seem to be left out. The findings here, however, go further in finding that not only are African Americans worse off relative to whites, but social capital appears to increase black incarcerations in absolute terms as well. This supports a small but important literature that has explored the so-called “dark side” of social capital (Portes 1998; Putnam 2007). Future work should continue to explore this conditional relationship and further uncover the causal mechanisms of this relationship.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Daniel Hawes received his PhD at Texas A&M University and is currently an associate professor in the Department of Political Science at Kent State University. His research is interested in how social and government institutions can better address the inequalities in policy outcomes for disadvantaged groups.

References

Anderson

Alistair R.

Jack

Sarah L.

2010. “The Articulation of Social Capital in Entrepreneurial Networks: A Glue or a Lubricant?” Entrepreneurship & Regional Development 14 (3): 193–210.

Beck

Nathaniel

Katz

Jonathan N.

1995. “What to Do (and Not to Do) with Time-Series Cross-Section Data.” American Political Science Review 89 (3): 634–47.

Berry

William D.

Fording

Richard C.

Ringquist

Evan J.

Hanson

Russell L.

Klarner

Carl

. 2013. “A New Measure of State Government Ideology, and Evidence that Both the New Measure and an Old Measure Are Valid.” State Politics & Policy Quarterly 13:164–82.

Blau

Judith R.

Blau

Peter M.

1982. “The Cost of Inequality: Metropolitan Structure and Violent Crime.” American Sociological Review 47:114–29.

Brambor

Thomas

Clark

William Roberts

Golder

Matt

. 2006. “Understanding Interaction Models: Improving Empirical Analyses.” Political Analysis 14 (1): 63–82.

Bursik

Robert J.

Grasmick

Harold G.

1993. Neighborhoods and Crime: The Dimensions of Effective Community Control. New York: Lexington Books.

Burton

Velmer S.

Jr. Dunaway

R. Gregory

. 1994. “The Importance of Relative Deprivation Theory on Middle-Class Delinquency.” Pp. 79–97. In Varieties of Criminology: Readings from a Dynamic Discipline, ed. Barak

Gregg

. New York: Praeger.

Coleman

James S.

1988. “Social Capital in the Creation of Human Capital.” American Journal of Sociology. 94(S): S95–S120.

Compton

Mallory

Meier

Kenneth J.

2016. “The Specter of Social Capital: Bringing New Evidence to the Debate about Equity.” Presented at the 2016 MPSA Annual Conference, Chicago, April.

10.

Epp

Charles R.

Maynard-Moody

Steven

Haider-Markel

Donald

. 2014. Pulled Over: How Police Stops Define Race and Citizenship. Chicago: University of Chicago Press.

11.

Fossett

Mark

Kiecolt

K. Jill

. 1989. “The Relative Size of Minority Populations and White Racial Attitudes.” Social Science Quarterly 70 (4): 820–35.

12.

Giles

Michael W.

Buckner

Melanie A.

1993. “Beyond Racial Threat: Failure of an Old Hypothesis in the New South.” Journal of Politics 58 (4): 1171–80.

13.

Glaeser

Edward L.

1994. “The Formation of Social Capital.” Working Paper. https://search.oecd.org/edu/innovation-education/1824983.pdf

14.

Grootaert

Christiaan

. 1998. “Social Capital: The Missing Link?” Social Capital Initiative. Working Paper No. 3. The World Bank, Washington, DC.

15.

Hagan

John

. 1994. Crime and Disrepute. Thousand Oaks: Pine Forge Press.

16.

Harris-Perry

Melissa

. 2004. Barbershops, Bibles, and BET: Everyday Talk and Black Political Thought. Princeton: Princeton University Press.

17.

Hawes

Daniel P.

Rocha

Rene R.

2011. “Social Capital, Racial Diversity, and Equity: Evaluating the Determinants of Equity in the United States.” Political Research Quarterly 64:924–37.

18.

Hawes

Daniel P.

Rocha

Rene R.

Meier

Kenneth J.

2013. “Social Capital in the 50 States: Measuring State-Level Social Capital, 1986–2004.” State Politics & Policy Quarterly 13:121–38.

19.

Hero

Rodney E.

1998. Faces of Inequality: Social Diversity in American Politics. New York: Oxford University Press.

20.

Hero

Rodney E.

2003. “Multiple Theoretical Traditions in American Politics and Racial Policy Inequality.” Political Research Quarterly 56:401–408.

21.

Hero

Rodney E.

2007. Racial Diversity and Social Capital. New York: Cambridge University Press.

22.

Hero

Rodney E.

Tolbert

Caroline J.

1996. “A Racial/Ethnic Diversity Interpretation of Politics and Policy in the States of the U.S.” American Journal of Political Science 40:851–71.

23.

Hirschi

Travis

. 1969. Causes of Delinquency. Berkeley: University of California Press.

24.

Janowitz

Morris

. 1976. Social Control of the Welfare State. New York: Elsevier.

25.

Kawachi

Ichiro

Kennedy

Bruce P.

Wilkinson

Richard G.

1999. “Crime: Social Disorganization and Relative Deprivation.” Social Science & Medicine 48:719–31.

26.

Kennedy

Bruce P.

Kawachi

Ichiro

Prothrow-Stith

Deborah

Lochne

Kimberly

Gupta

Vanita

. 1998. “Social Capital, Income Inequality, and Firearm Violent Crime.” Social Science & Medicine 1:7–17.

27.

Land

Kenneth C.

McCall

Patricia L.

Cohen

Lawrence E.

1990. “Structural Covariates of Homicides Rates: Are There Any Invariances across Time and Social Space?” American Journal of Sociology 95:922–63.

28.

Merton

Robert K.

1938. “Social Structure and Anomie.” American Sociological Review 3:672–82.

29.

Merton

Robert K.

1968. Social Theory and Social Structure. New York: Free Press.

30.

Nguyen

Minh

Nguyen

Hoa.

2010. “Stata Module: Estimates of System of Regression Equations with Unbalanced Panel Data and Random Effects.” Working Paper. http://fmwww.bc.edu/RePEc/bocode/x/xtsur.html

31.

Oliver

J. Eric

Wong

Janelle

. 2003. “Intergroup Prejudice in Multiethnic Settings.” American Journal of Political Science 47 (4): 567–82.

32.

Pattillo

Mary

Weiman

David

Western

Bruce

. 2004. Imprisoning America: The Social Effects of Mass Incarceration. New York: Russell Sage Foundation.

33.

Portes

Alejandro

. 1998. “Social Capital: Its Origins and Applications in Modern Sociology.” Annual Review of Sociology 24:1–24.

34.

Putnam

Robert D.

2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster.

35.

Putnam

Robert D.

2007. “E Pluribus Unum: Diversity and Community in the Twenty-First Century.” Scandinavian Political Studies 30:137–74.

36.

Rocha

Rene R.

Espino

Rodolfo

. 2009. “Racial Threat, Residential Segregation, and the Policy Attitudes of Anglos.” Political Research Quarterly 62:415–26.

37.

Rose

Dina R.

Clear

Todd R.

1998. “Incarceration, Social Capital, and Crime: Implications for Disorganization Theory.” Criminology 36 (3): 441–80.

38.

Rosenfeld

Richard

Messner

Steven F.

Baumer

Eric P.

2001. “Social Capital and Homicide.” Social Forces 80 (1): 283–310.

39.

Sampson

Robert J.

Groves

W. Byron

1989. “Community Structure and Crime: Testing Social Disorganization Theory.” American Journal of Sociology 94:774–802.

40.

Sampson

Robert J.

Julius Wilson

William.

1995. “Toward a Theory of Race, Crime, and Urban Inequality.” Pp. 37–54. In Crime and Inequality, eds. Hagan

John

Peterson

Ruth D.

Stanford: Stanford University Press.

41.

Scheve

Kenneth

Stasavage

David

. 2006. “Religion and Preferences for Social Insurance.” Quarterly Journal of Political Science 1:255–86.

42.

Shaw

Clifford R.

McKay

Henry D.

1942. Juvenile Delinquency and Urban Areas. Chicago: University of Chicago Press.

43.

Smith

Kevin B.

2004. “The Politics of Punishment: Evaluating Political Explanations of Incarceration Rates.” Journal of Politics 66 (3): 925–38.

44.

Sorens

Jason

Muedini

Fait

Ruger

William.

2008. “State and Local Public Politices in 2006: A New Database.” State Politics and Policy Quarterly 8(3): 309–26.

45.

Useem

Bert

Piehl

Anne Morrison

. 2008. Prison State: The Challenges of Mass Incarceration. New York: Cambridge University Press.

46.

Western

Bruce

. 2006. Punishment and Inequality in America. New York: Russell Sage Foundation.

47.

Yates

Jeff

Fording

Richard

. 2005. “Politics and State Punitiveness in Black and White.” Journal of Politics 67 (4): 1099–121.

48.

Zhu

Ling

. 2013. “Panel Data Analysis in Public Administration: Substantive and Statistical Considerations.” Journal of Public Administration Research and Theory 23 (2): 395–428.