Over the past three decades, U.S. criminal justice policy has become increasingly punitive. Developments in law enforcement, prosecution, adjudication and sentencing, and post-conviction sanctions have tended to the imposition of stricter regimes on those convicted or suspected of crime (Reference Sabol and HarrisonSabol & Harrison 2007; Reference ChristieChristie 2000). Garland has labeled this unprecedented expansion of the criminal justice apparatus the “punitive turn” (Reference GarlandGarland 2001). He offers a compelling account for this turn, attributing it to historical changes in the way crime has been represented, its increasing role in partisan politics, and more fundamental social and economic shifts (Reference GarlandGarland 2001). Garland's thesis is intellectually parallel to earlier work by Reference ScheingoldScheingold (1984). It intersects with historical inquiries into the accelerants of the punitive turn, such as popular media (Reference Beckett and SassonBeckett & Sasson 2004), as well as contemporary accounts of the infiltration of crime imagery and control mechanisms into everyday institutions (Reference SimonSimon 2007).
This growing body of scholarship has clarified theory about changes in crime policy and their consequences, but it has not yet been the subject of thorough empirical investigation. While this theoretical perspective emerged primarily as an explanation for changes in punishment practices, particularly the meteoric increase in incarceration rates, its scope is considerably broader. It should apply as well to policies that are not expressly connected to punishment or control, particularly those that define the parameters of due process in criminal proceedings. Further, because the political conditions that foreshadowed the punitive turn did not equally obtain in all states, the study of state politics and policy might provide an opportunity to begin to test the theory. In this article, we develop and test hypotheses regarding state-level variation in indigent defense, a criminal justice policy the primary purpose of which is to protect individual rights. We seek to contribute empirical knowledge about how such systems vary and the causes of such variation. Additionally, and more fundamentally, exploring the conditions that produce different indigent defense systems offers an opportunity to add to our more general knowledge about the causes of criminal justice policy variation. Existing scholarship on criminal justice policy has implications for due process as well as crime control policies, and we take it as our point of departure in attempting to explain variability in states' indigent defense programs. In the following sections, we (1) briefly describe the historical background and contemporary contours of indigent defense systems in the states; (2) place these developments in the context of the past three decades of increasingly punitive crime policy; (3) extend theoretical propositions on the causes of the punitive turn to state variations in due process policies; and (4) test the hypotheses derived from these propositions.
Due Process Rights and the Dimensions of Indigent Defense Policy
The assistance of counsel in court is a signature right in the contemporary criminal adjudication system, albeit one that is meaningful only when representation is readily available. Following the Supreme Court's decisions in Powell v. Alabama (1932), Gideon v. Wainwright (1963), and Argersinger v. Hamlin (1972), states have been obliged to provide counsel to all indigent defendants facing the possibility of incarceration. Recent estimates suggest that the vast majority of people accused of crimes in the United States—more than 80 percent of accused felons—are deemed indigent for purposes of assigning publicly funded legal counsel (Bureau of Justice Statistics 2001). However, the Court has left the implementation of this mandate up to states and communities. As a result, indigent defense services vary in terms of program structure, division of responsibility between state and local governments for underwriting program costs, and funding levels.
Program Types
Indigent defense programs can be divided into three organizational types. First, assigned counsel programs, typically administered by local court staff, involve assignment of indigent clients to private practitioners on a rotating basis. Second, public defender programs entail “[a] salaried staff or full-time or part-time attorneys render[ing] defense services through a public or private non-profit organization or as direct government employees” (Reference De Frances and LitrasDe Frances & Litras 2000:2). Although the term public defender is commonly used to describe all attorneys working in indigent defense, we adopt the definitions employed by the 1999 National Survey of Indigent Defense Systems and therefore distinguish these salaried employees from those retained under other arrangements. Third, contract programs entail financial agreements between governments and individual attorneys, law firms, consortiums, or bar associations to provide representation, typically for a specified number of cases or for a stipulated duration of time.
The earliest accommodations for indigent defendants were ad hoc assigned counsel systems where counsel was typically assigned to represent a defendant pro bono. Following Gideon, and the mandate for representation in felony cases for indigents, some (though not all) of these arrangements were augmented or replaced by more structured delivery systems. Advocates argued that public defender programs brought camaraderie and professionalism to the practice of defense work by guaranteeing oversight, facilitating the sharing of expertise, and helping to foster an ideological commitment to defense work (Reference McInytreMcIntyre 1987; Reference WeissWeiss 2004; Reference EmmelmanEmmelman 1993; Reference SnellenbergSnellenberg 1985). Others argued that organized public defender offices could effectively advocate for defendants' rights and needs through policy litigation (Reference BohneBohne 1978; and see Reference Sarat and ScheingoldSarat & Scheingold 1998). Assigned counsel and contract models were criticized for failing to guarantee that advocates would have either the expertise or the financial incentive to mount a vigorous defense (Reference WordenWorden 1991, Reference Worden1993). By 2002, 20 states had mandated the establishment of public defender offices in all counties.Footnote 1
Centralization of Funding
States not only choose whether or not to prescribe a specific indigent defense program type, but they also decide the extent to which they share responsibility for funding these programs with local governments. The balance of responsibility for funding indigent defense varies considerably across the states. In 2002, 23 states chose to fund fully all their indigent defense programs. Thirteen states provided less than 25 percent of the costs, relying on county governments to carry the lion's share of the burden. Two states, Utah and Pennsylvania, did not contribute any resources to indigent defense from the state budget, instead obliging counties to fund these programs entirely themselves (Spangenberg 2003:36).
Advocates argue that responsibility for funding should be taken out of the hands of localities for several reasons (National Legal Aid and Defender Association 1976; American Bar Association 1978). At the community level, local funding processes are politically problematic since defenders must compete head-to-head for resources with their more politically popular opposition, as well as with other court services such as victims' programs and specialized courts. Further, communities face risks. One unexpected prolonged trial, or a capital case, can wipe out an entire year's projected indigent defense budget (Reference DieterDieter 1994). Moreover, wealthier communities can afford to spend more on indigent defense (Reference Worden and WordenWorden & Worden 1989), raising equity issues across communities. Hence, some maintain, the states should take on responsibility for funding indigent defense.
Program Expenditures
Finally, expenditures for indigent defense vary significantly across states. Counting expenditures at all levels of government, in 2000 they ranged from slightly less than $2,000,000 to nearly $500,000,000, with a median of $30,000,000—figures that of course reflect variation in states' populations. A more useful perspective comes from comparing expenditures per capita, which is the metric we examine in this article. Overall, the median per capita state-level expenditures for indigent defense is about $6 (ranging from $3 to $37), compared with median figures of $118 for per capita welfare expenditures, and $780 for per capita corrections expenditures (U.S. Census Bureau 2000). When local expenditures are added to state outlays, the median per capita expenditure is $9 (Spangenberg 2003:36, U.S. Census Bureau 2000).
Indigent defense programs are widely regarded as severely underfunded (President's Commission 1967; see also Reference MoranMoran 1982; Harvard Law Review 2005). State laws stipulating reimbursement rates cast them far below market rates for privately retained criminal defense work; budgets for public defenders typically lag far behind those of prosecutors' offices. Estimates of actual expenditures per case are not comparable to the fees charged by private attorneys. Some states prescribe reasonable payment schedules in legislation but fail to appropriate enough money to cover real costs. Contracting programs run the risk of low-balling budgetary authorities, but then short-shrifting indigent clients to cover the shortfall. Just as a close look at actual welfare benefits tends to convince one that welfare hardly brings people out of poverty, a close look at expenditures on indigent defense suggests that the services provided to poor defendants cannot be consistently comparable with what most Americans would want or pay for if they were accused of crimes. Hence a better system would be one that enjoyed significantly higher expenditures than is now currently the case in most states.
In summary, over the past three decades, legal services for the indigent became institutionalized as a state responsibility, but states were left to their own devices in designing programs and funding plans to fulfill that constitutional mandate. As we document below, states interpreted this mandate in diverse ways, and the era during which these adaptations took place was one that marked tumult and change in the larger arena of crime policy. We now move to a discussion of theoretical perspectives on the conditions that might influence state policy choices when formulating and developing such policies, in the context of these contemporary shifts in criminal justice.
Theoretical Perspectives on the Provision of Indigent Defense
Scholarship on recent developments in criminal justice has described a punitive turn in policy approaches that does not appear wholly explicable by changes in the level of crime itself (Reference GarlandGarland 2001; see also Reference ScheingoldScheingold 1984; Reference BeckettBeckett 1997; Reference TonryTonry 2004). Garland proposes that, beginning in the 1960s, U.S. society became destabilized by seismic changes in economic and social forces: the unraveling of affluent postwar capitalism, social disorganization, income disparity, rising consumer expectations, and an increasingly visible welfare class (2001: Ch. 4). These shifts were accompanied by an increase in crime rates, fueled by the maturation of the postwar baby boomers. Other scholars have built upon this thesis and added insight into both the contributing factors, and the consequences, of these social changes. Tonry writes that “A succession of upsetting incidents has produced a series of moral panics that, among other things, has led to artificially heightened anxieties and fierce overreactions. Current policies are a predictable and understandable, but regrettable, result” (2004:25). Recently, Simon has argued that declines in public confidence in professional policy experts, alongside increasing anxiety about crime, not only opened the door to politicians' manipulation of crime as a campaign issue, but also, more insidiously, created a rationale for increasing executive authority, infiltration of crime control techniques into social institutions such as schools and workplaces, and public acquiescence to this state of affairs as the price to be paid for public safety (Reference SimonSimon 2007).
Adding to this picture, Reference Beckett and SassonBeckett and Sasson (2004) observe that the seemingly quiet 1950s presaged tremendous changes in Southern politics that rippled across the nation, as the civil rights movement challenged whites' views about race relations as well as the resettled post-Reconstruction politics of the early twentieth century. They observed that the civil rights movement galvanized a regional, and eventually national, resistance to liberal changes in the status quo that found voice in national election campaigns in the 1960s. The “war on poverty” declared by President Lyndon Johnson raised awareness about the pervasiveness of poverty, but it also educated the working- and middle-class public about the existence of the large welfare population. Faced with the intractable problems of race relations, poverty, and economic dislocations of the working class, politicians—especially Republican, conservative politicians—capitalized on the opportunity to attract voters by attributing the rising crime rates of the 1970s to these broader social changes. In short, by the 1970s, politicians had learned to recode the problems of poverty, welfare dependency, racial tension, and working-class anxiety into the problem of crime and disorder.
There is growing evidence at the national level that escalating punitiveness in criminal justice policy tracked with these political shifts. There is also evidence that the shifts of greatest interest to scholars of the punitive turn—dimensions such as racial intolerance, hostility to the disadvantaged, and political conservatism—have been associated with punitive policies at the state level. While many of the scholars studying both national and state policy shifts have focused on punishment, they have observed that due process policies were also at risk during this era (Reference TonryTonry 2004:136; Reference ScheingoldScheingold 1984; Reference GarlandGarland 2001; Reference Beckett and SassonBeckett & Sasson 2004). Indigent defense is not a punitive policy as such, but researchers, practitioners, and policy makers have recognized that decisions about the right to counsel can be highly politicized. Studies of public defenders (and defense counsel more generally) have chronicled ideological differences underlying the roles of defense and prosecution (e.g., Reference McInytreMcIntyre 1987; Reference WeissWeiss 2004; Reference Eisenstein and JacobEisenstein & Jacob 1977; Reference EmmelmanEmmelman 1993; Reference OgletreeOgletree 1993). Calls for improved defense services are justified by dual appeals to due process standards and compassion for the disadvantaged (American Bar Association 2004; Reference MoranMoran 1982; Harvard Law Review 2005). Following the literature of the punitive turn, then, we develop hypotheses regarding the relationship between indigent defense policy and three dimensions of states' political climate: Republican politics, racial heterogeneity and tolerance, and welfare policy climate.
Statehouse Politics: Conservative Republicanism
Among political scientists, it has long been axiomatic that party politics, at the state level, shapes policy outcomes (Reference KeyKey 1964; Reference BarrilleauxBarrilleaux 1997). Two theses dominate this literature. First, theorists suggest that electoral competition—the margin of risk or safety experienced by individual politicians, or parties—shapes the propensity to adopt policies that respond to the needs or wants of mainstream voters (Reference Holbrook and Van DunkHolbrook & Van Dunk 1993). Specifically, this thesis suggests that where margins of safety are very thin, elected officials accede to the imputed preferences of working-class and lower-class voters. ‘Close races are argued to require candidates to expand their voting blocs by appealing to members of socio-economic groups that do not traditionally take part in politics.’ The corollary to this thesis is that tax policies, welfare policies, and the like will favor the interests of lower-income constituents where neither party can count on large electoral margins. It is difficult to extrapolate this thesis neatly to criminal justice policy, particularly rights policy, however. Criminal defendants, and crime victims, are disproportionately lower-income people; yet the larger public identifies emotionally with the suffering of the latter, and is skeptical of the rights of the former.
The second, more directly useful thesis from this literature focuses on the ideological positions of the two major political parties and their ability to dominate policymaking bodies. At the national level, the Democratic Party has long been identified as the more liberal party, favoring the interests and needs of the disadvantaged, and the rights of the overlooked and repressed. The Republican Party has positioned itself, since the 1970s, as the party of the “silent majority,” the working- and middle-class voters who hold to traditional values about family structure, race relations, and social order. The economic dislocations and the social upheavals of the 1960s and 1970s confronted the Republican Party with the challenge of attracting and keeping voters who were unlikely to benefit from that party's economic platforms, but also with the opportunity to capitalize on working- and middle-class fears and anxieties about changing social conditions. The solution they offered was a package of social policy reforms that promised to reify traditional values about family structure and social hierarchie, and to crack down on drugs and crime (Reference Jacobs and HelmsJacobs & Helms 1996; Reference ScheingoldScheingold 1991). This message sold well at the national level, particularly in presidential campaigns. It was not long before it was adapted to state-level politics as well.
As a result, by the 1980s and certainly by the 1990s, the Republican Party had successfully polarized not only debates on social issues (such as abortion and legal tolerance of homosexuality), but also debates on crime, with the result of escalating rhetoric and policy initiatives aimed at punishing criminals and protecting the public. At the national level, of course, Republican Party domination of policy resulted in criminal justice policies that emphasized expansion of criminal codes, increased criminal sanctions, and institutionalized support for state-level initiatives that increased the scope of criminal law. At the state level, the politics of conservative Republicanism have been less thoroughly documented, although there is evidence that, over time, Republican political strength has tracked with higher incarceration rates (Reference Beckett and WesternBeckett & Western 2001; Reference Greenberg and WestGreenberg & West 2001; Reference Jacobs and HelmsJacobs & Helms 1996; and see Reference Jacobs and CarmichaelJacobs & Carmichael 2001). We hypothesize that where Republicans dominate the governor's office and the statehouse, defendant protections such as indigent defense will be a low priority. Furthermore, the New Federalism of the Reagan administration emphasized the devolution of policy responsibilities, which would suggest that social functions such as indigent defense would be left in the hands of local governments, rather than aggregated to state agencies.
The Racial Threat Hypothesis
There is ample evidence in both welfare and criminal justice research that race and ethnicity issues are entangled in public values and policy in these areas. Researchers theorize that states with larger minority populations are less generous in recognizing the needs and rights of minorities, and this spills over into policies that disproportionately affect minority groups, such as welfare policy (Reference GilensGilens 1999; Reference SossSoss et al. 2001). Specifically, to the extent that significant nonwhite populations pose a threat to the securities and sensibilities of whites, researchers predict that legislators (state as well as local) who allocate funding for programs largely associated with minority populations will adjust their budgets to accord with local preferences.
Race is arguably even more closely linked to criminal behavior than to welfare eligibility in the (white) public mind. Researchers refer to this linkage as the “racial threat hypothesis,” which states that majority populations suffer in general from a fear of minority populations. While the level of this fear is hypothesized to be variable, the result is that where minority populations are larger, whites' fears of minorities crystallize around punitive attitudes and sentencing practices (Reference Myers and TalaricoMyers & Talarico 1986; Reference Ulmer and JohnsonUlmer & Johnson 2004; Reference Kramer and UlmerKramer & Ulmer 2002; Reference Sorenson and StemenSorenson & Stemen 2004). There is empirical evidence to support this prediction (Reference Jacobs and HelmsJacob & Helms 1996; Reference Beckett and WesternBeckett & Western 2001; Reference JacksonJackson 1989).
Most tests of this hypothesis define racial threat simply in terms of the proportion of minority (typically African American) population, positing that up to a majoritarian tipping point, there is a positive correlation between proportions of racial minorities and repressive policies (e.g., Reference Jacobs and TopeJacobs & Tope 2007; Reference Jacobs and CarmichaelJacobs & Carmichael 2001; Reference Greenberg and WestGreenberg & West 2001; see also Reference StuckyStucky et al. 2005; Reference McGarrell and DuffeeMcGarrell & Duffee 1995, Reference McGarrell, Duffee, Duffee and Maguire2007; Reference SossSoss et al. 2001). It has been expressed in more subtle ways, however. Jacobs and Helms locate the threat in economic disparities between whites and blacks, positing that the greater the divide in income, the greater the “potential dangers presented by a predatory minority underclass” (1996:325). Percival measures racial threat not only as states' racial heterogeneity, but also more directly, as state residents' racial attitudes (Reference PercivalPercival 2007; and see Reference JohnsonJohnson 2001). We follow his lead here, and hypothesize both that states with larger minority populations will fail to pursue progressive funding of indigent defense, and that states with lower levels of public tolerance for racial equality and integration will do the same.
Welfare Climate
Because providing legal representation involves establishing and funding programs that rely on public resources to benefit low-income populations, it can be characterized as a redistributive policy as well as a criminal justice one. Many criminal justice policies share this characterization: community rehabilitation programs, specialized courts that mandate treatment, and prison education are examples of policy choices that supply services (not merely supervision and punishment) to eligible (if not always enthusiastic or willing) populations (Reference DuffeeDuffee 1990; Reference GuetzkowGuetzkow 2004; Reference PercivalPercival 2007). Indigent defense further resembles public welfare insofar as in both instances the federal government has established entitlement to services, but has left many of the details of program eligibility, delivery, and funding to states and localities. To understand how and why states have settled on different indigent defense policies, therefore, we draw on theories that have proven useful in accounting for both crime and justice policies, as well as welfare policies.
While the nexus of welfare policy and crime policy, broadly defined, has been the focus of theorists for 25 years (for example, see Reference DuffeeDuffee 1990), this linkage has recently become the subject of empirical research as well. Most of this research examines relations between welfare and correctional policy, measured in terms of both expenditures and population affected (Reference Beckett and WesternBeckett & Western 2001; Reference Greenberg and WestGreenberg & West 2001; Reference GuetzkowGuetzkow 2004; Reference StuckyStucky et al. 2007). Less attention has been given to whether or not social services and entitlement programs for defendants and offenders (such as indigent defense, rehabilitative programs, job training, and the like) track with more conventional welfare programs for the poor. We hypothesize that the “welfare climate” in a state will be related to indigent defense provision.
Students of public policy have long argued that welfare policies are not value-neutral, but, rather, that they construct rules about who deserves public assistance, and not simply what constitutes economic need (Reference Barrilleaux and BernickBarrilleaux & Bernick 2003; Reference Beckett and WesternBeckett & Western 2001). Through the establishment of explicit and implicit requirements about what one must do (or refrain from doing) to get benefits, researchers have argued, welfare policies reflect the normative values of policy makers and constituents.
During the past decade, states have had the opportunity (or mandate) to reconstruct their welfare policies under federal statutes establishing the Temporary Assistance for Needy Families program (TANF). As a result, contemporary state welfare rules offer a useful portrait of states' deliberate choices about welfare eligibility and services. A state's welfare climate can be measured in terms of the level of inclusiveness and generosity created by its policies, including policies on eligibility, maximum benefit levels, expenditures by taxpayers, and stipulations for penalizing families for noncompliance with program rules (such as work requirements).Footnote 2 A more accommodating welfare climate not only supports recipients at more generous levels, but also casts its eligibility net more widely, supporting more “marginal” clients, not merely those for whom the public feels sympathy, such as children, the disabled, and the elderly.
How might welfare climate shape indigent defense policy? We approach this question in an exploratory fashion. We hypothesize that states with more generous welfare climates will be more likely to subsidize local indigent defense programs, inasmuch as those states have assumed greater responsibility for disadvantaged populations. In short, states with generous welfare climates may choose to use discretionary resources to enhance local programs, while more-restrictive states may withhold such support. We hypothesize that a more accommodating welfare climate is also one that will provide more support for providing representation for people accused of crimes: states that are sparing with welfare benefits in general are unlikely to fund legal services for defendants at anything more than minimal levels.
Reference PetersonPeterson's classic (1981) study of federalism and redistributive policymaking suggests an additional possible linkage between state subsidies and overall expenditures. Peterson posited that local governments must compete with each other for resources as well as for taxpaying residents, visitors, and businesses; on the other side of the ledger, expenditures for redistributive policies subsidize residents who may contribute little to the community's economic health. State governments, while not immune to these competitive pressures, do not experience them as directly. As a result, Peterson hypothesized that where state governments take on a larger share of responsibility for such programs, overall expenditures will be higher (Reference Worden and WordenWorden & Worden 1989). Therefore, we predict that, all else equal, states that take on a greater proportion of indigent defense costs will also have higher expenditures for those programs.
Resources and Demand
We have been guided by the theoretical perspectives outlined by Garland and other scholars of the punitive turn as we have outlined our expectations for state-level variation in indigent defense policy. However, patterns in indigent defense policy may be affected by economic factors as well as political ones. Indeed, it would be surprising to discover that states' economic capacity, and demand for legal services, do not shape expenditure patterns. Since states are obliged to offer defense services to most defendants, demand levels will be determined by the number of poor persons accused of crimes, conditioned by local standards for determining indigence. There exist no reliable contemporary data on caseloads at the state level, but we suggest that indigent caseload is a function, at least in part, of crime and poverty rates. As either increases, the number of defendants and their rate of indigence will increase respectively, creating increased burdens upon the system. On the other hand, states maintain discretion over the generosity with which they fund indigent defense services. Some welfare scholars theorize that higher caseloads lead states to put brakes on welfare expenditures to reduce dependent populations (see Reference SossSoss et al. 2001).
States' spending may also reflect levels of available resources. Evidence exists to show that states with greater resources tend to spend more on redistributive enterprises, while poorer states minimize costs (Reference TweedieTweedie 1994; Reference Koven and MausoloffKoven & Mausoloff 2002). This might be especially true for programs such as indigent defense, which serve clienteles that do not garner public sympathy, and which permit officials considerable discretion in controlling the costs of the service.
These economic considerations may be relevant to several features of indigent defense policy. Reference Worden and WordenWorden and Worden (1989) found that county size in Georgia was related to the choice of delivery system for indigent defense services. Larger counties were more likely to establish public defender offices, leading the authors to speculate that institutionalized defender arrangements offered economies of scale. Regarding funding responsibility, states with higher demand for public services may fund indigent defense services centrally in order to avoid potential inequities across localities, though such generosity is likely to be tempered where budgets are limited (Reference SossSoss et al. 2001; Reference Kelleher and YackeeKelleher & Yackee 2004). Total funding levels are likely to be affected most directly, with the volume of need for representation setting something of a lower boundary to funding levels, while available resources establish a ceiling.
We do not expect these economic limitations to be fully determinative of state policy choices, however. Since eligibility standards for indigent defense services are not uniform across the nation (Reference MoranMoran 1982; Spangenberg Group 2000), and case processing costs and levels of compensation also differ, we do not expect levels of demand for these services to determine expenditures entirely, as would be expected if representation were a commodity that could be standardized, priced, and fixed to easily verified criteria. Nevertheless, they are important to control in order to avoid the possibility of unmeasured heterogeneity within the sample.
Interrelationships Among Indigent Defense Measures
We examine relationships among program structure, state subsidy, and overall funding dimensions of indigent defense policy. These three variables have an implicit temporal ordering. States must first decide how to deliver indigent defense services—either by leaving matters to local authorities or by positing a standard model for localities to implement—before deciding whether to subsidize programs centrally. Having mandated a service delivery system, however, legislatures may face higher expectations from local governments to fund their mandates. Then, decisions to subsidize indigent defense may then have implications for overall expenditures. Although we do not posit strong causal relationships among these variables (except in the case of the Reference PetersonPeterson (1981) hypothesis noted above), the temporal ordering of the variables itself would suggest that variables occurring earlier in the model ought to be controlled when attempting to predict those at later stages. Thus, when predicting the percentage of funding contributed by the state, we control for the decision to mandate public defender programs, and when predicting the overall level of system funding, we control for both prior state decisions on system type and the level of state commitment to funding the system.
Bivariate analysis suggests that these three variables are somewhat interrelated. States that mandate the establishment of public defender programs (regardless of the variability in program features that term includes) tend to invest more state than local money in programs: on average, states that dictate this program type fund more than 80 percent of expenditures, while those that leave program type up to local authorities, or permit a mix of program types, fund less than 50 percent.Footnote 3 There is only a very weak relationship between established program type and overall total expenditures. However, states in which legislatures shoulder greater responsibility for funding indigent defense provide more resources on average than those that leave funding up to county governments.
Data, Measures, and Analysis Plan
Dependent Variables
Data on indigent defense program types in use across states were obtained from published surveys of indigent defense programs.Footnote 4 We created measures of two state-level variables describing indigent defense funding: (1) percentage of costs borne by state rather than local governments, and (2) total expenditures on indigent defense per capita. Table 1 summarizes the distributions on these variables.
Table 1. Descriptive Statistics
Data on the proportion of indigent defense expenditures funded by state governments were gathered from reports produced by the Spangenberg Group, a consulting firm that has contracted with the American Bar Foundation and a variety of other organizations in the evaluation of indigent defense systems (Spangenberg 2003, 2004). These data include information on the total dollar amounts spent on indigent defense in each state from all sources, from which we were able to construct measures of both the proportion of such funding shouldered by state government, and the total expenditures in each state for these services. The measure of total expenditures was divided by the population of each state to control for differences in population size.Footnote 5
Independent Variables
Program Type
Data on program characteristics in each state were gathered from a variety of sources including program Web sites, state defender associations, and the Spangenberg reports. These data permitted us to determine whether or not the state had formally established a public defender system as the primary or sole mechanism for providing indigent defense in 2002. Twenty states had established statewide public defender systems; the others rely on a mix of programs, determined largely at the local level.
Welfare Climate
Welfare climate is defined here as the receptivity of a state to the needs of disadvantaged populations; this is significant for our purposes insofar as indigent defendants are, by definition and by class, a disadvantaged group placing claims on court services. As stated earlier, in 1996, all states “reset” their welfare policies under federal requirements, creating an unusual opportunity to measure purposive (rather than incremental) decisions about policy. Welfare climate is measured as a factor scale of four items: the rate of welfare assistance among the state's population of households, welfare expenditures per capita, the maximum benefit level allowed for welfare families of one adult and two children, and a three-point scale capturing severity of sanctions of noncompliance with TANF rules (General Accounting Office 2000).Footnote 6 Higher values on this scale represent more inclusive, more generous, and less punitive welfare climates.
Racial Threat
We use three measures to test the racial threat hypothesis. First, consistent with most previous research, we measure states' racial makeup as the percentage of the population that self-reports a racial identity as black or African American, according to 2000 U.S. Census data. Previous research suggests that to the extent the racial threat hypothesis garners empirical support in explaining criminal justice and welfare policy, it is the presence of African Americans, not all minorities, that matters. Because some research suggests that the effect of race may be curvilinear—that racial threat exerts an increasing effect on policy outcomes as the minority group moves toward 50 percent—we also include a quadratic term (Reference Jackson and CarrollJackson & Carroll 1981; Reference StuckyStucky et al. 2007).Footnote 7
Second, we recognize that the racial threat hypothesis, while typically tested with a measure of racial heterogeneity, is in fact a hypothesis about the attitudes and fears. There are no readily available contemporary state-level measures of racial anxiety. However, two sets of state-level opinion measures have been developed by Reference NorranderNorrander (2000) and Reference BraceBrace and colleagues (2002) that could provide components for constructing such a measure. These items, extracted respectively from the National Elections Studies (NES) and the General Social Survey (GSS), were aggregated over a series of years from 1974 to 1998. The authors of these studies observed that the measures were “remarkably stable” over time (Reference BraceBrace 2002:181), and more powerfully predicted policy outcomes at the state level than did more general measures of ideology. We draw upon their methodology to construct a measure of public opinion, which we label “illiberality,” and which we hypothesize is negatively associated with progressive indigent defense policies.
In order to test this hypothesis, we constructed a measure of illiberality by factor analyzing 10 items that were plausibly related to views on race, equality, tolerance, crime, and justice. (Descriptions of these items are included in the Appendix.) Five of these—items about racial equality, affirmative action, tolerance of diverse viewpoints, rating on a liberal-conservative scale, and conventional religiosity—produced a single strong scale that we employ as our measure of illiberality.Footnote 8 Together, these items produce an alpha of 0.77. Higher scores on this measure indicate more conservative values.Footnote 9
Party Politics
We include two measures of statehouse party politics. The first is a dummy variable, coded “1” if the governor in office in 2002 was a Republican. The second is a measure of party control of the legislature, coded “2” if Republicans controlled both houses; “1” if each party controlled one house, or party strength was at or very near 50 percent in both houses, or the state (Nebraska) had a nonpartisan, unicameral legislature; and “0” if Democrats controlled both houses. These two variables are not empirically related: the party of the governor offers no predictive power about control of the legislature, and hence we include both variables in the analyses.
Economic Capacity
We operationalize state economic capacity as the quantity of resources available for state programs and policies, adjusted by population. Some state policy researchers have relied on proxy measures of citizen wealth to capture capacity, such as median household income (Reference Koven and MausoloffKoven & Mausoloff 2002) or expenditures (D. Reference MillerMiller 1991). However, states and localities are obliged almost without exception to produce balanced budgets (National Conference of State Legislatures 1999), which means that governments at those levels can spend only what they can take in. Therefore, although governments are obliged to provide legal services to indigents, they may (and, we predict, do) adjust their spending depending upon the amount of money available. From a theoretical perspective, therefore, expenditures in any policy domain or program area (adjusted for state size or population) are a more meaningful measure of a state's commitment to that policy in the context of the funds available for all policy areas (Reference McGarrell, Duffee, Duffee and MaguireMcGarrell & Duffee 2007; Reference GuetzkowGuetzkow 2004). Further, research in welfare policy has found that expenditures are more closely related to state wealth than to measures of citizen wealth such as the cost of living or median income (Reference TweedieTweedie 1994). We measure state capacity as total taxable revenue per capita in thousands of dollars (U.S. Census Bureau 2000). Although lawmakers can dig deeper into taxpayers' pockets if they feel the need, in reality a policy issue such as indigent defense is likely to be defined in terms of available resources.
Demand for Indigent Defense
No state-level data exist on indigent defense caseload. However, given that judges, court administrators, and policy makers can in theory adjust eligibility standards (as well as reimbursement practices) to match budgetary limits, it is not at all clear that caseload would be the best measure for demand anyway. We use two measures to estimate need or demand for indigent defense. The first is an index of poverty, based on the premise that states with larger impoverished populations face more claims for indigent defense. This measure is a factor score scale of three items: poverty rate, unemployment rate, and an estimate of the state's unserved poverty population (the difference between poverty rate and public assistance rates, in 2000).Footnote 10 The second measure is the 2000 UCR rate of property and violent crime per 1,000 residents. These two variables are correlated, at the state level, at 0.340, suggesting that while they are related, they make separate contributions to indigent defense caseload.
Analysis Plan
Table 1 reports descriptive statistics on all these variables, including, where appropriate, the component variables of scales and indexes.
We examined the data for collinearity problems among the independent variables, and for the presence of possible outliers. Regarding collinearity, we analyzed series of models for each dependent variable, allowing us to inspect the stability of coefficients as potentially confounding variables were added to the model. First, the model was estimated using only economic (supply and demand) variables; second, the political climate variables were added (welfare climate, proportion African American, public conservatism); third, the measures of elite party politics were added (Republican control of the governor's office and the statehouse); and finally, where appropriate, the measures of prior indigent defense policy decisions were added (Tables 2 and 3). We found little evidence that intercorrelation resulted in unstable coefficients. We also ran all analyses with collinearity diagnostics; none presented results that were problematic.Footnote 11 Last, we ran reduced equations for all three dependent variables, including only those independent variables that had proven substantively significant in full equations; those results do not differ from the full models.
Table 2. OLS Regression of Political and Economic Variables on State Share of Indigent Defense Expenditures
*** p<0.01;
** p<0.05;
p<.10. n=47: Alaska and Michigan are excluded due to lack of data on dependent variable; Oregon is omitted due to outlying values.
Table 3. OLS Regression of Political and Economic Variables on Total Indigent Defense Expenditures Per Capita
*** p<.01;
** p<.05;
* p<.10
n=47: Alaska and Michigan excluded due to lack of data on dependent variable; Oregon omitted due to outlying values.
Regarding outliers, one state, Alaska, was an obvious outlier on the expenditure variable; however, we were missing component items on the welfare and poverty scales for this state. Data on dependent variables were missing for Michigan, so both these states are excluded from the analyses. Although we have a population rather than a sample, it is nonetheless possible that outliers might produce distorted estimates of relationships in the data. To explore this possibility, we examined scatterplots and estimated Cook's D for both models. One state, Oregon, proved to be potentially problematic using this conventional test, so the results presented here do not include that state.Footnote 12 However, our results and conclusions are not significantly altered by its presence or absence.
A third issue to consider was our approach to assessing the substantive significance of coefficients. We note that our dataset of 50 cases represents a universe, not a sample, of states. Therefore, while we present information on the statistical significance of coefficients, that information is not directly meaningful insofar as it cannot represent the probabilities of getting the results by chance. Instead, we offer this information as an intuitive guide to the substantive significance of predictor variables, and the strength of hypothesized relationships. We emphasize here that, with a small n, it is important to attend to the magnitude (and relative magnitudes) of relationships, lest we inadvertently cast aside a hypothesis that, with better measures or models, might prove theoretically and empirically sound.Footnote 13
Results
Tables 2 and 3 present the results of our analyses of state subsidization of indigent defense costs, and of total expenditures per capita, respectively.
State-Level Funding Responsibility
We find limited evidence that state subsidy of indigent defense is shaped by a state's welfare policy climate. We also find no support for the racial threat hypothesis for this policy variable: the association between proportion of African Americans and state subsidy is nil, and states with higher levels of public illiberality in fact take on more responsibility for centralized funding than do others (although this association is very small). However, statehouse partisanship appears to restrict legislatures' willingness to relieve localities of indigent defense; Republican control of both houses, compared with Democratic control of both houses, is associated with a 29 percent decrease in state subsidy. (The coefficient for the presence of a Republican governor is also negative, but nearly negligible.)
However, a critical determinant of state-level responsibility for the funding of indigent defense programs is the existence of a mandate to use public defender programs in all counties of a state. We suspect that the existence of such a mandate, regardless of when it was originally promulgated, is closely associated with centralized administration of indigent defense programming (although we had no measure of centralized administration). We also note that although we expected that wealthier states would tend to shoulder a larger proportion of the responsibility for funding indigent defense, this association is negligible. More important, however, states with higher rates of poverty rates leave more of the burden to local governments. The latter observation contradicts our hypothesis that states with higher poverty rates would fund these services more generally at the center to compensate for high levels of demand upon these services, perhaps suggesting that indigent defendants are not a constituency deemed deserving of state-level assistance.
Total Expenditures Per Capita
The results of these analyses suggest that expenditures on indigent defense are part politics and part structural. Again, we find scant evidence that welfare climate shapes generosity in providing services. We do find evidence for the racial threat thesis, however. While the coefficient for a state's African American population is very small, the quadratic term exerts a significant and negative effect on expenditures per capita. The behavior of these variables in the multivariate model does not differ much from bivariate associations: as African American populations rise beyond 15 percent, expenditures drop increasingly dramatically. These are, of course, largely states in the Deep South. Furthermore, the coefficient for public illiberality exerts the strongest effect on spending: expenditures are markedly lower in states whose publics hold to traditional conservative values that include resistance to racial inclusion and tolerance for diverse ideas. The estimated difference in expenditures between the least and most tolerant states is equivalent to almost a full standard deviation from the mean on this variable. However, Republican domination of the governor's office, or of the legislature, bears little relationship to expenditures, once other political and economic factors are controlled.
We had predicted that states that prescribed public defender systems for most or all of their jurisdictions would spend more on indigent defense, because institutionalized public defenders can become a constituency for better funding. We find otherwise, however. The results do suggest that states that fund indigent defense from state rather than local coffers spend more overall, consistent with our prediction that states have more latitude in subsidizing services than local governments do. Last, while we find no association between a state's wealth and crime rate and its expenditures, states with higher rates of poverty spend more on defense services.
Discussion
The analyses presented are a preliminary foray: we attempt to examine the consequences of the punitive turn for states' due process policies, and by so doing, contribute to the expansion of the empirical knowledge base about contemporary crime and justice policy beyond punishment issues. We emphasize that this study is a first step, and we outline briefly some of the study's limitations to help guide researchers who may wish to continue this line of inquiry.
This study is cross-sectional, and the punitive turn literature emphasizes short- to medium-term historical trends. Much of the research that has attempted to test the propositions advanced by Garland and others have tracked changes in political values and policy outcomes over time, first at the national but increasingly at the state level. We present a snapshot of contemporary policy conditions. Future researchers might profitably trace these variables backward, ideally to the 1960s, when indigent defense emerged as a key due process issue.
We rely on inferences about political processes in interpreting our findings about the links between public opinion, welfare politics, and party dominance. In reality, the politics of budgeting and expenditures is incremental, and it is probably influenced in many states by professional associations (such as state bars, the Legal Services Corporation, and criminal justice professionals; Reference Worden and WordenWorden & Worden 1989). We lack appropriate measures of these actors' and organizations' activities, although we might speculate that they capitalize on (or are constrained by) the more general political conditions that we do measure. Future researchers might follow the lead of Reference BarkerBarker (2006) and follow up aggregate studies such as ours with more intensive case studies of states that vary on more subtle political dimensions, such as interest group activity and public participation in the policy process.
Our measures of key constructs are limited, although this leads us to expect that with more precise measures, the associations we found might prove even more substantial. We have no measures of caseload, for example, and must rely on proxies of crime and poverty rates.
That said, we turn to interpretations of our findings, with some further thoughts on future research agendas. Our hypotheses about the political contexts of indigent defense policy involve three dimensions of politics, derived from scholarship on the recent history of crime policy at the national level, and its implications for contemporary variation in state policy. The first dimension directs us to the partisan politics of criminal justice policy. We hypothesized that indigent defense would be shaped by the same forces that seem to have influenced incarceration, capital punishment, and juvenile justice policies over the past two decades. Those are predominantly the politics of ideology and partisanship: more conservative states have consistently produced more punitive policies. The second dimension involves the politics of race. Research on both welfare and criminal justice policies has raised the “racial threat hypothesis”: the notion that larger minority populations produce a politics of fear among whites, and the result is policies that provide minimal benefits and protections to poor populations (which, in many states, constitute largely minority populations). The last dimension is the politics of welfare: we hypothesized that indigent defense, as a redistributive policy, would track the dynamics of welfare policy, insofar as states that had adopted more generous and inclusive programs since 1996 would practice the same generosity in welfare politics. On which, if any, of these stages are the politics of indigent defense policy played out?
We expected to find that these dimensions were entangled with each other, and that expectation guides our interpretations of our findings. We found that state policy on indigent defense is not a function of welfare policy climate. We commenced inquiry into this relationship in an exploratory fashion, and we acknowledge that as plausible as it might appear, there is not a causal connection between a general climate of inclusion and generosity to welfare clients, and extension of the same to criminal defendants. It is possible that indigent defendants are regarded differently than other beneficiaries of redistributive policies, even in generous states. Welfare provisions have traditionally been aimed only at the “deserving poor” (children, the elderly, and the mentally and physically disabled, among others) from whom were traditionally distinguished the “undeserving” poor—those who could work but chose not to. The distinction achieved statutory status in English law (Reference SlackSlack 1988) and is argued to be of continuing significance by critics of the welfare system today (Reference Forsythe and JordanForsythe & Jordan 2002).
Another explanation is more prosaic but can be empirically substantiated: political climate accounts for both welfare policies and indigent defense policies, so that when the political climate is included in models, welfare policies recede as a predictor variable. In fact, this appears to be the case.Footnote 14 Therefore, the stronger explanation is that political conservatism—specifically, racial threat (operationalized as public tolerance and attitudes toward diversity and, specifically, racial equality) and Republican politics—accounts for both welfare climate and indigent defense policy.
The racial threat hypothesis merits careful consideration, especially insofar as it empirically accounts for not only punishment policy, but also appears to influence defendants' rights policy. This hypothesis often has been presented as a relatively simple mathematical function: where racial minorities are a more significant presence, whites are more alarmed, and adopt broadly punitive welfare and crime policies in an attempt to control the “dangerous classes.” The most obvious historical expression of this thesis, in the United States, is of course the Southern states, where the historical legacy of slavery, reconstruction, the reemergence of a fragile white political domination, and the civil rights movement are thought to have galvanized Southern whites to disdain, fear, and resist African Americans' social and political equality. Reference Beckett and SassonBeckett and Sasson (2004) argue that the mid-century upheavals in the civil rights movement, validated by the federal government, gave Republican politicians an opportunity to resurrect post-Reconstruction racial antagonisms and resentments, reupholstered in the more respectable fabric of public safety and fighting crime.
However, our results suggest that the picture is not so simple. These analyses suggest that the higher the percentage of African Americans in a state, the lower the indigent defense expenditures—but this result is curvilinear, such that states with particularly large minority populations show a greater decline in expenditures. The more direct measure of public attitudes, our scale of illiberality, exerts a stronger pull on policy. A closer examination of the states helps us understand this result. It is true that Deep South states register high on illiberality and low on indigent defense policy, but it is also true that (1) some urban states with large minority populations score low on illiberality yet are reasonably generous on indigent defense, and (2) some states with minuscule minority populations score comparatively high on illiberality and low on indigent defense. Examples of the former include Maryland and New Jersey; the latter is exemplified by Idaho. This finding suggests a subtle but intriguing development in racial attitudes (and their consequences): racial fears are no longer the sole province of the beleaguered South (if they ever were) but instead may be part and parcel of conservative political values. On the other hand, some Southern and border states may be succeeding in leaving the politics of racial divisions behind. Based on this limited policy example, we cautiously suggest that further research on the racial threat hypothesis be directed by more carefully calibrated measures of public ideology and tolerance, in addition to conventional measures of racial heterogeneity.
Party politics is not as strongly associated with policy as we had predicted, although all the associations between progressive funding policies and Republican power are negative. Further, these associations are not diminished by the presence of other political variables in the models. We suggest that despite Republican successes in monopolizing conservative public order and crime agendas at the national level, at the state level partisan politics is not so closely linked to ideology and policy preferences. A cursory review of other measures of state ideology—votes for President George W. Bush in the 2000 presidential election, the Berry indicators of elite and citizen ideology—reveals substantial but not determinative correlations with Republican party control of state legislatures, but no associations with the party of the governor. We cautiously conclude that party politics is a somewhat attenuated influence on due process policies, or at least on expenditures for those policies, but we do not consider this important matter resolved by the analyses presented here.
Last, we note that simple economic accounts for defense services—the need for representation, and the capacity to pay for it—do not account for much state variation in subsidies or expenditures once political factors are controlled. Wealthier states do not spend more on indigent defense, nor do they boost overall expenditures above average, all else equal—a somewhat surprising finding given that indigent defense is a small part of states' budgets, and therefore one that might fairly painlessly draw on extra resources. Where poverty and crime are higher, states contribute less, leaving even more responsibility to local governments (which themselves are, of course, more taxed by the costs of higher welfare caseloads and crime problems). States that mandated, at some point, public defender models are inclined to contribute more to indigent defense operations, but that does not directly translate into overall higher expenditures. In short, the politics of indigent defense are driven less by straightforward economic factors than by the forces that appear to have influenced punishment policies over the last three decades.
Future Directions: Longitudinal Research, Additional Theories, and a Second Look at Policy Context
A number of other theoretical perspectives that have developed in the area of state (and particularly welfare) policy research invite application to the area of indigent defense but were not explored here. Changes over time in state policy in relation to both welfare (Reference SossSoss et al. 2001; Reference MeadMead 2004) and criminal justice policies (Reference CallCall et al. 1991; Reference McGarrell and DuffeeMcGarrell & Duffee 1995; Reference WilliamsWilliams 2003; Reference StuckyStucky et al. 2005) have attracted much scholarly attention but could not be addressed here due to the cross-sectional nature of our data. Indigent defense policy has indeed developed over time, however, including in the ways described above, whereby formalized state policies and institutions came in some places to replace preexisting ad hoc, pro bono arrangements as the latter proved unsatisfactory.
Our analysis has already shown that even with cross-sectional data, political ideology appears to have a significant impact on the likelihood of a state having made reforms in indigent defense policy. With longitudinal data, one could also examine the influence of many other time-contingent factors. Appeals court decisions, for example, have been responsible for solidifying much of the reform that has occurred in states in this area (Harvard Law Review 2005). Equally, the presence of active advocacy groups such as defender associations, civil liberties groups, or other policy entrepreneurs in a locality has been shown to have an association with indigent defense reforms (Reference Worden and WordenWorden & Worden 1989).
The importance of landmark events such as elections was also referred to above but not discussed in detail. In fact, building on classic work on state politics in the 1960s (e.g., Reference KeyKey 1964, Reference Key1967), a sophisticated literature on the subject of the impact of electoral competition on policymaking has been developed by Reference BarrilleauxBarrilleaux (1997, Reference Barrilleaux, Holbrook and Langer2002) and others (Reference Holbrook and Van DunkHolbrook & Van Dunk 1993). Barrilleaux argues that the impact of electoral competition upon state-level policymaking depends upon district-level electoral competition. Where this is tight, legislators can be shown to stick closely to predictable party lines in the area of welfare policy (Reference Barrilleaux, Holbrook and LangerBarrilleaux et al. 2002). Where competition is looser, legislators can drift from these positions. Barrilleaux's model was applied to the criminal justice policy of imprisonment rates between 1978 and 1996 by Reference StuckyStucky and colleagues (2005), who concluded that “partisan politics is most consequential for punishment policies only under particular circumstances, namely, when both parties have faced relatively competitive elections” (2005:232). Our data did not permit us to test this thesis and, as noted previously, it is not clear that the politics of due process galvanize marginal voters the way the politics of punishment may. However, that this approach has relevance both for welfare and criminal justice policy is suggestive of its applicability to indigent defense.
On a related note, some observers express skepticism about exclusive reliance on both cross-sectional and longitudinal research for drawing inferences about political dynamics (Reference WilliamsWilliams 2003; E. Reference MillerMiller 2004). These scholars remind us that the policy process involves elite decision processes, which themselves have been the subject of careful theory and modeling among students of policy. Future research might include legislative history studies that document the actors and conditions that have resulted in critical policy choices, such as the decision to centralize indigent defense programming around a public defender model.
Finally, we observe once again that the past 20 years of research on crime and justice policy, particularly policies involving the courts and corrections, have revolved largely around the upswing in punitiveness: prison-building, incarceration rates, three-strikes and mandatory minimum sentencing laws, and capital punishment, to name a few. These are all important matters, and they certainly merit the work that has been invested in understanding them. But less attention has been paid to the politics of policies and programs that are aimed at improving the functioning of the courts, and those that are aimed at increasing fairness, justice, or services for those accused of crimes. It may not be enough to simply assume that the politics of rights and services are the obverse of the politics of punishment: different dynamics may be at work. We suggest that future research might profitably catalog changes and innovations in these areas, such as we have attempted in this analysis of indigent defense.
Conclusion
This study sought to describe the determinants of indigent defense policy across states on three dimensions. We find that economic capacity and need play a secondary role to political ideology and culture in accounting for state-level characteristics of policy: prescribed public defender models, state subsidy of total expenditures, and total expenditures per capita. We find that indigent defense policy, although redistributive insofar as it confers an entitlement on a means-tested population, has little to do with welfare politics at the state level. We conclude by observing that the characteristics of indigent defense policy may have important implications for equity, effective representation, and therefore for justice, and as such this topic merits continued investigation. Future research could further explore the politics that produce different policy packages, as well as the implications of program type and funding for outcomes.
Appendix: Illiberality Scale
The measure of public intolerance is a factor scale constructed from the following indexes—Tolerance, Racial Integration, Religiosity, Ideology, and Affirmative Action—from the GSS and the NES. The GSS measures were originally devised by Reference BraceBrace and colleagues (2002). The NES measures were compiled by Reference NorranderNorrander (2001).
Tolerance (GSS): State-level average of individual responses to 15 questions, coded no (0), don't know (0.5), yes (1):
1 Should a person opposed to church and religion be allowed to speak?
2 Should a person opposed to church and religion be allowed to teach in college?
3 Should a book by a person opposed to church and religion be removed from the library?
4 Should a person who thinks blacks are inferior be allowed to speak?
5 Should a person who thinks blacks are inferior be allowed to teach in college?
6 Should a book by a person who thinks blacks are inferior be removed from the library?
7 Should a person who is an admitted Communist be allowed to speak?
8 Should a person who is an admitted Communist be allowed to teach in college?
9 Should a book by a person who is an admitted Communist be removed from the library?
10 Should a person who advocates letting the military run the country be allowed to speak?
11 Should a person who advocates letting the military run the country be allowed to teach in college?
12 Should a book by a person who advocates letting the military run the country be removed from the library?
13 Should a self-identified homosexual be allowed to speak?
14 Should a self-identified homosexual be allowed to teach in college?
15 Should a book by a self-identified homosexual be removed from the library?
Racial Integration (GSS): State-level averages of individual respondents' scale scores:
1 Do you think white and black students should go to the same school? (no, don't know, yes)
2 Blacks should not push themselves where they are not wanted. (agree, N/A, disagree)
3 Do you think there should be a law against marriages between members of different races? (no, don't know, yes)
4 White people have a right to keep black people out of their neighborhoods. (agree, N/A, disagree)
5 How strongly would you object if a member of your family wanted to bring a black person to dinner? (strongly/mild, don't know, not strongly)
Religiosity (GSS): State-level average of individual respondents' averaged scores:
1 Would you call yourself a strong (religious preference) or not a strong (religious preference)? (no, don't know, not very strong, somewhat strong, strong)
2 How often do you attend religious services? (less than once a year, once or several times a year, once a month to nearly every week, every week)
3 About how often do you pray? (never/once a week, once or several times a week, once a day, several times a day)
Ideology (NES):
1 We hear a lot of talk these days about liberals and conservatives. Think about a ruler for measuring political views that people might hold, from liberal to conservative. On this ruler, which goes from 1 to 7, a measurement of 1 means very liberal political views, and a measurement of 7 would be very conservative. Just like a regular ruler, it has points in between, at 2, 3, 4, 5, or 6. Where would you place yourself on this ruler, remembering that 1 is very liberal and 7 is very conservative, or have you not thought much about this?
Affirmative Action (NES):
1 Some people feel that the government in Washington should make every effort to improve the social and economic position of blacks. Others feel that the government should not make any special effort to help blacks because they should help themselves. Where would you place yourself on a scale from 1 to 7 where a measurement of 1 means you feel the government should make every effort to support blacks and 7 means you feel the government should not make any special effort to help blacks because they should help themselves?
