Saving Children,Controlling Families

Abstract

This study shows that state efforts at child protection are structured by the policy regimes in which they are enmeshed. Using administrative data on child protection, criminal justice, and social welfare interventions, I show that children are separated from their families and placed into foster care far more frequently in states with extensive and punitive criminal justice systems than in states with broad and generous welfare programs. However, large welfare bureaucracies interact with welfare program enrollment to create opportunities for the surveillance of families, suggesting that extensive and administratively complex welfare states engage in “soft” social control through the surveillance and regulation of family behavior. The article further shows that institutionalization, a particularly restrictive form of foster care placement, is least common in states with broad and generous welfare regimes and generally more common under punitive regimes. Taken together, these findings show that policy regimes influence the interaction between families and the state through their proximate effects on family structure and well-being and through institutional effects that delimit the routines and scripts through which policymakers and street-level bureaucrats intervene to protect children.

Keywords

child welfare criminal justice welfare state public policy political sociology

Child protection is the dominant means through which states seek to control the behavior of parents and ensure the welfare of children. Foster care, in which children are separated from their parents or guardians and placed with an alternative caregiver, is one of the principal tools states use to address child abuse and neglect. This form of coercive welfare intervention has an exceptionally broad reach. Between 2002 and 2011, an average of 1.4 percent of children in the United States came into contact with the foster care system each year.¹ Recent estimates suggest that 5.9 percent of all U.S. children experience foster care at some point during their childhood; 15.4 percent of Native American children and 11.5 percent of African American children enter foster care at some point between birth and age 18 (Wildeman and Emanuel 2014). States vary tremendously in their frequency of intervention. For example, children in Iowa enter foster care at a rate 4.5 times greater than do children in neighboring Illinois, and children in Wyoming enter foster care at a rate 4.8 times greater than do children in Virginia (see Figure 1).² This study shows that a key force driving variation in child protection intervention is the structure of a state’s social policy regime. The extent to which a state prefers punitive or redistributive strategies for addressing social problems affects both the frequency of child protection intervention and the character of those interventions.

Figure 1.

Average State Child Protection Intervention and Policy Regime Character, 2002 to 2011

Despite the consequential roles assigned to child welfare agencies and the impact that intervention has on families, few sociological investigations directly examine the causes and effects of child protection (exceptions include Perry 2006; Reich 2005; Swartz 2005; Wildeman and Emanuel 2014; Wildeman and Waldfogel 2014), and fewer still consider why states differ in their implementation of child protection (Swann and Sylvester 2006). Prior research has evaluated the role that incarceration and cash welfare programs have on increasing or decreasing rates of foster care entry by increasing or decreasing the likelihood of child abuse or affecting families’ capacity to care for children (Andersen and Wildeman 2014; Swann and Sylvester 2006). However, no prior studies have leveraged the insights of comparative social policy research to examine how differences in the general strategies states use to address social problems, frequently characterized as policy regimes (Beckett and Western 2001; Esping-Andersen 1990; Foster and Hagan 2015; Fox 2010; Sutton 2013), structure variation in the implementation of child protection between states. Child protection agencies are tasked with the dual, and frequently contradictory, tasks of coercive social control of parenting on the one hand, and provision of supportive services and resources for struggling families and children on the other (Gordon 1989; Pelton 1989). This ambivalent functional role and administrative structure makes child protection a novel case to explore how the character of a state’s policy regime may broadly structure social policy in domains beyond criminal justice (Sutton 2013) and entitlement or means-tested welfare programs (Esping-Andersen 1990).

Unlike prior studies that consider relationships between social policy and foster care intervention through their micro-level impacts on family structure, resources available to families, or parental behavior, this study suggests that responses to the problem of child abuse are institutionally and politically constructed through the formal and informal interactions of families, street-level bureaucrats, agency administrators, advocacy groups, politicians, courts, and the public (Burstein 1991; Ellermann 2009; Haney 2000; Swartz 2005). These structured relationships can become crystallized as a social policy regime, routinizing and institutionalizing particular styles of intervention (Esping-Andersen 1990; Foster and Hagan 2015; Sutton 2013). State preferences for social policy interventions can be characterized along a continuum, with exclusive and punitive policies at one extreme and inclusive and redistributive approaches at the other (Beckett and Western 2001; Bourdieu 1998). These institutionalized affinities for punitive or redistributive responses to social problems are significantly related to the frequency of family separation and the experiences children have while in the care of the state.

Using panel data on patterns of state child protection constructed from the Adoption and Foster Care Analysis and Reporting System (AFCARS), this study shows that after controlling for relevant demographic, social, and political contexts, states with expansive and generous welfare regimes place fewer children into foster care than do states with expansive and punitive criminal justice systems. However, the administrative complexity of a state’s welfare bureaucracy interacts with welfare program enrollment to produce higher rates of foster care entry, likely through increased opportunities for the surveillance of marginalized families by street-level bureaucrats. This analysis further shows that states with broad and generous redistributive programs place fewer children in restrictive institutional settings than do states that pursue punitive approaches to social problems. Placing a child into foster care is not merely a mechanical response to the incidence of child maltreatment. The frequency and coerciveness of child protection is institutionally structured by a state’s embrace of punishment or redistribution as an appropriate intervention to address social problems.

The findings of this study suggest that disruptive methods of child protection are institutionally aligned with punitive forms of social control. The results also provide evidence that coercive strategies for the regulation of families and children are generally suppressed in regimes that adopt inclusive or redistributive approaches to social problems. Despite ambitious theoretical claims of the broad effects of social policy regimes, comparative scholars typically examine only a small subset of social policy outcomes at the theoretical extremes of the social policy distribution—transfer-based welfare programs and criminal justice (Barker 2009; Beckett and Western 2001; Fox 2010; Soss, Fording, and Schram 2011; Sutton 2013). This study confirms that policy regimes have broad and powerful relationships to the strategies states pursue to achieve social control and provide social welfare.

The Structure of Child Protection in the United States

States conduct formal social control over parenting through child welfare agencies (Reich 2005). They do so by enforcing prohibitions against legally codified unacceptable parenting behavior, establishing a system for the surveillance of child maltreatment, providing a formal process through which violations can be investigated and classified, and intervening in a manner that both sanctions parents and protects children. These agencies also act as conduits through which services aimed at supporting and preserving families in crisis are delivered. Street-level child protection workers assist families in accessing medical, mental health, entitlement, and in-kind resources and are tasked with actively working to reunify children separated from their families when courts deem these efforts feasible and safe (Bartholet 1999; Lindsey 2004). Placing a child into foster care simultaneously punishes parents deemed inadequate or dangerous and provides children and families in crisis with protection and services.

Child welfare agencies rely on a diffuse surveillance network of voluntary observers and various categories of professionals required by law to report suspected child abuse or neglect. Statutes and procedures vary, but if children are perceived to be in imminent danger, caseworkers are empowered to remove children from their families and take them into state custody.³ After removal, the state generally assumes legal guardianship of the children and assigns them to a temporary home with either a member of their extended family (over 22 percent of cases), a non-relative foster home (approximately 40 percent of cases), a group home (approximately 7 percent of cases), or a more restrictive institutional setting such as a residential treatment center (approximately 9 percent of cases).⁴ A family or juvenile court decides on the appropriateness of either reunifying children with their families or permanently separating a child from the family and terminating parental rights, “freeing” a child for adoption (Bartholet 1999).

Within the constraints set by the funding incentives and requirements of federal policy, states have broad discretion in the legislation and implementation of child protection policy. States are guaranteed federal funding for foster care and limited family preservation services through Title IV-E and Title IV-B of the Social Security Act;⁵ many state and local governments commit additional resources to child protection or maltreatment prevention services (DeVooght 2014). Within states, child protection investigation and intervention can be either fully administered at the state-level or decentralized to county agencies. Some states take the further step of privatizing the administration and supervision of foster care services, which affects the relationship between politics, public concern, and agency activity (Gainsborough 2010).

Child protection agencies depend on the bureaucracies and street-level workers from neighboring social policy fields for their day-to-day operation. States have wide latitude in establishing statutory definitions of abuse and neglect, as well as in identifying categories of mandated reporters of child maltreatment (Connelly 2014). Surveillance of child abuse and neglect depends on the detection and reporting of suspected maltreatment to state agencies by police, teachers, social workers, and medical professionals. Between 2007 and 2011, these types of professionals accounted for 75 percent of non-anonymous reports and 58 percent of all reports (Children’s Bureau 2013b). The nature and frequency of contact between at-risk children and families and various categories of mandated reporters depends directly on the structure and breadth of state services in criminal justice, entitlement welfare, education, and public health.

Child Protection and Policy Regimes

Variation in state criminal justice and welfare programs affects the risk that families may abuse or neglect their children (Courtney et al. 2005; Swann and Sylvester 2006; Wakefield and Wildeman 2014). However, variation in social policy also signals differences in how states define and respond to social problems. Social policy is broadly structured by institutional and political forces that lead states to diverse packages of policy outcomes. These policy regimes create institutional pathways that make collaboration, isomorphism, or drift across domains more likely (DiMaggio and Powell 1983; Hacker 2004). Regimes structure the ideological frameworks and schemas that policymakers, bureaucrats, and the public use to narrate the causes of social problems and orient their attitudes toward particular styles of intervention or governance (DiMaggio 1997; Foucault 2003; Garland 2001; Simon 2009). Policy regimes have clear effects on states’ approaches to social welfare (Brown 2013b; Esping-Andersen 1990; Fording, Soss, and Schram 2011; Fox 2010; Hooks and McQueen 2010) and criminal justice (Barker 2006; Beckett 1997; Beckett and Western 2001; Campbell and Schoenfeld 2013; Gottschalk 2006; Lacey 2010; Lynch 2009; Page 2011).

Despite ample theoretical reason to suspect that the effects of policy regimes should extend broadly to the social policy landscape, we have little comparative evidence to assess how regimes structure state programs that are neither explicitly redistributive nor punitive. Comparative analysis of coercive welfare programs, such as child protection, can provide a useful test to evaluate whether the schematic logics and feedback effects produced by policy regimes drive heterogeneous policy outcomes. Historians and theorists have long argued that regimes broadly structure the diverse regulatory and disciplinary components of welfare states (Donzelot 1997; Foucault 1995, 2003; Garland 1987; Gordon 1986; Piven and Cloward 1993; Platt 1969; Polsky 1993; Roberts 2012), but to date few comparative studies empirically test these claims.

Child Protection and Social Welfare Policy

Child protection is motivated by a desire to prevent harm to children caused by parents or other caregivers. It has deep historical ties to progressive social policy reform (Gordon 1989; Tanenhaus 2004), and its professional discourses are grounded in improving children’s well-being (Lindsey 2004). Advocates of child protection argue that states have an affirmative obligation to provide safe and secure accommodations for children when parents are unable or unwilling to do so themselves (Bartholet 1999). They argue that these interventions are a regrettable but necessary component of a broad and generous welfare state and that current efforts do not go far enough to protect children (Epstein 1999). Critics counter that child protection is a vehicle through which states monitor and punish parents and families who fail to conform to hegemonic parenting standards (Abramovitz 1988; Roberts 2012). Given research demonstrating that the construction of the social problem of child abuse depends on the demonization of families that deviate from normative parenting ideals (Best 1993; Nelson 1986; Pfohl 1977), these critics suggest that child protection agencies’ primary task is the regulation of marginalized families and parents (Donzelot 1997; Roberts 2002). The links between welfare policy and child protection suggest a complex set of relationships between the scale and character of redistributive welfare services and child protection intervention.

As the ease with which families can access entitlement benefits and the value of benefits increase, the incidence of child abuse and neglect cases that are primarily a function of material deprivation should decrease, as poverty is among the key risk factors for child maltreatment (Sedlak et al. 2010). Indeed, many studies suggest that as welfare benefits decrease and welfare eligibility becomes more restrictive, foster care entries increase as a function of families’ increasing inability to care for their children (Courtney et al. 2005; Swann and Sylvester 2006). Generous social policy regimes are also likely to dedicate more resources to the prevention of child abuse and neglect as part of a general effort to address and ameliorate social problems associated with poverty (Epstein 1999).

Caseworkers and agencies may be faced with difficult decisions about child removal when supportive resources are scarce or funding for services is institutionally contingent on the state’s custody of a child. Because federal funding for many services for children and families is made available only when children are in out-of-home foster care (DeVooght 2014), street-level child protection workers frequently face a perverse incentive to remove children to trigger funding for needed services. States with expansive welfare services are less likely to rely on foster care as the sole mechanism to provide resources to children and families, as they provide street-level bureaucrats with a wider and more palatable menu of options for routine intervention to address abuse and neglect. These two potential mechanisms—the effects of welfare programs on the risk of child maltreatment and the availability of alternatives to foster care for caseworkers—suggest a negative relationship between welfare generosity and foster care entry.

Hypothesis 1: Foster care entries will decrease as the breadth and generosity of a state’s welfare programs increase.

However, the increasing administrative breadth of the welfare state, coupled with increased enrollment of eligible families, may increase the frequency of opportunities for the surveillance of low-income families by street-level bureaucrats. The development of social welfare policy creates a proliferation of state opportunities for the social control of populations (Bourdieu 1998; Donzelot 1997; Foucault 1995, 2003), particularly of marginalized groups (Gilliom 2001; Gordon 1998; Simon 1993). As more families come into contact with service providers, nearly all of whom are mandated reporters of child maltreatment, there is an increase in the opportunity for the detection and reporting of child abuse and neglect. Welfare state services may then become a gateway for more coercive and intrusive forms of social control. The detection of child abuse and neglect likely increases as a function of the interaction between the enrollment of families in state services and the administrative capacity of the state to surveil enrolled families.

Hypothesis 2: Foster care entries will increase as opportunities for surveillance of families by street-level welfare bureaucrats increase.

Child protection interventions will likely be less coercive and less disruptive under generous social policy regimes. These states are more likely to pursue minimally disruptive approaches that preserve family ties (Bartholet 1999; Gainsborough 2010), when feasible, and avoid the use of restrictive institutional settings for out-of-home placement of children in state custody. Redistributive regimes have historically not been averse to utilizing coercive institutional interventions (Polsky 1993), but critiques of institutionalization and coercive confinement (e.g., Platt 1969) have likely been effective in undermining contemporary support for institutionalization under redistributive regimes. Child welfare advocacy organizations have publicly argued against restrictive foster care settings; these efforts have likely been more successful in states with redistributive policy regimes (Shatzkin 2015).

Hypothesis 3: The rate of institutionalization of children in foster care will decrease as the breadth and generosity of a state’s welfare programs increase.

Child Protection and Criminal Justice Policy

Child protection also has a complex set of relationships to state criminal justice policy. First, a parent’s incarceration can have immediate and long-term consequences for a family’s stability and structure. Parental incarceration has a clear disruptive effect on a family’s ability to care for a child (Braman 2007; Comfort 2008; Geller et al. 2009; Wakefield and Wildeman 2014) and increases the likelihood of foster care entry by increasing the risk of child maltreatment (Andersen and Wildeman 2014; Swann and Sylvester 2006; Wildeman et al. 2014). Because incarceration strains the resources of individuals who share ties with multiple incarcerated individuals, foster care caseloads and entries are likely to increase as relatives available for informal kinship care reach their capacity for helping (Comfort 2008; Sykes and Pettit 2014; Western and Wildeman 2009). Moreover, incarceration is concentrated in low-income communities of color (Pettit 2012), so these effects are likely to disparately affect already marginalized groups. These direct impacts on families are not the only ways that state punishment structures foster care, however.

Punitive social policies seek to inflict pain in response to the violation of informal or formal rules (Golash 2006). Substantial evidence from qualitative research shows that punitive sanctions are a major part of the toolkits used by street-level child protection caseworkers and judges in family courts (Reich 2005; Roberts 2008). Deference to formal authority and compliance with case plans, rather than children’s best interests, are frequently used in official assessments of parental fitness and may have substantial impact on judicial decisions about foster care entry and parental reunification (Reich 2005). In communities that are under intense surveillance and subject to routine intervention, mothers of color frequently perceive child protection agencies as a threat rather than a resource (Roberts 2008). Both the negative impacts of incarceration on families and the institutional affinity between punishment and family separation suggest a positive relationship between foster care entry and the punitiveness of social policy regimes.

Hypothesis 4: Foster care entries will increase as the utilization of expressive punishment, the scale of policing, and incarceration rates increase.

By definition, states with punitive social policy regimes prefer coercive strategies to manage social problems. Most institutional settings for foster care rely on involuntary confinement, regimented discipline, and structured treatment. Coercive confinement in contexts beyond criminal justice will likely be considered less objectionable in states that more routinely rely on punishment than in states that do not. Policymakers operating within punitive regimes may also be less sympathetic to reformers’ efforts to reduce the use of these kinds of settings for foster care (e.g., Annie E. Casey Foundation 2009). As such, I expect punitive states will utilize institutional settings for the foster care of children in their custody more frequently than will states with less punitive social policy regimes.

Hypothesis 5: The rate of institutionalization of children in foster care will increase as the utilization of expressive punishment, the scale of policing, and incarceration rates increase.

Analytic Strategy

This analysis provides a first empirical test of the relationships between social policy regimes and child protection intervention. By modeling counts of foster care entries and the frequency of institutionalization of children in foster care, I provide evidence that policy regimes influence both the scale of child protection intervention and the character of care that children in state custody receive. These findings suggest that the institutionalization of redistributive or punitive approaches to social problems affect child protection outcomes both mechanically, through affecting the risk of child maltreatment, and institutionally, through structuring the preferences, norms, and rules that influence decisions by policymakers and street-level bureaucrats.

I provide a descriptive summary of the scale and character of foster care intervention nationally and between states in the United States for the years 2002 through 2011, followed by a series of multilevel regression models that test relationships between regime characteristics and foster care outcomes. First, I model foster care entries at the state-year level as a function of a state’s policy regime,⁶ controlling for relevant demographic and political variables. Second, I model the frequency of the institutionalization of children in foster care in residential treatment centers or other restrictive environments as a function of variation in social policy regimes. I then compute a series of predicted rates of foster care entries and rates of institutionalization based on theoretically motivated counterfactuals and parameter estimates drawn from simulated replications.⁷

Data and Measures

States are required by law to report information about all children involved in foster care to the U.S. Department of Health and Human Services through the Adoption and Foster Care Analysis and Reporting System (AFCARS), providing a case-level description of all children involved in state-supervised foster care or adoption (Children’s Bureau 2013a).⁸ For the years under consideration here, 2002 to 2011, AFCARS reports a national average of 1,026,583 contacts with foster care per year, or a total of 12,319,001 case-years. To measure the scale of foster care intervention, I constructed state-year level counts of foster care entries from case-year level AFCARS data. Entries are counted as all children that came into foster care during the reporting period. AFCARS reports the current setting of a child’s foster care placement, including a code for institutions such as detention centers, residential treatment centers, and inpatient mental hospitals. Disaggregating this range of institutional placement settings might reveal qualitative variation in institutionalization between places, but such distinctions are not possible with currently available data. I rely on these counts of children in institutional foster care to provide a measure of the frequency of a state’s utilization of restrictive forms of child protection, an important qualitative indicator of the kind of foster care states provide to children in their custody.

To triangulate a state’s social policy regime (Meyers, Gornick, and Peck 2001), I rely on three categories of measures: expressive policy action, policy extensiveness, and administrative breadth. The punitiveness of a social policy regime is measured through the issuance of new death sentences in state courts, incarceration rates, and the number of full-time state and local police officers employed in a state per capita. Incarceration rates, a measure of the extensiveness of a state’s punitive social policy, include persons held in state and local correctional facilities per capita, obtained from the Bureau of Justice Statistics’ National Prisoner Statistics (U.S. Department of Justice 2013). Counts of issued death sentences by state and year, a measure of a state’s utilization of expressive punishment, are compiled by the Death Penalty Information Center (2015) from the Bureau of Justice Statistics’ Capital Punishment Annual Reports, included here as rates of new death sentences issued per prison admission at the state-year level. Data on the number of full-time officers employed by state and local law enforcement agencies are provided by the Annual Survey of Public Employment and Payroll conducted by the U.S. Census Bureau, and are included here as officers per capita to capture the administrative breadth of state criminal justice regimes.

The expressive generosity of a state’s welfare regime is measured through the maximum benefit for a family of three available through TANF, adjusted for state regional price parity (obtained from the Bureau of Economic Analysis) and inflation. The extensiveness of a state’s welfare regime is measured by enrollment per children in poverty for TANF, and enrollment per persons in poverty for SNAP and Medicaid. These measures are compiled by the University of Kentucky Center for Poverty Research (UKCPR) (2014). Finally, I measure the administrative breadth of a state’s welfare regime with data on full-time public welfare workers per capita, obtained from the Annual Survey of Public Employment and Payroll. I handle the small proportion of missing data for policy regime measures through multiple imputation, using adjustments for the panel structure of these data (Honaker and King 2010; Honaker, King, and Blackwell 2011; Rubin 1996).

Regional and temporal heterogeneity in population characteristics associated with crime, poverty, child maltreatment, and political processes all likely influence social policy outcomes. Prior research has established that family structure, parental employment, and socioeconomic status are correlated with child abuse and neglect (Sedlak et al. 2010). To capture the exposure of a state’s population to risk factors for child maltreatment, I include measures for rates of child poverty, unemployment, single-parent-headed households, and adults over age 25 with less than a high school education using the 2000 Decennial Census and 2001 through 2011 American Community Survey (ACS) (Ruggles et al. 2010).⁹ I also include a measure of food insecurity, calculated as an index of responses to questions about access to food and hunger from the Current Population Survey and compiled as three-year rolling averages by the UKCPR (2014). Gross state products per capita, adjusted for inflation, also provided by the UKCPR through the U.S. Department of Commerce, act as a control for the relative prosperity of a state. I calculated crime rates from the FBI’s Uniform Crime Reports as the sum of violent and property index crimes known to police per capita.

Political ideology may play an important role in structuring a state’s response to perceived social problems, so I include a measure of state legislative ideology lagged by one year (Berry et al. 1998).¹⁰ Political ideologies play a key role in structuring criminal justice and welfare regimes (Gilens 2000; Jacobs and Helms 2001; King 2008), but child protection has a more ambiguous and under-examined relation to political partisanship (Gainsborough 2010). I also provide a measure of the proportion of a state’s population that is black, providing a limited control for the role of racial politics on foster care policy. During this time period, black children entered foster care nearly 3.9 times as frequently as white children,¹¹ and research suggests that racial hostility and paternalism play an important role in structuring social policy (Brown 2013a; Bruch, Ferree, and Soss 2010; Keen and Jacobs 2009; Roberts 2002; Welch and Payne 2010).

Statistical Methods

I model expected counts of total foster care intervention and counts of outcomes for children within foster care as Poisson distributed with state and year random intercepts, an exposure term to adjust for the size of the relevant eligible population, and an overdispersion term (Harrison 2014).¹² The inclusion of state and year random intercepts adjusts for clustering of observed outcomes caused by repeated measurements of states over time and correlations induced by year-to-year national trends (Gelman and Hill 2007). I chose these partially-pooled multilevel models to account for both between and within state variation in child protection outcomes in parameter estimates. This multilevel approach, in which state- and year-level random intercepts are simultaneously modeled with population average fixed effects, partially pools the estimates of random effects toward their group average, down-weighting extreme values that occur with classically estimated intercepts (as group dummy variables) due to the limited number of observations available in the relatively short time-span covered in these data. This approach is a better theoretical and methodological fit for this analysis than are alternative approaches that exclusively model within-state changes over time, or fully pooled models that fail to correct for autocorrelation and heteroscedasticity induced by including repeated measurements of states over time. These models enable an examination of how differences in the structure of policy regimes relate to foster care outcomes between states, as well as how shifts in the orientation of regimes within states affect child protection efforts.¹³

Expected values of outcome variables are modeled as follows:

y ~ Poisson (X, exposure)

\begin{array}{l} \log (E (y | X) = β_{0} + β X + δ + γ + θ \\ + \log (exposure) \end{array}

δ_{s} ~ N (0, σ_{δ}^{2})

γ_{t} ~ N (0, σ_{γ}^{2})

θ_{s t} ~ N (0, σ_{θ}^{2})

Where y is a vector of observed counts of child protection outcomes, X is the matrix of observed state-year level values for predictors, exposure is the size of the eligible population from which y is drawn (the state-year child population for foster care entries and the foster care caseload for institutionalization), $δ$ is a vector of state-level random intercepts, $γ$ is a vector of year-level random intercepts, and $θ$ is an observation-level overdispersion term.

Results from these regression models clarify the statistical significance and direction of relationships between social policy regimes and child protection, but they do not provide easily interpretable information about the magnitude of relationships in terms of expected changes in outcomes of interest for theoretically interesting counterfactual scenarios. I provide visualizations¹⁴ of expected rates of foster care entry and institutionalization in Figure 2 for a series of counterfactual states with varying policy regime configurations. These results are produced through an algorithm for estimating expected values of interest and model uncertainty with simulated replications developed by King, Tomz, and Wittenberg (2000).¹⁵

Figure 2.

Predicted Foster Care Entry Rates per 1,000 Child Population and Institutionalization Rates per 1,000 Children in Caseload from Counterfactual Scenarios and Simulated Replications

Findings

Descriptive Findings

States vary considerably in the frequency of foster care entry, the rate at which children in foster care are institutionalized, and the structure of their social policy regimes (see Table 1). To demonstrate the geographic patterns of this variation, Figure 1 displays a series of maps of state-level 2002 to 2011 average values of focal measures in quintiles. Table 1 provides a description of outcome and predictor variables between 2002 and 2011. An examination of the ranges and standard deviations of focal measures shows dramatic heterogeneity between states’ child protection systems and the contexts in which they make and implement policy.

Table 1.

Descriptive Statistics of U.S. States, 2002 to 2011

Variable	Mean	Std. dev.	Minimum	Maximum
Foster care entries per 1,000 child population	4.51	1.82	1.42	10.46
Foster care caseload size per 1,000 child pop.	15.65	5.49	5.67	31.16
Institutional foster care placements per 1,000 children in caseload	70.03	48.59	1.38	334.77
Proportion of adults over age 25 with less than high school education	.12	.04	.06	.22
Food insecurity per capita	.13	.03	.06	.22
Unemployment rate	.07	.02	.03	.14
Proportion of children in single-parent households	.36	.06	.21	.55
Child poverty rate	.18	.05	.08	.32
Gross state product per capita, regional price parity and inflation adjusted	$45214.83	$9692.41	$25192.49	$85947.80
Legislative ideology, larger values indicate more liberal state legislature	49.36	23.23	.00	91.03
Index crimes known to police per 1,000 pop.	36.07	9.70	14.94	66.28
Proportion black pop.	.10	.09	.002	.37
TANF benefits for a family of three, regional price parity and inflation adjusted	$477.45	$161.68	$196.57	$1087.25
Medicaid enrollment per persons in poverty	1.27	.41	.17	5.34
SNAP enrollment per persons in poverty	.76	.20	.37	1.42
TANF enrollment per children in poverty	.31	.17	.03	.93
Full-time state and local public welfare workers per 1,000 pop.	1.66	.52	.41	2.99
Incarceration per 1,000 adult pop.	5.04	1.91	1.91	11.49
New death sentences issued per 1,000 prison admissions	.15	.25	.00	1.52
Full-time state and local police officers per 1,000 pop.	2.19	.46	1.46	3.94

Note: Foster care data from AFCARS 2002 to 2011 (Children’s Bureau 2013a); child poverty, unemployment, education, family structure, and racial composition from 2000 U.S. Census and 2001 to 2011 American Community Survey, three-year rolling averages (Ruggles et al. 2010). Legislative ideology from Berry and colleagues (1998). Crime from FBI Uniform Crime Reports. Incarceration rates from BJS National Prisoner Statistics (U.S. Department of Justice 2013). TANF, SNAP, Medicaid, food security, and GSP from UKCPR poverty policy data (2014).

Foster care is a relatively common experience for children and families. Between 2002 and 2011, the average state placed about 4.5 children per thousand in its population into foster care per year, and had about 15.7 children per thousand on its caseload. Illinois had the lowest average rate of foster care entry during this period (1.7 entries per 1,000 child population); Wyoming had the highest average rate of foster care entry (8.4 entries per 1,000 child population) (see Figure 1).¹⁶ Virginia had the smallest average foster care caseload (7.1 per 1,000 child population), and Nebraska had the highest average caseload (28.1 per 1,000 child population). About 7 percent of children in foster care in the United States were institutionalized between 2002 and 2011. However, there is dramatic variation in the use of institutional foster care between states. Colorado had the highest average rate of institutionalization of children in its foster care system. Between 2002 and 2011, approximately 19 percent of children in foster care in Colorado, on average, resided in restrictive institutional settings, such as residential treatment centers, juvenile justice detention centers, or inpatient mental health facilities. Montana had the lowest average rate of institutionalization during this period, with only 1.2 percent of its foster care population residing in institutional facilities.

As Figure 1 shows, the rates of foster care entry and institutionalization are consistently high in a cluster of states: Minnesota, North and South Dakota, Nebraska, Nevada, Wyoming, and Colorado. Rates of entry are consistently low in one primary cluster of states in the South—Alabama, Mississippi, Louisiana, Georgia, and Texas—although Illinois, Maine, and Utah also have consistently low entry rates. A cluster of states in New England tend to have low institutionalization and entry rates.

A cluster of Northeastern, Midwestern, and West coast states have the highest TANF benefit levels available to a family of three after adjusting for regional price parity. States in the South have consistently low TANF benefits, even after adjusting for differences in the costs of living between places. Similar patterns hold in the per capita staffing rates of welfare bureaucracies. States in the Midwest and Northeast generally have much larger welfare bureaucracies than do states in the South. States in the South also tend to have higher rates of incarceration than do states in the Northeast or Midwest, and Nevada and Arizona have exceptionally high incarceration rates. In general, states in the Northeast and South have very high police staffing levels, although Illinois, Wyoming, and Arizona also have very large police forces. States in the Northwest and upper Midwest have relatively low police staffing levels. This confirms the frequently observed inverse relationship between punishment and redistribution in social policy regimes within the United States (Beckett and Western 2001; Soss et al. 2011). The bivariate relationships displayed in Figure 1 suggest that states with low welfare benefit levels and high incarceration rates in the South also generally have low foster care entry rates, and states with high TANF benefits and large welfare bureaucracies appear to have higher rates of institutionalization, the opposite direction predicted in this study’s hypotheses. However, underlying relationships between child protection outcomes and regime characteristics are likely masked by confounding demographic and political variables.

Statistical Models

Table 2 displays regression models examining the relationships between foster care entry, the institutionalization of children in foster care, and social policy regime characteristics. Results from Model 1 show that increases in the breadth and punitiveness of the criminal justice system are associated with more frequent entries of children into foster care. Expansive and generous welfare programs are negatively associated with foster care entry. However, the negative relationship between generous welfare regimes and foster care entry is attenuated through the interaction of administrative breadth and TANF enrollment. These results provide support for Hypotheses 1 and 2. Foster care entry rates decrease with increasing welfare generosity, but foster care entries are positively associated with the interaction of TANF enrollment and welfare staffing levels, suggesting that increasing opportunities for surveillance produce more frequent entries into foster care. These results also provide support for Hypothesis 4: incarceration rates, death sentences, and police staffing levels are positively associated with the scale of child protection intervention at the state-year level. The variance of estimated state intercepts is two orders of magnitude larger than the variance of estimated year intercepts, suggesting there is far more heterogeneity between states than between years in rates of foster care entry, conditional on model predictors.

Table 2.

Parameter Estimates and Standard Errors from Models of Foster Care Entries and Institutionalization, U.S. States 2002 to 2011

	Model 1. Entries	Model 2. Institutionalization
Unemployment rate	−.031	.028
	(.019)	(.038)
Single-parent-household rate	.021	−.134^*
	(.031)	(.055)
Food insecurity rate	−.036^*	−.042
	(.016)	(.033)
Child poverty rate	.016	.049
	(.039)	(.083)
Adults w/ less than high school education rate	.027	−.172^*
	(.034)	(.078)
GDP per capita	.032	.013
	(.020)	(.044)
Legislative ideology	−.002	−.046^*
	(.010)	(.025)
Crime per capita	.054^**	.066
	(.017)	(.041)
Percent black population	−.370^***	.197
	(.060)	(.117)
TANF benefit levels	−.054^*	.035
	(.027)	(.065)
SNAP enrollment rate	−.025	−.198^***
	(.020)	(.046)
Medicaid enrollment rate	−.029	.019
	(.025)	(.054)
TANF enrollment rate	.009	−.011
	(.018)	(.045)
Welfare workers per capita	.009	−.123^*
	(.022)	(.054)
TANF enrollment x welfare workers	.035^***	−.058^*
	(.010)	(.025)
Incarceration rate	.106^**	.061
	(.035)	(.078)
Death sentences per prison admission	.026^**	.028
	(.008)	(.021)
Police per capita	.053	.019
	(.030)	(.061)
Intercept	−5.499^***	−2.901^***
	(.051)	(.094)
Variance of state random effects $(σ_{δ}^{2})$	.12	.42
Variance of year random effects $(σ_{γ}^{2})$	.001	.00

Note: All covariates scaled to $\bar{x} = 0, s = 1$ , dependent measures are untransformed. Model 1 is offset by log (child pop) $\log (c h i l d p o p)$ , Model 2 is offset by log (caseload) $\log (c a s e l o a d)$ . All models include state (n = 50) and year (n = 10) random intercepts. Variances of random effects are averaged across imputations. N = 496.

p < .05; ^**p < .01; ^***p < .001 (two-tailed tests).

Results from Model 2 illustrate the relationships between policy regime characteristics and the institutionalization of children in a state’s foster care system. Hypothesis 3 predicts that states with more generous and expansive social policy regimes will institutionalize children in foster care less frequently than states with austere welfare programs, and Hypothesis 5 predicts that states with punitive social policy regimes will institutionalize children in foster care more frequently than will states with less punitive regimes. These results provide evidence in support of Hypothesis 3, but provide only weak support for Hypothesis 5. States with more generous social policy regimes are less likely to place children into restrictive institutional settings once children have entered foster care, and states with punitive policy regimes appear slightly more likely to place children into institutional foster care, although no single measure reaches the .05 significance threshold in this model. State random intercepts have high variability ( $σ_{δ}^{2} = . 42$ ), but year intercepts are effectively zero, suggesting that conditional institutionalization rates are strongly clustered within states but not within national time trends. Taken together, these results show that a state’s social policy regime is related to both the frequency of intervention and the kinds of foster care children receive.¹⁷

Counterfactual Simulation Results

Figure 2 displays predicted child protection outcomes based on six counterfactual regime configurations and parameter estimates from an arbitrarily large number of simulated replications. In scenario one, TANF benefit levels, SNAP enrollment per persons in poverty, Medicaid enrollment per persons in poverty, and TANF enrollment per children in poverty are incrementally increased from their observed minima to their observed maxima, while holding welfare staff per capita and all other measures at their observed mean values. In scenario two, incarceration rates, the issuance of new death sentences per prison admission, and police officers per capita are incrementally increased from their minimum observed values to their maximum observed values while holding all other values at their means. Scenario three systematically varies per capita welfare staff from the observed minimum to the observed maximum while holding welfare program enrollment and benefits at one standard deviation below the mean observed value, and all other values at their mean. Scenario four varies per capita welfare staffing from its minimum to its maximum while holding program enrollment and benefits at one standard deviation above the mean observed value. Scenario five incrementally varies all welfare dimensions, including staffing levels, to examine joint regime effects on institutionalization, and scenario six repeats this procedure for all criminal justice measures. This strategy enables an evaluation of the joint predicted effects of regime characteristics on the scale and character of child protection intervention. A counterfactual state with mean values for all independent variables is expected to have 4.1 foster care entries per 1,000 child population [95 percent CI 3.7, 4.5], and the same counterfactual mean state is predicted to have 55.4 children in institutional foster care per 1,000 children in that state’s foster care caseload [44.6, 66.1].

The results displayed in Figure 2, Plot 1 provide further support for Hypothesis 1. As welfare generosity and program inclusiveness systematically increase, the expected number of foster care entries drops. A state with a generous and inclusive welfare regime (SNAP, TANF, Medicaid enrollment and TANF benefits mean + 1 SD) is expected to place 3.7 children into foster care per 1,000 [3.3, 4.2], whereas a state with an austere welfare regime (mean – 1 SD for above measures) is expected to place 4.5 children per 1,000 into foster care [4.0, 5.1]. For a state with a mean child population (1.47 million), an increase in welfare generosity from one standard deviation below the mean to one standard deviation above the mean is expected to reduce foster care entries by approximately 1,100 per year. Plot 2 illustrates the strong positive relationship between punitive regime characteristics and foster care entry, providing clear support for Hypothesis 4. A state with a criminal justice regime that is less punitive than the average state (mean – 1 SD for incarceration, death sentences, police per capita) is expected to have 3.4 foster care entries per 1,000 children [3.0, 3.9], whereas states with broad and punitive criminal justice regimes are expected to place 4.9 children per 1,000 into foster care [4.3, 5.6]. These results predict an increase of 2,200 entries if a state with an average child population moved from a relatively lax criminal justice regime (mean – 1 SD) to a more punitive criminal justice regime (mean + 1 SD).

Plots 3 and 4 of Figure 2 illustrate the estimated interaction between TANF enrollments per children in poverty with the per capita number of full-time public welfare workers in a state. These results illustrate the countervailing effect that surveillance may have on the reductions in caseloads produced by expansive welfare regimes, providing support for Hypothesis 2. A counterfactual state with above average benefits and enrollment rates (+ 1 SD) and a larger than average welfare bureaucracy is expected to place 3.9 children per 1,000 into foster care in a given year [3.4, 4.4], an average difference of .2 placements per 1,000 children when compared to a generous welfare state with a welfare bureaucracy staffing rate at the observed mean. These estimates predict an increase of approximately 290 children per year in a state with a mean child population. An increase in the opportunities for bureaucratic surveillance of low-income families in generous welfare regimes moderately attenuates the otherwise clear negative relationship between welfare policy and foster care entry.

Foster care institutionalization is significantly lower in states with generous and expansive welfare policies than in states with austere welfare policies, as shown in Plot 5. A state with an expansive and generous welfare policy regime is expected to institutionalize children in foster care at a rate of 39.2 per 1,000 in its caseload [29.8, 50.3], compared to a rate of 78.1 per 1,000 [59.5, 100.4] for a state with narrow and ungenerous welfare programs, providing support for Hypothesis 3. No single measure of expansive and punitive criminal justice policy reaches conventional significance thresholds in models of foster care institutionalization, but joint variation in these measures predicts a positive, but highly uncertain, relationship between punitive policy regimes and foster care institutionalization. A state with a broad and punitive criminal justice regime is expected to institutionalize children in foster care at a rate of 64.9 per 1,000 in its caseload [49.5, 83.2], compared to a rate of 47.4 per 1,000 [35.9, 61.6] in a less punitive state, providing limited support for Hypothesis 5. Based on these models, a state with a robust welfare regime and an average foster care caseload is expected to institutionalize approximately 530 fewer children than a state with an expansive and punitive criminal justice regime.

These counterfactual simulations suggest that policy regime variation has dramatic impacts on foster care entry and institutionalization. They demonstrate that states with generous and broad welfare regimes are expected to have lower rates of foster care entry and lower rates of institutionalization, but increases in the size of welfare bureaucracies attenuate this relationship under generous and extensive welfare regimes. These simulations also demonstrate that punitive social policy regimes are expected to have significantly higher rates of foster care entry. On average, punitive policy regimes are also expected to have higher rates of foster care institutionalization, although estimates of this relationship are highly uncertain.

Discussion

These findings show that child protection intervention is most frequent in states with punitive social policy regimes and suppressed in states with social policy regimes that favor redistributive interventions. States with redistributive social policy regimes are also less likely to institutionalize children in foster care than are states with modest welfare policies. These results show that child protection intervention is structured both by the direct effects of social policy on the risk of child abuse and neglect and the influence of a state’s policy regime in defining appropriate styles of intervention and strategies of governance. States with redistributive policy regimes generally utilize disruptive child protection less frequently, but when these states have large welfare bureaucracies, foster care entries increase. This finding suggests that the administrative structure of agencies under generous welfare regimes expands or constrains the opportunities for state surveillance of low-income families.

No prior quantitative studies have evaluated child protection as a component of a complex policy regime for the management of social problems. Earlier studies have made important contributions in establishing causal relationships between incarceration, welfare provision, and foster care entry (Andersen and Wildeman 2014; Courtney et al. 2005; Swann and Sylvester 2006)—mediated by the increased or decreased likelihood of child abuse or neglect—but this study shows that a state’s policy regime structures the preferred or routinized modes of responding to social problems. That is, foster care entry is not only a reflexive response to the incidence of child abuse or neglect; disruptive child protection indicates a particular and paternalistic style of governance (Jackman 1994; Soss et al. 2011). Building on the sociological tradition of comparative welfare and comparative punishment scholarship (Barker 2009; Fox 2010; Sutton 2013), this study shows that identification of and response to social problems is a structured process dependent on social and political forces that are partially exogenous to the underlying issues a policy attempts to address.

Although identification of the forces and processes structuring a state’s policy regime is beyond the scope of this study, prior research suggests that systems of racial stratification and classification (Brown 2013a; Fox 2010; Muller 2012; Quadagno 1994; Soss et al. 2011), political institutions (Barker 2009; Tonry 2007; Weir, Orloff, and Skocpol 1988), the composition of electoral coalitions (Esping-Andersen 1990; Sutton 2013), path-dependent processes of legislation and policy implementation (Hacker 2002; Pierson 2000), and the discourses, organizational strength, and legitimacy of professional and advocacy organizations (Garland 2001; Gottschalk 2006; Page 2011) are important forces in structuring whether states will embrace redistributive, punitive, or paternalistic forms of policy intervention. Future research should consider child protection policy as an additional outcome in analyses of the formation of state policy regimes. Additional research should also evaluate whether deep racial inequities in child protection are a function of the character of social policy regimes. A growing body of evidence shows that racial inequality in foster care is poorly explained by differences in poverty and social disadvantage between groups (e.g., Wulczyn et al. 2013).

This study has some limitations due to the scope and precision of federal data collection. Placing children with a member of their extended family is the least disruptive option available to caseworkers and courts, but measuring the scale of kinship-based foster care is complex. Formal kinship placements are recorded by AFCARS, but as many as 75 percent of all placements arranged by child welfare officials with kin are done without taking a child into state custody (referred to as voluntary kinship care), and hence are not recorded in AFCARS (Annie E. Casey Foundation 2012). These procedures vary widely by jurisdiction (Macomber 2003). A count of kinship foster care placements would enable an alternative measure to examine regime effects on less-disruptive strategies of child protection, but such an analysis is not feasible without substantial new national data collection efforts. However, results presented here are robust to the exclusion of all kinship placements from foster care entry counts (see Table S1 in the online supplement). Additionally, disaggregation of the settings of institutional placements would allow for greater precision in identifying regime effects on qualitative variation in foster care placements. For example, inpatient mental institutionalization is likely subject to different processes than is placement at a residential treatment center focused on behavior modification through rigid discipline. However, without changes to federal reporting standards or substantial new data collection efforts, such distinctions are not possible.

This analysis shows that the extreme variation in rates of child protection intervention between states is not simply a function of differences in population composition that produce higher or lower rates of child abuse or neglect. States respond to the problem of child maltreatment in a manner that reflects deep differences in the ways they identify and manage social problems. Policy regimes structure the scripts policymakers and bureaucrats use to design and implement child protection systems and a state’s infrastructure for surveillance and case processing. Policy regimes play a central role in expanding or constraining coercive social welfare programs, such as child protection. Variation in the character of social policy regimes between states provides a compelling explanation of why some states remove children from their families more frequently than do others.

Footnotes

Acknowledgements

I am grateful to Becky Pettit, Jake Rosenfeld, Hedy Lee, Alexes Harris, Katherine Beckett, Stew Tolnay, Steve Pfaff, Sarah Quinn, Chris Wildeman, Heather Hill, Kyle Crowder, Kate Stovel, Catherine Clepper, Lindsey Beach, Christina Hughes, Andre Stephens, Jennifer Laird, Rebecca Ferrell, participants in the University of Washington’s Seminar on Culture, Organizations, Politics and Economics, the editors of ASR, and the anonymous reviewers for valuable feedback on this project.

Data Note

The data used in this analysis were made available by the National Data Archive on Child Abuse and Neglect, Cornell University, Ithaca, NY, and have been used with permission. Data from the Adoption and Foster Care Analysis and Reporting System (AFCARS) were originally collected by the Children’s Bureau. Funding for AFCARS was provided by the Children’s Bureau, Administration on Children, Youth and Families, Administration for Children and Families, U.S. Department of Health and Human Services. The collector of the original data, the funder, the Archive, Cornell University, and their agents or employees bear no responsibility for the analyses or interpretations presented here.

Notes

Frank Edwards is a graduate student in the Department of Sociology at the University of Washington. He studies how politics, institutions, and organizations affect social policy implementation. He is currently pursuing research on the structure of maltreatment surveillance and mechanisms of racial inequality in child welfare case processing.

References

Abramovitz

Mimi

. 1988. Regulating the Lives of Women: Social Welfare Policy from Colonial Times to the Present. Boston: South End Press.

Adolph

Christopher

. 2013a. Simcf: Counterfactuals and Confidence Intervals for Estimated Models. Seattle: University of Washington (http://faculty.washington.edu/cadolph/software).

Adolph

Christopher

. 2013b. Tile: Plot Multiple Graphical Displays in a Tiled Layout with Automated or Customizable Titles, Axes, and Annotations. Seattle: University of Washington (http://faculty.washington.edu/cadolph/software).

Andersen

Signe Hald

Wildeman

Christopher

. 2014. “The Effect of Paternal Incarceration on Children’s Risk of Foster Care Placement.” Social Forces 93(1):269–98.

Annie E. Casey Foundation. 2009. Rightsizing Congregate Care. Baltimore, MD: Annie E. Casey Foundation.

Annie E. Casey Foundation. 2012. Stepping Up For Kids. Baltimore, MD: Annie E. Casey Foundation.

Barker

Vanessa

. 2006. “The Politics of Punishing: Building a State Governance Theory of American Imprisonment Variation.” Punishment & Society 8(1):5–32.

Barker

Vanessa

. 2009. The Politics of Imprisonment: How the Democratic Process Shapes the Way America Punishes Offenders. New York: Oxford University Press.

Bartholet

Elizabeth

. 1999. Nobody’s Children: Abuse and Neglect, Foster Drift, and the Adoption Alternative. Boston: Beacon Press.

10.

Bates

Douglas

Maechler

Martin

Bolker

Benjamin M.

Walker

Steven

. 2014. “lme4: Linear Mixed-Effects Models Using Eigen and S4.” Journal of Statistical Software 67(1):1–48.

11.

Beckett

Katherine

. 1997. Making Crime Pay: Law and Order in Contemporary American Politics. New York: Oxford University Press.

12.

Beckett

Katherine

Western

Bruce

. 2001. “Governing Social Marginality: Welfare, Incarceration, and the Transformation of State Policy.” Punishment & Society 3(1):43–59.

13.

Berry

William D.

Ringquist

Evan J.

Fording

Richard C.

Hanson

Russell L.

1998. “Measuring Citizen and Government Ideology in the American States, 1960–93.” American Journal of Political Science 42(1):327–48.

14.

Best

Joel

. 1993. Threatened Children: Rhetoric and Concern about Child-Victims. Chicago: University of Chicago Press.

15.

Bolker

Benjamin M.

Brooks

Mollie E.

Clark

Connie J.

Geange

Shane W.

Poulsen

John R.

Stevens

M. Henry H.

White

Jada-Simone S.

2009. “Generalized Linear Mixed Models: A Practical Guide for Ecology and Evolution.” Trends in Ecology & Evolution 24(3):127–35.

16.

Bourdieu

Pierre

. 1998. Acts of Resistance: Against the Tyranny of the Market. New York: New Press.

17.

Braman

Donald

. 2007. Doing Time on the Outside: Incarceration and Family Life in Urban America. Ann Arbor: University of Michigan Press.

18.

Brown

Hana E.

2013a. “Racialized Conflict and Policy Spillover Effects: The Role of Race in the Contemporary U.S. Welfare State.” American Journal of Sociology 119(2):394–443.

19.

Brown

Hana E.

2013b. “The New Racial Politics of Welfare: Ethno-Racial Diversity, Immigration, and Welfare Discourse Variation.” Social Service Review 87(3):586–612.

20.

Bruch

Sarah K.

Ferree

Myra Marx

Soss

Joe

. 2010. “From Policy to Polity: Democracy, Paternalism, and the Incorporation of Disadvantaged Citizens.” American Sociological Review 75(2):205–226.

21.

Burstein

Paul

. 1991. “Policy Domains: Organization, Culture, and Policy Outcomes.” Annual Review of Sociology 17:327–50.

22.

Campbell

Michael C.

Schoenfeld

Heather

. 2013. “The Transformation of America’s Penal Order: A Historicized Political Sociology of Punishment.” American Journal of Sociology 118(5):1375–1423.

23.

Children’s Bureau. 2013a. Adoption and Foster Care Analysis and Reporting System (AFCARS), FFY 2012 [Dataset]. Washington, DC: U.S. Department of Health and Human Services, Administration for Children and Families (http://www.ndacan.cornell.edu).

24.

Children’s Bureau. 2013b. National Child Abuse and Neglect Data System (NCANDS) Child File, FFY 2012 [Dataset]. Washington, DC: U.S. Department of Health and Human Services, Administration for Children and Families (http://www.ndacan.cornell.edu).

25.

Comfort

Megan

. 2008. Doing Time Together: Love and Family in the Shadow of the Prison. Chicago: University of Chicago Press.

26.

Connelly

Dana

. 2014. “State Child Welfare Policy: Causes and Consequences.” Dissertation, University of Kentucky, Lexington, KY.

27.

Courtney

Mark E.

Dworsky

Amy

Piliavin

Irving

Zinn

Andrew

. 2005. “Involvement of TANF Applicant Families with Child Welfare Services.” Social Service Review 79(1):119–57.

28.

Death Penalty Information Center. 2015. Death Sentences in the United States from 1977 By State and By Year. Washington, DC. Retrieved August 10, 2015 (http://www.deathpenaltyinfo.org/death-sentences-united-states-1977-2008).

29.

DeVooght

Kerry

. 2014. Federal, State, and Local Spending to Address Child Abuse and Neglect in SFY 2012. Washington, DC: Child Trends.

30.

DiMaggio

Paul J.

1997. “Culture and Cognition.” Annual Review of Sociology 23:263–87.

31.

DiMaggio

Paul J.

Powell

Walter W.

1983. “The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields.” American Sociological Review 48(2):147–60.

32.

Donzelot

Jacques

. 1997. The Policing of Families. Baltimore, MD: Johns Hopkins.

33.

Ellermann

Antje

. 2009. States against Migrants: Deportation in Germany and the United States. New York: Cambridge University Press.

34.

Epstein

William M.

1999. Children Who Could Have Been: The Legacy of Child Welfare in Wealthy America. Madison: University of Wisconsin Press.

35.

Esping-Andersen

Gøsta

. 1990. The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press.

36.

Fording

Richard C.

Soss

Joe

Schram

Sanford F.

2011. “Race and the Local Politics of Punishment in the New World of Welfare.” American Journal of Sociology 116(5):1610–57.

37.

Foster

Holly

Hagan

John

. 2015. “Punishment Regimes and the Multilevel Effects of Parental Incarceration: Intergenerational, Intersectional, and Interinstitutional Models of Social Inequality and Exclusion.” Annual Review of Sociology 41:135–58.

38.

Foucault

Michel

. 1995. Discipline and Punish: The Birth of the Prison. New York: Vintage Books.

39.

Foucault

Michel

. 2003. Society Must Be Defended: Lectures at the Collège De France, 1975–76. New York: Picador.

40.

Fox

Cybelle

. 2010. “Three Worlds of Relief: Race, Immigration, and Public and Private Social Welfare Spending in American Cities, 1929.” American Journal of Sociology 116(2):453–502.

41.

Gainsborough

Juliet F.

2010. Scandalous Politics: Child Welfare Policy in the States. Washington, DC: Georgetown University Press.

42.

Garland

David

. 1987. Punishment and Welfare: A History of Penal Strategies. Aldershot, UK: Gower.

43.

Garland

David

. 2001. The Culture of Control: Crime and Social Order in Contemporary Society. Chicago: University of Chicago Press.

44.

Geller

Amanda

Garfinkel

Irwin

Cooper

Carey E.

Mincy

Ronald B.

2009. “Parental Incarceration and Child Well-Being: Implications for Urban Families.” Social Science Quarterly 90(5):1186–1202.

45.

Gelman

Andrew

Hill

Jennifer

. 2007. Data Analysis Using Regression and Multilevel/Hierarchical Models. New York: Cambridge University Press.

46.

Gilens

Martin

. 2000. Why Americans Hate Welfare: Race, Media, and the Politics of Antipoverty Policy. Chicago: University of Chicago Press.

47.

Gilliom

John

. 2001. Overseers of the Poor: Surveillance, Resistance, and the Limits of Privacy. Chicago: University of Chicago Press.

48.

Golash

Deirdre

. 2006. The Case against Punishment: Retribution, Crime Prevention, and the Law. New York: New York University Press.

49.

Gordon

Linda

. 1986. “Family Violence, Feminism, and Social Control.” Feminist Studies 12(3):453–78.

50.

Gordon

Linda

. 1989. Heroes of Their Own Lives: The Politics and History of Family Violence. New York: Penguin Books.

51.

Gordon

Linda

. 1998. Pitied but Not Entitled: Single Mothers and the History of Welfare 1890–1935. Cambridge, MA: Harvard University Press.

52.

Gottschalk

Marie

. 2006. The Prison and the Gallows: The Politics of Mass Incarceration in America. New York: Cambridge University Press.

53.

Hacker

Jacob S.

2002. The Divided Welfare State: The Battle over Public and Private Social Benefits in the United States. New York: Cambridge University Press.

54.

Hacker

Jacob S.

2004. “Privatizing Risk without Privatizing the Welfare State: The Hidden Politics of Social Policy Retrenchment in the United States.” American Political Science Review 98(2):243–60.

55.

Haney

Lynne A.

2000. “Feminist State Theory: Applications to Jurisprudence, Criminology, and the Welfare State.” Annual Review of Sociology 26:641–66.

56.

Harrison

Xavier A.

2014. “Using Observation-Level Random Effects to Model Overdispersion in Count Data in Ecology and Evolution.” PeerJ Computer Science 2:e616.

57.

Honaker

James

King

Gary

. 2010. “What to Do About Missing Values in Time Series Cross-Section Data.” American Journal of Political Science 54(3):561–81.

58.

Honaker

James

King

Gary

Blackwell

Matthew

. 2011. “Amelia II: A Program for Missing Data.” Journal of Statistical Software 45(7):1–47.

59.

Hooks

Gregory

McQueen

Brian

. 2010. “American Exceptionalism Revisited: The Military-Industrial Complex, Racial Tension, and the Underdeveloped Welfare State.” American Sociological Review 75(2):185–204.

60.

Jackman

Mary R.

1994. The Velvet Glove: Paternalism and Conflict in Gender, Class, and Race Relations. Berkeley: University of California Press.

61.

Jacobs

David

Helms

Ronald

. 2001. “Toward a Political Sociology of Punishment: Politics and Changes in the Incarcerated Population.” Social Science Research 30(2):171–94.

62.

Keen

Bradley

Jacobs

David

. 2009. “Racial Threat, Partisan Politics, and Racial Disparities in Prison Admissions: A Panel Analysis.” Criminology 47(1):209–238.

63.

King

Gary

Tomz

Michael

Wittenberg

Jason

. 2000. “Making the Most of Statistical Analyses: Improving Interpretation and Presentation.” American Journal of Political Science 44(2):347–61.

64.

King

Ryan D.

2008. “Conservatism, Institutionalism, and the Social Control of Intergroup Conflict.” American Journal of Sociology 113(5):1351–93.

65.

Lacey

Nicola

. 2010. “American Imprisonment in Comparative Perspective.” Daedalus 139(3):102–14.

66.

Lindsey

Duncan

. 2004. The Welfare of Children. New York: Oxford University Press.

67.

Lynch

Mona

. 2009. Sunbelt Justice: Arizona and the Transformation of American Punishment. Stanford, CA: Stanford University Press.

68.

Macomber

Jennifer Ehrle

. 2003. Kinship Foster Care: Custody, Hardships, and Services. Washington, DC: The Urban Institute.

69.

Meyers

Marcia K.

Gornick

Janet C.

Peck

Laura R.

2001. “Packaging Support for Low-Income Families: Policy Variation across the United States.” Journal of Policy Analysis and Management 20(3):457–83.

70.

Muller

Christopher

. 2012. “Northward Migration and the Rise of Racial Disparity in American Incarceration, 1880–1950.” American Journal of Sociology 118(2):281–326.

71.

Nelson

Barbara J.

1986. Making an Issue of Child Abuse: Political Agenda Setting for Social Problems. Chicago: University of Chicago Press.

72.

Page

Joshua

. 2011. The Toughest Beat: Politics, Punishment, and the Prison Officers Union in California. New York: Oxford University Press.

73.

Pelton

Leroy H.

1989. For Reasons of Poverty: A Critical Analysis of the Public Child Welfare System in the United States. New York: Praeger Publishers.

74.

Perry

Brea L.

2006. “Understanding Social Network Disruption: The Case of Youth in Foster Care.” Social Problems 53(3):371–91.

75.

Pettit

Becky

. 2012. Invisible Men: Mass Incarceration and the Myth of Black Progress. New York: Russell Sage Foundation Publications.

76.

Pfohl

Stephen J.

1977. “The ‘Discovery’ of Child Abuse.” Social Problems 24(3):310–23.

77.

Pierson

Paul

. 2000. “Increasing Returns, Path Dependence, and the Study of Politics.” American Political Science Review 94(2):251–67.

78.

Piven

Frances Fox

Cloward

Richard

. 1993. Regulating the Poor: The Functions of Public Welfare. New York: Vintage.

79.

Platt

Anthony M.

1969. The Child Savers. Chicago: University of Chicago Press.

80.

Polsky

Andrew J.

1993. The Rise of the Therapeutic State. Princeton, NJ: Princeton University Press.

81.

Quadagno

Jill

. 1994. The Color of Welfare: How Racism Undermined the War on Poverty. New York: Oxford University Press.

82.

R Core Team. 2014. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing (http://www.R-project.org).

83.

Reich

Jennifer A.

2005. Fixing Families: Parents, Power, and the Child Welfare System. New York: Routledge.

84.

Roberts

Dorothy E.

2002. Shattered Bonds: The Color of Child Welfare. New York: Basic Civitas Books.

85.

Roberts

Dorothy E.

2008. “The Racial Geography of Child Welfare: Toward a New Research Paradigm.” Child Welfare 87(2):125–50.

86.

Roberts

Dorothy E.

2012. “Prison, Foster Care, and the Systemic Punishment of Black Mothers.” UCLA Law Review 59:1474–99.

87.

Rubin

Donald B.

1996. “Multiple Imputation after 18+ Years.” Journal of the American Statistical Association 91(434):473–89.

88.

Ruggles

Steven

Genadek

Katie

Goeken

Ronald

Grover

Josiah

Sobek

Matthew

. 2010. Integrated Public Use Microdata Series: Version 5.0 [Machine-Readable Database]. Minneapolis: University of Minnesota.

89.

Sedlak

Andrea J.

Mettenburg

Jane

Basena

Monica

Petta

Ian

McPherson

Karla

Green

Angela

Spencer

. 2010. Fourth National Incidence Study of Child Abuse and Neglect (NIS-4). Washington, DC: US Department of Health and Human Services.

90.

Shatzkin

Kate

. 2015. Every Kid Needs a Family: Giving Children in the Child Welfare System the Best Chance for Success. Baltimore, MD: The Annie E. Casey Foundation.

91.

Simon

Jonathan

. 1993. Poor Discipline. Chicago: University of Chicago Press.

92.

Simon

Jonathan

. 2009. Governing through Crime: How the War on Crime Transformed American Democracy and Created a Culture of Fear. New York: Oxford University Press.

93.

Soss

Joe

Fording

Richard C.

Schram

Sanford

. 2011. Disciplining the Poor: Neoliberal Paternalism and the Persistent Power of Race. Chicago: University of Chicago Press.

94.

Sutton

John R.

2013. “The Transformation of Prison Regimes in Late Capitalist Societies.” American Journal of Sociology 119(3):715–46.

95.

Swann

Christopher A.

Sylvester

Michelle Sheran

. 2006. “The Foster Care Crisis: What Caused Caseloads to Grow?” Demography 43(2):309–335.

96.

Swartz

Teresa Toguchi

. 2005. Parenting for the State: An Ethnographic Analysis of Non-Profit Foster Care. New York: Routledge.

97.

Sykes

Bryan L.

Pettit

Becky

. 2014. “Mass Incarceration, Family Complexity, and the Reproduction of Childhood Disadvantage.” ANNALS of the American Academy of Political and Social Science 654(1):127–49.

98.

Tanenhaus

David

. 2004. Juvenile Justice in the Making. New York: Oxford University Press.

99.

Tonry

Michael

. 2007. “Determinants of Penal Policies.” Crime and Justice 36(1):1–48.

100.

United States Department of Justice. 2013. National Prisoner Statistics, 1978–2011. Washington, DC: Office of Justice Programs, Bureau of Justice Statistics.

101.

University of Kentucky Center for Poverty Research (UKCPR). 2014. UKCPR National Welfare Data, 1980–2013. Lexington, KY: Gatton College of Business and Economics, University of Kentucky.

102.

Wakefield

Sara

Wildeman

Christopher

. 2014. Children of the Prison Boom: Mass Incarceration and the Future of American Inequality. New York: Oxford University Press.

103.

Weir

Margaret

Orloff

Ann Shola

Skocpol

Theda

, eds. 1988. The Politics of Social Policy in the United States. Princeton, NJ: Princeton University Press.

104.

Welch

Kelly

Payne

Allison Ann

. 2010. “Racial Threat and Punitive School Discipline.” Social Problems 57(1):25–48.

105.

Western

Bruce

Wildeman

Christopher

. 2009. “The Black Family and Mass Incarceration.” ANNALS of the American Academy of Political and Social Science 621(1):221–42.

106.

Wildeman

Christopher

Andersen

Signe Hald

Lee

Hedwig

Karlson

Kristian Bernt

. 2014. “Parental Incarceration and Child Mortality in Denmark.” American Journal of Public Health 104(3):428–33.

107.

Wildeman

Christopher

Emanuel

Natalia

. 2014. “Cumulative Risks of Foster Care Placement by Age 18 for U.S. Children, 2000–2011.” PLoS ONE 9(3):e92785.

108.

Wildeman

Christopher

Waldfogel

Jane

. 2014. “Somebody’s Children or Nobody’s Children? How the Sociological Perspective Could Enliven Research on Foster Care.” Annual Review of Sociology 40(1):599–618.

109.

Wulczyn

Fred

Gibbons

Robert

Snowden

Lonnie

Lery

Bridgette

. 2013. “Poverty, Social Disadvantage, and the Black/White Placement Gap.” Children and Youth Services Review 35(1):65–74.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB