Severe Deprivation and System Inclusion Among Children of Incarcerated Parents in the United States After the Great Recession
The expansion of the criminal justice system over the last four decades and the corresponding rise of parental incarceration raises questions about whether the children of current and former inmates are at an increased risk of material hardship that necessitates social service intervention. Recent sociological scholarship finds that the greater surveillance experienced by former inmates and the criminally involved precludes them from seeking medical care and social services. Yet there is no scholarship that assesses health care and social service utilization among children exposed to parental incarceration. In this article, we investigate how race and educational inequality in parental incarceration were associated with markers of deprivation and social program enlistment after the Great Recession. Using data from the 2011–2012 National Survey of Children’s Health (NSCH), we not only find that children with an incarcerated parent experience greater levels of deprivation—material hardship, unmet health needs, and residential instability—but that these children are drawn into social service programs at a higher rate than the rate for children unexposed to parental incarceration. Nearly 2.1 million children (or 81 percent of minors) with an incarcerated parent are enrolled in at least one social service program. Our findings are consistent with a “system inclusion” perspective, which aligns with David Garland’s and William Julius Wilson’s theoretical and historical explanations of social service participation among disadvantaged minors.
incarceration, hardship, deprivation, program, participation, inequality, welfare
By the close of 2012, over 2.2 million Americans were incarcerated in local, state, and federal correctional facilities, and another 4.8 million were under some form of criminal justice supervision (Glaze and Herberman 2013), making the United States the world leader in incarceration. The rise of the penal state as a form of social exclusion has been likened to other historical configurations of race and social relations in the United States (Alexander 2010; Wacquant 2001). The contemporary incarceration of entire demographic groups has made the risk of imprisonment highly stratified by race and class, and incarceration now constitutes a new stage in the life course (Garland 2001a; Pettit and Western 2004): the lifetime risk of imprisonment among young, undereducated African American men hovers around [End Page 108] 70 percent (Pettit 2012; Pettit, Sykes, and Western 2009; Western and Wildeman 2009).
The social, economic, and familial consequences of criminal justice contact are staggering. Having a criminal record reduces employment and earnings (Pager 2003; Pager and Quillian 2005; Pettit 2012; Western 2006; Western and Pettit 2005), lowers marital prospects and heightens the risk of divorce (Edin and Kefelas 2005; Lopoo and Western 2005; Massoglia, Remster, and King 2011; Sampson, Laub, and Wimer 2006; Western, Lopoo, and McLanahan 2004), bars civic participation and obscures voter turnout (Pettit 2012; Uggen and Manza 2002), increases health disparities (Freudenberg 2001; Johnson and Raphael 2009; Wildeman and Muller 2012), worsens communities (Clear 2009), and creates new forms of nondischargeable legal debt owed to correctional and justice systems (Harris, Evans, and Beckett 2010). Spending time in prison or jail also excludes incarcerated men and women from national surveys that measure employment, high school completion and health (Ewert, Sykes, and Pettit 2014; Pettit 2012), particularly after the passage of civil rights legislation that was aimed at increasing the socioeconomic opportunities and civic engagement of racial and ethnic minorities (Pettit and Sykes 2015).
The prison boom has produced other collateral consequences that extend to the family members of current and former inmates (Hagan and Dinovitzer 1999). Over half of all prisoners have children under the age of eighteen (Pettit 2012; Pettit et al. 2009), and about 45 percent of fathers and two-thirds of mothers were living with their children at the time they were sent to prison (Glaze and Maruschak 2010; Mumola 2000). In 2012 nearly 2.6 million children had at least one parent in prison or jail (Sykes and Pettit 2014), and racial inequality in the risk of ever having a parent incarcerated is pronounced: 3.6 percent of white youth experienced parental incarceration by age fourteen in 2009, compared to 24.5 percent of blacks (Wildeman 2009). The consequences of parental incarceration exacerbate existing childhood disadvantage; children of incarcerated parents have increased behavioral challenges (Geller et al. 2012; Johnson 2009; Murray, Loeber, and Pardini 2012; Wakefield and Wildeman 2011; Wildeman 2010), and the learning difficulties and grade retention associated with parental incarceration place children at risk of early educational inequality (Cho 2009a, 2009b, 2010, 2011; Eddy and Poehlmann 2010; Foster and Hagan 2009; Hagan and Foster 2012; Turney 2014; Turney and Haskins 2014).
The social and economic consequences of criminal justice contact among adults raise important questions about the impact of parental incarceration on child well-being. Despite the burgeoning literature on the effects of incarceration on communities, families, and society as a whole, no empirical research exists on how the Great Recession may have placed children with an incarcerated parent at greater risk of government assistance. Past work on material hardship and paternal incarceration leverages data prior to the economic downturn (Schwartz-Soicher, Geller, and Garfinkel 2011; Sugie 2012); the fact that these studies are based on a cohort of American children in large U.S. cities may mask the relative needs of youth at different ages in the life course and in nonurban areas. This article fills a gap in the literature by examining the association between parental incarceration and severe deprivation—material hardship, unmet health needs, and residential instability—as well as enlistment in social programs after the Great Recession. We situate our study in the existing literature on poverty, material hardship, and childhood disadvantage.
POVERTY, MATERIAL HARDSHIP, AND CHILDHOOD DISADVANTAGE
Poverty remains deeply entrenched in American culture. Rising income inequality since the 1970s has increased the overall poverty rate, concentrating economic hardship and deprivation in urban, metropolitan areas (Massey 1996). In 2012, 21.8 percent of children were poor, and because of high unemployment rates, economic growth has been less effective in reducing poverty than during previous historical periods (Danziger and Wimer 2014). The spatial concentration of poverty and residential segregation creates a permanent underclass by isolating inner-city residents from mainstream social institutions known to have an impact on their economic well-being (Massey, Condran, and Denton 1987; Massey and Denton 1993; [End Page 109] Wilson 1987). Increasing family and household complexity among children is, in part, a response to growing social inequality since the latter half of the twentieth century (Furstenberg 2014; Guzzo 2014; Manning, Brown, and Stykes 2014).
Growth in the penal system has accentuated the material deprivation of minors. Children of incarcerated parents now face social exclusion and many hardships, including homelessness (Foster and Hagan 2007; Wildeman 2014), food insecurity (Cox and Wallace 2013), and political disengagement (Foster and Hagan 2007). Critically, for our purposes, Holly Foster and John Hagan (2007, 411) find that youth classified as “socially excluded” owing to paternal incarceration have a 77 percent chance of lacking health insurance, and Amanda Geller and her colleagues (2009) show that children of incarcerated parents experience more economic and residential instability than their counterparts. The economic hardship of children exposed to parental incarceration is in part due to the low remittances, if any, from formerly incarcerated fathers, and the loss of additional contributions is principally driven by fathers residing outside the household (Geller, Garfinkel, and Western 2011). Other scholarship shows that incarceration places children at significant risk of material hardship using a variety of indicators that include residential instability, having utilities disconnected, having unmet medical needs, and receiving free food (Schwartz-Soicher, Geller, and Garfinkel 2011).
Official measures of the poverty rate do not account for many government benefits that the poor receive (Danziger and Wimer 2014). Although researchers have called for increased social service participation to ameliorate the growing poverty and inequality associated with parental incarceration (Foster and Hagan 2007; Geller et al. 2009; Schwartz-Soicher, Geller, and Garfinkel 2011), and past work documents that childhood poverty would have been significantly higher during economic downturns if not for social service programs (Fox et al. 2015), we know of no existing research that quantifies how parental incarceration affects program participation among children after a severe economic recession. This article examines severe deprivation through four measures of hardship and how enrollment in five different needs-based programs may attenuate levels of disadvantage.
Parental incarceration—as a stressful life event—is one of many factors that contribute to the social exclusion of children. Having a parent in prison or jail deprives household members of much-needed resources (Geller, Garfinkel, and Western 2011), producing strain and opportunities for youth to engage in crime and delinquency (Cloward and Ohlin 1960; Merton 1938). The adverse socioemotional conditions that children experience during their formative years accumulate across the life course to reinforce and extend individual and familial disadvantage (Sampson and Laub 1997), with delinquency serving as a transitional state that results from repeated stressful life conditions (Hagan and Foster 2003). Stressful events like parental incarceration are known to produce a host of childhood behavioral challenges (Wakefield and Wildeman 2011, 2013; Wildeman 2010), health limitations (Turney 2014), and residential instabilities, such as homelessness and foster care placement (Andersen and Wildeman 2014; Wildeman 2014).
Despite these hardships, one-half to two-thirds of fathers expect to live with their children and families when they reenter society (Foster and Hagan 2009), requiring a particular public policy intervention aimed at increasing both child well-being and familial connectedness during periods of parental absence and reunification. The positive or protective factors that may increase well-being among children of incarcerated parents have not been studied in great detail. Caroline Lanskey and her colleagues (2014) argue that much of the research around the well-being of prisoners’ children does not account for the intersectionality of time, space, and agency, and they contend that external manifestations of well-being (wealth, environmental conditions, and so on) are neither accounts of nor substitutes for the subjective experiences (happiness or life satisfaction) of youth. However, because the material [End Page 110] deprivation and insecure attachments that children experience are likely to affect their cognitive development and emotional states (Eddy and Poehlmann 2010), governmental intervention is needed to ensure minimal levels of subsistence.
Yet recent scholarship shows that the mark of a criminal record facilitates a behavioral change among prospective, current, and former inmates. Alice Goffman’s (2014) ethnographic work details how men on the run avoid institutions that may increase their likelihood of arrest and incarceration. Hospitals, for instance, are surveilled in ways that make health care utilization problematic for the families and friends of men who are and have been criminally involved. Sarah Brayne (2014) shows that such surveilling institutions (hospitals, banks, formal employment, and schools) produce a form of “system avoidance” by former inmates because they expose men with criminal records to increased risk of formal monitoring. Both of these works draw on data that were collected before the recent economic recession, and this body of research largely focuses on the avoidance by the person incarcerated. Thus, it is unknown whether system avoidance extends to children with a parent currently or previously incarcerated.
There are several theoretical reasons why parental incarceration should be associated with increased governmental assistance, especially after the Great Recession. First, Wilson (1987) originally posited that the underclass in urban areas endure heightened levels of poverty, imprisonment, and unemployment that place them at risk of increased welfare and social services. He showed that the United States relies on public assistance to aid poor families, contrary to social policy in some European nations, which supplement income through “incentives to work” programs that allow for “alternative income transfers” in the forms of family, housing, and un(der)employment allowances (Wilson 1987, 156–57). The poverty and joblessness of young, undereducated men who experience criminal justice contact trickle down into their families and communities (Western and Wildeman 2009). Thus, one should expect levels of government assistance and program participation to have been higher among children of incarcerated parents after the Great Recession.
Second, historical understandings of crime control implicate social workers in uncovering a host of poverty-related social ills within households. Garland (2001b) argues that the economic and social progress that the welfare state ushered in during the 1980s ultimately undermined the effectiveness and legitimacy of welfarist reforms. Institutions created to meet the housing, health care, educational, and social service needs of the poor found that the most-disadvantaged members of our society have a constellation of needs and not a singular problem (Garland 2001b). Thus, the tattered fabric of American life for residents in the urban core revealed severe deprivation requiring immediate government assistance, which increased budgets and made problems appear larger than previously imagined. Nearly thirty years later, the economic recession of the early twenty-first century would again usher in a moment of extreme need.
Although the Great Recession ended in June 2009, the poverty associated with it continues. The national poverty rate remained at 14.5 percent for 2013, with black and Hispanic poverty rates two to three times higher than rates for whites (DeNavas-Walt and Proctor 2014, 13). With the rise of punitive justice and increased social control in America, having a criminal record now threatens the ability of poor families to enroll in, or continue the benefits of, various public assistance programs (Sugie 2012; Uggen and McElrath 2014). Such barriers to government assistance for needy families compromise the general welfare of children and accentuate the existing hardships of disadvantaged families.
We posit three different hypotheses about inequality in deprivation and program participation. First, we anticipate that children of an incarcerated parent will be more likely to live in poverty and reside in a working poor household. Second, we hypothesize that differences in levels of poverty during the Great Recession exposed children with an incarcerated parent to greater levels of deprivation than children unexposed to parental incarceration, net of social background effects. Finally, we expect higher levels of program participation among children [End Page 111] who have experienced parental incarceration because their households and communities are more likely to be visited by social workers (Garland 2001b) and to receive assistance (Wilson 1987) to buffer or attenuate these forms of deprivation. Such findings would support a “system inclusion” perspective for children of the prison boom during the economic downturn.
We use data from the 2011–2012 National Survey of Children’s Health (NSCH) to investigate race and class differences in markers of deprivation and program participation by parental incarceration status. NSCH data are collected by the National Opinion Research Center (NORC) at the University of Chicago on behalf of the Centers for Disease Control (CDC). The NSCH randomly sampled (both cell-phone and landline) telephone numbers to locate households with children ages zero to seventeen, and within each household one child was randomly selected to be the subject of the interview. The landline and cell-phone samples have national response rates of 38.2 and 15.5 percent, respectively, and the combined dual-frame sample has a response rate of 23.0 percent. Differences between landline and cellphone samples do not present nonresponse bias between sampling methods because the assumed rates of eligibility as well as the definitions to determine eligibility differed for the landline and cell-phone samples. The NSCH interview completion rate, defined as the proportion of households known to include children that completed all sections, was 54.1 percent for the landline sample and 41.2 percent for the cell-phone sample.
Despite the low response rate, these data are appropriate for several reasons. First, while longitudinal data sources, like the Fragile Families and Child Well-Being (FFCW) study, follow a cohort of children, such data do not describe the current condition of material hardship or program participation for all children outside particular age groups for any calendar year. Cross-sectional data are necessary to investigate how parental incarceration is related to hardship and the social service needs for all minors. Second, the NSCH data are the only data available, to our knowledge, that allow for post–Great Recession analyses. The low response rates may mean that the estimates reported here are quite conservative. Thus, while the response rates are not ideal, there are very few alternative data sources that describe levels of deprivation and program participation for all children after the recent economic recession.
The NSCH asked the same demographic and health questions of all children; however, because children experience the social world at different stages of development, specific questions were asked about early childhood (under age five) and later childhood (ages six to seventeen). Parents were asked about family functioning, parental health, neighborhood and community characteristics, health coverage, and other demographic information. Children over age six were asked about self-control and educational attachment. Each interview lasted, on average, about twenty-seven minutes, and data were collected between February 28, 2011, and June 25, 2012. Over 95,600 child-level interviews were completed, with the number of interviews ranging from 1,800 to 2,200 per state. When weighted, NSCH results represent the social experiences and familial conditions of non-institutionalized minors.
There are several reasons why these data are appropriate for studying severe deprivation and participation in needs-based social programs for children who have a parent currently or formerly incarcerated. First, the data contain many measures of program participation and several markers of social disadvantage (such as parental stress, economic hardship, and parental incarceration). Second, the data are nationally representative of all youth and their socioeconomic experiences, although recent research shows that survey measures of parental incarceration in the NSCH are lower than estimates derived from population data (Sykes and Pettit 2014). These differences may matter for estimating the relative magnitude of the impact of parental imprisonment on childhood deprivation and program participation—that is, we may underestimate within- and between-group differences—but the effects of having a parent in prison or jail on measures of social disadvantage are in the anticipated direction. Our [End Page 112] estimates of racial and educational inequality are likely to be statistically conservative because the number of children with a parent in prison or jail is underreported. Lastly, the NSCH data provide a rich set of parental indicators that are known to affect enrollment in needs-based programs.
Table 1 displays the operationalization and coding of measures in our study. The central variables in our study are measures of deprivation and program participation among children of incarcerated parents. The NSCH data contain several markers of deprivation: material hardship, unmet health needs, and residential instability. Material hardship is measured as the parent often finding it hard to cover the basics like food or housing; over one-quarter of U.S. children resided in such a household in 2011–2012. Similarly, approximately 7 percent of children had an unmet or delayed health care need, with minors utilizing an average of two services in the previous year. Residential instability is defined as the number of times the child has ever moved; parents reported that their children had made, on average, two relocations.
The NSCH data also contain measures of enrollment in social services programs. The Children’s Health Insurance Program (CHIP), benefits from the Women, Infants, and Children (WIC) program, cash assistance from a state or county welfare program (Temporary Assistance for Needy Families [TANF]), Supplemental Nutrition Assistance Program (SNAP) benefits, and programs for free or reduced-price lunches at school are our measures of government assistance that constitute program participation. Table 1 shows that the percentage of American children enlisted for these benefits varies substantially by program. Roughly 40 percent of all children are enrolled in CHIP; 17 percent are on WIC; 7 percent are in households that receive welfare; one in four minors is a SNAP participant; and over one-third of all children receive a free or reduced-price lunch.1 These programs respond to primary forms of economic need and are necessary to attenuate material hardship and deprivation.
We construct two composite measures that summarize the overall system inclusion of children based on their economic need. Program participation is the sum of all programs for which the child or household has been enrolled. Using these variables, we create a scale of system inclusion that ranges from low (0) to high (5) based on the total number of programs (alpha = 0.73). Additionally, we create a measure of whether a child has enrolled in any program. On average, children are enrolled in 1.25 programs, with 51 percent of all youth exposed to at least one program.
According to NSCH data, roughly 7.1 percent of children have experienced the incarceration of a parent, almost double the percentage observed in other studies (Pettit, Sykes, and Western 2009; Wildeman 2009). Rapid changes in family life, data, methods, and social policy may partially explain discrepant estimates of parental incarceration in population and survey data (Sykes and Pettit 2014). At the same time, racial-ethnic and class-based inequality in parental incarceration is similar to that found in other surveys (Sykes and Pettit 2014; Wildeman 2009).
Race is measured as non-Hispanic white, non-Hispanic black, Hispanic, and non-Hispanic other. The racial distribution of children in our sample is broadly representative of American minors; roughly 53 percent of white, 14 percent of black, and 23 percent of Hispanic children are included in these data.
Educational attainment is operationalized as whether the mother has less than a high school education, has received a high school diploma, or has received some college education. [End Page 113]
[End Page 114]
The fraction of mothers with some college education is more than twice the percentage of women with a four-year degree or more (U.S. Census Bureau 2012). These differences may reflect either delays in the timing of motherhood due to college enrollment or the growth in two-year and nontraditional educational programs for returning students.
Additional measures of social background reflect, to varying degrees, the social inequality in contemporary America. Past research shows that social institutions, neighborhood disorder, [End Page 115] and social cohesion are relevant to understanding inequality in residential stability and crime reduction, particularly within communities exposed to high levels of incarceration (Patillo 1998; Peterson, Krivo, and Harris 2000; Sampson, Raudenbush, and Earls 1997; Wilson 1987; Wilson and Kelling 1982). Therefore, we construct four distinct measures of social and neighborhood background to include as controls. “Social institutions” include scaled measures of whether the neighborhood has sidewalks, playgrounds, recreational centers, and libraries (alpha = 0.64); “neighborhood disorder” is the count of whether litter, dilapidated housing, and broken windows are present in the neighborhood (alpha = 0.59); “social cohesion” measures the connectedness of residents through their ability to help each other, watch each other’s kids, trust one another, and count on each other (alpha = 0.82); and “concentrated disadvantage” contains six measures of adverse childhood experiences and parental stress related to caring for the child (alpha = 0.52).2
We include an indicator of full-time employment that represents whether a parent or guardian was employed fifty out of fifty-two weeks in the past year. Approximately 84 percent of children live in households that had at least one adult employed for the full year, and nearly one in four minors live in poverty. The survey also includes a measure of the working poor, which is defined as a parent or guardian who is employed full-time but lives below the poverty line. According to these data, over one in eight children in America live in a working poor household.
We employ two different methods to investigate the relationship between parental incarceration and our measures of severe deprivation and program participation. First, we fit two incarceration-specific probit models to estimate racial differences in the likelihood that nonwhite children will experience program participation. We report marginal effects, which express the rate of change in the dependent variable (the predicted probability) relative to a unit change in an independent variable (Long 1997; Powers and Xie 2000). All models are evaluated at their mean values.
Second, because the likelihood of experiencing parental incarceration is not randomly distributed across the population, we use propensity score matching methods to reduce the bias associated with observable social background characteristics. The propensity score is the conditional probability of having a parent incarcerated given a set of demographic, social background, and labor market covariates that predict severe deprivation and program participation and are also potential confounders in the association between parental incarceration and material hardship. The method balances the background characteristics of children exposed to parental incarceration with the characteristics of children unexposed to parental incarceration to ensure that any differences in severe deprivation and enlistment in social programs are not due to significant variation in the likelihood of having a parent in prison or jail (that is, to ensure that there are no “pretreatment” differences, in the language of Paul Rosenbaum and Donald Rubin 1983, 1984).
SEVERE DEPRIVATION AMONG AMERICAN CHILDREN
Figure 1 plots racial and educational inequality in the percentage of minor children living in poverty and working poor households by parental incarceration status. This two-by-two figure displays the measure of hardship (poverty and working poor) against the parental incarceration status by levels of educational attainment for each racial or ethnic group. The top half of the graph measures poverty, while the bottom half measures the percentage of children in households designated as the working poor. The left side of the figure is for children who have never had a parent incarcerated, while the right side of the graph depicts the proportion of children who had a parent incarcerated.
Overall, children who have had a parent incarcerated are twice as likely to live in poverty [End Page 116] as children who have not experienced parental incarceration. Roughly one in five children who have not experienced parental incarceration live in poverty, compared to 40 percent of minors who have experienced parental incarceration. Pooled estimates conceal extraordinary racial and educational inequality in the experience of parental incarceration. Among children who have not had a parent behind bars, black children at every educational level have the highest likelihood of living in poverty, followed by Hispanic and white children.
Race and educational gradients in poverty are so profound that the overall percentage of children living in poverty with a parent incarcerated (40 percent) is on par with the percentage of white children living in poverty with a parent who has not been incarcerated but has less than a high school diploma or the percentage of black or Hispanic children living in poverty with a parent who has not been incarcerated but has a high school degree.
The bottom half of figure 1 displays the relative percentage of children who have a parent designated as working poor. One in eight children who have not had a parent incarcerated live in working poor homes, compared to almost one in four minors who have had a parent [End Page 117] behind bars. Interestingly, for African American children with mothers who have low levels of education (a high school diploma or less), there is no difference in living in a working poor household between children exposed and unexposed to parental incarceration; roughly 39 percent of black children with a mother who dropped out of high school live in working poor homes. This is the highest percentage observed for any of the race-education groupings. Higher levels of maternal education are associated with a lower fraction of children living in working poor households in both panels. However, there appears to be racial inequality among children classified as working poor who live with parents with some college education if those parents spent time behind bars. The wage and employment inequality associated with having a criminal record (Pager 2007; Pettit 2012; Western 2006; Western and Pettit 2005) suggests that children of incarcerated parents face heightened risks of severe hardships that may require increased social service resources.
Figure 2 displays the racial differences in material hardship of minor children by parental incarceration status. Consistent with Sara Wakefield and Christopher Wildeman’s (2013, 51) findings, having a parent incarcerated is associated with greater economic hardship. Approximately one in four children without an incarcerated parent encounter material hardship, compared to 46 percent of children who have experienced parental incarceration. White children of current and former inmates are 27.2 percentage points more likely to live in households experiencing material hardship than similarly situated youth whose parents [End Page 118] have not come into contact with the criminal justice system. For black and Hispanic children, percentage-point differences in material hardship between children with and without parents behind bars (15.3 and 12.5, respectively) are much lower, largely owing to the greater levels of hardship among children of never-incarcerated parents. Parental incarceration appears to worsen material hardship above and beyond the levels experienced by nonwhite minors.
Figure 3 displays estimates of the percentage of minor children with unmet and delayed health needs. Overall, children with an incarcerated parent are almost twice as likely (12.3 versus 6.3 percent) to have an unmet or delayed health need than children without a parent in prison or jail. Racial differences abound. Almost one in eight black children with an incarcerated parent have an unmet or delayed health need, compared to one in eleven African American children without a parent behind bars. Yet white children have the largest absolute difference in unmet health needs due to parental incarceration (5.9 percentage points).
Goffman’s (2014) and Brayne’s (2014) finding that hospitals are surveillance sites for criminal justice agents indicates that health care services will be underutilized by men who have been or are likely to be incarcerated. Yet no scholarship has examined whether health care service deprivation extends to the lives of children exposed to parental incarceration. Figure 4 shows that the average number of health services utilized by children with an incarcerated parent is slightly greater than and significantly different from the average for youth without a parent entangled in the criminal [End Page 119] justice system (b = 0.14, p < 0.001). This finding suggests that surveillance within health care institutions is limited to parents with a criminal or arrest record and does not extend to their children.
Residential stability is important for fostering increased attachments to neighborhoods, peers, and institutions, in addition to greater physical and mental health. Sarah Burgard, Kristin Seefeldt, and Sarah Zelner (2012) show that housing instability increases levels of depression and anxiety among renters, the homeless, and mortgage holders who are behind on payments. Wakefield and Wildeman (2013) show that paternal incarceration increases the risk of homelessness. However, there is no research that documents how increased foreclosures and greater unemployment during and after the Great Recession have affected the residential stability of children with an incarcerated parent. Figure 5 highlights the variation in residential instability among American children. Children who have experienced parental incarceration make twice as many moves, compared to children of parents never incarcerated. White youth who have a parent with a criminal record moved the most—4.26 times, on average—followed by Hispanics (3.75 moves) and blacks (3.53 relocations).
We assess whether these patterns of severe deprivation hold after addressing selection into families at risk of parental incarceration. Table 2 displays matched estimates of the effect of parental incarceration on measures of deprivation. Even after balancing our analysis for social background differences, we continue to find disparities in markers of need. Material hardship—that is, often finding it difficult to cover the basics like food or housing—is 17.9 percentage points higher [End Page 120]
[End Page 121]
among children who have had a parent incarcerated. Similarly, children exposed to parental incarceration are 4.8 percentage points more likely to have an unmet or delayed health need despite their slightly greater utilization of health services. Yet they experience pronounced residential instability. After adjusting for social background bias between minors exposed and unexposed to parental incarceration, children of the prison boom are associated with 1.7 more moves (or twice as many relocations), on average, than children who never had a parent under correctional supervision.
SYSTEM INCLUSION AMONG THE DISADVANTAGED
The foregoing analyses illustrate severe deprivation across a variety of measures. Yet policy findings are thin on how children with an incarcerated parent may have been included in social programs during the Great Recession or excluded from them. Although some research indicates that prospective and former inmates may avoid system involvement to circumvent additional surveillance, there is no evidence to suggest the same is true for their children, particularly if the other biological parent or residential guardian has no criminal record and requires program assistance to attenuate household material hardship.
Figure 6 presents estimates of the percentage of minor children who participate in social services programs on at least one measure. About half of all children who have not experienced parental incarceration are enrolled in at least one social program. Among whites, nearly one-third participate in a program, compared to 70 percent of black children and 75 percent of Hispanic children.
[End Page 122]
Having an incarcerated parent is associated with a higher likelihood of experiencing at least one measure of program participation. Overall, roughly 85 percent of all children who have experienced parental incarceration receive at least one form of social support. Roughly eight out of ten white children are enrolled in at least one program, and that figure rises to around 91 percent for Hispanic and black children.
Figure 7 displays the average number of social programs that minor children are enrolled in by race and parental incarceration status. Children who have not had a parent in jail or prison participate in 1.1 programs on average. White children are enrolled in less than one program, while black and Hispanic children are enrolled in an average of 1.8 and 1.9 services, respectively.
Children with an incarcerated parent participate in more social programs than youth who have not had a parent in prison or jail. Overall, minors who have had a parent behind bars have program enrollment levels twice those of children with never-incarcerated parents. Again, there is striking racial inequality in government assistance programs. White minors participate in 2.0 social programs, on average, while black and Hispanic youth participate in 2.6 and 2.5 programs, respectively.
Racial differences in system inclusion may be explained by factors beyond parental incarceration. Thus, we estimate incarceration-specific probit models to examine racial differences in participation rates. Table 3 presents the marginal change in the probability of a child being enrolled in any governmental program, net of social background characteristics. [End Page 123] Model 1 includes a limited set of controls (race, maternal education, age of child, sex of child, and parental employment status), and model 2 further adjusts for community institutions, neighborhood disorder, social cohesion, concentrated disadvantage, and other social background characteristics. In all models, white children are the reference group.
Among children who have not experienced parental incarceration, nonwhite children are significantly more likely to participate in social programs than white youth. Blacks and Hispanics are 36.9 and 21.7 percentage points more likely, respectively, to have received government aid. Including the additional set of controls (model 2) reduces these point estimates by 21.7 percent for blacks and by 18.9 percent for Hispanic children, but the differences remain statistically significant in comparison to white minors.
Racial differences between youth with an incarcerated parent are smaller, in part because the overall baseline level of participation is much higher than that for children who have not experienced parental incarceration. Black children with an incarcerated parent are 9.1 percentage points more likely than whites to experience any form of deprivation; Hispanic youth are at a slightly greater risk, with a 10.2 percentage-point difference. Again, the ecological contexts in which children reside (model 2) explain roughly two-fifths (39.2 percent) to over one-half (50.5 percent) of the point estimates in model 1 for Hispanics and blacks, respectively.
Percentage differences in point-estimate reductions between models 1 and 2 for children who did and did not experience parental incarceration suggest that strong selection effects may be operating. Table 4 investigates this line of inquiry and its implications for understanding how the exposure and extent of program participation is associated with parental incarceration. We estimate the effect of parental incarceration on social program involvement after balancing pretreatment differences in background characteristics among children. Unmatched estimates provide a relative baseline for understanding how selection matters in differences between children with and without an incarcerated parent.
Children who experience parental incarceration are significantly more likely to experience at least one measure of assistance. Unmatched estimates indicate that minors of incarcerated parents are 44.1 percentage points more likely to enroll in one program, relative to children who do not have a parent in the correctional system. Matching on observable characteristics reveals that children who have had a parent behind bars are 35.6 percentage points more likely to report participation in at [End Page 124] least one program. Social background differences between these groups account for 19.3 percent of the difference between matched and unmatched estimates. Using information contained in table 4, we estimate that of the 2.6 million minors who experienced parental incarceration in 2012 (Sykes and Pettit 2014), at least 2.10 million (2.6*0.808) of them are enrolled in at least one social program.
The severe deprivation experienced by children of incarcerated parents is profound. On average, these children participate in 2.04 social programs, whereas children who do not experience parental incarceration participate in 1.06 needs-based programs. The overall difference is that children of incarcerated parents have significantly higher program needs—they are enrolled in almost 1.24 more programs on average. Matching on observed characteristics reduces this difference to almost one program, indicating that balancing social background effects reduces bias in parental incarceration differences by 20.9 percent.
We also investigate program involvement among children of parents who work but remain in poverty. Table 5 presents matched estimates of government assistance enrollment by parental incarceration status among households designated as the working poor. Among non–working poor households, children of incarcerated parents participate in 0.31 more social programs, on average, than children without an incarcerated parent. Minors from working poor families have much lower program participation rates; however, the absolute difference between the children exposed and unexposed to parental incarceration is larger (0.39 programs) among the working poor because the non-incarcerated group has much lower program enrollment. Nevertheless, parental incarceration remains significantly associated with participation in more needs-based programs.
Perhaps the most important finding comes from our measure of any program participation. Among the non–working poor and working poor households, between 94 and 99 percent of children participate in at least one social program. In both types of households, having an incarcerated parent increases program participation by two to four percentage-points. This finding suggests that for the most economically disadvantaged children in America, program participation continues to depend on parental incarceration, although to a lesser extent. In fact, parental incarceration may have drawn children from the margins and into government assistance to alleviate the deprivation and material hardship associated with the Great Recession.
Finally, we examine whether social service outreach increases program enrollment [End Page 125] among children who are impoverished or have experienced parental incarceration, as posited by Wilson (1987) and Garland (2001b). Table 6 tests this proposition by examining the percentage of children under age three who have been visited by social service workers. Overall, children of incarcerated parents receive more social service visits than children who have not experienced parental incarceration. Almost two-fifths of black youth (38.4 percent), nearly one-quarter of Hispanic youth (24.7 percent), and over one-fifth (22.6 percent) of white youth with a parent under correctional supervision have had a nurse, social worker, or health professional visit their home, compared to 16.3, 12.1, and 12.9 percent of black, Hispanic, and white children unexposed to parental incarceration, respectively.
Maternal educational attainment reveals stark differences in social service visitation. Among children without a parent exposed to criminal justice contact, the educational patterns [End Page 126] are consistent with theoretical expectations: minors in households with lower levels of maternal education are more likely to have home visitations. Yet, among youth who have a parent under correctional supervision, the educational patterns vary by race. For instance, there is little difference between white children with a mother who has a high school diploma and those whose mother failed to complete high school. At the other extreme, Hispanic children with mothers who completed high school have the highest visitation rates relative to those whose mother has no high school degree or those whose mother has some college education. It is unclear, however, why Hispanic children in households where the mother has no high school degree have the lowest levels of visitation (6.4 percent). One possibility is that these families are avoiding the double jeopardy associated with both immigration and criminal justice surveillance. Lastly, there are large educational gradients for black children: children of mothers who dropped out of high school have the highest rates of visitation (53.3 percent), followed by those whose mother is college-educated (40.2 percent) or a high school graduate (34.5 percent).
CONCLUSIONS AND IMPLICATIONS
Michel Foucault (1977, 268) famously wrote that the prison system creates delinquents “by throwing the inmate’s family into destitution,” in part, through child abandonment, vagabondage, and familial begging. Although some scholarship indicates that there is system avoidance among prospective and former inmates (Brayne 2014; Goffman 2014), our findings are broadly consistent with the theories and historical accounts of Wilson (1987) and Garland (2001b): severe deprivation and economic disadvantage are triggers for system inclusion through government programs that mollify the harsh economic conditions of the poor and marginalized. Children of current and former inmates greatly participate in these social programs, particularly in the wake of the Great Recession. Our findings indicate that the criminal justice surveillance of former inmates does not upend the inclusive nature of program participation among children of incarcerated parents. In fact, the enrollment of children in government programs may expose and extend services to former inmates depending on their state of residence. A recent Pew report finds that states are beginning to lift bans on the receipt of food stamps and welfare for drug felons, with a majority of states having partial or no bans for former felons (Beitsch 2015).
Racial and educational differences in incarceration among adults have significant implications for understanding the persistence of poverty and intergenerational inequality. A significant body of work now highlights growing race and class disparities in the likelihood of incarceration (Pettit 2012; Pettit and Western 2004; Tonry 1995; Western 2006), and recent scholarship has linked the expansion of the criminal justice system to growing inequality in the risk of having a parent incarcerated (Pettit, Sykes, and Western 2009; Wildeman 2009). Yet researchers are only beginning to understand how these inequalities cascade across generations and structure particular forms of disadvantage for children later in life (Wakefield and Wildeman 2011, 2013).
In this article, we investigated how parental incarceration has affected exposure to and levels of deprivation and program participation among children in America since the Great Recession. Using data from the NSCH, we show that children who have experienced parental incarceration exhibit elevated rates of deprivation and greater involvement in a range of needs-based programs. Clear majorities of children who have had a parent incarcerated are involved in at least one needs-based social program (such as CHIP, TANF, free or reduced-price lunch, WIC, and SNAP), and many experience material hardship, unmet health needs, and increased residential instability. Black and Hispanic children who experience parental incarceration are the most likely to enroll in government assistance programs, but it is white children with a parent in prison or jail who encounter slightly greater material deprivation and residential instability.
Although our findings reveal systematic differences in children’s social program inclusion based on whether they had a parent incarcerated, our results cannot speak to the [End Page 127] practical effects of such inclusion on the poverty status of disadvantaged children or the material conditions of their lives. Nor can our findings address the questions raised by Goffman (2014) and Brayne (2014) about whether increased contact with welfare and social programs in fact leads to increased criminal justice surveillance and ultimately to an increased risk that children, like their parents, will become incarcerated. These limitations render the normative implications of our findings unclear and somewhat unsettling. Although increased system inclusion of the nation’s most-disadvantaged children seems promising for the larger goal of alleviating child poverty, the specter of mass incarceration looms large: it may be that the very social programs offering a path out of poverty simultaneously set a path leading to more formal surveillance into adulthood. Research shows that the coexistence between welfare and penal states is fundamentally orthogonal within modern industrialized societies (Beckett and Western 2001; Garland 1985; Wacquant 2010), raising the possibility that the most- disadvantaged families in America face a tragic dilemma: the price they must pay for inclusion in the social safety net may further their risk of getting caught in the widening carceral net. To this point, Kaaryn Gustafson (2011) illustrates the intersectional nature of welfare and penal institutions when poverty becomes criminalized because some social welfare recipients employ devious methods to cheat the system. Other impoverished families, however, pay the price of punitive social welfare policies meant to curtail the fraudulent behavior of some program participants.
This article joins a chorus of recent research drawing attention to the collateral consequences of mass incarceration for families and their children. Incarceration levies significant penalties against those who spend time behind bars. Yet the costs of incarceration extend to the children of current and former inmates, who are at heightened risk of severe deprivation. Although many of the children who experience parental incarceration may have already been at risk of severe deprivation, our research indicates that incarceration independently and significantly contributes to children’s need for social support to meet basic living standards.
Our findings are important and timely for several reasons. First, the American public is deeply concerned that government policies since the beginning of the Great Recession have done little to help the poor and the middle class. A recent Pew report shows that 71 percent of Americans believe that government economic policies since the recession have helped large banks and financial institutions a fair amount or a great deal, and roughly two-thirds of individuals polled said the same about laws and programs that have enabled large corporations to navigate difficult financial times. By comparison, 72 percent say that economic policies since the recession have done little or nothing to help the middle class, and slightly fewer (65 percent) feel the same when asked about the poor (Pew Research Center 2015). In light of childhood poverty associated with having an incarcerated parent, social policies must be expanded to aid the children of poor families, including those who are considered the working poor.
Second, our findings highlight the importance of continued government spending on social programs aimed at attenuating poverty and deprivation. Yet recent debates in Congress have focused on the need to reduce spending on Medicare and the food stamps programs that were expanded during the Great Recession (CBS 2015; Tracy 2015). The proposed cuts are in addition to the $8.7 billion food stamp reduction contained in the 2014 farm bill, which resulted in the loss of about $90 per month in food stamp benefits for 850,000 households (Resnikoff 2014). These cuts risk deepening and entrenching poverty among the children of ex-prisoners currently being served by such programs.
Children’s deprivation and program involvement are largely driven by both parental incarceration and growing economic inequality: children of the working poor who have not had a parent in prison or jail are suffering fates similar to those of their counterparts with a parent in prison or jail. Inmates are drawn from the most-disadvantaged segments of society and are often banished from particular domains of social life (Beckett and Herbert 2009). Thus, it should come as no surprise that [End Page 128] children of inmates also experience agony and distress. But parental incarceration is more than a symbol of disadvantage. Parental incarceration serves to crystallize social inequality by exposing children to additional risks, hardships, and severe deprivation that may further fuel the intergenerational transmission of disadvantage.
Bryan L. Sykes is assistant professor of criminology, law, and society at the University of California, Irvine.
Becky Pettit is professor of sociology at the University of Texas at Austin.
An earlier version of this article was presented at the 2014 Russell Sage Foundation conference “Severe Deprivation in America.” We thank Sheldon Danziger, Matthew Desmond, anonymous reviewers, Anjuli Verma, Naomi Sugie, participants of the Socio Legal Workshop (SLW) at the University of California, Irvine, and the “Severe Deprivation” conference participants for suggested revisions on previous drafts. Direct correspondence to: Bryan L. Sykes at email@example.com, 3317 Social Ecology II, Criminology, Law and Society, University of California, Irvine, Irvine, CA 92697; and Becky Pettit at firstname.lastname@example.org, Department of Sociology, The University of Texas at Austin, 305 E. 23rd St., A 1700, CLA 3.306, Austin TX 78712.
1. Unfortunately, the U.S. Department of Agriculture (USDA) Food and Nutrition Service (FNS) reports program data in aggregate counts of children who receive WIC and SNAP benefits (see USDA, FNS, “Overview,” at: http://www.fns.usda.gov/pd/overview). However, the USDA does report the percentage of children who receive reduced-price or free lunches—8.6 percent and 59.6 percent, respectively, in 2012. Our estimate of 33 percent is between these figures, suggesting that the NSCH question on food limitation, as an aggregate of these two programs, is reasonable given the unknown distribution of children receiving free or reduced-price lunches.