Anti-poverty Policy Innovations: New Proposals for Addressing Poverty in the United States

The 2016 presidential election has brought to the fore proposals to fundamentally restructure the U.S. anti-poverty safety net. Even though much of the current debate centers on shrinking or eliminating federal programs, we believe it is necessary and useful to explore alternatives that represent new approaches and significant innovations to existing policy and programs. This double issue of RSF: The Russell Sage Foundation Journal of the Social Sciences builds on and extends the scholarly conversation on the state of current U.S. anti-poverty policy by highlighting a collection of related innovative and specific policy proposals for the United States. Well before the election, the authors of the articles in this volume were explicitly tasked with proposing substantially new policies solidly grounded in social science evidence that have the potential to transform anti-poverty policy. Assuming the goal to be reducing poverty among the U.S. population, we asked what new ideas should be seriously considered. The authors responded with carefully crafted proposals that tackle poverty from a variety of perspectives. Some of these proposals are more of a departure from existing policies than others, some borrow from other countries or revive old ideas, some are narrow in focus and others much broader, but all seek to move anti-poverty efforts into new territory.

Background

Just over fifty years ago, the War on Poverty marked a significant expansion of the scope and scale of anti-poverty programs, as well as a considerable change in their financing and administration. The federal government required, and fully or substantially funded, new entitlement programs—including Food Stamps and Medicaid—and a broad range of related programs and services such as Head Start, Legal Services, and Job Corps. From the 1960s until the mid-1990s, most changes to anti-poverty programs were arguably incremental, although there were notable exceptions including the establishment and major expansions of the Earned Income Tax Credit (EITC) and the development of the child support enforcement system. Passed in 1996, the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), along with accompanying expansions in the EITC and childcare subsidies (administered through the Child Care Development Block Grant, CCDBG), represented a significant redirection for anti-poverty policy. Perhaps most notably, it eliminated the entitlement to cash assistance provided by Aid to Families with Dependent Children (AFDC) and replaced it with Temporary Assistance for Needy Families (TANF), a time-delimited benefit contingent on meeting work requirements. It also decoupled Food Stamps and Medicaid from cash welfare.

Advocates suggested that the 1996 welfare reform, by eliminating the entitlement to cash assistance and freeing states to substantially restructure their welfare program, would lead to fifty active laboratories of innovation—states experimenting with different approaches to helping low-income families, and the best models being disseminated and adopted. Although clearly variation in state program characteristics is greater under TANF than AFDC, the most striking result of the change in rules and funding may be the shrinking proportion of poor families who participate in TANF and the declining share of program funds spent on cash assistance and employment services, despite stubbornly persistent levels of poverty among vulnerable populations as assessed by the official poverty measure, which includes only pretax cash income. As we write this introduction in the first months of the Trump administration, attention has turned away from expansive proposals for new government programs toward greater reliance on market-oriented approaches to poverty, an approach embraced by the Clinton administration in the 1990s. The Republican-controlled Congress and recent cabinet appointees advocate dismantling the Affordable Care Act and are considering restrictions or block grants to replace the Supplemental Nutrition Assistance Program (SNAP, or Food Stamps), and Medicaid.

It is not only federal policy that shapes poverty policy. State- and local-level support for government anti-poverty programs and market-oriented interventions varies considerably. The variation across states in the character of many major programs is significant—from the expansion of Medicaid eligibility (Rose 2015; Courtemanche et al. 2017; Buettgens, Holahan, and Recht 2015), to the availability of a state EITC (Williams 2017; Cooper, Lutz, and Palumbo 2015), to variation in the scope and generosity of cash benefits under TANF (Schott, Pavetti, and Floyd 2015; Floyd 2017). Major increases in the minimum wage (Autor, Manning, and Smith 2016; Neumark 2015), as well as paid sick leave (Isaacs, Healy, and Peters 2017; Ahn 2016), and fair scheduling (National Women’s Law Center 2017) continue to garner support in many states and localities. In addition, some states have invested more heavily in supporting low-income students who attend postsecondary education or training (Barr and Turner 2013; U.S. Department of Education 2016; McLendon and Perna 2014), as well as providing support for the youngest learners through universal or targeted prekindergarten programs (Bartik and Hershbein 2017; Friedman-Krauss, Barnett, and Nores 2016).

The fractious nature of national and state politics—reflecting stark differences in world views between the politicians in our major political parties—makes even modest policy changes that require legislative approval challenging. Nonetheless, given the clear need for better policy options, we believe that it is worth-while—and, indeed, necessary—to propose, develop, and refine innovative anti-poverty policies. The early months of the Trump administration illustrate the challenges of major social policy innovation, even when one party controls the White House and both houses of Congress. On the other hand, signals are clear that states will be given more latitude, along with the risks and potential for innovation that implies. For example, President Trump has ordered that all agencies consider “whether some or all of the functions of an agency . . . are appropriate for the federal government or would be better left to state or local governments or to the private sector” (White House 2017), and the administration has signaled enthusiasm for state waivers, for example, related to Medicaid.

The current political uncertainty makes it difficult to judge the scope and most likely context for potential change. We argue that this is an appropriate juncture to again consider innovations in anti-poverty policy that push beyond marginal changes to existing programs to consider new and different approaches to the major challenges that persist despite fifty years of focused anti-poverty policy. This is what the articles in this volume aim to do. Not even a double issue can hope to touch on all the critical components of a comprehensive anti-poverty strategy, and many issues—including issues related to immigration, incarceration, and health care, for example—are not systematically addressed here. A range of innovative policies addressing income support, employment, housing, and education and training, among other topics, are included, however. Before turning to the specific policy proposals, we provide context by reviewing current and expected poverty-related trends, evidence on the causes and consequences of poverty, and evidence on existing anti-poverty policies.

Defining poverty

The word poverty brings to mind many differing images, and has been used to describe a variety of contexts of scarcity. In public conversations, poverty typically refers to a lack of economic resources; sometimes, however, it is defined more broadly as social exclusion (particularly in the European context). For some, it evokes images of poor children and families from economically developing countries, who struggle to meet their most basic needs. Yet, even in a nation as wealthy as the United States, the word characterizes the living conditions of a substantial share of the population. The overall economic conditions in the United States have cycled between growth and recession, but even the extensive economic expansion of the past seventy-five years has failed to lift millions of citizens out of poverty.

Measuring poverty with economic resources is complicated because it requires defining both which types of resources should be counted and the minimum threshold below which individuals and families should be deemed to have insufficient resources. For poverty scholars, the term poverty in the United States has a very specific meaning. In the 1960s, the U.S. federal government developed a method for generating a dollar amount of pretax cash income that, if not exceeded, could be used to designate an individual or family as poor. The resulting poverty thresholds, which differ according to family size, are used for tracking trends in poverty rates. They also inform the poverty guidelines issued each year by the U.S. Department of Health and Human Services, which are used for determining social program eligibility. The poverty thresholds have been updated annually using the Consumer Price Index (CPI) to track inflation. In 2016, the official poverty guideline was just over $24,000 for a family of two parents and two children, and just under $12,000 for an individual adult living alone (U.S. Department of Health and Human Services 2016).

Concern that the official poverty measure (OPM) was outdated—because it fails to account for contemporary family expenses and in-kind public benefits and tax transfers, which have increasingly become the primary means of combatting poverty, and because it does not take into account geographic differences in the cost of living—led the Census Bureau to create a supplemental poverty measure (SPM). The SPM differs in several ways from the OPM, including the measure of resources, the measure of need, the household members whose resources and needs are considered, and adjustments for geographic differences (Renwick and Fox 2016). Perhaps most importantly for this discussion, the SPM considers post-tax income, and includes noncash benefits. Thus, unlike the official measure, the SPM accounts for income from the EITC and the Child Tax Credit, as well as the value of SNAP and Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) benefits, in addition to direct cash assistance (such as TANF and Social Security). Especially for low-income families with children, who increasingly receive assistance from tax credits and SNAP rather than from TANF, the SPM provides a better indicator of poverty and the effectiveness of current policy—as well as of the potential effects of the proposals in this volume that center on tax credits and in-kind benefits. In addition, the SPM accounts for work-related expenses, out-of-pocket medical expenses, and child support paid to other households. It also takes into account the incomes of spouses and cohabiting partners, and considers all resident children regardless of their relationships to the household head or heads. It therefore provides a more complete accounting of household resources than the OPM. Moreover, it uses poverty thresholds that are updated to reflect the current cost of a basic set of necessities, with different thresholds for different living arrangements, for renters versus owners, and for different cost of living levels across geographic areas (Garner 2010; see also Fox et al., “Waging Wars,” 2015; Wimer et al. 2016). Given these advantages, this double issue’s final article by Christopher Wimer, Sophie Collyer, and Sara Kimberlin (2018), which estimates the effects of many of the proposals on poverty, relies principally on the SPM. However, here and in other articles, we also reference the OPM, given its continued prominence and importance for policy.

Poverty trends

The official poverty rate fell precipitously during the 1960s. Since the 1960s, it has fluctuated between about 11 percent and 15 percent, increasing during economic downturns, and decreasing during times of economic expansion. In 2015, about 43.1 million individuals (13.5 percent) lived in poverty, as measured by the OPM (Proctor, Semega, and Kollar 2016); this was a decline from the peak of the Great Recession in 2010 (15.1 percent). The supplemental poverty rate indicated that 45.7 million people were poor in 2015, a rate slightly higher than the official measure (14.3 percent), reflecting, among other things, that SPM thresholds are generally higher than the official poverty thresholds.

Average poverty rates, however, mask considerable variation in poverty across populations of interest. Reflecting historical and current experiences of oppression and discrimination, African Americans and Hispanics face considerably higher rates of poverty than non-Hispanic whites (26.2 percent and 23.1 percent, respectively, compared with 10.1 percent). Moreover, children have significantly higher rates of poverty (21.1 percent) than adults (13.5 percent) or the elderly (age sixty-five or older; 10.0 percent). Poverty also differs by nativity (14.2 percent for native born, 18.5 percent for foreign born), family structure (28.2 percent for single female-headed households compared with 6.2 percent for households headed by a married couple), and educational attainment (28.9 percent for those without a high school degree compared with 5.0 percent for those with at least a bachelor’s degree) (Proctor, Semega, and Kollar 2016). Finally, official poverty rates differ by location. Poverty is higher in the South and West compared to the rest of the United States, and in urban and rural areas, compared to suburban areas.

Most individuals and families who experience poverty do so for a short time. Data from 2009 to 2012 suggest that more than 30 percent of the population experienced a spell of poverty lasting two or more months during this period (Proctor, Semega, and Kollar 2016). However, many poverty spells are short lived; less than 3 percent of the population experienced poverty in all forty-eight months of this period. Of course, persistence of poverty differs across sociodemographic groups too. For example, whereas only about 10 percent of all children experience persistent poverty throughout childhood (for half or more years from birth to age eighteen), 37 percent of African American children do (Ratcliffe and McKernan 2013).

Finally, some scholars and policymakers argue for more attention on those at the very bottom of the U.S. income distribution—the highly disadvantaged. This group received attention following welfare reform in the 1990s, with a focus on welfare-leaving families who did not find stable work and had limited public supports available to them. Since then, the highly disadvantaged groups of interest have broadened to include individuals and families in “deep poverty” (below 50 percent of the poverty line), as well as those experiencing “disconnection” from employment, schooling, and public assistance, low food security, other forms of severe material hardship (such as housing instability, eviction), and the $2-a-day poor (Edin and Shaefer 2015). These populations may face economic hardship that is acute, compounded across dimensions, and persistent over the life course or even generations (Desmond 2015; Seefeldt 2016). Estimates of the size and growth of this population differ depending on the measures used (DeNavas-Walt and Proctor 2015; Sherman and Trisi 2015; Short 2015). However, the bulk of research suggests a growth in deep poverty in the past twenty years as measured by the official federal poverty measure (Fox et al., “Trends in Deep Poverty,” 2015; Shaefer and Edin 2013; Shaefer, Edin, and Talbert 2015; but, for a substantially different conclusion, see also Winship 2016). Although additional research is required to refine estimates of the size of the highly disadvantaged population, it is clear that a significant number of Americans are living on very little cash income.

Causes and consequences of poverty

Two key institutions that shape economic fortunes—the labor market and the family—have dramatically changed over the last half-century in ways that leave large segments of the population increasingly vulnerable to poverty and its effects. The low-wage labor market has been characterized by stagnation with little growth in wages and few opportunities for advancement (Osterman 2014). At the same time, young adults with low levels of education have increasingly transitioned into parenthood in the context of unmarried romantic partnerships that often dissolve shortly after their child’s birth. We discuss these important and related changes and what it means for the next generation of children born into vulnerable economic conditions.

For many, falling into or avoiding poverty largely turns on success in the labor market (Fox et al., “Trends in Deep Poverty,” 2015; Shaefer, Edin, and Talbert 2015). Employment challenges faced by less-educated workers are both structural and cyclical (Autor 2010; Farber 2011). Changes in the structure of the economy have diminished the importance of the manufacturing sector, traditionally a source of relatively high-wage jobs for men with low levels of education (Autor 2010) and, more recently, the housing crisis during the Great Recession diminished the construction industry as a source of new jobs (Glaeser 2010). Globalization, skill-biased technical change, and changes in union influence have also reduced employment and wage-growth opportunities for less-educated workers. Job growth for workers without a college education is now concentrated in the low-wage personal service sector (Damme 2011). These changes have resulted in stagnant earnings for less-educated workers and limited their ability to earn their way out of poverty.

Between 1990 and 2005, poverty was characterized more by low wages than by joblessness, but the picture has changed since 2007 (Smeeding 2006). Although low wages are still an important factor, unemployment and unstable work are primary causes of non-elderly poverty today (Levy and Kochan 2012). In 1998, about 67 percent of the U.S. population sixteen years of age or older was participating in the labor force; this rate declined significantly during the Great Recession, and has not increased much since, rates remaining no higher than 63 percent since 2014. Furthermore, nearly 19 percent of adults between the ages of twenty-five and fifty-four were not participating in the labor force in 2016, again reflecting an incomplete recovery to levels prior to the Great Recession (Bureau of Labor Statistics 2017).

As is the case for poverty more generally, labor market experiences are crucial determinants of falling into or avoiding deep poverty (Fox et al., “Trends in Deep Poverty,” 2015; Shaefer, Edin, and Talbert 2015). Notably, unstable employment, combined with low hours and wages, rather than a total disconnection from employment, appears to be driving deep poverty for many families (Shaefer, Edin, and Talbert 2015). Irregular or unpredictable hours, split shifts, and contingent labor arrangements leave many low-wage workers with variable and inadequate incomes (Lambert, Fugiel, and Henly 2014). The increase in precarious employment is characterized by decreased job tenure and increases in long-term unemployment, nonstandard work hours, and contingent employment in which workers are temporary or work on limited contracts (Lambert 2008; Kalleberg 2009). Low-skill workers have been especially affected by these trends, resulting in high levels of job insecurity (Kalleberg 2009; Lambert, Fugiel, and Henly 2014) and income instability (Morduch and Schneider 2017).

Challenges to sustained employment, including physical and mental illness and disability, addiction, and lack of transportation, are widespread in deep poor populations (Fox et al., “Trends in Deep Poverty,” 2015; Turner, Danziger, and Seefeldt 2006). Whereas the majority of families in deep poverty are headed by a single parent, a substantial proportion of the deep poor (now nearly 40 percent) are unemployed working-age adults without dependent children (Fox et al., “Trends in Deep Poverty,” 2015). Less-educated men, particularly those of color and who have criminal justice histories, are disproportionately likely to experience deep poverty as a result of low levels of labor force participation and high unemployment (Cuddy, Venator, and Reeves 2015; Holzer, Raphael, and Stoll 2006; Jacobs 2015; Council of Economic Advisers 2014) and limited access to income supports.

Whereas the economy is likely to continue to recover from the recession, and unemployment will decrease as a result, the fundamental polarization between high- and low-skill jobs is not expected to end (Autor 2010; Manyika et al. 2011). The wages received by those entering the formal labor market with modest levels of human capital are low. Although men continue to have higher earnings than women, less-educated men have seen much sharper declines in compensation than women (Blank 2009a; Bureau of Labor Statistics 2015). In addition, no evidence indicates that earnings growth alone will be enough to raise incomes above poverty for those with low human capital. Even with the exceptionally strong economy and rapid job creation in the 1990s, real wage growth among families leaving welfare was estimated to range between 2.0 percent and 4.5 percent per year (Card, Michalopoulos, and Robins 2001; Pavetti and Acs 2001). The challenge now is to support sustained labor market participation, increase opportunities for workers to improve skills, encourage earnings growth among all low-skilled workers, and effectively assist low-wage workers who remain poor.

Several determinants of labor market outcomes—related to the structure of employment opportunities as well as to workers’ skills—are key to understanding the labor market. First, whereas higher returns to education are expected to spur less-educated workers to greater human capital investments, growth in educational attainment has been meager (Goldin and Katz 2008). As a result, too few young people, particularly males, are acquiring the degrees and skills required to succeed in the labor market. Demographic shifts in immigration patterns also affect the skill level of the workforce. Over the past forty years, legal immigrants have increasingly arrived from countries with lower levels of human capital and higher rates of poverty, such as Mexico, Central America, and Asia (Raphael and Smolensky 2009). Moreover, the population of undocumented immigrants, who have especially low levels of skills, has grown by more than 300 percent since 1990, though growth declined sharply after 2001 (Warren and Warren 2013).

Finally, incarceration is a key poverty-relevant issue. Nearly 1.6 million individuals were in a prison facility at the end of 2012, the majority of whom were black males younger than forty (Carson and Golinelli 2013). The substantial variation in imprisonment rates by race and gender has been well established. For example, in 2012, incarceration rates for black and white adult males were 2.84 percent and 0.46 percent, respectively, compared to 0.12 percent and 0.05 percent for black and white adult females (Carson and Golinelli 2013). Indeed, incarceration—and criminal justice involvement more generally—is particularly common for black males. Evidence suggests that about half of all black men will be arrested by age twenty-three (Brame et al. 2012; Brame et al. 2014) and that 68 percent of black men without a high school degree will experience incarceration between the ages of twenty and thirty-four, which is true for about 28 percent of white men without a high school degree and 21 percent of black men with a high school degree (Pettit 2012). High rates of incarceration raise important unresolved questions about the implications for labor market opportunities, both for those with incarceration histories and those in affected communities (Holzer, Raphael, and Stoll 2004). Evidence suggests that having a criminal background creates substantial barriers to employment when individuals return to their communities. Criminal history has been identified as the biggest barrier to employment, even more so than failing to complete high school (Peterangelo and Henken 2016). Yet, recidivism is strongly related to whether former inmates get jobs quickly and maintain steady work (Council of Economic Advisers 2016).

Against this backdrop of a difficult labor market for low-skilled adults, major demographic transitions related to family formation have also occurred. First, young adults are now more likely to partner with individuals of comparable education, contributing to greater household income inequality by increasing the pairing of higher (and lower) earners (Schwartz 2013). Second, whereas marriage rates have stabilized for more educated adults, they have declined among the less educated. Of particular consequence, unmarried births among disadvantaged families are now common, representing 57 percent of births to women with less than a high school degree, but only 9 percent of births to women with at least a bachelor’s degree (Shattuck and Kreider 2013). Moreover, nearly three-quarters of unmarried births are unplanned (Sawhill 2014).

The so-called drift into parenthood by low-income young adults (Sawhill 2014) is particularly problematic because their romantic and parental relationships are often short lived. The majority of cohabiting parents break up within a few years of their child’s birth. Many low-income children are then raised with limited involvement with and financial support from their fathers. Whereas child support contributions from noncustodial parents (NCPs) have the potential to reduce poverty, a large proportion of low-income custodial parents receive only partial or no support, often because NCPs have low incomes themselves (Cancian, Meyer, and Han 2011; Smeeding, Garfinkel, and Mincy 2011). Over time, mothers and fathers repartner and have additional children, creating complex families that are likely to remain socially and economically disadvantaged. Children in complex families are then disproportionately likely to experience ongoing family instability, low income, and poverty; moreover, public benefit programs are challenged in designing supports and services that meet the needs of these complex families (Carlson and Meyer 2014).

Addressing the increasing divergence in the fortunes and life trajectories of advantaged and disadvantaged groups (defined by socioeconomic factors such as education, income, race, and wealth) is a fundamental challenge. Sara McLanahan describes the resulting “diverging destinies” as especially consequential for individual well-being and economic mobility (2004; see also McLanahan and Jacobsen 2015). Higher-income individuals have advantages in nearly every relevant institution—the family, neighborhoods, schools, and the labor market (where they encounter primarily other higher-income individuals)—whereas low-income individuals face compounding disadvantages in all of these domains. Economic resources and parental investments are increasing for advantaged children and youths, whereas their disadvantaged counterparts experience comparatively fewer investments.

Parents’ economic disadvantage plays a formative role in shaping children’s opportunities for success and acquisition of skills. The degree of intergenerational transmission of poverty and inequality varies across studies, but the correlation between parent and child income is typically estimated to be about 0.5 (Corak 2006; Jäntti 2009; for recent estimates suggesting higher persistence, see Mitnik et al. 2015). Such persistence in economic positions across generations, coupled with strong theory about why poor children fare worse than their more advantaged peers and accumulating empirical evidence about how poverty affects families and children’s daily experiences, implies that poverty may be determinative in children’s life chances.

Theoretical models of how poverty affects children encompass both what money can buy and how poverty harms relationships. Economic models view families with greater economic resources as being better able to purchase or produce important “inputs” into their children’s development, such as books and educational materials at home, high-quality childcare settings and schools, and safe neighborhoods (Becker 1991). Economically disadvantaged parents may also have less time to invest in children, owing to higher rates of single parenthood, nonstandard work hours, and less flexible work schedules (Smolensky and Gootman 2003). Psychologists and sociologists point to the quality of family relationships and stress to explain poverty’s detrimental effects on children. These theoretical models posit that higher income may improve parents’ psychological well-being and family processes, in particular the quality of parents’ interactions with their children. A long line of research has found that low income is associated with more punitive and less nurturing, stimulating, and responsive parenting. Finally, sources of everyday stress that poor children encounter outside of their family relationships, such as violent or polluted neighborhoods, may also have far-reaching negative consequences in their development (Evans 2001, 2004).

Research on the effects of poverty have focused largely on children’s academic achievement and educational attainment, perhaps because these are strong predictors of subsequent economic well-being. Income gaps and associated socioeconomic status-based gradients in academic skills are present when children enter school and persist through adolescence (Magnuson, Waldfogel, and Washbrook 2012). Poor children complete a year less of schooling than those who have family incomes between one and two times the federal poverty line, and two years less than those who have family incomes more than twice the federal line (Duncan, Ziol-Guest, and Kalil 2010). As described, far too many young adults are entering the labor market without the skills needed to secure stable employment at wages high enough to keep themselves and their family out of poverty.

Despite debate about whether and how much of the estimated associations between poverty and achievement outcomes are causal, several quasi-experimental studies point to substantively meaningful effects (Akee et al. 2010; Dahl and Lochner 2012; Morris, Duncan, and Rodrigues 2011; Milligan and Stabile 2011; see also Duncan, Magnuson, and Votruba-Drzal 2015). Deep and early poverty is particularly strongly associated with lower levels of educational achievement and attainment, holding constant other family advantages (Brooks-Gunn and Duncan 1997). Emerging research in neuroscience and developmental psychology suggests that poverty early in a child’s life may be particularly harmful (Miller and Chen 2013). Not only does the astonishingly rapid development of young children’s brains leave them sensitive (and vulnerable) to environmental conditions, but the family context (as opposed to schools or peers) dominates their everyday lives.

Increasingly, scholars have recognized the importance of appropriate behavior, self-regulation, and mental health in determining labor market and other important adult outcomes, such as criminal activity (Cunha et al. 2006). Many of the same environmental factors and resource constraints that contribute to differential educational attainment may also limit social and emotional development. Low-income children demonstrate less self-regulation, poorer mental health, and more problem behaviors than their higher-income counterparts in childhood and throughout adolescence (Magnuson and Votruba-Drzal 2009). These factors may contribute to criminal activity and incarceration, further compounding lower levels of education and job skills, thus limiting low-income children’s later labor market prospects (Cunha et al. 2006).

In short, a myriad of factors, including changes in labor market opportunities that disadvantage less-skilled workers, demographic trends that increase disparities in the family resources available to children of more- and less-advantaged parents, and changes in public policy, have converged in ways that are creating and exacerbating inequality in many aspects of contemporary American life. These factors have widespread implications with respect to both the current causes and consequences of poverty and for the intergenerational transmission thereof. They suggest the need to review current policies and consider new alternatives that are responsive to the twin challenges of poverty and inequality. In the following section, we review the effectiveness of current policies to provide context for the innovations developed in the articles that follow.

The effectiveness of current anti-poverty policies

Over the past twenty-five years, anti-poverty policies and related social welfare benefits have largely shifted from a system of guaranteed income support to a work-based safety net. These changes were solidified in PRWORA, which reflected a long-standing debate about the adverse effects of income transfers and the effectiveness of job training programs and work supports, as well as a shift toward a cultural norm of parental employment, even for mothers of young children. In addition to emphasizing work, PRWORA also included provisions to encourage marriage and bolster child support enforcement. The shift during the 1990s to work-conditioned benefits reallocated public benefits from nonworking to working households. Those with the lowest market incomes (less than 50 percent of the poverty line) once received substantially more in benefits than those with higher incomes (Moffitt 2015). This, however, is no longer the case: for single-parent families under 50 percent of the poverty line, increases in earnings now result in larger public benefit transfers; moreover, families that are near or just above the poverty line receive substantially larger transfers than in the past (Scholz, Moffitt, and Cowan 2009). Whereas low-income working families have benefited, the shift has left families increasingly vulnerable to periodic unemployment; it also coincides with an increase in the proportion of families that experience very little cash income, deep poverty, or high rates of material hardship—because families without income from formal employment are ineligible for many forms of public assistance, and cash assistance in particular (Sherman and Trisi 2014; Ziliak 2016).

State policies regarding work requirements, lifetime limits on program participation, family caps, and time-limited cash benefits, as well as diversionary tactics for applicants, appear to have affected rates of deep poverty (Hetling, Kwon, and Saunders 2015) or, at the very least, resulted in a considerable segment of the poor population having very little access to cash income (Shaefer and Edin 2013; Shaefer, Edin, and Talbert 2015). At the same time, funding for work supports, such as childcare subsidies, subsidized health insurance, nutrition assistance, and wage supplements (in the form of the EITC) grew extensively. Income support programs thus now function as complements to, rather than substitutes for, formal employment.

The policy changes associated with the 1996 welfare reform have been studied extensively, although much of the evidence was collected during a period of economic expansion. On the whole, welfare “reform generally raised earnings, although not by amounts that are likely to raise many poor families out of poverty” (Grogger and Karoly 2005, 153). Specific aspects of TANF’s work-based safety net have also been evaluated and generally been found to be associated with anticipated labor market effects, though effects on poverty are less evident. For example, mandatory work requirements (or requirements to participate in work-related activities) are associated with reduced welfare use and increased employment (Blank 2002, 2009b; Grogger, Karoly, and Klerman 2002), as are family caps, sanctions, and time limits. More generous childcare subsidies have also been found to promote maternal employment (Dunifon 2010; Grogger and Karoly 2005).

Although, on average, employment increased and there were limited improvements in economic well-being in the wake of welfare reform, as noted above, limited cash support may have increased economic hardship and deepened poverty for those who were not able to find steady work. Moreover, welfare sanctions and reduced access to cash welfare were associated with other negative outcomes, including child welfare involvement (Slack, Lee, and Berger 2007). There are also concerns that the end of the entitlement to cash assistance has contributed to disparities in access to economic support, for example, by race and ethnicity (Fording, Soss, and Schram 2011). Finally, the limited effectiveness of TANF as a safety net program was made clear during the Great Recession, when unemployment rates rose sharply, but TANF participation did not. As a result, trends in poverty and especially deep poverty are now more closely aligned with the business cycle than in the past (Bitler and Hoynes 2016).

Of course, patterns of public program participation look very different today than they did twenty years ago. Specifically, TANF has become much less salient, whereas SNAP, subsidized health insurance, and the EITC have grown dramatically in importance. Indeed, the largest social welfare expenditures today are for means-tested entitlements from Medicaid and SNAP, as well as the EITC. In 2015, the EITC program paid approximately $67 billion (U.S. Department of the Treasury 2017) and SNAP paid over $74 billion (U.S. Department of Agriculture 2017) in benefits to low-income families, versus total spending of $29 billion for TANF and its related childcare components, including state maintenance of effort spending (U.S. Department of Health and Human Services 2017, table A1). Whereas TANF caseloads dropped considerably in the wake of welfare reform and saw limited growth during the Great Recession, SNAP participation expanded significantly, real expenditures increasing over 200 percent between 1980 and 2010. SNAP also assisted the poor much more than TANF during the Great Recession. If counted as an income equivalent, SNAP benefits have reduced the depth and severity of poverty substantially over the last two decades (Shaefer and Edin 2013; Tiehen, Jolliffe, and Smeeding 2016). Likewise, EITC participation and expenditures have grown dramatically over the past several decades and were instrumental for working families during the Great Recession. Liana Fox and her colleagues estimate that the EITC and SNAP reduced child poverty by approximately 8 percentage points, leading them to argue that anti-poverty programs have been more effective in reducing poverty than previously thought (“Waging War,” 2015). Finally, Medicaid and subsidies for health insurance expanded for children as a result of federal funding for State Child Health Insurance Program beginning in 1997, and for adults significantly as a result of state options to expand Medicaid under the Affordable Care Act of 2010.

As first conceived, work-based welfare benefits were created as a way to push (and pull) welfare-dependent single mothers into the labor market. Yet, the increasingly apparent limitations and volatility of the low-wage labor market raise the question of how a work-based safety net can effectively bolster the employment and economic well-being of all workers, both those with and without families to support. Of additional concern, many disadvantaged men, who have in recent decades all but lost their earnings advantage relative to women, often do not have co-resident dependent children and are therefore ineligible for programs limited to resident parents. Disadvantaged nonresident fathers may have few connections to agencies outside of the penal and child support systems, suggesting that reforms to leverage these systems to increase employment and responsible fatherhood may hold promise. Policymakers continue to confront questions about the right mix of policies to increase the availability of family-supporting employment, enable low-wage workers to support their families when working, encourage reemployment when work is scarce, and provide an adequate safety net for those not currently able to work.

Why innovate now?

The challenges are clear. The United States has experienced significant economic growth, yet the fruits of productivity and labor market participation are not being experienced by a large proportion of our population. Inter- and intra-generational inequalities in both opportunities and outcomes by socioeconomic status and race-ethnicity are significant across a wide range of social institutions, spanning neighborhoods, housing, education, the labor market, and the criminal justice system. Less-educated workers face low (and stagnant) wages, instability in employment and hours, minimal employer-provided benefits, and limited opportunities for advancement. The large population of disadvantaged individuals with criminal justice histories—particularly black men—faces substantial labor market barriers and has little access to public benefits. Assortative mating and differences in family formation, fertility, and stability between advantaged and disadvantaged groups have contributed to diverging patterns of family life, with striking implications for the next generation. In particular, children born to disadvantaged parents are highly likely to spend time in a single-mother household and to experience parental multi-partner fertility and associated family instability and fluidity. Low-income children continue to experience lower-quality neighborhoods, housing, and schools. Within this context, anti-poverty policy has increasingly offered work-conditioned benefits and, particularly for nonworking individuals and families, has increasingly offered in-kind rather than cash assistance. Thus, a considerable portion of the low-income population relies on little cash income. Taken together, these factors suggest that anti-poverty innovations are warranted.

The current political divides may undercut the potential for bipartisan initiatives to address poverty. And, the early days of the Trump administration suggest more support for limiting or dismantling programs designed to address poverty, than for expanding their reach. On the other hand, class divides have received new attention, and there remains some bipartisan support for evidence-based policy change. Social science scholars have developed a significant base of research to inform what more could be done and what the effects of new innovations might be. Social science theory and empirical evidence have continued to accumulate and point to key ways in which policy innovations could better support the current generations of workers, both those who struggle to find steady employment and earn a family-sustaining wage, as well as those completing their education, starting to work and, often, also starting their families. Such evidence further points to how to help the younger generation—the children of low-income adults—who may be harmed by the experience of deep and persistent economic hardship, and who often miss out on experiences they will need to thrive later in their lives. Finally, it points to innovative ways to build on the current safety net to better assist low-income individuals and families, both those who are strongly and those who are weakly attached to the labor market.

Innovative anti-poverty approaches

With policymakers’ need for evidence-based solutions in mind, each article in this volume focuses on a specific social problem or population and presents a detailed, actionable response. The articles leverage the best available theoretical and empirical social science research to present evidence-based arguments for implementing a set of novel and potentially transformational policy innovations. The proposals span a wide range of policy domains, including cash transfers, employment-related policies, postsecondary education, housing support, food security, family planning, and two-generation human capital development. Despite this range, the proposals do not explicitly address several important policy topics: for example, immigration, incarceration, child-care, child welfare, transportation, subsidized health care, and Medicaid.

As Wimer, Collyer, and Kimberlin discuss in detail in the last article in this double issue (2018), the costs and impact of these proposals vary widely. So do the scale and ambition of the innovations, and the authors’ attention to details of policy implementation and the organizational contexts of programmatic implementation. Some call for wholesale transformation of programs or institutions, such as establishing a universal child allowance (Bitler, Hines, and Page 2018; Shaefer et al. 2018), or guaranteeing universal access to an above-poverty wage job (Paul et al. 2018). Others leverage existing programs to address key challenges facing the poor, for example, a minimum benefit to reduce elderly poverty (Herd et al. 2018), and a renter’s tax credit (Kimberlin, Tach, and Wimer 2018) to address housing costs. Others advocate for expanding investments to improve education and training—whether for children and their parents (Sommer et al. 2018), or adult workers (Holzer 2018; Strumbos, Linderman, and Hicks 2018). Many of the proposals respond to the needs of families with children, or to the particular vulnerability of children in single-parent families. However, only Lawrence Wu and Nicholas D. E. Mark (2018) consider an effort to directly alter family structure. They review evidence on the potential for improved access to contraception—specifically long-acting reversible contraceptives (LARCs)—to reduce unintended and nonmarital pregnancy and improve economic and social well-being. In contrast to most of the other articles in this double issue, Wu and Mark call for a pilot to test the impact of their proposal.

Whereas all the proposals address concerns with poverty, they vary substantially in their proximate goals, and at least implicitly, in their theory of poverty or anti-poverty policy. Mark Paul and his colleagues (2018), who propose a federal job guarantee whereby all American adults would be assured of full-time employment in a “public works” job with above-poverty wages and benefits, is the most costly, and also the most transformative. By assuring employment—with adequate wages and benefits—the authors argue that the need for many other programs would be eliminated. The proposal may have limited policy relevance, but it provides an important counterpoint to other articles in this double issue, which implicitly assume the current structure of the labor market, with a low minimum wage and limited required benefits.

A number of the proposals aim to reduce the proportion of workers relying on low-wage jobs by improving education and training in an effort to provide access to better jobs. Teresa Eckrich Sommer and her colleagues (2018) propose redesigning the Head Start program to more fully serve the needs of both parents and children by combining parental education, training, and employment opportunities with the existing early childhood education components of the program. Diana Strumbos, Donna Linderman, and Carson C. Hicks (2018) argue for a national community college model based on the CUNY ASAP model, which provides students with extensive advising, academic, career, and financial supports while requiring full-time enrollment in a highly structured degree program. Harry Holzer (2018) lays out a competitive grant program to states to implement performance-based community college programs to improve academic and employment outcomes for disadvantaged populations. Indivar Dutta-Gupta and his colleagues (2018) argue for a national on-the-job training and subsidized employment program for low-skilled workers.

Other proposals aim to supplement the earnings of low-wage workers in ways that generally improve the returns to work, and target particular challenges. For example, Jennifer Romich and Heather Hill (2018) describe a plan to “couple” minimum wage hikes with changes in existing benefit programs to avoid high marginal tax rates and benefit cliffs and, thereby, to offer a package of increased wages and sustained benefit (particularly SNAP and EITC) receipt. Sarah Halpern-Meekin and colleagues (2018) propose enabling workers to defer 20 percent of their EITC refund—with a 50 percent savings match—for a six-month period to encourage emergency saving among low-wage workers.

In contrast, this issue also includes a range of proposals to directly support individuals and families with insufficient resources, regardless of work status. Three proposals target families with children through new cash transfer mechanisms: both Marianne Bitler, Annie Laurie Hines, and Marianne Page (2018) and H. Luke Shaefer and his colleagues (2018) propose an unconditional child allowance, albeit in very different forms; Maria Cancian and Daniel Meyer (2018) propose a public guarantee of private child support payments available to all children not living with both parents. Sara Kimberlin, Laura Tach, and Christopher Wimer (2018) also propose a new transfer program that would provide a refundable renter’s tax credit for families facing high rental costs relative to their income; Pamela Herd and her colleagues (2018) argue for a targeted minimum benefit plan that would provide a guaranteed benefit through the Social Security system to bring the income of poor elderly individuals to the poverty threshold, regardless of their work history. Other proposals build off and expand existing programs. For example, Craig Gundersen, Brent Kreider, and John V. Pepper (2018) present a plan to change the SNAP benefit formula to increase benefits and substantially reduce food insecurity among SNAP recipients.

In short, the proposals highlighted in this double issue all aim to reduce poverty, but they offer markedly different solutions, in many ways solving different problems. If the problem is insufficient resources, the solution may be to make more money available to families with limited earnings, though the authors here have very different ideas about who should get what, and under what conditions. A few would not require work, but would provide support only to resident parents with children (child allowance proposals of Bitler, Hines, and Page 2018 and of Shaefer et al. 2018), or with children living apart from a parent (child support assurance proposal of Cancian and Meyer 2018). These authors highlight the needs of children and the returns to public investments to reduce their vulnerability to poverty. The child allowance proposals offer a universal benefit. The child support assurance proposal emphasizes the need to support nonresident parents’ contributions—children receive government support not when their custodial parent (usually their mother) is poor, but when their nonresident parent (usually their father) has inadequate earnings or otherwise is unable or unwilling to meet parental obligations.

Other proposals target the inadequacy of work and earnings—but the various proposals are based on very different assumptions about the nature of the problem. If work is not available, and earnings are the preferred resource, then providing jobs is a logical solution. But, should there be a universal and unlimited guarantee of government work at above-poverty wages, as Paul and colleagues (2018) propose, or time-limited, targeted subsidized jobs, paying minimum wages and often provided by private-sector employers, as Dutta-Gupta and colleagues (2018) propose? The answer depends in part on whether low wages reflect the structure of the labor market and of workers in today’s economy—which argues for a job guarantee to change workers’ options—or a skills mismatch that can be addressed with investment in human capital, at least in the medium to long run.

The articles in this double issue respond to a call for innovative policy proposals intended to reduce poverty and improve economic well-being. The individual responses differ across many dimensions and, as Wimer, Collyer, and Kimberlin (2018) stress in the final article, these differences can make comparisons challenging. At the same time, the range of approaches is instructive in highlighting the scope and diversity of potential innovations. In the face of continued high rates of poverty, growing inequality, and significant dissatisfaction with current efforts, there are reasons to substantially broaden the range of policies under discussion. The proposals that follow offer an important set of options to seed the debate.