THE EARNED INCOME TAX CREDIT AS AN ANTI-POVERTY PROGRAM: PALLIATIVE OR CURE? (draft)

Transcription

1 THE EARNED INCOME TAX CREDIT AS AN ANTI-POVERTY PROGRAM: PALLIATIVE OR CURE? (draft) Phyllis Jeroslow Doctoral Candidate School of Social Welfare University of California at Berkeley East Asian Social Policy Association Research Network/ United Kingdom Social Policy Association Conference July 18, 2012 York, England Copyright 2012 by Phyllis Jeroslow. All rights reserved.

2 Table of Contents Abstract 4 I. Introduction 6 II. Scope of the EITC 9 Purpose of the EITC 9 EITC Recipients 9 EITC Benefit Structure 10 Participation Rates 12 III. EITC Successes 14 IV. EITC Shortcomings 18 The Federal Poverty Measure 18 Poverty Rates 21 Structural and Functional Ramifications 26 Spending Patterns 28 Economic Mobility 30 Summary of EITC Shortcomings 32 V. Outcomes for Low-Income Children 34 The Effects of Poverty 34 Effects of Welfare to Work 36 Implications for the EITC in Relation to Child Development Outcomes 37 DRAFT 2

3 VI. From Palliative to Cure 39 EITC-based Solutions 39 Redirecting Anti-Poverty Policy 41 Bigger Picture Solutions 44 References 48 List of Tables and Figures Figure 1: Earned Income Tax Credit by Filing Status and Number of Children 12 Table 1: Poverty Status of Families by Type of Family, For All Races With Children Under 18 Years, for Selected Years 24 DRAFT 3

4 ABSTRACT The Earned Income Tax Credit (EITC), a federal tax rebate program of the United States, is heralded as that nation s largest antipoverty program for low-income, working families. After more than 35 years of implementation, poverty still plagues families and children, and entrenched pockets of concentrated disadvantage persist in the wake of lost manufacturing jobs and the internationalization of markets. Although the EITC has been lauded for increasing labor force participation, particularly for single mothers, and for lifting working families up to the federal poverty level, the short-term benefits of an EITC tax reduction do not appear to support long-term gains for upward mobility. The EITC is primarily aimed at families with children; two-child families are the most common type of recipients. When the EITC was first enacted in 1975, it was intended to lessen a tax burden that fell disproportionately on the poor. The EITC functioned to decrease welfare dependency by providing an incentive for low-income parents to work. The anti-welfare effect of the EITC has long served as a source for bipartisan support, and the EITC is often considered an integral component of the new workfare state. At several points in the long history of the EITC, coverage was expanded and the amounts of tax rebates were increased. In 2010, $59.5 billion was paid to 26.8 million families in EITC supplements. Contrary to what is conveyed by the successes of the EITC as measured by the official poverty level, there is growing consensus that poverty is not defined by income alone; it is more aptly described by a constellation of deprivation. Most EITC funds are spent for consumption needs, and few recipients are able to develop assets or make human capital investments in training and education to benefit themselves or their children. In the absence of a coordinated, national social policy, particularly regarding investments in early childhood that produce high DRAFT 4

5 economic returns, many Americans may continue to lead lives marked by deprivation and diminished opportunities. The way forward will require a more comprehensive solution to impoverishment than is afforded by the EITC. DRAFT 5

6 I. INTRODUCTION The Earned Income Tax Credit (EITC) is heralded as the largest antipoverty program for low-income, working families in the United States (Berube, 2006; Holt, 2006). As a provision of the federal tax code, the EITC operates primarily as a tax rebate. The EITC is lauded for lifting working families up to the federal poverty level and for increasing labor force participation, particularly for single mothers (Acs & Toder, 2006; Eissa & Hoynes, 2011; Greenstein, 2005; Hotz & Scholz, 2001; Liebman, 1997; Meyer, 2010). After welfare reform in 1996, the EITC played a major role as a support and incentive for those transitioning from welfare to work and for reducing welfare entries (Acs & Toder, 2006; Grogger, 2003; Holt, 2006; Meyer, 2010). The anti-welfare effect of the EITC has long served as a source for bipartisan support, and the EITC is often viewed as an integral component of the new workfare state (Ventry, 2000). The EITC is the largest federal cash transfer program for low-income households (Acs & Toder, 2006). The program is principally aimed at families with children, with only a modest credit, implemented in 1993, for childless workers (Holt, 2006). Benefits increase as the number of children rises from one to three or more (IRS, 2012). Two-child families are the most common type of recipients (Holt, 2006). When the EITC was first enacted in 1975, it was intended to lessen a tax burden that fell disproportionately on the poor (Ventry, 2000). Upon initial implementation, 6.2 million families claimed the credit for a total government expenditure of $1.25 billion and an average tax credit per household of $201 (in nominal dollars; Beamer, 2005). At several points in the long history of the EITC, coverage was expanded and the amounts of tax rebates were increased (Ventry, 2000). By 2003, EITC relief totaled $34.4 billion and was distributed among 19.3 million DRAFT 6

7 families; the average credit to households with children in 2003 was $1,784 (Beamer, 2005; Holt, 2006). In 2010, $59.5 billion was paid to 26.8 million families (IRS, 2012). The EITC demonstrates success when measured against program goals, namely, to lift families up to or over the federal poverty level, increase workforce participation by single mothers, and decrease welfare caseloads. In these respects, the EITC has functioned to provide a subsistence income for many American families, year by year, and simultaneously fill the economic need for low-wage workers. Upon deeper probing, however, the EITC is less successful when examined against the inadequacies of the federal definition of poverty, the relatively consistent rate of poverty in the U.S. since the mid-1970s, and the resolution of poverty as a transgenerational problem. After more than 35 years of EITC implementation and substantial cash transfers, poverty still plagues American families and children. Entrenched pockets of concentrated disadvantage persist in the wake of lost manufacturing jobs and the internationalization of markets (Lindsey, 2009; Wilson, 1987). The short-term benefits of an EITC tax rebate may not be sufficient to support longer-term gains needed for upward mobility (Berger, Paxson, & Waldfogel, 2009; Dowd & Horowitz, 2011; Rainwater & Smeeding, 2003; Simpson, Tiefenthaler, & Hyde, 2010). Additionally, there is growing consensus that poverty is not defined by income alone; it is more aptly described by a constellation of deprivation (Wilson, 1987). Most EITC funds are spent for consumption needs, and few recipients are able to develop assets or make human capital investments in training and education to benefit themselves or their children (Simpson et al., 2010). In the absence of an anti-poverty agenda that includes human capital-building, many Americans may continue to lead lives marked by deprivation and diminished opportunities. DRAFT 7

8 This paper begins its exploration of the EITC by explaining the program s scope and basic benefit structure. Successes of the EITC program with respect to cash transfers and labor force participation will be followed by a discussion of the policy s shortcomings against broader anti-poverty considerations. A deeper examination of EITC deficiencies will include definitions of poverty, poverty rates, EITC spending patterns, and prospects for economic mobility based on EITC refunds. The EITC primarily targets parents and calibrates benefits to the number of children in a household. Based on the eligibility and benefit structure of the EITC, it is fair to presume that the policy aims to assist children by increasing family income. The effects of deprivation on children in low-income and poor families will be considered in relation to finding policy solutions that emphasize building children s human capital and alleviating transgenerational poverty. In the 1960s, President Johnson s vision of the Great Society included universal opportunities for training and education and pro-work policies as integrated solutions to poverty (Economic Opportunity Act of 1964). This paper takes the position that the EITC provides much needed income support to working families, but that income alone, as a perennial supplement to low wages, is not enough to prevent many children of EITC recipients from experiencing deprivation that is likely to limit their life chances and perpetuate the cycle of poverty. As President Johnson envisioned, if poverty is to be surmounted in future generations, the federal anti-poverty agenda will need to combine income supports to parents with other forms of social assistance, and include policies that build the human capital of their children. DRAFT 8

9 II. SCOPE OF THE EITC Purpose of the EITC The EITC is a tax credit that functions as a wage supplement for low-income workers (Berube, 2006). At its most basic level, the EITC is intended to lift families up to or above the federal poverty line. The EITC also joins with other government programs as an integral part of an anti-poverty agenda to move people from welfare to work (Ventry, 2000). Since the EITC does not provide benefits for the non-working poor, it is not considered as a substitute for the safety net (Hotz & Scholz, 2001). EITC Recipients The EITC implicitly targets single mothers through its benefit structure (Meyer, 2010). Using data from the Current Population Survey, Meyer (2010) estimates that nearly 50 percent of EITC dollars go to single mothers and seven percent to single fathers, even though they together represent slightly over ten percent of U.S. households. Single parents claim a majority of EITC refunds because they are disproportionately represented among poor families with children. Thirty-eight percent of EITC dollars go to married couples with children, and only five percent go to childfree individuals, since payments to workers without children are comparatively minimal (Meyer, 2010). The IRS estimates that seventy-three percent of recipients are single parents, a considerably higher proportion than that indicated by the Current Population Survey (Meyer, 2010). Viewed from another perspective, sixty percent of single mothers and fifty percent of single fathers receive the credit, compared to twenty percent of married couples (Hoffman & Seidman, 2003). According to Meyer (2010), the EITC is targeted toward larger families, with at least sixty-one percent of all payments provided to families with more than one child. In comparison, DRAFT 9

10 a 1996 profile of EITC recipients by number of children indicates that 18.8 percent had no children, 30 percent had one child, 28.9 percent had two children and 22.3 percent had three or more children (Holt, 2006). EITC Benefit Structure The EITC is a federal, means-tested, cash assistance program for wage-earners administered through the Internal Revenue Service as a redistributive transfer. There are three main criteria for eligibility. Families must (1) have a wage earner; (2) be low income; and, in order to receive a significant credit, (3) have a resident child (Meyer, 2010). For those who qualify, the EITC reduces tax liability based on annual family income (Holt, 2006; Hotz & Scholz, 2001). Claimants must file a tax return in order to be eligible. Most of the IRS expenditure for the EITC is paid to families with at least two qualifying children and, for any given family, the amount of the cash transfer rises as the number of qualifying children increases up to the category designated three or more (Holt, 2006; U.S. Department of the Treasury, 2012). In 1993, EITC eligibility was extended to childfree workers between the ages of twentyfive and sixty-four, but the benefit amount is minimal in comparison to families with children (Holt, 2006). The EITC is refundable. If the amount of the credit exceeds the tax liability, the excess is refunded as a direct transfer payment (U.S. House of Representatives Committee on Ways and Means, 2004). If the tax liability is zero, then the entire amount is paid as a refund (Holt, 2006). In most cases, the credit is taken as a lump sum based on annual earnings, and received as part of a tax refund usually issued in the first few months following the filing year (Meyer, 2010). EITC benefit levels are structured by three ranges of annual income: (1) phase-in; (2) plateau; and (3) phase-out. In the phase-in range, the credit increases for each additional dollar DRAFT 10

11 earned up to a certain maximum level. At the maximum level, the amount of the credit reaches a plateau; more income does not result in an increase in the tax credit. During the phase-out range, the tax credit gradually decreases to zero as income rises (Holt, 2006). The amount of the credit also varies according to the number of qualified children in the household, with the category of three or more children providing the maximum credit levels (U.S. Department of the Treasury, 2012). The dollar amounts of the EITC benefits, and the specification of phase-in, plateau, and phase-out points, have changed over time. In 1975, the first year of EITC implementation, the phase-in range for annual income was $0-$4,000; the phase-out range was between $4,000 - $8,000; and the maximum credit was $400 (in nominal dollars; Hotz & Scholz, 2001). For tax year 2011, earned income could not exceed $36,052 ($41,132 married filing jointly) for one qualifying child, and $40,964 ($46,044 married filing jointly) for two qualifying children, and the maximum credits were $3,094 for one child and $5,112 for two children (U.S. Department of the Treasury, 2012). Limits were higher for families with three or more children. The proportion of EITC participants differs by phase. Roughly fifteen percent of recipients are in the phase-in range, twenty-five percent in the plateau range, and sixty percent in the phase out range (Ventry, 2000; Hoffman & Seidman, 1990; Scholz, 1994). In a 30-year retrospective of EITC policy, Holt (2006) explained that the EITC benefits moderate-income families as well as low-income families, and that claimants cycle in and out of eligibility through the years. More moderate-income families benefit from the EITC than from other government assistance programs, such as cash welfare or food stamps (Liebman, 2001). DRAFT 11

12 Figure 1. Participation Rates The participation rate (also known as the take-up rate) refers to the proportion of eligible workers and families that actually file a claim for the credit (Berube, 2006). The participation rate for the EITC is difficult to determine accurately (Holt, 2011). The U.S. General Accounting Office (2001) reported a take-up rate of seventy-five percent. Additionally, DRAFT 12

13 Holt (2011) found evidence of an increasing take-up rate, with seventy-six percent participation in tax year 2005, seventy-seven percent in 2006, and seventy-nine percent in These estimates indicate a higher participation rate for the EITC in comparison to other anti-poverty programs (Holt, 2011). The high take-up rate may be a consequence of recipients being able to avoid the stigma usually associated with a cash handout by instead receiving benefits through the tax system. The higher rate has also been attributed to the higher educational level of EITC claimants who may have greater awareness about government assistance programs, compared to the population of welfare recipients (Eissa & Hoynes, 2011; Holt, 2011; Liebman, 1997). Additionally, participation rates for the phase-in range may be lower than the total participation rate (Liebman, 1997). Approximately one in seven households (15%) nationwide qualify for the EITC (Hoffman & Seidman, 2003). Based on EITC requirements, ninety-five percent of single mothers with two or more children and no high school graduation would be eligible. Additionally, cohabiting couples with children are more likely than married couples to qualify for the credit because their incomes are comparatively lower on average (Holt, 2006). Depending upon income, EITC claimants may be eligible for other in-kind government assistance. For example, in a study conducted in a rural New York county, 40 percent of EITC recipients also received Food Stamps (Simpson et al., 2010). Relative to economic downturns, a person who lost his/her job could not benefit from the EITC if there were no earnings during a single tax year, but a cut back in work hours could render someone eligible for the EITC who previously was earning too much to qualify (Holt, 2011). DRAFT 13

14 III. EITC SUCCESSES The EITC has received ongoing acclaim for its success on several accounts. Most accolades concern the ability of the EITC to lift people out of poverty (Holt, 2006; Meyer, 2010; Simpson et al., 2010). The EITC increases after-tax wages for low-income earners by up to fortyfive percent (Meyer, 2010), delivering its greatest poverty-reducing impact to single parent families (Simpson et al., 2010). Each year, the EITC program is estimated to lift 2.5 million children in working families out of poverty (U.S. Department of the Treasury, 2010). Meyer (2010) notes that for tax year 2007, the EITC lifted over 1.1 million families above the poverty line, including 2.1 million children, reducing the poverty rate by ten percent and the child poverty rate by sixteen percent, the largest reductions of any antipoverty program. Without the EITC, the child poverty rate would be twenty-five percent higher (Greenstein, 2005). Estimates of EITC success with poverty reduction by Hoynes, Page, and Stevens (2006) are more conservative. For 2003, when EITC payments are added to post-tax incomes, Hoynes et al. (2006) estimate that the EITC lowers the poverty rate by 1.7 percentage points, and decreases the child poverty rate by 3.1 percentage points. In comparison, means-tested cash assistance for all non-elderly recipients lowers the overall poverty rate by 0.8 percent (Hoynes et al., 2006). The poverty-reduction effects of the EITC operate during times of economic contraction as well as expansion (Berube, 2006; Caputo, 2010; Dowd & Horowitz, 2011). In economic downturns, the EITC serves as a type of wage insurance for temporary lack of work during part of a tax year or reduction in work hours for those near the bottom of the wage distribution (Berube, 2006; Bluestone & Ghilarducci, 1996). When unemployment increased during the years between 2000 and 2003, EITC receipt far exceeded the rise in unemployment benefits DRAFT 14

15 (Berube, 2006). Particularly with federal assistance programs devolving into block grants administered at the state and local levels, the EITC has become a key safety net for lower-wage workers during economic recessions (Berube, 2006). In addition to serving as a buffer for economy-wide shocks, the EITC has also served as a safety net for individually-specific employment cutbacks (Dowd & Horowitz, 2011). In addition to poverty alleviation, the EITC has also been credited with reducing the burden on public welfare programs, especially for Aid to Families with Dependent Children (AFDC) and its 1996 successor, Temporary Assistance to Needy Families (TANF). Following welfare reform in the late 1990s, the EITC was found to exert particularly strong effects in reducing TANF caseloads by playing an important role in reducing welfare entries (Grogger, 2003, p. 25). The EITC functions as the main redistribution program of the government by extending benefits well into the middle of the income distribution (Eissa & Hoynes, 2011). Liebman (1997) calculates that the EITC offset the decline in income between 1976 and 1996 by twentythree percent for households with children in the lowest quintile of the income distribution, and by nine percent for households in the second lowest quintile. Additionally, the EITC counteracts the tax burden that disproportionately affects the working poor, and somewhat diminishes the gap between rich and poor (Ventry, 2000). Many studies attest that the EITC has achieved unequivocal success in its other major thrust to increase participation in the labor force, particularly among low-skilled, less-educated single mothers (Acs & Toder, 2006; Caputo, 2010; Dahl, DeLeire, & Schwabish, 2009; Eissa & Hoynes, 2011; Hotz & Scholz, 2001; Liebman, 1997; Meyer, 2001, 2010; Ventry, 2000). Estimates of the EITC s impact on increasing the employment of single mothers vary widely and DRAFT 15

16 cover overlapping time periods. Corresponding with substantial expansions in the EITC benefit structure, the labor force participation of single mothers between 1989 and 2002 increased remarkably by 14 percentage points (Eissa & Hoynes, 2011). The Tax Reform Act of 1986, which corrected the inflationary erosion of the EITC and extended its phase-out point, was found responsible for increasing labor-force participation among single mothers by 2.8 percentage points, and by 6.1 percentage points for those with less than a high school education (Liebman, 1997). For the years between 1984 and 1996, the EITC and other tax changes are considered responsible for more than 60 percent of the increase in labor participation among single mothers (Meyer, 2001). Grogger (2003) asserts that expansion of the EITC appeared to exert a substantial effect on reducing welfare entries. In the 1990s, the EITC was found responsible for fifteen percent of the prodigious decline in welfare receipt (Grogger, 2003, 2004). Decreases in welfare participation were also related to welfare reform, a decline in the real value of welfare benefits, and economic conditions favoring work (Grogger, 2003). The success of the EITC in promoting work is attributed to its incentive to make work pay by providing more combined income through work and tax benefits than could be acquired through welfare payments and other public assistance (Meyer, 2010; Ventry, 2000, p. 10). Large increases in the employment of single mothers relative to single women without children correspond to significant expansions of the EITC benefits in the early and mid-1990s. The gain of a few thousand dollars proves sufficient to influence the extensive margin in the single mother s decision to work or not to work (Acs & Toder, 2006, p. 335; Meyer, 2010). The decision to work would typically launch welfare-to-work mothers into the phase-in or plateau phases of EITC payments (Meyer, 2010). Employment rates also increased with the number of DRAFT 16

17 children, corresponding to legislative changes that provided more generous EITC benefits for families with more than one child (Meyer, 2010). The increases in employment rates for single mothers are also related to their education level, with the largest increases demonstrated by single mothers without a high school diploma. When single mothers with children are compared to single women without children, the differences in employment rates fall sharply between the two groups as education levels rise (Meyer, 2010). The EITC is also associated with increases in the labor participation of married men (Ventry, 2000). Although the EITC has been found to diminish work incentives for secondary earners in married couples and to offset the increase in work hours of married men, on balance the vast increase in the labor participation of single mothers outweighs the decreased employment of married mothers (Eissa & Hoynes, 2004; Ellwood, 2001; Hotz & Scholz, 2003). Contrary to what is expected by economic theories of motivation, the lesser returns of the phaseout period of EITC receipt do not appear to diminish hours worked on the intensive margin (i.e., the decision to work another hour or not) (Acs & Toder, 2006; Hotz & Scholz, 2001). DRAFT 17

18 IV. EITC SHORTCOMINGS This section will examine the EITC against standards that may veer from the program s goals, yet are no less relevant to a federal anti-poverty agenda. The discussion will include alternative definitions of poverty and deprivation, poverty rates, EITC structure and function, refund spending patterns, and prospects for economic mobility for EITC recipients. The Federal Poverty Measure A basic understanding of the poverty measure is essential to the analysis of the EITC as an anti-poverty program. Pressman (2011) notes, many people are not counted as poor but live in difficult circumstances that most of us would consider poor. One reason for this is that the official U.S. definition of poverty has not been updated in the past half-century (p. 118). The poverty thresholds were developed by economist Mollie Orshansky in based on the cost of the Department of Agriculture s economy food plan for families of different sizes. The 1955 Household Food Consumption Survey indicated that one-third of family after-tax income was spent on food. Consequently, poverty thresholds for families of different sizes were established by multiplying the cost of the economy food plan by three. Other than cost of living adjustments specified by the Consumer Price Index (DPI), the official poverty measure has changed little since it was established in 1969 (Hoynes et al., 2006, p. 52). For a family of two adults and two children the poverty threshold for 2011 is $22,811 and the poverty threshold for a single parent and two children is $18,123 (U.S. Bureau of the Census, 2012a). A family is designated as poor when its gross income, according to family size, falls below the poverty line. Pressman (2011) maintains that the poverty lines have never been updated to reflect realworld circumstances such as consumer debt, interest payments, payday lenders, and child care DRAFT 18

19 expenses. Housing, medical, and transportation expenses have also increased substantially since the 1960s (Citro & Michael, 1995). Further, Pressman (2011) indicates that Orshansky and the Social Security Administration where she worked advocated using the low-cost food budget that applied to annual costs rather than the emergency food budget that was not nutritionally adequate for an entire year. However, President Johnson and his Council of Economic Advisers chose to use the emergency food budget to establish poverty levels because they did not want to report that such a large proportion of the United States was poor. Pressman s (2011) point is that the United States poverty lines for families of different sizes were never adequate from the start. Statistics on the incidence of material hardship indicate that the low dollar amount of the federal poverty line masks the reality that families with incomes of up to twice that associated with poverty status continue to face difficulties in making ends meet (Gershoff et al., 2007, p. 71). Thus, when the EITC is touted as lifting families out of poverty, the fiscal realities underlying this rhetoric must be more carefully examined. The federal poverty line is not adjusted for regional variations in the cost of living, unemployment rates, and economic downturns. Further, the guidelines stymie efforts to determine the efficacy of anti-poverty policies because the poverty thresholds are based on taxable income and do not include in-kind government assistance such as food stamps, housing subsidies, or tax credits (Aber et al., 1997; Citro & Michael, 1995; Gershoff, Aber, & Raver, 2003). In contrast to the official poverty threshold of the United States, European and other nations use a relative measure to designate poverty thresholds that is usually based on fifty DRAFT 19

20 percent of a country s median income. With this metric, it is estimated that an additional eight percent of Americans would be considered poor (UNICEF Innocenti Research Centre, 2000). Absolute measures of poverty indicate deprivation, while relative measures emphasize inequality and social exclusion. The demarcation of fifty percent of median income indicates whether or not a family has an acceptable standard of living in reference to the entire society at a given point in time (Gershoff et al., 2003). Incomes far below standard levels signify that a family is unable to participate fully in their society (Townsend, 1992; UNICEF Innocenti Research Centre, 2000). Rainwater and Smeeding (2003) assert that the debate over the merits of absolute vs. relative measures obscures the distinction between economic and social definitions of poverty. Economic definitions concern consumption, while social definitions focus on the ability to participate in social activities and social roles as family members, workers, and citizens. Social standards of poverty change and societal poverty lines must be redrawn to keep pace (Rainwater & Smeeding, 2003, p. 11). The self-sufficiency standard provides another alternative measure of poverty, founded on the real costs of living for a family s basic needs, independent of subsidies (Gershoff et al., 2003, p. 83). Basic needs are conceived as housing, food, transportation, childcare, health care, and miscellaneous items such as clothing, medicine, and household items. According to the selfsufficiency standard, the federal poverty level is gravely inadequate and needs to be doubled (Gershoff et al., 2003). Boushey et al. (2001) estimate that United States poverty rates would increase two-and-a-half times according to the self-sufficiency standard. Boushey et al. (2001) find 200 percent of the poverty level as a useful dividing line to demarcate frequencies of hardships among families. Of families who lived below 200 percent of DRAFT 20

21 the federal poverty line in 1996, nearly thirty percent experienced one critical hardship (e.g., missing meals, being evicted, inability to access medical care) and over seventy-two percent suffered at least one serious hardship (e.g., food insecurity, missed mortgage or rent payment, reliance on emergency room medical care, or inadequate child care) (Boushey et al., 2001, p. 2). The self-sufficiency standard also has implications for social exclusion based on those who are able to partake of a safe and decent (Boushey et al., 2001, p. 2) living standard and those who cannot. Gershoff et al. (2003) find that exclusion from safe neighborhoods, adequate housing, and quality education may compromise children s healthy development and limit their future roles as learners, workers, and citizens (p. 85). The widely-criticized official poverty threshold (Gerhsoff et al., 2003) is the standard by which the EITC is assessed. If the EITC s success is predicated on lifting people out of poverty and the definition of poverty is in question, then the success of the EITC relative to overcoming poverty is also in question. Alternative standards provide other conceptions of poverty and deprivation that claim a firmer basis in human experience. Poverty Rates The EITC was first implemented in 1975 in response to a recession, an accompanying rise in unemployment rates, and increasing social security payroll tax burdens on the poor (Hotz & Scholz, 2001; Liebman, 1997; Ventry, 2000). The U.S. poverty rate for families with children for that year was relatively low compared to the preceding and following years. Since 1975, the poverty rate for families with children has fluctuated from a low of 12.7 percent in 2000 to a high of 18.5 percent in 1993 (See Table 1; U.S. Bureau of the Census, 2012). During the same period, the poverty rate for single women with children ranged from a low of 33.0 percent (2000) to a high of 46.1 percent in Official United States poverty rates do not take into account DRAFT 21

22 public assistance or tax credits received. Nevertheless, the overall trend reveals that poverty has persisted, and that interventions to alleviate poverty have not succeeded in significantly lowering poverty rates for families with children. Data from the OECD Income Distribution questionnaire covering the mid-to-late 2000s indicate that the United States has a poverty rate higher than what prevails in virtually all other rich nations (Stanford University, 2012). Approximately thirty-five percent of individuals who enter a period of poverty will remain in poverty for at least five of the next ten years (Stevens, 1999). For families who are able to leave poverty, there is a substantial rate of re-entry (Hoynes et al., 2006). Children who are born into poverty are more likely to remain poor than older children first entering poverty. Analysis by race indicates that one-year-old African American children born to single mothers with less than a high school education have an 89.5 percent chance of being poor in the following five or ten years, compared to a White child under similar family circumstances whose likelihood for being poor is 63 percent during the same time frame (Hoynes et al., 2006). Children compose approximately 40 percent of the poor compared with 29 percent of their presence in the overall population (Hoynes et al., 2006). Single parent families comprise 39.1 percent of families with children who are poor (Hoynes et al., 2006). Half of the single mothers in the United States live below the poverty threshold, irrespective of race (Lindsey, 2009). When race is considered, White and Asian children raised in two-parent families experience a poverty rate of five percent, while African American and Latino children in oneand two-parent families exhibit a poverty rate that ranges from 35 to 40 percent. However, more than two-thirds of African American children live in single-parent households, and of this group, the child poverty rate is nearly 60 percent (Lindsey, 2009). DRAFT 22

23 Using the self-sufficiency standard of twice the poverty threshold, the National Center for Children in Poverty estimates that in 2009, 42 percent of American children resided in lowincome families, split between 21 percent designated as near-poor (between one and two times the poverty line), and 21 percent falling below the poverty line. In 2009, the number of children below the poverty line reached 15.3 million, and the number of children below twice the poverty line reached 31.3 million (National Center for Children in Poverty, 2012). Hoynes, Page, and Stevens (2006) examine why the poverty rate has largely been intractable, turning primarily to labor market conditions and government antipoverty programs for explanations. Government programs have a greater effect on closing the poverty gap between income and the poverty threshold than on reducing the proportion below the poverty line, but Hoynes et al. (2006) locate the main explanation for ongoing high poverty rates in labor market conditions as defined by median wages, unemployment rates, and inequality. Despite dramatic growth in the GDP per capita since 1967, the poverty rate has remained largely unchanged. Hoynes et al. (2006) observe that poverty rates fall in association with decreases in inequality. Conversely, they find that a virtually continuous increase in wage inequality below the median driven by declines in wages for less-skilled workers provides the most cogent explanation for the upward trend in poverty rates (Hoynes et al., 2006, p. 53). While the poverty rate among single-mother households is generally three to four times that of the general population, the doubling of female-headed households between 1967 and 2003 only partially accounts for the continuation of poverty rates during that period (Hoynes et al., 2006). Poverty rates during this time frame were lower than expected due to the greater participation of single women in the labor force and an increase in their earnings (Hoynes et al., 2006). DRAFT 23

24 Acs and Loprest (2004) studied the effects of the EITC on poverty reduction in relation to the program s goal of incentivizing parents to move from welfare to work. They found that sixty percent of welfare-leavers worked, and their earnings combined with government subsidies and tax credits provided more income than was previously obtained through welfare. However, 85 percent of this group still had incomes below twice the poverty threshold (Acs & Loprest, 2004). Another study after welfare reform found that low-income mothers were less likely to rely on welfare, but were no more likely than their pre-reform counterparts to have incomes that exceeded 150 percent of the poverty threshold (Acs et al., 2005). These studies indicate that the goal of increasing low-wage work by former welfare recipients was not a panacea for reducing poverty rates. Table 1. Poverty Status of Families by Type of Family, For All Races With Children Under 18 Years, for Selected Years Note: Numbers in thousands. Families as of March of the following year All Families Female Householder Year No Husband Present Total Below Poverty Level Total Below Poverty Level Number Percent Number Percent 1975 EITC begins 31,377 4, ,119 2, EITC expands 33,801 5, ,094 3, EITC expands 34,503 5, ,707 3, Highest poverty 36,456 6, ,758 4, PWORWA begins 37,204 6, ,957 3, Lowest poverty 38,190 4, ,813 2, EITC expands 38,427 5, ,171 3, EITC expands 38,820 6, ,872 3, Latest data 38,274 6, ,036 4, U.S. Census, retrieved from DRAFT 24

25 Between 1995 and 2000, the economic situation of the very poor worsened (Blank, 2002). While overall poverty rates for single mothers fell dramatically during this period, the incomes of the poorest twenty percent of single mothers declined (Primus et al., 1999). About one in five welfare leavers did not work, did not live with someone who worked, or did not receive any public assistance (Loprest, 2001). Lindsey (2009) concurs that economic hardship for the poor became more severe after welfare reform, and that reliance on food stamps, WIC, and federally subsidized lunches increased to serve the growing numbers of children in poverty. According to Lindsey (2009), the increase in labor force participation after welfare reform was due to the improving economy. Concurrent with a contracting economy, parent employment declined between 2000 and 2004 (Lindsey, 2009). Another explanation for the deterioration of the economic position of low-income families is found in the federal tax code. By 1990, the income tax rate on the wealthiest Americans had become the lowest of any industrialized nation (Reich, 1992). Many benefits conferred by the IRS tax code continue to exacerbate inequality by favoring the higher income quintiles (Lindsey, 2009). The tax code provides several major deductions (e.g., deductions for self-employed health insurance, mortgage interest, student loan interest, charitable contributions, retirement savings, state and local property taxes, the child and dependent care credit, and preferential rates on capital gains and dividends). Of the deductions listed, the EITC contributes only a small fraction to the total value, and is the only benefit that helps low-income and poor families (Lindsey, 2009). Lindsey (2009) extends the analysis of Hoynes et al. (2006) by underscoring the effects of economic inequality to explain the persistence of poverty. In Lindsey s (2009) perspective, the pre-1980 era of the middle class has been replaced by the era of the wealthy class, and it DRAFT 25

26 has become increasingly difficult to break the cycle of poverty (p. 5). Economic conditions of the poor have worsened, while middle class income and wealth have stagnated, and the income and assets of the wealthiest, i.e., the top one percent and top ten percent, have more than doubled. These massive shifts in income and wealth have created the most dramatic increase in inequality in the nation s history (Lindsey, 2009, p. 6). Structural and Functional Ramifications In order to understand more fully what lies behind the claim that the EITC lifts people out of poverty, it is helpful to explore how the EITC differentially affects families at various points in the low- and moderate-income spectrum. The EITC is available only to those who work. Consequently, the poverty rate for p0or families with no earnings is not improved by the EITC (Acs &Toder, 2006). About one-third of families below the poverty line do not meet the earnings requirement even with a qualifying child (Hoffman & Seidman, 2003). In total, only about 35 percent of poor households can qualify for the EITC (Hoffman & Seidman, 2003). Holt (2006) observes, the EITC-eligible population overlaps with, but is distinct from the population defined as poor (p. 8). The EITC sharply reduces the numbers of families at targeted levels of 50 percent and 200 percent of the poverty line, with its strongest povertyreduction effects at 75 percent of the poverty threshold (Meyer, 2010). The average income for EITC-eligible households is about 125 percent of the poverty line (Hoffman & Seidman, 2003). In 1999, over half of EITC refunds benefitted families with incomes above the poverty threshold (Hotz &Scholz, 2001). In the 1990s, Liebman (1997) found that the EITC was an inefficient way of combating poverty because most EITC payments were transferred to householders with incomes spanning between fifty and 150 percent of the poverty level. Simpson et al. (2010) estimate that EITC DRAFT 26

27 eligibility extends to 35 percent of households below the poverty line, 35 percent of households between 100 percent and 150 percent of the poverty line, and 25 percent of families between 150 and 200 percent of the poverty threshold. EITC benefits are not exclusive to low-income families. While three-quarters of EITC dollars in 2004 aided households in the second and third deciles of the income distribution, benefits extended well into the fifth decile (Eissa & Hoynes, 2011). The EITC strengthens the incentive to leave welfare for full-time, low-wage work by phasing out the credit over a fairly wide swath of the income distribution (Liebman, 1997). Nearly two-thirds of EITC-eligible families are in the phase-out range of the program (Simpson et al., 2010). Generally, the hourly wages and annual earnings of low-wage workers are expected to increase through the years. However, even with a six percent annual increase, it would take ten years for a worker who started at $8 per hour to attain the median wage of $14.15 per hour (Acs & Toder, 2006; United States Department of Labor, 2012). Further, as wages increase, working families lose public supports, such as child care subsidies, and their EITC benefits decline during the phase-out period (Acs & Toder, 2006). Gershoff et al. (2003) find that work first approaches may be misguided if they are not accompanied by opportunities for skill development that could enable parents to land higher-paying jobs. A study based on 1989 data by Cossa, Heckman and Lochner (1999) found that the overall effects of the EITC on skill formation were negligible at best, controlling for the effects of the EITC on labor force entry and hours of work (Hotz & Scholz, 2001). Hotz and Scholz (2001) conclude that such findings indicate a need to understand the life cycle implications of EITC expansions relative to improving the skills and well-being of disadvantaged populations. DRAFT 27

28 Spending Patterns Another way of assessing how the EITC affects the economic situations of families is to examine how recipients spend EITC refunds. Tax refunds constitute 90 percent of EITC program costs; ten percent of EITC benefits reduce tax liability (Goodman-Bacon & McGranahan, 2008). Studies of family spending patterns for EITC refunds are limited. One line of inquiry concerns differentiating consumption expenditures from asset development or other forms of social mobility (Goodman-Bacon & McGranahan, 2008; Holt, 2006; Romich & Weisner, 2001, Smeeding, Phillips, & O Connor, 2000). The majority of EITC recipients use the additional income for short- and medium-term consumption needs, with fewer able to contribute to asset development (Holt, 2006). Holt (2006) suggests framing asset building more broadly to encompass savings aimed at avoiding future indebtedness (p. 17). Very poor families are usually not able to save any EITC refunds (Holt, 2006). Recipients with the largest EITC refunds are those least in a position to build assets, and at best may be able use the EITC as a hedge against future emergencies. For those few with higher incomes, the EITC can contribute to long-term savings (Holt, 2006). For example, in a survey conducted at free tax preparation sites in Milwaukee, seven percent indicated plans to use EITC refunds for investing in themselves or their children, and another seven percent intended to save for a house or other large expenditure (Goldberg & Percy, 2005). While it is possible for families to elect EITC refunds on a monthly basis, the vast majority of claimants choose to receive the refund as a lump sum consequent to their annual IRS filing, most often in February following the tax year (Barrow & McGranahan, 2000; Holt, 2006; Meyer, 2010; Romich & Weisner, 2001). The average amount refunded by the EITC is $1,500 (Meyer, 2010). Families are more likely to exert spending discipline with a lump sum EITC DRAFT 28

29 refund because it is considered as wealth in contrast to other forms of income, and its receipt is an anticipated annual event (Romich & Weisner, 2001, p. 212). EITC refunds are used more often to purchase durable goods than non-durable goods (Goodman-Bacon & McGranahan, 2008). Vehicles were the most popular durable expenditure, and transportation costs were the most common non-durable expense. The purchasing of vehicles and transportation were considered consistent with the overall goal of the EITC work ethic (Goodman-Bacon & McGranahan, 2008). In a study conducted among EITC recipients in Chicago for the filing year 1997, respondents indicated multiple uses for their individual refunds. Sixty-five percent of the study sample (N=1,226) used EITC refunds averaging $1850 per claimant and a smaller amount of other tax refunds for consumption spending to make ends meet (e.g., to pay bills and buy food). Seventy percent planned to use the refund for what was deemed economic and social mobility purposes. The authors identified car-related expenses, tuition, savings, and moving as economic and social mobility uses of the EITC and concluded that the EITC plays a large role in improving economic and social mobility (Smeeding et al., 2000, p. 1198). However, billpaying was the single most important priority for half the sample, and among the top three priorities for 83 percent of the sample (Smeeding et al., 2000). Paying bills was also the highest priority for 60 percent of family and individual EITC recipients for tax years between 2002 and 2004 in a rural county in upstate New York (N=432) (Simpson, Tiefenthaler, & Hyde, 2010). Purchasing items and paying housing costs were the next highest priorities for this group. Savings was only a second or third priority for about ten percent of respondents. Credit card payments and medical bills were among the most important bills to be paid, indicating that diverting EITC funds toward saving could result in late fees or DRAFT 29

30 increased interest charges for the recipients (Simpson et al., 2010). Overall, two-thirds of the sample intended to use their refunds to meet basic needs (Simpson et al., 2010). About one-fifth of claimants reported that they absolutely could not live without the refund (Simpson et al., 2010, p. 853). In contrast to Smeeding et al. (2000), Simpson et al. (2010) find that the EITC helps the majority of recipients get by but does not increase their economic mobility Clearly, the EITC is not generally viewed as a way to save significant amounts of money to make investments that will lead to economic mobility (pp. 843, 860). Economic Mobility While the EITC is lauded for lifting people out of poverty, few studies differentiate its function as a temporary safety net versus as a long-term support. Dowd and Horowitz (2011) investigate the EITC program relative to these functions for the years between 1989 and 2006, primarily for claimants with children. During this period, approximately 50 percent of taxpayers with children were EITC recipients, if only for a year or two (Dowd & Horowitz, 2011). The EITC program is characterized by churning, i.e., the ongoing movement of families into and out of the parameters of EITC eligibility (Dowd & Horowitz, 2011). Dowd and Horowitz (2011) identify four areas of change that prompt such cycling: (1) earnings; (2) eligibility of children; (3) family structure relevant to spousal income; and (4) EITC expansions. Entries into the EITC program primarily derive from two sources: households who earn less than $100 a year and then earn more (19%), and families whose earnings decrease after previously earning income that exceeded the phase-out point of the EITC (53%). The remaining entries relate to wives becoming household heads or the arrival of an eligible child (Horowitz, 2002). Families exit the EITC program primarily because they increase their earnings beyond the phase-out point (43%) or exit the labor force (15%). A common reason for entries is the DRAFT 30

31 birth of a baby, which is associated with decreased income and employment hiatuses (Horowitz, 2002). Forty-three percent of families are likely to claim the EITC after the birth of a child (Dowd & Horowitz, 2011). Family income generally increases in the years following the birth (Dowd & Horowitz, 2011). Sixty-one percent of recipients claim EITC benefits for one or two years, while the average spell is roughly three years (Dowd & Horowitz, 2011). Spells are typically longer for young families with single mothers and families with more children. Average spells are shortest for families who enter the EITC program in the phase-out range (2.7 years), and longest for families who enter at the plateau phase (3.9 years). Almost one-third of EITC recipients exit the program each year. The percentage of re-entries is high, but decreases as years of non-eitc participation pass. Twenty percent of those who leave reclaim the EITC after one year, and a total of 44 percent reclaim eligibility over time. EITC expansions lead to increases in re-entries (Dowd & Horowitz, 2011). The largest growth in income for EITC recipients is among those who claim benefits only once or twice, and occurs most often after the first year of claiming the credit. For those who claim the EITC for short intervals due to a cutback in work hours or a temporary job loss, the EITC functions as a short-term safety net. But for those who receive EITC benefits for many years and who have many spells of cycling in and out of eligibility, the EITC serves as a longterm support. There is less income mobility among EITC claimants than in the general population (Dowd & Horowitz, 2011). For the rural sample of EITC claimants in New York State between 2002 and 2004, Simpson et al. (2010) conclude: DRAFT 31