Testimony before the Committee on the Budget, U.S. House of Representatives, Washington, D.C. January 28, 2014

Chairman Ryan, Ranking Member Van Hollen, and Members of the Committee, thank you for inviting me to appear today to discuss the central issues of poverty and opportunity in America. The fiftieth anniversary of Lyndon Johnson’s declaration of war on poverty provides an occasion to reflect on federal anti-poverty policy to date and policies to promote opportunity and to consider how to build on the progress we have made while improving on those features of policy that have not served us well.

I will begin with a discussion of trends in poverty and focus specifically on child poverty, since that is the most relevant consideration for thinking about opportunity. I will argue that trends in poverty are brighter than many believe, but the war on poverty deserves limited credit for poverty reduction among children and in fact, by discouraging work during economic expansions and reducing the cost of single parenthood, it may have prevented poverty from falling more.

Regardless, even as poverty has declined, upward mobility has failed to increase, so the war on poverty has not been an effective war on immobility. An opportunity agenda that places too much emphasis on how much money the poor have—and on how much the rich have—will continue to be ineffective.

I will close with some policy ideas for an opportunity agenda that draws from the lessons of the last fifty years that economic growth and a work-promoting safety net are the most effective ways to reduce poverty. Because moving the needle on upward mobility will likely require more than poverty reduction, I believe that an opportunity agenda must also empower poor parents to invest in the skills of their children. But soft-hearted policies must also be hard-headed ones. We will have to discover how to increase the school readiness of poor children and keep them on track. A system of opportunity grants would make markets for child investment services, uncover successful models, facilitate the dismantling of ineffective ones, and potentially raise parental aspirations in communities with deficits of hope for their children.

A Limited Victory in the War on Poverty

It is surprisingly difficult assessing how we have done reducing poverty in America. Official figures from the Census Bureau indicate that after dropping like a rock in the 1960s, the poverty rate barely budged in the 1970s, rose in the 1980s, and was no better in 2007 than in 1989.[2] Twelve percent of Americans were poor in 1979 and 12 percent were poor in 2007. The poverty rate was 15 percent from 2010 through 2012, matching the levels in the early-1980s recession but otherwise the highest since 1965.

The trends for Americans differ, however, depending on their age. Among those over 65, poverty fell dramatically through the 1970s, fell some more in the 1980s and 1990s, then flattened out at an historically low rate.[3] One in three senior citizens was poor in 1959, but just one in 11 was in 2012, a remarkable decline.

Among children, however, the poverty rate rose during the 1970s and 1980s and at 20 percent in 1989 was only slightly lower than in 1965.[4] Then it fell in the 1990s before rising again since then. At 22 percent, the 2012 child poverty rate was worse than in 1965. This suggests we have made little progress on opportunity, to the extent that child poverty either affects upward mobility or stands in for things that do.

However, the official figures are limited in a number of well-known ways. To determine if someone is poor, the Census Bureau counts income received from private sources or from government employment and adds federal benefits if they take the form of a check—so-called “cash transfers” like the Temporary Assistance for Needy Families and Supplemental Security Income programs for poor families, Unemployment Insurance for the jobless, and Social Security for senior citizens. It compares a person’s family income—or their own income if they do not live with relatives—to a poverty line that varies depending on family size and the age of its head. The original poverty line was constructed in the mid-1960s and has been updated annually to reflect increases in the cost of living. While it was originally based in real-world analyses of what families needed to get by, it is best thought of today as an arbitrary but relatively low level of material well-being meant to be held constant over time.

The official measure does not count employer-provided benefits as income, and it does not count non-cash benefits from the federal government, such as Medicare, Medicaid, food stamps, or housing subsidies. Nor does it deduct taxes from income, which means that if the tax burden falls or the value of benefits provided through the tax code increases, improvement in living standards will be understated. But perhaps most importantly, because the official poverty line is adjusted for the cost of living every year by a measure, the “CPI-U,” that is known to overstate inflation, it represents a higher living standard than it used to, making the change in poverty look too dour.[5]

Efforts to improve the official poverty measure suggest that we have made considerable progress not just in reducing the poverty rate but in making child poverty rarer. Researchers from Columbia University recently extended the Census Bureau’s new “Supplemental Poverty Measure” back in time from 2010.[6] This measure corrects most of the official measure’s problems though it too probably understates improvement over time.[7] The “SPM” indicates that poverty fell in the 1970s and fell by more in the 1990s than the official measure suggests. In 1969, 22 percent of Americans were poor, according to the new measure, while just 16 percent were in 2012.[8]

Nevertheless, child poverty was still higher in 1989 than in 1969 under the new measure. In the 1990s, however, it fell to a historic low and was basically unchanged until the Great Recession, when it inched up. While a quarter of children were poor in both 1969 and 1989, just 18 percent were in 2000 and 19 percent in 2012.

We have, therefore, made progress reducing poverty. However, the liberal New Deal and Great Society programs we associate with the war on poverty turn out to play a more limited role than we might have guessed, at least when it comes to child poverty. The Columbia SPM series suggests that Social Security was an important reason for declining poverty among senior citizens during the 1960s and 1970s, and an important reason why their poverty rates remained low thereafter.[9] (The same could be said of Medicare and Medicaid, though they are not counted as income in the SPM.) There are two important caveats here, however. Other ways of providing federal retirement benefits, such as through personal retirement accounts or premium support for health insurance, might have achieved the same effect or an even stronger reduction in poverty. Furthermore, senior entitlements threaten to engulf the rest of the federal budget, which would crowd out any federal spending to expand child opportunity.

In contrast to the trend in poverty among seniors, nearly the whole story for children is the period between 1993 and 2000. This was the strongest period of income growth since the 1960s and points to the importance of robust economic growth in reducing poverty. But the Columbia SPM research suggests that child poverty would not have fallen as much during these years if not for federal taxes and transfers. That suggests that the work-oriented welfare reforms of the 1990s helped to reduce poverty by encouraging low-income adults to enter the workforce.

The safety net for non-working families became less generous during these years even as it became more generous for working families. The New Deal-era Aid to Families with Dependent Children program was replaced by Temporary Assistance to Needy Families through landmark 1996 legislation, but for years before that states had been experimenting with welfare reforms through the expanded use of federal waivers under Presidents George H. W. Bush and Bill Clinton. After 1996, cash assistance was block-granted and time-limited, and work and job-search requirements were mandated. Teen mothers were required to live at home and to receive education or training to receive benefits, and states could cap the number of children who were eligible for benefits.

However, welfare reform during the 1990s included generous carrots as well as sticks in attempting to move low-income adults into work. In particular, the Earned Income Tax Credit was expanded, child care subsidies increased, and it became easier to keep health insurance benefits upon taking work, reducing the marginal tax rates faced by the non-working poor when they entered the job market.

One reason to think that it was the shift to work promotion that lowered child poverty rather than traditional expectations-free safety net programs is that the Columbia SPM research indicates that subtracting out federal benefits and looking at pretax income does not alter the trend in child poverty much prior to the 1990s. Furthermore, in research I conducted with Harvard sociologist Christopher Jencks, we found that the 1990s expansion differed from those of the 1960s, 1970s, and 1980s in that the poverty rate for families headed by a single mother fell by more than it did among two-parent families.[10]

The child poverty story also merits two caveats. First, the traditional safety net probably did prevent child poverty from rising in recent years, starting with the 2000 recession and continuing through 2012. The Columbia SPM figures indicate that poverty among children would have begun rising if not for federal benefits and the impact of tax policy. Instead, it fell a little more. Our federal safety nets should promote work, but we must be prepared to assist those who cannot find work, a delicate balancing act.

Second, the Great Society liberalization of welfare rules may have contributed to the increase in single motherhood and even the decline of work participation rates among less-skilled men. The evidence for such an effect is weak, but the question itself is difficult to answer convincingly. Whether welfare reform was a major factor behind the dramatic drop in teen pregnancy during and after the 1990s is an under-studied research question. One study found that welfare reform in the 1990s reduced teenage motherhood among the daughters of single or less-educated parents and encouraged them to live with a spouse or parent if they did become mothers.[11]

A War on Poverty is Not a War on Immobility

For the policymakers behind the Great Society, a war on poverty was expected to produce more upward mobility. The point was not simply to make each generation less poor, it was to give poor children more opportunity to reach the middle class. However, parental income was given too prominent a role in explaining why the upward mobility of poor children was not greater. Indeed, policymakers in the 1960s did not even have data allowing them to say anything about the extent of upward mobility. Today we can say that upward mobility was, in fact, limited fifty years ago. But despite reducing child poverty, it remains no greater today.

The academic literature on intergenerational income mobility trends has become fairly sizable, but no research shows a significant increase in mobility since the mid-twentieth century. The most common finding is a change so modest as to be statistically indistinguishable from no change at all.[12] Earlier this month, Harvard economist Raj Chetty and a team of researchers confirmed this consensus with a paper finding that children born in 1993 likely had experienced the same mobility as those born in 1971.[13]In my own forthcoming research, I find that today’s thirty year olds—who experienced rising income inequality between the middle class and poor during early childhood and have witnessed rising income concentration at the top through their entire lives—have experienced no more and no less mobility than did thirty year olds in the mid-1970s.[14]

Many find it difficult to believe that mobility has not declined given that inequality has increased, but even today we exaggerate the importance for mobility of how much one’s parents—or someone else’s parents—make. The link between income inequality and mobility has been greatly overstated, as I have argued elsewhere.[15]

But the fact that mobility has not fallen should provide little comfort given how limited upward mobility remains. Only thirty percent of today’s adults who were raised in the bottom fifth of household incomes managed to make it into the middle fifth or higher.[16]

An Opportunity Agenda for a War on Immobility

What can we learn from the war on poverty and trends in poverty and mobility over the past fifty years? The experience of the 1990s demonstrates, first, the importance of economic growth to create a rising tide that will lift all boats. Economic growth is crucial to efforts to expand opportunity in two senses.

First, even if we fail to more children raised in the lower ranks to higher ranks by the time they are adults, there is a lot to be said for simply ensuring they are better off than their parents were. That is, stagnant mobility—with children from lower ranks remaining there as adults and a corresponding absence of movement among higher-ranking children—is less severe when everyone is moving up in absolute terms. Indeed, while mobility across ranks of the income distribution remains too low, 84 percent of adults today have higher incomes than their parents did, including 93 percent of those who were raised in the bottom fifth.[17]

There is a second reason to prioritize economic growth as part of an opportunity agenda. Moving more people up the ranks, frankly, requires that others move down. To the extent that this promotes economic efficiency, making the country more of a meritocracy ultimately will enrich the country. But if the economy is growing robustly, the dislocations experienced by those downwardly mobile in rank will be mitigated by continued upward mobility in income. That is to say, the person who falls from the top tenth as a child to the next-highest tenth in adulthood might still end up better off in absolute terms. Economic growth greases the wheels of meritocracy.

I have elsewhere discussed a number of ways to promote economic growth, but they might include:[18]

·       Pairing cuts in corporate and individual investment taxes to encourage job creation with increases in federal research-and-development spending to promote innovation. Both policies would be likely to pay for themselves down the road.

·       Greater high-skilled immigration to increase the stock of human capital from which our economy can draw for innovative ideas and to ensure competitive labor markets among professionals.

·       Reform of senior entitlements to bring down future deficits and keep our debt-to-GDP ratio manageable.

·       Health care reform, both as part of deficit reduction (because scheduled provider cuts are unlikely to be implemented) and to prevent the excessive health care inflation that Obamacare’s subsidies and mandated benefits are likely to create.

The experience of the 1990s offers another lesson for a successful opportunity agenda. Safety-net reforms that encourage work can reduce poverty by fostering initiative and lowering marginal tax rates. Welfare reform was successful by replacing a program with minimal reciprocal expectations of recipients and severe work disincentives with a social policy regime in which work clearly paid off. Yet, many of our safety-net policies still ask little of beneficiaries and retain high marginal tax rates. For most people, they serve as a temporary stopgap measure in hard times, but for others, especially during economic expansions, they become poverty traps, discouraging work, marriage, and saving. As Charles Murray long ago argued, the problem is not so much one of personal failure but that people are responding to the incentives embedded in our safety-net policies as any of us would in the same situation.

Two ideas that address the work disincentives of our sprawling and uncoordinated safety net regime have recently been offered. Oren Cass, former domestic-policy director of Mitt Romney’s 2012 presidential campaign, has proposed block-granting our means-tested programs and sending them to the states as a “flex fund”. The flex fund would be coupled with an expansion and reconfiguration of the work subsidies currently offered through the Earned Income Tax Credit. Senator Marco Rubio is developing a proposal along these lines, which would have the benefit of encouraging policy experimentation at the state level while increasing the incentives to work among those whose wage prospects are modest.

Relatedly, Congressman Paul Ryan, chairman of this committee, has spoken sympathetically toward the United Kingdom’s new Universal Credit program. The Universal Credit packages a number of scattered safety-net programs together and delivers a simplified benefit to those who qualify. Most importantly, it is carefully designed so that the size of the credit tapers as earnings become a bigger share of household income, but in such a way as to encourage beneficiaries to work more.

These are encouraging generalizations of the 1990s welfare reform. It is worth highlighting two issues regarding work-promoting safety net reforms, however. First, while the 1990s reforms clearly reduced poverty, we lack the evidence so far—because the 1990s were not that long ago—that they increased upward mobility. If, as I have argued, more money is not enough to expand mobility, then safety net reforms may have the biggest impact on child mobility to the extent that they affect the aspirations, values, and family lives of poor children and their parents.

The decline in teen pregnancy offers a possible hint that safety-net reforms can affect behavior in a way that might promote upward mobility. In addition to promoting work through the tax code or a universal credit, it may be desirable to promote marital childbearing as well. Out of wedlock childbearing has increased markedly, and I believe we are approaching the problem from the wrong perspective. Though research overwhelmingly suggests that children who grow up with a single parent tend to have poor outcomes, it does not establish that the children would typically have done any better if their parents—their specific parents, not generalized parents with some other set of skills, values, and assets—had married or stayed married. Anyone who has seen MTV’s “Sixteen and Pregnant” and the complicated lives of the show’s protagonists ought to recognize that the children born to these couples have a lot working against their success regardless of whether their parents marry or not.

However, even if it were the case that the children born to single mothers face long odds regardless of their parents’ marital status, by encouraging young men and women to delay childbearing until they are in a better place, policies might have an important impact on mobility rates. In that case, we would not be improving the opportunities of children in disrupted families by encouraging their parents to stay together. Instead, we would be improving child opportunities by preventing children from being born into disrupted families and encouraging childbearing among those same men and women when their lives are more conducive to successfully raising a child. Just as a work supplement can send a clear signal about the rewards for behaving in opportunity-enhancing ways, so too a married-parent credit could affect thinking at the margins when young men and women contemplate their birth timing.

A second issue related to work-based safety-net reforms is that we will always need a safety net to catch those who cannot secure stable employment. The overlap between children with low mobility prospects and children whose parents will struggle to find and keep work is likely to be substantial. One easy way to minimize the number of people who fall through the cracks is to build a counter-cyclical element into any block-grant or universal credit regime, so that the system can respond appropriately in downturns as unemployment worsens.

A modern opportunity agenda should include a third component alongside growth-promoting policies and safety net reforms.[19] Encouraging economic growth will not help those parents with the worst jobs keep their kids from filling those bad jobs someday themselves. Welfare reforms may expand child opportunity, but only indirectly, and they may only reduce poverty without actually nudging mobility. Winning the war on immobility, I believe, will require a program to help poor parents invest in their young children. If parents are unable to ensure their children’s school-readiness and keep them on track academically, the federal government can empower them to find the help they need.

The problem is that we have astonishingly few early- and middle-childhood models that have been shown to work on a large scale. But we should nevertheless commit substantial resources to discovering successful models, even as we also commit to shuttering existing programs that have not proved effective. A system of opportunity grants for poor children would allow low-income parents to pay qualifying providers for any of a range of eligible child investment services—after-school programs, tutoring, summer enrichment programs or other strategies. Providers would have to agree to be evaluated, and consistently ineffective providers—and approaches—would be excluded from receiving grants as the evidence comes in.

This approach would be “market making” in the sense that it would incentivize the supply of child investment services and encourage parents to seek them out. Ideally, the circulation of opportunity grants in low-income communities would inspire competition among parents to ensure they are doing right by their children, potentially altering community norms and aspirations. To be sure, many—probably most—models initially would be revealed to be ineffective, but that will build consensus about the limits of what social policy can do and about the need not to waste money on approaches that do not work. And such a program would discover workable models and seed successful ventures like the KIPP schools, becoming much more obviously cost-effective in time. If the opportunity grant program succeeds, it would open up the conversation around K-12 reform as well and point to a new federal role in education.

We can do better than a limited win against child poverty with stagnant upward mobility. We can and should wage a war against immobility that builds on the lessons of the past fifty years, embraces fundamentally modern policy approaches, and achieves something that the Great Society could not—a restoration of faith in the American Dream.