As usual, on the 50th anniversary of the War on Poverty Washington has fallen into partisan bickering over whether it was successful. Most Republicans say it was a bust and most Democrats say it was a smashing success. The truth is that the War on Poverty produced mixed results but on balance has been good for the nation, especially the elderly.
On the good side of the ledger, I would count the dramatic decline in poverty and improvement in health among the elderly as a solid success of American social policy. Poverty among the elderly declined primarily because of Social Security. Although Social Security was established during the New Deal in 1935, President Johnson increased its cash benefits during the War on Poverty. Subsequent Presidents and Congresses followed the Johnson pattern and also increased Social Security benefits, almost always on a bipartisan basis. Primarily as a result of these increased benefits, poverty among the elderly fell to 25 percent in 1968 from 35 percent in 1959, a reduction of almost 30 percent during the Johnson Presidency. Elderly poverty continued to fall after 1968. In 2012, a year in which the overall poverty rate was 15 percent and the child poverty rate nearly 22 percent, the poverty rate among the elderly was about 9 percent.
Similarly, Johnson’s Medicare program has protected the health of the elderly, and redistributes resources to the elderly overall and especially the poor elderly, most of whom would not be able to pay for health insurance out of their savings or their income. Work by Henry Aaron and Gary Burtless of Brookings shows that the cash value of the health services received by the elderly in the bottom 10 percent of income distribution would cost an amount equal to about 85 percent of their income if they had to pay the full cost out of pocket – and for many, it would be well above 100 percent of their income in some years. Medicare has been successful because it has eased both the physical burden of poor health and the psychological burden the aging would bear if they had to worry about paying to maintain their health. But there are serious financing problems with Medicare, and with Social Security to a lesser degree.
Democrats often question the nation’s commitment to the poor and urge expansion of existing means-tested programs or adoption of new programs. However, as Republicans need to acknowledge the successes of the War on Poverty, Democrats should acknowledge that federal and state spending on the poor has increased almost every year in inflation adjusted dollars since the War on Poverty began. As I showed in testimony before this Committee in 2012, between the federal government and the states we now spend around a trillion dollars a year on programs for poor and low-income people. Counting only the federal spending, between 1980 and 2011spending per person in poverty increased from about $4,300 to about $13,000 or more than $3 spent in 2011 for every dollar spent in 1980. These figures do not include any money from Unemployment Compensation, Social Security, or Medicare, all of which also redistribute resources down the income distribution.
One measure of what we purchased with all this spending is provided by examining poverty rates. Using the official poverty rate, poverty declined about 30 percent within five years of President Johnson’s declaration of his War in 1964, but there has been little progress since the 1960s (except among the elderly). For the last two decades, poverty has averaged well above the 12.1 percent achieved when Johnson left office in 1969. For the last three years, the poverty rate has been at or above 15 percent. However, the official measure of poverty is deeply flawed, especially because it only counts cash transfer payments and ignores the billions of dollars the federal and state governments pay to low-income households through the tax code in refundable credits. As a result of these flaws, the Census Bureau has wisely developed a much better poverty measure, called the Supplemental Poverty Measure (SPM), which does take these government transfer payments into account. A group of scholars at Columbia has recently used the SPM and income and expenditure data from federal surveys to provide what most analysts would agree is a more accurate estimate of U.S. poverty rates since the 1960s. They estimate that without assistance from major government programs, the poverty rate in 2012 would have been 31percent. But when government benefits (not including health programs) are included in income, the poverty rate was 16 percent, thereby demonstrating that government benefits reduced poverty by around half in 2012. By comparison, government benefits in 1967 cut poverty only by about one-quarter. It must be granted that cutting poverty by half is an important achievement. Moreover, analyses of the impact of government benefits during the recession, also using a broader measure of benefits, showed that poverty did not increase at all in 2008 in the teeth of the most severe economic downturn since the Depression.
Although fighting poverty is a worthy goal, it is by no means the sole measure of the effectiveness of the nation’s social policy. As suggested by the title of this hearing, ensuring that every American child has the opportunity to get ahead is also a major goal of the nation’s social policy. In turning to a discussion of economic opportunity, it is useful to begin by reviewing the status of this debate, then turning to a brief overview of major trends in the American economy and society that are pressuring poverty rates to rise and reducing the odds that young people can move up the income distribution, and finally offering a few ideas of how the federal government could adopt policies that would reduce poverty and increase opportunity.
In recent speeches, President Obama has been emphasizing inequality. I hope that members of this committee and of Congress more generally will not fall into the inequality trap. From a political perspective, inequality is a red flag, in large part because, as demonstrated by the President’s framing of the issue, discussion quickly turns to criticizing the rich. The problems with this framing are not only political but substantive as well. If our goal is to increase opportunity, it seems unlikely that limiting income at the top of the distribution or taking more money from the rich will increase opportunity. We already have a highly progressive income tax code in which the top twenty percent of earners pay about 93 percent of income taxes and the top one percent of earners pay nearly 40 percent of income taxes, while the bottom forty percent pay a negative 5 percent of income taxes (the federal government sends them money through refundable tax credits). Meanwhile, we spend the $1 trillion on programs for poor and low-income Americans discussed above, demonstrating that we already have government redistribution of income on a large scale. Some argue that we should tax the rich even more and provide even more means-tested benefits to poor and low-income people, but the debate must acknowledge that our current system already redistributes a great deal of money and dramatically reduces poverty rates.
Similarly, the President and many on the left argue that the middle class is emptying out and that people at the bottom of the distribution have suffered from declining income in recent decades. But the Congressional Budget Office (CBO) has now published several reports presenting estimates of how much income has changed since 1979 all along the income distribution. Their 2011 report divided the income distribution into five parts (called “quintiles”) with an equal number of households in each quintile. CBO found, using a broad definition of income that included an inclusive measure of the value of health care benefits provided by government, that the average income of households in all five quintiles increased between 1979 and 2007: the bottom quintile increased by 18 percent, the average of the middle three quintiles (the “middle class” broadly defined) by 40 percent, the top quintile (not including the top 1 percent) by 65 percent, and the top 1 percent by 275 percent. The major reason for the large increases in income at the top of the distribution is that income from capital gains and business income increased so much more than other types of income and households in the top quintile have lots of income of both types. Thus, the CBO study shows that income inequality increased all along the distribution, especially at the top, but also that every quintile did better.
Another complaint on the left is that income mobility has declined. This claim is not consistent with multigenerational income data on parents and their children from the Panel Study of Income Dynamics and other data sources. Most recently, a study conducted by Raj Chetty and his colleagues at Harvard concluded that “measures of intergenerational mobility have remained extremely stable for the 1971-1993 cohorts.”
It would increase the productivity and reduce the heat of the Washington debate on poverty and opportunity if all sides agreed to base their understanding of poverty and mobility on the basic facts laid out above, all from reliable sources. Here is a summary of these facts:
- Government spending on poor and low-income families has increased almost every year for five decades; since 1980 spending has tripled as measured on a per person in poverty basis in constant dollars
- An improved measure of poverty shows that government spending focused on poor and low-income households cuts the poverty rate by about half; government spending on these programs and the Unemployment Compensation program prevented poverty from increasing during the most severe recession since the Great Depression
- A CBO analysis shows that when government benefits are counted at their full value, households all along in income distribution, including those in the bottom 20 percent, enjoyed increased income between 1979 and 2007
- The same CBO report shows that because income increased more the higher up we look in the income distribution between 1979 and 2007, income inequality has also increased since 1979; by far the biggest increase in inequality is between the top and the rest of the distribution
- Several studies show that claims that intergenerational income mobility has slowed down in the U.S. are false; the U.S. has less income mobility that many European nations, but mobility has remained constant over the past four decades or so; nonetheless, children whose parents were in the bottom 20 percent of the income distribution have more than a 40 percent chance of staying in the bottom themselves.
Examining these generalizations, it seems reasonable to conclude that the nation has made some progress against poverty, that people all along the income distribution are doing better in recent years due in part to government policy, and that despite the big increase in income inequality, economic mobility has not changed. So what’s the problem? The problem is that poverty rates, especially among children, are too high. But even more important, especially for a nation that prides itself on the opportunities we offer to the “the tired, the poor, the huddled masses,” hardly anyone thinks we should be satisfied with the opportunities we offer to poor children when they are twice as likely to wind up at the bottom of the income distribution as children from more affluent families and when many other nations with advanced economies have more economic mobility than we do.
Powerful Forces Pushing Poverty and Mobility in the Wrong Direction
In searching for ways to reduce poverty and promote mobility, a good place to start is with the trends in American society that, in effect, constitute a tide moving against attempts to reduce poverty and increase opportunity. Four trends are especially important: dissolution of the traditional family, declines in work rates among males, failure of parents and the public schools to close the gap in learning between children from poor and affluent families; and the increasing fraction of well-paying jobs that require more education and greater skills than in the past. A word about each of these trends is in order.
Changes in marriage rates over the past decade are astounding. Between 1970 and 2010, the percentage of 35 year old women living in married-couple families with children fell from about 78 percent to 50 percent while the percentage of women who were single and living with children more than doubled from 9 percent to over 20 percent. This demographic trend has two major effects that work against our goals of reducing poverty and increasing mobility. First, children in single-parent families are four times as likely to be poor as children in married-couple families. Thus, the rising share of children in female-headed families is a major force pushing up the child poverty rate. No one thinks being reared in poverty is good for children and their development. Second, there is now all but universal agreement that the best rearing environment for children is a married-couple family. Not only do married-couple families have more money to invest in their children, but they also spend more time with their children and use child rearing techniques that are more conductive to child development. Research shows that disadvantaged parents, usually single mothers, spend less time with their children, talk with them less, and are more likely to use child-rearing techniques that are associated with poor developmental outcomes, especially corporal punishment.
The decline in work rates among males is also a trend that pushes up poverty rates and may also be one of the underlying reasons that marriage rates have been declining. The work rates of men with a high school degree or less between ages 25 and 55, the prime earning years, fell from over 90 percent in 1970 to under 75 percent in 2010. For young black males between ages of 20 and 24, the work rate is now well below 50 percent. These figures are especially discouraging in view of the fact that the poverty rate among those who work full time is less than 3 percent while the poverty rate for part-time workers is about 17 percent and the rate for those who don’t work at all is 33 percent. Men with low work rates are not likely to provide a firm economic basis for family formation. Thus, the declining work rates of men and the declining marriage rates of women are likely to be related.
Given the rising importance of education to enabling young adults to find well-paying jobs, the link between education and both poverty and mobility is accepted by virtually everyone who has studied the impact of education. But the trends, again, are similar to those for marriage rates and work rates because they are moving in a direction that undermines our policies to reduce poverty and increase opportunity. A recent comprehensive study of the available data by Sean Reardon of Stanford found that disadvantaged children enter the public schools already significantly behind their more advantaged peers in both literacy and numeracy and the public schools fail to help them catch up. In fact, the gap in achievement scores between poor and more advantaged children has been increasing for more than four decades.
The most straightforward way to reduce poverty and increase opportunity would be to reverse these trends. However, the changes in family composition have been proceeding for more than four decades and show no signs of abating, despite a host of efforts by policy makers. Similarly, the schools have been in a state of almost permanent reform since publication of A Nation at Risk in 1984, but the trend of poor children falling further behind in achievement has not been reversed. In fact, as technology continues to increase the skills needed to secure well-paying jobs, we can expect this trend to intensify and to further reduce the odds against children from poor families using education as a ladder to success. The decline in work rates among males also shows no signs of abating. Changes in personal behavior, including completing high school, getting a job, and marrying before child bearing would have an enormous impact on poverty and opportunity, at least as much as government benefits. But it turns out that these behaviors are resistant to change, leaving government to help individuals and families to overcome their mistakes.
What to Do
Government is already doing a lot to fight poverty and increase opportunity. We spend more money per student on education that any nation in the world, we have reformed out tax code on several occasions to help reduce marriage penalties, we spend considerable sums of money to help women (including single women) control their fertility which reduces the number of children in single-parent families, and we have dramatically increased government support for low-income workers with children. There is good evidence that some of these strategies produce benefits, but the review above shows that unless we’re satisfied with high poverty rates and great differences in the economic opportunity we provide for poor children based on their family income, we need to strengthen the policies that are showing success and try new policies. In what follows, I divide recommendations into two categories: policies for which there is strong evidence that they would work and policies for which there is more modest evidence of success.
Work and Work Supports. Other than giving money to the poor as we did with Social Security benefits for the elderly, arguably the best strategy the federal government and the states have adopted to reduce poverty is to encourage cajole, or force poor mothers on welfare to work. In 1996 congress passed bipartisan welfare reform legislation that was signed by President Clinton that provided very strong incentives, both negative and positive, for mothers on welfare and even low-income mothers who were not on welfare, to work. That was half the new strategy for reducing poverty. The other half was a series of legislative reforms, enacted over more than a decade, which greatly expanded the incentive to work by making work pay. Reforms in the Earned Income Tax Credit, Medicaid for children, child care, and food stamps ensured not only that mothers would not lose all their benefits when they left welfare but that they would be better off financially if they got even a low-wage job. The year after welfare reform, Congress enacted the Child Tax Credit and made it refundable, thereby proving even more incentive for low-income families to work.
The result was that low-income mothers went to work in droves. Following welfare reform, there was a 40 percent increase in the percentage of never-married mothers (the most disadvantaged group of single mothers and the most likely to be on welfare) with jobs. Moreover, the poverty rated among both children in female-headed families and black children declined to their lowest level ever. Today, after two recessions, one of which was the most serious since the Great Depression, never-married mothers are still more likely to have a job that before welfare reform and poverty rates for children in female-headed families and for black children are still lower than they were before welfare reform.
There is virtually no question that if we could help more poor mothers work, even at low wages, child poverty would decline and there is some chance that children would have a better chance to get ahead. There is a considerable amount of evidence, for example, that reducing the poverty rate in early childhood would promote child development.
To promote work among poor mothers and other low-income adults, Congress should strengthen work requirements in both housing programs and the food stamp program. Especially today with the nation’s high unemployment rate, just imposing work requirements would be likely to have only a modest effect. Thus, Congress should experiment with policies that would stimulate job growth in the low-wage job market. When the federal government gave states $5 billion in emergency funds in 2009 during the recession to help families on welfare, many states used the money to subsidize jobs. In fact, states used the funds to subsidize 260,000 jobs, many of them in the private sector, for mothers on welfare. The federal government should offer matching funds to states to repeat this performance and pay for studies in several states to determine the most effective way to subsidize these jobs.
Senator Marco Rubio has proposed major reforms of how government supports the poor. His general idea, following a proposal made by Oren Cass, the policy director for the presidential campaign of Mitt Romney, is to create two systems of benefits, one for those who don’t work and one for those who do. Senator Rubio’s proposal needs greater specificity, but his basic idea is to create more work incentive by reforming the Earned Income Tax Credit (EITC) and by converting noncash benefits now received by the working poor (perhaps including food stamps and housing) into cash benefits. Further, whereas the EITC is now paid only once a year, Rubio would pay it every month, thereby increasing work incentive because workers would get more immediate cash benefits. Converting in-kind benefits like food stamps into cash would also increase work incentive because workers prefer cash so they maximize their own preferences. Depending on the administrative arrangements, much of this policy could be revenue neutral. If the reform results in greater work incentive without spending additional federal dollars, this is a policy well worth trying out in several states while carefully studying its implementation. It can be expected that implementation will raise a host of thorny administrative issues.
Minimum Wage/EITC. Recently, my Brookings colleagues Isabel Sawhill and Quentin Karpilow have proposed combining an increase in the minimum wage with cost-neutral reforms of the Earned Income Tax Credit. Among other changes, they would greatly increase the cash benefit for single workers and strengthen the work requirement for EITC eligibility. They propose combining the EITC reforms with an increase in the minimum wage to $10.10 per hour. Many Republicans might find it difficult to support this large increase in the minimum wage, but there is no reason the proposal could not be combined with an increase in the minimum wage to something in the $9 per hour range. In any case, their EITC reform to increase the wage subsidy for single workers harmonizes well with the proposal made below to help single workers. The minimum wage/EITC package analyzed by Sawhill and Karpilow would reduce poverty by 7 percent by raising 3.4 million people above the poverty line; the poverty rate among single-parent families (mostly females) would fall by 8 percent. As pointed out below, increasing the income of young single workers could have an impact on marriage rates by making young males more attractive as potential marriage partners.
Reducing Nonmarital Births. Nonmarital births increase the nation’s poverty rate, increase income inequality, and have deleterious effects on children’s development. Parents, children, and the nation as a whole would benefit from a reduction in the nation’s nonmarital birth rate. The good news is that we know something about how to achieve this goal. The nation is making good progress against teen pregnancy rates which have declined in all but two years since 1991. Despite this progress, the U.S. still has higher teen pregnancy rates than any European nation. The Obama Department of Health and Human Services has been pursuing an evidence-based teen pregnancy reduction initiative, called Teen Pregnancy Prevention, since the first year of the administration. Local projects around the nation, many of which emphasize both abstinence and birth control, are now using model programs that have been determined by rigorous research to have impacts on teen pregnancy. In addition, every project is being carefully evaluated with an eye toward improving performance. Congress should be certain to continue funding of these programs. They have an excellent chance of further reducing teen pregnancy rates and abortion rates. They would also save money in the short run because most of these births are paid for with Medicaid funds and the children are eligible for Medicaid benefits.
Moving from a specific focus on teen pregnancy prevention to the wider problem of nonmarital births involving partners of any age, there are many fewer model programs and much less research involving women over age 19 than there is on teens. Even so, there is reasonable evidence on ways to reduce nonmarital births among young adults. Between 1993 and 2006, 25 states implemented a waiver option under the Medicaid program to provide family planning coverage to women who were not eligible for the regular Medicaid program. Research showed that this policy, by making more low-income mothers eligible for free or reduced price birth control, was associated with a reduction in births that may have been as high as 9 percent. More recently, a study in the St. Louis area showed that free access to long-acting reversible birth control (LARC) substantially reduced both pregnancy and abortion rates.
Increasing the access of poor and low-income women to effective, long-lasting methods of birth control would reduce nonmarital births and the poverty rates as well. The federal government should adopt policies, beginning with carefully planned role out in selected states, that use media campaigns and work with state and local organizations to bring attention to the availability of free birth control, especially long-acting methods.
The three strategies above have good evidence that they will reduce poverty and perhaps increase opportunity. But they are by no means the only policies that have at least some evidence that they would effectively fight poverty.
Young Males. One of the nation’s most pressing problems is that young males, especially minority males, are highly likely to be unemployed or out of the labor force. Data from the Bureau of Labor Statistics show that the employment-to-population ratio for black males between the ages of 20 to 24 has hovered around 50 percent since at least 1980 and has now dipped to just above 40 percent. This is the lowest EPR of any demographic group. To make matters worse, young black males are very likely to serve time in prison. Nearly 60 percent of black high school dropouts born between 1965 and 1969 had been in prison by the time they reached their early thirties. Having a prison record makes it more difficult to find work when they leave prison and disrupts their relationships with their kin and friends, including in many cases with their girlfriends and children.
Three public policies could improve the life situation of these young males, increase the chances that they could find work, and improve their opportunity to develop a normal relationship with their children and the mother of their children. The first two policies are to create more opportunities for disadvantaged young men to prepare for employment and to reduce their rates of incarceration. There are a number of programs that have been tested by gold standard evaluations that show positive impacts on the employment of young males. Foremost among them are the Career Academies program in high school and apprenticeship programs that give young people a skill and a certificate that can greatly increase their employment rates, often at good jobs. The Career Academies program even led to higher marriage rates. Participating in a good job preparation program and finding work will help reduce incarceration rates and may increase marriage rates. States and the federal government should also change mandatory sentencing laws and thereby reduce the number of nonviolent offenders who serve long prison sentences. Many states, sometimes forced by budget shortages, are already beginning to change their mandatory sentencing laws, although little is known yet about the effects of these changes.
Beyond these two policies, a policy that has been receiving attention for at least a decade is to provide an earnings supplement, similar to the EITC, to single adults. Even if job training and employment programs are successful, there will still be many young men who can only find jobs that pay modest wages. If a government program supplemented these low wages along the lines recommended by Sawhill and Karpilow and reviewed above, more young men might be drawn into the job market because they could earn a reasonable income when their earnings and the wage supplement are combined. An EITC policy of this type is now being tested in New York City. Congress and the President should give the Department of Health and Human Services the authority to plan and conduct demonstrations of this type in states or large cities that are willing to bear up to a quarter of the costs.
Preschool Programs. Washington is now in the midst of a heated debate about whether preschool programs actually help poor children prepare for school, touched off by President Obama’s proposal to greatly expand federal spending on preschool. Those who doubt that preschool produces lasting effects point to the national evaluation of Head Start which found impacts at the end of the Head Start year on several intellectual and socioemotional outcomes. However, the impacts were gone by the end of the first year of schooling. Concerns that the effects of good preschool programs might fade rapidly were increased by the results from a recent high quality study in Tennessee. Results from the Tennessee Voluntary Pre-K Program were similar to those for the national Head Start evaluation with fairly substantial impacts on both intellectual and socioem