My name is Ron Haskins and I’m pleased and privileged to have the opportunity to testify before the Finance Committee about poverty. Few topics have enjoyed as much attention from federal policymakers over the past half century as poverty and what can be done to reduce it. After a brief review of our success, such as it is, in reducing poverty, I examine the major causes of poverty, trends in spending to help poor and low-income Americans, and strategies Congress has adopted to fight poverty. Poverty has shown great if unfortunate staying power, but we have learned useful lessons about how to fight it.
Figure 1 shows poverty trends since the 1960s for the elderly, children, and all people. After some initial progress in the 1960s, and continuing progress for the elderly, the nation has made surprisingly little progress against poverty. The nation’s inability to reduce children’s poverty is especially troublesome. A review of the leading causes of poverty shows why trends in the economy, demography, and education make progress against poverty so difficult to achieve.
Causes of Poverty
In the United States, with the important exception of those on Social Security, the only way for most adults and families to avoid poverty is to work. Yet between 1980 and 2009, work rates for men declined from 74.2 percent to 67.6 percent, a fall of around 9 percent. The trend for young black men (ages 20-24) is even worse. Starting from the very low base of 60.9 percent, their ratio declined to the startling level of 46.9 percent, a decline of nearly 23 percent. Work among young black males is a national crisis. 
The work rate of women stands in sharp contrast to that of men. In 2007 before the Great Recession set in, 58.1 percent of women were working, a 25 percent increase since 1980. These figures reflect the post-World War II trend of the relentlessly increasing participation by women – including mothers of young children – in the nation’s economy. Equally impressive is the 20 percent rise in work by lone mothers over the same period, a trend that bears directly on child poverty rates because children in female-headed families are four or five times (depending on the year) more likely to be in poverty than children in married-couple families.  Even more important for the nation’s poverty rate, work by never-married mothers rose more sharply than that of any other group during the 1990s. These mothers and their children have always been the group most likely to be in poverty, including long-term poverty, in large part because their work rates have been so low.  In 1983, for example, only about only 35 percent of never-married mothers worked. After the welfare reform legislation of 1996, their work rate exploded, increasing from 46.5 percent in 1995 to 66.0 percent in 1999, an increase of more than 40 percent in just four years. Equally surprising, after a lengthy period of employment stagnation and decline associated with the mild recession of 2001 and the deep recession of 2007-2009, in 2010 their work rate was still more than 25 percent higher than it had been before welfare reform in the mid-1990s.
The fact that in 2007 – before the Great Recession – the work rates of males and females were 72 percent and 58 percent respectively, combined with the fact that the poverty rate for individuals in families in which no one works is nearly eight times as high as the poverty rate for individuals in families with at least one full-time, year-round worker, shows that there is plenty of room for improvement.  This uneven record of maintaining high levels of work is a leading cause of poverty in America. Without high work levels, it will be difficult to mount an effective fight against poverty.
Wage rates are a second work-associated factor that has a major impact on poverty. Based on data from the U.S. Census Bureau, trends in wages since 1979 can be succinctly summarized. Wages at the 10th percentile fell and then recovered and ended the nearly three decade period almost exactly where they were in 1979. The general trend of wages at the 50th percentile was a slow increase amounting to about a 10 percent rise over the entire period. At the top, by contrast, wages did very well, increasing 32 percent over the period at the 90th percentile. If we were to plot wages higher up in the distribution, they would rise even higher. 
In 2007, wages at the 10th percentile were about $8 per hour, more or less where they were in 1979 if inflation is taken into account. Working at this wage for 35 hours a week year round, a person would earn $14,560, $2,145 under the poverty level for a family of three. It is an amazing mathematical fact that 10 percent of all workers will always be at the 10th percentile of earnings or below. Thus, if wages do not improve at the bottom, all single parents with two or more children at or below the 10th percentile – and even many above the 10th percentile – will always be in poverty if earnings are their only income.
In 2009, the poverty rate for children in married-couple families was 11.0 percent. By contrast, the poverty rate for children in female-headed families was 44.3 percent.  The difference between these two poverty rates is a specter haunting American social policy because the percentage of American children who live in female-headed families has been increasing relentlessly for over five decades. In 1950, 6.3 percent of families with children were headed by a single mother. By 2010, 23.9 percent of families with children had single-mother heads.  That a higher and higher fraction of children live in the family type in which they are about four times as likely to be poor exerts strong upward pressure on the poverty rate. One way to think of the shift to female-headed families is that even if government policy were successful in moving people out of poverty, the large changes in family composition serve to offset at least part of the progress that otherwise would be made. In fact, a Brookings analysis shows that if we had the marriage rate we had in 1970, the poverty rate would fall by more than 25 percent. 
There now appears to be universal agreement that the combination of technological advances and globalization have resulted in education being a major factor in determining the employment and earnings of many American workers.  Census Bureau data on the relationship between education and family income since the 1960s show that families headed by adults with more education make more money. Some of the differences are huge. In 2009, the difference in median family income between families headed by an individual who dropped out of high school and families headed by an individual with a bachelor’s degree or higher was about $68,600 ($31,100 compared with $99,700).  Even more pertinent for examining the causes of poverty, family income for those with less than a college degree has been stagnant or declining for three decades. Without a college degree, 45 percent of the children from families in the bottom fifth of income will themselves be mired in the bottom fifth as adults. By contrast, with a college degree, adult children cut their odds of staying in the bottom fifth all the way down to 16 percent from 45 percent. The odds of making it to the top quintiles indicate similar abrupt changes if youngsters from poor families manage to achieve a college degree. 
Despite the great advantages of having a college degree, James Heckman has demonstrated that the high school graduation rate reached its highest level at about 80 percent in the late 1960s and has since decreased by 4 to 5 percentage points. A high school degree is usually required for college admission. Moreover, a significant gap remains between the graduation rate of white students (above 80 percent) and black and Hispanic students (both about 65 percent).  Ethnic gaps such as these are a continuing plague on the nation’s social policy.
The four-year college enrollment and graduation rates of students from families with varying levels of income renders the education picture discouraging. Youngsters from higher-income families are more likely both to enroll in and graduate from college than youngsters from poorer families. For example, 79 percent of children whose parents were in the top income fifth enrolled in college and 53 percent earned a four-year degree. But only 34 percent of children whose parents were in the bottom income quintile enrolled in college and only 11 percent received a four-year degree. If education is one of the routes out of poverty, the American educational system seems to be perpetuating poverty and income distinctions as much as it facilitates movement up the income scale.
The effectiveness of the nation’s K-12 education system is cast into serious doubt by comparing the performance of U.S. students with students from other OECD nations. In the most recent version of the Program for International Student Assessment (PISA), the U.S. was tied with two other countries for 27th in math, was 17th in science, and tied for 12th in reading.  A recent volume by Claudia Golden and Lawrence Katz of Harvard presents a strong case that past U.S. achievements in international competitiveness were due in large part to the superiority of the nation’s system of universal education and excellent colleges and universities.  The U.S. now seems to be mired in a situation in which the nation’s young people are at a level of educational achievement that is inferior to that of young people from many other nations. Thus, not only will the modest educational achievement of many Americans continue to make progress against poverty difficult, but American competitiveness in the global economy seems threatened.
Until the recent recession, America had been experiencing one of the greatest waves of immigration in its history. For the past two decades, an average of about one million immigrants has obtained legal permanent resident status in the U.S. each year.  In addition, according to the Pew Hispanic Center, in the seven years before the Great Recession, the population of undocumented immigrants grew by an average of a little over 500,000 per year.  In a nation that prides itself on being built by immigrants, these large numbers alone are not particularly daunting. However, as George Borjas of Harvard shows, about 20 percent of immigrants have less than a 9th grade education as compared with a little less than 3 percent of non-immigrants.  Consistent with the relatively large number of immigrants who lack even minimally adequate education, Borjas also finds a long-term trend toward lower wages among immigrants. In 1940, the age-adjusted average wage of first-generation male immigrants was 5.8 percent above the average wage of non-immigrant males. This figure fell to 1.4 percent above the average wage of non-immigrant males in 1970 and then dropped dramatically to 20 percent below the non-immigrant male wage in 2000. 
It comes as little surprise, then, that the poverty rate among immigrants is higher than the poverty rate among native-born Americans. In 2009, the immigrant poverty rate was 19.0 percent as compared with 13.7 percent for native-born Americans. Given that the overall poverty rate for the nation was 14.3 percent, the poverty rate would be lower by about 0.6 percentage points (or around 1.9 million people) if the immigrant poverty rate were the same as the poverty rate for native-born citizens. 
Reflecting on these five major causes of poverty leads one to understand why it has been so hard for the U.S. to make much progress against poverty despite the proliferation of social programs and the substantial increases in spending (see below) since President Johnson first declared war on poverty in the mid-1960s. Declining work rates, stagnant wages, the rise of female-headed families, inferior education, and the arrival of millions of immigrants with poor education and low skills are little engines pushing up the poverty rate. Conditions in the U.S. virtually ensure high poverty rates because the underlying factors that cause poverty have remained very strong. Even so, the nation has done a lot to reduce poverty and has even achieved a few victories as we shall now see.
Trends in Spending on Means-Tested Programs
Many analysts think that the nation could greatly reduce poverty by spending more money on programs for the poor. Figure 2, based on a Brookings analysis of federal budget data published by the Office of Management and Budget, shows that federal spending since 1962 in the ten biggest means-tested federal programs has increased dramatically. Since 1980, by which time all but two of the ten programs that spent the most money in 2011 were in place, spending has increased by about $500 billion, from $126 billion to $626 billion after adjusting for inflation. Similarly, spending per person in poverty between 1980 and 2011 increased from about $4,300 to $13,000 or more than $3 spent per person in poverty in 2011 for every dollar spent in 1980. More recently, means-tested spending increased from about $477 billion to $626 billion in the first three years of the Obama administration, an increase of about 31 percent.
The Congressional Research Service has conducted a study that divides means-tested spending into eight categories (health, cash, nutrition, employment and training, etc.). Health is by far the biggest category of means-tested spending at $319 billion in 2009, around 2.5 times as much as cash programs, the second biggest category. Employment and training at $9 billion is the smallest of the eight categories.
These spending data are for only the ten largest means-tested programs. The Congressional Research Service estimates that in 2009, spending on these ten programs represented about 75 percent of total federal means-tested spending.  If that percentage remained roughly the same for 2011, total federal means-tested spending in that year was closer to $835 billion than the $626 billion spent on the ten biggest programs. Even $835 billion is an underestimate of total means-tested spending because state and local governments also spend money on many of these programs. The Congressional Research Service has estimated that state and local governments supplemented federal spending on means-tested programs by around 27 percent in 2004.  If we assume that the 27 percent has remained roughly constant, we can estimate that total federal, state, and local government spending on means-tested programs was probably more than $1 trillion in 2011.
On a per-person in poverty basis, that figure represents about $23,700 in spending by federal, state, and local governments. But this estimate should be considered in light of several caveats. The first is that not all of the spending on means-tested programs goes directly to individuals and families. Some of the money is spent on programs, such as the $14.5 billion spent on Title I of the No Child Left Behind Act and the $9 billion in spending on employment and training programs, that provide services rather than direct cash or in-kind benefits to individuals and families. Second, some of the money in programs that provide cash or in-kind benefits directly to households goes to individuals and families that are not below the poverty level. Children in families of up to 200 percent of the poverty level, for example, are eligible for Medicaid or the Child Health Insurance Program (CHIP) in almost every state.  Similarly, people in households with incomes up to 130 percent of poverty are eligible for SNAP benefits (Supplemental Nutrition Assistance Program, previously food stamps). In the case of the Earned Income Tax Credit (EITC), in 2010 a single mother with two children could receive benefits if the mother’s income was below $40,964, about 225 percent of the poverty level for this family.
Thus, means-tested spending has increased enormously no matter how it is measured. Although there have been some periods of comparatively rapid growth, such as during the recession of 2007 to 2009, Figure 2 shows that spending has grown almost every year for the last five decades. The increase in spending has been the most rapid in health programs, but cash, nutrition, and several other types of spending have also increased rapidly. Spending per person in poverty has also increased substantially, although not quite as rapidly as total spending. Unfortunately, as we have seen, the high level and remarkable increases in means-tested spending have not led to consistent declines in poverty, although we have learned a lot about what works and what doesn’t, a topic to which we now turn our attention.
Strategies to Reduce Poverty
Although the dramatic increase in federal spending has not led to an overall reduction in the nation’s poverty rate, at least two strategies have been successful in reducing poverty within specific demographic groups. Both should be considered major successes of the nation’s social policy and both could be extended. The first is to give money to people who are not expected to work and the second is to use welfare policy to strongly encourage work and then to subsidize earnings because so many of the poor have low skills and often cannot earn enough to escape poverty.
Before reviewing these and other strategies for reducing poverty, I want to emphasize the importance of individual initiative in reducing poverty and promoting economic success. My Brookings colleague Isabel Sawhill and I have spent years emphasizing the importance of individual responsibility in reducing poverty and increasing opportunity. One of our arguments, based in part on a Brookings analysis of Census Bureau data, is that young people can virtually assure that they and their families will avoid poverty if they follow three elementary rules for success – complete at least a high school education, work full time, and wait until age 21 and get married before having a baby. Based on an analysis of Census data, people who followed all three of these rules had only a 2 percent chance of being in poverty and a 72 percent chance of joining the middle class (defined as above $55,000 in 2010). These numbers were almost precisely reversed for people who violated all three rules, elevating their chance of being poor to 77 percent and reducing their chance of making the middle class to 4 percent.  Individual effort and good decisions about the big events in life are more important than government programs. Call it blaming the victim if you like, but decisions made by individuals are paramount in the fight to reduce poverty and increase opportunity in America. The nation’s struggle to expand opportunity will continue to be an uphill battle if young people do not learn to make better decisions about their future.
Strategy 1: Give Them Money
The most straightforward way to help people escape poverty, primarily when they belong to a group, such as the elderly or disabled, who are not expected to work is to give them cash and in-kind benefits that will bring their income above the poverty threshold. The Social Security program, for example, is designed specifically to help the elderly avoid destitution. Although in its early decades it provided benefits that were quite modest by today’s standards, in the early 1970s Congress enacted laws that increased the Social Security cash benefit.  These reforms had an immediate impact in driving down the poverty rate among the elderly. Indeed, research shows that virtually the entire decline in poverty among the elderly is accounted for by the rise in Social Security benefits.  In addition, nearly all the elderly are covered by the Medicare health program and the poor and low-income elderly are qualified for many other programs including housing and nutrition. If the value of taxes, in-kind benefits (except health insurance), and the imputed return on home equity are taken into account, poverty among the elderly drops even further, from the official rate of 8.9 percent to as low as 5.3 percent. 
The strategy of reducing poverty by providing government benefits touches on one of the major fault lines in American politics. As polls consistently show, Americans think able-bodied, non-elderly people should earn their own way. Americans simply don’t like welfare, even when someone calls it by a different name (e.g., “food stamps” or more recently “SNAP”).  In 1995 and 1996 during the welfare reform debate, when Republicans were proposing that the cash welfare benefit of lone mothers who didn’t meet work requirements be reduced or terminated and that the benefit be time limited, polls showed that the American public supported these tough policies.  Giving money and other goods to the poor might work with the elderly or the disabled, because Americans do not expect either group to work. But it seems likely that simply giving welfare to the able-bodied poor, even if they are single mothers, will never be an effective strategy for reducing poverty in the U.S. because it will be difficult to enact legislation authorizing the necessary spending. 
Strategy 2: Increase Work Rates and Work Supports
At the other end of the continuum from policies that give money and other benefits to the poor are policies that encourage work. Well over 75 percent of families with children that lack a full-time, year-round worker are in poverty.  An individual or family in the U.S. whose only source of income is welfare benefits cannot escape poverty.  It follows that an effective anti-poverty strategy would be to increase work rates. This was precisely the goal of the welfare reform legislation enacted in 1996, which replaced the Aid to Families with Dependent Children program with the Temporary Assistance for Needy Families (TANF) program. Passed on a strong bipartisan basis, and signed by Democratic President Bill Clinton, the new law required individuals to meet work requirements in order to qualify for welfare benefits. Mothers on welfare had to participate in state-designed welfare-to-work programs that provided training, job search assistance, or actual work experience. If the mothers did not participate, states were required by federal law to impose financial sanctions on them in the form of reduced or even terminated cash welfare (although they remained eligible for non-cash benefits). In addition to work requirements and sanctions, the reform law imposed a 5-year time limit on benefit receipt for most mothers who accepted welfare.
After passage of the 1996 reforms, poor mothers entered the workforce in unprecedented numbers. Between 1995 and 1999, for example, there was an increase of more than 40 percent in the number of never-married mothers, the poorest of the poor, who found employment.  In large part due to this increased employment of never-married mothers, poverty among all single mothers and their children fell by 30 percent, from a 1991 peak of 47.1 percent to 33.0 percent in 2000, its lowest level ever. Similarly, poverty among black children, who live disproportionately in female-headed families, reached its lowest level ever in 2001. This example demonstrates what is possible if government policy encourages and even pressures adults to go to work and then subsidizes the incomes of those who earn low wages. The combination of work requirements and earned public benefits has the appearance of an approach to reducing poverty that has strong bipartisan overtones.
It is notable that even during and after the recessions of 2001 and 2007 to 2009, work rates among never-married mothers did not return to their pre-welfare reform level. Although their work rates fell from the 1999 peak (and highest ever) of 66.0 percent to 58.7 percent in 2010, the 2010 level is still about 25 percent higher than the pre-welfare reform level of 46.5 percent in 1995.  On its face, the strategy of emphasizing work, even by mothers whose education and experience usually limit them to low-wage jobs, results in more mothers working and lower poverty levels than the low employment levels that prevailed during the pre-welfare reform years in which millions of mothers spent many years on the welfare rolls.  Further, giving benefits to mothers who work full time in order to help them and their children escape poverty is more politically popular – and therefore sustainable – than giving welfare to mothers who don’t work.
The predominance of low-wage work for poorly educated workers is the reason encouraging work is only half the federal strategy for increasing work rates to attack poverty. Millions of Americans have low-wage and part-time jobs that do not provide them with enough money to support a family at or above the poverty level. If a lone mother worked year-round, full-time at the minimum wage ($7.25 per hour), with no vacations and no time off for illness or to care for sick children, she would earn $15,080, about $2,500 below the poverty level for a mother and two children. Realizing the problem of low wages, and hoping to increase work incentives at the bottom of the wage scale, Congress and three presidents, beginning roughly in the mid-1980s, passed a series of laws that created, modified, or expanded programs that provide cash and in-kind benefits to poor and low-income working families.  These programs, often referred to as the nation’s work support system, are structured in such a way that they provide benefits to families that have low earnings. Some of the programs, like the Child Tax Credit and the Earned Income Tax Credit (EITC), provide their benefits only to families with earnings. All of the programs allow at least some of their benefits to flow to families that avoid or leave welfare for work.
Federal policymakers developed the work support system over a period of roughly two decades.  If the laws on child care, medical assistance, the child tax credit, and the Earned Income Tax Credit (EITC) had not changed after 1984, a study by the Congressional Budget Office shows that in 1999 working families would have received a mere $5.6 billion in benefits. But because Congress expanded, modified, or created all these work support programs after 1984, in 1999 working families qualified for $51.7 billion in benefits. A more recent estimate of benefits from these programs for working families by Sheila Zedlewski of the Urban Institute shows that between 1996 and 2002, work support benefits grew by 27 percent in real dollars.  Since 2002, the EITC has been expanded twice, Medicaid coverage has expanded, food stamps benefits have been reformed to make it easier for working families to get them, and day care funding has been expanded. Federal policy does a lot to increase income and reduce poverty among low-income working families.
The increase in work by single mothers following enactment of welfare reform in 1996 activated the work support system for millions of these families, both those leaving welfare and those who never went on welfare. Work and work support, in short, functioned together to reduce poverty and welfare dependency. Figure 3 provides a clear view of the impact of government programs on children’s poverty rates in families headed by never-married mothers in 1989, before welfare reform, and in 2006, after welfare reform. The raw poverty rate (before any government transfers or taxes) in 2006 was nearly 20 percent lower than in 1989 (39.6 percent vs. 48.3 percent). Undoubtedly, this decline in raw poverty before any government assistance was due to increased work and earnings by these never-married mothers. As shown by the second bar graph in each set, social insurance and non-cash benefits reduced poverty by between 20 and 25 percent in both years. But in 2006, cash benefits provided through the EITC and Child Tax Credit (CTC) sent the poverty rate down another 13 percent while in 1989 the mothers’ work rate was not high enough to attract substantial support from the EITC and CTC. In short, both more work by the mothers and more work support benefits from government contributed to the final poverty rate being almost 40 percent lower in 2006 (26.1 percent) than in 1989 (39.1 percent). 
Although controversial, a reasonable implication of these results is that federal policy should encourage work. One way to achieve this end would be to strengthen work requirements in both the SNAP program and the means-tested housing programs. This recommendation is controversial because the 1996 reforms showed that some mothers either do not find jobs or have difficulty holding down a job for an extended period. As often happens in these situations, an argument has broken out among researchers and pundits about whether the finding that many mothers do not retain employment is a major problem. There are two primary facts that are accepted by both sides.
First, the percentage of single mothers who are “disconnected” from both cash welfare and employment has more than doubled since welfare reform was enacted in 1996. Second, the percentage of poor single mothers and children who receive cash welfare from the TANF program is the lowest ever; in 1979, about 82 families were receiving cash from the old Aid to Families with Dependent Children program for every 100 families with children in poverty; by 2010 that number had fallen to 27 families receiving TANF for every 100 families in poverty.  One side argues that these data show that these mothers and their children are on the edge of destitution and that policy should be changed so that they can qualify for cash welfare benefits.  The other side argues that these mothers are choosing not to work, that they get other means-tested benefits (especially food stamps), and that they usually live with other adults who have income.  At the very least, the large number of mothers with very low income who are not receiving cash welfare calls for caution. Caution in this case means that strengthening work requirements in SNAP and housing programs should be done on a limited basis and studied carefully during implementation. Caution may also call for Congressional hearings and studies by Congressional agencies of how states are implementing the work requirements established by the 1996 welfare reform law and the Deficit Reduction Act.
If giving money to the elderly and incentivizing work combined with supplementing earnings with work support benefits have proven to be effective in reducing poverty, at least four other strategies hold promise for reducing poverty.
Immigration. History shows that immigrants are often a hard working and creative group of people who move from their home country specifically because they want to get ahead.  In the U.S., however, legislation enacted in 1965 gave preference in admitting immigrants to relatives of those who are already in the U.S., regardless of their education or skill levels. It appears from recent debates that there is now widespread recognition that it would be wise to shift immigration policy to reduce the importance of family relationships in favor of increasing the emphasis on skills and on giving employers more flexibility in allowing valuable employees to stay in the U.S. The hope for legislation to reflect this recognition is being held up, however, primarily by continuing disagreements concerning what to do about undocumented immigrants. If we shifted immigration poverty to place a greater emphasis on education and skills, we would see a shift in the percentage of immigrants who earn higher wages and thereby avoid poverty. In addition, immigrants with higher education and skills would contribute more to the American economy which would in turn contribute to economic growth and increase both employment and tax revenues. There does not appear to be any downside to immigration reforms of this sort.
Reducing Nonmarital Births. One of the engines driving poverty in the U.S. is the fragmentation of families. Around a quarter of children are living in female-headed families at any given moment and about half experience at least some time during their childhood in a female-headed family. When children live in female-headed families, they are at least four times as likely to be poor as when they live in a married-couple family.  But poverty is not the only risk faced by these children. Since Sara McLanahan and Gary Sandefur published Growing up with a Single Parent in 1994,  social science research has repeatedly shown that children reared in female-headed families are more likely to fail in school, more likely to be arrested, more likely to get pregnant as teens, more likely to have mental health problems and to commit suicide, more likely to get a divorce when they grow up, and more likely to experience other negative outcomes.  In addition, as Kathy Edin has shown, these parents tend to separate within a few years, whereupon both the mother and father usually go on to form new relationships. Thus, their children experience a series of changes in household composition as their mothers form new cohabiting relationships.  The mother might even have a baby with one or more of these new men, creating a household with complex and often difficult relationships among the adults and usually making it hard for the children to establish a close relationship with their fathers.  The point is that life in female-headed families imposes both a high likelihood of poverty and of household instability that can produce negative impacts on child development.  If the share of children born into and living in married-couple families could be increased, poverty and childhood education, health, and mental health problems would decline, increasing the human capital of the nation’s children and having a long-term impact on the nation’s poverty rate. 
Several programs have proven successful in reducing teen pregnancy.  In part because of the prevalence of these programs, the U.S. teen birthrate has declined in all but three years since 1991.  It is difficult, however, to be too optimistic about the declining teen birthrate because as the teen birthrate has declined, the nonmarital birthrate for young women in the twenties and early thirties has increased more than enough to offset the decline in the teen rate. 
Even so, investments in programs aimed at reducing nonmarital births have been shown not only to actually reduce such births among women in their 20s and 30s, but to save government money.  The programs are a mass media campaign that encourages men to use condoms, a program for teens that both encourages abstinence and instructs on the proper use of contraceptives, and expansion of family planning services provided by Medicaid, mostly birth control for low-income females. Similarly, the Obama administration has initiated a number of new evidence-based initiatives that could reduce the number of nonmarital births even more. Additional investments in these programs would reduce the number of nonmarital births and in doing so reduce the nation’s poverty rate. but with 72 percent of black babies, 53 percent of Hispanic babies, and over 40 percent of all babies born outside marriage, there is a long way to go.  The nonmarital birth machine that expands poverty and produces children with less human capital than their peers being reared in married-couple families is disrupting and will continue to disrupt the nation’s drive to curb poverty.
Increasing Human Capital. The most basic reason rich nations need welfare programs is because a significant fraction of their populations do not earn enough money to support themselves and their families. Since the 1960s, a major goal of anti-poverty policy has been helping the poor acquire the education and skills needed to achieve earnings that will support a decent standard of living. In other words, the goal has been to create more opportunity for economic and social achievement by helping people improve their human capital, which in turn would increase their earnings and reduce their dependency on welfare. There are four major types of programs that the U.S. uses to develop human capital: preschool programs, K-12 education, post-secondary education, and employment and training programs. All have the potential to increase human capital, thereby making the economy more efficient and competitive while simultaneously providing participants with the education and skills needed to find a productive and rewarding place in the American economy. Here I focus attention only on preschool programs.
No human capital program is so widely believed to be effective as preschool education for children from poor and low-income families.  The Perry Preschool Program in Michigan, the Abecedarian program in North Carolina, and the Child-Parent Centers in Chicago have all produced both immediate and long-term impacts on the development of poor children.  Similarly, a number of state pre-K programs have shown that they boost the development of preschool children from poor families and increase their school readiness.  Reviewers regularly cite these and similar programs to argue that there is rigorous evidence that preschool programs can have broad and long-lasting effects that boost human capital. 
Between the federal government and the states, the U.S. now spends around $31 billion on preschool programs (including child care and home visiting), mostly for children from poor and low-income families.  At least some of this money is being spent on programs, particularly the state pre-K programs, that are of high enough quality to produce some of the impacts achieved by Perry, Abecedarian, and the Chicago Parent-Child Centers. Unfortunately, a substantial portion of the money is being spent on preschool programs that lack an educational focus or on Head Start (about $7 billion in 2010), which has been shown by a recent random-assignment evaluation to have almost no impacts by the end of first grade. Thus, until the U.S. figures out how to achieve bigger impacts in the programs supported by our current $31 billion in expenditures on preschool programs, it cannot be expected that poor children are going to receive enough of a boost from preschool programs to make a long-term difference in their school performance, employment, or earnings.
Two courses of action now seem desirable. First, the Obama administration is subjecting Head Start to the most important and far-reaching reforms in its history. Each Head Start program is being carefully evaluated; those that fail must re-compete against other willing program operators to retain their funding. This reform should be followed carefully by Congress over the next several years through hearings and assessments from the Government Accountability Office or the Congressional Budget Office. If the Obama reforms improve the average impact of Head Start on school readiness, the school performance of poor children in grades K-12 could improve, increasing the chances that they will succeed in the American economy. Second, Congress should work with states to maintain funding of both state pre-K programs and state child care programs and to help states improve the average quality of child care, much of which is woefully inadequate.
 In technical papers, work rates are called “employment to population ratios.” They are calculated by putting the number in the demographic group of interest in the denominator and the number of members of this group who have jobs in the numerator.
 Harry J. Holzer, Paul Offner, and Elaine Sorensen, “Declining employment among young black less-educated men: The role of incarceration and child support,” Journal of Policy Analysis and Management 24, no. 2 (2005): pp. 329-350.
 For example, in 2009, 11.0 percent of children in married-couple were in poverty compared to 44.3 percent of children in female-headed families. Similarly, in 2008 the poverty rates for children in married-couple families and female-headed families were 9.9 percent and 43.5 percent, respectively. And in 2007, these respective child poverty rates were 8.5 percent and 42.9 percent. U.S. Census Bureau, Detailed Poverty Tables, “POV03: People in Families with Related Children Under 18 by Family Structure, Age, and Sex, Iterated by Income-to-Poverty Ratio and Race: 2007” (http://pubdb3.census.gov/macro/032008/pov/new03_100_01.htm), “POV03: People in Families with Related Children Under 18 by Family Structure, Age, and Sex, Iterated by Income-to-Poverty Ratio and Race: 2008” (http://www.census.gov/hhes/www/cpstables/032009/pov/new03_100_01.htm), and “POV03: People in Families with Related Children Under 18 by Family Structure, Age, and Sex, Iterated by Income-to-Poverty Ratio and Race: 2009” (http://www.census.gov/hhes/www/cpstables/032010/pov/new03_100_01.htm).
 As Wu and Stojnic (2007, 7-8) note, during the period between 1975 and 2000, “individuals in never-married female-headed households were roughly 4 to 5 times more likely to be in poverty relative to a randomly chosen individual in the U.S. population; similarly, individuals in never-married female-headed households were, on average, about twice as likely to live below the poverty line as individuals in ever-married female-headed households…” Lawrence L. Wu and Miodrag Stojnic, “Poverty among the Poorest-Poor in the United States: Trends for Never-Married Women and Their Children,” Paper Presented at the Inaugural Conference, Center for Research on Inequalities and the Life Course, Yale University (May 2007). Available at http://www.yale.edu/ciqle/INAUGURAL%20PAPERS/pp070502.pdf.
 U.S. Census Bureau, Detailed Poverty Tables, “POV10: People in Families by Number of Working Family Members and Family Structure: 2009”
 Based on Social Security income records, the top 1 percent of workers earned 12.3 percent of all income in 2004, up from 5.9 percent in 1960. See Wojciech Kopczuk, Emmanuel Saez, and Jae Song, “Earnings Inequality and Mobility in the Untied States: Evidence from Social Security Data since 1937,” Quarterly Journal of Economics 125, no. 1, 2010, pp. 91-128.
 U.S. Census Bureau, Detailed Poverty Tables, “POV03: People in Families with Related Children Under 18 by Family Structure, Age, and Sex, Iterated by Income-to-Poverty Ratio and Race: 2009” (http://www.census.gov/hhes/www/cpstables/032010/pov/new03_100_01.htm).
 U.S. Census Bureau, Families and Living Arrangements, “Table FM-1: Families by Presence of Own Children Under 18: 1950 to Present” (http://www.census.gov/population/www/socdemo/hh-fam.html).
 Ron Haskins and Isabel Sawhill, Creating an Opportunity Society (Washington, DC: Brookings, 2009).
 Claudia Golden and Lawrence F. Katz, The Race between Education and Technology (Cambridge, MA: Harvard, 2008).
 U.S. Census Bureau, Historical Income Tables, “Table F-18: Educational Attainment of Householder--Families with Householder 25 Years Old and Over by Median and Mean Income” and “Table F-19: Years of School Completed--Families with Householder 25 Years Old and Over by Median and Mean Income” (http://www.census.gov/hhes/www/income/data/historical/families/index.html).
 Ron Haskins, “Education and Economic Mobility,” in Julia B. Isaacs, Isabel, V. Sawhill, and Ron Haskins, editors, Getting Ahead or Losing Ground: Economic Mobility in America (Washington, DC: Brookings and Pew Charitable Trusts, 2009).
 James J. Heckman and Paul A. LaFontaine, “The American High School Graduation Rate: Trends and Levels,” National Bureau of Economic Research Working Paper No. 13670, December 2007 and James J. Heckman and Paul A. LaFontaine, “The Declining American High School Graduation Rate: Evidence, Sources, and Consequences,” NBER Reporter, Research Summary, 2008, No. 1.
 Organization of Economic Cooperation and Development, PISA 2009 Results: What Students Know and Can Do: Student Performance in Reading, Mathematics, and Science, vol. 1 (Paris: Author, 2010). Professor Daniel Koretz of Harvard recommends caution in simply comparing the average PISA scores of American students with the average scores of students from other nations in part because the scores are so variable from test occasion to test occasion. Nonetheless, the performance of U.S. students shows clearly that they are not near the top on any of the subtests. Rather, American students are closer to the average scores for OECD nations than to the top score; see Daniel Koretz, “How Do American Students Measure Up? Making Sense of International Comparisons,” The Future of Children 19, no. 1 (Spring 2009): 37-51.
 Golden and Katz, 2008.
 Department of Homeland Security, Yearbook of Immigration Statistics: 2010, Table 6 (http://www.dhs.gov/files/statistics/publications/LPR10.shtm).
 Pew Hispanic Center, “Unauthorized Immigrant Population: National and State Trends, 2010. Washington, DC: Author, February 2011.
 U.S. Census Bureau, “Foreign-Born Population of the United States Current Population Survey - March 2009: Detailed Tables,” Table 1.5 (http://www.census.gov/population/www/socdemo/foreign/cps2009.html).
 George J. Borjas, “Making It in America: Social Mobility in the Immigrant Population,” The Future of Children 16, no. 2 (2006): 55–71. More recently, Borjas has noted that there was an unexplained rise in entry wages during the 1990s. See George J. Borjas and Rachel M. Freidberg, “Recent Trends in the Earnings of New Immigrants to the United States” (Working Paper 15406), Cambridge: National Bureau of Economic Research, October 2009.
 U.S. Census Bureau, Income, Poverty and Health Insurance Coverage in the United States: 2009 (Washington, D.C.: U.S. Department of Commerce, September 2010), Table 4.
 The number of people in poverty in 2011 will not be published by the Census Bureau until next fall. Based on a model designed by Richard Bavier, a former senior official at OMB, that has successfully predicted the poverty level for the past several years, poverty increased by .6 percentage points in 2011, bringing the poverty rate to 15.7 percent in that year. According to the American Community Survey, the population of the United States in 2011 was 306 million. Thus, around 48,042,000 people were poor in 2011. This is the figure I used to compute means-tested spending per person in poverty for 2011.
 Karen Spar, “Federal Benefits and Services for People with Low Income: Programs, Policy, and Spending, FY2008-FY2009” (R41625), Washington, DC: Congressional Research Service, January 2011, p. 12.
Domestic Social Policy Division, “Cash and noncash Benefits for Persons with Limited Income: Eligibility Rules, Recipient and Expenditure Data, FY2002-FY2004” (RL33340), Washington, DC: Congressional Research Service, March 2006.
 According to the Kaiser Family Foundation, only North Dakota, Oklahoma, Arkansas, and Idaho set the child eligibility level for Medicaid and CHIP below 200 percent of poverty; half the states are at 250 percent or above. See http://www.kff.org/medicaid/upload/7993-02.pdf
 The Brookings analysis is suggestive rather than definitive because of something called selection effects. People who at least complete high school, get a job, and get married and wait until age 21 to have babies may differ in other ways that could account for their ability to avoid poverty and join the middle class. There may be four rules, or five rules, or a dozen rules, most of which those who follow the three rules we specify also follow. If this were true, then attributing success to our three rules would be misleading.
 Social Security Administration, Social Security: A Brief History (Washington DC: Author, 2005); available at http://www.ssa.gov/history/pdf/2005pamphlet.pdf.
 Jonathan Gruber and Gary Engelhardt, “Social Security and the Evolution of Elderly Poverty,” in Public Policy and the Income Distribution, edited by Alan Auerbach, David Card, and John Quigley (New York: Russell Sage Foundation, 2006), pp. 259–87.
 U.S. Census Bureau, Definitions 1 and 15 from, “Table 2: Percent of Persons in Poverty, by Definition of Income and Selected Characteristics: 2009,” from the Current Population Survey, Annual Social and Economic Supplement; available at http://www.census.gov/hhes/www/cpstables/032010/rdcall/2_001.htm.
 Martin Gilens, Why American’s Hate Welfare (Chicago: University of Chicago, 2000).
 R. Kent Weaver, Ending Welfare as We Know It (Washington, DC: Brookings, 2000), Chapter 7; Ron Haskins, Work over Welfare: The Inside Story of the 1996 Welfare Reform Law (Washington, DC: Brookings, 2006), p. 55.
 R. Kent Weaver, “The Politics of Low-Income Families in the United States,” in Making the Work-Based Safety Net Work Better: Forward-Looking Policies to Help Low-Income Families, edited by Carolyn J. Heinrich and John Karl Scholz (New York: Russell Sage, 2009).
 U.S. Census Bureau, Poverty Division CPS ASEC, “Detailed Tables, Table POV07: Families With Related Children Under 18 by Number of Working Family Members and Family Structure: 2009,” from Income, Poverty, and Health Insurance Coverage in the United States: 2009, (Washington, DC: Author, 2010); available at http://www.census.gov/hhes/www/poverty/data/index.html.
 For example, according to Liz Schott and Zachary Levinson, TANF Benefits Are Low and Have Not Kept Pace with Inflation, (Washington DC: Center on Budget and Policy Priorities, 2008); available at http://www.cbpp.org/pdf/11-24-08tanf.pdf; the combined benefits of TANF and Food Stamps (SNAP) remained under 75 percent of the federal poverty line in all but three states in 2008.
 The percentage increase is based on our analysis of data from the Annual Social and Economic Supplement to the Current Population Survey. The percentage of never-married mothers employed in 1995 was 47; by 1999, the percentage had increased to 66.
 Brookings tabulations of data from the Annual Social and Economic Supplement to the Current Population Survey.
 Mary Jo Bane and David Ellwood, “Slipping into and out of Spells of Poverty,” Journal of Human Resources 21, no. 1 (1986): 23.
 Carolyn J. Heinrich and John Karl Scholz, editors, Making the Work-Based Safety Net Work Better: For ward-Looking Policies to Help Low-Income Families (New York: Russell Sage, 2009).
 U.S. Congressional Budget Office, Policy Changes Affecting Mandatory Spending for Low-Income Families Not Receiving Welfare (Washington, DC: 1998).
 Sheila Zedlewski and others, “Is there a System Supporting Low-Income Working Families,” Low Income Work Families Paper 4, The Urban Institute (2006); available at http://www.urban.org/UploadedPDF/311282_lowincome_families.pdf.
 Christopher Bollinger et al., Welfare Reform and the Level and Composition of Income, in Welfare Reform and its Long-Term Consequences for America’s Poor, edited by James P. Ziliak (Forthcoming); Robert F. Schoeni & Rebecca M. Blank, What Welfare Reform Accomplished? Impacts on Welfare Participation, Employment, Income, Poverty, and Family Structure, NBER Working Paper 7627, 2000; Robert A Moffitt, The Effect of Pre-PRWORA Waivers on AFDC Caseloads and Female Earnings, Income, and Labor Force Behavior, in Economic Conditions and Welfare Reform, edited by Sheldon Danziger (W.E. Upjohn 1999).
 Danilo Trisi and LaDonna Pavetti, “TANF Weakening as a Safety Net for Poor Families,” Washington, DC: Center on Budget and Policy Priorities, March 2012; Pamela J. Loprest, “Disconnected Families and TANF,” Washington, DC: Urban Institute, OPRE Brief #2, November 2011.
 Rebecca M. Blank and Brian K. Kovak, “The Growing Problem of Disconnected Single Mothers,” in Making the Work-Based Safety Net Work Better, edited by Carolyn J. Heinrich and John Karl Scholz (New York: Russell Sage, 2009).
 Douglas J. Besharov, “Leaving Welfare Without Working: How Do Mothers Do it? And What Are the Implications?” College Park, MD: University of Maryland, Welfare Reform Academy, February 2002.
 George J. Borjas, Heaven’s Door: Immigration Policy and the American Economy (Princeton, NJ: Princeton, 1999); Jacob L. Vigdor, From Immigrants to Americans: The Rise and Fall of Fitting In (New York: Rowman & Littlefield, 2009).
 Census Bureau, Families, “Table C2: Household Relationship and Living Arrangements of Children Under 18 Years, by Age and Sex: 2010,” available at http://www.census.gov/population/www/socdemo/hh-fam/cps2010.html;” Katherine Magnuson and Lawrence M. Berger, “Family Structure States and Transitions: Associations with Children’s Well-Being During Middle Childhood,” Journal of Family and Marriage 71, no. 3 (2000): 575-591.
 Sara McLanahan and Gary Sandefur, Growing up with a Single Parent: What Hurts, What Helps (Cambridge: Harvard, 1994).
 Paul R. Amato, “The Impact of Family Formation Change on the Cognitive, Social, and Emotional Well-Being of the Next Generation,” The Future of Children 15, no. 2 (2005): 75-96; Mitch Pearlstein, Shortchanging Student Achievement: The Educational, Economic, and Social Costs of Family Fragmentation (Forthcoming); Haskins and Sawhill, Creating an Opportunity Society, Chapter 10.
 Reference Edin study of transitions in Fragile Families study
 Kathy Edin, personal communication, April 28, 2011; see https://crcw.princeton.edu%2fworkingpapers%2fWP11-10-FF.pdf.
 Sara McLanahan and others, “Strengthening Fragile Families,” Future of Children Policy Brief, Princeton-Brookings, Fall 2010.
 Adam Thomas and Isabel Sawhill, “For Richer or Poorer: Marriage as an Antipoverty Strategy,” Journal of Policy Analysis and Management 21, no. 4 (2002): 587-599; Ron Haskins and Isabel Sawhill, “Work and Marriage: The Way to End Poverty and Welfare,” Welfare Reform and Beyond Brief 28, The Brookings Institution (September 2003).
 Douglas Kirby, Emerging Answers 2007: Research Findings on Programsto Reduce Teen Pregnancy and Sexually Transmitted Diseases (Washington DC: National Campaign to Prevent Teen and Unplanned Pregnancy, 2007).
 The National Campaign to Prevent Teen and Unplanned Pregnancy, “National Birth Rates for Teens, aged 15-19,” (Accessed May 19, 2011); available at http://www.thenationalcampaign.org/national-data/NBR-teens-15-19.aspx.
 Stephanie J. Ventura, “Changing Patterns of Nonmarital Childbearing in the United States,” National Center for Health Statistics Data Brief, No. 18, May 2009.
 Adam Thomas, "Three Strategies to Prevent Unintended Pregnancy,” Journal of Policy Analysis and Management 31, No. 2 (2012).
 Brady Hamilton, Joyce Martin, and Stephanie Ventura, “Table 1: Total Births and birth, fertility, and total fertility rates and nonmarital births, by race and Hispanic origin of mother: United States, final 2008 and preliminary 2009,” from National Vital Statistics Reports: Births: Preliminary Data for 2009 59, no. 3 (2010); available at http://www.cdc.gov/nchs/products/nvsr.htm.
 James J. Heckman, “Catch ‘Em Young,” Wall Street Journal, January 10, 2006, p. A14.
 Craig Ramey, Frances Campbell, and Clancy Blair, “Enhancing the Life Course for High-Risk Children: Results from the Abecedarian Project,” in Social Programs that Work, edited by Jonathan Crane (New York: Russell Sage Foundation, 1998); Lawrence J. Schweinhart, and others, Lifetime Effects: The High/Scope Perry Preschool Study through Age 40 (Ypsilanti, MI: High/Scope Press, 2005); Arthur J. Reynolds, Success in Early Intervention: The Chicago Child-Parent Centers (Lincoln, NB: University of Nebraska Press, 2000).
 W. Steven Barnett and others, Effects of Five State Pre-Kindergarten Programs on Early Learning (Rutgers University, National Institute for Early Education Research, 2007); William T. Gormley, Jr. and Deborah Phillips, “The Effects of Universal Pre-K in Oklahoma: Research Highlights and Policy Implications,” The Policy Studies Journal 33, no. 1 (2005): 65-82.
 Haskins and Sawhill, Creating an Opportunity Society, Chapter 8; Gregory Camilli and others, “Meta-Analysis of the Effects of Early Education Interventions on Cognitive and Social Development,” Teachers College Record 112, no. 3 (2010): 579-620; Lynn A. Karoly, M. Rebecca Kilburn, and Jill Cannon, Early Childhood Interventions: Proven Results, Future Promise, RAND Corporation (MG-341), 2005
 Ron Haskins and W. Steven Barnett eds., Investing in Young Children: New Directions in Federal Preschool and Early Childhood Policy (Washington DC: Brookings and National Institute for Early Education Research, 2010), p. 4.