Although you might have thought it would be easy to reauthorize one of the most successful government programs ever enacted, in Washington everything is hard. A year ago, the reformed federal welfare program expired. Ever since, Congress has been passing short-term extensions and arguing. The House has passed one bill, and last week the Senate Finance Committee approved another. Whether Congress can settle the argument this year is an open question.
Still, the bigger picture is encouraging. Since the mid-1990s, almost everyone has accepted that welfare should be linked to work. Only the most reactionary of liberals want to go back to providing cash as a substitute for employment. The new consensus is a good thing, because it matches a new reality. No feasible amount of cash assistance could solve America's poverty problem, even in principle. The problem has changed. It has become more behavioral than economic.
I grew up, in the 1960s and 1970s, taking for granted that the poor were just like you and me, only with less money. They were victims of a stingy government and a harsh economy. Poverty could be abolished by writing checks. America's unwillingness to rise to the task showed an unconscionable lack of compassion and common sense.
In recent years, several lines of evidence have converged to suggest that the "it's the economy, stupid" view of poverty is plain wrong. One is research by a sociologist named Susan Mayer, whose work I described in detail in an earlier column (November 13, 1999). She performed a variety of innovative statistical tests and found that lack of money was more an effect of poor people's other, more-fundamental problems than a cause in its own right; and so handing out more cash would be of little help. Other research similarly pointed away from money and toward the importance of two-parent families, education, and work.
A second line of evidence comes from the 1996 welfare reform itself. From 1965 to 1995, notes Ron Haskins of the Brookings Institution, the federal and state governments increased their means-tested spending, adjusted for inflation, by a multiple of seven. Yet child poverty increased. After the mid-1990s, when state and federal programs switched their emphasis from writing checks to encouraging work, welfare caseloads fell by a stunning 60 percent, an unprecedented number of single mothers found and kept jobs (even through the recent recession), and child poverty dropped—for black children, Haskins says, to its lowest level ever. "I don't think any public policy has ever had this kind of impact," Haskins says.
Now comes a third line of evidence, in a Brookings paper just published by Haskins and his colleague Isabel V. Sawhill.
Sawhill is a liberal, and a former Clinton administration official, who began studying the behavioral roots of poverty in the days when most liberals condemned all such thinking as "blaming the victim." Then came welfare reform, which "was far more successful than most people—including me—anticipated," Sawhill says. The strong economy helped, she says, but "we've also got to stop thinking of people as passive victims of the economy and whatever the social safety net provides. Liberals have too often emphasized the income-to-behavior link without also recognizing that there's a behavior-to-income link as well."
In their new paper, she and Haskins use detailed census data and statistical modeling to simulate what would happen if the poor worked as many hours as the nonpoor, at jobs matching the workers' actual qualifications. The result: Full-time work would reduce the poverty rate from today's 13 percent to 7.5 percent—almost half.
Separately, they next ask what would happen to the poverty rate if the poor were as likely to marry and stay married as they were in 1970, to real-world partners of similar age, education, and race. (Contrary to a widely held assumption, Sawhill and Haskins found no shortage of marriageable men in most segments of the population, the important exception being in some African-American education and age categories.) Increased marriage alone, by combining two adults' incomes, reduced poverty to 9.5 percent.
Getting a high school degree and having no more than two children also reduced poverty, though not as much. And if the poor did all four—worked full-time, got married, stayed in school, and stopped at two kids—the poverty rate would drop to less than 4 percent.
The bigger surprise, however, was yet to come. Sawhill and Haskins then simulated a doubling of all welfare benefits, much more than anyone seriously contemplates. The result? Poverty dropped from 13 percent to 12 percent. The meter barely jiggled. Even a massive welfare increase would have less effect than any one of four kinds of behavioral change.
"If people did a few things—graduated from high school, got a job, and delayed having a baby until they married—our analysis shows that would eliminate a huge chunk of poverty in this country," says Sawhill, "and that would be far more effective than anything we could feasibly do through the welfare system alone."
If liberals have averted their eyes from the behavioral sources of today's poverty, conservatives, with a prominent exception, have averted their gaze from the corrosive effects of growing inequality. The exception is Charles Murray, whose 1984 book, Losing Ground: American Social Policy, 1950-1980, played a part in discrediting the old welfare system. In a 2002 article in The American Enterprise, he argued that the crumbling of the two-parent family within lower-income strata implies that "class segregation will increase and social mobility at the bottom will decline. And America's image of itself as one big middle-class society will wither away."
Coming from the liberal side, Sawhill now lends him some support. Since the late 1960s, she notes in the fall 2003 issue of The Public Interest, the American population has been working more hours as women entered the labor force; but the poor, uniquely, have been working less. A "growing salary gap," she writes, "has been greatly amplified by a growing hours gap." Likewise, since the 1960s single parenthood has nearly tripled among the least-educated women (from 7 percent to almost 20 percent), while barely changing among the best-educated (at about 5 percent).
The result, Sawhill finds, is a growing bifurcation of children's life prospects. More kids are born into high-risk homes, with a single mother who dropped out of high school, an often absent father, fitful employment. More kids are also born into low-risk homes, where married parents with college degrees earn high incomes. Fewer kids are born in the middle, to married parents with high school degrees and at least one full-time job—the sort of home that formed the backbone of the postwar middle class. "In other words," Sawhill writes, "as a result of changes in work and family patterns, today's children are getting a much less equal start in life than the children who were born a few decades ago."
That would not be quite so disturbing if people readily moved up from the bottom over the course of their lives. Alas, there is reason to fear that impoverishing behavior is harder to escape than economic disadvantage. Kids who don't know their fathers, and whose teenage mothers dropped out of school, and who grow up in neighborhoods where married parents are curiosities, are much more likely to become unmarried parents and dropouts themselves. Kids who are doted on by two college-educated parents never even think about failure.
"We know we have growing income inequality," Sawhill says, "but there's a lot less focus on another gap that's opening between rich and poor, and that's in the behavioral domain, and it involves both work and marriage." This new gap, she adds, "doesn't bode well for the future of social relations in America." Unchecked, it might lead to permanent class barriers, something America has, until now, been mercifully spared.
Forget about the haves and the have-nots. America now faces a divide between do's and do-nots. Coping requires conservatives to see that inequality threatens mainstream values, and liberals to see that mainstream values are the key to reducing inequality. Conservatives, Sawhill argues, will need to spend more generously on child care subsidies and wage supplements and last-resort jobs to get the poor working (jobs bring mainstream values as well as money). Liberals will need to accept that money without behavioral change is useless or worse.
The good news is that Congress is moving in the right direction: toward more emphasis on work, more child care support, and new efforts to promote marriage and deferred childbearing. What remains to be seen is whether the policy can outpace the problem.