While I was moving, the tenth anniversary of welfare reform came and went without too much fanfare. Many observers simply declared that welfare reform had "worked," because caseloads had declined. But what kind of metric is that? The whole point of reform was to push people off welfare. Of course caseloads were going to decline—that's what the law did. We should be asking if families are now better off because of it. I'm sure you can guess my answer but I'll go through the numbers anyway.

Nation-wide, welfare caseloads declined from 4.4 million families in 1996 to about 2 million families today, as states began kicking people off the rolls. So what happened to all of those families? The Urban Institute conducted a massive survey on this, and found that in 2002, of those who left welfare:

57 percent were working (about 40 percent full-time)

26 percent had returned to welfare

14 percent had "no employment income, no working spouse, and no cash welfare or public disability benefits." (Presumably some of these families receive other benefits from America's stingy safety net, such as housing assistance, food stamps, or WIC grants.)

Welfare "reform" has obviously failed the last group, people who by and large are in poor physical and mental health and unable to work. That's about 300,000 families so far, and the number has grown since 2000. Many more families who still receive assistance will likely face a similar situation once they run up against the five-year time limit and get kicked off welfare. That's especially true if Republicans in Congress get their way and pass rules that would make it harder for states to evade these time limits.

Now consider the first group, those who went back to work. Their median hourly wage was about $8 per hour in 2002, or about $16,000 per year. Many families do better, many do worse. For a single mother, though, that's not enough to raise children on—it's hardly a wonder that the poverty rate among single working mothers increased slightly during the "booming" economy in the late 1990s, as women were pushed off welfare to find jobs that paid little. And those jobs are often precarious; an illness or a broken car can easily mean getting fired, with only a shredded safety net to fall back on.

So has work made families better off? The MDRC has done some valuable case studies on the subject and found that, for instance, in Los Angeles "families were not substantially better off financially, even though many parents went to work." And California has one of the more generous state programs around. The flip-side, too, is that now those parents—many of them single mothers—have less time to spend with their kids and someone has to pay for child care.

And what about the children? Well, in Minnesota, which also has one of the more generous TANF programs in the country, reform "had no overall effect on the elementary school achievement of very young children." (Some disadvantaged children in the state saw gains, although note that Minnesota was one of the few states that spent more under its reformed TANF program than it did under the old AFDC system.). The Urban Institute study also noted that "children's outcomes were largely unchanged" nationwide. Just because welfare reform wasn't nearly as disastrous as critics predicted doesn't mean it's done a lot of good.

And what about poverty reduction? In 1999, after welfare reform plus a roaring economy plus new work-support programs like the EITC, the poverty rate was 11.8 percent, which was... the exact same rate as in 1979, under the "bad old welfare" system. (1979, note, was only two years before the Reagan administration started slashing AFDC benefits). Moreover, the black poverty rate—which "reform" was supposed to help reduce—declined faster in the three years before 1996 than it did in the three years after. So what, exactly, did welfare reform accomplish, apart from pushing people who genuinely need assistance off the rolls and saving the government a few bucks?

Many liberals will say that welfare reform can and will be a stunning success story if only we increase the Earned-Income Tax Credit and raise the minimum wage and provide heaps more funds for child care and heaps more money for job-training and so on. Well, no kidding. Cup Noodle makes a great meal if it comes with a side of steak. If people are going to be forced to work then the government should make work pay. If society isn't willing to do this, then cash assistance is the way to go.

Over at TNR, meanwhile, Jon Cohn argued this week that welfare reform was a positive thing because it got an unpopular monkey off the Democratic Party's back and gave them "more political room" to propose new anti-poverty programs. That could be, although it seems like a difficult counterfactual to prove. I'll note, however, that the flagship anti-poverty program being pushed by Democrats right now is a proposal to raise the minimum wage to $7.15 an hour by 2008. In real terms, this will be equivalent to about $5.60 in 1997 dollars, when the minimum wage was last raised to... $5.15. A radical shift? Really?