"What's the flaw in Nina Bernstein's front-page NYT story that seemingly ties welfare reform with a decline in marriage? The studies Bernstein publicizes looked at people who were on welfare. Those who kept getting the old-style welfare checks got married at higher rates than those who went through welfare-to-work programs. But the studies didn't look at what happened to people who never go on welfare in the first place once they know that if they do they will have to go to work. Indeed, even if welfare reform mildly discourages those who are on welfare from getting married that factor could be outweighed by the fact that there are less than half as many people on welfare as here used to be. That's why national figures might show, as they do, that a smaller percentage of children are living in single-parent homes, while marriage is increasing among African-Americans. ... Bernstein acknowledges the conflict between her two studies and the national figures, but then suggests that her studies are superior because they "focus on welfare applicants." No! They present an incomplete picture because they "focus on welfare applicants," not on the much larger group of people who might go on welfare -- or, more precisely, who might have gone on welfare under the old system. ... It's the overall national marriage numbers we want to affect, not the numbers for those who happen to be on welfare, and since welfare reform in 1996 the overall national numbers have been going in the right direction for the first time in generations. (Bernstein also implies that the economy caused this effect, but previous good economies didn't.) ...
Nor do Bernstein's two studies of welfare applicants in two states seem well-positioned measure the larger cultural changes that can occur when all potential welfare recipients in all states know the rules have changed acrosss the nation. It matters less if there's a differential between those in welfare programs and those in welfare-to-work programs if both are moving in a pro marriage direction. Who's to say that even those in the "control" group receiving traditional welfare after the 1996 reform weren't affected by the reform? Did they stop reading the papers or listening to the news or to popular music? ...
That said, it's not crazy to think that greater self-sufficiency for single mothers might lead to less marriage. (Women in these programs are often told "Yes, you can make it on your own!") It's just that this effect doesn't seem to be winning out over the other pro-marriage effects of welfare reform (like wanting to have someone to help you out, if you're going to have to juggle work and parenting). ... Nor can Bernstein have it both ways -- if welfare reform is making women less likely to marry because it is making them richer and more self-sufficient, then it can't be making them less likely to marry because they "are now doing much worse" economically than before. (Bernstein skillfully tries to fudge this contradiction for her sources in the first paragraph on this Web page. She's still "the Advocates' Advocate"!). ...
To see the anti-reform broadside that is the press release for the larger study from which Bernstein plucks her marriage angle, click here. The release mentions that "[i]n Connecticut, for example, women participating in the state's Jobs First program showed a lower marriage rate." But the study also looked at Florida and California. Hmmm. What happened to marriage rates there? Answer: Only Connecticut appears to have had a "control group" of women who stayed on old-style welfare. Still in California and Florida, as far as I can see, two different "waves" of interviews, two years apart, were undertaken, and the text says "self-reported marital status did change significantly between waves 1 and 2, rising from ...7% to ... 12%." This statistically signifcant increase is then waved away as a possible "artifact of sampling design and mother age." But is it? ... The full 102-page PDF-format study is here. You, the reader, decide whether my suspicion -- that if indicia of marriage had gone down they would have made a huge deal of it and maybe gotten Nina Bernstein interested! -- is correct. (See esp. page 60.) ... P.S.: The share of these single mothers who lived in "a household with one other adult" also rose, from 29% to 34%."