Kausfiles Special

Crock of Goldstein

Two month’s ago the Washington Post’s Amy Goldstein made the front page with a story describing how welfare caseloads had stopped shrinking in Indiana and several other states. The piece didn’t come close to justifying its ominous hed (“Welfare Reform’s Progress is Stalled”) since caseloads were still falling nationally. But Goldstein had at least latched onto an interesting fact, namely that in some states the people left on the rolls were those euphemistically described as ‘hard to serve’–people with “overwhelming problems.” Goldstein seemed to imply that some new federal initiative was required, although it wasn’t clear what this might be, since Indiana was already redirecting its program (and one point of the 1996 welfare reform was to let states try various approaches and find out which ones work).

Last week, Goldstein was back on the front page with another vaguely portentous piece, “Geography of Welfare Is Changing.” This story is so muddled, ill-informed and dishonest that it justifies the return of the notorious kausfiles Snipe-o-Meter. As usual, the primary text (Goldstein’s piece) is printed in boldface, followed by context and perspective from the kausfiles staff.

As welfare rolls have plummeted in recent years, families that still rely on government assistance are increasingly concentrated in big cities, according to a new study, which cautions that welfare reform is failing to ease the burden of poverty on urban America.

This is Goldstein’s lede. She’s hyping a study from the Brookings Institution’s Center on Urban and Metropolitan Policy, a relatively new wing of Brookings that was pushed by former Fannie Mae chief (and Mondale aide) Jim Johnson. For the past few years the “Urban Center” has been putting out reports documenting with seeming precision what everyone has known about welfare for decades: that the most troubled, hard-core welfare recipients are those concentrated in urban ghettos (the so-called “underclass”). As welfare has been revamped, the national welfare rolls have fallen by about half–but urban welfare recipients haven’t left the rolls as quickly as non-urban recipients, so the remaining caseload has became more “concentrated” in cities.

This isn’t a surprise. But the Urban Center seems to regard it as a blinding revelation. “In 1994 … 48 percent of welfare recipients lived in the counties containing the nation’s 100 largest cities. By contrast, in 1999, these counties were home to 58 percent.” You see, it’s an urban problem! (“Cities fundamentally matter to welfare reform,” says the Center’s director, Bruce Katz.) And urban problems require urban solutions, and Urban Centers!

Reporters usually know enough about welfare to know that the Urban Center’s banal statistics aren’t news. (I once attended a Katz press event with a dozen reporters who sat around slack-jawed, asking themselves, “Do they really expect us to write about this?”) But this year the Center added a twist, implying that there was something close to a crisis at hand: “Current welfare policies may exacerbate cities’ burden of poverty, and cities cannot bear this burden alone,” warns the 2000 Report. Cities are “shouldering a disproportionate welfare burden.”

Goldstein fell for it. The only problem is that there is no evidence, in Goldstein’s piece or anywhere in the Brookings study–or anywhere else, that I know of–that welfare reform is, as her lede says, “failing to ease the burden of poverty on urban America.” Quite the opposite. Poverty has fallen across the nation since 1993. The poverty rate in central cities has fallen from 21.5 percent to 18.5 percent, which translates into almost 2 million fewer poor people. Welfare caseloads have fallen too. The urban share of the dwindling caseload has grown because urban caseloads have fallen more slowly than non-urban caseloads–but they’ve still fallen, and not by just a bit. The Center itself says, “Between 1994 and 1999, the urban counties’ welfare caseloads dropped by 40.6 percent.” [Emphasis added.]

That’s not just an easing of the “burden” of poverty, it’s a dramatic easing–made all the more dramatic because, under the 1996 reform, the amount of federal money available to assist the poor remains fixed, meaning that there is more money for each welfare case left. True, a 40 percent decline may not be as big an easing as the 51.5 percent by which the overall national caseload has shrunk, but so what? Things have still gotten much better in the cities, as far as the “welfare burden” goes.

It’s amazingly deceptive for a respectable institution like Brookings to suggest otherwise, and equally amazing that a respectable paper like the Post would fall for that perverse spin. Under the reasoning of Goldstein and the Urban Center, if there were only three welfare recipients left in the entire nation, but two of them lived downtown, then cities would be in trouble because they were “shouldering a disproportionate welfare burden.”

The evidence that welfare recipients are becoming clustered in big cities follows other recent findings that the program is becoming racially isolated, with African Americans and Hispanics accounting for a growing share of the families who remain on the rolls.

The air of crisis in Goldstein’s piece includes the worry that the welfare caseload is becoming not only increasingly urban but increasingly black and Hispanic. It would be perfectly understandable if that were happening, since African-Americans and Hispanics have always made up a disproportionate share of long-term welfare recipients (when compared with their share of all welfare recipients, long- and short-term). People who’ve been on welfare a long time presumably will have the most trouble getting off.

Perfectly understandable. Except that the Brookings study itself finds it isn’t happening. The Urban Center investigated the race issue, and reports: “Since 1996, the overall racial composition of the welfare caseload in 20 of the largest urban counties has changed only slightly.” The share of the caseload that was African-American increased by only “0.6 percentage points between 1996 and 1999.” That’s because “black and Hispanic welfare recipients in urban counties have left the welfare rolls at roughly the same rate as white urban recipients.”

In other words, it’s not just that Goldstein’s scare story about racial isolation isn’t really that scary–since even if the black share of the falling welfare rolls were increasing, the number of blacks on welfare would still be falling quite rapidly. It’s not just that Goldstein’s scare story about racial isolation may not be true. It’s that Goldstein’s scare story is refuted by the very study she’s reporting on and (in other contexts) relying on. She can write the sentence she wrote only because she either didn’t read, or suppressed, the Urban Center’s evidence. Astoundingly, her story doesn’t even mention that the Brookings study didn’t find the “racial isolation” trend.

P.S.: I call Goldstein’s reports “dishonest.” I don’t use that word lightly. In her earlier welfare story, discussing the people left on the caseload in Indiana, Goldstein’s lede sentence says that “neither Texas Gov. George W. Bush nor Vice President Gore speaks of anyone like Connie Galbraith,” a struggling Indiana mother “diagnosed with a number of psychiatric disorders” who has had trouble staying employed. Instead, Goldstein says, “[b]oth candidates … would redirect the government’s attention from those who remain tethered to the welfare system to those who have joined the ranks of the working poor.”

I don’t know about Gore, but this seemingly evenhanded condemnation recklessly, if not willfully, mischaracterizes Bush’s push for “faith-based” efforts to “save and change lives” (something he calls the “next bold step of welfare reform”). Whatever you might say about Bush’s plan–I tend to think it’s an ineffective alternative to a big public jobs program–it is directed precisely at people like Connie Galbraith, at people “tethered” to the welfare system, people with multiple problems, “young mothers without self-respect and education” (Bush’s words), people whose “life is broken” (Bush again).

There’s a pattern here. In each of the two welfare stories, Goldstein simply ignores or suppresses central, inconvenient truths that don’t fit at all into her thesis (whether the thesis is that both Gore and Bush are ignoring the ‘hard-to-serve’ welfare recipients or that the caseload is growing increasingly black.) The Post’s editors wouldn’t tolerate this sort of misreporting on an issue they really cared or knew something about, like campaign finance reform or electoral strategy. The poverty beat remains foreign territory. But must it bear the disproportionate burden of Goldstein?