Welfare Rolls Inching Up? It's Not the Economy, USAT

In early May, Richard Wolf at USA Today tried to make a big deal over a very small statistic, and wrote one of those "signs of hard times" pieces that have become all the rage these days in Old Media (previous examples are here and here).

Wolf's piece was hampered by a possibly excusable math error, courtesy of the data supplied. But he also showed no curiosity as to why there have been such wide variations in state-by-state changes in the number of those receiving "welfare" (now known as Temporary Assistance for Needy Families, or TANF).

Here's how his report began:

States' welfare caseloads starting to rise

State welfare rolls, which declined for more than a decade after a 1996 overhaul of the nation's cash-assistance program, are beginning to rise, due in part to the struggling economy.

Federal data for the last half of 2007 show welfare rolls rose about 0.6%, and 27 states reported increases. That follows a decline of 68% since the federal law imposed work requirements, time limits and penalties for recipients who don't follow the rules.

..... The reversal of a downward trend that began in 1994 reflects a hard reality facing the Temporary Assistance for Needy Families program: The 3.9 million people who remain on welfare are mostly adults with physical, mental or emotional barriers to employment, as well as children being raised by someone other than their parents — often grandparents, who are not expected to get jobs.

..... Caseloads plummeted in the late 1990s, and while the 2001 recession caused some increases, they continued to drop through 2007. All but five states have cut caseloads by more than 50%. Wyoming leads with a 95% reduction; fewer than 500 recipients remain.

I'm not convinced that what Wolf claims in the third excerpted paragraph is correct. But if it is, wouldn't it negate the "it's the economy" argument as to why the caseload is increasing?

In any event, there is no overall evidence at the yearly caseload table accompanying the story that the 2001 sort-of recession interrupted the caseload decline (see this link; two negative quarters in a row never occurred; it's the National Bureau of Economic Research which, in July 2003, declared that a recession had occurred from March-November, 2001):

As to the most recent data, Wolf, perhaps understandably, missed the fact that the Administration for Children & Families (ACF) caseload table did not, and still does not, have a value in June 2007 for Maine. I verified that the total caseload at the table is the sum of all values listed, including Maine's zero.

Adding in 23,600 cases for Maine, roughly the average of the preceding and following months, totally eliminates the increase he reported:

3,858,204 - June 2007 caseload as reported by ACF
3,881,804 - June 2007 caseload, including Maine's estimate

3,880,321 - December 2007 caseload
22,117 - Reported caseload increase (0.57%)
- 1,483 - Actual caseload decrease (-0.04%)

Wolf's error is "perhaps understandable" because, while ACF's error is pretty deeply buried, he somehow had to learn that 27 states increased their caseload between June and December 2007. If someone simply told him, then the error is understandable. If he actually counted up the increases himself, he would have to have seen Maine's zero value, but not questioned it.

Giving Wolf's premise the benefit of the doubt for a moment (but as you will see, that moment will pass) -- It is true that the total welfare caseload began increasing after July 2007, from its alltime post-Reform low of 3.859 million to 3.911 million, or about 1.3%, during the next three months. But most of that increase was given back in November, as the caseload declined to 3.880 million, and December's increase was almost imperceptible.

Moreover, Wolf paid way too little attention to what has happened in individual states. From July's trough through December of 2007, here are the states whose caseloads (i.e., recipients) went up or down by what I see as either significant numbers or percentages:

In light of the above, I don't see how anyone can make a case that changes in TANF caseloads are significantly tied to the economy. If that were so, Ohio, Pennsylvania, West Virginia, and Illinois would be moving in the same direction. But, as you can see, Ohio's caseload is way up, while the others are way down. I would instead suggest that the changes, even in Welfare Reform's twelfth year, have a lot more to do with how aggressively each state is in trying to move recipients off the dole and into productive work.

California alone had a bigger caseload increase than the countrywide net change. A somewhat useful USA Today table that accompanies the article shows the percentage of nationwide caseload by state. For comparative purposes, the table would have been much more useful if it had also told us the percentage of the nation's population each state has. California, as an awful example, has 30% of the caseload even though its 36 million people make up only 12% of the country's population. If the not-so-Golden State's TANF caseload reflected the rest of the US, it would be a staggering 800,000 lower.

But I suppose that instead of looking at the detail as I have begun to, Wolf, like so many other Old Media reporters, found it easier to make breezy assumptions that fit his preconceived notions, one of which appears to be, "the economy has tanked." For real evidence to support that notion, Mr. Wolf will have to look elsewhere.

Cross-posted at BizzyBlog.com.