Note: This article is a sidebar to "Media Moguls on Board: Murdoch, Malone and the Cato Institute"
In 1995, during the national debate over welfare reform, the Cato Institute released a study called "The Work Vs. Welfare Trade-Off." The report was packed with statistics culled from Cato's research: "In 40 states welfare pays more than an $8 an hour job.. .. In nine states welfare pays more than the average first-year salary for a teacher. . . . And in the six most generous states it pays more than the entry-level salary for a computer programmer." On average, the study reported, welfare "pays" an average of $17,500 a year.
Not surprisingly, "recipients are likely to choose welfare over work, thus increasing long-term dependence." Therefore, "the most promising welfare reforms are those that substantially cut back on the level of benefits." These findings made a deep impression on the nation's editors and editorialists.
The Associated Press (1/19/96) ran a piece by its "Business Analyst," John Cuniff, lamenting that "as matters stand, it pays to be on welfare—and Americans, after all, have been taught to respond to incentives. A Cato Institute study shows that in 40 states welfare pays more than an $8 an hour job."
The Detroit News ran an editorial (9/24/95) questioning the assumption that "ending benefits would supposedly cause more destitution. The Cato Institute has rebutted the argument," it said, showing that welfare pays $10.20 an hour in Detroit. Therefore, Washington "should cut off all federal funding for welfare and reduce the tax rate accordingly."
The Arizona Republic (10/6/95) expressed similar feelings: "Astonishingly, in 16 states, welfare payouts turn out to be more generous than a $10 an hour job. . . Without a fundamental shift in these dependency incentives . . . it becomes difficult to envision a depopulation of the welfare state, much less its demise."
Columnists Tony Snow (USA Today, 9/25/95), Paul Craig Roberts (Business Week, 12/4/95), Walter Williams (Dallas Morning News, 12/9/95) and William F. Buckley (Sacramento Bee, 2/15/97) all relied on the Cato study to help illuminate their thinking on welfare reform.
And the Cleveland Plain Deader (7/25/96) led its piece on the study ("Welfare Package Pays Better Than Minimum Wage") with this grabber: "Anybody drawing the maximum welfare who'd give it up for a minimum wage job is probably too stupid to get one."
But on closer inspection, Cato's report turns out to be an embarrassing mess of bad math and dishonest numbers. In April 1996, seven months after the report was released, the liberal Center on Budget and Policy Priorities (CBPP) issued a critique of the Cato study, pointing out numerous analytical errors in the report, as well as simple faulty arithmetic.
For example, in calculating the value of welfare benefits for an AFDC recipient, Cato included the value of housing assistance (such as rent subsidies) which, depending on the state, added between $2,500 and $8,600 to the income of Cato's "typical" welfare family. But, as CBPP pointed out, only 23 percent of AFDC recipients receive any housing assistance whatsoever.
In defending its decision to count housing assistance as welfare income even though less than a quarter of welfare families receive it, Cato argued in a later report that "if a disease were killing 25 percent of Americans, few would dismiss the epidemic as minor."
Many benefits to the poor are given to all families below a certain income level, whether they are working or not--a fact that Cato failed to acknowledge. For instance, Cato calculated that Medicaid--the health insurance program for the poor--added $3,800 to $5,400 to the annual income of the typical welfare recipient. But Cato did not count Medicaid (or employer-provided health insurance) as income for working-poor families--even though 62 percent of children in working-poor families receive such coverage. In its response to CBPP, Cato did not even address this criticism.
After correcting for Cato's errors (including a miscalculation of the value of food stamps that Cato had to redo in its response to the critique), the CBPP concluded that a typical welfare family receives an average of $9,000 a year in benefits, instead of $17,500 as Cato reported. Cato responded that "quibbling over a few dollars' worth of benefits is unimportant."