Fighting for Reliable Evidence

Chapter 7: Waiver Evaluations: How Random Assignment Evaluation Came to Be the Standard for Approval

CHAPTER 7 Waiver Evaluations: How Random Assignment Evaluation Came to Be the Standard for Approval* To accomplish these things we had to go to Washington and kiss someone’s ring to say, “Please give us a chance to try something that might work.” —Governor Tommy Thompson1 Paraphrasing Senator Daniel Patrick Moynihan, political officials are entitled to their own policies, but not their own science. —Richard Bavier2 Throughout the first Reagan administration, the Department of Health and Human Services (HHS) applied the approval standard described in chapter 5 to section 1115 waiver applications: it approved waiver requests from states whose policy proposals were consistent with administration policy and denied those from states that wished to try anything different. Evaluation requirements were minimal, which was plainly permissible as a legal matter, given the simplicity of the standard and the broad secretarial discretion accompanying it. In this chapter I recount how that approval standard became increasingly untenable, as the federal stance on welfare reform shifted to the more federalist approach of encouraging states to test their alternative options for reform. *Chapter 7 authored by Howard Rolston. 218 FIGHTING FOR RELIABLE EVIDENCE In the process, federal officials gradually changed the waiver approval standard to one that focused primarily on holding the federal cost of state-initiated reform initiatives no higher than a ceiling—which they eventually decided to enforce by requiring inclusion of random assignment evaluations, using the control group experience to establish the cost cap (see table 7.1 for the major events in the welfare waiver story). THE FIRST SKIRMISH On April 19, 1985, less than three months into the second Reagan administration , the governor of North Carolina, James G. Martin, sent a letter to HHS secretary Margaret Heckler requesting a waiver of the state Aid to FamTable 7.1 Timeline of the Welfare Waiver Story and Related Events Date Event July 1985 HHS offers North Carolina waivers for random assignment February 1987 Reagan administration budget proposes waiver authority for a number of low-income programs February 1987 Wisconsin governor Tommy Thompson meets with HHS secretary Otis Bowen to describe upcoming waiver request July 1987 President Reagan creates ILIOAB August 1987 OMB proposes randomly assigned cost control group September 1987 HHS approves New Jersey demonstration without random assignment November 1987 ILIOAB adopts guidelines with random assignment as the preferred design April 1989 President George H. W. Bush creates the LIOB and directs it to provide substantial deference to state requests for nonrandom assignment designs February 1992 Bush administration budget announces states as laboratories initiative April 1992 Wisconsin governor Tommy Thompson agrees to random assignment for Bridefare February 1993 President Bill Clinton tells governors of his support for waiver flexibility August 1996 Temporary Assistance for Needy Families (TANF) becomes law, rendering waivers irrelevant Source: Author’s compilation. HOW RANDOM ASSIGNMENT EVALUATION CAME TO BE THE STANDARD 219 ilies with Dependent Children (AFDC) plan to permit an increase in child care for welfare mothers. The requested approach was as follows: Child Care Resources, Inc., which administered public day care funds in Mecklenburg County, North Carolina, had received a $300,000 matching grant from the county, with which it had successfully raised an additional $300,000 from a number of private and public funders to establish a Child Care Recycling Fund. The purpose of the fund was to increase the availability of subsidized child care for AFDC-eligible parents who wanted to go to work by “recycling ” the federal, state, and county AFDC (and food stamp) funds that would be “saved” because of reduced benefits to the AFDC parents who would go to work. To us at HHS, this was an all too familiar claim—that spending more money to provide new benefits to welfare parents would increase employment and reduce welfare expenditures sufficiently to offset the cost of the new benefit . It was clear that every AFDC case that left the rolls and received subsidized child care to support employment would generate transfers to the fund, whether or not that mother would have found employment without the subsidized child care. In the vocabulary of program evaluation, the proposed recycling fund would rely on an overstatement of savings to AFDC because gross outcomes, not net impacts, would be the basis for the calculation. A further mark against approval, in the eyes of the administration, was that the child care benefit was simply continuing dependency, albeit in another form. Thus it would have been...