Thank you Chairman Lugar and good morning.

A. INTRODUCTION

 My colleagues and I at Mathematica Policy Research has been studying the Food Stamp Program for many years, and this research has often focused on participation levels.

 This testimony discusses how participation in the Food Stamp Program fluctuates with the economic cycle.  Let me start by reviewing recent program fluctuations, followed by a discussion of the research on why the program grew so dramatically between late 1988 and 1994, and why participation has since declined rapidly.  Then I'd like to close with a few thoughts on how the program might "behave," so to speak,  in the next recession.

B. PARTICIPATION IN THE FOOD STAMP PROGRAM FLUCTUATES DRAMATICALLY

 Participation in the Food Stamp Program has tended to rise sharply during economic recessions and then fall as the labor market tightens.  Participation peaked at 22 million during the 1983 recession and then fell steadily with the stronger economy until 1988.  From 1988 to 1994, participation increased by 9.5 million, a 52 percent increase, as the economy went into recession as shown in Figure 1.  From 1994 to this February, participation fell by 7.8 million, a 28 percent decrease, as the economy picked up.

C. WHY DID THE PROGRAM GROW SO SHARPLY FROM 1988 TO 1994?

 Based on studies Mathematica did in 1991 and 1993, there appear to be three reasons for program growth from 1988 to 1994.  I believe there is no doubt that the most important reason for the caseload increase between 1988 and 1994 was the 1990-91 economic recession.   As more Americans became unemployed and found themselves living in poverty, the FSP performed its traditional role of providing these people with important safety net resources. The second most important reason for FSP caseload growth was the growth of the Medicaid program.  Congress earlier had broadened Medicaid eligibility, and a number of states were attempting to improve maternal and child health by enrolling mothers and children in both Medicaid and other safety net programs including the FSP.  Our research suggests legislative changes in the FSP (IRCA and the Homeless Assistance Act) played a much smaller, but significant, role.

 The mechanism for caseload growth emerged from Mathematica's 1993 study on the dynamics of program growth.  Monthly entrants were 40 percent higher in the early 1990s than in the late 1980s.   At the same time, exits from the program slowed during the recession, and time in the program increased.  This confluence of more entrants and slower exit caused the program to grow rapidly from 1988 to 1994.

D. WHY HAVE 7.8 MILLION PARTICIPANTS LEFT THE PROGRAM
 SINCE 1994?

 There are three plausible reasons for the sharp decline in FSP participation since 1994: the 1996 legislation, which limited FSP eligibility; a strong economy; and the 1996 reform of cash assistance.  So we have to ask, how did the 1996 FSP changes contribute to the decline in participation?  These changes imposed a strict time limit on able-bodied adults without dependents.  They made most legal aliens ineligible for the FSP, slightly reduced the value of the maximum benefit, and limited certain deductions.  While we will know more about the effects of the legislation as program data that are just now becoming available are analyzed, in my opinion, these changes will be found to account for no more than 10 percent of the decline in FSP participation since 1994.

 How plausible is the second reason for the decline in participation since 1994?  Was the economy operating through the strong labor market truly responsible for the decline?  The number of unemployed persons fell by 3.6 million during this period, and the unemployment rate fell from 7.8 to 4.6 percent of the labor force.  Previously developed models of the relationship between the economy and FSP participation indicate that economic trends since 1994 could account for anywhere from 28 to 45 percent of the decline in participation.

 Finally, was welfare reform responsible for the decline in FSP participation?  The number of AFDC/TANF recipients has fallen by over 4 million since 1994, and the vast majority of the AFDC/TANF recipients were also receiving food stamps.  However, some substantial portion of that decline was also caused by the strong economy, and there is little agreement among researchers about how much of the shrinking caseload resulted from welfare reform and how much resulted from the economy.  Whatever the cause, the reduction in the AFDC/TANF rolls is associated with about half the decline in FSP participation.  That part of the reduction in FSP participation comes from a combination of the strong labor market and state and local programs to move recipients into jobs, along with welfare diversion and similar efforts.

E. HOW WILL THE NEXT RECESSION AFFECT FSP PARTICIPATION?

 Both the recent and the more distant history of FSP participation allows us to say with confidence that FSP participation will increase when the economy next falls into recession.  The 1996 changes that limited FSP eligibility will slightly moderate the absolute increase.  It is unclear how former TANF and FSP recipients who have moved into jobs will fare in the next recession.  It seems likely that many of them will lose these jobs, again needing to rely on the FSP and other elements of the safety net.