Museado / NARA Collection wjc-dpc — Collection wjc-dpc
Recent Folios
Login Register
English español (Spanish) magyar (Hungarian) हिन्दी (Hindi)

Survos folio page chat

Welfare - Food Stamps - Caseloads I [1]

Ask the Scholar

Page 102 of 102
I can add historical knowledge about this page.

Page image

Page 102

OCR

Who Receives Food Stamps Today? Most food stamp recipients are children or elderly. Over half (51 percent) are children and another 8 percent are age 60 or older. Working- age women represent 29 percent of the caseload, while working-age men represent 12 percent. The majority of food stamp households do not receive AFDC or TANF benefits. Only 35 percent do so. Other cash assistance received by food stamp households includes Supplemental Security Income (27 percent of households), Social Security (21 percent) and State General Assistance benefits (6 percent). Nine percent of households have no income of any kind. Many food stamp recipients work. Twenty-four percent of food stamp households have earnings, and for these households, earnings are the primary source of income. Food stamp households have little income. Only 9 percent are above the poverty line, while 40 percent have incomes at or below half the poverty line. The typical food stamp household had gross income of $558 per month and received a monthly food stamp benefit of $169. Nearly one-fourth of monthly available funds (cash income plus food stamps) available to a typical household comes from food stamps. Food stamp households possess few resources. The average food stamp household possesses only about $92 in countable resources (including vehicles, checking and savings accounts, and other savings). Most food stamp households are small. The average food stamp household size was 2.4, but varied considerably by household composition. Households with children were relatively large, averaging 3.4 members. Households with elderly members tended to be smaller, with an average size of 1.3 people. 12/1/98 FY 1997 data FAX COVER SHEET OFFICE OF ANALYSIS AND EVALUATION FOOD AND CONSUMER SERVICE, USDA (Office: 703-305-2019) (Fax: 703-305-2576) DATE: 3-26-99 TO: ERIC GOULD OFFICE: PHONE #: FAX PHONE #: FROM: STEVEN CARLIUN PHONE: # PAGES (Including cover page): 9 COMMENTS: SECOND TRY se 6/T'd S60'ON WH88 :8 6661. '97 "AHW Who Leaving the Food Stamp Program? An Analysis of Caseload Changes from 994'to 1997 United States Department of Agriculture Office of Analysis, Nutrition, and Evaluation Food and Nutrition Service March 1999 Background The number of people receiving food stamps These questions cannot be answered fell by over 5.9 million between summer conclusively yet. However, an analysis of 1994 and summer 1997, with most of the food stamp administrative data provides decline occurring in the year between some initial insights into the changes that September 1996 and September 1997. This have occurred in the Food Stamp Program. decline occurred during a period of strong This analysis examines FSP administrative economic growth - unemployment fell, data from 1994, when FSP caseloads inflation stayed low, and the percentage of peaked, and 1997, the most recent year when Americans living in poverty fell slightly. In full data are available. 1997 was a transition the same period, Congress enacted and year as States replaced the Aid to Families States implemented sweeping reforms to the with Dependent Children (AFDC) program Food Stamp Program (FSP) and to the with Temporary Assistance for Needy nation's welfare programs. Families (TANF) and when changes to the FSP took effect. Thus this report provides Some point to the participation decline as information about caseload changes during proof that the strong economy is lifting all the early stages of implementation This boats, providing job opportunities and higher analysis focuses on those groups most wages for all, including low-income families affected by welfare reform - single parents, Welfare reform is credited with moving legal immigrants, and unemployed childless millions of families from dependence on the adults. state to work. General Trends However, others raise more troubling questions. Suppose familles are not leaving As Table 1 shows, three major groups the safety net for paid employment and self- accounted for almost all the drop in the food sufficiency but rather remain poor but stamp caseload between 1994 and 1997. without assured access to sufficient food to The number of legal immigrants fell by 54 meet their basic needs? Suppose people percent, accounting for 14 percent of the leaving welfare remain eligible for food total decline. The number of childless stamps but don't realize that these benefits unemployed adults fell by 44 percent, are still available to them? These critics cite accounting for 8 percent of the total decline. the increased demand for assistance at food Finally, the number of families receiving pantries and soup kitchens as evidence that welfare benefits fell by 28 percent. Because declining caseloads are not necessarily a sign this group accounts for such a large share of of success. the FSP caseload, they represented 61 This report was prepared by Jenny Genser. Office of Analysis. Nutrition, and Evaluation, based on data prepared by Scott Cody and Laura Castner of Mathematica Policy Research, Inc. The information presented in this report is based on data collected by Food Stamp Program Quality Control data for fiscal years 1994 and 1997. 6.2% S60'ON W988:8 6661 percent of the decline. Only 17 percent of while the number of single-parents receiving the decline came from other groups, AFDC or TANF benefits fell overall, the including elderly, the disabled, and number of those who combined welfare with households with earnings and no welfare work also rose by 9 percent. This points to income. an increased reliance on work among single parents. Thus, the steepest declines in participation occurred among legal immigrants and However, it is important to note that the unemployed childless adults, the two groups number of single-parent households with no affected by tough new restrictions under income from either welfare or work also rose welfare reform. However, most of the by 9 percent. Also, the number of single- reduction occurred among the large group of parent food stamp households receiving cash welfare households. welfare dropped by nearly 900,000, while the number not receiving welfare rose by only Trends Among Single Parent Families 120,000. Because we only have data on food stamp participants, we don't know Among other changes, welfare reform was whether the households leaving both welfare designed to move families from welfare to and food stamps found jobs and are work. The Aid to Families with Dependent economically independent, or are Program (AFDC), which was an entitlement unemployed and in need. for needy single parents with children, was replaced by block grants to States to operate Trends Among Legal Immigrants the Temporary Assistance to Needy Families (TANF), which was designed to provide Welfare reform legislation made most legal short term assistance to help poor families immigrants ineligible for food stamps. Those become economically self-sufficient who were participating at the time the law went into effect could participate until As Table 2 shows, the number of single- September 1997. New applicants became parent families, the group most affected by ineligible starting October 1996. welfare reform, fell by 17 percent. Within this group, the number receiving AFDC or In 1994, nearly 1.5 million legal immigrants TANF benefits fell by 27 percent but the received food stamps. This number dropped number of those not receiving welfare rose sharply after welfare reform was enacted. by 9 percent. This suggests that welfare The number of legal immigrants receiving reform has been partially responsible for the food stamps declined steadily throughout drop in welfare caseloads Had it been solely late 1996 and most of 1997 (Figure 1). The the strong economy, the number of food decline was gradual throughout the year, stamp households with welfare and without rather than falling sharply between August welfare would have both fallen by similar and September 1997. This indicates that as amounts. It also suggests that at least some current immigrants left the program, they families that no longer receive welfare were not replaced by new immigrant continue to receive food stamps. participants. Among single-parent families, the number of Restrictions on participation by legal those with earnings rose by 10 percent And immigrants appear to have deterred 2 6/8'd NO.095 USDA/FCS/OAE WHAR:8 6661 participation by their children, many of caseload that they represent fluctuates. whom retained their eligibility for food During periods of caseload expansion, stamps. Participation among U.S. born elderly households represent a declining children living with their legal immigrant share of the food stamp population. During parents fell faster than participation among periods of caseload decline, they represent a children living with native-born parents larger share. (Table 3). The number of children living with legal immigrants fell by 37 percent, Changes in Ethnic/Racial Composition versus 15 percent for children living with native-born parents. Some have raised a concern that those finding work and leaving welfare are The number of naturalized citizens receiving predominantly non-Hispanic whites, leaving food stamps rose by 173,000 between 1994 the welfare caseload even more and 1997, an increase of 66 percent (Table disproportionately minority. The data do not 4). This reflects the surge in naturalizations support this claim. The racial composition of starting in 1993. the food stamp caseload as a whole is virtually unchanged between 1994 and 1997 Trends Among Childless Unemployed (Figure 3), despite steep drops in Adults participation overall. Welfare reform restricted most childless The same is true for AFDC/TANF recipients. unemployed adults to no more than three Whites left the food Stamp Program and months of food stamps in a 36-month period, welfare at the same rate as minorities. The unless they were employed or participating in number of whites receiving welfare and food qualified work programs. Many parts of the stamps between 1994 and 1997 fell by 32 country were exempt from the work percent, compared to a 31 percent drop for requirement and time limit, due to waivers African Americans and a 27 percent drop for granted to areas with high unemployment Hispanics. As a result, the caseload rates or insufficient jobs. composition is very similar for both years. As expected, the number of unemployed The one group where the racial/ethnic childless adults fell by 476,000, a drop of composition did change significantly is the one third, between August 1996 and unemployed childless adult group subject to September 1997 (Figure 2). This decline time limits. While the proportion of blacks was sharpest in the period between January in this group remained unchanged between and March, 1997, as States implemented the 1994 and 1997, the proportion of whites fell, time limits. while the proportion of Hispanics doubled. However, this is a relatively small group that Trends Among the Elderly may be heavily affected by State waiver policy. The number of households with aged members dropped by 86,000 from 1994 to About the Data 1997, a decline of less than five percent. Over time, the number of elderly receiving The data come from Food Stamp Quality benefits is very steady, while the share of the Control records. The cases are derived from 3 S60'ON W868:8 MAR. 26. 1999 State samples of caseloads pulled each month for a review on payment accuracy. Records from all States for all months during a fiscal year are combined into one file. Each year, there are about 50,000 households represented in the data. The files are then edited for consistency and weights are assigned. We analyzed data from two years -- 1994, the year participation peaked; and 1997, the most recent year that we have complete data. Limitations of the Analysis The analysis compares snapshots of the food stamp population at different points in time. It does not follow individuals or families over a course of time, showing their movements on and off jobs, welfare, or food stamps. Nor does this analysis provide any information about low-irlcome households not receiving food stamps including former participants. 4 6/S'd 560'0N USDA/FCS/OAE WHAD:8 6661 '97 XHW TABLE 1 Participation Changes from 1994 to 1997 (in thousands) Participants: Participants Participation Percent Share of Summer 1994 Summer 1997 Change Change Decline Legal Permanent Residents 1,537 706 -831 54% 14% Childless Unemployed Adults 1,148 648 -500 -44% 8 % AFDC/TANF Participants 13,052 9,442 -3,610 -28% 61% All Other Participants 11,697 10,707 -990 -8% 17% TOTAL 27,434 21,503 -5,931 -22% 100% TABLE 2: Single Parent Households by Presence of Earnings and Welfare Receipt: 1994 and 1997 Households: Households: Participation Percent 1994 1997 Change Change All Single Parents 4.595 3,816 -779 -17% With AFDC/TANF 3,319 2,422 -897 -27% Without earnings 3.022 2,097 -925 -31% With earnings 297 325 28 9 % Without AFDC/TANF 1.276 1,394 118 9% Without earnings 568 614 46 8 % With earnings 708 780 72 10% With earnings 1.005 1,10S 100 10% 5 6/9'd S60'ON USDA/FCS/OAE WH00:8 6661.92'86W FIGURE 1: Legal Immigrant Participants: August 1996 through September 1997 1.800 , 400 ,00 000 BOD Number on 000 400 200 0 $ a My was Saptember-06 October-Df Homema 8 Der Derector's 10 - Fatury With April # May 187 Are-91 and A Sectimes BY . TABLE 3: Number of Children Participating by Citizenship Status of Parents: October 1996 and September 1997 (in thousands) Participants: Participants: Participation Percent October 1996 September 1997 Change Change Children Living with Legal Immigrants 1,251 742 434 37% Children Not Living with Legal Immigrants 11,034 9,804 1,682 15% 6 P.7/9 S60'ON WHOD 8 MAR. 26. 1999 TABLE 4: Number of Legal Immigrants by Status: 1994 and 1997 (in thousands) Participants: Participants: Participation Percent 1994 1997 Change Change Permanent Resident Aliens 1,453 547 -906 - 62 % Refugees 359 235 -104 - 31 % Naturalized Citizens 264 437 +173 66% All Legal Immigrants 2,056 1,219 -837 - 41 % Figure 2: Childless Unemployed Adults: August 1996 through September 1997 1.200 1 000 800 Unemployed Chedres AGUES Humber in thousands) 600 400 200 O X & August 96 September 96 Dctober-96 November 96 Cocamber = = Amount = Fathery = Byd April May91 June-97 rev-97 August-97 September-97 - --- 7 6/8'd S60'ON USDA/FCS/OAE MAR.26.1999 8:41AM Figure 3 Racial Composition of Food Stamp Participants Receiving Welfare: 1994 and 1997 Other FY 1994 5% Hispanic 19% White 41% Black 35% Other FY 1997 4% Hispanic 19% White 41% Black 36% 8 6/6'd d S60'ON 8:41AM MAR.26.1999 FAX COVER SHEET OFFICE OF ANALYSIS AND EVALUATION FOOD AND CONSUMER SERVICE, USDA (Office: 703-305-2019) (Fax: 703-305-2576) DATE: 5-25-99 TO: EXIC GOULA OFFICE: DPC PHONE #: FAX PHONE #: (202)456-7431 FROM: STEVEN CARLSON PHONE: (703) 305-2134 8 # PAGES (Including cover page): COMMENTS: ERIC: I JUST GOT YOUR 5/20 B-MAIL ASKING FOR FAXOO COPY OF MATORIAL on PARTICIPATION DECLINE. HERE. IS Thto ENTIRE PACK760 some OF UNICH I FAX50 TO YOU AND KOWIN LAST . THURSDAY. AA se NO.024 P.1/8 WH80:8 NOT FOR ATTRIBUTION, CITATION, OR PUBLICATION May 20, 1999 -DRAFT-- Understanding Food Stamp Program Caseload Decline In February 1999, the Food Stamp Program served 18.29 million people, the fewest in nearly 20 years. The number of people receiving food stamps over the past five years has fallen by 9.7 million, a drop of over one-third. This trend far outpaces the decline in poverty. Between 1995 and 1997, food stamp participation fell five times as fast as poverty. A forthcoming GAO report will note that the decline in the number of children participating in the program between 1994 and 1997 substantially outpaced the decline in the number of children living in poverty, resulting in a substantial fall in the percentage of poor children receiving food stamps. At least five factors contribute to the decline in food stamp participation: Changes in program rules under welfare reform restricted the participation of immigrants and unemployed childless adults. The strength of the Nation's economy allowed some participants to find work, reducing their need for food stamps. The success of the Temporary Assistance for Needy Families (TANF) Program moved some participants from welfare to work, with an increase in income sufficient to eliminate the need for food stamps. Barriers to participation erected by some States and localities may have deterred some eligible individuals from applying, or resulted in some participants leaving the program despite their continued eligibility. Lack of awareness of eligibility for the program may have led some participants to leave the program unnecessarily and discouraged others from applying for benefits. There is little reason for concern about falling participation if it is largely due to the first three of these factors - the expected consequences of economic growth and welfare reform. There is greater reason for concern and a potential need to explore policy options to redress problems to the extent that falling participation is attributed to the last two factors. Understanding the relative importance of these factors represents a complex analytical program. By piecing together information from a variety of sources (with different degrees of reliability), we can begin a preliminary assessment of their contribution. This analysis is limited to the period between March 1994, the month in which food stamp participation peaked, and September 1997, the last month for which we have relatively complete information Over this period, food stamp participation dropped by just under 7 million people. As shown in Table 1, this preliminary assessment suggests that roughly 20 percent of the decline in food stamp participation over this period was due to new limits on immigrants and unemployed jobless adults, roughly 30 percent was due to success of economic expansion and welfare reform in moving people into jobs, roughly 32 percent was due to a decline in participation rates among otherwise eligible individuals, and roughly 18 percent was left unexplained. 8/2'd NO.024 USDA/FCS/OAE WH82:8 6661 Table 1 Explaining the Drop in Food Stamp Participation: March 1994 - September 1997 Number of Relative Participants Contribution Source Net change in participation 6,974,000 100% FNS administrative records Welfare reform limits on: Immigrants 1,053,000 15 FNS administrative records Unemployed adults 327,000 5 FNS administrative records Employment growth 2,117,000 30 FNS forecast model Lower participation rates 2,198,000 32 Preliminary estimates from Current Population Survey All other/unknown 1,279,000 18 Changes in Program Rules The food stamp provisions of welfare reform restricted the participation of many legal immigrants and unemployed childless adults. The effect of these provisions can be seen in the drop in participation among these groups in the period immediately following enactment of the 1996 welfare reform law. From September 1996 until September 1997, 1,053,000 million legal immigrants and 327,000 childless adults ceased to participate in the Food Stamp Program. Together, these two groups account for 20 percent of the March 1994 through September 1997 decline. There may also have been a spill-over effect of the restrictions on legal immigrants. From October 1996 to September 1997, the number of U.S.-born children of legal immigrants receiving food stamps fell by 509,000, a drop of 41 percent, despite the fact that these children remained eligible. Had the number of these children fallen at the same rate as children of citizen parents, 370,000 more children of immigrants would have participated in September 1997. Employment Growth The sustained strength of the U.S. economy has created millions of new jobs and welfare reform has successfully moved many from welfare to work. As income rises, fewer people are eligible for benefits; and among those who remain eligible, some may feel less need to seek government assistance. One model of the relationship between unemployment and participation suggests that about 30 percent of the decline in food stamp participants between March 1994 and September 1997 (2,117,000 people) may be traced to the reduction in the number of unemployed over the 2 8/8' 200.00 380/SD H/HASA WH87:8 FAST . same period. It is not possible to separate the effects of a strong economy from the independent effects of welfare reform's success.¹ There is reason to celebrate the significant expansion of job opportunities among low-income households. There is also reason to be cautious. Because baseline levels of employment and earnings for welfare participants were so low, even substantial improvements leave most families exiting welfare poor. And, as yet, there are few signs of substantial improvement in the count of families and individuals officially poor. With the food stamp income limit set at 130 percent of the poverty line, many newly working families may continue to be eligible for food stamp benefits. Falling Participation Rates Not all persons eligible for benefits actually participate in the program. Falling participation rates may be symptomatic of barriers to participation erected by some States or localities² or a lack of awareness of continued eligibility by some. Preliminary evidence suggests that participation rates among those who are neither non-citizens nor unemployed childless adults fell significantly over this period (from 82 percent in 1994 to 69 percent in 1997), accounting for 32 percent (or 2,198,000 people) of the caseload decline. Table 2 shows that the drop in participation rates was sharpest among households with children and more than one adult. The Characteristics of Leavers As Table 3 shows, three major groups accounted for almost all the drop in food stamp caseload between 1994 and 1997. The number of legal immigrants fell by 54 percent, accounting for 14 percent of the total decline. The number of childless unemployed adults fell by 44 percent, accounting for 8 percent of the total decline. Finally, the number of persons in families receiving welfare benefits fell by 28 percent. Because they make up such a large share of the food stamp caseload, they account for 61 percent of the overall decline. Thus, the steepest declines in food stamp participation occurred among legal immigrants and unemployed childless adults, the two groups whose eligibility for food stamps was directly curtailed by the 1996 welfare reform law. However, most of the reduction occurred among the large group of families receiving cash assistance. I This estimate is based on a simple forecast model relating the number of unemployed in the current and lagged quarter to the number of participants with a set of seasonal dummy variables. While not developed to explain historical participation levels, it suggests the magnitude of the relationship between changes in unemployment and changes in program participation. This estimate is consistent with Wallace and Blank (1999) who suggest that declining unemployment rates can explain 28 to 44 percent of the food stamp caseload decline through 1996. 2 During the past year, FNS regional offices have reviewed access to programs in New York, Oregon, and Wisconsin in response to complaints from advocacy groups and a member of Congress. Each review has identified barriers that make the Food Stamp Program less accessible to cligible people. 3 P.4/8 20 ON 360/S0 H/HASO WH67:8 6661 Table 2 Participation Rates Among Those in Households with No Non-citizens or Unemployed Childless Adults Number of Eligibles Participation Rate Individuals in: 1994 1997 1994 1997 Change Households with single parents 12,995,000 12,118,000 95.8 86.2 -9.6 with earnings 4,166,000 4,893,000 63.0 67.7 4.7 without earnings 8,829,000 7,225,000 111.3 98.7 -12.6 Other households with children 7,550,000 6,887,000 96.3 74.8 -21.5 with earnings 4,667,000 4,460,000 77.2 64.0 -13.1 without earnings 2,883,000 2,427,000 127.3 94.6 -32.6 Households with elderly members and no children 5,274,000 5,261,000 30.6 27.9 -2.7 Other households 2,681,000 3,222,000 70.2 57.3 -13.0 All Individuals 28,500,000 27,489,000 81.5 68.9 -12.6 Source: Food Stamp Program Quality Control (1994 and 1997); Current Population Survey (March 1995 and March 1998). Table 3 Composition of Food Stamp Caseload: 1994 - 1997 Number of Participants Summer Summer Change Percent Share of 1994 1997 Change Decline Legal permanent residents 1,537,000 706,000 -831,000 -54% 14% Childless unemployed adults 1,148,000 648,000 -500,000 -44% 8 % AFDC/TANF participants 13,052,000 9,442,000 -3,610,000 -28% 61% All other participants 11,697,000 10,707,000 -990,000 -8% 17% All participants 27,434,000 21,503,000 -5,931,000 -22% 100 % Source: Reproduced from Table 1 in Who is Leaving the Food Stamp Program? An Analysis of Caseload Changes from 1994 to 1997, Food and Nutrition Service, USDA, March 1999. 4 8/S'd NO.024 360/SD3/60S0 WHAT:8 6661 - - DRAFT- - P.6/8 Explaining the Drop in Food Stamp Participation: March 1994 - September 1997 Factor Effect Contribution Source/Derivation (000's) NO.024 FNS National Data Bank (number of participants): Participation Change 6,974 100% March 1994 27,965 September 1997 20,991 Change 6,974 Characteristics of Food Stamp Households: Fiscal Year 1997, Tables B-1, B-3: Welfare Reform Permanent Resident Unemployed Childless Limits on: Aliens Adults Immigrants 1,053 15 September 1996 1,600 989 September 1997 547 662 Unemployed 327 5 Change 1,053 327 Adults w/o children Derived from FNS forecast model (based on current and lagged quarterly number of seasonally-unadjusted unemployed and seasonal dummy variables). Employment Growth 2,117 30 Model implies a net change of 83 participants for every 100-person change in Due to Economic the number unemployed: MAY.25.1999 8:50AM USDA/FCS/OAE Strength and Number Unemployed Welfare Reform 1994:Q2 9,209 1997:Q4 6,659 Change 2,550 X .83 = 2,117 Note: Wallace and Blank (1999) suggest that declining unemployment rates can explain about 28 to 44 percent of the food stamp caseload decline through 1996. Working Draft: Do Not Attribute, Cite or Publish P.7/8 Approximates change in participation decisions of persons eligible for benefits independent of economic improvements. Assumes improved economy (1) reduces Participation 2,198 32 the number of persons eligible for benefits and (2) reduces the perceived need NO.024 Reductions among for assistance for some, reducing participation among otherwise eligible people. Eligibles Independent of Derived from preliminary CPS-based estimates of participation rates for Economic Growth individuals in households with neither noncitizens nor unemployed adults without children (since participation effects of welfare reform restrictions on these two groups are captured earlier): Eligible Participating Participation Rate August 1994 28,500 23,223 81.5 September 1997 27,489 18,912 68.8 Change 1,011 4,311 12.7 Reduction in participation due to falling number of eligibles (arguably due to improved economy): 1,011 X .815 = 824 Reduction in participation due to falling participation rates (some of which may be due to improved economy): 27,489 X 127 = 3,491 8:50AM USDA/FCS/OAE Total reduction (equal to 824 + 3,491) 4,315 Previously assigned to economy 2,117 Net reduction 2,198 All Other/Unknown 1,279 18 MAY.25.1999 May 19, 1999 DRAFT TABLE I P.8/8 FSP PARTICIPATION RATES IN 1994 AND 1997 FOR SELECTED SUBGROUPS 1994 1997 Change Participation Purticipation Participation Eligibles Participants Rate Eligibles Participants Rate Eligibility Participation Rate NO. NO.024 Individuals in (000s) (000s) (percen!) (000s) (000x) (percent) (percent) (percent) (points) Households with permanent resident aliens 6.373 3.072 48.2 1,167 968 82.9 -81.7 -68.5 34.7 ) louseholds with ABAWDs¹ 1,680 1,220 72.6 649 582 89.7 -61.4 -52.3 17.1 Houscholds with single parents² 12,995 12,455 95.8 12,118 10,446 86.2 -6.7 -16.1 -9.6 with earnings 4,166 2,624 63.0 4,893 3,313 67.7 17.5 26.3 47 with no earnings 8,829 9,831 111.3 7,225 7,133 98.7 -18.2 -27.4 -12.6 Other households with children² 7.550 7,271 96.3 6,887 5,154 74.8 -X.R -29.1 -21.5 with carnings 4.667 3.601 77.2 4,460 2.856 64.0 -4.4 -20.7 -13.1 with no earnings 2.883 3,669 127.3 2,427 2,297 94.6 -15.8 -37.4 -12.6 Households with elderly and no children 5.274 1.615 30.6 5,261 1,468 27.9 -0.2 -9.1 -27 Other households 2,681 1.8M3 70.2 3,222 1.846 57.3 20.2 -2.0 -13.0 Total Individuals 36,553 27.515 75.3 29,305 20,462 69.8 -19.8 -25 n -5 4 Sources: 1994 and 1997 IQCS Datalites. 1995 and 1998 CPS Datailes No permanent resident aliens 'N₁ ABAWDs or permanent resident aliens USDH/F US/UHE МНИС:Р CD. 1999 J. Eric Gould 04/05/99 06:01:54 PM Record Type: Record To: Cynthia A. Rice/OPD/EOP CC: Subject: FS Caseloads I put together a group from OMB, CEA and USDA to discuss FS caseloads and we reconvened today for our second meeting. I wanted to focus on 1) comparing take up rates over recent years; and 2) economic data that could help explain the caseload anomaly. I have people more focused now but we are creating questions faster than we are coming up with answers. 1) Take up rates. March CPS data is in (covers FY97 caseloads). USDA is really concerned that the numbers are really screwed up since it shows that there are 6 million less eligibles than in FY96. The good news is that if the numbers are correct than the take-up rate is actually going up a little from 68 % in FY96 to 70% in FY97. But at this point that type of drop in eligibility is very difficult to explain since poverty levels didn't change to near that extent. The #1 assignment for the week is to try to get a handle on the accuracy of the new CPS data. USDA said that they would provide the resources to do so. The data indicated that of the 6 million decrease in eligibles, there were 3 million fewer eligible legal immigrants. Again, this was much higher than expected and might be attributable to a) bad data; or b) underestimation of the # of eligible LI to begin with, which is possible because the old methodology (pre 1994) was fairly poor. Because the estimate on eligibles is still so shaky we didn't spend a lot of time discussing what could have caused a sharp decrease in those numbers but not reflect increased income through poverty stats. A preliminary response was that these low-income folks could be doing a better job collecting assets, like a car. 2) Economic data. The effect of unemployment rates on caseloads needs to be completely reexamined. The basis for these models is 25 years old. Currently, we have no way to measure the effect of prolonged economic expansion and the added effect of low-unemployment pushing up wage scales. We are at a point were the bottom of the income bracket is doing better than they have ever done before but we are not sure how to measure what that would do to FS caseloads. Bob and Becky are looking at this question and they are talking to outside economists about it but they claim that this is a complicated question that makes a quick answer difficult. So for next week we will examine: 1) the believability of the eligible (specifically legal immigrants) data from March, 1998 CPS; and 2) what we can say or discern from changes in the wage structure for FS eligibles and the effects of very low unemployment. From the info. we have right now, it's possible that the same % of folks receiving FSs is staying steady and even going up a little but that the number of those eligible is taking a big whack. MINTHIA FAX COVER SHEET OFFICE OF ANALYSIS AND EVALUATION FOOD AND CONSUMER SERVICE, USDA (Office: 703-305-2019) (Fax: 703-305-2576) DATE: 3-29-99 TO: EMC Gould OFFICE: PHONE #: FAX PHONE #: FROM: Snova CARLSON PHONE: # PAGES (Including cover page): COMMENTS: FYI THIS was PULLISO TO60THX R LAST WOOK FUK KEASOUS UNKORAND TO THE OFFICE TO ORPLAIN No FSP BLOP, BUT IT HAS sumis VALUE AS BACKGROUND a P.1/2 NO.134 USDA/FCS/OAE W802:21 6661'62'860 March 30, 1999 NOTE FOR ERIC GOULD FROM: Steven Carlson SUBJECT: Food Stamp Participation Decline FYI, I tried to implement the framework that I think we've all agreed to pursue to quantify the factors causing the decline in food stamp participants (attached). The basic structure is as follows: (1) Calculate the drop in participation for a particular period (March 1994 to September 1997 in the attached table; March is selected as the peak, September as the last month for which we have detailed characteristic information). (2) Estimate the number of persons made ineligible by welfare reform restrictions (chiefly on immigrants and unemployed childless adults). These figures are drawn directly from two tables in the recently released Characteristics of Food Stamp Households: Fiscal Year 1997; they differ somewhat from the figures sent last week since they are the observed differences in the number of these participants between the two selected months (rather than fiscal year or quarterly differences). (3) Estimate the effect of the economy. This step is, I believe, poorly estimated here, but hopefully might be improved with the work you and Bob are undertaking. The estimate presented is derived from an internal Agency model of food stamp participation. It is sparsely specified, consisting of measures of the number of unemployed people (not seasonally adjusted) in the current and a lagged quarter plus a set of quarterly dummy variables. The model is estimated from a quarterly time series dating from 1977. From the model forecast, we derived an estimate of the effect of changes in unemployment on food stamp participation (83 participants for every 100-person change in the number unemployed). There are few reasons to be confident in this particular adjustment. The forecast model did not perform particularly well during the run-up in food stamp participation during the late 1980s and early 1990s, nor has it performed well during the more recent decline. The model is sparsely specified, providing a meager proxy for economic changes and failing to account for significant program changes. I've included it here only for the purpose of illustration. (4) Net out effect of eligibility restrictions and economy, leaving an unexplained residual. Call with questions on (703) 305-2134. Attachment P.2/3 NO. 161 USDA/FCS/OAE 3:44PM MAR. 30. 1999 Preliminary Decomposition of Food Stamp Participation Change Permanent Resident Unemployed All Participants Aliens Childless Adults March 1994 27,965 1,501 1,320 September 1997 20,991 -6,974 547 -954 662 -658 December 1998 18,618 -9,347 n/a n/a n/a n/a # Unemployed 1994 Q1 9,209 1997 Q4 6,659 -2,550 X .83 = -2,117 March - Sep Drop -6,974 Immigrants -954 0.14 ABAWDS -658 0.09 Subtotal Unemployment -2,117 0.30 Unknown -3,246 0.47 1.00 Notes: (1) Estimates of permanent resident aliens and unemployed childless adults from Characteristics of Food Stamp Households, Fiscal Year 1997, Tables B-1 and B-3. 1pt UR: (pt fs/ (2) Estimate of effect of unemployment on food stamp participation derived from FNS forecast model (based on quarterly number of unemployed (seasonally unadjusted and set of quarterly dummy variables). Model implies that a net change of 83 participants for every 100-person change in number unemployed. ! Participation rate of eligibles Chepser 3 March 30, 1999 2. # g total eligibles, and tabe up- 3. Laged effects of wage VS. UR, P.3/3 191.19 USDA/FCS/OAE 3:44PM 1999 "0E MARK Food and Nutrition Service US Department of Agriculture National Data Bank V 5.0 March-94 December-98 3/23/99 State / FS PROGRAM FS PROGRAM Difference Territory PART ACTUAL PART ACTUAL Dec 98 Mar 94 ALABAMA 561,803 415,270 -146,533 ALASKA 51,692 15,680 -36,012 ARIZONA 512,288 259,935 -252,353 ARKANSAS 290,212 256,378 -33,834 CALIFORNIA 3,191,890 2,062,101 -1,129,789 COLORADO 275,765 180,083 -95,682 CONNECTICUT 225,044 183,118 -41,926 DELAWARE 61,717 42,482 -19,235 DISTRICT OF COL $2,427 85,344 -7,083 FLORIDA 1,469,271 967,349 -501,922 GEORGIA 826,295 633,555 -192,740 GUAM 15,058 18,895 3,837 HAWAII 115,461 126,249 10,788 IDAHO 88,094 58,411 -29,683 ILLINOIS 1,207,279 840,288 -366,993 INDIANA 543,248 301,485 -241,763 IOWA 203,883 128,493 -75,390 KANSAS 197,340 113,104 -84,236 KENTUCKY 536,337 396,813 -139,524 LOUISIANA 763,280 531,415 -231,865 MAINE 141,742 109,656 -32,086 MARYLAND 399,877 295,722 -104,155 MASSACHUSETTS 451,344 274,751 -176,593 MICHIGAN 1,047,450 723,807 -323,643 MINNESOTA 323,939 207,606 -116,331 MISSISSIPPI 519,006 295,042 -223,964 MISSOURI 606,179 408,118 -198,061 MONTANA 74,280 61,069 -13,211 NEBRASKA 113,484 98,019 -15,465 NEVADA $8,977 64,777 -34,200 NEW HAMPSHIRE 63,760 37,931 -25,829 NEW JERSEY 554,501 394,452 -160,049 NEW MEXICO 251,655 182,559 -69,096 NEW YORK 2,175,154 1,577,854 -597,300 NORTH CAROLINA 646,230 498,053 -148,177 NORTH DAKOTA 47,459 33,654 -13,805 OHIO 1,269,050 661,747 -607,303 OKLAHOMA 382,926 280,798 -102,128 OREGON 298,653 224,513 -74,140 PENNSYLVANIA 1,243,533 842,332 -401,201 RHODE ISLAND 96,336 76,913 -19,423 SOUTH CAROLINA 390,940 320,011 -70,929 SOUTH DAKOTA 55,200 43,529 -11,671 TENNESSEE 747,371 520,315 -227,056 TEXAS 2,764,395 1,460,671 -1,303,724 UTAH 131,124 90,222 -40,902 VERMONT 65,475 44,953 -20,522 VIRGINIA 571,254 370,069 -201,185 VIRGIN ISLANDS 19,632 17,062 -2,570 WASHINGTON 483,004 326,240 -156,764 WEST VIRGINIA 333,260 250,995 -82,265 WISCONSIN 334,900 183,752 -151,148 WYOMING 34,698 24,037 -10,661 United States 27,965,172 18,617,677 -9,347,495 2/2'd NO.134 USDA/FCS/OAE W802:21 6661.62.86W CFA Sheet1 Number of Participants Unemployed Permanent Childless Adult Participants Resident Aliens (ABAWDS) Mar-94 27,965 1,501 1,320 Sep-97 20,991 547 662 Dec-98 18,618 n/a n/a Change from 3/94 to 9/97 Participants: 3/97 27965 C17 (1) Drop due to Aliens -954 C18 (2) Drop due to ABAWDS -658 C19 (3) Drop due to economy -2530 (1)-(3) -4142 Actual decline -6974 % of change explained by: Aliens 13.7% ABAWDS 9.4% Economic conditions 36.3% All other factors 40.6% Other potential factors: 1. Reform of AFDC in 1996 (C17+C18+C19) (65-4.9)*(0.06) Page 1 J. Eric Gould 03/31/99 02:30:47 PM Record Type: Record To: Cynthia A. Rice/OPD/EOP CC: Subject: Food Stamp Caseloads Our research group is making a little headway. Unfortunately, 40-45 percent of the caseload reduction can not be explained through any modeling techniques available. A problem is that we can't just attribute it to "whatever welfare reform, be it good or bad, is doing" to the caseload. There are unknown possible economic effects that could have significant influence but we are not sure. These effects caused by "great prolonged economic conditions" are hotly debated in economic circles right now and also at CEA. Becky is one of the leading proponents of the theory that prolonged economic success creates affects among low-income individuals that we are not yet able to measure. For example, the standard economic model to predict FS caseloads is the unemployment rate (for every 1% increase in unempl. the FS caseload is supposed to increase 6%) but in a tight labor market the unemplo. rate has gone so low that wages get pushed up and have a different but powerful effect on FS caseloads. We are also looking at the changes in the income levels of all FS eligibles (below 130% poverty) - which should be going up meaning more people might be eligible for less benefit. We are also looking at the decline in the total amount of eligibles over time (the group of below 130% is getting smaller). We are trying to make USDA get updated numbers on the participation rate of eligibles, armed with that and the total number of eligibles we can work out take-up rates and compare them over time. J. Eric Gould 03/26/99 06:27:28 PM Record Type: Record To: Cynthia A. Rice/OPD/EOP CC: Subject: Food Stamp Chart I had a long conversation with USDA about the chart (Table 2) you and I discussed this morning. The chart reviews data on single parent households by presence of earnings and welfare receipt between 1994-1997. From my conversation with USDA I gathered: 1. The number of single parent households receiving TANF fell, which is consistent with overall declines in welfare caseloads. 2. The number of single parent households not receiving TANF increased modestly, which is consistent with the notion that as folks leave welfare, some continue to receive food stamps as NPA cases. 3. The number of single parent households not receiving TANF who have earnings increased slightly, which is consistent with recent efforts to move single parent families off of welfare and into work. Conclusions from the data: 1. The data show that some households appear to leave welfare and food stamps completely. We expect that some had higher earnings. However, the data does not provide any information about these households. 2. Others find work and leave welfare, but continue to combine work with food stamps as a nutritional supplement, which is how the Program is supposed to work. One thing to keep in mind is that these numbers represent people who continue participation and people who come on to the program. For example, when the data reflects a single mom with earnings and no TANF, we don't know if she got TANF and food stamps before and is now a food stamp only recipient, or is she never got TANF and came to the State agency as an NPA household. This means that when the chart shows an increase in single parents with earnings and no TANF, we don't know how many of them worked their way off of TANF. Similarly, when there is no TANF and no earnings, these could be people who were kicked off TANF or who never received it in the first place. What is Causing Caseload Reductions in the Food Stamp Program ? 3/10 In November 1998, the Food Stamp Program served 18.6 million people. The number of people receiving food stamps fell by over 9 million people, a drop of one third, since March 1994. Part of this drop is due to the strength of the economy and the success of welfare reform. Part of the drop is due to new restrictions on the participation of certain legal immigrants and able-bodied unemployed adults without dependents. However, other factors may also be at work. Between 1995 and 1997, Food Stamp Program participation fell five times as fast as poverty, suggesting that many poor families have left the program despite their continuing eligibility. While the program is available to households with gross incomes of up to 130 percent of poverty, 91 percent of program participants have income below the poverty level. There are three important factors that must be considered in addressing the reasons for the sharp decline in Food Stamp participation since 1994: 1) specific Food Stamp eligibility changes in PRWORA; 2) broader PRWORA changes requiring work and discouraging dependence on cash assistance; and 3) a strong economy. The relation of these three factors is a topic of continuing research but preliminary information shows that economic factors can not fully explain recent caseload declines and further research is needed to look more closely at behavioral changes in take-up as well as state-specific changes in eligibility. How ave these different -97? 1. Who is Leaving the Food Stamp Program? Subgroup Participants Participants Participation Percent Subgroup as Subgroup as Percent of Decli in 1994 in 1997 Change 1994- Change in Percent of Percent of from Subgroup (thousands) (thousands) 1997 Participation 1994 1997 caseload 1994-1997 (thousands) caseload Legal 1,537 706 -831 -54% 6% 3.4% 14% Permanent Resident Aliens Childless 1,148 648 -500 -43.5% 4% 3.1% 8.4% bet Unemployed Adults AFDC/TANF 13,052 9,442 -3,610 -27.6% 48% 43% 60.8% Recipients All Other 11,697 10,707 -990 -8.5% 43% 49% 16.7% Participants Total 27,434 21,503 -5,931 -21.6% 100% 100% 100% Participants July-September Food Stamp Program Quality Control Data for 1994 and 1997 Narrow columns 50thenal FH The above chart shows that while there were large reductions within the legal immigrant and enpoye childless unemployed adult populations, their combined effect only contributes to less than a quarter of the entire food stamp caseload reductions. It must also be noted that the above percentage changes also reflect economic influences. Therefore, even without policy changes the subgroups would still have reflected decreased caseloads. The bulk of the caseload changes comes from AFDC/TANF recipients, who make up approximately 43 percent of the Food Stamp caseload. Between 1994-1997, this group experienced a 27 percent decrease in participation, which resulted in 60 percent of the entire caseload reduction between 1994-1997. The group identified as "All Other Participants" consists of working poor, aged poor and disabled poor. This group is the second largest segment of the Food Stamp population, making up 43 percent of the entire caseload in 1997. (alth immigo most Feable 2. How are specific changes in PRWORA affecting the Food Stamp Program? Pium (Lb/ At least 20 percent of the decline between 1994-1997 came from two subgroups directly affected by PRWORA. Participation of legal immigrants declined 54 percent between 1994-1997 and reflected 14 percent of the entire caseload's decline. Changes in BBA and last year's Agricultural Research Act will may have án affect on these numbers as certain groups of legal immigrants become eligible for Food Stamps. The group of childless unemployed adults without dependents made up 8.5 percent of the decline in the entire Food Stamp caseload, as the subgroup registered a 44 percent decline between 1994-1997. Remind people of policy eg, 18to sos only eligible for smonths of wratthnu unless working Adding the percentage caseload declines from these two subgroups suggests that 22.5 percent of 220hpm the entire caseload reductions between 1994-1997 were attributable to changes in eligibility for these populations. However, economic factors also have to be taken into consideration. Experts' estimates of the economy's impact on the Food Stamp caseload during this period varies between 25-45 percent. Taking into consideration economic forces, the direct impact of PRWORA on these two groups and their relation to their total impact on Food Stamp caseload declines has been estimated to be approximately 10 percent. 3. What affect do the broader changes in PRWORA have on Food Stamp caseload changes? Regardless of the impact that specific policy changes in PRWORA have had on Food Stamp caseloads, the bulk of the decline between 1994 and 1997 came from AFDC/TANF recipients, who as a group attribute for 60 percent of the decline The number of AFDC/TANF recipients has fallen by over 1-million since 199 and the vast majority of the AFDC/TANF recipients were also receiving Food Stamps Historifally The Food Stamp Program has been closely coordinated in local offices with cash assistance programs.nd this can dramatically affect the household' assessment of the costs and benefits of participation. chrifort, individuals receiving cash ass stance nave always been fai more likely to recewe Food Stamps tram other eligible endi ideals PERCENTAGE OF are are these they all non-earners! PARTICIPATION OF FOOD STAMP ELIGIBLE INDIVIDUALSUN THE FOOD STAMP PROGRAM Household with earnings 45.8% Add you Identy in get fn stat non- Households with children 85.8% Single parents with children 96.8% what of earners is rate not on AFDC? Households with earnings and children 55.2% Single parents with earnings/children 59% 1994, SIPP public arrythnce These numbers indicate that families with a single adult, who were receiving are almost certain to participate, while households with earnings have a participation rate under 50 percent. At some level, the presence of children has a positive influence on participation even among the subset that works and being a single parent raises the likelihood of participation further. As the TANF roles decrease and individuals continue.to obtain more earnings, these numbers indicate that they are less likely to retain their Food Stamps even though they are eligible. Although participation rates among single parents and their children have been historically high, the recent declines in TANF caseloads suggest they might fall in the future. For example, in Indiana, the state's welfare reforms have reduced Food Stamp participation without a corresponding reduction in the number of potentially-eligible families. In short, some families that leave or avoid TANF and remain in poverty fail to apply for Food Stamp benefits. Knowing that the number of households with earning is increasing and that they are less likely to retain their Food Stamps, even though they may still be eligible, an important issue is the relationship between the percentage of working eligibles who obtain Food Stamps as compared to the present percentage. The trend has been that an increasing amount of Food Stamp participating households have some earnings. From 1994-1997 almost three percent more participating households reported earnings. But M people are 60 likely toget FS orce they have earnings, wfat los this mem? PERCENTAGE OF FOOD STAMP PARTICIPATING HOUSEHOLDS WITH EARNINGS 1992 20.2% would be more useful to 1993 20.6% compane I Aolto This to include all leavers 1994 21.4% FS studies (surveys guilat 1995 21.4% eligibles % of resports leasing To all 1996 22.5% go FST Medicaed) see 1997 24.2% deveropm s> 19 email from Andria 1997 USDA Food Stamp Quality Control Samples Do some calcula tions based on these data and the leavers studies 2. whechshow bet 50-60% of those leaving the rolls go to work i.e., if bet 28-45 90 left due to economy, courlte rest of the decline from 194- 197 be attributable to foid that 55% of Those who left are now only talf as likely to get FS forwhechtley we elear!? 4. What factors contribute to nonparticipation? USDA has studied the reasons for nonparticipation and identified three common factors that are likely to apply: Insufficient or incorrect information about the program. Working-poor families with children often do not know about their eligibility for Food Stamps given that many such families are not eligible for cash assistance. Lack of accurate program information is also prevalent among illiterate and non-English speaking individuals. toget to office Problems of Program access and administrative Influialty difficulties with the application, during rabdtothis working open hours/ process. Historically, participation rates are often lower among rural and elderly incentives error populations because of physical access and transportation barriers. However, recent rates), for reports, similar to the illegal NYC access situation, have created questions whether wide- states than spread diversion or discouragement of Food Stamp participation is taking place on a these fam broader level. Give an ex= some one w/ su/hr Small size of the benefit or lack of desire for benefits. As an individual's or 106 t 2 cheldren is eleaptle forxx FS in household's income rises the potential Food Stamp benefit declines. At some point, program eligibles decide the small amount of the benefit is worth less than the cost of applying. Other people forgo benefits regardless of the amount because of personal objections to income transfer programs and / or the stigma associated with receiving or using Food Stamps. The increasing use of EBT could decrease this stigma in the future as the transaction will be similar to credit card use. 5. How does the decline in AFDC/TANF participation compare to declines in food stamp participation? Make this CHANGE IN WELFARE / FOOD STAMP CASELOADS SINCE ENACTMENT OF PRWORA Change in AFDC/TANF Percent Change in Change in Food Stamp Percent Change AFDC/TANF #2 Participation Participation in Food Stamps Aug. 96 - Sep. 98 Aug. 96 - Sep. 98 Aug. 96 - Sep. 98 Aug. 96 - Sep. 98 4,286,000 -35% 6,229,107 -25% Comparing data between 1980 through 1998, Food Stamp caseloads are usually about twice as high as AFDC/TANF caseloads. This reflects the broader eligibility rules for Food Stamps. AFDC caseloads which were largely flat from the mid-1970s through 1990 rose 27 percent between 1990 and 1994. But, between 1994 and mid-1998 they fell by 40 percent. In September, 1998, they were at their lowest since 1969. Food stamp caseloads follow a similar trend. Participation peaked at 22 million during the 1983 recession and then fell steadily with a stronger economy until 1988. From 1988 to 1994, participation increased by 9.5 million, a 52 percent increase. From 1994 to February 1998, Let's discus what best compromounsibe -1994-1997 ? 1996-1997? 96-98? Food Stamp participation fell by 7.8 million, a 28 percent decrease. 6. Is the economy operating through the strong labor market responsible for the decline? The relationship between the economy and Food Stamp participation has been fairly strong over the last twenty years. Even when forecasters have not been able to predict AFDC caseloads accurately on the basis on economic factors, there are still corresponding patterns of rise and fall in unemployment and the Food Stamp caseload. But even though economic factors are important, they are not the entire story. There have been periods where Food Stamp caseload trends varied significantly from unemployment rates. The unemployment rate has fallen from 6.6 percent in 1994 to 4.4 percent in 1999. Previously developed models of the relationship between the economy and Food Stamp participation indicate that economic trends since 1994 could account for anywhere from 28 to 45 percent of the change in Food Stamp participation. With the caseload declining by almost eight million individuals since 1994, this 17 percent difference (difference between 45 percent and 28 percent) is significant in understanding what happened to approximately 1.4 million individuals. Recent research done by Rebecca Blank suggests that a one-point rise in the unemployment rate will produce a 6 percent increase in the food stamp caseload over a three year period. This might be an understatement if a lack of TANF-funded assistance produces greater need among low-income families in the future. On the other hand, it may be an overestimation if smaller numbers of people in TANF-funded programs mean that fewer families are aware of their food stamp eligibility and take-up rates fall. Still, economic data that takes into consideration unemployment and median wages, can only account for 40 percent of the caseload changes that have taken place between 1994 and 1998. F52 What are we doing about this? We will continue to monitor caseload trends and keep you informed. Moreover, there are important questions we need to further examine with the assistance of USDA and CEA. If the sharp decline in Food Stamp participation can not be fully explained through economic forces, what relationship does the decreasing TANF caseload have on it and can we expect both caseloads to continue to decline? How much influence do the common factors of non-participation have on program eligibles and how can we address these specific factors? What can we do through Administrative action to address this situation? on the work departmy What is Causing Caseload Reductions in the Food Stamp Program (ESP)? Food stamp digfility There are three important factors Sta that must be considered in addressing the plausible reasons for the sharp decline in FST participation since 1994: 1) specific FOP changes in PRWORA; 2) broader PRWORA changes to the cash system; and 3) a strong economy. The relation of these three factors is a topic of continuing research and debate. This memo addresses current easeload trends and research but proliminary mathon shows 1. Who is Leaving the Food Stamp Program? 1994 1997 Subgroup Participants in Participants Participation Percent Change Subgroup Percent of Decline 1994 in 1997 Change 1994- in Participation as Percent from Subgroup (thousands) (thousands) (thousands) 1997 associal caselord of 1997 Legal 1,537 706 -831- -54% 3.4% 14% Permanent 6.0% Resident Aliens Childless 1,148 648 -500 -43.5% 3.1% 8.4% Unemployed 4.0% Adults aye18.50 AFDC/TANF 13,052 9,442 -3,610 -27.6% 48% 43% 60.8% Recipients All Other 11,697 10,707 -990 -8.5% 43% 49% 16.7% Participants how Total 27,434 21,503 -5,931 21.6% 100% 100% 100% manyol Participants These July-September Food Stamp Program Quality Control Data for 1994 and 1997 towmany work 2. How does the decline in AFDC/TANF participation compare to declines in food stamp participation? update & Decline in AFDC/TANF Percent Changen AFOC/TAIN Decline in Food Stamp Percent Change intiod Stamps Sept " chu Cha nge Participation 1998 March 1994 to June 1998 Participation March 1994 to June 1998 March 1994 - June 1994 March 1994 - June 1998 -5,957,854 -41.8% -8,953,651 -31.4% Comparing data between 1980 through 1998, food stamp caseloads are usually about twice as high as AFDC/TANF caseloads. This reflects the broader eligibility rules in the FSP AFDC caseloads which were largely flat from the mid-1970s through 1990 rose 27 percent between 1990 and 1994. But, between 1994 and mid-1998 they fell by 40 percent. In June, 1998, they were at their lowest since 1969 Spt Food stamp caseloads follow a similar trend. They decline slightly faster than AFDC caseloads what are As? her about a araph ! your but noclat above ? PRWORA gote need # in the mid-1990s, in part because ofPRWORA Still, the rapidity of these hanges is unprecedented and hard to explain. with Bortne Food stamp elizablety Food stump gulat 3. Did the 1996 FSP changes in PRWORA contribute to the decline in FSP participation? 0\0 stamp other These changes imposed a strict time limit on able-bodied adults without dependents. They made X most legal aliens ineligible for the ESP (even though much of these cuts have been restored), "D' slightly reduced the value of the maximum benefit, and limited certain deductions. With little hard data available, there is no way to know for sure what affect these changes actually have had on FSP participation. Some experts agree that these changes would have affected easeloads no more than 10 percent (Beebout, Mathematica 1998). Think This Dostino mean 109. of dertne bowe Johnny the Throughanges would study is good 4. How do the non-FSP changes in PRWORA relate to FSP participation? through # appeople reversity The number of AFDC/TANF recipients has fallen by over 4 million since 1994, and the vast banfts. 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 how 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. ?? Jowe Vor 5. Is the economy operating through the strong labor market truly responsible for the This decline? From 1994 to February 1998, FSP participation fell by 7.8 million, a 28 percent decrease. In comparison 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 of the decline in participation percent (Blank, CEA 1998; Beebout 1998; Hamilton, Abt Associates 1998). meaning AFDC The relationship between the economy and FSP caseload has been fairly strong over the last 20 years. Even when forecasters have not been able to predict caseloads accurately on the basis on economic factors, there are still corresponding patterns of rise and fall in unemployment and the FSP caseload. However, although economic factors are important, they are not the entire story. There have been periods where FSP caseload trends took off from unemployment rates. ?? - Simply explained, there are other broader factors that help explain how many people participate in the FSP: the size and composition of the population and the eligibility rules in place interact with economic factors. 6 Still, what do economic models tell us about the future of food stamp participation rates? Recent research done by Rebecca Blank (CEA) suggests that a one-point rise in the unemployment rate will produce a 6 percent increase in the food stamp caseload over a three year period. This might be an understatement if a lack of TANF-funded assistance produces greater need among low-income families in the future. On the other hand, it may be an overestimation if smaller numbers of people in TANF-funded programs mean that fewer families are aware of their food stamp eligibility and take-up rates fall. Instead 7. What "links" exist between food stamps and TANF participation? Participation in AFDC/TANF plays a special role in food stamp participation patterns. The FSP 90 wk. has been closely coordinated in local offices with cash assistance programs. This can dramatically affect the household's assessment of the costs and benefits of participation. dezplar Participating in both programs has only marginally higher costs than participating in one, and as obtains a result, FSP participation is strongly associated with cash assistance participation. The vast (and majority of single-parent households who receive cash assistance also receive food stamps, but currently food stamp participation is much less common among single-parent households who do not receive cash assistance even when households are grouped by income levels. The result is that, what for a substantial part of the food stamp caseload, food stamp participation closely parallels to 1846 AFDC/TANF participation. Therefore, recent reports that some states are diverting, delaying and terminating individuals from TANF, and sometimes food stamps, in violation of federal law puthis is potentially a critical issue in terms of better understanding participation rates. aspant Somewhat surprisingly, even though more than half of all food stamp cases do not receive cash of assistance, the national food stamp caseload has closely tracked the national AFDC/TANF caseload since the early 1980s. This may mean that, beyond the administrative linkage, both Q+A programs are influenced by some underlying pattern in households' assessments of the costs and benefits of participation. This might reflect economic factors, such as perceptions about job opportunities, or non-economic factors. 8. What conclusions can be drawn from all of this? Although it is clear that many factors are clearly correlated with food stamp caseload changes, they do not explain the recent trends well. The uncertainty over the causes of recent powerful declines in the caseload are leaving most economists scratching their heads and causing other researchers to look more closely at behavioral changes in take-up as well as state specific eligibility changes that might be influencing these changes. Some economists (Blank and Beebout) and experts (Greenstein) believe that recipients are strongly influenced by a host of less-measurable factors (including their own sense of the "accessibility" of utilizing public assistance) when deciding whether or not to participate in the FSP. 3/4/99 SUMMARY OF RECENT RESEARCH ON WELFARE LEAVERS Study: Percent of Leavers Percent of Leavers Receiving Food Stamps Receiving Medicaid Delaware - 1/98 65% 75% Indiana - 1997 38% 53% 11/98 52% (after Year 1) 66% offered health insurance by 40% (after Year 2) employer, but 50% declined for various reasons including cost (5% declined because of Medicaid coverage) Iowa - 1993-5 66% 66% 1999 59% South Carolina - 1997 66% 75% - 10/98 52% of adults had health insurance (40% of the total had medicaid and 12% had private insurance); 91% of the kids had health insurance (80% medicaid and 11% private) Washington - 1/99 45% 44% (Adults) 64% (Kids) Wisconsin - 1/99 49% 87% A2 SUNDAY. MARCH 7. 1999 S NATIONAL NEWS THE WASHINGTON Drop in Food Stamp Rolls Is Worrying Officials A Sharp Decline By JUDITH HAVEMANN mittee, said the declines are "quite The welfare application process food stamp usage has been pro- called for a review of the rules that The number of people who Washington Post Staff Writer mysterious" and need to be investi- traditionally has been the main en- pelled by a booming economy and govern food stamp and Medicaid eli- receive food stamps fell from a gated because they "far exceed ex- try point for food and medical assis- changes in the law that have made gibility, while Rep. Robert W. Goo- high of 27.5 million in 1994 to Advocates for the poor and some pectations." Johnson said that the tance, as well as cash. Since the 1996 about 700,000 legal immigrants inel- dlatte (R-Va.), chairman of the Agri- 19.8 million last year. members of Congress are alarmed availability of health care benefits is welfare reform law was passed, igible. The rest of the drop, they say, culture subcommittee that oversees by a steep drop in the number of "core" to whether welfare reform some states have been "diverting" is a result of states discouraging po- food stamps, said he would favor ef- Number of recipients, in millions Americans receiving food stamp as- succeeds and that it is unclear poor families from welfare tential applicants, intentionally or forts to turn over the food stamp 30 sistance and have begun pressing whether states are doing enough to inadvertently, because workers are program to the states in the same for changes to ensure that poor fami- inform poor families of their eligibil- The decline in so focused on moving them into manner as welfare if there is broad 28 lies have enough to eat. ity for assistance. She plans to hold jobs. support for such a move. Nearly 8 million Americans have hearings to draw attention to the Moreover, welfare reform has Douglas Besharov, a senior fellow left the food stamp rolls during the problem. food stamp usage 26 stigmatized all forms of aid, they at the American Enterprise Insti- four years, a 28 percent reduc- State welfare officials, mean- say, making many poor families re- tute, said the decline in Medicaid 24 tion that is far greater than the de- while, will converge on Washington and food stamp rolls is not fully un- 1994: today to ask Congress to abolish suggests a lot of luctant to apply for government help cline in poverty for roughly the same of any kind. Some Americans are derstood. But he said some of it rep- 27.5 million 22 period, according to new federal fig- rules that they say are an obstacle to simply unaware that they are poor ures. poor families left resents a common-sense decision by getting food stamps and are in- enough to qualify for food stamps if recipients who don't want to miss a 20 Agriculture Secretary Dan Glick- compatible with welfare reform. For they earn less than $1,783 a month day of work every six months to visit 1991: man said that suggests "that a lot of example, they say, it is absurd that a the program even to support a family of four. Also, welfare offices to qualify for a small 22.6 million 18 poor families left the program even low-income family with a car valued many former welfare recipients are allotment of food stamps. Moreover, though the need was still there." above $4,650 is almost automatical- not told they are entitled to at least though the need because poor people can sign up for six months of Medicaid benefits to 16 1998: 19.8 million The combination of plummeting ly ineligible for food stamps when Medicaid when they visit an emer- food stamp rolls and a similar, but some state welfare agencies are help them make the transition to gency room, they are less inclined to smaller decline in the number of helping welfare families obtain cars was still there." work, these advocates say. 14 go through the sign-up process people in Medicaid, the health insur- to get to work. Kinder, gentler' is a thing of the when they are not in immediate ance program for the poor, has fu- The state officials, members of -Dan Glickman, past," said Rep. Fortney "Pete" need of care. 0 eled concerns among advocates for the American Public Human Servic- secretary of agriculture Stark (D-Calif.). "Some states "There is no evidence of a signif- 1991 '92 '93 '94 '95 '96 '97 '98 the poor that the 1996 federal wel- es Association, also will urge the Ag- have been illegally telling people icant increase in the level of hard- SOURCE: Agriculture Department fare reform law may be discouraging riculture Department to reduce by making them search for jobs or they are no longer qualified." ship the poor are facing, and evi- THE WASHINGTON some families from seeking needed some of the program's red tape, in- giving them a onetime payment to Second Harvest, which operates dence, in fact, that the condition of aid. Private food pantries and chari- cluding requirements that states tide them over. Food advocacy the largest network of food banks in the poor has improved," Besharov ures are available. table organizations have reported continually summon recipients to groups say that welfare rules have al- the country, said that demand is up, said. The Medicaid rolls fell from 41.4 an increased demand for food assis- county welfare offices to recertify so "diverted" families from the food particularly among families, but said Food stamp usage fell from 27.5 million to 40.3 million during the tance. their eligibility. stamps and medical insurance that it was unable to compare the 26 mil- million in 1994 to 19.8 million last same period, even as federal and Rep. Nancy L. Johnson (R- But advocates for the poor charge the families need more than ever. lion people who sought help during year, while those living in poverty state health officials beat the bushes Conn.), chairman of the Ways and that the real problem may be the The groups cite studies showing 1997 with earlier years. fell from 38.1 million to 36.6 million to extend health insurance, to the Means human resources subcom- states themselves. that only about half of the decline in Rep. Robert T. Matsui (D-Calif.) in 1997, the last year for which fig- many qualified uninsured families. APHSA American Public Human Services Association Marva Livingston Hammons, President William Waldman, Executive Director March 2, 1999 Cynthia Rice Special Assistant to the President, Domestic Policy Council 1600 Pennsylvania Ave, NW 212R OEOB Washington, DC 20500 Dear Ms. Rice: The American Public Human Services Association (APHSA) invites you to come hear the Commissioners of three state human services agencies offer their perspectives on "Food Stamps, TANF and the Working Poor." The forum will be held on Tuesday, March 9, 1999 from 1:30 to 3:00 p.m. in the Senate Agriculture, Nutrition and Forestry Committee Room in 328A of the Russell Senate Building. Please see the attached brochure for more information. If you have any questions regarding this forum, please call Anna Lovejoy at APHSA, 202-682-0100. We look forward to your attendance at this important policy discussion. APHSA nonprofit, bi-partisan organization of individuals and agencies concerned with human services. Our mission is to develop, promote, and implement public human service policies that improve the health and well-being of families, children, and adults. Sincerely, Williuab William A. Waldman Executive Director Representing Public Human Services Since 1930 810 First Street, NE, Suite 500, Washington, DC 20002-4267 (202) 682-0100 fax: (202) 289-6555 http://www.aphsa.org Why the Food Stamp Program No Longer Meets the Needs of the Working Poor The strict regulations governing the Food Stamp Program. These two separate The Food Stamp Program also requires administration of the federal Food Stamp agencies periodically promulgate inconsistent families to report normally every three months Program pose serious obstacles to states' policies that increase the cost of administering for in-person interviews, immediately strategies to reduce dependence on welfare, the Food Stamp Program and impede state pass along nearly every variation in their increase the employment of Temporary welfare reform innovations. This was true under wages, and have their employers fill out a Assistance for Needy Families (TANF) clients, AFDC and, unfortunately, is increasingly the steady stream of paperwork verifying and meet the needs of the working poor. The case under TANF. their work status. These requirements are due clash between the TANF and Food Stamp to the historical focus on total and predictable programs is costly and complex. State Examples of Program Conflicts accuracy of payment levels, whereby administrators see the conflicts of the two individuals and states are both held programs on a daily basis because the TANF For example, under the state TANF programs, responsible for "errors" and face penalties for and Food Stamp programs are often clients are receiving assistance to purchase or inaccurate wage reporting. But the nature of administered by the same human service repair cars to get to work. However, under the the entry-level job market is that fluctuations agency charged with moving welfare clients Food Stamp Program rules, if a recipient owns a are frequent, and many jobs do not last very off assistance and into employment. This is car worth more than $4,650, the client is ineligible long. Moreover, as states become more not the case at the federal level where the U.S. for food stamps. Similarly, many states have effective in moving TANF clients into jobs, Department of Health and Human Services increased income and asset disregards they run the risk of increasing "errors" due to oversees the TANF block grant and the U.S. and established individual development accounts the uncertainty of the entry-level job market. Department of Agriculture administers the for TANF clients. These assets may make families While all state administrators hold the ineligible for food stamps. efficient and effective administration of food stamps as a goal, adjustments are needed to help support those food stamp clients who are making this effort, rather than burden them FOOD COUPON and penalize states. This forum will present the states' ideas for a more flexible and innovative approach to serving working families in the Food Stamp Program. State human service agency commissioners will share their experiences in facing these policy challenges and will offer their thoughts on what policy changes might help them serve families better. Food Stamps, TANF, and the Founded in 1930, the American Public Working Poor Human Services Association (APHSA) is a Food Stamps, nonprofit, bipartisan organization of state Tuesday, March 9, 1999 and local agencies concerned with human TANF, and the 1:30-3:00 p.m. services. Members include all state and 328A Russell Senate Building many territorial human service agencies, Working Poor more than 800 local agencies, and several Welcome and Introductions thousand individuals who work in or otherwise have an interest in human William Waldman, Executive Director, APHSA service programs. APHSA, which recently changed its name from the American Public Panel Discussion of State Welfare Association, serves as a conduit State human service commissioners call Commissioners between members of Congress, the media, for changes in the Food Stamp Program and the broader public on what is Gary K. Weeks, Director, Oregon to meet the needs of the working poor. happening in the states and localities Department of Human Resources Hill staff will have an opportunity to concerning welfare, child welfare, health question the panelists and hear state Clarence Carter, Commissioner, care reform, and other issues involving perspectives. Virginia Department of Social Services families, children, and the elderly. Lynda Fox, Secretary, Maryland Department of Human Resources The mission of APHSA is to develop, promote, and implement public human service policies Questions from Congressional that improve the health and well-being of Staff Panel families, children, and adults. Mike Ruffner, Majority Professional FOOD COUPON Staff Member, Senate Agriculture, Nutri- tion and Forestry, Subcommittee on Research, Nutrition, and General Legisla- tion Mark Halverson, Minority Chief Coun- cil, Senate Agriculture, Nutrition and For media inquiries, contact Forestry Committee Amy Tucci, Communications Kevin Kramp, Majority Staff Director, Director, at (202) 682-0100, Sponsored by the House Agriculture Subcommittee on ext. 288, or e-mail to American Public Human Department Operations, Nutrition and [email protected] Foreign Agriculture Services Association Kimberly Barnes O'Connor, Children's 810 First Street, NE Suite 500 Policy Director, Senate Health, Education, Washington, DC 20002-4267 Tuesday, March 9, 1999 Labor and Pensions Committee (202) 682-0100 1:30 p.m. to 3:00 p.m. Debbie Banks, Senior Policy Advisor, fax: (202) 289-6555 328A Russell Senate Rep. Robert T. Matsui, (D-Calif.) http://www.aphsa.org Office Building Who is Leaving the Food Stamp Program? Subgroup Participants Participants Participation Percent Subgroup as Subgroup as Percent of in 1994 in 1997 Change 1994- Change in Percent of Percent of Decline (thousands) (thousands) 1997 Partici- 1994 caseload 1997 caseload from (thousands) pation Subgroup 1994-1997 Legal 1,537 706 -831 -54% 6% 3.4% 14% Permanent Resident Aliens Childless 1,148 648 -500 -43.5% 4% 3.1% 8.4% Unemployed Adults AFDC/TANF 13,052 9,442 -3,610 -27.6% 48% 43% 60.8% Recipients All Other 11,697 10,707 -990 -8.5% 43% 49% 16.7% Participants Total 27,434 21,503 -5,931 -21.6% 100% 100% 100% Participants July-September Food Stamp Program Quality Control Data for 1994 and 1997 drugs and their implications for whether the United States is serious to other federal entitlement pro- Medicaid. Dr. Garth L. Splinter of the about providing health care to unin- grams. She points out some new and Oklahoma Health Care Authority sured children in immigrant families. unanticipated challenges for program shares his views on Viagra and the She calls our attention to the current administrators. dilemmas Medicaid directors face conflict between public health and I hope you'll find this issue inter- when dealing with quality-of-life immigration policy, and she points esting. Our goal is for POLICY & drugs. He offers much food for out its impact on children's health. PRACTICE to be a useful resource for thought as he attempts to answer: In our "View from the Nation's you in your daily work. I'm open to How do we determine what Medicaid Capital," Elaine M. Ryan, our govern- your suggestions and comments, so should and should not cover? ment affairs director at APHSA, looks don't hesitate to contact me. Claudia Schlosberg of the National at the Temporary Assistance for Best wishes for a happy and safe Health Law Program questions Needy Families program and its link holiday season and new year. Travel without good planning, and you're bound to get lost. However far you wander. tive rates. If you're less than Term Lite you need an insurance plan enchanted with your current Member Assistance that stays in place. Pick one insurance carrier. call us at High Limit Accident up from APHSA and you're 800 424-9883. or in Wash- covered wherever you go. ington, DC at 202 457-6820 As long as you're " member, We promise you'll never get we'll help you find your way " crumby deal. And you'll to the best of plans, and never have to go into the always at the most competi- woods alone. INSURANCE FOR APHSA MEMBERS This plan is administered by Scabury & Smith, a Marsh & McLennan Company. The term life plan is underwritten by the New York Life Insurance Company. APHSA 51 Madison Avenue, New York, NY 10010. on Policy Form GMR DECEMBER 1991 View from Nation's the Capital Parcy Practice Jamue 12/98 The Unfinished Agenda: Two Years after TANF or more than six decades, the connections. Today, we know that gies also provide families with F Aid to Families with Dependent those intricate interconnections that short-term, lump-sum assistance Children (AFDC) program was the once worked well (in a system the aimed at reducing dependence on entryway poor nation grew to despise) now welfare. More states are granting a families walked threaten the long-term success mix of cash, benefits, and services through to gain of welfare reform. tailored to meet the individual by Elaine M. Ryan access to a con- Welfare reform marked the end of needs and circumstances of eligible catenation of fed- an era of an individual entitlement to families. Unlike the previous systen eral benefit pro- cash and child care assistance in administered primarily by human grams. Under the this country and the beginning of a service agencies, today states are AFDC program, new era of time-limited, work-focused changing the way they deliver ser- families were eligi- assistance and services. The vices to TANF families by devolving ble for food Temporary Assistance for Needy resources and responsibilities to stamps, Medicaid, Families (TANF) block grant that new public, private, and not-for- child support. child care, and, in replaced AFDC provides states with a profit partners. some instances, federal housing and fixed amount of federal funds and The five-year federal lifetime limit energy assistance. permits broad flexibility in setting eli- on the receipt of TANF assistance Over time, AFDC and other federal gibility levels by disregarding earned has created a new imperative to entitlement programs were inextrica- income and assets and by providing move clients into employment as bly entwined, bound together by a families with a wide array of cash quickly as possible and to support complex set of federal regulations, and supportive work services. their transition into the world of work funding streams, and policies. Since In response, states are adopting Food stamps, Medicaid, child care, the enactment of the Personal a variety of innovative policies and and housing are key components of Responsibility and Work Opportunity strategies which allow families on state job retention strategies for Reconciliation Act of 1996 that welfare to keep more of their these new low-income workers. Over repealed AFDC, we have begun to earned income and to accumulate time, families may increase their fully appreciate those intricate inter- assets. These policies and strate- earnings such that they no longer 8 POLICY & PRACTICE require this assistance; however, in policy and program conflicts Food Stamp program. These two the short-run, these critical programs described below. separate agencies periodically pro- provide families the stability they mulgate inconsistent policies that need to stay in the workforce and off The Food Stamp Program increase the cost of administering of welfare. The strict regulations governing the the Food Stamp program and impede As TANF has diverged from the administration of the federal Food state welfare reform innovations. rigid rules of the former AFDC pro- Stamp program pose serious obsta- This was true under AFDC and, unfor- gram, its link to the remaining federal cles to states' strategies to reduce tunately, is increasingly the case entitlement programs has produced dependence on welfare and increase under TANF. new and unanticipated challenges for the employment of TANF clients. The Today, more than 30 states cur- program administrators as they clash between the TANF and Food rently operate TANF diversion pro- implement welfare reform measures. Stamp programs is historic, costly. grams that provide short-term assis- In reviewing the first two years of and complex. State administrators tance to low-income families as an implementing TANF, we find the inter- see the conflicts of the two programs alternative to TANF receipt. Diversion actions between program rules and on a daily basis because the TANF or welfare avoidance payments often policy objectives of TANF and such and Food Stamp programs are often take the form of lump-sum cash federal programs as Food Stamps administered by the same human assistance to help families overcome and Medicaid. for example, no longer service agency charged with moving a crisis or to remove an obstacle to make sense in the post-AFDC world welfare clients off assistance and their employment. Under diversion and threaten the long-term success into employment. This is not the programs, families may receive three of welfare reform. case at the federal level where the months of benefits in one month or This unfinished agenda of welfare U.S. Department of Health and assistance to repair an automobile reform will require significant federal Human Services oversees the TANF needed to transport a client to his statutory and regulatory change in block grant and the U.S. Department or her job. order to untangle the conflicting of Agriculture (USDA) administers the Earlier this year, the USDA Food and Nutrition Service (FNS) mandat- ed that when states determine eligi- bility for food stamps, they must con- sider diversion payments as income. Under this mandate, many of the families who received diversion payments would be ineligible for food stamps. In response to this ruling. states argued that diversion assistance pro- vided to families in crisis or in an effort to remove an obstacle to employment should not disqualify families from receiving the food assistance they need. Lump-sum assistance should not be equated to income because a family may have resources in one month and be desti- tute the next. More important, once the link to food stamps is broken, families may be discouraged from applying for benefits they are entitled to receive in the future. Later in the year, FNS revised its original guidance and replaced it with a moderately more generous but administratively cumbersome one. The new rules disregard diversion payments as income if states docu- ment in writing that a payment was provided to a family "only once in a 12 month period for needs not to exceed a 90 day period.' Therefore. 9 DECEMBER 1998 if two or more payments are made to maintenance system. and must be streamlined. Absent a family within a year, the state must As a result, the more effectively such an effort. the TANF and Food consider these payments as income. states move TANF clients into jobs, it Stamp programs will continue on the Clearly, states need greater flexibil- is more likely states will be penalized current collision course with costly ity to operate diversion programs for payment errors. The efficient and and unintended consequences for than this policy allows. It is impossi- effective administration of this feder- the families we serve. ble to predict when and if families al program is a goal of all state may need short-term assistance. administrators and the federal food Medicaid While states could avoid this policy stamp error rate should reflect seri- The welfare reform law also broke the obstacle by assigning all families to ous patterns of payment inaccuracy, link between TANF and Medicaid eligi- the welfare rolls. this would be con- not minor errors attributable to suc- bility. Despite efforts to manage the trary to the goal of reducing long- cessful welfare-to-work efforts. two programs together, several term dependence on assistance. These are but two examples of the administrative challenges have If this is the extent of flexibility the clash between the TANF and Food emerged that could have serious administration believes it can exer- Stamp programs, yet many others implications for families in need. cise within the Food Stamp statute, exist. Although the welfare reform Under the former AFDC program, then perhaps congressional action is law includes an option for states to clients were categorically eligible for necessary to afford states greater coordinate the two programs together Medicaid. Now. the law requires state flexibility in aligning the TANF and under the Simplified Food Stamp pro- administrators to refer to the income Food Stamp programs. gram, few states have opted to do and resource standards that were in Another illustration of the need for so. In fact, the scope of the "simpli- effect in the state's AFDC plan as of increased program flexibility is the fied program" is so limited that it is July 16, 1996, to determine appli- calculation of the food stamp error rate. Under federal law, food stamp benefits are adjust- ed as recipients' incomes and earn- ings change. States are measured on their ability to accu- BORSPINDLE rately calculate the food stamp benefit amount. If the so- called "error rate" increases, states will be penalized. Calculating the amount of the benefit was sim- pler under AFDC because far fewer welfare clients were earning income from employment and remaining on assis- tance. In the TANF program, however, clients' employ- more complex to administer than the cants' eligibility. Section 1931 of the ment may be episodic and the current program. With the shared welfare reform law permits states to amount of earned income may fluctu- objective of ensuring greater access increase income and resource stan- ate significantly from month to to the Food Stamp program for all dards (but not above the Consumer month. The federal method used to needy families, state and federal offi- Price Index), decrease income stan- calculate the food stamp error rate. cials must work together in crafting a dards (but not below those in place however. has remained virtually sweeping agenda for reform. on May 1. 1998), or use "less restric- unchanged despite the fundamental Moreover. the Food Stamp program tive" methods. The so-called "look- shift away from the AFDC income is increasingly difficult to administer back" is administratively burdensome 10 POLICY & PRACTICE and confusing to clients and case- Statutory or regulatory change is For example, since the enactment workers alike. On a practical level, it necessary to allow states to use of welfare reform, 42 states have is unreasonable to expect Medicaid TANF savings when calculating the changed TANF income and eligibility eligibility workers to refer to state cost-neutrality of future waivers. rules that disregard earned income plan provisions of an AFDC entitle- Perhaps combined savings attribut- and 38 states have permitted clients ment program long-since repealed. A able to unobligated TANF block grant to retain more resources and still new standard that is simpler to dollars and Medicaid could be used remain on assistance. These policies administer and easier for families to or a modest copayment for working are intended to provide clients with understand must be established. poor families could generate the an incentive to work and a chance to Statutory or regulatory action is resources needed to meet the new accumulate savings. At the same also necessary to preserve states' federal test. Without a creative solu- time. if these clients live in public option to provide more than 12 tion, thousands of families will lose housing, the earned income is often months of Medicaid assistance to the critical medical assistance they absorbed by an increase in rent. families making the transition from may need to make the successful Landlords receive higher payments welfare to work. Under the welfare transition to long-term self-sufficiency. but the net effect to the TANF family reform law, states are required to pro- As the differences between the is negligible. The Quality Housing and vide 12 months of transitional med- TANF and Medicaid programs grow, it Work Responsibility Act of 1998 ical assistance to individuals whose will become administratively costly to approved by Congress this year pro- earned incomes increase beyond a manage these programs together and vides some relief to families who exit point that would otherwise make separate eligibility systems may be TANF for work by restricting rent them eligible for Medicaid under developed. In addition, families who increases for up to two years. Section 1931. Prior to the enactment once sought assistance through the Although this is a welcomed reform, of the welfare reform law, however, agencies that administered the for- over time, however, we cannot pre- only 12 states had been granted fed- mer AFDC program in the future may dict the effect of changing TANF and eral transitional Medicaid waivers that gain access to the Medicaid program public housing policies. We do not allowed them to provide Medicaid through a variety of public, nonprofit, know the number of families whose assistance to families for longer than or private entities which administer incomes will fall to zero when their one year. Under the current calcula- the TANF program. Without action to lifetime time limits are reached and tion of cost-neutrality, however, states afford states greater flexibility in the what effect their lack of income may will not be permitted to continue to administration of the Medicaid pro- have on the future cash flow and extend this additional Medicaid cover- gram, the result may be fewer eligible capacity of public housing authori- age to families when their waivers families receiving the medical assis- ties. Therefore, we must monitor the expire. Under the AFDC program, tance they are entitled to receive. relationship between welfare reform states were able to achieve cost-neu- and housing to preserve the stability trality of their waivers by calculating Public Housing TANF clients need to prosper in savings to the federal government in Beyond the Food Stamp and the workforce. both the AFDC and Medicaid pro- Medicaid programs, there are various TANF is a new program constantly grams. States calculated that the federal benefit programs that will changing to meet the diverse needs cost of extending transitional enable states to achieve success in of clients on assistance and clients Medicaid for more than one year welfare reform. One such program is who have left the caseload, and to would be offset by the savings the housing assistance. Safe and afford- prevent dependence on the welfare federal government would achieve by able housing provides the stability system. We must act now to ensure reducing the likelihood that clients families need to exit welfare for work. that key federal programs and poli- would return to welfare. Since the fed- Though limited in scope, the feder- cies are aligned to meet the goals eral cost of TANF is capped, the al government funds housing assis- and objectives of welfare reform and Health Care Financing Administration tance for low-income families living in provide states all the flexibility they informed states that the cost-neutrali- public housing authorities and need to support state TANF program ty of future transitional waiver demon- through Section 8 certificates and innovations. If we fail to do so, mil- strations must be achieved within the vouchers. On average, families lions of dollars will be spent navigat- Medicaid program alone. Under this receiving assistance pay approxi- ing through the federal entitlement new test, states will not be able to mately one-third of their incomes in bureaucracy. separate eligibility sys- demonstrate sufficient savings and rent. The interplay between TANF tems will be established, and fami- subsequent transitional Medicaid and the rules governing the adminis- lies may lose eligibility and essential waivers will not be approved. Indeed, tration of these federal housing pro- benefits in the process. in three of the 12 states, the transi- grams often produce unintended neg- tional Medicaid waivers have expired ative outcomes for families trying to Elaine M. Ryan is director of and have not been renewed. work their way off welfare. government affairs at APHSA. 11 DECEMBER 1998 NO.662 P002/005 01/07/99 16:47 CBPP 4567431 Decline in Food Stamp Participation from March 1994 to July 1998 Arizona Minnesota Wisconsin Ohio Texas Indiana Kansas Mississippi New Hampshire Massachusetts Colorado Florida Iowa Virginia Missouri Idaho Rhode Island Nevada Wyoming Louisiana Utah California Pennsylvania North Dakota New Mexico Delaware Tennessee Washington Vermont Michigan Illinois Oklahoma New York New Jersey Alabama Kentucky Oregon Georgia West Virginia Maine Maryland North Carolina South Dakota Montana South Carolina Connecticut Nebraska Alaska Arkansas Virgin Islands Dist. of Columbia Hawaii Guam TOTAL US -20 0 20 40 60 % Reduction Center on Budget and Policy Priorities F:/sip/karen/presentn/1103DS.ps NO.662 P003/005 01/07/99 16:47 CBPP 4567431 Decline in Food Stamp Participation Since March 1994 Percent Decline - March 1994 to STATE March 1994 July 1998 July 1998 ALABAMA 561,803 415,612 -26.0% ALASKA 51,692 43,834 -15.2% ARIZONA 512,288 229,316 -55.2% ARKANSAS 290,212 254,142 -12.4% CALIFORNIA 3,191,890 2,216,978 -30.5% COLORADO 275,765 177,222 -35.7% CONNECTICUT 225,044 189,268 -15.9% DELAWARE 61,717 43,394 -29.7% DIST. OF COLUMBIA 92,427 82,640 -10.6% FLORIDA 1,469,271 946,660 -35.6% GEORGIA 826,295 635,794 -23.1% GUAM 15,058 16,931 12.4% HAWAII 115,461 121,388 5.1% IDAHO 88,094 58,368 -33.7% ILLINOIS 1,207,279 875,297 -27,5% INDIANA 543,248 301,515 -44.5% IOWA 203,883 134,046 -34.3% KANSAS 197,340 113,625 -42.4% KENTUCKY 536,337 402,139 -25.0% LOUISIANA 763,280 525,729 -31.1% MAINE 141,742 110,764 -21.9% MARYLAND 399,877 314,126 -21.4% MASSACHUSETTS 451,344 276,984 -38.6% MICHIGAN 1,047,450 755,230 -27.9% MINNESOTA 323,939 173,818 -46.3% MISSISSIPPI 519,006 309,173 -40.4% MISSOURI 606,179 399,431 -34.1% MONTANA 74,280 61,610 -17.1% NEBRASKA 113,484 95,726 -15.6% NEVADA 96,977 67,694 -31.6% NEW HAMPSHIRE 63,760 38,518 -39.6% NEW JERSEY 554,501 409,894 -26.1% NEW MEXICO 251,655 176,779 -29.8% NEW YORK 2,175,154 1,592,673 -26.8% NORTH CAROLINA 646,230 509,793 -21.1% NORTH DAKOTA 47,459 33,118 -30.2% OHIO 1,269,050 693,748 -45.3% OKLAHOMA 382,926 279,962 -26.9% OREGON 298,653 229,435 -23.2% PENNSYLVANIA 1,243,533 866,465 -30.3% RHODE ISLAND 96,336 63,874 -33.7% SOUTH CAROLINA 390,940 326,527 -16.5% SOUTH DAKOTA 55,200 44,574 -19.3% TENNESSEEE 747,371 527,488 -29.4% TEXAS 2,764,395 1,529,964 -44.7% UTAH 131,124 90,618 -30.9% VERMONT 65,475 46,837 -28,5% VIRGIN ISLANDS 19,632 17,501 -10.9% VIRGINIA 571,254 376,137 -34.2% WASHINGTON 483,004 344,136 -28.8% WEST VIRGINIA 333,260 257,976 -22.6% WISCONSIN 334,900 181,049 -45.9% WYOMING 34,698 23,797 -31.4% TOTAL US 27,965,172 10 317 01/07/99 16:47 CBPP -> 4567431 NO.662 P004/005 Decline in AFDC/TANF Participation compared to Decline in Food Stamp Participation From March 1994 - June 1998 Percentage decline in number of cases Idaho Wyoming Wisconsin West Virginia Mississippi Kansas Florida Oregon Alabama South Carolina Colorado Oklahoma Arkansas Texas Georgia North Carolina New Hampshire Tennessee Ohio Michigan Virginia South Dakota Louisiana Arizona North Dakota Massachusetts Indiana Maryland Missouri Kentucky Utah Pennsylvania lowa Montana New Jersey Maine Delaware Connecticut Illinois Nevada Vermont Washington New York New Mexico Dist. of Columbia California Minnesota Alaska Nebraska Rhode Island Hawaii TOTAL US -40 -20 0 20 40 60 80 100 % Reduction % Reduction Center on Budget and Policy Priorities F:/sfp/karen/presentn/1104SD.ps AFDC/TANF FOOD STAMP Alabama AFDC / TANF Caseload 70,000 160% Caseload relative to 1996:8 (left axis) 60,000 Fight 140% 120% 50,000 Caseload (left axis) 100% 40,000 Caseload 80% 30,000 60% Caseload relative to 1998:6 20,000 40% 10,000 20% 0 0% 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 Year casr_fin.wpd Page 1 October 21, 1998 MEMORANDUM FOR THE PRESIDENT FROM: Bruce Reed Andrea Kane SUBJECT: Minority Welfare Caseloads You asked us to evaluate recent reports that African-American and Hispanic families are leaving welfare more slowly than whites, and to consider what more we could do to ensure all welfare recipients are making the successful transition from welfare to work. Here is a brief summary of the trends, along with some new ideas we are developing for consideration in your FY 2000 budget. I. Caseload Trends We have worked closely with NEC, CEA, OMB, HHS, and the Census Bureau to examine the most recent welfare caseload data (generally through 1997). As detailed in the attached tables, the data show: 1. Most of the changes in the welfare caseload can be attributed to changes in the composition of the population as a whole -- specifically, population growth explains nearly two-thirds of the difference in caseload decline between Hispanics and whites and nearly half the difference between blacks and whites. Since 1994, the number of welfare cases has indeed dropped more among whites (26 percent) than among blacks (18 percent) and Hispanics (9 percent). However, when population growth is taken into account, the difference narrows dramatically. The rate of welfare dependency has dropped sharply for all populations - -- by 26 percent among whites, 21 percent among blacks, and 20 percent among Hispanics. In other words, minorities are leaving or staying off welfare at nearly the same rate as whites, but make up a growing share of the welfare population because they make up a growing share of the population as a whole. [See Table 1.] 2. The difference in caseload decline is even narrower among adults. Since 1994, the adult rate of welfare dependency has declined by 30% among whites, 26% among blacks, and 24% among Hispanics. casr_fin.wpd Page 3 A. Current Initiatives As you know, you have put forward many initiatives to help the hardest-to-serve welfare recipients and those living in concentrated areas of poverty. Many of these initiatives were implemented in 1998, too late to influence 1994-1997 trends outlined above. The $3 Billion Welfare-to-Work Fund you fought for in the Balanced Budget Act is designed specifically to help long-term welfare recipients (and non-custodial parents) in high-poverty areas obtain jobs and move up a career ladder. While it is too early to have demographic data on the individuals served by these funds, the distribution formula and individual eligibility criteria ensure that most of these funds will be spent on minorities. The first of these funds were awarded in January 1998 and are just now starting to provide individual services. Welfare to Work Transportation Funds enacted in the TEA-21 transportation reauthorization bill will help welfare recipients and low-income workers get to where the jobs are, often in suburban areas not served by public transportation. The omnibus budget bill includes $75 million for this year. Welfare-to-Work Housing Vouchers we've proposed will help families in isolated urban or rural areas move closer to job opportunities. Congress has funded our request for 50,000 vouchers. Community Empowerment Initiatives. The Administration's Community Empowerment initiatives -- empowerment zones, enterprise communities, Brownfields, CDFIs -- will spur economic development and job creation in distressed neighborhoods and help address the geographic isolation faced by minorities on welfare. B. New Initiatives In preparation for next year's budget, we are developing a number of options to address the particular challenges faced by minorities in making the transition from welfare to work: Increasing Investments in English-Language and Literacy Training. We hope to recommend targeted new investments in two areas that directly affect minority and long-term recipients: learning English and learning to read. This could be done by expanding existing Department of Education adult education programs, or better yet, by dedicating Welfare-to-Work funds for job-related literacy and ESL casr_fin.wpd Page 4 programs, provided either in the workplace or by community organizations preparing individuals for employment. Expanding Work-Related Drug Treatment. Since many of those remaining on welfare suffer from drug or alcohol dependencies, we are exploring ways to provide drug treatment for those who agree to go to work. Targeting Welfare-to-Work Funds to the Toughest Areas. While the current Welfare-to-Work formula favors high-poverty areas, we are going to examine whether the funds could be even more targeted. Increasing Work and Child Support Among Noncustodial Fathers. We may be able to attract bipartisan support for an effort to help states increase the employment and child support payments of noncustodial parents. Minority Caseloads Analysis and Tables Table 1: Population-Adjusted Change in Rate of Welfare Dependency Since 1994, the number of welfare cases has dropped more among whites (26 percent) than among blacks (18 percent) and Hispanics (9 percent). However, when population growth is taken into account, the difference narrows dramatically. The rate of welfare dependency has dropped sharply for all populations -- by 26 percent among whites, 21 percent among blacks, and 20 percent among Hispanics. Specifically, population growth explains nearly two-thirds of the difference in caseload decline between Hispanics and whites and nearly half the difference between blacks and whites. Table 1: Change from 1994 to 1997* Race/Ethnicity Number of Welfare Population Rate of Welfare Cases Aged 15-49 Dependency (caseload adjusted for population) White -26% -0.1% -26% Black -18% 4.4% -21% Hispanic -9% 13.0% -20% casr_fin.wpd Page 5 * National data is only available through June 1997. We do not yet have more recent data, or state-specific data, that we consider accurate. in July, The New York Times reported more recent data provided by some states, but HHS believes that data, particularly for New York and California, may contain significant reporting errors due to states implementation of the new TANF data reporting system. The trends in population aged 15-49 are used here because this is the population group most likely to be a welfare head of household, whose race/ethnicity would be counted when tallying the case demographics. Table 2: Population-Adjusted Adult Rate of Decline The difference in caseload decline among groups is even narrower for adults. Child-only cases are decreasing more slowly than the overall welfare caseload and are disproportionately minority; in fact, between 1994 and 1997 they increased (though they declined slightly between 1996 and 1997). Child-only cases are those in which the parent or adult is not part of the case, (e.g., adult is not a citizen but the child is; child is being cared for by a relative who is not part of the case; parent receives SSI rather than welfare). Therefore, child-only cases are not significantly affected by welfare to work efforts. After adjusting for population growth, the rate of welfare dependency for adults (percent of 15-49 year old population on welfare) has declined 30% among whites, 26% among blacks, and 24% among Hispanics. Table 2: Population-Adjusted Rate of Decline in Adult Welfare Dependency: 1994 - 1997* Rate of decline Rate of Rate of decline Population-adjust for all cases increase for for adult-headed ed rate of welfare child-only cases dependency for cases adult cases White - 26% 7% - 30% - 30% Black - 18% 3% - 23% - 26% Hispanic - 9% 9% - 15% - 24% *National data is only available through June 1997. Table 3: Racial Breakdown of Adult Cases The composition of the welfare caseload have changed gradually over the past 25 years, driven largely by population changes. Despite differing rates of caseload decline since 1994, the composition of the adult welfare caseload has remained relatively constant. Table 3: Racial Breakdown of Adult Cases casr_fin.wpd Page 6 Race/Ethnicity 1994 1997* White 40% 37% Black 36% 37% Hispanic 19% 21% Asians, Native Americans, and those designated "Unknown" comprise the rest of the caseload. *National data is only available through June 1997. Table 4: Employment Rate of Welfare Recipients There is encouraging evidence that the employment rates of minority welfare recipients (people on welfare in one year who were working the following year) are catching up with the employment rate for whites. Table 4: Employment Rate of Welfare Recipients: 1996-98 Race/Ethnicity 1996 1998 Percent Change 96-98 White 36% 38 % +5% Black 23 % 33 % + 44% Hispanic 19 % 29 % + 49% All Recipients 27% 34% + 28% Table 5: Trends in Marriage Rates and Births The trends in marriage rates and births to unmarried women could exacerbate the increasing proportion of Hispanic families on welfare. While the proportion of never-married single mothers is increasing for the entire population, the rate of increase is largest for Hispanic women. Also, the birth rate to unmarried teenagers remains much higher for blacks and Hispanics than for whites. While the rate is decreasing for blacks and slightly for whites, it continues to increase for Hispanics. For example, between 1991 and 1996, the rate of births to unmarried teenagers decreased 18% for blacks and 4% for whites, but increased 3% for Hispanics. 1992 1997 % Change % of all single mothers who were 30% 35% + 17% never married Never-married single mothers by race/ethnicity: White 17% 21% + 24% Black 51% 55% + 8% casr_fin.wpd Page 7 Hispanic 33% 42% +27% Table 6: Characteristics of Minorities on the Caseload Minorities on welfare are more likely to have characteristics associated with long-term welfare recipiency. Blacks and Hispanics on welfare tend to have lower educational levels, marriage rates, and larger families than whites, and are more than twice as likely to live in central cities and areas of concentrated poverty. Hispanics also have less recent work history than whites or blacks. Table 6: Characteristics of AFDC/TANF Recipients by Race/Ethnicity* TOTAL WHITE BLACK HISPANIC % without HS 43% 30% 43% 64% diploma % never married 47% 33% 69% 43% > 2 children 29% 20% 33% 39% Worked during the 45% 49% 48% 33% year Live in area w/ 48% 29% 67% 58% poverty rate > 20% Live in central city 49% 29% 68% 60% *These data are from the March 1998 Current Population Survey, showing characteristics of recipients in 1997. Minorities are more likely to be long-term welfare recipients. For example, in 1997, 20 percent of blacks on welfare had been on the rolls for at least five continuous years, compared to 19 percent for Hispanics and 14 percent for whites. 12/21/98 09:17 PROGRAM DEVELOPEMENT DIVISON 912024567431 NO. 937 P001/004 USDA:FCS:FSP:PDD:PDB FAX 3101 PARK CENTER DRIVE ROOM 718 ALEXANDRIA, VA 22302 Number of pages 12/21/98 including cover sheet Date To: From: Cynthico Rice ARTHUR T. FOLEY DIRECTOR PROGRAM DEVELOPMENT DIVISION Phone Phone 703-305-2490 Fax Phone 456-7431 Fax Phone 703-305-2486 CC: REMARKS: Urgent For your review Reply ASAP Please comment Who is Leaving the Food Stamp Program? Detailed Table 12/21/98 09:17 Participants Participants Participation Percent Percent of Subgroup in 1994 in 1997 Change in Decline from (In thousands) (in thousands) (in (ii) Participation Subgroup Legal Permanent Resident Aliens 1,537 706 831 54.07 % 14.01 % Childless Unemployed 1,148 648 500 43.55 % 8.43 % Adults AFDC/TANF Reciplents 13,052 9,442 3,610 27.66% 60.87 % PROGRAM DEVELOPEMENT DIVISON -> 912024567431 All Other Participants 11,697 10,707 990 8.46% 16.69% Total Participants 27,434 21,503 5,931 21.62% 100.00 % Notes: 1. Participation obtained from July-September Food Stamp Program Quality Control data for 1994 and 1997. 2. All subgroups are mutually exclusive. Childless unemployed adults do not include Legal Permanent Resident aliens. Likewise, AFDC/TANF recipients do not include aliens or childless unemployed adults. NO.937 P002/004 Who has left the FSP? 09:17 12/21/98 1994 Caseload Distribution Aliens Share of 1994-1997 Decline ABAWDs Others Aliens 17% 14% ABAWDs 8% 0 PROGRAM DEVELOPEMENT DIVISON > AFDC/TANF 47% AFDC/TANF 912024567431 61% NO.937 P003/004 Are Children of Legal Immigrants Leaving the Program Faster than Other Children? 12/21/98 09:17 Child Liveswith Child Does Not Live lien:, with Legal- Alien Number of Child Jumber of Children October 1996 1,251,000 11,034,000 September 1997 742,000 9,804,000 PROGRAM DEVELOPEMENT DIVISON -> 912024567431 Drop 434,000 1,682,000 Percent Drop 36.90% 14.64% Notes: 1. Numbers obtained from Fiscal Year 1997 Food Stamp Quality Control data. 2. Children in both groups are citizens. Those living with a legal alien are those whose household contains a legal permanent resident who was affected by the restrictions on benefits to immigrants as a result of welfare reform. NO.937 P004/004 931 TABLE 15-5.-COUNTED (NET) AND BASIC (GROSS) MONTHLY INCOME ELIGIBILITY LIMITS FOR THE FOOD STAMP PROGRAM, FISCAL YEAR 1998 48 States, the District of Household size Alaska Hawaii Columbia, and the territories Counted (net) monthly income eligibility lim- its 1. 1 person $658 $823 $756 2 persons 885 1,106 1,017 3 persons 1,11D 1,390 1,278 4 persons 1,338 1,673 1,539 5 persons 1,565 1,956 1,800 6 persons 1,791 2,240 2,060 7 persons 2,018 2,523 2,321 8 persons 2,245 2,806 2,582 Each additional person +227 +284 +261 Basic (gross) monthly income eligibility lim- its 2: 1 person 855 1,070 983 2 persons 1,150 1,438 1,322 3 persons 1,445 1,806 1,661 4 persons 1,739 2,175 2,000 5 persons 2,034 2,543 2,339 6 persons 2,329 2,911 2,678 7 persons 2,623 3,280 3,018 8 persons 2,918 3,648 3,357 Each additional person +295 +369 +340 1 Set at the applicable Federal poverty guidelines, updated for inflation through calendar 1996. 2 Set at 130 percent of the applicable Federal poverty guidelines, updated for inflation through cal- endar 1996. Source: U.S. Department of Agriculture, Food and Consumer Service. Counted liquid assets include cash on hand, checking and sav- ings accounts, savings certificates, stocks and bonds, individual re- tirement accounts (IRAs) and "Keogh" plans (less any early with- drawal penalties), and nonrecurring lump-sum payments such as insurance settlements. Certain "less liquid" assets are also counted: a portion of the value of vehicles (generally, the fair market value in excess of $4,650) and the equity value of property not producing income consistent with its value (e.g., recreational property). Counted assets do not include the value of the household's resi- dence (home and surrounding property), business assets, personal property (household goods and personal effects), lump-sum earned income tax credit payments, burial plots, the cash value of life in- surance policies and pension plans (other than Keogh plans and IRAs), and certain other resources whose value is not accessible to the household or are required to be disregarded by other Federal laws. Work-related requirements Unless exempt, most able-bodied adults must (to gain or retain eligibility) (1) register for work (typically with the welfare agency 266 / THE RAND JOURNAL OF ECONOMICS This solution is implemented by a tariff P(q) such that the marginal price P'(q) is larger than the marginal cost C'(q) for all relevant consumption levels. Moreover, it is immediate to verify that P(q) C(q). All consumers thus contribute to profits, both in absolute and in marginal terms. 3. Competitive bypass The market structure. We now assume that competitors can sell direct access to long distance, using the same bypass technology as the operator. We assume perfect competition between bypass operators, so that consumers opting for the bypass tech- nology face a tariff cq + ƒ, in which case a type-Θ consumer buys q(Θ) and gets a utility S(Θ). If the local network operator wants to attract a type-Θ consumer, it must give him a utility above S(Θ). This imposes a new participation constraint: U(Θ) ≥ max {S(Θ), 0}. (IR) The optimal tariff now maximizes W subject to (IC) and this new participation constraint (IR). We denote by qᶜ(Θ) the quantity consumed by type Θ under the optimal tariff Pᶜ(q), and by Uc(Θ) the consumer's utility. As before, the incentive-compatibility conditions (IC) imply that consumption in- creases with Θ. Together with technical efficiency, this implies that consumers can be divided among three groups: those below some threshold, Θᶜ, are excluded; those above another threshold, Θᶜ, are served through the bypass technology and consume more than q;; and those in the middle range, between Θᶜ and Θᶜ, are connected to the network and consume less than q.- We start the analysis with a result that is of general validity for the problems addressed here and in particular does not depend on the precise cost structure. Proposition 1. No quantity is served at a tariff below (minimum) cost. As a conse- quence, bypass is priced at cost. The basic intuition is as follows. Suppose that some consumers are served at a price below cost. Raising the price of the consumption of all these consumers up to cost obviously reduces their utility, without affecting the choices of the others. How- ever, it leaves open for these consumers the option of gaining at least the total surplus generated before (by keeping their consumption unchanged). Hence, given that profit is now positive or null on all types, the surplus levels generated by the new choices of these agents must then be higher than the previous levels. As a consequence, since utilities have decreased, total welfare (W = S AU) increases. Proposition 1 has an interesting implication: Corollary 1. The optimal strategy is the same whether the operator is allowed to sell bypass or not. Proof. From Proposition 1, one optimal strategy is to sell no bypass (i.e., to leave bypass, which generates no profit, to competitors) and to sell qᶜ(Θ) to types Obviously this strategy remains optimal if the operator is not allowed to sell bypass. Q.E.D. So far, we have assumed that competition among bypass operators leads to mar- ginal cost pricing (for both access and use). Using a standard Bertrand type of argu- ment, the next result shows that, more generally, the same pricing policy would emerge in equilibrium as soon as the incumbent monopoly and at least one competitor have access to the bypass technology. 934 tion); and States may impose food stamp disqualification when an individual is disqualified from another public assistance program. Automatic disqualification is required for those applying in mul- tiple jurisdictions, fleeing arrest, or convicted of a drug-related fel- ony. And States may disqualify individuals not cooperating with child support enforcement authorities or in arrears on their child support obligations. BENEFITS Food stamp benefits are a function of a household's size, its net monthly income, and maximum monthly benefit levels (in some cases, adjusted for geographic location). An eligible household's net income is determined (i.e., deductions are subtracted from gross in- come), its maximum benefit level is established, and a benefit is calculated by subtracting its expected contribution (30 percent of its counted net income) from its maximum allotment. Thus, a 3- person household with $400 in counted net income (after deduc- tions) would receive a monthly allotment of $201 (the maximum 3- person benefit in the 48 States, $321, less 30 percent of net income, $120). Allotments are not taxable and food stamp purchases may not be charged sales taxes. Receipt of food stamps does not affect eligi- bility for or benefits provided by other welfare programs, although some programs use food stamp participation as a "trigger" for eligi- bility and others take into account the general availability of food stamps in deciding what level of benefits to provide. In fiscal year 1996, monthly benefits averaged $73 a person and about $183 a household. Maximum monthly allotments Maximum monthly food stamp allotments are tied to the cost of purchasing a nutritionally adequate low-cost diet, as measured by the Agriculture Department's Thrifty Food Plan (TFP). Maximum allotments are set at: the monthly cost of the TFP for a four-person family consisting of a couple between ages 20 and 50 and two school-age children, adjusted for family size (using a formula re- flecting economies of scale developed by the Human Nutrition In- formation Service), and rounded down to the nearest whole dollar. Allotments are adjusted for food price inflation annually, each Oc- tober, to reflect the cost of the TFP in the immediately previous June. Maximum allotments are standard in the 48 contiguous States and the District of Columbia; they are higher, reflecting substan- tially different food costs, in Alaska, Hawaii, Guam, and the Virgin Islands (table 15-6). Minimum and prorated benefits Eligible one- and two-person households are guaranteed a mini- mum monthly food stamp allotment of $10. Minimum monthly ben- efits for other household sizes vary from year to year, depending on the relationship between changes in the income eligibility limits and the adjustments to the cost of the TFP. In a few cases, benefits can be reduced to zero before income eligibility limits are exceeded. CURIEN, JULLIEN, AND REY / 267 Corollary 2. The operator's optimal pricing strategy, the solution to the above program, is also part of a Nash equilibrium, in which the bypass competitor(s) prices (price) access and usage at marginal cost. Proof. By construction, the above optimal pricing policy is a best response to such pricing behavior from bypass operators. But this policy leaves no room for positive profits for the bypass operators, which thus cannot do better than marginal cost pricing. Q.E.D. Tariff, consumptions and market shares. The precise analysis of the optimi- zation problem is given in the Appendix. The first-order condition with respect to q(Θ) yields where φ(Θ) is the shadow value of the incentive constraint. This condition expresses the familiar tradeoff between informational rents and efficiency. φ(Θ) measures the cost of the rents that must be given to consumers to keep types close to Θ from consuming q(Θ). As usual, this cost induces a deviation from the "first-best" level; however, in contrast with the standard case, φ may be either positive or negative. Underconsumption is associated with φ(Θ) < 0 and a marginal price above marginal cost. This occurs when the problem faced is to prevent larger consumers from reducing their consumption, as in the standard nonlinear pricing model. Incentive com- patibility is then best achieved by reducing the consumption q(Θ)-along with the total price P(q(Θ))-since low-quantity/low-price bundles are less attractive for higher-val- uation consumers. But φ can also be positive, in which case q(Θ) is larger than q(Θ). This can be understood as follows. The operator has to match the bypass tariff, which is particularly attractive for large levels of consumption. This not only benefits large users, it also gives medium-size users an incentive to increase their consumption. To prevent this, the operator raises the consumption of large consumers compared to the first best-along with an increase of the total bill-to make the quantity/price bundle less attractive to consumers with lower willingness to pay. As a result, Proposition 2. There exists σᶜ and Θᶜ such that Θᶜ < Θᶜ < Θᶜ ≤ Θᶜ and qᶜ(Θ) < qFB(Θ) on 10c, σ[, qᶜ(Θ) > qFB(Θ) on ]Φ, Θ°[, > S(Θ) for all Θ < To prevent medium-range consumers (between Θᶜ and Θᶜ) from using bypass, the optimal tariff induces overconsumption-the marginal price P'(q) on the local network is thus below the marginal cost c for those consumers' consumption levels. This occurs before the participation constraint starts to be binding (for all consumers between Θᶜ and Θc). It is even possible that all consumers connected to the network, including those who overconsume, receive a strictly positive net surplus (i.e., Öᶜ = Θc). Lemma 1. If c is small enough, Uc(Θ) > S(Θ) for all Θ < Θᶜ. Clearly, given the consumption profile, the allocation of consumers between the two technologies is technically efficient. However, the allocation involves global in- efficiencies in market shares due to the quantity distortions. 935 At present, minimum monthly allotments for households of three or more persons range from $2 to over $80. In addition, a household's calculated monthly allotment can be prorated (reduced) for 1 month. On application, a household's first month's benefit is reduced to reflect the date of application. If a previously participating household does not meet eligibility recer- tification requirements in a timely fashion, but does become cer- tified for eligibility subsequently, benefits for the first month of its new certification period normally are prorated to reflect the date when recertification requirements were met. TABLE 15-6.-MAXIMUM MONTHLY FOOD STAMP ALLOTMENTS, FISCAL YEAR 1998 48 States and the Household size Alaska I Hawaii Guam Virgin District Islands of Co- lumbia 1 person $122 $154 $197 $180 $157 2 persons 224 283 361 331 288 3 persons 321 405 517 474 413 4 persons 408 514 657 602 525 5 persons 485 611 780 715 623 6 persons 582 733 936 858 748 7 persons 643 810 1,035 948 827 8 persons 735 926 1,183 1,083 945 Each additional person +92 +116 +148 +135 +118 1 Maximum monthly allotments for designated urban areas of Alaska. Two separate higher allotment levels are applied in remote rural areas of Alaska. They are 28 and 55 percent higher than the urban allotments shown here. Source: U.S. Department of Agriculture. Application, processing, and issuing food stamps Food stamp benefits normally are issued monthly. The local wel- fare agency must either deny eligibility or make food stamps avail- able within 30 days of initial application and must provide food stamps without interruption if an eligible household reapplies and fulfills recertification requirements in a timely manner. Households in immediate need because of little or no income and very limited cash assets, as well as the homeless and those with extraordinarily high shelter expenses, must be given expedited service (provision of benefits within 7 days of initial application). Food stamp issuance is a welfare agency responsibility, and issuance practices differ among welfare agencies. Most food stamp coupons are issued by: (1) providing (usually mailing) recipients an authorization-to-participate (ATP) card that is then turned in at a local issuance point (e.g., a bank or post office) when picking up their monthly allotment; or (2) mailing food stamp coupon allot- ments directly to recipients. However, in a growing number of States, electronic benefit transfer (EBT) systems are used. EBT systems replace coupons with an ATM-like card used to make food purchases at the point of sale by deducting the purchase amount from the recipient's food stamp benefit account. EBT issuance is 268 / THE RAND JOURNAL OF ECONOMICS Proposition 3. ΘFB and Θᶜ < Compared with the full-information case, the operator excludes more small users and the range of bypass increases. For small consumers, the analysis is similar to the pure (protected) monopoly: to reduce rents, the operator excludes the smallest consum- ers and reduces the consumption of the other ones. But for large users, the operator must leave the bypass surplus to the largest ones and increase the consumption of the other ones, which tends to reduce the market share of the network technology. The nature of the inefficiencies is illustrated by the boundary conditions f = 0 + f = cq(Θ) f, which reduce to = ) S(Φ, qᶜ(Θᶜ)) S(Θc, = g(Θ) (V(q()) V(q°(Θᶜ))). These conditions express the tradeoff between the marginal benefit of serving one more consumer through the local network and either not serving him at all (first condition) or serving him using the bypass technology. Transversality conditions, moreover, imply φ() < 0 and φ(Θ) > 0. Hence S(Θc, qᶜ(Θᶜ)) > 0 and S(Θ, q°(Θ)) - S(Θ, q())) ≥ 0 (> 0 if q(Θᶜ) > qᶜ(Θᶜ)). Including more types at the lower margin always raises the total surplus but forces the operator to reduce the price offered to all inframarginal consumers and reduces profits. The same is true at the higher margin if consumers just below Θᶜ strictly prefer to stay in the network (then > qᶜ(Θᶜ)). If large customers of the network are just indifferent between the network and bypass, they consume as under bypass (qc(Θ) = q(Θ)), and their allocation between the network and bypass is only a matter of technical efficiency; in this case, the market share of bypass is locally efficient. Although there are inefficiencies, Proposition 1 shows that no consumer is globally subsidized (in absolute terms). The optimal tariff can be interpreted as follows. The operator lets competitors serve consumers under bypass and uses only the network technology. It sells only up to a maximal quantity q°(Θᶜ) < qₑ, with an access fee P(0) > 0 and a nonlinear variable tariff P(q) - P(0). The marginal price P (g) is above the network marginal cost for low quantities and below the network marginal cost for large quantities, and it is actually possible to show that P(q) > C(q) for q<q Comparative statics. We present here some comparative statics that will help explain the impact of bypass. We assume here that the marginal cost of bypass is small, so that Lemma 1 applies. (All claims are proved in the Appendix.) To illustrate the nature of the externality imposed by the competitive pressure, let us consider an increase in the connection cost of bypass f. This weakens the competitive pressure on the operator and, as intuition suggests, reduces the market share of bypass. Obviously, the utility of bypass consumers decreases because of the direct increase in the competitive price for bypass, cq + f. As an indirect consequence, lowering the Dear Commissioner: The Administration is committed to ensuring that all eligible low-income households have access to vital Federal food assistance programs such as the Food Stamp program. The Food Stamp program, which provides food assistance benefits to more than 19 million eligible individuals nationwide, is the cornerstone of our efforts to guarantee a healthy and nutritious diet to all low- income families. Due to welfare reform and to today's strong economy, many households are finding employment and leaving the Temporary Assistance for Needy Families (TANF) program. We want to continue to work together to ensure that policy innovations in TANF do not inadvertently result in diminished access to food assistance for those who still need assistance, including working families. Recently we polled all the State agencies to find out if they were in compliance with food stamp application processing regulations. We learned that most State agencies are continuing to encourage households to apply for food stamps even if they have been diverted from applying for TANF. They are continuing to process food stamp applications even for households that have been denied TANF. Moreover, they are continuing to keep eligible households on the food stamp rolls after they have left TANF. Finally, many State agencies have taken special steps to avoid improperly denying or delaying food stamp receipt by providing special training for eligibility workers and/or sending notices to local offices reiterating the need to follow food stamp application processing regulations. While most of the feedback has been positive, your continued cooperation with federal law will ensure that low income families have the opportunity to apply for food stamps and receive benefits to which they are entitled. We want to highlight several circumstances where State agencies need to pay particular attention to the food stamp application processing regulations. Application for Benefits: Many States are diverting households from going on the TANF rolls. Households may be required to engage in job search prior to filing an application for TANF. These pre-application requirements for TANF cannot be applied to food stamps. Therefore, in these instances, it is particularly important that the State agency make sure that applicants are fully aware of their right to file an application for food stamps. Processing the Application: In some States, applications for TANF are put on hold until the applicant has completed the pre-application requirements. In these circumstances, it is imperative that the State agency continue to process the food stamp application and screen it for expedited service despite the delay in processing the TANF application. Determining Eligibility: In some circumstances, TANF-required job search may result in an applicant getting a job and therefore being denied TANF based on his/her earnings. Or, the applicant may accept a lump sum diversion payment CHARACTERISTICS OF FOOD STAMP HOUSEHOLDS: FISCAL YEAR 1997 (Advance Report) United States Department of Agriculture Office of Analysis and Evaluation Food and Nutrition Service October 1998 T he FSP is the nation's largest food children received an average monthly food stamp assistance program. In fiscal year 1997; it benefit of $234, reflecting their relatively large served an average of 22.9 million people average size (3.4 persons). Single-parent homes per month. Almost $19.6 billion was paid out in (3.1 persons on average) received an average of food stamps that year. $228 in food stamps, and multiple-adult homes with children (4.5 persons on average) received an Food stamps are made available to virtually all average of $268. low-income households with few resources to supplement their food purchases and help them Households containing elderly persons maintain a healthy diet. In fiscal year 1997, food represented 17.6 percent of all food stamp stamps were approximately one-fourth of a households. Slightly over three-quarters of them participating household's total monthly income were single-person households, which received an (cash plus food stamps). If the value of food average monthly benefit of $47. Households stamps was counted in addition to cash as gross containing elderly and other persons received an income, over one-fifth of food stamp households average benefit of $118. A substantial proportion would move from below to above half the poverty of food stamp households contained disabled line (Figure 1). persons (22.3 percent); these households received an average benefit of $104. In fiscal year 1997, slightly over half of all food stamp participants were children, most of whom II. Characteristics of Food Stamp Participants lived in single-parent households (Figure 2). The remaining participants were nonelderly adults Among adult participants (age 18 or older), (40.6 percent) or elderly adults--age 60 or older-- women outnumbered men by over two to one (7.9 percent). (Table 2). Of the children participating, 34.1 percent were of preschool age (0 to 4 years), and I. Composition of Food Stamp Households 65.9 percent were of school age (5 to 17 years). In fiscal year 1997, the majority (58.3 percent) of The largest proportion of food stamp participants food stamp households contained children (Table were white, non-Hispanic (40.3 percent); about 1). Of these households, over two-thirds (69.1 one-third were African-American, non-Hispanic percent) were single-parent homes, approximately (34.9 percent); and approximately one-fifth were one-quarter (24.7 percent) were multiple-adult Hispanic (19.2 percent). The remaining homes, and the remaining households contained participants were Asian, Native American, or of no members over the age of 17. Households with another race or ethnicity (Table 3). This brief was prepared by Scott Cody and Jacquelyn Anderson of Mathematica Policy Research, Inc. Programming support was provided by Mark Brinkley. The information presented in this brief is based on data collected by the Food and Nutrition Service for quality control purposes for fiscal year 1997. Page 2 FIGURE 1 POVERTY STATUS OF FOOD STAMP HOUSEHOLDS WITH AND WITHOUT FOOD STAMPS, 1997 51 to 100% of Poverty 51 to 100% of Poverty 51.5% 66.6% 101% + of Poverty 8.6% 15.3% 101% + of 18.1% 39.9% Poverty Poorest Households Poorest Households (50% of Poverty or Less) (50% of Poverty or Less) CASH ONLY CASH AND FOOD STAMPS FIGURE 2 DISTRIBUTION OF FOOD STAMP PARTICIPANTS, 1997 Children in Single Parent Households 34.4% Nonelderly Adults 40.6% 2.8% Children in Other Households Elderly Adults 7.9% 14.3% Children in Multiple Adult Households Source: 1997 Food Stamp Quality Control sample. Page 3 III. Benefits and Income of Food Stamp Minnesota, New Hampshire, Ohio, and Vermont-- Households states with typically above-average gross household income and below-average household The average monthly food stamp benefit in fiscal size. year 1997 was $169. The majority (68.6 percent) of food stamp households received monthly V. Change in Participation Over Time benefits of over $100, and 36.4 percent received over $200 (Table 4). In fiscal year 1997, 80.7 The number of persons participating in the FSP percent of all benefits went to households with grew substantially between 1989 and 1994 and children, and 6.6 percent of all benefits went to has declined steadily since 1994 (Figure 3). households with elderly members (not shown). Specifically, the number of food stamp participants grew from 18.9 million persons in The average monthly gross income of food stamp 1989 to a record high of 28.0 million in March households was $558 (Table 5). The majority of 1994 and has decreased to 22.1 million in June households (78.4 percent) received unearned 1997. The change in the number of Food Stamp income from one or more of the following Program participants is associated with changes in sources: Aid to Families with Dependent the economy, as indicated by the number of Children/Temporary Assistance to Needy unemployed persons and the number of persons Families (AFDC/TANF), General Assistance living in poverty. For example, Figure 3 shows (GA), Social Security, Supplemental Security that the surge in FSP participation between 1989 Income (SSI), Unemployment Compensation, or and 1994 was associated with a worsening some other payment such as veterans benefits or economy, and the drop in participation since 1994 educational loans. Slightly less than one-forth of has been associated with an improving economy. all households (24.2 percent) received earned income. Almost a tenth (9.2 percent) received The changes in program participation have been zero gross income. accompanied by changes in the composition of the caseload (Table 7). Although households with IV. Characteristics of Food Stamp Households children remain a large proportion of the caseload, by State that proportion has dropped from a high of 62.2 percent in 1992 to 58.3 percent in 1997. On the In fiscal year 1997, over half of all food stamp other hand, the proportion of households with households were located in eight states: elderly has risen from a low of 15.4 percent in California, Florida, Illinois, Michigan, New York, 1992 to 17.6 percent in 1997. However, the most Ohio, Pennsylvania, and Texas (Table 6). The substantial change in caseload composition has average monthly food stamp benefit varied by been the dramatic rise in the proportion of state, reflecting differences in income, expenses, participating households with disabled persons, household size, and composition. Average rising from a low of 8.9 percent in 1990 to 12.5 monthly benefits were largest in Alaska, Hawaii, percent in 1994, and from 18.9 percent in 1995 to Guam, and the Virgin Islands because the 22.3 percent in 1997. maximum allotment is set higher in those areas to accommodate higher costs of living. Within the Changes Under PRWORA. The Personal contiguous United States, food stamp benefits Responsibility and Work Opportunity were highest (greater than $200) in Arizona and Reconciliation Act of 1996 (PRWORA) took Texas--states with above-average household size. effect in FY 1997. This legislation, enacted Conversely, average monthly food stamp benefits August 22, 1996, made the following significant were lowest (less than $140) in Connecticut, modifications to the FSP:2 Page 4 FIGURE 3 FOOD STAMP PROGRAM PARTICIPANTS, UNEMPLOYED PERSONS, AND POOR PERSONS Millions 40 Persons in poverty a 30 b FSP participants 20 10 Unemployed persons c 0 84 85 86 87 88 89 90 91 92 93 94 95 96 97 Year "Annual values. The number of persons in poverty in Fiscal Year 1997 was not available when this report went to print. Source: Bureau of the Census, Poverty in the United States: 1996, Series P60-198. ᵇAverage monthly values. Source: Food and Nutrition Service, Fiscal Years 1984-1997 FSP Participation and Issuance. °Average monthly values. Source: Economic Report of the President, 1998 Table B-36. Page 5 Most legal permanent resident aliens are households, regardless of almost any nonfinancial disqualified from the FSP.³ categorical criteria. Second, instead of cash, it provides benefits in the form of coupons or Most able-bodied, nonworking, childless adults electronic benefits which can be redeemed for are limited to 3 months of FSP benefits in any food in any of about 200,000 authorized stores 36-month period. across the nation. The cost of providing food stamps to needy persons is funded fully by the The maximum food stamp benefit is reduced federal government. Administrative costs are from 103 percent to 100 percent of the Thrifty shared by federal, state, and local governments. Food Plan. Eligibility. To be eligible for food stamps, a The standard deduction is frozen indefinitely at household's assets, gross income, and net income, fiscal year 1996 levels. which is based on gross income less deductions permitted under the FSP, must not exceed New shelter deduction caps are established for specified levels that vary by household size, fiscal years 1997 through 2001, with the cap composition, and location. Most permanent frozen at fiscal year 2001 levels thereafter. resident aliens are ineligible to join the FSP, and most able-bodied, nonworking, childless adults The fiscal year 1997 data provide the first picture are limited to 3 months of FSP benefits in any 36- of food stamp participants since PRWORA's month period. implementation. The data show that the percentage of permanent resident aliens Food Stamp Benefit Computation and participating in the FSP fell from 5.6 percent in Issuance. Benefits are computed by subtracting fiscal year 1996 to 4.4 percent in fiscal year 1997. 30 percent of a household's net income from the In addition, the percentage of able-bodied, maximum benefit amount, which is based on 100 nonworking, childless adults participating in the percent of the June cost of the Thrifty Food Plan FSP fell from 3.6 percent in fiscal year 1996 to (TFP) for a family of four, adjusted for household 2.9 percent in fiscal year 1997. PRWORA is size and location. The TFP is based on the cost of likely a contributing factor behind the continued a market basket of food that provides an overall decline in the FSP caseload and costs economical and nutritious diet. In fiscal year since 1996. However, the extent to which these 1997, the maximum benefit for a family of four in trends are caused by PRWORA rather than other the contiguous United States was $400 per month. factors such as the economy is difficult to Program participants receive their monthly determine. Furthermore, PRWORA was benefits through the mail, directly from the local implemented in stages throughout fiscal year office, or through an electronic benefit transfer 1997, and the average annual estimates provided (which is similar to a bank card). here include data from before major components of PRWORA were in effect. VII. Data VI. Description of the Food Stamp Program The estimates presented here are based on data extracted from the Integrated Quality Control The FSP is unique among income maintenance System, which is an ongoing review of food stamp programs in two important ways. First, it offers households designed to measure the accuracy with assistance to nearly all financially needy which eligibility and benefit determinations are Page 6 made. All estimates are based on a full-year sample of 48,854 households. Food and Nutrition Service administrative records indicate that the FSP served 22.9 million persons in fiscal year 1997, and food stamp households received an average benefit of $173 per month. The figures in the attached tables of 23.1 million participants with an average household benefit of $169 vary from the administrative figures because they are estimates from the Food Stamp Quality Control sample, which weights data by households rather than persons or benefits. Administrative figures are based on a monthly census of actual FSP participation and benefit issuance. Notes ¹The 1994-to-1995 increase in the proportion of households with disabled persons is due in part to a change in the definition of households with disabled persons. However, using the old definition, the proportion with disabled still increases from 12.5 percent to 13.3 percent. ²A summary of the PRWORA provisions that affect the FSP is available from the FNS World Wide Web site (http://www.usda.gov/fcs/fcs.htm). 3 The Agricultural Research Bill, enacted on June 23, 1998, restored eligibility to a substantial number of legal immigrants who lost eligibility under PRWORA. Specifically, child, elderly, and disabled permanent resident aliens will be able to begin receiving federal food stamp benefits on November 1, 1998. Page 7 Table 1 - Selected Characteristics of Participating Food Stamp Households, 1997 Participating households Average monthly values (dollars) Average household Household type size Number Percent Food stamp Gross Net income (thousands) benefit income (persons) Total 9,452 100.0 169 558 299 2.4 Children 5,508 58.3 234 648 364 3.4 Single-parent households 3,806 40.3 228 576 302 3.1 Multiple-adult households 1,360 14.4 268 917 590 4.5 Other 342 3.6 169 378 162 2.0 Elderly 1,667 17.6 63 577 319 1.3 Living alone 1,292 13.7 47 521 266 1.0 Not living alone 375 4.0 118 767 503 2.5 Disabled 2,108 22.3 104 687 418 2.1 Living alone 1,110 11.7 52 525 244 1.0 Not living alone 998 10.6 162 868 612 3.4 Other 1,307 13.8 118 185 52 1.1 Single-person 1,168 12.4 110 159 38 1.0 Multiple-person 139 1.5 185 407 170 22 Source: 1997 Food Stamp Quality Control sample. Page 8 Table 2 - Gender and Age of Food Stamp Participants, 1997 All participants Female Male Age Number Number Number Percent Percent Percent (thousands) (thousands) (thousands) Total 23,117 100.0 13,880 100.0 9,233 100.0 Children (0-17) 11,871 51.4 5,950 42.9 5,918 64.1 0-4 4,046 17.5 2,017 14.5 2,026 21.9 5-17 7,825 33.8 3,933 28.3 3,892 42.2 Adults (18 or more) 11,219 48.5 7,916 57.0 3,302 35.8 18-35 5,332 23.1 4,007 28.9 1,324 14.3 36-59 4,053 17.5 2,582 18.6 1,472 15.9 60 or more 1,834 7.9 1,328 9.6 506 5.5 Unknown 27 0.1 14 0.1 13 0.1 Source: 1997 Food Stamp Quality Control sample. Table 3 - Race/Ethnicity of Food Stamp Participants, 1997 Participants Race Number Percent (thousands) Total 23,117 100.0 White, Non-Hispanic 9,323 40.3 African-American, Non-Hispanic 8,072 34.9 Hispanic 4,441 19.2 Asian 705 3.0 Native American 313 1.4 Other 263 1.1 Source: 1997 Food Stamp Quality Control sample. Page 9 Table 4 - Monthly Food Stamp Benefits of Participating Food Stamp Households, 1997 Participating households Monthly food stamp benefit Number Percent (thousands) Total 9,452 100.0 - - $0-50 1,726 18.3 $51-100 1,236 13.1 $101-150 2,085 22.1 $151-200 969 10.3 $201-250 1,258 13.3 $251-300 660 7.0 $301 + 1,519 16.1 Source: 1997 Food Stamp Quality Control sample. Table 5 - Selected Economic Characteristics of Participating Food Stamp Households, 1997 Participating Persons in households Average income Average Average households with source (dollars) food house- Income source stamp hold Number Number From benefit size Percent Percent Gross (thousands) (thousands) source (dollars) (persons) Total 9,452 100.0 23,117 100.0 558 (n/a) 169 2.4 Earned income 2,284 24.2 7,533 32.6 879 708 187 3.3 Wages and salaries 2,139 22.6 7,105 30.7 898 728 185 3.3 Self-employment 137 1.5 424 1.8 620 324 225 3.1 Other earned income 35 0.4 106 0.5 817 421 178 3.0 Unearned income 7,415 78.4 18,036 78.0 580 492 163 2.4 AFDC/TANF 3,270 34.6 10,649 46.1 569 372 240 3.3 General Assistance 588 6.2 899 3.9 411 252 123 1.5 Supplemental Security Income 2,504 26.5 4,782 20.7 642 372 93 1.9 Social Security 1,999 21.1 3,377 14.6 647 483 77 1.7 Unemployment 156 1.7 509 2.2 755 491 190 3.3 Other unearned income 1,476 15.6 4,128 17.9 666 211 176 2.8 No Income 868 9.2 1,408 6.1 0 0 180 1.6 Source: 1997 Food Stamp Quality Control sample. Page 10 Table 6 - Selected Characteristics of Participating Food Stamp Households by State, 1997 Average monthly amount Average house- Total Percent of all State Food stamp Total Countable hold (thousands) households Gross income Net income benefit deduction assets size (dollars) (dollars) (dollars) (dollars) (dollars) (persons) Total 9,452 100.0 169 558 299 291 92 2.4 Alabama 186 2.0 174 553 302 292 78 2.5 Alaska 15 0.2 273 928 607 382 117 3.1 Arizona 133 1.4 203 569 305 294 76 2.9 Arkansas 105 1.1 159 568 339 263 163 2.5 California 1,045 11.1 187 594 333 296 122 2.8 Colorado 91 1.0 169 563 277 319 48 2.4 Connecticut 94 1.0 139 596 335 278 135 22 Delaware 20 0.2 175 577 305 309 105 2.6 Dist. of Col. 40 0.4 186 390 201 227 16 2.3 Florida 514 5.4 168 549 276 306 139 2.4 Georgia 284 3.0 177 535 279 298 80 2.5 Guam 6 0.1 408 567 251 396 118 3.3 Hawaii 57 0.6 291 629 355 308 169 2.3 Idaho 27 0.3 175 622 338 322 175 2.7 Illinois 434 4.6 173 476 252 260 61 2.4 Indiana 140 1.5 167 572 308 294 123 2.5 Iowa 67 0.7 148 625 360 290 85 2.4 Kansas 64 0.7 156 586 323 290 125 2.4 Kentucky 174 1.8 165 551 341 243 109 2.6 Louisiana 220 2.3 185 526 282 277 75 2.6 Maine 58 0.6 149 597 269 356 109 2.1 Maryland 152 1.6 184 445 220 264 24 2.4 Massachusetts 149 1.6 148 635 333 318 95 2.3 Michigan 366 3.9 161 579 283 330 57 2.3 Minnesota 110 1.2 139 573 331 270 159 2.2 Mississippi 155 1.6 160 550 332 247 161 2.5 Missouri 199 2.1 161 542 312 264 115 2.4 Montana 27 0.3 171 600 302 325 236 2.5 Nebraska 41 0.4 150 630 361 290 238 2.4 Nevada 38 0.4 165 520 257 305 67 2.3 New Hampshire 21 0.2 116 596 394 239 173 2.2 New Jersey 212 2.2 176 512 224 316 42 2.3 New Mexico 75 0.8 179 571 348 252 107 2.7 New York 899 9.5 150 603 283 346 24 2.2 North Carolina 250 2.6 156 513 289 256 85 23 North Dakota 16 0.2 156 637 364 304 351 2.5 Ohio 389 4.1 137 556 325 253 78 2.2 Oklahoma 131 1.4 168 534 299 277 79 2.5 Oregon 121 1.3 141 516 289 276 249 2.1 Pennsylvania 440 4.7 161 531 265 302 93 2.3 Rhode Island 37 0.4 158 562 304 279 95 2.3 South Carolina 140 1.5 160 550 346 234 104 2.5 South Dakota 18 0.2 185 552 288 313 206 2.6 Tennessee 253 2.7 154 564 312 285 135 2.3 Texas 751 7.9 201 535 296 278 52 2.8 Utah 38 0.4 167 661 375 312 185 2.7 Vermont 25 0.3 125 598 339 286 164 2.1 Virgin Islands 6 0.1 322 490 292 223 93 3.2 Virginia 206 2.2 156 548 309 268 122 2.3 Washington 198 2.1 157 547 272 306 56 2.2 West Virginia 117 1.2 168 509 280 269 90 2.4 Wisconsin 87 0.9 151 689 427 287 148 2.7 Wyoming 11 0.1 174 583 314 313 180 2.6 Source: 1997 Food Stamp Quality Control sample. Page 11 Table 7 - Selected Characteristics of Food Stamp Households Over Time Fiscal Year 1989-Fiscal Year 1995 Total Children Elderly Disabled Children Households Elderly Households Households (percent with children (percent with elderly with disabled Fiscal Participants Households of all (percent of all (percent (percent Year (thousands) (thousands) participants) of all) participants) of all) of all) 1997 23,117 9,452 51.4 58.3 7.9 17.6 22.3 1996 25,926 10,552 51.0 59.5 7.3 16.2 20.2 1995 26,955 10,883 51.5 59.7 7.1 16.0 18.9° 1994 28,009 11,091 51.4 61.1 7.0 15.8 12.5 1993 27,595 10.791 51.5 62.1 6.8 15.5 10.7 1992b 25,743 10,049 51.9 62.2 6.6 15.4 9.5 1991b 22,963 8,855 52.0 60.4 7.0 16.4 9.0 1990b 20,411 7,803 49.6 60.3 7.7 18.1 8.9 19892b 18,925 7,209 49.8 60.4 8.2 19.3 9.1 "Full year analysis files were not developed for the years prior to 1989. ᵇPrior to fiscal year 1993, food stamp cases from Guam and the Virgin Islands were excluded from the analysis files. °Beginning In 1995, disabled households are defined as households with at least one member under age 65 who received SSI, or at least one member age 18 to 61 who received Social Security, veterans benefits, or other government benefits as a result of a disability. For years prior to 1995, disabled households are defined as households with SSI but no members over age 59. The substantial increase in the percentage of households with a disabled member between 1994 and 1995 is due in part to the change in the definition of disabled households. Using the previous definition, 13.3 percent of households included a disabled person in fiscal year 1995. Source: Food Stamp Quality Control samples. FSP Participants: FY 1988 - 1998 29,000,000 27,000,000 25,000,000 23,000,000 People 21,000,000 19,000,000 17,000,000 15,000,000 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998* Year 12/1/98 Who is Leaving the Food Stamp Program? Detailed Table Participants Participants Paticipation Percent Percent of Subgroup as Subgroup in 1994 in 1997 Change in Decline from % of 1997 (In thousands) (in thousands) (in Participation Subgroup Legal Permanent Resident Aliens 5-6% 1,537 706 831 54.07% 14.01% Childless Unemployed 4,2% 1,148 648 500 43.55% 8.43% Adults AFDC/TANF Reciplents 13,052 9,442 47.5% 3,610 27.66% 60.87% All Other Participants 11,697 10,707 990 8.46% 16.69% 39% Total Participants 27,434 21,503 5,931 21.62% 100.00% Notes: 1. Participation obtained from July-September Food Stamp Program Quality Control data for 1994 and 1997. 2. All subgroups are mutually exclusive. Childless unemployed adults do not include Legal Permanent Resident aliens. Likewise, AFDC/TANF recipients do not include aliens or childless unemployed adults. SINGLE PARENT FAMILIES: 1989, 1994, and 1997 Number of Single Parent Families on Food Stamps by Welfare Receipt and Presence of Earnings: 1989, 1994, and 1997 5,000 DRAFT 4,500 568 4,000 708 3,500 614 297 3,000 Households (in thousands) 324 780 2,500 334 208 325 2,000 1,500 3,022 1,000 2,094 2,097 500 0 1989 1994 1997 Year Single parent households with no earnings and no AFDC/TANF Single parent households with earnings and no AFDC/TANF Single parent households with AFDC/TANF and earnings Single parent households with AFDC/TANF and no earnings 12/1/98 Changes in the Number of Single Parent Participants on Welfare and Not on Welfare 3,500 DRAFT 3,000 2,500 Number (in thousands) 2,000 1,500 1,000 500 0 1989 1994 1997 Single parent households on welfare Single parent households not on welfare 12/1/98 includes is 490mp I "day" TWO GROUPS AFFECTED BY WELFARE REFORM: LEGAL IMMIGRANTS AND UNEMPLOYED CHILDLESS ADULTS AGED 18-59 1989, 1994, and 1997 Decline in participation among legal immigrants and childless unemployed adults in the wake of welfare reform 1,600 DRAFT 1,400 1,200 1,000 Number (in thousands) 800 600 400 200 0 July-96 August-96 September-96 October-96 November-96 December-96 January-97 February-97 March-97 April-97 May-97 June-97 July-97 August-97 September-97 Permanent Resident Aliens Unemployed Childless Adults 12/1/98 Desired Dres not sholfund 3 Goof 150mg road wood Immigrant Participants by Citizen/Alien Status: 1989, 1994, and 1997 2,500 DRAFT 2,000 339 1,500 Number 235 1,000 1,453 234 547 500 582 437 264 131 0 1989 1994 1997 Year Naturalized citizens Permanent Residents Refugees 12/1/98 Changes in the Number of Childless Unemployed Adult Participants: 1989, 1994, and 1997 DRAFT 1,400 1,309 1,200 1,000 833 800 Number 652 600 400 200 0 1989 1994 1997 Year 12/1/98 ELDERLY FOOD STAMP PARTICIPANTS 1989, 1994, and 1997 Number of Aged Food Stamp Participants: 1989, 1994, and 1997 2,000 DRAFT 1,800 1,600 1,400 1,200 Number (in thousands) 1,000 800 600 400 200 0 1989 1994 1997 Year 12/1/98 CHANGES IN THE RACIAL/ETHNIC COMPOSITION OF THE FOOD STAMP CASELOAD 1989, 1994, and 1997 The Racial/Ethnic Composition of FSP Participants: 1994 and 1997 FY 1994 Other 5% RAFT Hispanic 19% White 41% Black 35% FY 1997 Other 4% Hispanic 19% White 41% Black 36% 12/1/98 Racial Composition of AFDC/TANF Participants: 1994 and 1997 FY 1994 Other 7% DRAFT Hispanic White 21% 34% Black 38% FY 1997 Other 6% Hispanic White 22% 33% Black 39% 12/1/98 Welfore Reups pereary f S, Racial Composition of Childless Unemployed Adults: 1994 and 1997 FY 1994 DRAFT Other Hispanic 4% 7% White 44% Black 45% FY 1997 Other 3% Hispanic 14% White 37% Black 46% 12/1/98 TABLES WITH UNWEIGHTED QUALITY CONTROL DATA FROM 1989, 1994, 1997, and PRELIMINARY 1998 FILES Sources of Income Using Unweighted Quality Control Data: 1989, 1994, 1997, and 1998 preliminary data 45.00% DRAFT 40.00% 35.00% 30.00% 25.00% Percent 20.00% 15.00% 10.00% 5.00% 0.00% 1989 1994 1997 1998 Year Welfare for Families 'Welfare for Single Adults Supplemental Security Income Social Security Earnings 12/1/98 Percent of Caseload that are Legal Immigrants or Childless Unemployed Adults (Using Unweighted Data): 1989, 1994, 1997, and 1998 Preliminary Data DRAFT 4.00% 3.50% 3.00% 2.50% Percent 2.00% 1.50% 1.00% 0.50% 0.00% 1989 1994 1997 1998 Year Aliens ABAWDs 12/1/98 Jenny genser 703- 305- 2152 Jenny genser @fcs. @ usda . gov

Page data

Page
102
Source index
0
Type
document
Media ID
818a99ebb79d8dee
Size
unknown

Document data

ID
26412201
Core
doc
Type
document
DTO data 
{
    "id": "26412201",
    "sourceUrl": "https://catalog.archives.gov/id/26412201",
    "contentType": "document",
    "title": "Welfare - Food Stamps - Caseloads I [1]",
    "citationUrl": "https://catalog.archives.gov/id/26412201",
    "collections": [
        "Records of the Domestic Policy Council (Clinton Administration)",
        "Cynthia Rice's Subject Files"
    ],
    "iiifBase": "https://s3.amazonaws.com/NARAprodstorage/opastorage/live/1/4122/26412201/content/presidential-libraries/clinton/foia/2009-1414-S/2009-1414-S-JPG/Box_038/42-t-647851-20091414-S-038-006-2015/42_t_647851_20091414_S_038_006_2015_Page_001.JPG",
    "thumbnailUrl": "https://s3.amazonaws.com/NARAprodstorage/opastorage/live/1/4122/26412201/content/presidential-libraries/clinton/foia/2009-1414-S/2009-1414-S-JPG/Box_038/42-t-647851-20091414-S-038-006-2015/42_t_647851_20091414_S_038_006_2015_Page_001.JPG",
    "largeImageUrl": "https://s3.amazonaws.com/NARAprodstorage/opastorage/live/1/4122/26412201/content/presidential-libraries/clinton/foia/2009-1414-S/2009-1414-S-JPG/Box_038/42-t-647851-20091414-S-038-006-2015/42_t_647851_20091414_S_038_006_2015_Page_001.JPG",
    "imageCount": 102,
    "hasImages": true,
    "source": "import",
    "hasTranscription": false
}

Context sent to Scholar

Document identity 
{
    "localId": "26412201",
    "label": "Welfare - Food Stamps - Caseloads I [1]",
    "core": "doc",
    "dtoType": "document",
    "citationUrl": "https://catalog.archives.gov/id/26412201"
}
Document source metadata 
{
    "id": "26412201",
    "sourceUrl": "https://catalog.archives.gov/id/26412201",
    "contentType": "document",
    "title": "Welfare - Food Stamps - Caseloads I [1]",
    "citationUrl": "https://catalog.archives.gov/id/26412201",
    "collections": [
        "Records of the Domestic Policy Council (Clinton Administration)",
        "Cynthia Rice's Subject Files"
    ],
    "iiifBase": "https://s3.amazonaws.com/NARAprodstorage/opastorage/live/1/4122/26412201/content/presidential-libraries/clinton/foia/2009-1414-S/2009-1414-S-JPG/Box_038/42-t-647851-20091414-S-038-006-2015/42_t_647851_20091414_S_038_006_2015_Page_001.JPG",
    "thumbnailUrl": "https://s3.amazonaws.com/NARAprodstorage/opastorage/live/1/4122/26412201/content/presidential-libraries/clinton/foia/2009-1414-S/2009-1414-S-JPG/Box_038/42-t-647851-20091414-S-038-006-2015/42_t_647851_20091414_S_038_006_2015_Page_001.JPG",
    "largeImageUrl": "https://s3.amazonaws.com/NARAprodstorage/opastorage/live/1/4122/26412201/content/presidential-libraries/clinton/foia/2009-1414-S/2009-1414-S-JPG/Box_038/42-t-647851-20091414-S-038-006-2015/42_t_647851_20091414_S_038_006_2015_Page_001.JPG",
    "imageCount": 102,
    "hasImages": true,
    "source": "import",
    "hasTranscription": false
}
Document source extras 
{
    "url": "https://catalog.archives.gov/id/26412201",
    "naId": 26412201,
    "levelOfDescription": "fileUnit",
    "otherTitles": [
        "42-t-647851-20091414-S-038-006-2015"
    ],
    "recordType": "description",
    "ocrSource": "nara-archive"
}
Page context 
{
    "seq": 102,
    "pageIndex": 0,
    "type": "document",
    "url": "https://s3.amazonaws.com/NARAprodstorage/opastorage/live/1/4122/26412201/content/presidential-libraries/clinton/foia/2009-1414-S/2009-1414-S-PDF/Box_038/42-t-647851-20091414-S-038-006-2015.pdf",
    "mediaId": "818a99ebb79d8dee",
    "ocrText": "Who Receives Food Stamps Today?\nMost food stamp recipients are children or elderly. Over half (51\npercent) are children and another 8 percent are age 60 or older. Working-\nage women represent 29 percent of the caseload, while working-age men\nrepresent 12 percent.\nThe majority of food stamp households do not receive AFDC or\nTANF benefits. Only 35 percent do so. Other cash assistance received\nby food stamp households includes Supplemental Security Income (27\npercent of households), Social Security (21 percent) and State General\nAssistance benefits (6 percent). Nine percent of households have no\nincome of any kind.\nMany food stamp recipients work. Twenty-four percent of food stamp\nhouseholds have earnings, and for these households, earnings are the\nprimary source of income.\nFood stamp households have little income. Only 9 percent are above\nthe poverty line, while 40 percent have incomes at or below half the\npoverty line. The typical food stamp household had gross income of $558\nper month and received a monthly food stamp benefit of $169. Nearly\none-fourth of monthly available funds (cash income plus food stamps)\navailable to a typical household comes from food stamps.\nFood stamp households possess few resources. The average food stamp\nhousehold possesses only about $92 in countable resources (including\nvehicles, checking and savings accounts, and other savings).\nMost food stamp households are small. The average food stamp\nhousehold size was 2.4, but varied considerably by household\ncomposition. Households with children were relatively large, averaging\n3.4 members. Households with elderly members tended to be smaller,\nwith an average size of 1.3 people.\n12/1/98 FY 1997 data\nFAX COVER SHEET\nOFFICE OF ANALYSIS AND EVALUATION\nFOOD AND CONSUMER SERVICE, USDA\n(Office: 703-305-2019)\n(Fax: 703-305-2576)\nDATE: 3-26-99\nTO:\nERIC GOULD\nOFFICE:\nPHONE #:\nFAX PHONE #:\nFROM: STEVEN CARLIUN\nPHONE:\n# PAGES (Including cover page):\n9\nCOMMENTS:\nSECOND TRY\nse\n6/T'd S60'ON\nWH88 :8\n6661. '97 \"AHW\nWho Leaving the Food Stamp Program?\nAn Analysis of Caseload Changes from 994'to 1997\nUnited States Department of Agriculture\nOffice of Analysis, Nutrition, and Evaluation\nFood and Nutrition Service\nMarch 1999\nBackground\nThe number of people receiving food stamps\nThese questions cannot be answered\nfell by over 5.9 million\nbetween summer\nconclusively yet. However, an analysis of\n1994 and summer 1997,\nwith\nmost\nof\nthe\nfood stamp administrative data provides\ndecline occurring in the year between\nsome initial insights into the changes that\nSeptember 1996 and September 1997. This\nhave occurred in the Food Stamp Program.\ndecline occurred during a period of strong\nThis analysis examines FSP administrative\neconomic growth - unemployment fell,\ndata from 1994, when FSP caseloads\ninflation stayed low, and the percentage of\npeaked, and 1997, the most recent year when\nAmericans living in poverty fell slightly. In\nfull data are available. 1997 was a transition\nthe same period, Congress enacted and\nyear as States replaced the Aid to Families\nStates implemented sweeping reforms to the\nwith Dependent Children (AFDC) program\nFood Stamp Program (FSP) and to the\nwith Temporary Assistance for Needy\nnation's welfare programs.\nFamilies (TANF) and when changes to the\nFSP took effect. Thus this report provides\nSome point to the participation decline as\ninformation about caseload changes during\nproof that the strong economy is lifting all\nthe early stages of implementation This\nboats, providing job opportunities and higher\nanalysis focuses on those groups most\nwages for all, including low-income families\naffected by welfare reform - single parents,\nWelfare reform is credited with moving\nlegal immigrants, and unemployed childless\nmillions of families from dependence on the\nadults.\nstate to work.\nGeneral Trends\nHowever, others raise more\ntroubling\nquestions. Suppose familles are not leaving\nAs Table 1 shows, three major groups\nthe safety net for paid employment and self-\naccounted for almost all the drop in the food\nsufficiency but rather\nremain\npoor\nbut\nstamp caseload between 1994 and 1997.\nwithout assured access to\nsufficient food to\nThe number of legal immigrants fell by 54\nmeet their basic needs?\nSuppose people\npercent, accounting for 14 percent of the\nleaving welfare remain\neligible for food\ntotal decline. The number of childless\nstamps but don't realize\nthat these benefits\nunemployed adults fell by 44 percent,\nare still available to them?\nThese critics cite\naccounting for 8 percent of the total decline.\nthe increased demand for\nassistance at food\nFinally, the number of families receiving\npantries and soup kitchens as evidence that\nwelfare benefits fell by 28 percent. Because\ndeclining caseloads are not necessarily a sign\nthis group accounts for such a large share of\nof success.\nthe FSP caseload, they represented 61\nThis report was prepared by Jenny Genser. Office of Analysis. Nutrition, and Evaluation, based on data prepared\nby Scott Cody and Laura Castner of Mathematica Policy Research, Inc. The information presented in this report is\nbased on data collected by Food Stamp Program Quality Control data for fiscal years 1994 and 1997.\n6.2%\nS60'ON\nW988:8\n6661\npercent of the decline. Only 17 percent of\nwhile the number of single-parents receiving\nthe decline came from other groups,\nAFDC or TANF benefits fell overall, the\nincluding elderly, the\ndisabled,\nand\nnumber of those who combined welfare with\nhouseholds with earnings\nand no welfare\nwork also rose by 9 percent. This points to\nincome.\nan increased reliance on work among single\nparents.\nThus, the steepest declines in participation\noccurred among legal\nimmigrants\nand\nHowever, it is important to note that the\nunemployed childless adults, the two groups\nnumber of single-parent households with no\naffected by tough new restrictions under\nincome from either welfare or work also rose\nwelfare reform. However, most of the\nby 9 percent. Also, the number of single-\nreduction occurred among the large group of\nparent food stamp households receiving\ncash welfare households.\nwelfare dropped by nearly 900,000, while the\nnumber not receiving welfare rose by only\nTrends Among Single Parent Families\n120,000. Because we only have data on\nfood stamp participants, we don't know\nAmong other changes, welfare reform was\nwhether the households leaving both welfare\ndesigned to move families from welfare to\nand food stamps found jobs and are\nwork. The Aid to Families with Dependent\neconomically independent, or are\nProgram (AFDC), which was an entitlement\nunemployed and in need.\nfor needy single parents with children, was\nreplaced by block grants to States to operate\nTrends Among Legal Immigrants\nthe Temporary Assistance to Needy Families\n(TANF), which was designed to provide\nWelfare reform legislation made most legal\nshort term assistance to help poor families\nimmigrants ineligible for food stamps. Those\nbecome economically self-sufficient\nwho were participating at the time the law\nwent into effect could participate until\nAs Table 2 shows, the number of single-\nSeptember 1997. New applicants became\nparent families, the group most affected by\nineligible starting October 1996.\nwelfare reform, fell by 17 percent. Within\nthis group, the number receiving AFDC or\nIn 1994, nearly 1.5 million legal immigrants\nTANF benefits fell by 27 percent but the\nreceived food stamps. This number dropped\nnumber of those not receiving welfare rose\nsharply after welfare reform was enacted.\nby 9 percent. This suggests that welfare\nThe number of legal immigrants receiving\nreform has been partially responsible for the\nfood stamps declined steadily throughout\ndrop in welfare caseloads Had it been solely\nlate 1996 and most of 1997 (Figure 1). The\nthe strong economy, the\nnumber\nof\nfood\ndecline was gradual throughout the year,\nstamp households with welfare and without\nrather than falling sharply between August\nwelfare would have both fallen by similar\nand September 1997. This indicates that as\namounts. It also suggests that at least some\ncurrent immigrants left the program, they\nfamilies that no longer\nreceive welfare\nwere not replaced by new immigrant\ncontinue to receive food stamps.\nparticipants.\nAmong single-parent families, the number of\nRestrictions on participation by legal\nthose with earnings rose by 10 percent And\nimmigrants appear to have deterred\n2\n6/8'd\nNO.095\nUSDA/FCS/OAE\nWHAR:8\n6661\nparticipation by their children, many of\ncaseload that they represent fluctuates.\nwhom retained their eligibility for food\nDuring periods of caseload expansion,\nstamps. Participation among U.S. born\nelderly households represent a declining\nchildren living with their legal immigrant\nshare of the food stamp population. During\nparents fell faster than participation among\nperiods of caseload decline, they represent a\nchildren living with native-born parents\nlarger share.\n(Table 3). The number of children living\nwith legal immigrants fell by 37 percent,\nChanges in Ethnic/Racial Composition\nversus 15 percent for children living with\nnative-born parents.\nSome have raised a concern that those\nfinding work and leaving welfare are\nThe number of naturalized citizens receiving\npredominantly non-Hispanic whites, leaving\nfood stamps rose by 173,000 between 1994\nthe welfare caseload even more\nand 1997, an increase of 66 percent (Table\ndisproportionately minority. The data do not\n4). This reflects the surge in naturalizations\nsupport this claim. The racial composition of\nstarting in 1993.\nthe food stamp caseload as a whole is\nvirtually unchanged between 1994 and 1997\nTrends Among Childless Unemployed\n(Figure 3), despite steep drops in\nAdults\nparticipation overall.\nWelfare reform restricted most childless\nThe same is true for AFDC/TANF recipients.\nunemployed adults to no more than three\nWhites left the food Stamp Program and\nmonths of food stamps in a\n36-month\nperiod,\nwelfare at the same rate as minorities. The\nunless they were employed or participating\nin\nnumber of whites receiving welfare and food\nqualified work programs.\nMany parts of the\nstamps between 1994 and 1997 fell by 32\ncountry were exempt from the work\npercent, compared to a 31 percent drop for\nrequirement and time limit, due to waivers\nAfrican Americans and a 27 percent drop for\ngranted to areas with high unemployment\nHispanics. As a result, the caseload\nrates or insufficient jobs.\ncomposition is very similar for both years.\nAs expected, the number of unemployed\nThe one group where the racial/ethnic\nchildless adults fell by 476,000, a drop of\ncomposition did change significantly is the\none third, between August 1996 and\nunemployed childless adult group subject to\nSeptember 1997 (Figure 2). This decline\ntime limits. While the proportion of blacks\nwas sharpest in the period between January\nin this group remained unchanged between\nand March, 1997, as States implemented the\n1994 and 1997, the proportion of whites fell,\ntime limits.\nwhile the proportion of Hispanics doubled.\nHowever, this is a relatively small group that\nTrends Among the Elderly\nmay be heavily affected by State waiver\npolicy.\nThe number of households with aged\nmembers dropped by 86,000 from 1994 to\nAbout the Data\n1997, a decline of less than five percent.\nOver time, the number of elderly receiving\nThe data come from Food Stamp Quality\nbenefits is very steady, while the share of the\nControl records. The cases are derived from\n3\nS60'ON\nW868:8\nMAR. 26. 1999\nState samples of caseloads pulled each\nmonth for a review on payment accuracy.\nRecords from all States for all months during\na fiscal year are combined into one file. Each\nyear, there are about 50,000 households\nrepresented in the data.\nThe files are then\nedited for consistency and weights are\nassigned.\nWe analyzed data from two years -- 1994,\nthe year participation peaked; and 1997, the\nmost recent year that we have complete data.\nLimitations of the Analysis\nThe analysis compares snapshots of the food\nstamp population at different points in time.\nIt does not follow individuals or families\nover a course of time, showing their\nmovements on and off jobs, welfare, or food\nstamps. Nor does this analysis provide any\ninformation about low-irlcome households\nnot receiving food stamps including former\nparticipants.\n4\n6/S'd\n560'0N\nUSDA/FCS/OAE\nWHAD:8\n6661 '97 XHW\nTABLE 1\nParticipation Changes from 1994 to 1997\n(in thousands)\nParticipants:\nParticipants\nParticipation\nPercent\nShare of\nSummer 1994\nSummer 1997\nChange\nChange\nDecline\nLegal\nPermanent\nResidents\n1,537\n706\n-831\n54%\n14%\nChildless\nUnemployed\nAdults\n1,148\n648\n-500\n-44%\n8 %\nAFDC/TANF\nParticipants\n13,052\n9,442\n-3,610\n-28%\n61%\nAll Other\nParticipants\n11,697\n10,707\n-990\n-8%\n17%\nTOTAL\n27,434\n21,503\n-5,931\n-22%\n100%\nTABLE 2:\nSingle Parent Households by Presence of Earnings and Welfare Receipt: 1994 and 1997\nHouseholds:\nHouseholds:\nParticipation\nPercent\n1994\n1997\nChange\nChange\nAll Single Parents\n4.595\n3,816\n-779\n-17%\nWith AFDC/TANF\n3,319\n2,422\n-897\n-27%\nWithout earnings\n3.022\n2,097\n-925\n-31%\nWith earnings\n297\n325\n28\n9 %\nWithout AFDC/TANF\n1.276\n1,394\n118\n9%\nWithout earnings\n568\n614\n46\n8 %\nWith earnings\n708\n780\n72\n10%\nWith earnings\n1.005\n1,10S\n100\n10%\n5\n6/9'd\nS60'ON\nUSDA/FCS/OAE\nWH00:8\n6661.92'86W\nFIGURE 1:\nLegal Immigrant Participants:\nAugust 1996 through September 1997\n1.800\n, 400\n,00\n000\nBOD\nNumber on\n000\n400\n200\n0\n$\na\nMy\nwas\nSaptember-06\nOctober-Df\nHomema 8\nDer Derector's 10\n-\nFatury\nWith\nApril #\nMay 187\nAre-91\nand\nA\nSectimes BY\n.\nTABLE 3:\nNumber of Children Participating by Citizenship Status of Parents:\nOctober 1996 and September 1997\n(in thousands)\nParticipants:\nParticipants:\nParticipation\nPercent\nOctober 1996\nSeptember 1997\nChange\nChange\nChildren Living\nwith Legal\nImmigrants\n1,251\n742\n434\n37%\nChildren Not\nLiving with Legal\nImmigrants\n11,034\n9,804\n1,682\n15%\n6\nP.7/9\nS60'ON\nWHOD 8\nMAR. 26. 1999\nTABLE 4:\nNumber of Legal Immigrants by Status: 1994 and 1997\n(in thousands)\nParticipants:\nParticipants:\nParticipation\nPercent\n1994\n1997\nChange\nChange\nPermanent\nResident Aliens\n1,453\n547\n-906\n- 62 %\nRefugees\n359\n235\n-104\n- 31 %\nNaturalized\nCitizens\n264\n437\n+173\n66%\nAll Legal\nImmigrants\n2,056\n1,219\n-837\n- 41 %\nFigure 2:\nChildless Unemployed Adults:\nAugust 1996 through September 1997\n1.200\n1 000\n800\nUnemployed Chedres AGUES\nHumber in thousands)\n600\n400\n200\nO\nX &\nAugust 96\nSeptember 96\nDctober-96\nNovember 96\nCocamber =\n= Amount\n= Fathery\n= Byd\nApril\nMay91\nJune-97\nrev-97\nAugust-97\nSeptember-97\n- ---\n7\n6/8'd\nS60'ON\nUSDA/FCS/OAE\nMAR.26.1999 8:41AM\nFigure 3\nRacial Composition of Food Stamp Participants Receiving Welfare: 1994 and 1997\nOther\nFY 1994\n5%\nHispanic\n19%\nWhite\n41%\nBlack\n35%\nOther\nFY 1997\n4%\nHispanic\n19%\nWhite\n41%\nBlack\n36%\n8\n6/6'd d\nS60'ON\n8:41AM\nMAR.26.1999\nFAX COVER SHEET\nOFFICE OF ANALYSIS AND EVALUATION\nFOOD AND CONSUMER SERVICE, USDA\n(Office: 703-305-2019)\n(Fax: 703-305-2576)\nDATE: 5-25-99\nTO: EXIC GOULA\nOFFICE: DPC\nPHONE #:\nFAX PHONE #: (202)456-7431\nFROM: STEVEN CARLSON\nPHONE: (703) 305-2134\n8\n# PAGES (Including cover page):\nCOMMENTS:\nERIC: I JUST GOT YOUR 5/20 B-MAIL ASKING\nFOR FAXOO COPY OF MATORIAL on PARTICIPATION\nDECLINE. HERE. IS Thto ENTIRE PACK760 some\nOF UNICH I FAX50 TO YOU AND KOWIN LAST .\nTHURSDAY. AA\nse\nNO.024 P.1/8\nWH80:8\nNOT FOR ATTRIBUTION, CITATION, OR PUBLICATION\nMay 20, 1999\n-DRAFT--\nUnderstanding Food Stamp Program Caseload Decline\nIn February 1999, the Food Stamp Program served 18.29 million people, the fewest in nearly 20\nyears. The number of people receiving food stamps over the past five years has fallen by 9.7\nmillion, a drop of over one-third. This trend far outpaces the decline in poverty. Between 1995\nand 1997, food stamp participation fell five times as fast as poverty. A forthcoming GAO report\nwill note that the decline in the number of children participating in the program between 1994 and\n1997 substantially outpaced the decline in the number of children living in poverty, resulting in a\nsubstantial fall in the percentage of poor children receiving food stamps.\nAt least five factors contribute to the decline in food stamp participation:\nChanges in program rules under welfare reform restricted the participation of immigrants\nand unemployed childless adults.\nThe strength of the Nation's economy allowed some participants to find work, reducing their\nneed for food stamps.\nThe success of the Temporary Assistance for Needy Families (TANF) Program moved\nsome participants from welfare to work, with an increase in income sufficient to eliminate the\nneed for food stamps.\nBarriers to participation erected by some States and localities may have deterred some\neligible individuals from applying, or resulted in some participants leaving the program despite\ntheir continued eligibility.\nLack of awareness of eligibility for the program may have led some participants to leave the\nprogram unnecessarily and discouraged others from applying for benefits.\nThere is little reason for concern about falling participation if it is largely due to the first three of\nthese factors - the expected consequences of economic growth and welfare reform. There is\ngreater reason for concern and a potential need to explore policy options to redress problems to\nthe extent that falling participation is attributed to the last two factors.\nUnderstanding the relative importance of these factors represents a complex analytical program.\nBy piecing together information from a variety of sources (with different degrees of reliability),\nwe can begin a preliminary assessment of their contribution. This analysis is limited to the period\nbetween March 1994, the month in which food stamp participation peaked, and September 1997,\nthe last month for which we have relatively complete information Over this period, food stamp\nparticipation dropped by just under 7 million people.\nAs shown in Table 1, this preliminary assessment suggests that roughly 20 percent of the decline\nin food stamp participation over this period was due to new limits on immigrants and unemployed\njobless adults, roughly 30 percent was due to success of economic expansion and welfare reform\nin moving people into jobs, roughly 32 percent was due to a decline in participation rates among\notherwise eligible individuals, and roughly 18 percent was left unexplained.\n8/2'd\nNO.024\nUSDA/FCS/OAE\nWH82:8\n6661\nTable 1\nExplaining the Drop in Food Stamp Participation: March 1994 - September 1997\nNumber of\nRelative\nParticipants\nContribution\nSource\nNet change in participation\n6,974,000\n100%\nFNS administrative records\nWelfare reform limits on:\nImmigrants\n1,053,000\n15\nFNS administrative records\nUnemployed adults\n327,000\n5\nFNS administrative records\nEmployment growth\n2,117,000\n30\nFNS forecast model\nLower participation rates\n2,198,000\n32\nPreliminary estimates from\nCurrent Population Survey\nAll other/unknown\n1,279,000\n18\nChanges in Program Rules\nThe food stamp provisions of welfare reform restricted the participation of many legal immigrants\nand unemployed childless adults. The effect of these provisions can be seen in the drop in\nparticipation among these groups in the period immediately following enactment of the 1996\nwelfare reform law. From September 1996 until September 1997, 1,053,000 million legal\nimmigrants and 327,000 childless adults ceased to participate in the Food Stamp Program.\nTogether, these two groups account for 20 percent of the March 1994 through September 1997\ndecline.\nThere may also have been a spill-over effect of the restrictions on legal immigrants. From\nOctober 1996 to September 1997, the number of U.S.-born children of legal immigrants receiving\nfood stamps fell by 509,000, a drop of 41 percent, despite the fact that these children remained\neligible. Had the number of these children fallen at the same rate as children of citizen parents,\n370,000 more children of immigrants would have participated in September 1997.\nEmployment Growth\nThe sustained strength of the U.S. economy has created millions of new jobs and welfare reform\nhas successfully moved many from welfare to work. As income rises, fewer people are eligible\nfor benefits; and among those who remain eligible, some may feel less need to seek government\nassistance. One model of the relationship between unemployment and participation suggests that\nabout 30 percent of the decline in food stamp participants between March 1994 and September\n1997 (2,117,000 people) may be traced to the reduction in the number of unemployed over the\n2\n8/8'\n200.00\n380/SD H/HASA\nWH87:8\nFAST .\nsame period. It is not possible to separate the effects of a strong economy from the independent\neffects of welfare reform's success.¹\nThere is reason to celebrate the significant expansion of job opportunities among low-income\nhouseholds. There is also reason to be cautious. Because baseline levels of employment and\nearnings for welfare participants were so low, even substantial improvements leave most families\nexiting welfare poor. And, as yet, there are few signs of substantial improvement in the count of\nfamilies and individuals officially poor. With the food stamp income limit set at 130 percent of the\npoverty line, many newly working families may continue to be eligible for food stamp benefits.\nFalling Participation Rates\nNot all persons eligible for benefits actually participate in the program. Falling participation rates\nmay be symptomatic of barriers to participation erected by some States or localities² or a lack of\nawareness of continued eligibility by some. Preliminary evidence suggests that participation rates\namong those who are neither non-citizens nor unemployed childless adults fell significantly over\nthis period (from 82 percent in 1994 to 69 percent in 1997), accounting for 32 percent (or\n2,198,000 people) of the caseload decline. Table 2 shows that the drop in participation rates was\nsharpest among households with children and more than one adult.\nThe Characteristics of Leavers\nAs Table 3 shows, three major groups accounted for almost all the drop in food stamp caseload\nbetween 1994 and 1997. The number of legal immigrants fell by 54 percent, accounting for 14\npercent of the total decline. The number of childless unemployed adults fell by 44 percent,\naccounting for 8 percent of the total decline. Finally, the number of persons in families receiving\nwelfare benefits fell by 28 percent. Because they make up such a large share of the food stamp\ncaseload, they account for 61 percent of the overall decline. Thus, the steepest declines in food\nstamp participation occurred among legal immigrants and unemployed childless adults, the two\ngroups whose eligibility for food stamps was directly curtailed by the 1996 welfare reform law.\nHowever, most of the reduction occurred among the large group of families receiving cash\nassistance.\nI\nThis estimate is based on a simple forecast model relating the number of unemployed in the current and lagged\nquarter to the number of participants with a set of seasonal dummy variables. While not developed to explain\nhistorical participation levels, it suggests the magnitude of the relationship between changes in unemployment and\nchanges in program participation. This estimate is consistent with Wallace and Blank (1999) who suggest that\ndeclining unemployment rates can explain 28 to 44 percent of the food stamp caseload decline through 1996.\n2 During the past year, FNS regional offices have reviewed access to programs in New York, Oregon, and\nWisconsin in response to complaints from advocacy groups and a member of Congress. Each review has identified\nbarriers that make the Food Stamp Program less accessible to cligible people.\n3\nP.4/8\n20 ON\n360/S0 H/HASO\nWH67:8\n6661\nTable 2\nParticipation Rates Among Those in Households\nwith No Non-citizens or Unemployed Childless Adults\nNumber of Eligibles\nParticipation Rate\nIndividuals in:\n1994\n1997\n1994\n1997\nChange\nHouseholds with single parents\n12,995,000\n12,118,000\n95.8\n86.2\n-9.6\nwith earnings\n4,166,000\n4,893,000\n63.0\n67.7\n4.7\nwithout earnings\n8,829,000\n7,225,000\n111.3\n98.7\n-12.6\nOther households with children\n7,550,000\n6,887,000\n96.3\n74.8\n-21.5\nwith earnings\n4,667,000\n4,460,000\n77.2\n64.0\n-13.1\nwithout earnings\n2,883,000\n2,427,000\n127.3\n94.6\n-32.6\nHouseholds with elderly\nmembers and no children\n5,274,000\n5,261,000\n30.6\n27.9\n-2.7\nOther households\n2,681,000\n3,222,000\n70.2\n57.3\n-13.0\nAll Individuals\n28,500,000\n27,489,000\n81.5\n68.9\n-12.6\nSource: Food Stamp Program Quality Control (1994 and 1997); Current Population Survey (March 1995 and\nMarch 1998).\nTable 3\nComposition of Food Stamp Caseload: 1994 - 1997\nNumber of Participants\nSummer\nSummer\nChange\nPercent\nShare of\n1994\n1997\nChange\nDecline\nLegal permanent residents\n1,537,000\n706,000\n-831,000\n-54%\n14%\nChildless unemployed adults\n1,148,000\n648,000\n-500,000\n-44%\n8 %\nAFDC/TANF participants\n13,052,000\n9,442,000\n-3,610,000\n-28%\n61%\nAll other participants\n11,697,000\n10,707,000\n-990,000\n-8%\n17%\nAll participants\n27,434,000\n21,503,000\n-5,931,000\n-22%\n100 %\nSource:\nReproduced from Table 1 in Who is Leaving the Food Stamp Program? An Analysis of Caseload\nChanges from 1994 to 1997, Food and Nutrition Service, USDA, March 1999.\n4\n8/S'd\nNO.024\n360/SD3/60S0\nWHAT:8\n6661\n- - DRAFT- -\nP.6/8\nExplaining the Drop in Food Stamp Participation: March 1994 - September 1997\nFactor\nEffect\nContribution\nSource/Derivation\n(000's)\nNO.024\nFNS National Data Bank (number of participants):\nParticipation Change\n6,974\n100%\nMarch 1994\n27,965\nSeptember 1997\n20,991\nChange\n6,974\nCharacteristics of Food Stamp Households: Fiscal Year 1997, Tables B-1, B-3:\nWelfare Reform\nPermanent Resident\nUnemployed Childless\nLimits on:\nAliens\nAdults\nImmigrants\n1,053\n15\nSeptember 1996\n1,600\n989\nSeptember 1997\n547\n662\nUnemployed\n327\n5\nChange\n1,053\n327\nAdults w/o\nchildren\nDerived from FNS forecast model (based on current and lagged quarterly\nnumber of seasonally-unadjusted unemployed and seasonal dummy variables).\nEmployment Growth\n2,117\n30\nModel implies a net change of 83 participants for every 100-person change in\nDue to Economic\nthe number unemployed:\nMAY.25.1999 8:50AM USDA/FCS/OAE\nStrength and\nNumber Unemployed\nWelfare Reform\n1994:Q2\n9,209\n1997:Q4\n6,659\nChange\n2,550 X .83 = 2,117\nNote: Wallace and Blank (1999) suggest that declining unemployment rates can\nexplain about 28 to 44 percent of the food stamp caseload decline through\n1996.\nWorking Draft: Do Not Attribute, Cite or Publish\nP.7/8\nApproximates change in participation decisions of persons eligible for benefits\nindependent of economic improvements. Assumes improved economy (1) reduces\nParticipation\n2,198\n32\nthe number of persons eligible for benefits and (2) reduces the perceived need\nNO.024\nReductions among\nfor assistance for some, reducing participation among otherwise eligible people.\nEligibles\nIndependent of\nDerived from preliminary CPS-based estimates of participation rates for\nEconomic Growth\nindividuals in households with neither noncitizens nor unemployed adults without\nchildren (since participation effects of welfare reform restrictions on these\ntwo groups are captured earlier):\nEligible\nParticipating\nParticipation Rate\nAugust 1994\n28,500\n23,223\n81.5\nSeptember 1997\n27,489\n18,912\n68.8\nChange\n1,011\n4,311\n12.7\nReduction in participation due to falling number of eligibles (arguably due to\nimproved economy):\n1,011 X .815 = 824\nReduction in participation due to falling participation rates (some of which may\nbe due to improved economy): 27,489 X 127 = 3,491\n8:50AM USDA/FCS/OAE\nTotal reduction (equal to 824 + 3,491)\n4,315\nPreviously assigned to economy\n2,117\nNet reduction\n2,198\nAll Other/Unknown\n1,279\n18\nMAY.25.1999\nMay 19, 1999\nDRAFT\nTABLE I\nP.8/8\nFSP PARTICIPATION RATES IN 1994 AND 1997 FOR SELECTED SUBGROUPS\n1994\n1997\nChange\nParticipation\nPurticipation\nParticipation\nEligibles\nParticipants\nRate\nEligibles\nParticipants\nRate\nEligibility\nParticipation\nRate\nNO. NO.024\nIndividuals in\n(000s)\n(000s)\n(percen!)\n(000s)\n(000x)\n(percent)\n(percent)\n(percent)\n(points)\nHouseholds with permanent resident aliens\n6.373\n3.072\n48.2\n1,167\n968\n82.9\n-81.7\n-68.5\n34.7\n) louseholds with ABAWDs¹\n1,680\n1,220\n72.6\n649\n582\n89.7\n-61.4\n-52.3\n17.1\nHouscholds with single parents²\n12,995\n12,455\n95.8\n12,118\n10,446\n86.2\n-6.7\n-16.1\n-9.6\nwith earnings\n4,166\n2,624\n63.0\n4,893\n3,313\n67.7\n17.5\n26.3\n47\nwith no earnings\n8,829\n9,831\n111.3\n7,225\n7,133\n98.7\n-18.2\n-27.4\n-12.6\nOther households with children²\n7.550\n7,271\n96.3\n6,887\n5,154\n74.8\n-X.R\n-29.1\n-21.5\nwith carnings\n4.667\n3.601\n77.2\n4,460\n2.856\n64.0\n-4.4\n-20.7\n-13.1\nwith no earnings\n2.883\n3,669\n127.3\n2,427\n2,297\n94.6\n-15.8\n-37.4\n-12.6\nHouseholds with elderly and no children\n5.274\n1.615\n30.6\n5,261\n1,468\n27.9\n-0.2\n-9.1\n-27\nOther households\n2,681\n1.8M3\n70.2\n3,222\n1.846\n57.3\n20.2\n-2.0\n-13.0\nTotal Individuals\n36,553\n27.515\n75.3\n29,305\n20,462\n69.8\n-19.8\n-25 n\n-5 4\nSources: 1994 and 1997 IQCS Datalites. 1995 and 1998 CPS Datailes\nNo permanent resident aliens\n'N₁ ABAWDs or permanent resident aliens\nUSDH/F US/UHE\nМНИС:Р\nCD. 1999\nJ. Eric Gould\n04/05/99 06:01:54 PM\nRecord Type:\nRecord\nTo:\nCynthia A. Rice/OPD/EOP\nCC:\nSubject: FS Caseloads\nI put together a group from OMB, CEA and USDA to discuss FS caseloads and we reconvened\ntoday for our second meeting. I wanted to focus on 1) comparing take up rates over recent years;\nand 2) economic data that could help explain the caseload anomaly. I have people more focused\nnow but we are creating questions faster than we are coming up with answers.\n1) Take up rates. March CPS data is in (covers FY97 caseloads). USDA is really concerned that\nthe numbers are really screwed up since it shows that there are 6 million less eligibles than in\nFY96. The good news is that if the numbers are correct than the take-up rate is actually going up\na little from 68 % in FY96 to 70% in FY97. But at this point that type of drop in eligibility is very\ndifficult to explain since poverty levels didn't change to near that extent. The #1 assignment for\nthe week is to try to get a handle on the accuracy of the new CPS data. USDA said that they\nwould provide the resources to do so. The data indicated that of the 6 million decrease in eligibles,\nthere were 3 million fewer eligible legal immigrants. Again, this was much higher than expected\nand might be attributable to a) bad data; or b) underestimation of the # of eligible LI to begin with,\nwhich is possible because the old methodology (pre 1994) was fairly poor.\nBecause the estimate on eligibles is still so shaky we didn't spend a lot of time discussing what\ncould have caused a sharp decrease in those numbers but not reflect increased income through\npoverty stats. A preliminary response was that these low-income folks could be doing a better job\ncollecting assets, like a car.\n2) Economic data. The effect of unemployment rates on caseloads needs to be completely\nreexamined. The basis for these models is 25 years old. Currently, we have no way to measure\nthe effect of prolonged economic expansion and the added effect of low-unemployment pushing up\nwage scales. We are at a point were the bottom of the income bracket is doing better than they\nhave ever done before but we are not sure how to measure what that would do to FS caseloads.\nBob and Becky are looking at this question and they are talking to outside economists about it but\nthey claim that this is a complicated question that makes a quick answer difficult.\nSo for next week we will examine: 1) the believability of the eligible (specifically legal immigrants)\ndata from March, 1998 CPS; and 2) what we can say or discern from changes in the wage\nstructure for FS eligibles and the effects of very low unemployment.\nFrom the info. we have right now, it's possible that the same % of folks receiving FSs is staying\nsteady and even going up a little but that the number of those eligible is taking a big whack.\nMINTHIA\nFAX COVER SHEET\nOFFICE OF ANALYSIS AND EVALUATION\nFOOD AND CONSUMER SERVICE, USDA\n(Office: 703-305-2019)\n(Fax: 703-305-2576)\nDATE: 3-29-99\nTO:\nEMC Gould\nOFFICE:\nPHONE #:\nFAX PHONE #:\nFROM: Snova CARLSON\nPHONE:\n# PAGES (Including cover page):\nCOMMENTS:\nFYI\nTHIS was PULLISO TO60THX R LAST WOOK FUK\nKEASOUS UNKORAND TO THE OFFICE TO\nORPLAIN No FSP BLOP, BUT IT HAS sumis\nVALUE AS BACKGROUND\na\nP.1/2\nNO.134\nUSDA/FCS/OAE\nW802:21 6661'62'860\nMarch 30, 1999\nNOTE FOR ERIC GOULD\nFROM:\nSteven Carlson\nSUBJECT:\nFood Stamp Participation Decline\nFYI, I tried to implement the framework that I think we've all agreed to pursue to quantify the factors\ncausing the decline in food stamp participants (attached). The basic structure is as follows:\n(1) Calculate the drop in participation for a particular period (March 1994 to September 1997 in the\nattached table; March is selected as the peak, September as the last month for which we have\ndetailed characteristic information).\n(2) Estimate the number of persons made ineligible by welfare reform restrictions (chiefly on\nimmigrants and unemployed childless adults). These figures are drawn directly from two tables in\nthe recently released Characteristics of Food Stamp Households: Fiscal Year 1997; they differ\nsomewhat from the figures sent last week since they are the observed differences in the number of\nthese participants between the two selected months (rather than fiscal year or quarterly\ndifferences).\n(3) Estimate the effect of the economy. This step is, I believe, poorly estimated here, but hopefully\nmight be improved with the work you and Bob are undertaking.\nThe estimate presented is derived from an internal Agency model of food stamp participation. It is\nsparsely specified, consisting of measures of the number of unemployed people (not seasonally\nadjusted) in the current and a lagged quarter plus a set of quarterly dummy variables. The model\nis estimated from a quarterly time series dating from 1977. From the model forecast, we derived\nan estimate of the effect of changes in unemployment on food stamp participation (83 participants\nfor every 100-person change in the number unemployed).\nThere are few reasons to be confident in this particular adjustment. The forecast model did not\nperform particularly well during the run-up in food stamp participation during the late 1980s and\nearly 1990s, nor has it performed well during the more recent decline. The model is sparsely\nspecified, providing a meager proxy for economic changes and failing to account for significant\nprogram changes. I've included it here only for the purpose of illustration.\n(4) Net out effect of eligibility restrictions and economy, leaving an unexplained residual.\nCall with questions on (703) 305-2134.\nAttachment\nP.2/3\nNO. 161\nUSDA/FCS/OAE\n3:44PM\nMAR. 30. 1999\nPreliminary Decomposition of Food Stamp Participation Change\nPermanent Resident\nUnemployed\nAll Participants\nAliens\nChildless Adults\nMarch 1994\n27,965\n1,501\n1,320\nSeptember 1997\n20,991\n-6,974\n547\n-954\n662\n-658\nDecember 1998\n18,618\n-9,347\nn/a\nn/a\nn/a\nn/a\n# Unemployed\n1994 Q1\n9,209\n1997 Q4\n6,659\n-2,550 X .83 =\n-2,117\nMarch - Sep Drop\n-6,974\nImmigrants\n-954\n0.14\nABAWDS\n-658\n0.09\nSubtotal\nUnemployment\n-2,117\n0.30\nUnknown\n-3,246\n0.47\n1.00\nNotes:\n(1) Estimates of permanent resident aliens and unemployed childless adults from\nCharacteristics of Food Stamp Households, Fiscal Year 1997, Tables B-1 and\nB-3.\n1pt\nUR:\n(pt\nfs/\n(2) Estimate of effect of unemployment on food stamp participation derived from\nFNS forecast model (based on quarterly number of unemployed (seasonally\nunadjusted and set of quarterly dummy variables). Model implies that a net\nchange of 83 participants for every 100-person change in number unemployed.\n! Participation rate\nof eligibles\nChepser 3\nMarch 30, 1999\n2. # g total eligibles,\nand tabe up-\n3. Laged effects of wage VS. UR,\nP.3/3\n191.19\nUSDA/FCS/OAE\n3:44PM\n1999 \"0E MARK\nFood and Nutrition Service\nUS Department of Agriculture\nNational Data Bank V 5.0\nMarch-94\nDecember-98\n3/23/99\nState /\nFS PROGRAM\nFS PROGRAM\nDifference\nTerritory\nPART ACTUAL\nPART ACTUAL\nDec 98 Mar 94\nALABAMA\n561,803\n415,270\n-146,533\nALASKA\n51,692\n15,680\n-36,012\nARIZONA\n512,288\n259,935\n-252,353\nARKANSAS\n290,212\n256,378\n-33,834\nCALIFORNIA\n3,191,890\n2,062,101\n-1,129,789\nCOLORADO\n275,765\n180,083\n-95,682\nCONNECTICUT\n225,044\n183,118\n-41,926\nDELAWARE\n61,717\n42,482\n-19,235\nDISTRICT OF COL\n$2,427\n85,344\n-7,083\nFLORIDA\n1,469,271\n967,349\n-501,922\nGEORGIA\n826,295\n633,555\n-192,740\nGUAM\n15,058\n18,895\n3,837\nHAWAII\n115,461\n126,249\n10,788\nIDAHO\n88,094\n58,411\n-29,683\nILLINOIS\n1,207,279\n840,288\n-366,993\nINDIANA\n543,248\n301,485\n-241,763\nIOWA\n203,883\n128,493\n-75,390\nKANSAS\n197,340\n113,104\n-84,236\nKENTUCKY\n536,337\n396,813\n-139,524\nLOUISIANA\n763,280\n531,415\n-231,865\nMAINE\n141,742\n109,656\n-32,086\nMARYLAND\n399,877\n295,722\n-104,155\nMASSACHUSETTS\n451,344\n274,751\n-176,593\nMICHIGAN\n1,047,450\n723,807\n-323,643\nMINNESOTA\n323,939\n207,606\n-116,331\nMISSISSIPPI\n519,006\n295,042\n-223,964\nMISSOURI\n606,179\n408,118\n-198,061\nMONTANA\n74,280\n61,069\n-13,211\nNEBRASKA\n113,484\n98,019\n-15,465\nNEVADA\n$8,977\n64,777\n-34,200\nNEW HAMPSHIRE\n63,760\n37,931\n-25,829\nNEW JERSEY\n554,501\n394,452\n-160,049\nNEW MEXICO\n251,655\n182,559\n-69,096\nNEW YORK\n2,175,154\n1,577,854\n-597,300\nNORTH CAROLINA\n646,230\n498,053\n-148,177\nNORTH DAKOTA\n47,459\n33,654\n-13,805\nOHIO\n1,269,050\n661,747\n-607,303\nOKLAHOMA\n382,926\n280,798\n-102,128\nOREGON\n298,653\n224,513\n-74,140\nPENNSYLVANIA\n1,243,533\n842,332\n-401,201\nRHODE ISLAND\n96,336\n76,913\n-19,423\nSOUTH CAROLINA\n390,940\n320,011\n-70,929\nSOUTH DAKOTA\n55,200\n43,529\n-11,671\nTENNESSEE\n747,371\n520,315\n-227,056\nTEXAS\n2,764,395\n1,460,671\n-1,303,724\nUTAH\n131,124\n90,222\n-40,902\nVERMONT\n65,475\n44,953\n-20,522\nVIRGINIA\n571,254\n370,069\n-201,185\nVIRGIN ISLANDS\n19,632\n17,062\n-2,570\nWASHINGTON\n483,004\n326,240\n-156,764\nWEST VIRGINIA\n333,260\n250,995\n-82,265\nWISCONSIN\n334,900\n183,752\n-151,148\nWYOMING\n34,698\n24,037\n-10,661\nUnited States\n27,965,172\n18,617,677\n-9,347,495\n2/2'd\nNO.134\nUSDA/FCS/OAE\nW802:21 6661.62.86W\nCFA\nSheet1\nNumber of Participants\nUnemployed\nPermanent\nChildless Adult\nParticipants\nResident Aliens\n(ABAWDS)\nMar-94\n27,965\n1,501\n1,320\nSep-97\n20,991\n547\n662\nDec-98\n18,618\nn/a\nn/a\nChange from 3/94 to 9/97\nParticipants: 3/97\n27965\nC17\n(1) Drop due to Aliens\n-954\nC18\n(2) Drop due to ABAWDS\n-658\nC19\n(3) Drop due to economy\n-2530\n(1)-(3)\n-4142\nActual decline\n-6974\n% of change explained by:\nAliens\n13.7%\nABAWDS\n9.4%\nEconomic conditions\n36.3%\nAll other factors\n40.6%\nOther potential factors:\n1. Reform of AFDC in 1996\n(C17+C18+C19) (65-4.9)*(0.06)\nPage 1\nJ. Eric Gould\n03/31/99 02:30:47 PM\nRecord Type:\nRecord\nTo:\nCynthia A. Rice/OPD/EOP\nCC:\nSubject: Food Stamp Caseloads\nOur research group is making a little headway. Unfortunately, 40-45 percent of the caseload\nreduction can not be explained through any modeling techniques available. A problem is that we\ncan't just attribute it to \"whatever welfare reform, be it good or bad, is doing\" to the caseload.\nThere are unknown possible economic effects that could have significant influence but we are not\nsure. These effects caused by \"great prolonged economic conditions\" are hotly debated in\neconomic circles right now and also at CEA. Becky is one of the leading proponents of the theory\nthat prolonged economic success creates affects among low-income individuals that we are not yet\nable to measure. For example, the standard economic model to predict FS caseloads is the\nunemployment rate (for every 1% increase in unempl. the FS caseload is supposed to increase 6%)\nbut in a tight labor market the unemplo. rate has gone so low that wages get pushed up and have a\ndifferent but powerful effect on FS caseloads. We are also looking at the changes in the income\nlevels of all FS eligibles (below 130% poverty) - which should be going up meaning more people\nmight be eligible for less benefit. We are also looking at the decline in the total amount of eligibles\nover time (the group of below 130% is getting smaller).\nWe are trying to make USDA get updated numbers on the participation rate of eligibles, armed with\nthat and the total number of eligibles we can work out take-up rates and compare them over time.\nJ. Eric Gould\n03/26/99 06:27:28 PM\nRecord Type:\nRecord\nTo:\nCynthia A. Rice/OPD/EOP\nCC:\nSubject: Food Stamp Chart\nI had a long conversation with USDA about the chart (Table 2) you and I discussed this morning.\nThe chart reviews data on single parent households by presence of earnings and welfare receipt\nbetween 1994-1997. From my conversation with USDA I gathered:\n1. The number of single parent households receiving TANF fell, which is consistent with overall\ndeclines in welfare caseloads.\n2. The number of single parent households not receiving TANF increased modestly, which is\nconsistent with the notion that as folks leave welfare, some continue to receive food stamps as\nNPA cases.\n3. The number of single parent households not receiving TANF who have earnings increased\nslightly, which is consistent with recent efforts to move single parent families off of welfare and\ninto work.\nConclusions from the data:\n1. The data show that some households appear to leave welfare and food stamps completely. We\nexpect that some had higher earnings. However, the data does not provide any information about\nthese households.\n2. Others find work and leave welfare, but continue to combine work with food stamps as a\nnutritional supplement, which is how the Program is supposed to work.\nOne thing to keep in mind is that these numbers represent people who continue participation and\npeople who come on to the program. For example, when the data reflects a single mom with\nearnings and no TANF, we don't know if she got TANF and food stamps before and is now a food\nstamp only recipient, or is she never got TANF and came to the State agency as an NPA household.\nThis means that when the chart shows an increase in single parents with earnings and no TANF,\nwe don't know how many of them worked their way off of TANF. Similarly, when there is no\nTANF and no earnings, these could be people who were kicked off TANF or who never received it\nin the first place.\nWhat is Causing Caseload Reductions in the Food Stamp Program ?\n3/10\nIn November 1998, the Food Stamp Program served 18.6 million people. The number of people\nreceiving food stamps fell by over 9 million people, a drop of one third, since March 1994. Part\nof this drop is due to the strength of the economy and the success of welfare reform. Part of the\ndrop is due to new restrictions on the participation of certain legal immigrants and able-bodied\nunemployed adults without dependents.\nHowever, other factors may also be at work. Between 1995 and 1997, Food Stamp Program\nparticipation fell five times as fast as poverty, suggesting that many poor families have left the\nprogram despite their continuing eligibility. While the program is available to households with\ngross incomes of up to 130 percent of poverty, 91 percent of program participants have income\nbelow the poverty level.\nThere are three important factors that must be considered in addressing the reasons for the sharp\ndecline in Food Stamp participation since 1994: 1) specific Food Stamp eligibility changes in\nPRWORA; 2) broader PRWORA changes requiring work and discouraging dependence on cash\nassistance; and 3) a strong economy. The relation of these three factors is a topic of continuing\nresearch but preliminary information shows that economic factors can not fully explain recent\ncaseload declines and further research is needed to look more closely at behavioral changes in\ntake-up as well as state-specific changes in eligibility.\nHow ave these different\n-97?\n1. Who is Leaving the Food Stamp Program?\nSubgroup\nParticipants\nParticipants\nParticipation\nPercent\nSubgroup as\nSubgroup as\nPercent of Decli\nin 1994\nin 1997\nChange 1994-\nChange in\nPercent of\nPercent of\nfrom Subgroup\n(thousands)\n(thousands)\n1997\nParticipation\n1994\n1997 caseload\n1994-1997\n(thousands)\ncaseload\nLegal\n1,537\n706\n-831\n-54%\n6%\n3.4%\n14%\nPermanent\nResident Aliens\nChildless\n1,148\n648\n-500\n-43.5%\n4%\n3.1%\n8.4%\nbet\nUnemployed\nAdults\nAFDC/TANF\n13,052\n9,442\n-3,610\n-27.6%\n48%\n43%\n60.8%\nRecipients\nAll Other\n11,697\n10,707\n-990\n-8.5%\n43%\n49%\n16.7%\nParticipants\nTotal\n27,434\n21,503\n-5,931\n-21.6%\n100%\n100%\n100%\nParticipants\nJuly-September Food Stamp Program Quality Control Data for 1994 and 1997\nNarrow columns 50thenal FH\nThe above chart shows that while there were large reductions within the legal immigrant and\nenpoye\nchildless unemployed adult populations, their combined effect only contributes to less than a\nquarter of the entire food stamp caseload reductions. It must also be noted that the above\npercentage changes also reflect economic influences. Therefore, even without policy changes the\nsubgroups would still have reflected decreased caseloads.\nThe bulk of the caseload changes comes from AFDC/TANF recipients, who make up\napproximately 43 percent of the Food Stamp caseload. Between 1994-1997, this group\nexperienced a 27 percent decrease in participation, which resulted in 60 percent of the entire\ncaseload reduction between 1994-1997.\nThe group identified as \"All Other Participants\" consists of working poor, aged poor and disabled\npoor. This group is the second largest segment of the Food Stamp population, making up 43\npercent of the entire caseload in 1997.\n(alth\nimmigo most\nFeable\n2. How are specific changes in PRWORA affecting the Food Stamp Program?\nPium (Lb/\nAt least 20 percent of the decline between 1994-1997 came from two subgroups directly affected\nby PRWORA. Participation of legal immigrants declined 54 percent between 1994-1997 and\nreflected 14 percent of the entire caseload's decline. Changes in BBA and last year's Agricultural\nResearch Act will may have án affect on these numbers as certain groups of legal immigrants become\neligible for Food Stamps. The group of childless unemployed adults without dependents made up\n8.5 percent of the decline in the entire Food Stamp caseload, as the subgroup registered a 44\npercent decline between 1994-1997.\nRemind people of policy eg, 18to sos only\neligible for smonths of wratthnu unless working\nAdding the percentage caseload declines from these two subgroups suggests that 22.5 percent of 220hpm\nthe entire caseload reductions between 1994-1997 were attributable to changes in eligibility for\nthese populations. However, economic factors also have to be taken into consideration. Experts'\nestimates of the economy's impact on the Food Stamp caseload during this period varies between\n25-45 percent. Taking into consideration economic forces, the direct impact of PRWORA on\nthese two groups and their relation to their total impact on Food Stamp caseload declines has been\nestimated to be approximately 10 percent.\n3. What affect do the broader changes in PRWORA have on Food Stamp caseload\nchanges?\nRegardless of the impact that specific policy changes in PRWORA have had on Food Stamp\ncaseloads, the bulk of the decline between 1994 and 1997 came from AFDC/TANF recipients,\nwho as a group attribute for 60 percent of the decline The number of AFDC/TANF recipients has\nfallen by over 1-million since 199 and the vast majority of the AFDC/TANF recipients were\nalso receiving Food Stamps Historifally The Food Stamp Program has been closely coordinated in local\noffices with cash assistance programs.nd this can dramatically affect the household' assessment\nof the costs and benefits of participation.\nchrifort, individuals receiving cash ass stance nave always\nbeen fai more likely to recewe Food Stamps tram other\neligible endi ideals\nPERCENTAGE OF\nare are these they all non-earners!\nPARTICIPATION OF FOOD STAMP ELIGIBLE\nINDIVIDUALSUN THE FOOD STAMP PROGRAM\nHousehold with earnings\n45.8%\nAdd you Identy in get fn stat non-\nHouseholds with children\n85.8%\nSingle parents with children\n96.8%\nwhat of earners is rate not on AFDC?\nHouseholds with earnings and children\n55.2%\nSingle parents with earnings/children\n59%\n1994, SIPP\npublic arrythnce\nThese numbers indicate that families with a single adult, who were receiving are almost\ncertain to participate, while households with earnings have a participation rate under 50 percent.\nAt some level, the presence of children has a positive influence on participation even among the\nsubset that works and being a single parent raises the likelihood of participation further.\nAs the TANF roles decrease and individuals continue.to obtain more earnings, these numbers\nindicate that they are less likely to retain their Food Stamps even though they are eligible.\nAlthough participation rates among single parents and their children have been historically high,\nthe recent declines in TANF caseloads suggest they might fall in the future. For example, in\nIndiana, the state's welfare reforms have reduced Food Stamp participation without a\ncorresponding reduction in the number of potentially-eligible families. In short, some families\nthat leave or avoid TANF and remain in poverty fail to apply for Food Stamp benefits.\nKnowing that the number of households with earning is increasing and that they are less likely to\nretain their Food Stamps, even though they may still be eligible, an important issue is the\nrelationship between the percentage of working eligibles who obtain Food Stamps as compared to\nthe present percentage. The trend has been that an increasing amount of Food Stamp participating\nhouseholds have some earnings. From 1994-1997 almost three percent more participating\nhouseholds reported earnings.\nBut M people are 60 likely toget FS orce they have\nearnings, wfat los this mem?\nPERCENTAGE OF FOOD STAMP PARTICIPATING HOUSEHOLDS WITH EARNINGS\n1992\n20.2%\nwould be more\nuseful to\n1993\n20.6%\ncompane\nI\nAolto This to include all leavers\n1994\n21.4%\nFS\nstudies (surveys\nguilat\n1995\n21.4%\neligibles\n% of resports leasing To all\n1996\n22.5%\ngo FST Medicaed) see\n1997\n24.2%\ndeveropm s> 19\nemail from Andria\n1997 USDA Food Stamp Quality Control Samples\nDo some calcula tions based\non these data and the leavers studies 2. whechshow bet 50-60% of those leaving\nthe rolls go to work i.e., if bet 28-45 90 left due to economy, courlte\nrest of the decline from 194- 197 be attributable to foid that\n55% of Those who left are now only talf as likely to get FS\nforwhechtley we elear!?\n4. What factors contribute to nonparticipation?\nUSDA has studied the reasons for nonparticipation and identified three common factors that are\nlikely to apply:\nInsufficient or incorrect information about the program. Working-poor families with\nchildren often do not know about their eligibility for Food Stamps given that many such\nfamilies are not eligible for cash assistance. Lack of accurate program information is also\nprevalent among illiterate and non-English speaking individuals.\ntoget\nto\noffice\nProblems of Program access and administrative Influialty difficulties with the application, during\nrabdtothis\nworking\nopen\nhours/\nprocess. Historically, participation rates are often lower among rural and elderly\nincentives\nerror\npopulations because of physical access and transportation barriers. However, recent\nrates), for\nreports, similar to the illegal NYC access situation, have created questions whether wide- states than\nspread diversion or discouragement of Food Stamp participation is taking place on a\nthese fam\nbroader level.\nGive an ex= some one w/ su/hr\nSmall size of the benefit or lack of desire for benefits. As an individual's or\n106 t 2 cheldren is eleaptle forxx FS in\nhousehold's income rises the potential Food Stamp benefit declines. At some point,\nprogram eligibles decide the small amount of the benefit is worth less than the cost of\napplying. Other people forgo benefits regardless of the amount because of personal\nobjections to income transfer programs and / or the stigma associated with receiving or\nusing Food Stamps. The increasing use of EBT could decrease this stigma in the future as\nthe transaction will be similar to credit card use.\n5. How does the decline in AFDC/TANF participation compare to declines in food stamp\nparticipation?\nMake this\nCHANGE IN WELFARE / FOOD STAMP CASELOADS SINCE ENACTMENT OF PRWORA\nChange in AFDC/TANF\nPercent Change in\nChange in Food Stamp\nPercent Change\nAFDC/TANF\n#2\nParticipation\nParticipation\nin Food Stamps\nAug. 96 - Sep. 98\nAug. 96 - Sep. 98\nAug. 96 - Sep. 98\nAug. 96 - Sep. 98\n4,286,000\n-35%\n6,229,107\n-25%\nComparing data between 1980 through 1998, Food Stamp caseloads are usually about twice as\nhigh as AFDC/TANF caseloads. This reflects the broader eligibility rules for Food Stamps.\nAFDC caseloads which were largely flat from the mid-1970s through 1990 rose 27 percent\nbetween 1990 and 1994. But, between 1994 and mid-1998 they fell by 40 percent. In\nSeptember, 1998, they were at their lowest since 1969.\nFood stamp caseloads follow a similar trend. Participation peaked at 22 million during the 1983\nrecession and then fell steadily with a stronger economy until 1988. From 1988 to 1994,\nparticipation increased by 9.5 million, a 52 percent increase. From 1994 to February 1998,\nLet's discus what best compromounsibe\n-1994-1997 ?\n1996-1997?\n96-98?\nFood Stamp participation fell by 7.8 million, a 28 percent decrease.\n6. Is the economy operating through the strong labor market responsible for the decline?\nThe relationship between the economy and Food Stamp participation has been fairly strong over\nthe last twenty years. Even when forecasters have not been able to predict AFDC caseloads\naccurately on the basis on economic factors, there are still corresponding patterns of rise and fall\nin unemployment and the Food Stamp caseload. But even though economic factors are\nimportant, they are not the entire story. There have been periods where Food Stamp caseload\ntrends varied significantly from unemployment rates.\nThe unemployment rate has fallen from 6.6 percent in 1994 to 4.4 percent in 1999. Previously\ndeveloped models of the relationship between the economy and Food Stamp participation\nindicate that economic trends since 1994 could account for anywhere from 28 to 45 percent of\nthe change in Food Stamp participation. With the caseload declining by almost eight million\nindividuals since 1994, this 17 percent difference (difference between 45 percent and 28\npercent) is significant in understanding what happened to approximately 1.4 million individuals.\nRecent research done by Rebecca Blank suggests that a one-point rise in the unemployment rate\nwill produce a 6 percent increase in the food stamp caseload over a three year period. This\nmight be an understatement if a lack of TANF-funded assistance produces greater need among\nlow-income families in the future. On the other hand, it may be an overestimation if smaller\nnumbers of people in TANF-funded programs mean that fewer families are aware of their food\nstamp eligibility and take-up rates fall. Still, economic data that takes into consideration\nunemployment and median wages, can only account for 40 percent of the caseload changes that\nhave taken place between 1994 and 1998.\nF52\nWhat are we doing about this?\nWe will continue to monitor caseload trends and keep you informed. Moreover, there are\nimportant questions we need to further examine with the assistance of USDA and CEA.\nIf the sharp decline in Food Stamp participation can not be fully explained through\neconomic forces, what relationship does the decreasing TANF caseload have on it and\ncan we expect both caseloads to continue to decline?\nHow much influence do the common factors of non-participation have on program\neligibles and how can we address these specific factors?\nWhat can we do through Administrative action to address this situation?\non\nthe work departmy\nWhat is Causing Caseload Reductions in the Food Stamp Program (ESP)?\nFood stamp digfility\nThere are three important factors Sta that must be considered in addressing the plausible reasons for\nthe sharp decline in FST participation since 1994: 1) specific FOP changes in PRWORA; 2)\nbroader PRWORA changes to the cash system; and 3) a strong economy. The relation of these\nthree factors is a topic of continuing research and debate. This memo addresses current easeload\ntrends and research\nbut proliminary mathon shows\n1. Who is Leaving the Food Stamp Program?\n1994 1997\nSubgroup\nParticipants in\nParticipants\nParticipation\nPercent Change\nSubgroup\nPercent of Decline\n1994\nin 1997\nChange 1994-\nin Participation\nas Percent\nfrom Subgroup\n(thousands)\n(thousands)\n(thousands) 1997\nassocial\ncaselord of 1997\nLegal\n1,537\n706\n-831-\n-54%\n3.4%\n14%\nPermanent\n6.0%\nResident Aliens\nChildless\n1,148\n648\n-500\n-43.5%\n3.1%\n8.4%\nUnemployed\n4.0%\nAdults aye18.50\nAFDC/TANF\n13,052\n9,442\n-3,610\n-27.6%\n48%\n43%\n60.8%\nRecipients\nAll Other\n11,697\n10,707\n-990\n-8.5%\n43%\n49%\n16.7%\nParticipants\nhow\nTotal\n27,434\n21,503\n-5,931\n21.6%\n100%\n100%\n100%\nmanyol\nParticipants\nThese\nJuly-September Food Stamp Program Quality Control Data for 1994 and 1997\ntowmany work\n2. How does the decline in AFDC/TANF participation compare to declines in food stamp\nparticipation?\nupdate\n&\nDecline in AFDC/TANF\nPercent Changen AFOC/TAIN\nDecline in Food Stamp\nPercent Change intiod Stamps\nSept\n\"\nchu\nCha nge\nParticipation\n1998\nMarch 1994 to June 1998\nParticipation\nMarch 1994 to June 1998\nMarch 1994 - June 1994\nMarch 1994 - June 1998\n-5,957,854\n-41.8%\n-8,953,651\n-31.4%\nComparing data between 1980 through 1998, food stamp caseloads are usually about twice as\nhigh as AFDC/TANF caseloads. This reflects the broader eligibility rules in the FSP AFDC\ncaseloads which were largely flat from the mid-1970s through 1990 rose 27 percent between\n1990 and 1994. But, between 1994 and mid-1998 they fell by 40 percent. In June, 1998, they\nwere at their lowest since 1969\nSpt\nFood stamp caseloads follow a similar trend. They decline slightly faster than AFDC caseloads\nwhat are As?\nher about a araph !\nyour but\nnoclat above\n?\nPRWORA\ngote\nneed\n#\nin the mid-1990s, in part because ofPRWORA Still, the rapidity of these hanges is\nunprecedented and hard to explain.\nwith Bortne\nFood stamp elizablety\nFood stump\ngulat\n3. Did the 1996 FSP changes in PRWORA contribute to the decline in FSP participation?\n0\\0\nstamp\nother\nThese changes imposed a strict time limit on able-bodied adults without dependents. They made\nX\nmost legal aliens ineligible for the ESP (even though much of these cuts have been restored),\n\"D'\nslightly reduced the value of the maximum benefit, and limited certain deductions. With little\nhard data available, there is no way to know for sure what affect these changes actually have had\non FSP participation. Some experts agree that these changes would have affected easeloads no\nmore than 10 percent (Beebout, Mathematica 1998).\nThink This\nDostino mean 109. of dertne\nbowe Johnny the\nThroughanges\nwould\nstudy is good\n4. How do the non-FSP changes in PRWORA relate to FSP participation?\nthrough\n# appeople\nreversity\nThe number of AFDC/TANF recipients has fallen by over 4 million since 1994, and the vast\nbanfts.\nmajority of the AFDC/TANF recipients were also receiving food(stamps. However, some\nsubstantial portion of that decline was also caused by the strong economy, and there is little\nagreement among researchers about how much of the shrinking caseload resulted from welfare\nhow\nreform and how much resulted from the economy. Whatever the cause, the reduction in the\nAFDC/TANF rolls is associated with about half the decline in FSP participation.\n?? Jowe Vor\n5. Is the economy operating through the strong labor market truly responsible for the\nThis\ndecline?\nFrom 1994 to February 1998, FSP participation fell by 7.8 million, a 28 percent decrease. In\ncomparison the number of unemployed persons fell by 3.6 million during this period, and the\nunemployment rate fell from 7.8 to 4.6 percent of the labor force. Previously developed models\nof the relationship between the economy and FSP participation indicate that economic trends\nsince 1994 could account for anywhere from 28 to 45 of the decline in participation percent\n(Blank, CEA 1998; Beebout 1998; Hamilton, Abt Associates 1998).\nmeaning\nAFDC\nThe relationship between the economy and FSP caseload has been fairly strong over the last 20\nyears. Even when forecasters have not been able to predict caseloads accurately on the basis on\neconomic factors, there are still corresponding patterns of rise and fall in unemployment and the\nFSP caseload. However, although economic factors are important, they are not the entire story.\nThere have been periods where FSP caseload trends took off from unemployment rates. ??\n-\nSimply explained, there are other broader factors that help explain how many people participate\nin the FSP: the size and composition of the population and the eligibility rules in place interact\nwith economic factors.\n6 Still, what do economic models tell us about the future of food stamp participation\nrates?\nRecent research done by Rebecca Blank (CEA) suggests that a one-point rise in the\nunemployment rate will produce a 6 percent increase in the food stamp caseload over a three year\nperiod. This might be an understatement if a lack of TANF-funded assistance produces greater\nneed among low-income families in the future. On the other hand, it may be an overestimation if\nsmaller numbers of people in TANF-funded programs mean that fewer families are aware of\ntheir food stamp eligibility and take-up rates fall.\nInstead\n7. What \"links\" exist between food stamps and TANF participation?\nParticipation in AFDC/TANF plays a special role in food stamp participation patterns. The FSP\n90 wk.\nhas been closely coordinated in local offices with cash assistance programs. This can\ndramatically affect the household's assessment of the costs and benefits of participation.\ndezplar\nParticipating in both programs has only marginally higher costs than participating in one, and as\nobtains\na result, FSP participation is strongly associated with cash assistance participation. The vast\n(and\nmajority of single-parent households who receive cash assistance also receive food stamps, but\ncurrently\nfood stamp participation is much less common among single-parent households who do not\nreceive cash assistance even when households are grouped by income levels. The result is that,\nwhat\nfor a substantial part of the food stamp caseload, food stamp participation closely parallels\nto\n1846\nAFDC/TANF participation. Therefore, recent reports that some states are diverting, delaying\nand terminating individuals from TANF, and sometimes food stamps, in violation of federal law\nputhis\nis potentially a critical issue in terms of better understanding participation rates.\naspant\nSomewhat surprisingly, even though more than half of all food stamp cases do not receive cash\nof\nassistance, the national food stamp caseload has closely tracked the national AFDC/TANF\ncaseload since the early 1980s. This may mean that, beyond the administrative linkage, both\nQ+A\nprograms are influenced by some underlying pattern in households' assessments of the costs and\nbenefits of participation. This might reflect economic factors, such as perceptions about job\nopportunities, or non-economic factors.\n8. What conclusions can be drawn from all of this?\nAlthough it is clear that many factors are clearly correlated with food stamp caseload changes,\nthey do not explain the recent trends well. The uncertainty over the causes of recent powerful\ndeclines in the caseload are leaving most economists scratching their heads and causing other\nresearchers to look more closely at behavioral changes in take-up as well as state specific\neligibility changes that might be influencing these changes. Some economists (Blank and\nBeebout) and experts (Greenstein) believe that recipients are strongly influenced by a host of\nless-measurable factors (including their own sense of the \"accessibility\" of utilizing public\nassistance) when deciding whether or not to participate in the FSP.\n3/4/99\nSUMMARY OF RECENT RESEARCH\nON WELFARE LEAVERS\nStudy:\nPercent of Leavers\nPercent of Leavers\nReceiving Food Stamps\nReceiving Medicaid\nDelaware - 1/98\n65%\n75%\nIndiana - 1997\n38%\n53%\n11/98\n52% (after Year 1)\n66% offered health insurance by\n40% (after Year 2)\nemployer, but 50% declined for\nvarious reasons including cost\n(5% declined because of\nMedicaid coverage)\nIowa - 1993-5\n66%\n66%\n1999\n59%\nSouth Carolina - 1997\n66%\n75%\n-\n10/98\n52% of adults had health\ninsurance (40% of the total\nhad medicaid and 12% had\nprivate insurance);\n91% of the kids had health\ninsurance (80% medicaid and\n11% private)\nWashington - 1/99\n45%\n44% (Adults)\n64% (Kids)\nWisconsin - 1/99\n49%\n87%\nA2\nSUNDAY. MARCH 7. 1999\nS\nNATIONAL NEWS\nTHE WASHINGTON\nDrop in Food Stamp Rolls Is Worrying Officials\nA Sharp Decline\nBy JUDITH HAVEMANN\nmittee, said the declines are \"quite\nThe welfare application process\nfood stamp usage has been pro-\ncalled for a review of the rules that\nThe number of people who\nWashington Post Staff Writer\nmysterious\" and need to be investi-\ntraditionally has been the main en-\npelled by a booming economy and\ngovern food stamp and Medicaid eli-\nreceive food stamps fell from a\ngated because they \"far exceed ex-\ntry point for food and medical assis-\nchanges in the law that have made\ngibility, while Rep. Robert W. Goo-\nhigh of 27.5 million in 1994 to\nAdvocates for the poor and some\npectations.\" Johnson said that the\ntance, as well as cash. Since the 1996\nabout 700,000 legal immigrants inel-\ndlatte (R-Va.), chairman of the Agri-\n19.8 million last year.\nmembers of Congress are alarmed\navailability of health care benefits is\nwelfare reform law was passed,\nigible. The rest of the drop, they say,\nculture subcommittee that oversees\nby a steep drop in the number of\n\"core\" to whether welfare reform\nsome states have been \"diverting\"\nis a result of states discouraging po-\nfood stamps, said he would favor ef-\nNumber of recipients, in millions\nAmericans receiving food stamp as-\nsucceeds and that it is unclear\npoor families from welfare\ntential applicants, intentionally or\nforts to turn over the food stamp\n30\nsistance and have begun pressing\nwhether states are doing enough to\ninadvertently, because workers are\nprogram to the states in the same\nfor changes to ensure that poor fami-\ninform poor families of their eligibil-\nThe decline in\nso focused on moving them into\nmanner as welfare if there is broad\n28\nlies have enough to eat.\nity for assistance. She plans to hold\njobs.\nsupport for such a move.\nNearly 8 million Americans have\nhearings to draw attention to the\nMoreover, welfare reform has\nDouglas Besharov, a senior fellow\nleft the food stamp rolls during the\nproblem.\nfood stamp usage\n26\nstigmatized all forms of aid, they\nat the American Enterprise Insti-\nfour years, a 28 percent reduc-\nState welfare officials, mean-\nsay, making many poor families re-\ntute, said the decline in Medicaid\n24\ntion that is far greater than the de-\nwhile, will converge on Washington\nand food stamp rolls is not fully un-\n1994:\ntoday to ask Congress to abolish\nsuggests a lot of\nluctant to apply for government help\ncline in poverty for roughly the same\nof any kind. Some Americans are\nderstood. But he said some of it rep-\n27.5 million\n22\nperiod, according to new federal fig-\nrules that they say are an obstacle to\nsimply unaware that they are poor\nures.\npoor families left\nresents a common-sense decision by\ngetting food stamps and are in-\nenough to qualify for food stamps if\nrecipients who don't want to miss a\n20\nAgriculture Secretary Dan Glick-\ncompatible with welfare reform. For\nthey earn less than $1,783 a month\nday of work every six months to visit\n1991:\nman said that suggests \"that a lot of\nexample, they say, it is absurd that a\nthe program even\nto support a family of four. Also,\nwelfare offices to qualify for a small\n22.6 million\n18\npoor families left the program even\nlow-income family with a car valued\nmany former welfare recipients are\nallotment of food stamps. Moreover,\nthough the need was still there.\"\nabove $4,650 is almost automatical-\nnot told they are entitled to at least\nthough the need\nbecause poor people can sign up for\nsix months of Medicaid benefits to\n16\n1998: 19.8 million\nThe combination of plummeting\nly ineligible for food stamps when\nMedicaid when they visit an emer-\nfood stamp rolls and a similar, but\nsome state welfare agencies are\nhelp them make the transition to\ngency room, they are less inclined to\nsmaller decline in the number of\nhelping welfare families obtain cars\nwas still there.\"\nwork, these advocates say.\n14\ngo through the sign-up process\npeople in Medicaid, the health insur-\nto get to work.\nKinder, gentler' is a thing of the\nwhen they are not in immediate\nance program for the poor, has fu-\nThe state officials, members of\n-Dan Glickman,\npast,\" said Rep. Fortney \"Pete\"\nneed of care.\n0\neled concerns among advocates for\nthe American Public Human Servic-\nsecretary of agriculture\nStark (D-Calif.). \"Some states\n\"There is no evidence of a signif-\n1991\n'92\n'93\n'94\n'95\n'96\n'97\n'98\nthe poor that the 1996 federal wel-\nes Association, also will urge the Ag-\nhave been illegally telling people\nicant increase in the level of hard-\nSOURCE: Agriculture Department\nfare reform law may be discouraging\nriculture Department to reduce\nby making them search for jobs or\nthey are no longer qualified.\"\nship the poor are facing, and evi-\nTHE WASHINGTON\nsome families from seeking needed\nsome of the program's red tape, in-\ngiving them a onetime payment to\nSecond Harvest, which operates\ndence, in fact, that the condition of\naid. Private food pantries and chari-\ncluding requirements that states\ntide them over. Food advocacy\nthe largest network of food banks in\nthe poor has improved,\" Besharov\nures are available.\ntable organizations have reported\ncontinually summon recipients to\ngroups say that welfare rules have al-\nthe country, said that demand is up,\nsaid.\nThe Medicaid rolls fell from 41.4\nan increased demand for food assis-\ncounty welfare offices to recertify\nso \"diverted\" families from the food\nparticularly among families, but said\nFood stamp usage fell from 27.5\nmillion to 40.3 million during the\ntance.\ntheir eligibility.\nstamps and medical insurance that\nit was unable to compare the 26 mil-\nmillion in 1994 to 19.8 million last\nsame period, even as federal and\nRep. Nancy L. Johnson (R-\nBut advocates for the poor charge\nthe families need more than ever.\nlion people who sought help during\nyear, while those living in poverty\nstate health officials beat the bushes\nConn.), chairman of the Ways and\nthat the real problem may be the\nThe groups cite studies showing\n1997 with earlier years.\nfell from 38.1 million to 36.6 million\nto extend health insurance, to the\nMeans human resources subcom-\nstates themselves.\nthat only about half of the decline in\nRep. Robert T. Matsui (D-Calif.)\nin 1997, the last year for which fig-\nmany qualified uninsured families.\nAPHSA\nAmerican Public Human Services Association\nMarva Livingston Hammons, President\nWilliam Waldman, Executive Director\nMarch 2, 1999\nCynthia Rice\nSpecial Assistant to the President, Domestic Policy Council\n1600 Pennsylvania Ave, NW\n212R OEOB\nWashington, DC 20500\nDear Ms. Rice:\nThe American Public Human Services Association (APHSA) invites you to come hear\nthe Commissioners of three state human services agencies offer their perspectives on\n\"Food Stamps, TANF and the Working Poor.\" The forum will be held on Tuesday,\nMarch 9, 1999 from 1:30 to 3:00 p.m. in the Senate Agriculture, Nutrition and Forestry\nCommittee Room in 328A of the Russell Senate Building.\nPlease see the attached brochure for more information. If you have any questions\nregarding this forum, please call Anna Lovejoy at APHSA, 202-682-0100. We look\nforward to your attendance at this important policy discussion.\nAPHSA nonprofit, bi-partisan organization of individuals and agencies concerned with\nhuman services. Our mission is to develop, promote, and implement public human\nservice policies that improve the health and well-being of families, children, and adults.\nSincerely,\nWilliuab\nWilliam A. Waldman\nExecutive Director\nRepresenting Public Human Services Since 1930\n810 First Street, NE, Suite 500, Washington, DC 20002-4267\n(202) 682-0100\nfax: (202) 289-6555\nhttp://www.aphsa.org\nWhy the Food Stamp Program No Longer\nMeets the Needs of the Working Poor\nThe strict regulations governing the\nFood Stamp Program. These two separate\nThe Food Stamp Program also requires\nadministration of the federal Food Stamp\nagencies periodically promulgate inconsistent\nfamilies to report normally every three months\nProgram pose serious obstacles to states'\npolicies that increase the cost of administering\nfor in-person interviews, immediately\nstrategies to reduce dependence on welfare,\nthe Food Stamp Program and impede state\npass along nearly every variation in their\nincrease the employment of Temporary\nwelfare reform innovations. This was true under\nwages, and have their employers fill out a\nAssistance for Needy Families (TANF) clients,\nAFDC and, unfortunately, is increasingly the\nsteady stream of paperwork verifying\nand meet the needs of the working poor. The\ncase under TANF.\ntheir work status. These requirements are due\nclash between the TANF and Food Stamp\nto the historical focus on total and predictable\nprograms is costly and complex. State\nExamples of Program Conflicts\naccuracy of payment levels, whereby\nadministrators see the conflicts of the two\nindividuals and states are both held\nprograms on a daily basis because the TANF\nFor example, under the state TANF programs,\nresponsible for \"errors\" and face penalties for\nand Food Stamp programs are often\nclients are receiving assistance to purchase or\ninaccurate wage reporting. But the nature of\nadministered by the same human service\nrepair cars to get to work. However, under the\nthe entry-level job market is that fluctuations\nagency charged with moving welfare clients\nFood Stamp Program rules, if a recipient owns a\nare frequent, and many jobs do not last very\noff assistance and into employment. This is\ncar worth more than $4,650, the client is ineligible\nlong. Moreover, as states become more\nnot the case at the federal level where the U.S.\nfor food stamps. Similarly, many states have\neffective in moving TANF clients into jobs,\nDepartment of Health and Human Services\nincreased income and asset disregards\nthey run the risk of increasing \"errors\" due to\noversees the TANF block grant and the U.S.\nand established individual development accounts\nthe uncertainty of the entry-level job market.\nDepartment of Agriculture administers the\nfor TANF clients. These assets may make families\nWhile all state administrators hold the\nineligible for food stamps.\nefficient and effective administration of food\nstamps as a goal, adjustments are needed to\nhelp support those food stamp clients who are\nmaking this effort, rather than burden them\nFOOD COUPON\nand penalize states.\nThis forum will present the states' ideas for a\nmore flexible and innovative approach to\nserving working families in the Food Stamp\nProgram. State human service agency\ncommissioners will share their experiences in\nfacing these policy challenges and will offer\ntheir thoughts on what policy changes might\nhelp them serve families better.\nFood Stamps, TANF, and the\nFounded in 1930, the American Public\nWorking Poor\nHuman Services Association (APHSA) is a\nFood Stamps,\nnonprofit, bipartisan organization of state\nTuesday, March 9, 1999\nand local agencies concerned with human\nTANF, and the\n1:30-3:00 p.m.\nservices. Members include all state and\n328A Russell Senate Building\nmany territorial human service agencies,\nWorking Poor\nmore than 800 local agencies, and several\nWelcome and Introductions\nthousand individuals who work in or\notherwise have an interest in human\nWilliam Waldman, Executive Director,\nAPHSA\nservice programs. APHSA, which recently\nchanged its name from the American Public\nPanel Discussion of State\nWelfare Association, serves as a conduit\nState human service commissioners call\nCommissioners\nbetween members of Congress, the media,\nfor changes in the Food Stamp Program\nand the broader public on what is\nGary K. Weeks, Director, Oregon\nto meet the needs of the working poor.\nhappening in the states and localities\nDepartment of Human Resources\nHill staff will have an opportunity to\nconcerning welfare, child welfare, health\nquestion the panelists and hear state\nClarence Carter, Commissioner,\ncare reform, and other issues involving\nperspectives.\nVirginia Department of Social Services\nfamilies, children, and the elderly.\nLynda Fox, Secretary, Maryland\nDepartment of Human Resources\nThe mission of APHSA is to develop, promote,\nand implement public human service policies\nQuestions from Congressional\nthat improve the health and well-being of\nStaff Panel\nfamilies, children, and adults.\nMike Ruffner, Majority Professional\nFOOD COUPON\nStaff Member, Senate Agriculture, Nutri-\ntion and Forestry, Subcommittee on\nResearch, Nutrition, and General Legisla-\ntion\nMark Halverson, Minority Chief Coun-\ncil, Senate Agriculture, Nutrition and\nFor media inquiries, contact\nForestry Committee\nAmy Tucci, Communications\nKevin Kramp, Majority Staff Director,\nDirector, at (202) 682-0100,\nSponsored by the\nHouse Agriculture Subcommittee on\next. 288, or e-mail to\nAmerican Public Human\nDepartment Operations, Nutrition and\[email protected]\nForeign Agriculture\nServices Association\nKimberly Barnes O'Connor, Children's\n810 First Street, NE Suite 500\nPolicy Director, Senate Health, Education,\nWashington, DC 20002-4267\nTuesday, March 9, 1999\nLabor and Pensions Committee\n(202) 682-0100\n1:30 p.m. to 3:00 p.m.\nDebbie Banks, Senior Policy Advisor,\nfax: (202) 289-6555\n328A Russell Senate\nRep. Robert T. Matsui, (D-Calif.)\nhttp://www.aphsa.org\nOffice Building\nWho is Leaving the Food Stamp Program?\nSubgroup\nParticipants\nParticipants\nParticipation\nPercent\nSubgroup as\nSubgroup as\nPercent of\nin 1994\nin 1997\nChange 1994-\nChange in\nPercent of\nPercent of\nDecline\n(thousands)\n(thousands)\n1997\nPartici-\n1994 caseload\n1997 caseload\nfrom\n(thousands)\npation\nSubgroup\n1994-1997\nLegal\n1,537\n706\n-831\n-54%\n6%\n3.4%\n14%\nPermanent\nResident Aliens\nChildless\n1,148\n648\n-500\n-43.5%\n4%\n3.1%\n8.4%\nUnemployed\nAdults\nAFDC/TANF\n13,052\n9,442\n-3,610\n-27.6%\n48%\n43%\n60.8%\nRecipients\nAll Other\n11,697\n10,707\n-990\n-8.5%\n43%\n49%\n16.7%\nParticipants\nTotal\n27,434\n21,503\n-5,931\n-21.6%\n100%\n100%\n100%\nParticipants\nJuly-September Food Stamp Program Quality Control Data for 1994 and 1997\ndrugs and their implications for\nwhether the United States is serious\nto other federal entitlement pro-\nMedicaid. Dr. Garth L. Splinter of the\nabout providing health care to unin-\ngrams. She points out some new and\nOklahoma Health Care Authority\nsured children in immigrant families.\nunanticipated challenges for program\nshares his views on Viagra and the\nShe calls our attention to the current\nadministrators.\ndilemmas Medicaid directors face\nconflict between public health and\nI hope you'll find this issue inter-\nwhen dealing with quality-of-life\nimmigration policy, and she points\nesting. Our goal is for POLICY &\ndrugs. He offers much food for\nout its impact on children's health.\nPRACTICE to be a useful resource for\nthought as he attempts to answer:\nIn our \"View from the Nation's\nyou in your daily work. I'm open to\nHow do we determine what Medicaid\nCapital,\" Elaine M. Ryan, our govern-\nyour suggestions and comments, so\nshould and should not cover?\nment affairs director at APHSA, looks\ndon't hesitate to contact me.\nClaudia Schlosberg of the National\nat the Temporary Assistance for\nBest wishes for a happy and safe\nHealth Law Program questions\nNeedy Families program and its link\nholiday season and new year.\nTravel without good planning,\nand you're bound\nto get lost.\nHowever far you wander.\ntive rates. If you're less than\nTerm Lite\nyou need an insurance plan\nenchanted with your current\nMember Assistance\nthat stays in place. Pick one\ninsurance carrier. call us at\nHigh Limit Accident\nup from APHSA and you're\n800 424-9883. or in Wash-\ncovered wherever you go.\nington, DC at 202 457-6820\nAs long as you're \" member,\nWe promise you'll never get\nwe'll help you find your way\n\" crumby deal. And you'll\nto the best of plans, and\nnever have to go into the\nalways at the most competi-\nwoods alone.\nINSURANCE FOR APHSA MEMBERS\nThis plan is administered by Scabury & Smith, a Marsh & McLennan Company.\nThe term life plan is underwritten by the New York Life Insurance Company.\nAPHSA\n51 Madison Avenue, New York, NY 10010. on Policy Form GMR\nDECEMBER 1991\nView from Nation's\nthe\nCapital\nParcy\nPractice\nJamue\n12/98\nThe Unfinished\nAgenda:\nTwo Years after TANF\nor more than six decades, the\nconnections. Today, we know that\ngies also provide families with\nF\nAid to Families with Dependent\nthose intricate interconnections that\nshort-term, lump-sum assistance\nChildren (AFDC) program was the\nonce worked well (in a system the\naimed at reducing dependence on\nentryway poor\nnation grew to despise) now\nwelfare. More states are granting a\nfamilies walked\nthreaten the long-term success\nmix of cash, benefits, and services\nthrough to gain\nof welfare reform.\ntailored to meet the individual\nby Elaine M. Ryan\naccess to a con-\nWelfare reform marked the end of\nneeds and circumstances of eligible\ncatenation of fed-\nan era of an individual entitlement to\nfamilies. Unlike the previous systen\neral benefit pro-\ncash and child care assistance in\nadministered primarily by human\ngrams. Under the\nthis country and the beginning of a\nservice agencies, today states are\nAFDC program,\nnew era of time-limited, work-focused\nchanging the way they deliver ser-\nfamilies were eligi-\nassistance and services. The\nvices to TANF families by devolving\nble for food\nTemporary Assistance for Needy\nresources and responsibilities to\nstamps, Medicaid,\nFamilies (TANF) block grant that\nnew public, private, and not-for-\nchild support. child care, and, in\nreplaced AFDC provides states with a\nprofit partners.\nsome instances, federal housing and\nfixed amount of federal funds and\nThe five-year federal lifetime limit\nenergy assistance.\npermits broad flexibility in setting eli-\non the receipt of TANF assistance\nOver time, AFDC and other federal\ngibility levels by disregarding earned\nhas created a new imperative to\nentitlement programs were inextrica-\nincome and assets and by providing\nmove clients into employment as\nbly entwined, bound together by a\nfamilies with a wide array of cash\nquickly as possible and to support\ncomplex set of federal regulations,\nand supportive work services.\ntheir transition into the world of work\nfunding streams, and policies. Since\nIn response, states are adopting\nFood stamps, Medicaid, child care,\nthe enactment of the Personal\na variety of innovative policies and\nand housing are key components of\nResponsibility and Work Opportunity\nstrategies which allow families on\nstate job retention strategies for\nReconciliation Act of 1996 that\nwelfare to keep more of their\nthese new low-income workers. Over\nrepealed AFDC, we have begun to\nearned income and to accumulate\ntime, families may increase their\nfully appreciate those intricate inter-\nassets. These policies and strate-\nearnings such that they no longer\n8\nPOLICY & PRACTICE\nrequire this assistance; however, in\npolicy and program conflicts\nFood Stamp program. These two\nthe short-run, these critical programs\ndescribed below.\nseparate agencies periodically pro-\nprovide families the stability they\nmulgate inconsistent policies that\nneed to stay in the workforce and off\nThe Food Stamp Program\nincrease the cost of administering\nof welfare.\nThe strict regulations governing the\nthe Food Stamp program and impede\nAs TANF has diverged from the\nadministration of the federal Food\nstate welfare reform innovations.\nrigid rules of the former AFDC pro-\nStamp program pose serious obsta-\nThis was true under AFDC and, unfor-\ngram, its link to the remaining federal\ncles to states' strategies to reduce\ntunately, is increasingly the case\nentitlement programs has produced\ndependence on welfare and increase\nunder TANF.\nnew and unanticipated challenges for\nthe employment of TANF clients. The\nToday, more than 30 states cur-\nprogram administrators as they\nclash between the TANF and Food\nrently operate TANF diversion pro-\nimplement welfare reform measures.\nStamp programs is historic, costly.\ngrams that provide short-term assis-\nIn reviewing the first two years of\nand complex. State administrators\ntance to low-income families as an\nimplementing TANF, we find the inter-\nsee the conflicts of the two programs\nalternative to TANF receipt. Diversion\nactions between program rules and\non a daily basis because the TANF\nor welfare avoidance payments often\npolicy objectives of TANF and such\nand Food Stamp programs are often\ntake the form of lump-sum cash\nfederal programs as Food Stamps\nadministered by the same human\nassistance to help families overcome\nand Medicaid. for example, no longer\nservice agency charged with moving\na crisis or to remove an obstacle to\nmake sense in the post-AFDC world\nwelfare clients off assistance and\ntheir employment. Under diversion\nand threaten the long-term success\ninto employment. This is not the\nprograms, families may receive three\nof welfare reform.\ncase at the federal level where the\nmonths of benefits in one month or\nThis unfinished agenda of welfare\nU.S. Department of Health and\nassistance to repair an automobile\nreform will require significant federal\nHuman Services oversees the TANF\nneeded to transport a client to his\nstatutory and regulatory change in\nblock grant and the U.S. Department\nor her job.\norder to untangle the conflicting\nof Agriculture (USDA) administers the\nEarlier this year, the USDA Food\nand Nutrition Service (FNS) mandat-\ned that when states determine eligi-\nbility for food stamps, they must con-\nsider diversion payments as income.\nUnder this mandate, many of the\nfamilies who received diversion\npayments would be ineligible for\nfood stamps.\nIn response to this ruling. states\nargued that diversion assistance pro-\nvided to families in crisis or in an\neffort to remove an obstacle to\nemployment should not disqualify\nfamilies from receiving the food\nassistance they need. Lump-sum\nassistance should not be equated to\nincome because a family may have\nresources in one month and be desti-\ntute the next. More important, once\nthe link to food stamps is broken,\nfamilies may be discouraged from\napplying for benefits they are entitled\nto receive in the future.\nLater in the year, FNS revised its\noriginal guidance and replaced it with\na moderately more generous but\nadministratively cumbersome one.\nThe new rules disregard diversion\npayments as income if states docu-\nment in writing that a payment was\nprovided to a family \"only once in a\n12 month period for needs not to\nexceed a 90 day period.' Therefore.\n9\nDECEMBER 1998\nif two or more payments are made to\nmaintenance system.\nand must be streamlined. Absent\na family within a year, the state must\nAs a result, the more effectively\nsuch an effort. the TANF and Food\nconsider these payments as income.\nstates move TANF clients into jobs, it\nStamp programs will continue on the\nClearly, states need greater flexibil-\nis more likely states will be penalized\ncurrent collision course with costly\nity to operate diversion programs\nfor payment errors. The efficient and\nand unintended consequences for\nthan this policy allows. It is impossi-\neffective administration of this feder-\nthe families we serve.\nble to predict when and if families\nal program is a goal of all state\nmay need short-term assistance.\nadministrators and the federal food\nMedicaid\nWhile states could avoid this policy\nstamp error rate should reflect seri-\nThe welfare reform law also broke the\nobstacle by assigning all families to\nous patterns of payment inaccuracy,\nlink between TANF and Medicaid eligi-\nthe welfare rolls. this would be con-\nnot minor errors attributable to suc-\nbility. Despite efforts to manage the\ntrary to the goal of reducing long-\ncessful welfare-to-work efforts.\ntwo programs together, several\nterm dependence on assistance.\nThese are but two examples of the\nadministrative challenges have\nIf this is the extent of flexibility the\nclash between the TANF and Food\nemerged that could have serious\nadministration believes it can exer-\nStamp programs, yet many others\nimplications for families in need.\ncise within the Food Stamp statute,\nexist. Although the welfare reform\nUnder the former AFDC program,\nthen perhaps congressional action is\nlaw includes an option for states to\nclients were categorically eligible for\nnecessary to afford states greater\ncoordinate the two programs together\nMedicaid. Now. the law requires state\nflexibility in aligning the TANF and\nunder the Simplified Food Stamp pro-\nadministrators to refer to the income\nFood Stamp programs.\ngram, few states have opted to do\nand resource standards that were in\nAnother illustration of the need for\nso. In fact, the scope of the \"simpli-\neffect in the state's AFDC plan as of\nincreased program flexibility is the\nfied program\" is so limited that it is\nJuly 16, 1996, to determine appli-\ncalculation of the\nfood stamp error\nrate. Under federal\nlaw, food stamp\nbenefits are adjust-\ned as recipients'\nincomes and earn-\nings change. States\nare measured on\ntheir ability to accu-\nBORSPINDLE\nrately calculate the\nfood stamp benefit\namount. If the so-\ncalled \"error rate\"\nincreases, states\nwill be penalized.\nCalculating the\namount of the\nbenefit was sim-\npler under AFDC\nbecause far fewer\nwelfare clients\nwere earning\nincome from\nemployment and\nremaining on assis-\ntance. In the TANF\nprogram, however, clients' employ-\nmore complex to administer than the\ncants' eligibility. Section 1931 of the\nment may be episodic and the\ncurrent program. With the shared\nwelfare reform law permits states to\namount of earned income may fluctu-\nobjective of ensuring greater access\nincrease income and resource stan-\nate significantly from month to\nto the Food Stamp program for all\ndards (but not above the Consumer\nmonth. The federal method used to\nneedy families, state and federal offi-\nPrice Index), decrease income stan-\ncalculate the food stamp error rate.\ncials must work together in crafting a\ndards (but not below those in place\nhowever. has remained virtually\nsweeping agenda for reform.\non May 1. 1998), or use \"less restric-\nunchanged despite the fundamental\nMoreover. the Food Stamp program\ntive\" methods. The so-called \"look-\nshift away from the AFDC income\nis increasingly difficult to administer\nback\" is administratively burdensome\n10\nPOLICY & PRACTICE\nand confusing to clients and case-\nStatutory or regulatory change is\nFor example, since the enactment\nworkers alike. On a practical level, it\nnecessary to allow states to use\nof welfare reform, 42 states have\nis unreasonable to expect Medicaid\nTANF savings when calculating the\nchanged TANF income and eligibility\neligibility workers to refer to state\ncost-neutrality of future waivers.\nrules that disregard earned income\nplan provisions of an AFDC entitle-\nPerhaps combined savings attribut-\nand 38 states have permitted clients\nment program long-since repealed. A\nable to unobligated TANF block grant\nto retain more resources and still\nnew standard that is simpler to\ndollars and Medicaid could be used\nremain on assistance. These policies\nadminister and easier for families to\nor a modest copayment for working\nare intended to provide clients with\nunderstand must be established.\npoor families could generate the\nan incentive to work and a chance to\nStatutory or regulatory action is\nresources needed to meet the new\naccumulate savings. At the same\nalso necessary to preserve states'\nfederal test. Without a creative solu-\ntime. if these clients live in public\noption to provide more than 12\ntion, thousands of families will lose\nhousing, the earned income is often\nmonths of Medicaid assistance to\nthe critical medical assistance they\nabsorbed by an increase in rent.\nfamilies making the transition from\nmay need to make the successful\nLandlords receive higher payments\nwelfare to work. Under the welfare\ntransition to long-term self-sufficiency.\nbut the net effect to the TANF family\nreform law, states are required to pro-\nAs the differences between the\nis negligible. The Quality Housing and\nvide 12 months of transitional med-\nTANF and Medicaid programs grow, it\nWork Responsibility Act of 1998\nical assistance to individuals whose\nwill become administratively costly to\napproved by Congress this year pro-\nearned incomes increase beyond a\nmanage these programs together and\nvides some relief to families who exit\npoint that would otherwise make\nseparate eligibility systems may be\nTANF for work by restricting rent\nthem eligible for Medicaid under\ndeveloped. In addition, families who\nincreases for up to two years.\nSection 1931. Prior to the enactment\nonce sought assistance through the\nAlthough this is a welcomed reform,\nof the welfare reform law, however,\nagencies that administered the for-\nover time, however, we cannot pre-\nonly 12 states had been granted fed-\nmer AFDC program in the future may\ndict the effect of changing TANF and\neral transitional Medicaid waivers that\ngain access to the Medicaid program\npublic housing policies. We do not\nallowed them to provide Medicaid\nthrough a variety of public, nonprofit,\nknow the number of families whose\nassistance to families for longer than\nor private entities which administer\nincomes will fall to zero when their\none year. Under the current calcula-\nthe TANF program. Without action to\nlifetime time limits are reached and\ntion of cost-neutrality, however, states\nafford states greater flexibility in the\nwhat effect their lack of income may\nwill not be permitted to continue to\nadministration of the Medicaid pro-\nhave on the future cash flow and\nextend this additional Medicaid cover-\ngram, the result may be fewer eligible\ncapacity of public housing authori-\nage to families when their waivers\nfamilies receiving the medical assis-\nties. Therefore, we must monitor the\nexpire. Under the AFDC program,\ntance they are entitled to receive.\nrelationship between welfare reform\nstates were able to achieve cost-neu-\nand housing to preserve the stability\ntrality of their waivers by calculating\nPublic Housing\nTANF clients need to prosper in\nsavings to the federal government in\nBeyond the Food Stamp and\nthe workforce.\nboth the AFDC and Medicaid pro-\nMedicaid programs, there are various\nTANF is a new program constantly\ngrams. States calculated that the\nfederal benefit programs that will\nchanging to meet the diverse needs\ncost of extending transitional\nenable states to achieve success in\nof clients on assistance and clients\nMedicaid for more than one year\nwelfare reform. One such program is\nwho have left the caseload, and to\nwould be offset by the savings the\nhousing assistance. Safe and afford-\nprevent dependence on the welfare\nfederal government would achieve by\nable housing provides the stability\nsystem. We must act now to ensure\nreducing the likelihood that clients\nfamilies need to exit welfare for work.\nthat key federal programs and poli-\nwould return to welfare. Since the fed-\nThough limited in scope, the feder-\ncies are aligned to meet the goals\neral cost of TANF is capped, the\nal government funds housing assis-\nand objectives of welfare reform and\nHealth Care Financing Administration\ntance for low-income families living in\nprovide states all the flexibility they\ninformed states that the cost-neutrali-\npublic housing authorities and\nneed to support state TANF program\nty of future transitional waiver demon-\nthrough Section 8 certificates and\ninnovations. If we fail to do so, mil-\nstrations must be achieved within the\nvouchers. On average, families\nlions of dollars will be spent navigat-\nMedicaid program alone. Under this\nreceiving assistance pay approxi-\ning through the federal entitlement\nnew test, states will not be able to\nmately one-third of their incomes in\nbureaucracy. separate eligibility sys-\ndemonstrate sufficient savings and\nrent. The interplay between TANF\ntems will be established, and fami-\nsubsequent transitional Medicaid\nand the rules governing the adminis-\nlies may lose eligibility and essential\nwaivers will not be approved. Indeed,\ntration of these federal housing pro-\nbenefits in the process.\nin three of the 12 states, the transi-\ngrams often produce unintended neg-\ntional Medicaid waivers have expired\native outcomes for families trying to\nElaine M. Ryan is director of\nand have not been renewed.\nwork their way off welfare.\ngovernment affairs at APHSA.\n11\nDECEMBER 1998\nNO.662 P002/005\n01/07/99 16:47\nCBPP 4567431\nDecline in Food Stamp Participation\nfrom March 1994 to July 1998\nArizona\nMinnesota\nWisconsin\nOhio\nTexas\nIndiana\nKansas\nMississippi\nNew Hampshire\nMassachusetts\nColorado\nFlorida\nIowa\nVirginia\nMissouri\nIdaho\nRhode Island\nNevada\nWyoming\nLouisiana\nUtah\nCalifornia\nPennsylvania\nNorth Dakota\nNew Mexico\nDelaware\nTennessee\nWashington\nVermont\nMichigan\nIllinois\nOklahoma\nNew York\nNew Jersey\nAlabama\nKentucky\nOregon\nGeorgia\nWest Virginia\nMaine\nMaryland\nNorth Carolina\nSouth Dakota\nMontana\nSouth Carolina\nConnecticut\nNebraska\nAlaska\nArkansas\nVirgin Islands\nDist. of Columbia\nHawaii\nGuam\nTOTAL US\n-20\n0\n20\n40\n60\n% Reduction\nCenter on Budget and Policy Priorities\nF:/sip/karen/presentn/1103DS.ps\nNO.662\nP003/005\n01/07/99\n16:47\nCBPP\n4567431\nDecline in Food Stamp Participation Since March 1994\nPercent Decline -\nMarch 1994 to\nSTATE\nMarch 1994\nJuly 1998\nJuly 1998\nALABAMA\n561,803\n415,612\n-26.0%\nALASKA\n51,692\n43,834\n-15.2%\nARIZONA\n512,288\n229,316\n-55.2%\nARKANSAS\n290,212\n254,142\n-12.4%\nCALIFORNIA\n3,191,890\n2,216,978\n-30.5%\nCOLORADO\n275,765\n177,222\n-35.7%\nCONNECTICUT\n225,044\n189,268\n-15.9%\nDELAWARE\n61,717\n43,394\n-29.7%\nDIST. OF COLUMBIA\n92,427\n82,640\n-10.6%\nFLORIDA\n1,469,271\n946,660\n-35.6%\nGEORGIA\n826,295\n635,794\n-23.1%\nGUAM\n15,058\n16,931\n12.4%\nHAWAII\n115,461\n121,388\n5.1%\nIDAHO\n88,094\n58,368\n-33.7%\nILLINOIS\n1,207,279\n875,297\n-27,5%\nINDIANA\n543,248\n301,515\n-44.5%\nIOWA\n203,883\n134,046\n-34.3%\nKANSAS\n197,340\n113,625\n-42.4%\nKENTUCKY\n536,337\n402,139\n-25.0%\nLOUISIANA\n763,280\n525,729\n-31.1%\nMAINE\n141,742\n110,764\n-21.9%\nMARYLAND\n399,877\n314,126\n-21.4%\nMASSACHUSETTS\n451,344\n276,984\n-38.6%\nMICHIGAN\n1,047,450\n755,230\n-27.9%\nMINNESOTA\n323,939\n173,818\n-46.3%\nMISSISSIPPI\n519,006\n309,173\n-40.4%\nMISSOURI\n606,179\n399,431\n-34.1%\nMONTANA\n74,280\n61,610\n-17.1%\nNEBRASKA\n113,484\n95,726\n-15.6%\nNEVADA\n96,977\n67,694\n-31.6%\nNEW HAMPSHIRE\n63,760\n38,518\n-39.6%\nNEW JERSEY\n554,501\n409,894\n-26.1%\nNEW MEXICO\n251,655\n176,779\n-29.8%\nNEW YORK\n2,175,154\n1,592,673\n-26.8%\nNORTH CAROLINA\n646,230\n509,793\n-21.1%\nNORTH DAKOTA\n47,459\n33,118\n-30.2%\nOHIO\n1,269,050\n693,748\n-45.3%\nOKLAHOMA\n382,926\n279,962\n-26.9%\nOREGON\n298,653\n229,435\n-23.2%\nPENNSYLVANIA\n1,243,533\n866,465\n-30.3%\nRHODE ISLAND\n96,336\n63,874\n-33.7%\nSOUTH CAROLINA\n390,940\n326,527\n-16.5%\nSOUTH DAKOTA\n55,200\n44,574\n-19.3%\nTENNESSEEE\n747,371\n527,488\n-29.4%\nTEXAS\n2,764,395\n1,529,964\n-44.7%\nUTAH\n131,124\n90,618\n-30.9%\nVERMONT\n65,475\n46,837\n-28,5%\nVIRGIN ISLANDS\n19,632\n17,501\n-10.9%\nVIRGINIA\n571,254\n376,137\n-34.2%\nWASHINGTON\n483,004\n344,136\n-28.8%\nWEST VIRGINIA\n333,260\n257,976\n-22.6%\nWISCONSIN\n334,900\n181,049\n-45.9%\nWYOMING\n34,698\n23,797\n-31.4%\nTOTAL US\n27,965,172\n10 317\n01/07/99 16:47\nCBPP -> 4567431\nNO.662 P004/005\nDecline in AFDC/TANF Participation compared to Decline in Food Stamp Participation\nFrom March 1994 - June 1998\nPercentage decline in number of cases\nIdaho\nWyoming\nWisconsin\nWest Virginia\nMississippi\nKansas\nFlorida\nOregon\nAlabama\nSouth Carolina\nColorado\nOklahoma\nArkansas\nTexas\nGeorgia\nNorth Carolina\nNew Hampshire\nTennessee\nOhio\nMichigan\nVirginia\nSouth Dakota\nLouisiana\nArizona\nNorth Dakota\nMassachusetts\nIndiana\nMaryland\nMissouri\nKentucky\nUtah\nPennsylvania\nlowa\nMontana\nNew Jersey\nMaine\nDelaware\nConnecticut\nIllinois\nNevada\nVermont\nWashington\nNew York\nNew Mexico\nDist. of Columbia\nCalifornia\nMinnesota\nAlaska\nNebraska\nRhode Island\nHawaii\nTOTAL US\n-40\n-20\n0\n20\n40\n60\n80\n100\n% Reduction\n% Reduction\nCenter on Budget and Policy Priorities\nF:/sfp/karen/presentn/1104SD.ps\nAFDC/TANF\nFOOD STAMP\nAlabama AFDC / TANF Caseload\n70,000\n160%\nCaseload relative to\n1996:8 (left axis)\n60,000\nFight\n140%\n120%\n50,000\nCaseload (left axis)\n100%\n40,000\nCaseload\n80%\n30,000\n60%\nCaseload relative to 1998:6\n20,000\n40%\n10,000\n20%\n0\n0%\n80\n81\n82\n83\n84\n85\n86\n87\n88\n89\n90\n91\n92\n93\n94\n95\n96\n97\n98\n99\nYear\ncasr_fin.wpd\nPage 1\nOctober 21, 1998\nMEMORANDUM FOR THE PRESIDENT\nFROM:\nBruce Reed\nAndrea Kane\nSUBJECT: Minority Welfare Caseloads\nYou asked us to evaluate recent reports that African-American and Hispanic\nfamilies are leaving welfare more slowly than whites, and to consider what more\nwe could do to ensure all welfare recipients are making the successful transition\nfrom welfare to work. Here is a brief summary of the trends, along with some new\nideas we are developing for consideration in your FY 2000 budget.\nI. Caseload Trends\nWe have worked closely with NEC, CEA, OMB, HHS, and the Census Bureau\nto examine the most recent welfare caseload data (generally through 1997). As\ndetailed in the attached tables, the data show:\n1.\nMost of the changes in the welfare caseload can be attributed to changes in\nthe composition of the population as a whole -- specifically, population\ngrowth explains nearly two-thirds of the difference in caseload decline\nbetween Hispanics and whites and nearly half the difference between blacks\nand whites.\nSince 1994, the number of welfare cases has indeed dropped more among\nwhites (26 percent) than among blacks (18 percent) and Hispanics (9 percent).\nHowever, when population growth is taken into account, the difference narrows\ndramatically. The rate of welfare dependency has dropped sharply for all\npopulations - -- by 26 percent among whites, 21 percent among blacks, and 20\npercent among Hispanics. In other words, minorities are leaving or staying off\nwelfare at nearly the same rate as whites, but make up a growing share of the\nwelfare population because they make up a growing share of the population as a\nwhole. [See Table 1.]\n2.\nThe difference in caseload decline is even narrower among adults. Since\n1994, the adult rate of welfare dependency has declined by 30% among\nwhites, 26% among blacks, and 24% among Hispanics.\ncasr_fin.wpd\nPage 3\nA. Current Initiatives\nAs you know, you have put forward many initiatives to help the\nhardest-to-serve welfare recipients and those living in concentrated areas of\npoverty. Many of these initiatives were implemented in 1998, too late to influence\n1994-1997 trends outlined above.\nThe $3 Billion Welfare-to-Work Fund you fought for in the Balanced\nBudget Act is designed specifically to help long-term welfare recipients\n(and non-custodial parents) in high-poverty areas obtain jobs and move\nup a career ladder. While it is too early to have demographic data on\nthe individuals served by these funds, the distribution formula and\nindividual eligibility criteria ensure that most of these funds will be\nspent on minorities. The first of these funds were awarded in January\n1998 and are just now starting to provide individual services.\nWelfare to Work Transportation Funds enacted in the TEA-21\ntransportation reauthorization bill will help welfare recipients and\nlow-income workers get to where the jobs are, often in suburban areas\nnot served by public transportation. The omnibus budget bill includes\n$75 million for this year.\nWelfare-to-Work Housing Vouchers we've proposed will help families\nin isolated urban or rural areas move closer to job opportunities.\nCongress has funded our request for 50,000 vouchers.\nCommunity Empowerment Initiatives. The Administration's\nCommunity Empowerment initiatives -- empowerment zones,\nenterprise communities, Brownfields, CDFIs -- will spur economic\ndevelopment and job creation in distressed neighborhoods and help\naddress the geographic isolation faced by minorities on welfare.\nB. New Initiatives\nIn preparation for next year's budget, we are developing a number of options\nto address the particular challenges faced by minorities in making the transition\nfrom welfare to work:\nIncreasing Investments in English-Language and Literacy Training. We\nhope to recommend targeted new investments in two areas that\ndirectly affect minority and long-term recipients: learning English and\nlearning to read. This could be done by expanding existing\nDepartment of Education adult education programs, or better yet, by\ndedicating Welfare-to-Work funds for job-related literacy and ESL\ncasr_fin.wpd\nPage 4\nprograms, provided either in the workplace or by community\norganizations preparing individuals for employment.\nExpanding Work-Related Drug Treatment. Since many of those\nremaining on welfare suffer from drug or alcohol dependencies, we are\nexploring ways to provide drug treatment for those who agree to go to\nwork.\nTargeting Welfare-to-Work Funds to the Toughest Areas. While the\ncurrent Welfare-to-Work formula favors high-poverty areas, we are\ngoing to examine whether the funds could be even more targeted.\nIncreasing Work and Child Support Among Noncustodial Fathers. We\nmay be able to attract bipartisan support for an effort to help states\nincrease the employment and child support payments of noncustodial\nparents.\nMinority Caseloads Analysis and Tables\nTable 1: Population-Adjusted Change in Rate of Welfare Dependency\nSince 1994, the number of welfare cases has dropped more among whites\n(26 percent) than among blacks (18 percent) and Hispanics (9 percent). However,\nwhen population growth is taken into account, the difference narrows dramatically.\nThe rate of welfare dependency has dropped sharply for all populations -- by 26\npercent among whites, 21 percent among blacks, and 20 percent among Hispanics.\nSpecifically, population growth explains nearly two-thirds of the difference in\ncaseload decline between Hispanics and whites and nearly half the difference\nbetween blacks and whites.\nTable 1: Change from 1994 to 1997*\nRace/Ethnicity\nNumber of Welfare\nPopulation\nRate of Welfare\nCases\nAged 15-49\nDependency (caseload\nadjusted for population)\nWhite\n-26%\n-0.1%\n-26%\nBlack\n-18%\n4.4%\n-21%\nHispanic\n-9%\n13.0%\n-20%\ncasr_fin.wpd\nPage 5\n* National data is only available through June 1997. We do not yet have more recent data, or\nstate-specific data, that we consider accurate. in July, The New York Times reported more recent\ndata provided by some states, but HHS believes that data, particularly for New York and California,\nmay contain significant reporting errors due to states implementation of the new TANF data\nreporting system. The trends in population aged 15-49 are used here because this is the population\ngroup most likely to be a welfare head of household, whose race/ethnicity would be counted when\ntallying the case demographics.\nTable 2: Population-Adjusted Adult Rate of Decline\nThe difference in caseload decline among groups is even narrower for adults.\nChild-only cases are decreasing more slowly than the overall welfare caseload and\nare disproportionately minority; in fact, between 1994 and 1997 they increased\n(though they declined slightly between 1996 and 1997). Child-only cases are those\nin which the parent or adult is not part of the case, (e.g., adult is not a citizen but\nthe child is; child is being cared for by a relative who is not part of the case; parent\nreceives SSI rather than welfare). Therefore, child-only cases are not significantly\naffected by welfare to work efforts. After adjusting for population growth, the rate\nof welfare dependency for adults (percent of 15-49 year old population on welfare)\nhas declined 30% among whites, 26% among blacks, and 24% among Hispanics.\nTable 2: Population-Adjusted Rate of Decline in Adult Welfare Dependency: 1994 -\n1997*\nRate of decline\nRate of\nRate of decline\nPopulation-adjust\nfor all cases\nincrease for\nfor adult-headed\ned rate of welfare\nchild-only\ncases\ndependency for\ncases\nadult cases\nWhite\n- 26%\n7%\n- 30%\n- 30%\nBlack\n- 18%\n3%\n- 23%\n- 26%\nHispanic\n- 9%\n9%\n- 15%\n- 24%\n*National data is only available through June 1997.\nTable 3: Racial Breakdown of Adult Cases\nThe composition of the welfare caseload have changed gradually over the\npast 25 years, driven largely by population changes. Despite differing rates of\ncaseload decline since 1994, the composition of the adult welfare caseload has\nremained relatively constant.\nTable 3: Racial Breakdown of Adult Cases\ncasr_fin.wpd\nPage 6\nRace/Ethnicity\n1994\n1997*\nWhite\n40%\n37%\nBlack\n36%\n37%\nHispanic\n19%\n21%\nAsians, Native Americans, and those designated \"Unknown\" comprise the rest of the caseload.\n*National data is only available through June 1997.\nTable 4: Employment Rate of Welfare Recipients\nThere is encouraging evidence that the employment rates of minority welfare\nrecipients (people on welfare in one year who were working the following year) are\ncatching up with the employment rate for whites.\nTable 4: Employment Rate of Welfare Recipients: 1996-98\nRace/Ethnicity\n1996\n1998\nPercent Change\n96-98\nWhite\n36%\n38 %\n+5%\nBlack\n23 %\n33 %\n+ 44%\nHispanic\n19 %\n29 %\n+ 49%\nAll Recipients\n27%\n34%\n+ 28%\nTable 5: Trends in Marriage Rates and Births\nThe trends in marriage rates and births to unmarried women could\nexacerbate the increasing proportion of Hispanic families on welfare. While the\nproportion of never-married single mothers is increasing for the entire population,\nthe rate of increase is largest for Hispanic women. Also, the birth rate to unmarried\nteenagers remains much higher for blacks and Hispanics than for whites. While the\nrate is decreasing for blacks and slightly for whites, it continues to increase for\nHispanics. For example, between 1991 and 1996, the rate of births to unmarried\nteenagers decreased 18% for blacks and 4% for whites, but increased 3% for\nHispanics.\n1992\n1997\n% Change\n% of all single mothers who were\n30%\n35%\n+ 17%\nnever married\nNever-married single mothers by race/ethnicity:\nWhite\n17%\n21%\n+ 24%\nBlack\n51%\n55%\n+ 8%\ncasr_fin.wpd\nPage 7\nHispanic\n33%\n42%\n+27%\nTable 6: Characteristics of Minorities on the Caseload\nMinorities on welfare are more likely to have characteristics associated with\nlong-term welfare recipiency. Blacks and Hispanics on welfare tend to have lower\neducational levels, marriage rates, and larger families than whites, and are more\nthan twice as likely to live in central cities and areas of concentrated poverty.\nHispanics also have less recent work history than whites or blacks.\nTable 6: Characteristics of AFDC/TANF Recipients by Race/Ethnicity*\nTOTAL\nWHITE\nBLACK\nHISPANIC\n% without HS\n43%\n30%\n43%\n64%\ndiploma\n% never married\n47%\n33%\n69%\n43%\n> 2 children\n29%\n20%\n33%\n39%\nWorked during the\n45%\n49%\n48%\n33%\nyear\nLive in area w/\n48%\n29%\n67%\n58%\npoverty rate > 20%\nLive in central city\n49%\n29%\n68%\n60%\n*These data are from the March 1998 Current Population Survey, showing characteristics of\nrecipients in 1997.\nMinorities are more likely to be long-term welfare recipients. For example, in\n1997, 20 percent of blacks on welfare had been on the rolls for at least five\ncontinuous years, compared to 19 percent for Hispanics and 14 percent for whites.\n12/21/98 09:17\nPROGRAM DEVELOPEMENT DIVISON\n912024567431\nNO. 937 P001/004\nUSDA:FCS:FSP:PDD:PDB\nFAX\n3101 PARK CENTER DRIVE\nROOM 718\nALEXANDRIA, VA 22302\nNumber of pages 12/21/98 including cover sheet\nDate\nTo:\nFrom:\nCynthico Rice\nARTHUR T. FOLEY\nDIRECTOR\nPROGRAM\nDEVELOPMENT\nDIVISION\nPhone\nPhone\n703-305-2490\nFax Phone\n456-7431\nFax Phone\n703-305-2486\nCC:\nREMARKS:\nUrgent\nFor your review\nReply ASAP\nPlease comment\nWho is Leaving the Food Stamp Program?\nDetailed Table\n12/21/98 09:17\nParticipants\nParticipants\nParticipation\nPercent\nPercent of\nSubgroup\nin 1994\nin 1997\nChange in\nDecline from\n(In thousands)\n(in thousands)\n(in\n(ii)\nParticipation\nSubgroup\nLegal Permanent\nResident Aliens\n1,537\n706\n831\n54.07 %\n14.01 %\nChildless\nUnemployed\n1,148\n648\n500\n43.55 %\n8.43 %\nAdults\nAFDC/TANF\nReciplents\n13,052\n9,442\n3,610\n27.66%\n60.87 %\nPROGRAM DEVELOPEMENT DIVISON -> 912024567431\nAll Other\nParticipants\n11,697\n10,707\n990\n8.46%\n16.69%\nTotal\nParticipants\n27,434\n21,503\n5,931\n21.62%\n100.00 %\nNotes:\n1. Participation obtained from July-September Food Stamp Program Quality Control data for 1994 and 1997.\n2. All subgroups are mutually exclusive. Childless unemployed adults do not include Legal Permanent Resident\naliens. Likewise, AFDC/TANF recipients do not include aliens or childless unemployed adults.\nNO.937 P002/004\nWho has left the FSP?\n09:17 12/21/98\n1994 Caseload Distribution\nAliens\nShare of 1994-1997 Decline\nABAWDs\nOthers\nAliens\n17%\n14%\nABAWDs\n8%\n0\nPROGRAM DEVELOPEMENT DIVISON\n>\nAFDC/TANF\n47%\nAFDC/TANF\n912024567431\n61%\nNO.937 P003/004\nAre Children of Legal Immigrants Leaving\nthe Program Faster than Other Children?\n12/21/98 09:17\nChild Liveswith\nChild Does Not Live\nlien:,\nwith Legal- Alien\nNumber of Child\nJumber of Children\nOctober 1996\n1,251,000\n11,034,000\nSeptember 1997\n742,000\n9,804,000\nPROGRAM DEVELOPEMENT DIVISON -> 912024567431\nDrop\n434,000\n1,682,000\nPercent Drop\n36.90%\n14.64%\nNotes:\n1. Numbers obtained from Fiscal Year 1997 Food Stamp Quality Control data.\n2. Children in both groups are citizens. Those living with a legal alien are those whose household contains a legal\npermanent resident who was affected by the restrictions on benefits to immigrants as a result of welfare reform.\nNO.937 P004/004\n931\nTABLE 15-5.-COUNTED (NET) AND BASIC (GROSS) MONTHLY INCOME ELIGIBILITY\nLIMITS FOR THE FOOD STAMP PROGRAM, FISCAL YEAR 1998\n48 States, the\nDistrict of\nHousehold size\nAlaska\nHawaii\nColumbia, and the\nterritories\nCounted (net) monthly income eligibility lim-\nits 1.\n1 person\n$658\n$823\n$756\n2 persons\n885\n1,106\n1,017\n3 persons\n1,11D\n1,390\n1,278\n4 persons\n1,338\n1,673\n1,539\n5 persons\n1,565\n1,956\n1,800\n6 persons\n1,791\n2,240\n2,060\n7 persons\n2,018\n2,523\n2,321\n8 persons\n2,245\n2,806\n2,582\nEach additional person\n+227\n+284\n+261\nBasic (gross) monthly income eligibility lim-\nits 2:\n1 person\n855\n1,070\n983\n2 persons\n1,150\n1,438\n1,322\n3 persons\n1,445\n1,806\n1,661\n4 persons\n1,739\n2,175\n2,000\n5 persons\n2,034\n2,543\n2,339\n6 persons\n2,329\n2,911\n2,678\n7 persons\n2,623\n3,280\n3,018\n8 persons\n2,918\n3,648\n3,357\nEach additional person\n+295\n+369\n+340\n1 Set at the applicable Federal poverty guidelines, updated for inflation through calendar 1996.\n2 Set at 130 percent of the applicable Federal poverty guidelines, updated for inflation through cal-\nendar 1996.\nSource: U.S. Department of Agriculture, Food and Consumer Service.\nCounted liquid assets include cash on hand, checking and sav-\nings accounts, savings certificates, stocks and bonds, individual re-\ntirement accounts (IRAs) and \"Keogh\" plans (less any early with-\ndrawal penalties), and nonrecurring lump-sum payments such as\ninsurance settlements. Certain \"less liquid\" assets are also counted:\na portion of the value of vehicles (generally, the fair market value\nin excess of $4,650) and the equity value of property not producing\nincome consistent with its value (e.g., recreational property).\nCounted assets do not include the value of the household's resi-\ndence (home and surrounding property), business assets, personal\nproperty (household goods and personal effects), lump-sum earned\nincome tax credit payments, burial plots, the cash value of life in-\nsurance policies and pension plans (other than Keogh plans and\nIRAs), and certain other resources whose value is not accessible to\nthe household or are required to be disregarded by other Federal\nlaws.\nWork-related requirements\nUnless exempt, most able-bodied adults must (to gain or retain\neligibility) (1) register for work (typically with the welfare agency\n266 / THE RAND JOURNAL OF ECONOMICS\nThis solution is implemented by a tariff P(q) such that the marginal price P'(q) is\nlarger than the marginal cost C'(q) for all relevant consumption levels. Moreover, it is\nimmediate to verify that P(q) C(q). All consumers thus contribute to profits, both in\nabsolute and in marginal terms.\n3. Competitive bypass\nThe market structure. We now assume that competitors can sell direct access to\nlong distance, using the same bypass technology as the operator. We assume perfect\ncompetition between bypass operators, so that consumers opting for the bypass tech-\nnology face a tariff cq + ƒ, in which case a type-Θ consumer buys q(Θ) and gets a\nutility S(Θ).\nIf the local network operator wants to attract a type-Θ consumer, it must give him\na utility above S(Θ). This imposes a new participation constraint:\nU(Θ) ≥ max {S(Θ), 0}.\n(IR)\nThe optimal tariff now maximizes W subject to (IC) and this new participation\nconstraint (IR). We denote by qᶜ(Θ) the quantity consumed by type Θ under the optimal\ntariff Pᶜ(q), and by Uc(Θ) the consumer's utility.\nAs before, the incentive-compatibility conditions (IC) imply that consumption in-\ncreases with Θ. Together with technical efficiency, this implies that consumers can be\ndivided among three groups: those below some threshold, Θᶜ, are excluded; those above\nanother threshold, Θᶜ, are served through the bypass technology and consume more\nthan q;; and those in the middle range, between Θᶜ and Θᶜ, are connected to the network\nand consume less than q.-\nWe start the analysis with a result that is of general validity for the problems\naddressed here and in particular does not depend on the precise cost structure.\nProposition 1. No quantity is served at a tariff below (minimum) cost. As a conse-\nquence, bypass is priced at cost.\nThe basic intuition is as follows. Suppose that some consumers are served at a\nprice below cost. Raising the price of the consumption of all these consumers up to\ncost obviously reduces their utility, without affecting the choices of the others. How-\never, it leaves open for these consumers the option of gaining at least the total surplus\ngenerated before (by keeping their consumption unchanged). Hence, given that profit\nis now positive or null on all types, the surplus levels generated by the new choices\nof these agents must then be higher than the previous levels. As a consequence, since\nutilities have decreased, total welfare (W = S AU) increases.\nProposition 1 has an interesting implication:\nCorollary 1. The optimal strategy is the same whether the operator is allowed to sell\nbypass or not.\nProof. From Proposition 1, one optimal strategy is to sell no bypass (i.e., to leave\nbypass, which generates no profit, to competitors) and to sell qᶜ(Θ) to types\nObviously this strategy remains optimal if the operator is not allowed to sell bypass.\nQ.E.D.\nSo far, we have assumed that competition among bypass operators leads to mar-\nginal cost pricing (for both access and use). Using a standard Bertrand type of argu-\nment, the next result shows that, more generally, the same pricing policy would emerge\nin equilibrium as soon as the incumbent monopoly and at least one competitor have\naccess to the bypass technology.\n934\ntion); and States may impose food stamp disqualification when an\nindividual is disqualified from another public assistance program.\nAutomatic disqualification is required for those applying in mul-\ntiple jurisdictions, fleeing arrest, or convicted of a drug-related fel-\nony. And States may disqualify individuals not cooperating with\nchild support enforcement authorities or in arrears on their child\nsupport obligations.\nBENEFITS\nFood stamp benefits are a function of a household's size, its net\nmonthly income, and maximum monthly benefit levels (in some\ncases, adjusted for geographic location). An eligible household's net\nincome is determined (i.e., deductions are subtracted from gross in-\ncome), its maximum benefit level is established, and a benefit is\ncalculated by subtracting its expected contribution (30 percent of\nits counted net income) from its maximum allotment. Thus, a 3-\nperson household with $400 in counted net income (after deduc-\ntions) would receive a monthly allotment of $201 (the maximum 3-\nperson benefit in the 48 States, $321, less 30 percent of net income,\n$120).\nAllotments are not taxable and food stamp purchases may not be\ncharged sales taxes. Receipt of food stamps does not affect eligi-\nbility for or benefits provided by other welfare programs, although\nsome programs use food stamp participation as a \"trigger\" for eligi-\nbility and others take into account the general availability of food\nstamps in deciding what level of benefits to provide. In fiscal year\n1996, monthly benefits averaged $73 a person and about $183 a\nhousehold.\nMaximum monthly allotments\nMaximum monthly food stamp allotments are tied to the cost of\npurchasing a nutritionally adequate low-cost diet, as measured by\nthe Agriculture Department's Thrifty Food Plan (TFP). Maximum\nallotments are set at: the monthly cost of the TFP for a four-person\nfamily consisting of a couple between ages 20 and 50 and two\nschool-age children, adjusted for family size (using a formula re-\nflecting economies of scale developed by the Human Nutrition In-\nformation Service), and rounded down to the nearest whole dollar.\nAllotments are adjusted for food price inflation annually, each Oc-\ntober, to reflect the cost of the TFP in the immediately previous\nJune.\nMaximum allotments are standard in the 48 contiguous States\nand the District of Columbia; they are higher, reflecting substan-\ntially different food costs, in Alaska, Hawaii, Guam, and the Virgin\nIslands (table 15-6).\nMinimum and prorated benefits\nEligible one- and two-person households are guaranteed a mini-\nmum monthly food stamp allotment of $10. Minimum monthly ben-\nefits for other household sizes vary from year to year, depending\non the relationship between changes in the income eligibility limits\nand the adjustments to the cost of the TFP. In a few cases, benefits\ncan be reduced to zero before income eligibility limits are exceeded.\nCURIEN, JULLIEN, AND REY /\n267\nCorollary 2. The operator's optimal pricing strategy, the solution to the above program,\nis also part of a Nash equilibrium, in which the bypass competitor(s) prices (price)\naccess and usage at marginal cost.\nProof. By construction, the above optimal pricing policy is a best response to such\npricing behavior from bypass operators. But this policy leaves no room for positive\nprofits for the bypass operators, which thus cannot do better than marginal cost\npricing. Q.E.D.\nTariff, consumptions and market shares. The precise analysis of the optimi-\nzation problem is given in the Appendix. The first-order condition with respect to q(Θ)\nyields\nwhere φ(Θ) is the shadow value of the incentive constraint.\nThis condition expresses the familiar tradeoff between informational rents and\nefficiency. φ(Θ) measures the cost of the rents that must be given to consumers to keep\ntypes close to Θ from consuming q(Θ). As usual, this cost induces a deviation from the\n\"first-best\" level; however, in contrast with the standard case, φ may be either positive\nor negative. Underconsumption is associated with φ(Θ) < 0 and a marginal price above\nmarginal cost. This occurs when the problem faced is to prevent larger consumers from\nreducing their consumption, as in the standard nonlinear pricing model. Incentive com-\npatibility is then best achieved by reducing the consumption q(Θ)-along with the total\nprice P(q(Θ))-since low-quantity/low-price bundles are less attractive for higher-val-\nuation consumers. But φ can also be positive, in which case q(Θ) is larger than q(Θ).\nThis can be understood as follows. The operator has to match the bypass tariff, which\nis particularly attractive for large levels of consumption. This not only benefits large\nusers, it also gives medium-size users an incentive to increase their consumption. To\nprevent this, the operator raises the consumption of large consumers compared to the\nfirst best-along with an increase of the total bill-to make the quantity/price bundle\nless attractive to consumers with lower willingness to pay. As a result,\nProposition 2. There exists σᶜ and Θᶜ such that Θᶜ < Θᶜ < Θᶜ ≤ Θᶜ and\nqᶜ(Θ) < qFB(Θ) on 10c, σ[,\nqᶜ(Θ) > qFB(Θ) on ]Φ, Θ°[,\n> S(Θ) for all Θ <\nTo prevent medium-range consumers (between Θᶜ and Θᶜ) from using bypass, the\noptimal tariff induces overconsumption-the marginal price P'(q) on the local network\nis thus below the marginal cost c for those consumers' consumption levels. This occurs\nbefore the participation constraint starts to be binding (for all consumers between Θᶜ\nand Θc). It is even possible that all consumers connected to the network, including those\nwho overconsume, receive a strictly positive net surplus (i.e., Öᶜ = Θc).\nLemma 1. If c is small enough, Uc(Θ) > S(Θ) for all Θ < Θᶜ.\nClearly, given the consumption profile, the allocation of consumers between the\ntwo technologies is technically efficient. However, the allocation involves global in-\nefficiencies in market shares due to the quantity distortions.\n935\nAt present, minimum monthly allotments for households of three\nor more persons range from $2 to over $80.\nIn addition, a household's calculated monthly allotment can be\nprorated (reduced) for 1 month. On application, a household's first\nmonth's benefit is reduced to reflect the date of application. If a\npreviously participating household does not meet eligibility recer-\ntification requirements in a timely fashion, but does become cer-\ntified for eligibility subsequently, benefits for the first month of its\nnew certification period normally are prorated to reflect the date\nwhen recertification requirements were met.\nTABLE 15-6.-MAXIMUM MONTHLY FOOD STAMP ALLOTMENTS, FISCAL YEAR 1998\n48\nStates\nand the\nHousehold size\nAlaska I\nHawaii\nGuam\nVirgin\nDistrict\nIslands\nof Co-\nlumbia\n1 person\n$122\n$154\n$197\n$180\n$157\n2 persons\n224\n283\n361\n331\n288\n3 persons\n321\n405\n517\n474\n413\n4 persons\n408\n514\n657\n602\n525\n5 persons\n485\n611\n780\n715\n623\n6 persons\n582\n733\n936\n858\n748\n7 persons\n643\n810\n1,035\n948\n827\n8 persons\n735\n926\n1,183\n1,083\n945\nEach additional person\n+92\n+116\n+148\n+135\n+118\n1 Maximum monthly allotments for designated urban areas of Alaska. Two separate higher allotment\nlevels are applied in remote rural areas of Alaska. They are 28 and 55 percent higher than the urban\nallotments shown here.\nSource: U.S. Department of Agriculture.\nApplication, processing, and issuing food stamps\nFood stamp benefits normally are issued monthly. The local wel-\nfare agency must either deny eligibility or make food stamps avail-\nable within 30 days of initial application and must provide food\nstamps without interruption if an eligible household reapplies and\nfulfills recertification requirements in a timely manner. Households\nin immediate need because of little or no income and very limited\ncash assets, as well as the homeless and those with extraordinarily\nhigh shelter expenses, must be given expedited service (provision\nof benefits within 7 days of initial application).\nFood stamp issuance is a welfare agency responsibility, and\nissuance practices differ among welfare agencies. Most food stamp\ncoupons are issued by: (1) providing (usually mailing) recipients an\nauthorization-to-participate (ATP) card that is then turned in at a\nlocal issuance point (e.g., a bank or post office) when picking up\ntheir monthly allotment; or (2) mailing food stamp coupon allot-\nments directly to recipients. However, in a growing number of\nStates, electronic benefit transfer (EBT) systems are used. EBT\nsystems replace coupons with an ATM-like card used to make food\npurchases at the point of sale by deducting the purchase amount\nfrom the recipient's food stamp benefit account. EBT issuance is\n268 / THE RAND JOURNAL OF ECONOMICS\nProposition 3. ΘFB and Θᶜ <\nCompared with the full-information case, the operator excludes more small users\nand the range of bypass increases. For small consumers, the analysis is similar to the\npure (protected) monopoly: to reduce rents, the operator excludes the smallest consum-\ners and reduces the consumption of the other ones. But for large users, the operator\nmust leave the bypass surplus to the largest ones and increase the consumption of the\nother ones, which tends to reduce the market share of the network technology. The\nnature of the inefficiencies is illustrated by the boundary conditions\nf = 0\n+ f = cq(Θ) f,\nwhich reduce to\n= )\nS(Φ, qᶜ(Θᶜ)) S(Θc, = g(Θ) (V(q()) V(q°(Θᶜ))).\nThese conditions express the tradeoff between the marginal benefit of serving one\nmore consumer through the local network and either not serving him at all (first condition)\nor serving him using the bypass technology. Transversality conditions, moreover, imply\nφ() < 0 and φ(Θ) > 0. Hence S(Θc, qᶜ(Θᶜ)) > 0 and S(Θ, q°(Θ)) - S(Θ, q())) ≥ 0\n(> 0 if q(Θᶜ) > qᶜ(Θᶜ)). Including more types at the lower margin always raises the\ntotal surplus but forces the operator to reduce the price offered to all inframarginal\nconsumers and reduces profits. The same is true at the higher margin if consumers just\nbelow Θᶜ strictly prefer to stay in the network (then > qᶜ(Θᶜ)). If large customers\nof the network are just indifferent between the network and bypass, they consume as\nunder bypass (qc(Θ) = q(Θ)), and their allocation between the network and bypass is\nonly a matter of technical efficiency; in this case, the market share of bypass is locally\nefficient.\nAlthough there are inefficiencies, Proposition 1 shows that no consumer is globally\nsubsidized (in absolute terms). The optimal tariff can be interpreted as follows. The\noperator lets competitors serve consumers under bypass and uses only the network\ntechnology. It sells only up to a maximal quantity q°(Θᶜ) < qₑ, with an access fee\nP(0) > 0 and a nonlinear variable tariff P(q) - P(0). The marginal price P (g) is above\nthe network marginal cost for low quantities and below the network marginal cost for\nlarge quantities, and it is actually possible to show that P(q) > C(q) for q<q\nComparative statics. We present here some comparative statics that will help\nexplain the impact of bypass. We assume here that the marginal cost of bypass is small,\nso that Lemma 1 applies. (All claims are proved in the Appendix.)\nTo illustrate the nature of the externality imposed by the competitive pressure, let\nus consider an increase in the connection cost of bypass f. This weakens the competitive\npressure on the operator and, as intuition suggests, reduces the market share of bypass.\nObviously, the utility of bypass consumers decreases because of the direct increase in\nthe competitive price for bypass, cq + f. As an indirect consequence, lowering the\nDear Commissioner:\nThe Administration is committed to ensuring that all eligible low-income households have\naccess to vital Federal food assistance programs such as the Food Stamp program. The Food Stamp\nprogram, which provides food assistance benefits to more than 19 million eligible individuals\nnationwide, is the cornerstone of our efforts to guarantee a healthy and nutritious diet to all low-\nincome families. Due to welfare reform and to today's strong economy, many households are\nfinding employment and leaving the Temporary Assistance for Needy Families (TANF) program.\nWe want to continue to work together to ensure that policy innovations in TANF do not\ninadvertently result in diminished access to food assistance for those who still need assistance,\nincluding working families.\nRecently we polled all the State agencies to find out if they were in compliance with food\nstamp application processing regulations. We learned that most State agencies are continuing to\nencourage households to apply for food stamps even if they have been diverted from applying for\nTANF. They are continuing to process food stamp applications even for households that have been\ndenied TANF. Moreover, they are continuing to keep eligible households on the food stamp rolls\nafter they have left TANF. Finally, many State agencies have taken special steps to avoid\nimproperly denying or delaying food stamp receipt by providing special training for eligibility\nworkers and/or sending notices to local offices reiterating the need to follow food stamp application\nprocessing regulations.\nWhile most of the feedback has been positive, your continued cooperation with federal law\nwill ensure that low income families have the opportunity to apply for food stamps and receive\nbenefits to which they are entitled. We want to highlight several circumstances where State\nagencies need to pay particular attention to the food stamp application processing regulations.\nApplication for Benefits: Many States are diverting households from going on\nthe TANF rolls. Households may be required to engage in job search prior\nto filing an application for TANF. These pre-application requirements for\nTANF cannot be applied to food stamps. Therefore, in these instances, it is\nparticularly important that the State agency make sure that applicants are fully\naware of their right to file an application for food stamps.\nProcessing the Application: In some States, applications for TANF are put on\nhold until the applicant has completed the pre-application requirements. In\nthese circumstances, it is imperative that the State agency continue to process\nthe food stamp application and screen it for expedited service despite the delay\nin processing the TANF application.\nDetermining Eligibility: In some circumstances, TANF-required job search may\nresult in an applicant getting a job and therefore being denied TANF based on\nhis/her earnings. Or, the applicant may accept a lump sum diversion payment\nCHARACTERISTICS OF FOOD STAMP HOUSEHOLDS:\nFISCAL YEAR 1997\n(Advance Report)\nUnited States Department of Agriculture\nOffice of Analysis and Evaluation\nFood and Nutrition Service\nOctober 1998\nT\nhe FSP is the nation's largest food\nchildren received an average monthly food stamp\nassistance program. In fiscal year 1997; it\nbenefit of $234, reflecting their relatively large\nserved an average of 22.9 million people\naverage size (3.4 persons). Single-parent homes\nper month. Almost $19.6 billion was paid out in\n(3.1 persons on average) received an average of\nfood stamps that year.\n$228 in food stamps, and multiple-adult homes\nwith children (4.5 persons on average) received an\nFood stamps are made available to virtually all\naverage of $268.\nlow-income households with few resources to\nsupplement their food purchases and help them\nHouseholds containing elderly persons\nmaintain a healthy diet. In fiscal year 1997, food\nrepresented 17.6 percent of all food stamp\nstamps were approximately one-fourth of a\nhouseholds. Slightly over three-quarters of them\nparticipating household's total monthly income\nwere single-person households, which received an\n(cash plus food stamps). If the value of food\naverage monthly benefit of $47. Households\nstamps was counted in addition to cash as gross\ncontaining elderly and other persons received an\nincome, over one-fifth of food stamp households\naverage benefit of $118. A substantial proportion\nwould move from below to above half the poverty\nof food stamp households contained disabled\nline (Figure 1).\npersons (22.3 percent); these households received\nan average benefit of $104.\nIn fiscal year 1997, slightly over half of all food\nstamp participants were children, most of whom\nII. Characteristics of Food Stamp Participants\nlived in single-parent households (Figure 2). The\nremaining participants were nonelderly adults\nAmong adult participants (age 18 or older),\n(40.6 percent) or elderly adults--age 60 or older--\nwomen outnumbered men by over two to one\n(7.9 percent).\n(Table 2). Of the children participating, 34.1\npercent were of preschool age (0 to 4 years), and\nI. Composition of Food Stamp Households\n65.9 percent were of school age (5 to 17 years).\nIn fiscal year 1997, the majority (58.3 percent) of\nThe largest proportion of food stamp participants\nfood stamp households contained children (Table\nwere white, non-Hispanic (40.3 percent); about\n1). Of these households, over two-thirds (69.1\none-third were African-American, non-Hispanic\npercent) were single-parent homes, approximately\n(34.9 percent); and approximately one-fifth were\none-quarter (24.7 percent) were multiple-adult\nHispanic (19.2 percent). The remaining\nhomes, and the remaining households contained\nparticipants were Asian, Native American, or of\nno members over the age of 17. Households with\nanother race or ethnicity (Table 3).\nThis brief was prepared by Scott Cody and Jacquelyn Anderson of Mathematica Policy Research, Inc. Programming\nsupport was provided by Mark Brinkley. The information presented in this brief is based on data collected by the Food\nand Nutrition Service for quality control purposes for fiscal year 1997.\nPage 2\nFIGURE 1\nPOVERTY STATUS OF FOOD STAMP HOUSEHOLDS\nWITH AND WITHOUT FOOD STAMPS, 1997\n51 to 100% of Poverty\n51 to 100% of Poverty\n51.5%\n66.6%\n101% + of\nPoverty\n8.6%\n15.3%\n101% + of\n18.1%\n39.9%\nPoverty\nPoorest Households\nPoorest Households\n(50% of Poverty or Less)\n(50% of Poverty or Less)\nCASH ONLY\nCASH AND FOOD STAMPS\nFIGURE 2\nDISTRIBUTION OF FOOD STAMP PARTICIPANTS, 1997\nChildren in Single\nParent Households\n34.4%\nNonelderly Adults\n40.6%\n2.8% Children in\nOther Households\nElderly Adults 7.9%\n14.3%\nChildren in Multiple\nAdult Households\nSource: 1997 Food Stamp Quality Control sample.\nPage 3\nIII. Benefits and Income of Food Stamp\nMinnesota, New Hampshire, Ohio, and Vermont--\nHouseholds\nstates with typically above-average gross\nhousehold income and below-average household\nThe average monthly food stamp benefit in fiscal\nsize.\nyear 1997 was $169. The majority (68.6 percent)\nof food stamp households received monthly\nV. Change in Participation Over Time\nbenefits of over $100, and 36.4 percent received\nover $200 (Table 4). In fiscal year 1997, 80.7\nThe number of persons participating in the FSP\npercent of all benefits went to households with\ngrew substantially between 1989 and 1994 and\nchildren, and 6.6 percent of all benefits went to\nhas declined steadily since 1994 (Figure 3).\nhouseholds with elderly members (not shown).\nSpecifically, the number of food stamp\nparticipants grew from 18.9 million persons in\nThe average monthly gross income of food stamp\n1989 to a record high of 28.0 million in March\nhouseholds was $558 (Table 5). The majority of\n1994 and has decreased to 22.1 million in June\nhouseholds (78.4 percent) received unearned\n1997. The change in the number of Food Stamp\nincome from one or more of the following\nProgram participants is associated with changes in\nsources: Aid to Families with Dependent\nthe economy, as indicated by the number of\nChildren/Temporary Assistance to Needy\nunemployed persons and the number of persons\nFamilies (AFDC/TANF), General Assistance\nliving in poverty. For example, Figure 3 shows\n(GA), Social Security, Supplemental Security\nthat the surge in FSP participation between 1989\nIncome (SSI), Unemployment Compensation, or\nand 1994 was associated with a worsening\nsome other payment such as veterans benefits or\neconomy, and the drop in participation since 1994\neducational loans. Slightly less than one-forth of\nhas been associated with an improving economy.\nall households (24.2 percent) received earned\nincome. Almost a tenth (9.2 percent) received\nThe changes in program participation have been\nzero gross income.\naccompanied by changes in the composition of the\ncaseload (Table 7). Although households with\nIV. Characteristics of Food Stamp Households\nchildren remain a large proportion of the caseload,\nby State\nthat proportion has dropped from a high of 62.2\npercent in 1992 to 58.3 percent in 1997. On the\nIn fiscal year 1997, over half of all food stamp\nother hand, the proportion of households with\nhouseholds were located in eight states:\nelderly has risen from a low of 15.4 percent in\nCalifornia, Florida, Illinois, Michigan, New York,\n1992 to 17.6 percent in 1997. However, the most\nOhio, Pennsylvania, and Texas (Table 6). The\nsubstantial change in caseload composition has\naverage monthly food stamp benefit varied by\nbeen the dramatic rise in the proportion of\nstate, reflecting differences in income, expenses,\nparticipating households with disabled persons,\nhousehold size, and composition. Average\nrising from a low of 8.9 percent in 1990 to 12.5\nmonthly benefits were largest in Alaska, Hawaii,\npercent in 1994, and from 18.9 percent in 1995 to\nGuam, and the Virgin Islands because the\n22.3 percent in 1997.\nmaximum allotment is set higher in those areas to\naccommodate higher costs of living. Within the\nChanges Under PRWORA. The Personal\ncontiguous United States, food stamp benefits\nResponsibility and Work Opportunity\nwere highest (greater than $200) in Arizona and\nReconciliation Act of 1996 (PRWORA) took\nTexas--states with above-average household size.\neffect in FY 1997. This legislation, enacted\nConversely, average monthly food stamp benefits\nAugust 22, 1996, made the following significant\nwere lowest (less than $140) in Connecticut,\nmodifications to the FSP:2\nPage 4\nFIGURE 3\nFOOD STAMP PROGRAM PARTICIPANTS,\nUNEMPLOYED PERSONS, AND POOR PERSONS\nMillions\n40\nPersons in poverty a\n30\nb\nFSP participants\n20\n10\nUnemployed persons c\n0\n84\n85\n86\n87\n88\n89\n90\n91\n92\n93\n94\n95\n96\n97\nYear\n\"Annual values. The number of persons in poverty in Fiscal Year 1997 was not available when this report went to print.\nSource: Bureau of the Census, Poverty in the United States: 1996, Series P60-198.\nᵇAverage monthly values. Source: Food and Nutrition Service, Fiscal Years 1984-1997 FSP Participation and Issuance.\n°Average monthly values. Source: Economic Report of the President, 1998 Table B-36.\nPage 5\nMost legal permanent resident aliens are\nhouseholds, regardless of almost any nonfinancial\ndisqualified from the FSP.³\ncategorical criteria. Second, instead of cash, it\nprovides benefits in the form of coupons or\nMost able-bodied, nonworking, childless adults\nelectronic benefits which can be redeemed for\nare limited to 3 months of FSP benefits in any\nfood in any of about 200,000 authorized stores\n36-month period.\nacross the nation. The cost of providing food\nstamps to needy persons is funded fully by the\nThe maximum food stamp benefit is reduced\nfederal government. Administrative costs are\nfrom 103 percent to 100 percent of the Thrifty\nshared by federal, state, and local governments.\nFood Plan.\nEligibility. To be eligible for food stamps, a\nThe standard deduction is frozen indefinitely at\nhousehold's assets, gross income, and net income,\nfiscal year 1996 levels.\nwhich is based on gross income less deductions\npermitted under the FSP, must not exceed\nNew shelter deduction caps are established for\nspecified levels that vary by household size,\nfiscal years 1997 through 2001, with the cap\ncomposition, and location. Most permanent\nfrozen at fiscal year 2001 levels thereafter.\nresident aliens are ineligible to join the FSP, and\nmost able-bodied, nonworking, childless adults\nThe fiscal year 1997 data provide the first picture\nare limited to 3 months of FSP benefits in any 36-\nof food stamp participants since PRWORA's\nmonth period.\nimplementation. The data show that the\npercentage of permanent resident aliens\nFood Stamp Benefit Computation and\nparticipating in the FSP fell from 5.6 percent in\nIssuance. Benefits are computed by subtracting\nfiscal year 1996 to 4.4 percent in fiscal year 1997.\n30 percent of a household's net income from the\nIn addition, the percentage of able-bodied,\nmaximum benefit amount, which is based on 100\nnonworking, childless adults participating in the\npercent of the June cost of the Thrifty Food Plan\nFSP fell from 3.6 percent in fiscal year 1996 to\n(TFP) for a family of four, adjusted for household\n2.9 percent in fiscal year 1997. PRWORA is\nsize and location. The TFP is based on the cost of\nlikely a contributing factor behind the continued\na market basket of food that provides an\noverall decline in the FSP caseload and costs\neconomical and nutritious diet. In fiscal year\nsince 1996. However, the extent to which these\n1997, the maximum benefit for a family of four in\ntrends are caused by PRWORA rather than other\nthe contiguous United States was $400 per month.\nfactors such as the economy is difficult to\nProgram participants receive their monthly\ndetermine. Furthermore, PRWORA was\nbenefits through the mail, directly from the local\nimplemented in stages throughout fiscal year\noffice, or through an electronic benefit transfer\n1997, and the average annual estimates provided\n(which is similar to a bank card).\nhere include data from before major components\nof PRWORA were in effect.\nVII. Data\nVI. Description of the Food Stamp Program\nThe estimates presented here are based on data\nextracted from the Integrated Quality Control\nThe FSP is unique among income maintenance\nSystem, which is an ongoing review of food stamp\nprograms in two important ways. First, it offers\nhouseholds designed to measure the accuracy with\nassistance to nearly all financially needy\nwhich eligibility and benefit determinations are\nPage 6\nmade. All estimates are based on a full-year\nsample of 48,854 households.\nFood and Nutrition Service administrative records\nindicate that the FSP served 22.9 million persons\nin fiscal year 1997, and food stamp households\nreceived an average benefit of $173 per month.\nThe figures in the attached tables of 23.1 million\nparticipants with an average household benefit of\n$169 vary from the administrative figures because\nthey are estimates from the Food Stamp Quality\nControl sample, which weights data by\nhouseholds rather than persons or benefits.\nAdministrative figures are based on a monthly\ncensus of actual FSP participation and benefit\nissuance.\nNotes\n¹The 1994-to-1995 increase in the proportion of\nhouseholds with disabled persons is due in part to a\nchange in the definition of households with disabled\npersons. However, using the old definition, the\nproportion with disabled still increases from 12.5\npercent to 13.3 percent.\n²A summary of the PRWORA provisions that affect the\nFSP is available from the FNS World Wide Web site\n(http://www.usda.gov/fcs/fcs.htm).\n3 The Agricultural Research Bill, enacted on June 23,\n1998, restored eligibility to a substantial number of\nlegal immigrants who lost eligibility under PRWORA.\nSpecifically, child, elderly, and disabled permanent\nresident aliens will be able to begin receiving federal\nfood stamp benefits on November 1, 1998.\nPage 7\nTable 1 - Selected Characteristics of Participating Food Stamp Households, 1997\nParticipating households\nAverage monthly values\n(dollars)\nAverage\nhousehold\nHousehold type\nsize\nNumber\nPercent\nFood stamp\nGross\nNet income\n(thousands)\nbenefit\nincome\n(persons)\nTotal\n9,452\n100.0\n169\n558\n299\n2.4\nChildren\n5,508\n58.3\n234\n648\n364\n3.4\nSingle-parent households\n3,806\n40.3\n228\n576\n302\n3.1\nMultiple-adult households\n1,360\n14.4\n268\n917\n590\n4.5\nOther\n342\n3.6\n169\n378\n162\n2.0\nElderly\n1,667\n17.6\n63\n577\n319\n1.3\nLiving alone\n1,292\n13.7\n47\n521\n266\n1.0\nNot living alone\n375\n4.0\n118\n767\n503\n2.5\nDisabled\n2,108\n22.3\n104\n687\n418\n2.1\nLiving alone\n1,110\n11.7\n52\n525\n244\n1.0\nNot living alone\n998\n10.6\n162\n868\n612\n3.4\nOther\n1,307\n13.8\n118\n185\n52\n1.1\nSingle-person\n1,168\n12.4\n110\n159\n38\n1.0\nMultiple-person\n139\n1.5\n185\n407\n170\n22\nSource: 1997 Food Stamp Quality Control sample.\nPage 8\nTable 2 - Gender and Age of Food Stamp Participants, 1997\nAll participants\nFemale\nMale\nAge\nNumber\nNumber\nNumber\nPercent\nPercent\nPercent\n(thousands)\n(thousands)\n(thousands)\nTotal\n23,117\n100.0\n13,880\n100.0\n9,233\n100.0\nChildren (0-17)\n11,871\n51.4\n5,950\n42.9\n5,918\n64.1\n0-4\n4,046\n17.5\n2,017\n14.5\n2,026\n21.9\n5-17\n7,825\n33.8\n3,933\n28.3\n3,892\n42.2\nAdults (18 or more)\n11,219\n48.5\n7,916\n57.0\n3,302\n35.8\n18-35\n5,332\n23.1\n4,007\n28.9\n1,324\n14.3\n36-59\n4,053\n17.5\n2,582\n18.6\n1,472\n15.9\n60 or more\n1,834\n7.9\n1,328\n9.6\n506\n5.5\nUnknown\n27\n0.1\n14\n0.1\n13\n0.1\nSource: 1997 Food Stamp Quality Control sample.\nTable 3 - Race/Ethnicity of Food Stamp Participants, 1997\nParticipants\nRace\nNumber\nPercent\n(thousands)\nTotal\n23,117\n100.0\nWhite, Non-Hispanic\n9,323\n40.3\nAfrican-American, Non-Hispanic\n8,072\n34.9\nHispanic\n4,441\n19.2\nAsian\n705\n3.0\nNative American\n313\n1.4\nOther\n263\n1.1\nSource: 1997 Food Stamp Quality Control sample.\nPage 9\nTable 4 - Monthly Food Stamp Benefits of Participating Food\nStamp Households, 1997\nParticipating households\nMonthly food stamp benefit\nNumber\nPercent\n(thousands)\nTotal\n9,452\n100.0\n-\n-\n$0-50\n1,726\n18.3\n$51-100\n1,236\n13.1\n$101-150\n2,085\n22.1\n$151-200\n969\n10.3\n$201-250\n1,258\n13.3\n$251-300\n660\n7.0\n$301 +\n1,519\n16.1\nSource: 1997 Food Stamp Quality Control sample.\nTable 5 - Selected Economic Characteristics of Participating Food Stamp Households, 1997\nParticipating\nPersons in households\nAverage income\nAverage\nAverage\nhouseholds\nwith source\n(dollars)\nfood\nhouse-\nIncome source\nstamp\nhold\nNumber\nNumber\nFrom\nbenefit\nsize\nPercent\nPercent\nGross\n(thousands)\n(thousands)\nsource\n(dollars)\n(persons)\nTotal\n9,452\n100.0\n23,117\n100.0\n558\n(n/a)\n169\n2.4\nEarned income\n2,284\n24.2\n7,533\n32.6\n879\n708\n187\n3.3\nWages and salaries\n2,139\n22.6\n7,105\n30.7\n898\n728\n185\n3.3\nSelf-employment\n137\n1.5\n424\n1.8\n620\n324\n225\n3.1\nOther earned income\n35\n0.4\n106\n0.5\n817\n421\n178\n3.0\nUnearned income\n7,415\n78.4\n18,036\n78.0\n580\n492\n163\n2.4\nAFDC/TANF\n3,270\n34.6\n10,649\n46.1\n569\n372\n240\n3.3\nGeneral Assistance\n588\n6.2\n899\n3.9\n411\n252\n123\n1.5\nSupplemental Security Income\n2,504\n26.5\n4,782\n20.7\n642\n372\n93\n1.9\nSocial Security\n1,999\n21.1\n3,377\n14.6\n647\n483\n77\n1.7\nUnemployment\n156\n1.7\n509\n2.2\n755\n491\n190\n3.3\nOther unearned income\n1,476\n15.6\n4,128\n17.9\n666\n211\n176\n2.8\nNo Income\n868\n9.2\n1,408\n6.1\n0\n0\n180\n1.6\nSource: 1997 Food Stamp Quality Control sample.\nPage 10\nTable 6 - Selected Characteristics of Participating Food Stamp Households by State, 1997\nAverage monthly amount\nAverage\nhouse-\nTotal\nPercent of all\nState\nFood stamp\nTotal\nCountable\nhold\n(thousands)\nhouseholds\nGross income\nNet income\nbenefit\ndeduction\nassets\nsize\n(dollars)\n(dollars)\n(dollars)\n(dollars)\n(dollars)\n(persons)\nTotal\n9,452\n100.0\n169\n558\n299\n291\n92\n2.4\nAlabama\n186\n2.0\n174\n553\n302\n292\n78\n2.5\nAlaska\n15\n0.2\n273\n928\n607\n382\n117\n3.1\nArizona\n133\n1.4\n203\n569\n305\n294\n76\n2.9\nArkansas\n105\n1.1\n159\n568\n339\n263\n163\n2.5\nCalifornia\n1,045\n11.1\n187\n594\n333\n296\n122\n2.8\nColorado\n91\n1.0\n169\n563\n277\n319\n48\n2.4\nConnecticut\n94\n1.0\n139\n596\n335\n278\n135\n22\nDelaware\n20\n0.2\n175\n577\n305\n309\n105\n2.6\nDist. of Col.\n40\n0.4\n186\n390\n201\n227\n16\n2.3\nFlorida\n514\n5.4\n168\n549\n276\n306\n139\n2.4\nGeorgia\n284\n3.0\n177\n535\n279\n298\n80\n2.5\nGuam\n6\n0.1\n408\n567\n251\n396\n118\n3.3\nHawaii\n57\n0.6\n291\n629\n355\n308\n169\n2.3\nIdaho\n27\n0.3\n175\n622\n338\n322\n175\n2.7\nIllinois\n434\n4.6\n173\n476\n252\n260\n61\n2.4\nIndiana\n140\n1.5\n167\n572\n308\n294\n123\n2.5\nIowa\n67\n0.7\n148\n625\n360\n290\n85\n2.4\nKansas\n64\n0.7\n156\n586\n323\n290\n125\n2.4\nKentucky\n174\n1.8\n165\n551\n341\n243\n109\n2.6\nLouisiana\n220\n2.3\n185\n526\n282\n277\n75\n2.6\nMaine\n58\n0.6\n149\n597\n269\n356\n109\n2.1\nMaryland\n152\n1.6\n184\n445\n220\n264\n24\n2.4\nMassachusetts\n149\n1.6\n148\n635\n333\n318\n95\n2.3\nMichigan\n366\n3.9\n161\n579\n283\n330\n57\n2.3\nMinnesota\n110\n1.2\n139\n573\n331\n270\n159\n2.2\nMississippi\n155\n1.6\n160\n550\n332\n247\n161\n2.5\nMissouri\n199\n2.1\n161\n542\n312\n264\n115\n2.4\nMontana\n27\n0.3\n171\n600\n302\n325\n236\n2.5\nNebraska\n41\n0.4\n150\n630\n361\n290\n238\n2.4\nNevada\n38\n0.4\n165\n520\n257\n305\n67\n2.3\nNew Hampshire\n21\n0.2\n116\n596\n394\n239\n173\n2.2\nNew Jersey\n212\n2.2\n176\n512\n224\n316\n42\n2.3\nNew Mexico\n75\n0.8\n179\n571\n348\n252\n107\n2.7\nNew York\n899\n9.5\n150\n603\n283\n346\n24\n2.2\nNorth Carolina\n250\n2.6\n156\n513\n289\n256\n85\n23\nNorth Dakota\n16\n0.2\n156\n637\n364\n304\n351\n2.5\nOhio\n389\n4.1\n137\n556\n325\n253\n78\n2.2\nOklahoma\n131\n1.4\n168\n534\n299\n277\n79\n2.5\nOregon\n121\n1.3\n141\n516\n289\n276\n249\n2.1\nPennsylvania\n440\n4.7\n161\n531\n265\n302\n93\n2.3\nRhode Island\n37\n0.4\n158\n562\n304\n279\n95\n2.3\nSouth Carolina\n140\n1.5\n160\n550\n346\n234\n104\n2.5\nSouth Dakota\n18\n0.2\n185\n552\n288\n313\n206\n2.6\nTennessee\n253\n2.7\n154\n564\n312\n285\n135\n2.3\nTexas\n751\n7.9\n201\n535\n296\n278\n52\n2.8\nUtah\n38\n0.4\n167\n661\n375\n312\n185\n2.7\nVermont\n25\n0.3\n125\n598\n339\n286\n164\n2.1\nVirgin Islands\n6\n0.1\n322\n490\n292\n223\n93\n3.2\nVirginia\n206\n2.2\n156\n548\n309\n268\n122\n2.3\nWashington\n198\n2.1\n157\n547\n272\n306\n56\n2.2\nWest Virginia\n117\n1.2\n168\n509\n280\n269\n90\n2.4\nWisconsin\n87\n0.9\n151\n689\n427\n287\n148\n2.7\nWyoming\n11\n0.1\n174\n583\n314\n313\n180\n2.6\nSource: 1997 Food Stamp Quality Control sample.\nPage 11\nTable 7 - Selected Characteristics of Food Stamp Households Over Time\nFiscal Year 1989-Fiscal Year 1995\nTotal\nChildren\nElderly\nDisabled\nChildren\nHouseholds\nElderly\nHouseholds\nHouseholds\n(percent\nwith children\n(percent\nwith elderly\nwith disabled\nFiscal\nParticipants\nHouseholds\nof all\n(percent\nof all\n(percent\n(percent\nYear\n(thousands)\n(thousands)\nparticipants)\nof all)\nparticipants)\nof all)\nof all)\n1997\n23,117\n9,452\n51.4\n58.3\n7.9\n17.6\n22.3\n1996\n25,926\n10,552\n51.0\n59.5\n7.3\n16.2\n20.2\n1995\n26,955\n10,883\n51.5\n59.7\n7.1\n16.0\n18.9°\n1994\n28,009\n11,091\n51.4\n61.1\n7.0\n15.8\n12.5\n1993\n27,595\n10.791\n51.5\n62.1\n6.8\n15.5\n10.7\n1992b\n25,743\n10,049\n51.9\n62.2\n6.6\n15.4\n9.5\n1991b\n22,963\n8,855\n52.0\n60.4\n7.0\n16.4\n9.0\n1990b\n20,411\n7,803\n49.6\n60.3\n7.7\n18.1\n8.9\n19892b\n18,925\n7,209\n49.8\n60.4\n8.2\n19.3\n9.1\n\"Full year analysis files were not developed for the years prior to 1989.\nᵇPrior to fiscal year 1993, food stamp cases from Guam and the Virgin Islands were excluded from the analysis files.\n°Beginning In 1995, disabled households are defined as households with at least one member under age 65 who\nreceived SSI, or at least one member age 18 to 61 who received Social Security, veterans benefits, or other\ngovernment benefits as a result of a disability. For years prior to 1995, disabled households are defined as\nhouseholds with SSI but no members over age 59. The substantial increase in the percentage of households with\na disabled member between 1994 and 1995 is due in part to the change in the definition of disabled households.\nUsing the previous definition, 13.3 percent of households included a disabled person in fiscal year 1995.\nSource: Food Stamp Quality Control samples.\nFSP Participants: FY 1988 - 1998\n29,000,000\n27,000,000\n25,000,000\n23,000,000\nPeople\n21,000,000\n19,000,000\n17,000,000\n15,000,000\n1988\n1989\n1990\n1991\n1992\n1993\n1994\n1995\n1996\n1997\n1998*\nYear\n12/1/98\nWho is Leaving the Food Stamp Program?\nDetailed Table\nParticipants\nParticipants\nPaticipation\nPercent\nPercent of\nSubgroup as\nSubgroup\nin 1994\nin 1997\nChange in\nDecline from\n% of 1997\n(In thousands)\n(in thousands)\n(in\nParticipation\nSubgroup\nLegal Permanent\nResident Aliens\n5-6%\n1,537\n706\n831\n54.07%\n14.01%\nChildless\nUnemployed\n4,2%\n1,148\n648\n500\n43.55%\n8.43%\nAdults\nAFDC/TANF\nReciplents\n13,052\n9,442\n47.5%\n3,610\n27.66%\n60.87%\nAll Other\nParticipants\n11,697\n10,707\n990\n8.46%\n16.69%\n39%\nTotal\nParticipants\n27,434\n21,503\n5,931\n21.62%\n100.00%\nNotes:\n1. Participation obtained from July-September Food Stamp Program Quality Control data for 1994 and 1997.\n2. All subgroups are mutually exclusive. Childless unemployed adults do not include Legal Permanent Resident\naliens. Likewise, AFDC/TANF recipients do not include aliens or childless unemployed adults.\nSINGLE PARENT FAMILIES:\n1989, 1994, and 1997\nNumber of Single Parent Families on Food Stamps by Welfare Receipt and Presence of Earnings:\n1989, 1994, and 1997\n5,000\nDRAFT\n4,500\n568\n4,000\n708\n3,500\n614\n297\n3,000\nHouseholds (in thousands)\n324\n780\n2,500\n334\n208\n325\n2,000\n1,500\n3,022\n1,000\n2,094\n2,097\n500\n0\n1989\n1994\n1997\nYear\nSingle parent households with no earnings and no AFDC/TANF\nSingle parent households with earnings and no AFDC/TANF\nSingle parent households with AFDC/TANF and earnings\nSingle parent households with AFDC/TANF and no earnings\n12/1/98\nChanges in the Number of Single Parent Participants on Welfare and Not on Welfare\n3,500\nDRAFT\n3,000\n2,500\nNumber (in thousands)\n2,000\n1,500\n1,000\n500\n0\n1989\n1994\n1997\nSingle parent households on welfare\nSingle parent households not on welfare\n12/1/98\nincludes is 490mp\nI \"day\"\nTWO GROUPS AFFECTED BY\nWELFARE REFORM:\nLEGAL IMMIGRANTS AND\nUNEMPLOYED CHILDLESS\nADULTS AGED 18-59\n1989, 1994, and 1997\nDecline in participation among legal immigrants and childless unemployed adults in the wake of\nwelfare reform\n1,600\nDRAFT\n1,400\n1,200\n1,000\nNumber (in thousands)\n800\n600\n400\n200\n0\nJuly-96\nAugust-96\nSeptember-96\nOctober-96\nNovember-96\nDecember-96\nJanuary-97\nFebruary-97\nMarch-97\nApril-97\nMay-97\nJune-97\nJuly-97\nAugust-97\nSeptember-97\nPermanent Resident Aliens\nUnemployed Childless Adults\n12/1/98\nDesired Dres not\nsholfund\n3 Goof 150mg road wood\nImmigrant Participants by Citizen/Alien Status:\n1989, 1994, and 1997\n2,500\nDRAFT\n2,000\n339\n1,500\nNumber\n235\n1,000\n1,453\n234\n547\n500\n582\n437\n264\n131\n0\n1989\n1994\n1997\nYear\nNaturalized citizens\nPermanent Residents\nRefugees\n12/1/98\nChanges in the Number of Childless Unemployed Adult Participants:\n1989, 1994, and 1997\nDRAFT\n1,400\n1,309\n1,200\n1,000\n833\n800\nNumber\n652\n600\n400\n200\n0\n1989\n1994\n1997\nYear\n12/1/98\nELDERLY FOOD STAMP\nPARTICIPANTS\n1989, 1994, and 1997\nNumber of Aged Food Stamp Participants: 1989, 1994, and 1997\n2,000\nDRAFT\n1,800\n1,600\n1,400\n1,200\nNumber (in thousands)\n1,000\n800\n600\n400\n200\n0\n1989\n1994\n1997\nYear\n12/1/98\nCHANGES IN THE\nRACIAL/ETHNIC\nCOMPOSITION OF THE FOOD\nSTAMP CASELOAD\n1989, 1994, and 1997\nThe Racial/Ethnic Composition of FSP Participants: 1994 and 1997\nFY 1994\nOther\n5%\nRAFT\nHispanic\n19%\nWhite\n41%\nBlack\n35%\nFY 1997\nOther\n4%\nHispanic\n19%\nWhite\n41%\nBlack\n36%\n12/1/98\nRacial Composition of AFDC/TANF Participants:\n1994 and 1997\nFY 1994\nOther\n7%\nDRAFT\nHispanic\nWhite\n21%\n34%\nBlack\n38%\nFY 1997\nOther\n6%\nHispanic\nWhite\n22%\n33%\nBlack\n39%\n12/1/98\nWelfore Reups\npereary f S,\nRacial Composition of Childless Unemployed Adults:\n1994 and 1997\nFY 1994\nDRAFT\nOther\nHispanic\n4%\n7%\nWhite\n44%\nBlack\n45%\nFY 1997\nOther\n3%\nHispanic\n14%\nWhite\n37%\nBlack\n46%\n12/1/98\nTABLES WITH UNWEIGHTED\nQUALITY CONTROL DATA\nFROM 1989, 1994, 1997, and\nPRELIMINARY 1998 FILES\nSources of Income Using Unweighted Quality Control Data:\n1989, 1994, 1997, and 1998 preliminary data\n45.00%\nDRAFT\n40.00%\n35.00%\n30.00%\n25.00%\nPercent\n20.00%\n15.00%\n10.00%\n5.00%\n0.00%\n1989\n1994\n1997\n1998\nYear\nWelfare for Families\n'Welfare for Single Adults\nSupplemental Security Income\nSocial Security\nEarnings\n12/1/98\nPercent of Caseload that are Legal Immigrants or Childless Unemployed Adults\n(Using Unweighted Data):\n1989, 1994, 1997, and 1998 Preliminary Data\nDRAFT\n4.00%\n3.50%\n3.00%\n2.50%\nPercent\n2.00%\n1.50%\n1.00%\n0.50%\n0.00%\n1989\n1994\n1997\n1998\nYear\nAliens\nABAWDs\n12/1/98\nJenny genser\n703- 305- 2152\nJenny genser @fcs. @ usda . gov"
}