December 4,1997

A Guide to Estimate the Cost of a State-Funded
Food Stamp
Program for Legal Immigrants
by Jennifer Daskal and Kelly Carmody

Note: This paper was written prior to the federal restoration of food stamps for some legal immigrants, and will be revised soon to reflect the changes enacted in the new legislation. Until the paper is revised, please see "New Federal Food Stamp Restoration for Legal Immigrants: Implications and Implementation Issues" for a description of which legal immigrants will be eligible for federal food stamp benefits effective November 1, 1998.

A state contemplating funding its own food stamp program for legal immigrants needs to be able to estimate the cost of providing state-funded food stamps to legal immigrants who are ineligible for federally-funded food stamps. Available information can help states develop fairly reliable projections. The following is a step-by-step guide to assist states with the necessary calculations.(1)

The text of this guide describes how to determine the number of immigrants to be served by a state-funded food stamp program and the cost of providing food assistance to these immigrants. To make the discussion more concrete, a sample worksheet shows these calculations for a hypothetical state. The text and worksheet continuously refer to each other, and the two pieces should be reviewed together.

The calculations begin with the number of non-exempt legal immigrants receiving federally-funded food stamps in a typical month in federal fiscal year (FFY) 1996. (Each state's number has been generated from a database compiled by the U.S. Department of Agriculture's Food and Consumer Service (FCS) and is provided in the following Table 1, Column A.) The guide then outlines each of the several steps involved in estimating how many non-exempt legal immigrants will be subject to the new eligibility restrictions two years later in FFY 1998. This estimate is used to calculate the cost of providing food stamps for these immigrants at the same benefit levels that would apply in the federal food stamp program. (All of the calculations are based on the assumption that states use the federal benefit structure in their state-funded programs.) The final section describes and evaluates options for states that are unable to allocate sufficient funding to provide full benefits to all of the legal immigrants losing federally-funded food stamps.

It is important to emphasize that the figures used are estimates, based on assumptions and data supplied by the Congressional Budget Office (CBO) and the Food and Consumer Service (FCS) of the United States Department of Agriculture. Rather, this guide represents Center's best attempt to estimate the number of legal immigrants subject to the food stamp restrictions. The Center recommends that states and state advocates to follow the methodology outlined in this guide.

I. Estimating the Numbers of Legal Immigrants in Need of State-Funded Food Stamps

The number of legal immigrants receiving food stamps at the time the welfare law was signed is not the same as the number of legal immigrants denied food stamps under the provisions of the welfare law. Some immigrants fall within an exempt category and remain eligible for federally-funded food stamps. Some will have naturalized and remain eligible as citizens. Others have exited the rolls as a growing economy pushes federal food stamp caseloads to recent lows. All of these factors reduce the number of legal immigrants who are now impacted by the welfare law's restrictions on federal food stamp eligibility.

On the other hand, other legal immigrants (either those in the country when the welfare law was signed but not receiving food stamps at that time or new arrivals) would have applied for and received federal food stamps if it were not for the welfare law's restrictions on eligibility.

Given these factors, states must make several adjustments to the FY1996 figure presented in Table 1 in order to estimate the number of legal immigrants actually subject to the restrictions on eligibility enacted in last year's welfare law and now in need of state-funded food stamps. The following sections make suggestions as to what these adjustments should be, and the worksheet shows the calculations for the hypothetical state. A blank worksheet is also included for state-specific calculations.

Step One: Adjustment for Exemptions

Step Two: Adjustment for Caseload Decline:

Step Three: Adjustment for Exemptions for Immigrants with 40 Quarters of Work

Step Four: Adjustment for Immigrants who Become Naturalized Citizens:

Step Five: Adjustment for Participation Rates

II. Costs at Full Benefit Level

III. Dealing With Budget Constraints: Some Options

Some states may not appropriate sufficient funds to provide the full benefit level to all legal immigrants denied federal food stamps. These states will have to make difficult choices about whom to serve. As shown in part III of the worksheet, our hypothetical state would need approximately $2.9 million to serve all legal immigrants at full benefit levels. If the state only appropriates $1.5 million for the fiscal year, it must either provide benefits at a reduced rate or deny assistance to some of the affected legal immigrant population.

Subtracting federal administrative costs. This is the reverse of the process described above. If a state only has a limited amount of money, it must first determine how much will be available for benefits by subtracting out the federal administrative costs ($2,800 per $ 1 million of food stamps). The calculations are shown in line 10 of the sample worksheet.

Distributing the benefits. The next section of this guide describes three possible choices that a state could make given limited resources. However, a state's choices are not limited to the three options outlined in this paper, and a state could combine two or more approaches. All of the options described are designed for states serving legal immigrants made ineligible for federally-funded food stamps, regardless of their date of entry into the country.

Option A: Reduced Benefits for All

Option B: Targeting Households with Children

Option C: Targeting the Elderly, Disabled, and Children

 

Summary

Estimating the number of legal immigrants to be served under a state-funded food stamp program, and the cost of doing so, is not an exact science. Yet the estimates outlined in this guide do provide the best information available for states thinking about the likely costs and numbers served under various programs. States which have not taken into account the impact of the exemptions, caseload decline, and increased rate of naturalizations on the number of legal immigrants denied federal food stamps may be surprised to find that the costs of running a state-funded food stamp program are much lower than anticipated.

Many states may follow the course that Nebraska and Washington have charted and offer full replacement for all legal immigrants subject to the new restrictions on eligibility for federally-funded food stamps. Some states, however, may be forced to make hard choices about what benefits to provide in light of budgetary limitations. Given the difficulties in identifying and targeting the most needy of the legal immigrant population subject to the restrictions on food stamp eligibility, it may make the most sense for states to take the option Massachusetts selected, that of serving all legal immigrants, but at a reduced benefit rate. After four to six months, the state can determine whether the participation rate assumptions used initially were accurate; if the spending is less than anticipated, the states could increase the benefit level provided (or vice-versa). Providing some benefits to all eligible legal immigrants denied federally-funded food stamps ensures that the elderly, disabled, children, as well as other extremely poor legal immigrants who might otherwise suffer severe hardship receive some supplemental food assistance. 

TABLE 1: NUMBERS OF NON-EXEMPT LEGAL IMMIGRANTS  
LOSING FOOD STAMPS, 1998
  Column A Column B
  Non-exempt Legal Immigrants Subject to FS Cuts (monthly estimates) Adjusted for Caseload Decline, 40 Quarters Exemption, and Naturalizations
STATES  
Alabama 551 331
Alaska 1,136 640
Arizona 27,812 15,683
Arkansas 1,088 613
California 494,162 278,659
Colorado 10,445 5,890
Connecticut 9,890 5,577
Delaware 967 546
Dist. of Col. 1,268 715
Florida 152,334 85,902
Georgia 8,991 5,070
Hawaii 7,324 4,130
Idaho 1,945 1,097
Illinois 43,284 24,408
Indiana 2,568 1,448
Iowa 1,989 1,121
Kansas 2,691 1,518
Kentucky 2,132 1,202
Louisiana 6,538 3,687
Maine 718 405
Maryland 14,466 8,157
Massachusetts 28,986 16,345
Michigan 15,187 8,564
Minnesota 17,845 10,063
Mississippi 678 382
Missouri 6,099 3,439
Montana 240 135
Nebraska 1,661 937
Nevada 3,997 2,254
New Hampshire 600 338
New Jersey 34,194 19,282
New Mexico 13,872 7,822
New York 262,825 148,207
North Carolina 5,907 3,331
North Dakota 685 371
Ohio 13,638 7,690
Oklahoma 2,493 1,406
Oregon 12,407 6,996
Pennsylvania 20,688 11,666
Rhode Island 5,774 3,256
South Carolina 454 256
South Dakota 652 368
Tennessee 3,176 1,791
Texas 236,149 133,165
Utah 2,312 1,304
Vemont 435 246
Virginia 10,384 5,856
Washington 36,047 20,327
West Virginia 1,064 600
Wisconsin 11,241 6,339
Wyoming 245 138
Total 1,542,208 869,654
Sources: 1996 Food Stamp QC file. CBO 1997 baseline assumptions. USDA Program Information Data, 1996 -1997. 
Assumptions: 17 percent of non-exempt legal immigrants will naturalize by 1998; a total of 25 percent of legal immigrants in the baseline meet the 40-quarter exemption ; the number of legal immigrants on food stamps falls 14 percent between FFY 1996 and FFY 1998 ; and about two-thirds of refugees, asylees, and those granted withholding of deportation remain eligible for food stamps because they have been in the country for five years or less, have accumulated 40 quarters of work, or have naturalized. All of these figures are based on CBO baseline assumptions and USDA Program Information data. 

     

TABLE 2: ESTIMATES OF NON-EXEMPT 
LEGAL IMMIGRANTS 
LOSING FOOD STAMPS, 1998
  Non-exempt Legal Immigrants Subject to FS Cuts (monthly estimates) Adjusted for Caseload Decline, 40 Quarters Exemption, and Naturalizations (A) 
Percent Who Are Children
(B) 
Percent Who Are Adults That Live With Children
(C) 
Percent Who Live in Families With Children
(D) 
Percent Who Are Elderly
(E) 
Percent Who Are Disabled (incl. those over age 59)
(F) 
Percent Who Are Disabled & Under Age 60
United States (except CA) 1,048,046 590,995 17.4% 47.8% 65.1% 20.1% 6.7% 4.6%
California 494,162 278,659 20.9% 65.1% 85.9% 4.3% 0.7% 0.7%
Other Selected States:
Arizona 27,812 15,683 17.9% 58.1% 76.0% 7.6% 4.3% 2.4%
Connecticut 9,890 5,577 24.3% 37.7% 62.0% 22.4% 7.5% 5.4%
Florida 152,334 85,902 15.4% 39.2% 54.6% 24.0% 8.6% 5.5%
Hawaii 7,324 4,130 9.5% 23.7% 33.2% 54.0% 4.0% 1.6%
Illinois 43,284 24,408 20.5% 36.1% 56.6% 29.9% 17.0% 12.2%
Massachusetts 28,986 16,345 20.3% 52.4% 72.7% 13.0% 10.0% 6.9%
New Jersey 34,194 19,282 26.8% 43.5% 70.3% 20.0% 3.7% 3.1%
New Mexico 13,872 7,822 9.5% 65.6% 75.1% 8.3% 4.6% 4.1%
New York 262,825 148,207 17.7% 38.8% 56.5% 27.1% 7.0% 3.7%
Rhode Island 5,774 3,256 18.3% 45.4% 63.7% 20.1% 10.0% 6.1%
Texas 236,149 133,165 14.7% 65.7% 80.3% 8.8% 3.9% 3.5%
Washington 36,047 20,327 20.1% 49.3% 69.4% 21.8% 8.9% 3.5%
Wisconsin 11,241 6,339 33.8% 52.8% 86.6% 8.5% 2.9% 2.9%
Sources: 1996 Food Stamp QC file. CBO 1997 baseline assumptions. USDA Program Information Data, 1996-1997

     

SAMPLE WORKSHEET 1:
CALCULATING NUMBERS AND COSTS OF LEGAL IMMIGRANTS 
SERVED BY THE STATE-FUNDED FOOD STAMP PROGRAM  
IN A HYPOTHETICAL STATE
  Calculations Hypothetical State
I. Numbers to Be Served
1. Number of non-exempt legal immigrants, 1996 
(from 1996 food stamp QC data base or state-specific number)
  10,000
2. Adjustment for caseload decline, 1996-1998 
(14 percent reduciton from number of non-exempt legal immigrants, line 1, or state-specific number)
- (0.14 x 10,000) 
-1,400
-1,400 
8,600
3. Adjustment for 40 quarters exemptions  
(20 percent reduction from line 2, or state-specific number)
- (0.20 x 8,600) 
-1,720
-1,720 
6,880
4. Adjustment for naturalizations 
(17 percent reduction from line 3 or state-specific number)
- (0.17 x 6,880) 
-1,170
-1,170 
5,710
5. Adjustment for participation rate  
(20 percent reduction from line 4 or state-specific number)
-(0.20 x 5,710) 
-1,142
-1,142 
4,568
II. Costs at Full Benefit Level
6. Monthly cost of serving immigrants at full benefit level 
(average benefit of $71.37, or state-specific number, multiplied by number of eligible legal immigrants, line 5)
$71.37 x 4,568 
$326,018
$326,018
7. Cost, FY 1998 
(monthly cost, line 6, multiplied by number of months of program operation)
9 x $326,018 
$2,934,163
$2,934,163
8. Federal administrative cost 
(add $2,800 per $1 million of total expenditure, line 7)
+ (2,800 x $2.93) 
+ 8,204
+8,204 
$2,942,367
III. Dealing with Budget Constraints: Some Options
9. Fixed amount available 
(amount appropriated by the state)
  $1,500,000
10. Federal administrative costs 
(subract $2,800 per $1 million of expenditures, line 9)
- (2,800 x 1.5) 
-4,200
-4,200 
$1,495,800
11. Option A: Reduced Benefits for All
a.) Average annual benefit per recipient, FY 1998 
(amount available, line 10, divided by number of recipients, line 5)
$1,495,800 / 4,568 
$327.45
$327.45
b.) Average monthly benefit per recipient  
(average annual benefit, line 11a, divided by number of months of program operation)
$327.45 / 9 
$36.38
$36.38
c.) Percent reduction per recipient 
(percent difference between average amount provided, line 11b, and average benefit at full-funded levels)
($36.38 / $71.31) - 1 
- 0 .49
-49.0%
12. Option B: Targeting Immigrant Households with Children
a.) Number of recipient served 
(65.1 percent live with children, from Table 2, multiplied by total number of eligible legal immigrants, line 5; use the state-specific estimate of the percent living with children from Table 2, Column C, if available)
0.651 x 4,568 
2,974
2,974
b.) Average fiscal year benefit per recipient 
(amount available, line 10, divided by number of recipients, line 12a)
$1,495,800 / 2,974 
$502.96
$502.96
c.) Average monthly benefit perprecipient 
(average fiscal year benefit per recipient, line 12b, divided by number of months of program operation)
$502.96 / 9 
$55.88
$55.88
d.) Percent reduction per recipient* 
(percent difference between average amount provided, line 12c, and average benefit at full-funded levels) 
* if excess funds, go to step 13d
($55.88 / $71.37) - 1 
- 0.217
-21.7%
13.) Option C: Targeting the Elderly, Disabled and Legal Immigrant Children
a.) Number of recipients served: 
(42.1% who are elderly, disabled, or children, sum of columns a,d, &f in Table 2, multiplied by the total number of eligible immigrants, line 5; use state-specific numbers from Table 2, if available)
0.421 x 4,568 
1,923
1,923
b.) Average total benefits per recipient: 
(amount available, line 10, divided by number of recipients, line 13a)
$1,495,800 / 1,923 
$777.85
$777.85
c.) Average monthly benefit per recipient: average fiscal year benefit per recipient, line 13b, divided by number of months of program operation) $777.85 / 9 
$86.43
$86.43
d.) excess funds
(i) amount available - (1,923 x $71.37) 
$137,245
$1,495,800
(ii) minus amount spent:  
(number served, line 13a, multiplied by average monthly benefit level, $71.37, multiplied by number of months of program operation)
- ($137,245 x 9) 
$1,235,205
- $1,235, 205
(iii) excess funds available   $260,595
e.) Redo steps 13a-c, targeting an additional category of legal immigrants, with amount from 13d ($263,169) as the total amount available to insert in 13b.

     

SAMPLE WORKSHEET 2:
CALCULATING NUMBERS AND COSTS OF LEGAL IMMIGRANTS 
SERVED BY THE STATE-FUNDED FOOD STAMP PROGRAM 
  Calculations Your State
I. Numbers to Be Served
1. Number of non-exempt legal immigrants, 1996 
(from 1996 food stamp QC data base or state-specific number)
   
2. Adjustment for caseload decline, 1996-1998 
(14 percent reduciton from number of non-exempt legal immigrants, line 1, or state-specific number)
   
3. Adjustment for 40 quarters exemptions  
(20 percent reduction from line 2, or state-specific number)
   
4. Adjustment for naturalizations 
(17 percent reduction from line 3 or state-specific number)
   
5. Adjustment for participation rate  
(20 percent reduction from line 4 or state-specific number)
   
II. Costs at Full Benefit Level
6. Monthly cost of serving immigrants at full benefit level 
(average benefit of $71.37, or state-specific number, multiplied by number of eligible legal immigrants, line 5)
   
7. Cost, FY 1998 
(monthly cost, line 6, multiplied by number of months of program operation)
   
8. Federal administrative cost 
(add $2,800 per $1 million of total expenditure, line 7)
   
III. Dealing with Budget Constraints: Some Options
9. Fixed amount available 
(amount appropriated by the state)
   
10. Federal administrative costs 
(subract $2,800 per $1 million of expenditures, line 9)
   
11. Option A: Reduced Benefits for All
a.) Average annual benefit per recipient, FY 1998 
(amount available, line 10, divided by number of recipients, line 5)
   
b.) Average monthly benefit per recipient  
(average annual benefit, line 11a, divided by number of months of program operation)
   
c.) Percent reduction per recipient 
(percent difference between average amount provided, line 11b, and average benefit at full-funded levels)
   
12. Option B: Targeting Immigrant Households with Children
a.) Number of recipient served 
(65.1 percent live with children, from Table 2, multiplied by total number of eligible legal immigrants, line 5; use the state-specific estimate of the percent living with children from Table 2, Column C, if available)
   
b.) Average fiscal year benefit per recipient 
(amount available, line 10, divided by number of recipients, line 12a)
   
c.) Average monthly benefit perprecipient 
(average fiscal year benefit per recipient, line 12b, divided by number of months of program operation)
   
d.) Percent reduction per recipient* 
(percent difference between average amount provided, line 12c, and average benefit at full-funded levels) 
* if excess funds, go to step 13d
   
13.) Option C: Targeting the Elderly, Disabled and Legal Immigrant Children
a.) Number of recipients served: 
(42.1% who are elderly, disabled, or children, sum of columns a,d, &f in Table 2, multiplied by the total number of eligible immigrants, line 5; use state-specific numbers from Table 2, if available)
   
b.) Average total benefits per recipient: 
(amount available, line 10, divided by number of recipients, line 13a)
   
c.) Average monthly benefit per recipient: average fiscal year benefit per recipient, line 13b, divided by number of months of program operation)    
d.) excess funds
(i) amount available    
(ii) minus amount spent:   (number served, line 13a, multiplied by average monthly benefit level, $71.37, multiplied by number of months of program operation)    
(iii) excess funds available    
e.) Redo steps 13a-c, targeting an additional category of legal immigrants, with amount from 13d ($263,169) as the total amount available to insert in 13b.

   


End Notes

1. See also Dean, Stacy and Carmody, Kelly, "States Now Have the Option to Provide Food Stamps to Legal Immigrants" (Center on Budget and Policy Priorities: October 18, 1997).

2. Earlier estimates provided by the Center on Budget and Policy Priorities were based on 1995 data. These estimates have now been updated based on newly available 1996 data. As a result, some of the underlying assumptions have changed to reflect the more recent data. States that have already calculated the cost of running a state-funded food stamp program and the number of legal immigrants to be served may wish to update their estimates based on the more up-to-date data.

3. CBO assumes that about 300,000 refugees, asylees, and those granted withholding of deportation eligible for food stamps have been in the country for less than five years, and are thus exempted from the welfare law's restrictions on legal immigrants' eligibility for food stamps. This amounts to about 55 percent of those labeled as refugees, asylees, and those granted withholding in the food stamp data base. Since the Cuban-Haitian entrants and Amerasians were added to the five-year exempt category by the Balanced Budget Act of 1997 and are not identified separately in the QC database, they were not included in these estimates. States with large numbers of either or both groups will need to make the necessary revisions.

4. This projection makes two adjustments to take into account the impact of the eligibility restrictions enacted as part of the 1996 welfare law on the food stamp caseloads. First, in March 1997 the three month time limit for 18- to 49-year olds went into effect. According to CBO estimates, approximately 500,000 individuals were denied food stamps by April due to these time limits. Second, after August 22, 1996, new legal immigrants could no longer join the rolls, and, in April 1997, non-exempt legal immigrants began being removed from the food stamp rolls at the time of their recertification. By August of FFY 1997, the Center estimates that approximately one-third of the affected legal immigrants (or 300,000 individuals) had been denied federally-funded food stamps. To determine the trend decline in food stamp caseloads (the portion of the decline that was not due to eligibility changes), 800,000 was added to the number reported as receiving food stamps in August of FFY 1997. The adjusted figure was used to determine the caseload decline through FFY 1998.

5. Although not all recipients of Social Security (such as those who get Social Security due to disability) have earned 40 quarters of work, the majority do.

6. The USDA estimate is similar; 23 percent of immigrants are estimated to be exempt because of their work histories.

7. Alternatively, states may have tabulated their own data on the number of legal immigrants with 40 quarters in the state. State-specific data, however, should be adjusted to reflect the fact that the initial estimate from Table 1, Column A already excludes five percent of legal immigrants under the assumption that they have accumulated 40 quarters. Therefore, if your state data shows that 20 percent of the legal immigrants receiving food stamps in 1996 had accumulated forty quarters, your state should make a 15 percent (20 minus the five percent already assumed exempt) adjustment in line 3 of the worksheet.

8. CBO assumes that 15 percent of legal permanent residents and 23 percent of refugees, asylees, and those granted withholding of deportation eligible for food stamps have been in the country for a minimum of five years, and are thus eligible to naturalize in 1998. Combined, 17 percent of all legal immigrants (permanent residents as well as refugees, asylees, and those granted withholding of deportation) are assumed to naturalize in 1998.

9. Even in those states implementing a state program prior to the final August 1997 cut-off date, immigrants in food stamp households that were recertified after April 1, 1997, will have been removed from the rolls and others will have been barred from applying since August 1996.

10. Stravianos, Michael, Food Stamp Program Participation Rates: January 1994. (Food and Consumer Service, United States Department of Agriculture: March 1997).

11. The average food stamp benefits for legal immigrants is about 98 percent of the average benefit for all recipients "Characteristics of Childless Unemployed Adult and Legal Immigrant Food Stamp Participants: Fiscal Year 1995": 34-35. (Mathematica Policy Research, Inc.: February 1997). According to CBO assumptions, food stamp benefits will rise by 2 percent between 1997 and 1998.

12. The average benefit for legal immigrants at full federal benefit levels may be much lower than $71.37 a month. This is due to the fact that a majority of legal immigrants 64 percent live in mixed households, comprised of both legal immigrants and citizens, ("Characteristics of Childless Unemployed Adult and Legal Immigrant Food Stamp Participants: Fiscal Year 1995":37). In a family comprised of two legal immigrant parents with no earnings and two citizen children, the citizen children still remain eligible for the maximum federally-funded food stamp allotment. Assuming that the benefit level for the eligible children is calculated by treating them as a household of two, they will receive the maximum benefit of $230. If the parents became eligible for state-funded food stamps, they will constitute a household of four, eligible for a maximum allotment of $419/month. While the average benefit received by each member of the household is $105/month, the additional benefit provided through the state-funded program is only $95/month for each of the parents. If the parents have earnings that are not deemed to the citizens in the household, the additional benefit provided by the state-funded program would be even less.

13. The numbers used in this guide are estimates, based on the best information available. A state that appropriates close to the amount estimated as necessary to provide full benefits for all recipients may want to initially provide full benefits to all eligible for the state-funded program rather than reduce benefit levels or restrict eligibility. Then, later in the fiscal year, the state could use actual participation rates to evaluate whether the appropriations are sufficient to provide full benefits for the remainder of the year. If the state finds that it will have a budget shortfall, it can reduce benefit levels at that point.

14. Tabulated from the 1996 food stamp quality control data.

15. These figures do not include data for California. Unlike SSI recipients in all other states, those in California are ineligible for food stamps because they receive a state supplementation payment which covers their food assistance. Since a large proportion of the poor who are elderly or disabled receive SSI, only four percent of food stamp recipients in California are elderly or disabled, as compared with 22 percent for the rest of the nation. Since California has such a large immigrant population, including California data with data from the rest of the nation skews the national averages. Therefore, data for the nation as a whole excluding California and data for California are shown separately.

16. For the purposes of this example, we have assumed that the average benefits received by legal immigrants who are children or live with children is equal to the average prorated benefit provided to all legal immigrants. Some states are considering treating the ineligible immigrant as a non-household member rather than as an ineligible household member. States using this method would need to make an appropriate adjustment in their estimate of the average monthly benefit provided to the legal immigrants living in mixed households.

17. The state-specific data is only provided for those states with large immigrant populations subject to the food stamp restrictions. In the other states, the sample of food stamp recipients included in the QC file only includes a very small number of legal immigrants. Without a larger sample size, the data are not reliable.

18. The Food Stamp Act defines the elderly as persons 60 years of age and older.