Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 317
Measuring Poverty: A New Approach 7 Use of the Poverty Measure in Government Assistance Programs The current official U.S. poverty measure has been not only an important statistical indicator; it has also had direct policy uses in government programs that are designed to help low-income families whose resources fall below a standard of need. Many programs have their own need standard for eligibility, but a significant number link their standard to the official poverty thresholds (or a multiple of them). In most cases, the link is actually to the poverty guidelines derived from the thresholds, and, consequently, we use the term guidelines in this chapter.1 Another program use of the poverty measure has been for allocation of federal funds to states and localities. For example, funds for educationally deprived children under the Elementary and Secondary Education Act are allocated to school districts on the basis of their share of children aged 5-17 who live in poor families. Head Start funds are also allocated to states by a formula that takes account of each state's share of children under age 18 in families receiving Aid to Families with Dependent Children (AFDC) and its share of children under age 6 in poor families. The share of poor people is also one factor in the formula for allocating Community Development Block Grant funds to cities and counties. In this chapter we consider the relationship of a poverty measure to eligibility and benefit standards for government means-tested programs that 1 The poverty guidelines are issued annually by the U.S. Department of Health and Human Services (HHS) by smoothing the official poverty thresholds for different-size families. The guidelines are higher than the thresholds for Alaska (by 25%) and Hawaii (by 15%).
OCR for page 318
Measuring Poverty: A New Approach provide assistance to individual families.2 In particular, we consider the implications of the changes we propose in the current poverty measure for program eligibility and benefit determination. To put the issue in perspective, we review in general the types of programs that are designed to help low-income people and consider a few specific examples. Different categories of programs pose somewhat different questions for the potential role of one or another poverty measure. Appendix D provides details on all such federal programs as they existed through 1994. In Chapter 8, we focus on the possible relationship of the proposed poverty measure to benefit levels in the AFDC program, and we also address the relationship of that measure to state AFDC standards of need, which, in many states, exceed actual benefit levels. RECOMMENDATION We argue throughout this report that the proposed poverty measure is a marked improvement over the current measure for use as a statistical indicator, and we recommend its adoption for this purpose. We believe that the proposed measure also deserves serious consideration for use as an income eligibility standard in government assistance programs that currently determine eligibility or benefit amounts by comparing family resources to the poverty guidelines derived from the official thresholds. However, we do not flatly recommend that the proposed measure be adopted in place of the current measure for program use. Rather, we urge program agencies to carefully review the proposed measure to determine whether it is appropriate and whether it may need to be modified in one or more respects to better serve program objectives. In their review, program agencies should consider the implications of the proposed measure in relation to the current measure. They should also keep in mind some important criteria for evaluating any measure of need. In particular, it is critical that the measure provide for consistency between the definition of family resources and the definition of the poverty threshold (or other need standard). This criterion is important for a statistical measure of poverty so that population groups are appropriately classified by poverty status; it is also important for program use so that program benefits are given to needy families. As we have noted above, the current poverty measure fails this consistency criterion in several important respects, for example, by not excluding 2 There are other questions about the role of a poverty measure and about the changes we propose to the current measure for fund allocation purposes that we do not address. Thus, our recommendation to adjust the official poverty thresholds for geographic area differences in the cost of housing has obvious implications for the distribution of program funds among jurisdictions. However, broadly speaking, the availability of reliable data for estimating poverty rates for small geographic units may be a more important concern for fund allocation than the properties of a specific poverty measure.
OCR for page 319
Measuring Poverty: A New Approach taxes from family resources even though the poverty thresholds were computed on an after-tax basis. Hence, some working families that pay taxes may be erroneously classified above the poverty line because their resources are defined as gross rather than net income. The proposed poverty measure embodies a definition of family resources as money and near-money disposable income that is consistent with the derivation of the poverty thresholds from expenditure data for such basic needs as food, clothing, and shelter. However, the proposed definition is considerably more demanding of data than the current definition: full implementation would require asking about in-kind benefits and several types of expenses as well as money income. For such assistance programs as food stamps and AFDC, which make a very detailed determination of financial eligibility and benefit amounts, implementing the proposed definition of family resources would not complicate program administration. Indeed, that definition, in concept if not in detail, is quite similar to the definitions already in use in these programs. However, other assistance programs currently have fairly simple application procedures that obtain a crude measure of gross money income and compare it with the relevant poverty guideline to determine program eligibility. For these programs, to implement the proposed resource definition could pose a burden on both applicants and program administrators. We believe there are ways to simplify that definition for programs for which a simple application process is valued and there is a willingness to give up some precision in classifying applicants' eligibility status (see below). With regard to the need standard component of the proposed poverty measure, program agencies should consider whether the cutoff for eligibility should be 100 percent of the guidelines or a multiple, as is now the case in many programs. Obviously, there are budget implications of this choice, particularly for those entitlement programs that use the guidelines and that must provide benefits for all applicants who meet the eligibility criteria. In this regard, it is important for program agencies to be aware of the implication of the proposal to update the poverty thresholds each year for real changes in basic consumption rather than to update them only for price inflation. The thresholds developed under the procedure will probably increase more rapidly than thresholds that are updated for price inflation only, even though they are not likely to increase as fast as a purely relative set of poverty thresholds. There are ways to address the budgetary consequences of using poverty thresholds that are updated in real terms for program purposes. For example, eligibility could be limited to families with resources below a fraction of the thresholds. This strategy is not a contradiction in terms. We have argued strongly that updating the poverty thresholds for real growth in spending on basic necessities makes a great deal of sense for a statistical measure. There is considerable evidence that poverty thresholds are relative to time and place,
OCR for page 320
Measuring Poverty: A New Approach and a regular, automatic adjustment for real growth seems preferable to an adjustment that occurs spasmodically. However, the design of government assistance programs must take into account many factors, only one of which is a statistical standard of need. Other considerations, such as funding constraints and competing uses for scarce tax dollars, may dictate that assistance program benefits be set at a level below the statistical poverty thresholds. RECOMMENDATION 7.1. Agencies responsible for federal assistance programs that use the poverty guidelines derived from the official poverty thresholds (or a multiple) to determine eligibility for benefits and services should consider the use of the panel's proposed measure. In their assessment, agencies should determine whether it may be necessary to modify the measure—for example, through a simpler definition of family resources or by linking eligibility less closely to the poverty thresholds because of possible budgetary constraints—to better serve program objectives. GOVERNMENT ASSISTANCE PROGRAMS Overview In 1994, 70 federal and federal-state programs were providing cash, in-kind benefits, or other types of services to families or individuals who were deemed needy on the basis of an explicit income test.3 Table 7-1 summarizes the number and expenditures of these programs in fiscal 1992 (see Burke, 1993, and Appendix D for details). Of the 70 programs, 27 (39%) have as one of their income eligibility criteria that income be compared with the poverty guidelines or some multiple of them; see Table 7-2. They run the gamut from small programs that spend only a few million dollars a year (e.g., Follow Through and Senior Companions) to two of the largest assistance programs, food stamps and Medicaid. Of these programs, 14 use the poverty guidelines (or a multiple) as the sole criterion of income eligibility; they account for 2 percent of expenditures by all assistance programs. Examples are the Maternal and Child Health Services Block Grant, Legal Services, and Foster Grandparents. The other 13 programs, which account for 56 percent of expenditures by all assistance programs, have several ways of determining income eligibility. For example, School Lunch and School Breakfast accord eligibility to children whose families already participate in AFDC or food stamps, and they also permit other 3 Assistance programs typically have other requirements for eligibility besides a comparison of income with a need standard: for example, they may provide benefits only to people in certain age categories or have a limit on assets in addition to income or have other restrictions or requirements. Our discussion focuses on programs' definitions of and limits on income.
OCR for page 321
Measuring Poverty: A New Approach TABLE 7-1 Government Assistance Programs That Link Eligibility to Income, Fiscal 1992 Programs Expenditures Program Type Number Percent Million $ Percent Programs that link eligibility solely to the federal poverty guidelines 14 20.0 6,510 2.3 Programs that link eligibility to the federal poverty guidelines and also to participation in other programs (e.g., AFDC, SSI, or food stamps) 13 18.6 156,580 56.1 Programs that link eligibility to a percentage of the local area (or state) median income 12 17.1 21,302 7.6 Programs that have their own income eligibility standards (or that link eligibility to participation in another program) 31 44.3 94,583 33.9 Total 70 100.0 278,975 100.0 SOURCE: Derived from Burke (1993). NOTES: Not included in the table are two assistance programs that are wholly supported by state and local funds: General Assistance (fiscal 1992 expenditures of $3,340 million) and General Assistance—medical care component (fiscal 1992 expenditures of $4,850 million). Also not included are eight programs that allocate benefits on some other basis (e.g., area of residence): Indian Health Services, Nutrition Program for the Elderly, State Legalization Impact Assistance Grants, Chapter 1 Migrant Education Program, Emergency Food and Shelter Program, Food Distribution Program on Indian Reservations, Migrant High School Equivalency Program, and College Assistance Migrant Program, which had total fiscal 1992 expenditures of $2,626 million. For details of the programs in each category, see Appendix D. SSI, supplemental Security Income. children to qualify on the basis of comparing their family income to a multiple of the poverty guidelines. Programs authorized by the Job Training Partnership Act (e.g., Job Corps and Summer Youth Employment) accord eligibility to people already participating in AFDC or food stamps and permit other people to qualify on the basis of comparing their family income to 100 percent of the poverty guidelines or 70 percent of the lower living standard income level determined by the Department of Labor, whichever amount is higher. The remaining 43 programs (61%) use some other income eligibility
OCR for page 322
Measuring Poverty: A New Approach TABLE 7-2 Government Assistance Programs That Link Eligibility or Benefits to the Current Poverty Measure, by Program Type and Poverty Cutoff for Eligibility, Fiscal 1992 Programs That Provide All-or-Nothing Service Poverty Cutoff for Eligibility (%) Commodity Supplemental Food Program 100 (for elderly people) Community Services Block Grant 100; 125 at state option Follow Through 100 Foster Grandparents 125 Head Starta 100 Job Corpsa 100 Legal Services 125 (up to 187.5 for people with excessive medical or child care expenses) Medicaida,b 100 for some people; 133 for others (up to 185 at state discretion for others) Senior Community Service Employment Programa 125 Senior Companions 100 Special Milk Program 130 Special Programs for Students with Disadvantaged Backgrounds (TRIO Programs) 150 Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) 100 to 185 at state discretion Summer Food Service Program for Children 185 (applies to service areas, not applicants) Summer Youth Employment Programa 100 Training for Disadvantaged Adults and Youtha 100 Vocational Education Opportunities, Disadvantaged Activitiesa 100 Weatherization Assistancea 125 Programs That Relate Benefits to Income or Charge for Services on a Sliding Scale Poverty Cutoff for Eligibility (%) Child and Adult Care Food Program 130 for free meals; 185 for reduced price Community Health Centers 100 for free care; sliding scale up to 200 Food Stamp Programa,b 130 (gross income); 100 (net income) Low-Income Home Energy Assistance Program (LIHEAP)a 150 Maternal and Child Health Services Block Grant 100 for free care; sliding scale for others Migrant Health Centers 100 for free care; sliding scale up to 200 School Breakfast Programa,b 130 for free meals; 185 for reduced price School Lunch Programa,b 130 for free meals; 185 for reduced price Title X Family Planning Services 100 for free care; sliding scale up to 250 SOURCE: Burke (1993). a Program also accords eligibility on bases other than the poverty guidelines (e.g., children on AFDC are automatically eligible for Head Start); see Appendix D. b Entitlement program: eligible applicants cannot be denied benefits.
OCR for page 323
Measuring Poverty: A New Approach criterion. Of these programs, 12 of them, which account for 8 percent of total expenditures, determine income eligibility on the basis of comparing household income to a percentage of state or local area median family income. Examples are Section 8 Low-Income Housing Assistance and Rural Housing Loans. Finally, 31 programs, which account for 34 percent of total expenditures, have their own income eligibility standards. Examples are AFDC, Supplemental Security Income (SSI), veterans' pensions, and Stafford Loans. The 27 programs that link eligibility for some or all applicants to the poverty thresholds or guidelines differ on a number of dimensions. These, in turn, have implications for using the proposed poverty measure for eligibility determination. One dimension is how the benefits are related to income. Some programs have a poverty-based income test simply to determine eligibility and do not further condition benefits for eligible people on the amount of their income. In other words, these programs provide an all-or-nothing service (examples are Head Start and Legal Services). Other programs do condition benefits on the amount of an applicant's income. For example, the Food Stamp Program reduces the dollar amount of the coupons provided to recipients in direct relationship to their ''countable" income. Such programs as Maternal and Child Health Services charge recipients for services on a sliding scale: some people pay nothing, others pay a fraction of the costs, and still others pay full costs, depending on broad income-to-poverty guideline categories. A second dimension is the complexity of the method for measuring applicants' incomes. Programs that provide an all-or-nothing benefit often have a fairly simple application form that does not ask applicants for extensive detail about income sources. Many programs that charge recipients for services on a sliding scale are also in this category. In contrast, the Food Stamp Program, which calibrates benefits quite closely to income, includes an elaborate process to determine applicants' gross income and their net income after allowable deductions. Another distinction is between entitlement and nonentitlement programs. Entitlement programs (e.g., Medicaid, food stamps, School Lunch, and School Breakfast) must provide benefits to all eligible applicants. However, many of the programs that link eligibility to the poverty guidelines (e.g., Head Start, Legal Services) are not entitlements. These programs do not guarantee to provide services to all eligible families; rather, legislatively set budget limits determine how many eligible people who apply for services will actually be assisted and how many will be put on a waiting list. Finally, programs vary in whether they use 100 percent or a multiple of the poverty guidelines as the basis for determining eligibility (see Table 7-2). For example, Head Start has an income cutoff of 100 percent of the poverty guidelines, but Legal Services has a cutoff of 125 percent, and Special Programs for Students with Disadvantaged Backgrounds has a cutoff of 150 per
OCR for page 324
Measuring Poverty: A New Approach cent. School Lunch and School Breakfast provide free meals to children of families with incomes below 130 percent of the poverty guidelines and charge a reduced price for families with incomes between 130 and 185 percent of the guidelines. Community and Migrant Health Centers provide free medical care to people with incomes below 100 percent of the poverty guidelines and charge reduced fees on a sliding scale to people with incomes between 100 and 200 percent of the guidelines. The Title X Family Planning Services Program operates in a similar manner except that the cutoff for reduced fees is 250 rather than 200 percent of the poverty guidelines. States have discretion to set income eligibility limits for the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) at the level used by state or local agencies for free health care, so long as the level is between 100 and 185 percent of the poverty guidelines. Determining Income Eligibility: Selected Programs To determine how families' incomes are estimated for comparison with the poverty guidelines, we examined application procedures for selected programs.4 In many cases—for example, for Community Health Centers and Title X Family Planning Services5—local centers or agencies have a good deal of discretion in how they determine income eligibility. In other cases, such as food stamps, federal regulations are very specific about the definitions and procedures used. As examples of current procedures and definitions, we summarize the income determination process for Head Start, school nutrition programs, WIC, and food stamps. From our analysis, we conclude that the proposed poverty measure is advantageous for program use in many respects, although it may need modification in some instances. Head Start Local Head Start agencies have discretion in determining income eligibility, although they must have on file documentation for participating families that certifies that they met the income eligibility criteria. Families participating in AFDC are automatically eligible for Head Start, and no additional verification or documentation of their income is required. AFDC families make up about one-half of Head Start participants; the remainder are largely working poor families. Head Start agencies typically ask to see paystubs for documentation of earnings. The income definition used is the same as for the current poverty measure, namely, gross money income.6 4 For more complete program descriptions, see Appendix D. 5 Information provided by Malvina Ford, Congressional Research Service. 6 Information provided by Craig Turner, Head Start Bureau, Administration for Children and Families, U.S. Department of Health and Human Services.
OCR for page 325
Measuring Poverty: A New Approach School Nutrition Programs For school nutrition programs administered by the U.S. Department of Agriculture (USDA)—School Lunch, School Breakfast, Special Milk Program—federal regulations are fairly specific, although states may institute additional policies that do not conflict with the federal requirements. Generally, schools are required to inform households of the availability of free or reduced-price school meals to those who meet eligibility requirements. Households already participating in food stamps or AFDC can be certified automatically through contact with the local food stamp and AFDC offices; other households must provide information on their previous month's income.7 A USDA manual (Food and Nutrition Service, 1991) specifies the types of income to be included and excluded. Federal law excludes various benefits from the calculation of income, such as food stamps and educational assistance received under means-tested programs (e.g., Pell Grants); and negative self-employment income is set to zero; otherwise, the definition of income is much the same as the gross money income definition used in the March Current Population Survey for the official poverty statistics. The specific information requested from households is in the form of a grid, with each household member listed down the side and the following sources of income listed across the top: gross monthly earnings (before deductions) for the first and second job; combined monthly payments from welfare, child support, alimony; combined monthly payments from pensions, retirement, Social Security; other monthly income. From the information provided, the school computes total income and compares it with a multiple of the appropriate poverty guideline (130% for free meals, 185% for reduced-price meals). Finally, the school is responsible for conducting annually a verification of income for a sample of participating households. WIC State agencies that operate the WIC program may adopt the income eligibility criteria for reduced-price school meals (i.e., 185% of the poverty guidelines), and, if they do so, they must follow the definition of income used by the school nutrition programs. Alternatively, state WIC agencies may adopt the income eligibility criteria used by state or local agencies for free or reduced-price health care, so long as the income limit is not less than 100 percent and not more than 185 percent of the poverty guidelines. Under this alternative, state WIC agencies may use the income definition of the state or local health care agencies. However, the value of in-kind housing or other in-kind benefits must not be counted as income; likewise, the value of various pay 7 Households in special circumstances (e.g., those that have money from seasonal work) may project their anticipated annual income rather than reporting previous month's income.
OCR for page 326
Measuring Poverty: A New Approach ments or benefits provided under certain federal programs as specified by law (e.g., Pell Grants) must be excluded. The intent of the option of tying income eligibility to the limits used by state or local health care agencies is to encourage coordination of WIC with health services and to simplify the administrative burden of determining eligibility (Food and Nutrition Service, 1988a). Food Stamps Households that receive AFDC or SSI, and so have already been through an eligibility determination process, are generally automatically eligible for food stamps. Other households can receive food stamps if they meet certain income and asset requirements. Because the program has a short (monthly) accounting period, it applies an asset test that is designed to screen out applicants who have savings and other liquid assets on which they can draw to cover a temporary period of low-income. The program also applies a gross and net income test that is similar in many respects to the proposed calculation of gross and disposable income for purposes of measuring poverty. Gross income for the Food Stamp Program includes all kinds of money income, with a few exceptions (e.g., the Earned Income Tax Credit [EITC] is not counted). Net income for households without an elderly or disabled member is gross income minus: a standard deduction that does not vary by household size and is adjusted for inflation each October ($131 a month in fiscal 1994); 20 percent of any earned income (to allow for taxes and work expenses); out-of-pocket dependent care expenses, when necessary for work or training, up to $200 per month for each dependent under age 2 and up to $175 for other dependents; and shelter expenses that exceed 50 percent of counted income after all other deductions up to a legislatively set ceiling ($231 a month as of July 1994). Net income for households with an elderly or disabled member is gross income minus: the standard, earned income, and dependent care deductions noted above; shelter expenses that exceed 50 percent of counted income after all other deductions, with no ceiling; and out-of-pocket medical care expenditures for the elderly or disabled member that exceed $35 a month.8 Gross and net income are compared with the current Department of Health and Human Services (HHS) poverty guideline appropriate for the family's size to determine eligibility. Households without elderly or disabled members must have gross monthly income below 130 percent of the HHS poverty guidelines and net monthly income below 100 percent of the poverty guidelines. Households with an elderly or disabled member need only meet the net income test. To determine benefits, a different cutoff is used because the Food Stamp Program is intended to supplement families' resources for food consumption 8 Different standard deductions and shelter expenses ceilings apply in Alaska and Hawaii.
OCR for page 327
Measuring Poverty: A New Approach only. Hence, the cutoff is the USDA Thrifty Food Plan (instead of the poverty guidelines), and the amount of food stamps that eligible applicants receive is the difference between 30 percent of their countable income and the Thrifty Food Plan value for their size family. As an example, if the Thrifty Food Plan value for a family is $400 a month and the family has $900 of countable income, the family will receive $100 in food stamps—$400 minus $300 (30% of $900). In effect, the Food Stamp Program expects that households will spend 30 percent of their net countable income on food, or roughly the amount that food represents of the official poverty thresholds (as originally developed); the program supplements families' food-consumption resources up to the level of the Thrifty Food Plan. USING THE PROPOSED POVERTY MEASURE In assessing whether and how to use the proposed poverty measure for determining income eligibility for benefits or services, program agencies must consider a number of issues. These issues relate to the thresholds, the family resource definition, and other aspects of the measure. The Thresholds We have recommended a method for deriving the poverty thresholds each year but not a specific threshold for the reference family of four with which to initate a new series of poverty statistics. If the proposed measure is adopted for statistical purposes and a specific initial threshold is designated by the U.S. Office of Management and Budget, agencies will need to assess the consequences for program costs and caseloads of any difference between a new threshold and the current poverty guideline for a four-person family. If the new threshold is higher, its use as an eligibility standard for programs will likely produce a larger pool of potential applicants. For nonentitlement programs (i.e., programs that do not guarantee services or benefits to all eligible applicants), there are no budgetary consequences from an increase in the applicant pool. However, should the newly eligible people apply for benefits or services, such programs would have to lengthen their waiting lists unless budget ceilings are raised. For entitlement programs that use the guidelines, there would be a direct effect on caseloads and costs if the applicant pool increases and the newly eligible people apply for assistance. Even if the new threshold is the same as the current threshold, changes in the family resource definition could still increase the pool of potential applicants. This could happen for programs that automatically accord eligibility to families receiving welfare benefits (e.g., AFDC, SSI, or food stamps) and also allow other families to qualify on the basis of comparing their income with the poverty guidelines. Because of such changes to the family resource definition
OCR for page 328
Measuring Poverty: A New Approach as deducting taxes and work-related expenses from income, the applicant pool for these programs could include a higher number of families not now receiving public assistance (plus the same number of families who are receiving public assistance). Given a particular initial threshold, there could be an effect on the distribution of the applicant pool by family size due to the differences between the proposed equivalence scale and the scale implicit in the current guidelines. We have argued that the proposed scale is an improvement over the current scale, so this effect would be appropriate in terms of targeting services to those types of families most in need.9 There might also be an effect on the size of the applicant pool in different areas of the country because of the recommended adjustment to the thresholds for geographic differences in housing costs. Depending on its magnitude, this effect could be temporarily disruptive to programs in various areas that were accustomed to higher or lower caseloads, but it should represent an improved overall targeting of services. The use of poverty thresholds that are adjusted for geographic differences in the cost of housing raises some special issues for the Food Stamp Program. The use of such thresholds for eligibility determination should, as just noted, represent an improved targeting of program benefits.10 For benefit determination, however, the assumption that households spend 30 percent of their income on food would need to be reexamined—otherwise, newly eligible households in more expensive areas would not, in fact, benefit from the program. For example, if the maximum benefit for a particular size household were $300 per month and the eligibility level for that size household were raised from $1,000 to $1,200 because of higher housing costs in the area, then a household with $1,100 of net countable income would be newly eligible but would receive no food stamp benefits (30% of its countable income would be $330—above the maximum benefit). For such a household to benefit, the assumed percentage of countable income available for food expenditures would need to be lowered. Alternatively, the maximum benefit could be raised (as is currently done for Alaska and Hawaii), if it is assumed that food as well as housing costs are higher in the area. A major issue with the use of the proposed method for determining 9 The proposed scale is an improvement over the scale implicit in the current poverty thresholds, which has many irregularities. It is also an improvement over the scale implicit in the poverty guidelines: that scale is smooth, but it assumes that children need as much as adults, and it also assumes that each family member beyond the first costs the same (i.e., that economies of scale do not increase for larger families; see Chapter 3). 10 It would have to be decided whether to adopt state-specific thresholds—to reflect the state involvement in administering the program—or the recommended breakdown by geographic division and size of metropolitan area; see Chapter 8.
OCR for page 329
Measuring Poverty: A New Approach poverty thresholds is that the method will generate new thresholds each year that reflect real changes in expenditures on food, clothing, and shelter. (We propose the use of 3-year moving averages to derive each year's thresholds, which will guard against big changes from one year to the next—see Chapter 2.) In years in which there is an economic downturn, the thresholds may decrease in real terms. In most years, however, given economic growth, they are likely to increase in real terms—that is, to increase more than the rate of inflation. Thresholds that rise in real terms will not necessarily result in a larger number or proportion of poor people compared with thresholds that are simply adjusted for price changes (like the official thresholds). (Similarly, thresholds that fall in real terms will not necessarily result in a smaller number or proportion of poor people compared with price-adjusted thresholds.) The outcome depends on a combination of factors, such as changes in government tax policies, that affect the distribution of income in the vicinity of the thresholds. However, it is likely that the use of thresholds developed by the proposed procedure will produce a larger pool of potential program applicants, which, could, in turn, produce higher program costs and caseloads (or longer waiting lists) compared with continued use of the poverty guidelines derived from the official thresholds. Program agencies must consider their response to this likely consequence. One option would be to periodically reconsider the multiple of the thresholds that a program uses as the cutoff for eligibility (or the cutoff for partial payment by the applicant in the case of programs that charge on a sliding scale). For example, the School Lunch Program might, at some future date, decide, on cost grounds, that it would lower eligibility for free lunches from 130 percent to 100 percent of the poverty level. Another option would be to use the proposed equivalence scale and geographic adjustments for housing costs but continue to update the initial threshold simply for price changes. This option is less attractive because it implies the continuance of two different poverty measures. It seems preferable to have one official measure and require decision makers to consider in a forthright manner the issues involved in determining the multiple—or fraction—of the official thresholds to use for program eligibility. In debating what multiple to choose, decision makers will necessarily have to acknowledge possibly competing goals, such as the desire to help people whose resources fall below a reasonable standard of need and the desire to contain program spending within specified limits.11 11 Chapter 8 discusses a range of factors that affect decisions about program eligibility standards and benefit levels, with specific reference to AFDC.
OCR for page 330
Measuring Poverty: A New Approach The Family Resource Definition Proper implementation of the proposed poverty measure requires not only using the revised thresholds, but also changing the definition of income to compare with those thresholds. As we have stressed throughout, a poverty measure is a package in which the two components—the budget or threshold concept and the definition of family resources—must be consistent. Although the initial poverty threshold for the proposed measure might well be set at a level close to the current threshold, it represents a different concept, namely, a basic budget for food, clothing, shelter, and a little more for other necessities. This budget explicitly excludes some kinds of expenses—such as taxes, work-related expenses, child support, and out-of-pocket medical care expenses—which are instead treated as deductions from income. The proposed definition of disposable income also includes the value of in-kind benefits. This change in definition has somewhat different implications for programs that currently have a fairly simple process for determining gross regular money income and programs that already collect extensive information with which to determine gross and net income. Simplified Determination of Disposable Income For programs that currently obtain a crude measure of gross money income, full implementation of the proposed disposable income definition would require collecting additional information from applicants about income and expenses. Hence, there could be increased administrative costs and an increased burden on applicants. We are certainly not in a position to provide detailed guidance to federal and state program agencies to determine how best they might implement the proposed disposable income definition. However, we have some ideas for ways to do so that could reduce the added burden on program agencies and applicants. It is important to note that the approaches we suggest, while minimizing burden, may increase the chance of an error in classifying an applicant's eligibility status in comparison with an approach that asks very detailed questions about applicants' income and expenses. (The assumption, based on survey research results, is that asking more detailed questions will elicit more complete responses; see Appendix B.) However, programs that at present obtain a fairly crude and hence less burdensome measure of gross money income probably already experience some classification errors. A simplified determination of disposable income might work as follows, by taking the School Lunch and School Breakfast Programs as examples. These programs currently provide automatic eligibility to AFDC and food stamp families and presumably would continue doing so. For other families, the program asks about monthly income by several broad categories, including
OCR for page 331
Measuring Poverty: A New Approach earnings for up to two jobs. A possibility for obtaining after-tax income would be to ask for net pay after deductions for Social Security and payroll taxes. A drawback is that such monthly pay information probably would not reflect the EITC. Another alternative would be for the Food and Nutrition Service, with guidance from the Census Bureau, to provide schools with a simple formula for calculating payroll and net income taxes from information on gross earnings and family composition. The specifications could indicate an income level above which it would not be necessary to estimate taxes; in other words, there should be no need to go through the calculation for families clearly above the thresholds. For child care costs and child support payments, it seems fairly straightforward to ask families if they pay for child care or child support and their typical monthly costs. The flat deduction for commuting and other work-related expenses would not require asking families for any added information. With regard to in-kind benefits, it would not be necessary to ask about food stamp income, because food stamp families are automatically eligible.12 Families could be asked if they receive housing assistance, although the value of such assistance is difficult to determine, and it might be wise, for administrative ease, to ignore this source of income.13 Finally, rather than asking families about last month's out-of-pocket medical costs, which might not be representative of their annual costs, it might be easier simply to ask whether they have public or private health insurance. The Food and Nutrition Service, with guidance from the Census Bureau, could provide schools with a formula for assigning average out-of-pocket expenses to applicants on the basis of their family composition (including ages of family members) and insurance coverage. The process just described for determining disposable income would be more involved than the current process for determining gross money income. However, we think that a "cookbook" (which might be computerized) could be developed for state and local agencies that would provide a reasonably straightforward way to calculate disposable income with acceptable accuracy with only a few added questions being asked of applicants. An alternative approach would be to develop a "menu" of poverty thresholds for different types of families—such as working families with and without child care expenses and with and without health insurance coverage—that are appropriate to compare with a gross money income definition of family resources. For example, the threshold for a working family of two adults and 12 However, programs that rely solely on comparing income with the poverty guidelines to determine eligibility and do not accord automatic eligibility to welfare families would need to ask about food stamps and, perhaps, other sources of in-kind income. 13 In fact, many public housing recipients are also receiving food stamps or AFDC and hence would not need to be queried about income. Data from the 1991 American Housing Survey showed that 54 percent of renters receiving housing assistance also received food stamps (Nelson and Redburn, 1994: Table 1).
OCR for page 332
Measuring Poverty: A New Approach two children that pays for child care and has health insurance could be based on the threshold that results from the panel's concept plus average amounts for such families for income and payroll taxes, child care expenses, and out-of-pocket medical care expenses. We do not recommend this approach for the statistical measure of poverty for a number of reasons. It would require the development of a large number of thresholds that, even so, would not likely provide an accurate measure of poverty status for the many families that are not "average." However, we believe the use of this approach has merit to determine eligibility for assistance programs in which the goal is a reasonable estimate that minimizes burden on applicants and program staff. To use Head Start as an example, families not on AFDC might be asked, as now, for gross money income and documentation of earnings. They might also be asked, on a simple yes-no basis, whether they pay child care or child support, whether they have health insurance, and whether they receive food stamps or live in public housing. Using this information, the Head Start agency could compare the family's gross money income to the appropriate threshold for that family's circumstances by consulting a menu of thresholds. The process would be similar to that performed now, except that the menu would contain thresholds that vary by factors (e.g, work status, presence of health insurance coverage, etc.) in addition to family type and geographic area. The Census Bureau could assist program agencies by developing the menu. In sum, we believe that there are reasonable strategies for program agencies that want to use the proposed poverty measure but, at the same time, retain a relatively simple application process. Whatever the strategy adopted to implement the proposed measure (e.g., a "cookbook" or "menu" approach or some other strategy), its use should improve the targeting of services to needy people compared to the current measure. Full Determination of Disposable Income A number of assistance programs already obtain a great deal of information about applicants' resources in order to calculate gross and net income. The definition of net family income that is used in many of these programs is similar in broad outline, if not in specific details, to the proposed definition of family resources for the poverty measure. Hence, such programs as food stamps or AFDC would not find it difficult to use the proposed disposable income definition, although they should still consider the particulars of the definition and their appropriateness for program use. As we have stressed previously, it is important that the concept underlying the eligibility cutoff for a program be consistent with the family resource definitions. For example, if a program's need standard makes no allowance for expenses required to earn income (e.g., taxes, child care, commuting costs),
OCR for page 333
Measuring Poverty: A New Approach then the determination of countable income should subtract any such expenses that are incurred before comparing income with the eligibility cutoff. In this regard, we note that poverty thresholds developed according to the proposed concept would be more appropriate in many ways for eligibility determination in the Food Stamp Program than would the current poverty guidelines. This program currently defines countable income to exclude child care expenses and an allowance for taxes and other work-related expenses, which is consistent with the proposed threshold concept (but not with the current guidelines). In addition, out-of-pocket medical care expenditures above a certain limit are excluded from income for the elderly and disabled.14 In contrast, however, the fact that EITC benefits cannot be counted as income (by law) for purposes of food stamp eligibility introduces an element of inconsistency with the proposed concept. Again, we are not in a position to provide specific guidance for programs. We repeat that the need concept and the definition of countable income in a program should be consistent. Other Issues There are some other features of the proposed poverty measure that may or may not be suitable for program use. For example, the proposal is that need be measured on an annual basis, that asset values not be included in resources, and that the unit for measuring need be the family as defined by the Census Bureau. Program agencies may well have sound reasons for reaching some other decision on these aspects of program design. Thus, some programs are intended to provide short-term assistance and hence use a shorter accounting period than a year: for example, the accounting period in food stamps and AFDC is 1 month. In order to ensure that people applying for benefits have used up their available resources and are genuinely in crisis, programs with short accounting periods typically limit the assets that applicants can have and still be eligible for assistance. With regard to the assistance unit, programs differ in their target populations and hence often differ in their definition of an eligible unit—for example, the Food Stamp Program generally defines eligible units to be the entire household, whereas AFDC generally defines eligible units to be families consisting of dependent children and their parent(s)—a narrower definition of family than that used by the Census Bureau. These differences from the proposed statistical poverty measure are certainly appropriate in light of program objectives. 14 Shelter costs in excess of 50 percent of income (up to a ceiling for households with no elderly or disabled members) are also deducted from income for purposes of food stamp eligibility. This provision benefits people who, whether they live in high-cost or low-cost areas, pay what is deemed an excessive amount for housing relative to their resources.
OCR for page 334
Measuring Poverty: A New Approach This page in the original is blank.
Representative terms from entire chapter: