I want to thank the Chair, Rep. Jim McDermott, and the Committee for considering this important issue. My name is Diana M. Pearce, and I am Senior Lecturer and Director of the Center for Women’s Welfare at the University of Washington. For more than a decade, I have worked on developing and disseminating an alternative measure of poverty, the Self-Sufficiency Standard.

In this brief testimony, I would like to address four topics:  (1) why we need a revised poverty measure, (2) the methodological and conceptual problems in the proposed National Academy of Sciences (NAS) approach, (3) the impact of using the NAS approach versus other alternative measures of poverty on both poverty rates and our understanding of poverty, and (4) recommendations for an alternative poverty measure and the difference an alternative measure can make. 

Please note that the following comments are grounded in my work over the last 12 years on the Self-Sufficiency Standard, which is now found in 37 states plus the District of Columbia and used in a wide array of settings and program applications. My conclusions reflect not just my opinions, but the experience of many who have applied this poverty standard in their work.

1. Why We Need a Revised Poverty Measure

The first and most important reason that a revision of the federal poverty measure (known widely as the Federal Poverty Line, or FPL) is that the FPL is too low. I will not recount the reasons it is now too low, as they have been well-detailed elsewhere, but only remind us that a measure that is too low has three serious consequences. First, because the FPL is too low, many people who do not have enough income to meet their needs are not counted as “poor”. Second, countless assistance programs have turned to using a multiple of the poverty line to determine eligibility. For example, eligibility for the Food Stamp Program is set at 130% of the FPL and eligibility for S-CHIP (State Child Health Insurance Program) is as high as 300% of the FPL. That is, in some states, costs are so high that families with incomes at 300% of the FPL are deemed unable to have enough to meet their children’s health needs. It is an oxymoron to have an eligibility level that is three times the poverty level. Third, because the FPL is too low, we have a skewed vision of poverty affecting our program and policymaking. Too many programs are focused only on getting people into employment, when in fact the great majority of families lacking adequate income according to the Self-Sufficiency Standard (80-85%) already have at least one worker in the workforce.[1] 

There are other problems with the FPL that should be mentioned briefly, as they have guided efforts to revise the poverty line:

• The FPL does not vary by place.
• The FPL does not, and cannot, reflect changing demographics, work patterns, and the emergence of new needs/costs, including child care, health care, transportation and taxes.
• The FPL does not show the impact of taxes, tax credits (such as EITC), or benefits including cash (such as TANF benefits), near-cash (such as Food Stamps), and in-kind benefits (such as child care assistance, Medicaid, or housing assistance.)

2. The Methodological and Conceptual Problems in the Proposed National Academy of Sciences (NAS) Approach

While I commend Representative McDermott and others for recognizing the need to reevaluate the way in which we measure the unmet needs of American households, I believe the NAS approach does not adequately address the critiques and concerns of the FPL described above. The heart of the problem with the NAS approach is the methodology. This is most clearly seen by comparing the methodology of the FPL, the Self-Sufficiency Standard, and the NAS approach (see attached chart). Measuring poverty is a three-step process (see table on page 10). The NAS approach has problems and unintended consequences within each of the steps, which are described in order below.

Step 1: Creating the Poverty Threshold 

While the FPL only specifies food, by implication it includes all other costs that were significant at the time such as housing, transportation, clothing, and miscellaneous. The Self-Sufficiency Standard explicitly includes all the costs implied in the FPL—food, housing (including utilities), transportation, miscellaneous (including clothing)—plus new costs such as health care, child care, and taxes/tax credits. In contrast, the NAS approach only includes food, shelter, clothing, and miscellaneous. Thus, the NAS approach only creates a partial poverty threshold.

Partial thresholds are problematic, and when misunderstood as full thresholds, are too low.  Thresholds quickly have a life of their own, and the “fine print” of the methodology gets lost.  Most people assume that a revised measure is comparable to the original measure, that is, it is a complete threshold, but in fact, the NAS approach is “apples” to the FPL “oranges”. The NAS threshold is only a partial threshold and cannot be used on its own. However, this crucial detail gets lost very quickly.

For example, the New York Times editorial on July 22, 2008 entitled “Poverty’s Real Measure” notes that the FPL for a family of four was only $20,444, and in the next sentence states that “The mayor raised New York’s poverty ceiling to a more believable $26,138” (p. A18). Note that nothing in this statement indicates that this number is a partial threshold, or that it needs to be adjusted for costs such as health care or child care.[2]

A partial threshold is also problematic because poverty thresholds are used in many ways beyond measuring poverty with datasets. Not only are they used for eligibility, but poverty thresholds are also used as benchmarks to measure individual and program achievement, to target resources such as job training and education resources, to set minimum and living wage levels, and for many other purposes. 

Step 2:  Calculating Household Income

In step 2, both the FPL and the Self-Sufficiency Standard simply count gross income from all sources (except refundable tax credits) to determine household income. In the NAS approach, gross income from all sources is also counted, however to determine household net income, certain expenses (actual expenditures only) are deducted from gross income (including health care and work-related expenses, such as child care and transportation).

Reducing household income with deductions is problematic, as it underestimates and constrains some costs, particularly for certain groups.  By including some costs, such as housing in the threshold itself, while assigning other costs to be deducted from household income, the NAS approach “privileges” some costs over others. When an amount is allotted in the threshold to meet a basic need, such as housing, this implies that that cost is essential. Conversely, other costs not given an allotment in the threshold itself, such as health care or child care, are implicitly labeled as non-essential, as optional or extra, for there is not an amount set aside to insure that there is enough to meet this need. However, employment and its associated costs are now the norm, not the exception: 62% of two-parent families have both parents in the workforce, and 72% of single parent families have the parent in the workforce, making work-related expenses such as child care essential, not optional, for the majority of families.[3]

The “deductions” approach hides the very poverty that should be made explicit because families too poor to afford these items will not have these deductions.  If a family has enough income for rent, food, clothing and miscellaneous, but not health care, they will forgo the latter, and thus may not appear to be poor income-wise, but lack health care, a basic need.  For example, a family of four with income of $27,000 could spend nothing on health care; in New York City, their income would be above the NAS-based poverty threshold of $26,138. However, an equivalent family, who had an income of $29,000, but spent $3,000 on health care, would have that expenditure deducted from their income, reducing their income to $26,000, putting them under the NAS threshold. Thus the former family would have less income and no health care, yet would not be considered poor, while the latter family would have more income (or virtually the same in net income), yet have health care and be counted as poor. 

The costs that are to be deducted from income may be capped at unrealistically low levels. For example, if the federal Child Care and Dependent Tax Credit (CCTC) caps are used to cap child care deductions, the amount a family could deduct for child care expenses may be much lower than the real cost of child care.  The CCTC caps the amount for child care to $3,000 per year for one child and $6,000 for two or more children (or $250 per month for one child, and $500 for two children).  Data from Self-Sufficiency Standard reports for recent years show that across a number of states the minimal cost of adequate child care ranges from $423 to $707 per month for infants in family care and from $544 to $805 for preschoolers in child care centers.  Only for part-time school-age child care in one state are costs as low as the CCTC cap level ($250 to $536).[4]  Note as well that capping deductions does not acknowledge the geographic variation in costs; as suggested here, child care costs vary geographically almost as much as housing costs, but low caps negate this variation. 

By not including these costs as necessities, the revised poverty measure hides substantial amounts of very real poverty.  Such hidden poverty affects working poor families who cannot afford these necessities, primarily families with young children who need child care and health care.

Step 3: Treatment of Taxes, Tax Credits, and Transfers

In the case of the FPL and the SSS, step 3 involves comparing the threshold determined in step 1 to the gross income determined in step 2.  (Note that in the Self-Sufficiency Standard, taxes and tax credits but not benefits have been included in the “costs” calculated in Step 1).  In step 3 of the NAS approach, however, net income is compared to the threshold both before and after taxes, tax credits, and transfers. The NAS treatment of taxes, tax credits, and transfers under-estimates need and over-estimates the impact of these three factors.

Taxes:  For most people, taxes are the first cost they must pay, in the form of payroll deductions before they even see their paycheck.  This means that the amount people need to meet their needs includes the amount they have to pay in taxes upfront, including federal payroll, federal income and state income tax.  Mathematically as well, taxes need to be included at the “beginning” not the “end”.  For example, to pay the payroll tax (for Social Security and Medicare) of 7.65%, one does not need to earn just $7.65 for each $100 (earned to cover other costs), but rather $108.28 (7.65% of $108.28 = $8.28) to fully cover the payroll tax. 

The NAS approach only accounts for federal taxes. State income taxes should be included as well as federal income taxes.  If a household is earning a Self-Sufficiency Standard level of income, state income taxes range from .6% to over 6% of income in states with taxes; the average state tax burden is about 3.6%.[5] Highly variable across states, the state income tax should be included in a poverty measure.

Tax Credits (Refundable):  In contrast to taxes, tax credits are last to be received, overwhelmingly (99%, according to one survey) received as lump sums when people file their taxes, early in the following year.[6]  Moreover, most people do not use their tax credits to pay for daily expenses, such as food or housing. While some costs can be “bought” with credit, such as putting groceries on a credit card, most people according to studies (1) do not know how much they will be getting because of uneven income and (2) use their tax refund/credit for lump sum purchases.[7] The other most common uses are to make major purchases, such as a car, pay tuition, and pay first and last month’s rent, or to pay off debts.[8]  The most common debt, however, is medical debt, not usually voluntarily taken on in anticipation of a refund.  In other words, tax credits are used as “forced savings.” Therefore continuing to assume that these credits are available to meet daily costs in the year in which they are earned needs to be reexamined.

Transfers and Benefits:  There is wide consensus that any revised poverty measure should reflect the impact of benefits such as child care or heath care assistance (Medicaid, S-CHIP) on family resources.  The Self-Sufficiency Standard accounts for child care and health care costs in the first step of determining the poverty threshold, and therefore the impact of receiving assistance can be modeled as lowering the threshold wages needed, or as increasing the “income adequacy” of a given wage. However, the NAS approach does not have any means of showing their lack of ability to meet these needs.  Since families receiving these benefits are by definition too poor to secure these resources at market rates, they will not be able to deduct the cost of these expenditures from their income (Step 2).  At the same time, the NAS approach only shows the  effect of food stamps and housing assistance, because only food and housing are included in the (partial) threshold.  Benefits that reduce the cost of health care or work-related expenses, such as child care assistance or transportation cannot be credited against the costs of housing. 

3. The Impact of Using the NAS Approach versus other Alternative Measures of Poverty on both Poverty Rates and Our Understanding of Poverty

Because of the methodology used, the NAS approach creates both lower thresholds, and overestimates available resources, resulting in underestimating the number of people who are below poverty.  Using the Self-Sufficiency Standard, which is based on a full rather than partial poverty threshold, results in a very different poverty rate and count of the poor. 

For example the count of the poor increases from 18.9% according to the FPL to 23% according to the NAS approach (in the NYC application), a 22% increase in the number of people counted as poor in New York City. However, the count of the poor increases from 7-10% according to the FPL to 20-30% using the Self-Sufficiency Standard in five select states (shown in the table below), which is a two-three fold increase in the number of people with inadequate income.  (Note that the differences are probably even greater, as the studies done with the Self-Sufficiency Standard excluded the elderly and disabled, as the Standard assumes all income is earned.)

Comparison of the Federal Poverty Level to the CEO Measure in New York City and the Self-Sufficiency Standard in Select States

Note that the NAS approach not only results in a reduced poverty rate and poverty count than when the Self-Sufficiency Standard “bare bones” budgets are used, but also results in a very different picture of poverty.  Not only are approximately four-fifths of households in these states “working poor,” but the Standard tells us that their income (almost all of it wages) is insufficient to meet the costs of working, meaning that they are forgoing meeting some needs such as health care and child care, to which the NAS thresholds and poverty rate analysis is “blind.” 

4. Recommendations for an Alternative Poverty Measure and the Difference an Alternative Measure Can Make

I would like to make the following recommendations to the subcommittee, based on experience with the Self-Sufficiency Standard and the analysis above:

1. Incorporate in the legislation that the Modern Poverty Measure consider developments in terms of data availability and the experience of alternative approaches since the 1995 NAS report when devising methodology for this revision of the poverty measure.  That is, the legislation should require that the Modern Poverty Measure be able to reflect changing needs, changing demographics, and changing data availability.

Furthermore, there is no allowance, implied or otherwise, in any of these poverty standards, for debt, savings, or large capital purchases. Although the size and scope of tax credits has increased since the development of the NAS approach, so has the problem of debt and predatory lending. Certainly it would be unbalanced to assume that tax credits are available to meet daily expenses while not allowing for necessary non-daily expenses for which these credits are largely being used.  In other words, if transfers are to be included in a poverty measure, debts and large purchases should also be included. Alternatively, neither tax credits nor debts and large purchases should be included in assessing resources available to families.  This is not a simple issue, and deserves careful research.

2. Although not addressed on in the hearing testimony, we recommend that sufficient resources be devoted to developing the Decent Living Standard as a realistic, but minimal, full measure of income adequacy. Given the inadequacies of the FPL, the Self-Sufficiency Standard was developed to meet the need for a complete, but modern standard of income adequacy. Now calculated in 37 states and the District of Columbia, this type of measure has been used by approximately 2000 organizations, and increasingly has been adopted by workforce councils, state labor and welfare departments, in prison/parole systems, and many more settings across the nation. Given this high level of demand, and its evident usefulness, developing an official equivalent measure would contribute greatly to both measuring need and focusing policies and programs on true poverty reduction.

The NAS approach proposes a means to more accurately measure poverty. Although the NAS approach does measure poverty, it is not designed to function as a tool. On the other hand, a Decent Living Standard can be—and in the case of the Self-Sufficiency Standard is being—used not only for measuring poverty but also as a policy tool. The Self-Sufficiency Standard is widely used by advocates, employers, policy makers, and service providers, to improve career counseling services, target employment and training programs towards higher wage jobs. The Self-Sufficiency Standard is also used to evaluate and design public policies with the goal of helping families reach self-sufficiency. Because it provides fully comprehensive and transparent thresholds, the Self-Sufficiency Standard is a more versatile and useful tool in battling poverty because it is more than a measure. Indeed, the Self-Sufficiency Standard has gained wide usage in three-quarters of our nation’s states.[9]

The Self-Sufficiency Standard

Below I will briefly outline a possible poverty measure alternative, the Self-Sufficiency Standard. I will explain how it addresses the issues raised above, and conclude with a discussion of the difference the Standard can and has already made in our understanding of poverty, as well as being a more useful tool to combat poverty.  I developed this measure in the early 1990s, and since 1996, it has been calculated in 37 states and the District of Columbia. State reports may be found at http://www.wowonline.org/ourprograms/fess/

The Self-Sufficiency Standard measures how much income is needed for a family of a certain composition in a given place to adequately meet its minimal basic needs without public or private assistance. The Standard is designed as a national measure, with a specific methodology that is tailored to the costs of each state and county within that state. The Self-Sufficiency Standard:

·        Assumes that all adults in the household work full-time and, thus, have work-related expenses such as taxes, transportation and child care, when children are present.

·        Assumes the employer provides employee and dependents’ health insurance and uses average premiums and out-of-pocket expenses.

·        Distinguishes by family size and type. The Standard accounts for differing costs not only by family size and composition (as does the official poverty measure), but also by the ages of children.

·        Varies costs geographically and does not assume there is a universal “equivalency” scale based on the size of place or urban versus rural areas. 

Seven Categories of Expenses

The Standard measures seven categories of expenses using scholarly and credible federal and state data sources. The Standard does not rely on the cost of a single item, such as food, to establish a ratio against which to calculate the total family budget.  The Self-Sufficiency Standard is based on the cost of each basic need by county – food, housing, health care, child care, transportation and taxes – determined independently using official and otherwise publicly available data.  We add 10 percent of these costs for miscellaneous necessary expenses such as clothing, phone, and household goods.

The Self-Sufficiency Standard includes all taxes, including state and local sales and use taxes, payroll tax, federal, state and local income taxes, along with the Earned Income Tax Credit, Child and Dependent Care Tax Credit and Child Tax Credit. This is a minimal amount and produces a bare bones budget that does not take into account entertainment, savings, or education. It does not include funds for one time purchases (e.g. furniture, appliances or a car). The Standard does not build in costs related to savings for a security deposit, down payment, emergencies, retirement, college or debt repayment that can be essential in today’s economy.

Cost Components of the Self-Sufficiency Standard

Costs for the Self-Sufficiency Standard are based on data such as HUD’s Fair Market Rent, the USDA Low-Cost Food Plan, and sub-state market rates for child care published by state welfare agencies. Transportation costs are figured using data from state and local transportation departments, the National Association of Insurance Commissioners, the American Automobile Association, and the IRS mileage allowance. Since families cannot be truly self-sufficient without health insurance, employer-sponsored coverage is assumed as the norm for full-time workers. For the family’s health insurance premium and out-of-pocket costs, we rely largely on data from the Medical Expenditure Panel Survey (MEPS).[10]

What Difference Does Using a Measure Such as the Self-Sufficiency Standard Make?

·        Using the Standard results in a substantially larger number of households who lack adequate income. Because it is a realistic measure of income adequacy, it results in a substantially larger number of households who lack adequate income to meet their needs. In several state studies, using coded Census data, we have found that about one-fifth to almost one-third of non-elderly, non-disabled households lack adequate income. Specifically in California, about 30% of working-age households have incomes below the Standard, while in Washington, Colorado, New Jersey and Connecticut it is approximately one-fifth.  This is almost three times as many households as are officially counted as poor – using the FPL - in each of these states.[11] (Note that when the proportion of households lacking minimally adequate income reaches one in five, or more as in California, then it is clear that this is a systemic problem, that the issues are widespread and not simply attributable to individual issues, such as lack of education, etc.) 

·        Using the Standard provides a different picture of who is poor in each of these states.  Although there is much variation by state, several themes emerge.  Those who are below the Standard, while disproportionately people of color, are racially and ethnically diverse.  Families maintained by women alone, those who are Hispanic, and those with young children, are especially likely to have incomes below the Standard.  Yet the majority of households are married couple households, and in most states, the majority are White.  Most important, in every state, 80% or more of households with incomes below the Standard have at least one worker in them, and in roughly half of these, there is a full-time year-round worker.  There is not space here to provide detailed pictures, but because it controls for cost of living differences (such as rural vs. urban counties in a given state), it reveals within-state geographic patterns and concentrations of those households with inadequate income.

·        Using the Standard provides the means for understanding which costs are contributing the most to family budget constraints, as well as which programs are helping families make ends meet. The Standard can, and is, being used with clients to help them determine how much income they need, and thus what training/education or jobs will meet their self-sufficiency needs. The Standard can, and is, being used by program providers and policymakers to evaluate the impact of their services, and/or to model the impact on family budgets of specific services and work supports, such as child care assistance or tax credits. 

For references to studies and reports cited above, or further information about the Standard, how it is calculated, and how it can be used, please contact the author, Dr. Diana M. Pearce at pearce@u.washington.edu or (206) 616-2850, or the Center for Women’s Welfare, School of Social Work, University of Washington, 4101 15^th Avenue NE, Seattle WA 98105.

Comparison of Three Poverty Standards

[1] For more information see the Overlooked and Undercounted reports for California, Colorado, Connecticut, and Washington State, available at http://www.wowonline.org/ourprograms/fess/

[2] In fact, the NAS-type threshold calculated for New York City could be considered almost literally “half a threshold”. According to the most recent available Consumer Expenditure Survey (Table B, 2005), on average housing, food and apparel and services are 49.6% of average expenditures (median expenditures were not available). Pulling out equivalent costs from the Self-Sufficiency Standard for New York City (for a variety of four-person families across the different boroughs) we find that the total of housing, food and miscellaneous [which includes clothing] ranges from 99% to 119% of the threshold cited by Mark Levitan (except in Lower Manhattan, with its very high housing costs).  But note again, this is only part of the costs families have to pay.

[3] U.S. Department of Labor. Bureau of Labor Statistics. Employment Characteristics of Families in 2007. (page 2). Retrieved from http://www.bls.gov/news.release/pdf/famee.pdf

[4] Self-Sufficiency Standard reports, including child care costs are available at http://www.wowonline.org/ourprograms/fess/

[5] This percentage is calculated for 20 states with Self-Sufficiency Standards and with state income taxes.

[6] Of federal returns filed in 2001, only 137,685 taxpayers reported having received advanced EITC payments out of more than 16 million families with children receiving the EITC. Numbers cited by John Wancheck of the Center on Budget and Policy Priorities, based on data reported in the IRS Income Tax Section, Monthly Operational Review of Earned Income Credit.

[7] Romich, J. L. & Weisner, T.  (2000). How families view and use the EITC: The case for lump-sum delivery. Paper delivered at Northwestern University, Joint Center for Poverty Research Conference.

[8] Some workers may be unaware of the advanced payment option, and others may have employers who do not participate. Also, research has shown that families make financial decisions based on receipt of the EITC (together with tax refunds) when they file their taxes early in the following year. Romich, J. L. & Weisner, T.  (2000). How families view and use the EITC: The case for lump-sum delivery. Paper delivered at Northwestern University, Joint Center for Poverty Research Conference.

[9] The Self-Sufficiency Standard is widely used by a variety of organizations and agencies across the states. For example, online Self-Sufficiency Calculators, used by counselors and the public, have been developed for Illinois, New York, Pennsylvania, Washington State, the San Francisco Bay Area in California, and Washington, DC. In 1999, Sonoma County, California was the first county in the country to adopt the Standard as its formal measure of self-sufficiency and benchmark for measuring success of welfare to work programs. In Connecticut, the Standard has been adopted at the state level since 1998 and has been used in planning state-supported job training, placement and employment retention programs, and has been distributed to all state agencies that counsel individuals who are seeking education, training, or employment. The Standard has been used in a number of states (including New York, New Jersey, and Hawaii) to advocate for higher wages through Living Wage Ordinances and in negotiating labor union agreements. Workforce Development Boards in Pennsylvania, Washington, and Oregon are using the Standard as a case management tool. In Colorado, the Colorado Center on Law and Policy successfully lobbied the Eastern Region Workforce Board to officially adopt the Standard to determine eligibility for intensive and training services. When the Oklahoma DHS proposed large increases in the child care co-payments, the Oklahoma Community Action Project used analysis based on the Standard in a report that resulted in a rescinding of the proposed increases. More information on these and other examples on how the Standard is and can be used are available in the Self-Sufficiency Standard reports, http://www.wowonline.org/ourprograms/fess/

1 A complete discussion of data sources and methodology for the Self-Sufficiency Standard can be found on WOW’s Website at: http://www.wowonline.org/ourprograms/fess/ and clicking the report for any state.

[11] To view the Standard’s state Demographic reports, see the Overlooked and Undercounted reports for California, Colorado, Connecticut, and Washington State, available at http://www.wowonline.org/ourprograms/fess/