Download the entire Issue Guide in PDF format

Posted August 2001

How does the federal government measure poverty?
Once a year, the Census Bureau determines the number of people in poverty. Those considered "poor" live in families with incomes below the poverty threshold for their family type (based on family size and number of children in the family). For example, in 1999 the poverty threshold for a family of four with two children was $16,895, making such a family officially "poor."

To measure income, the Census Bureau uses a nationally representative survey. The official poverty definition counts money income before taxes, including wages, salaries, interest, dividends, self-employment income, welfare payments (TANF), unemployment insurance, and social security payments. Realized capital gains and non-cash government benefits (such as public housing, Medicaid, and food stamps) are excluded from this definition of income. There are, however, no poverty measurements for people in military barracks and institutional group quarters or for unrelated individuals under age 15 who live together (such as foster children); these people are not counted as either "poor" or "non-poor."

What is the difference between the poverty thresholds and the federal poverty guidelines?
The poverty thresholds are determined by the Census Bureau and are the standard used in measuring poverty. The poverty guidelines are issued by the Department of Health and Human services and are used to determine eligibility for many government programs, including food stamps, legal services, and the school lunch program. For practical purposes, the main difference between the two sets of numbers is that the poverty guidelines are more current than the thresholds. The guidelines are updated for the current year in the winter/early spring of that year, whereas the thresholds aren't updated until the sometime in the following year. The poverty guideline for a family of four in 2000 is $17,050, while the poverty threshold (last calculated for 1999) for the same family is $17,029.

How does the federal government determine the poverty thresholds?
The original poverty thresholds were developed in the early 1960s by Molly Orshansky, an employee of the Social Security Administration. She based her thresholds on the U.S. Department of Agriculture's (USDA) Economy Food Plan (predecessor of the Thrifty Food Plan), which she used to determine how much a family needed to spend on food to meet their minimum food needs. Orshansky then determined the share of income a family spent on food. In 1955, families of three or more persons (all such families, not just low-income families) spent about one-third of their after-tax money income on food. Because food costs were a third of a family's expenses, Orshansky multiplied the costs of the food plan for different family sizes by three.

These measures were later adopted by the Census Bureau as the official poverty thresholds. The Census Bureau updates the thresholds for inflation each year using the consumer price index. There have been only minor changes to the way the thresholds are calculated since they were adopted.

Are the federal poverty thresholds an accurate measure of poverty today?
As described above, the methodology used to create the poverty thresholds assumes that one-third of family income was spent on food. But, over time, the relative prices of the items that families consume have changed considerably. For example, families spend more of their income on housing, health care, and transportation than they used to, and less on food. In addition, families now spend more on certain items, like child care, due to the increased number of women in the labor force.

Due to these changes, families now spend less than one-fifth of their income on food. So if we were to multiply food costs by the share of income families spend on food today, we would come up with poverty thresholds higher than those currently in use because food costs would be multiplied by five instead of three.

These official poverty thresholds are made even more out of date because they have not been adjusted to reflect improvements in standards of living. The failure to increase the poverty thresholds as incomes (adjusted for inflation) have grown means those who fall below the poverty line are worse off relative to the typical (median) family now than they were 30 years ago. For example, when the poverty thresholds were first introduced in the early 1960s, the threshold for a family of four-$16,530 in 1998 dollars-was 42% the median income for that family size. By 1998, because of real growth in the median family income, that value had fallen to 35.4%.

Finally, poverty thresholds have not kept up with public opinion. Around the time poverty thresholds were first introduced in the early 1960s, surveys showed that the public believed that families needed an amount that was 1.4 times the poverty threshold for that year just to get by (Citro and Michael 1995). Today, the majority of Americans say a family needs $35,000 to get by, almost twice the current poverty line of $17,184 (Jobs for the Future 2000).

What are the other criticisms of the way the federal government measures poverty?

• No adjustments for geographic differences. There is only one set of poverty thresholds for the whole nation. Most researchers agree that there needs to be adjustments to the thresholds based on cost-of-living differences between different regions of the U.S., and between urban, suburban, and rural areas. However, data for making these kinds of adjustments is limited, and researchers are still debating how these adjustments should be made.
  
• Exclusion of taxes. Whether families are able to meet their basic needs depends on their income after paying taxes. In addition, many low-income families receive the Earned Income Tax Credit, which offsets other tax payments, like payroll taxes. The Census Bureau, however, continues to use family income before taxes to measure official poverty.
  
• Exclusion of other government benefits. Many poor families receive government assistance that helps them meet their needs. However, the only forms of government assistance included in the Census Bureau's definition of income are cash payments, like welfare (TANF) payments and Social Security Insurance payments. Most researchers agree that other forms of government assistance, such as food stamps, housing subsidies, free or reduced school lunches for children, and energy assistance should be included when determining whether a family is in poverty.
  
• No way to account for differences in childcare and medical care costs. Currently, the federal poverty thresholds are based on the average share of income a family spent on food in the 1950s multiplied by the cost of food. This method ignores the fact that families have widely varying expenses for certain items, like childcare and medical care. If a family has young children and all of the adults in the family work, that family will have childcare expenses that other types of families don't have. Families who do not get health insurance through an employer will have different health costs than families who try to purchase non-group insurance. Families with a disabled member will have higher medical expenses than will families in which all members are relatively healthy. In addition, when a family member is seriously ill or injured, the family might have much higher medical expenses that year compared to other years. Although most researchers agree that poverty measurements should include adjustments for the varying expenses faced by families, there is still disagreement about how this should be done.
  
• Inadequate system for adjusting for family size. There are currently different poverty thresholds based on the size of the family and the number of children in the family. The idea that a larger family has greater expenses than smaller families is common sense; the difficulty is developing a system to measure these differences. Some researchers criticize the way the current poverty thresholds adjust for family size, but there is still no agreement on how to make this adjustment.

How do the Census Bureau's Experimental Poverty Measures improve on the current methods of measuring poverty?
The Experimental Poverty Measures are the Census Bureau's first attempt to analyze the effects of the National Research Council's recommended changes to the current poverty measure. The purpose of these Experimental Poverty Measure reports is to examine how the Council's recommendations will change the demographic composition of the poor (for example, elderly versus children) as well as the total number of poor. The report does not make recommendations for a new poverty threshold.

The Experimental Poverty Measures chart the effects of the following changes to the current way poverty is measured:

• Using a percentage of median expenditures by two adult, two child families for food, clothing, shelter, and utilities as the basis of the threshold;
• Using a different measure for adjusting for family size and number of children (called equivalence scales);
• Making geographic adjustments to the thresholds;
• Adding the value of non-cash government assistance (such as food stamps, housing subsidies, school lunches, and energy assistance) to the measure of income; and
• Subtracting expenses such as taxes, costs associated with working (such as childcare), and medical expenses from the measure of income.

The percentage of people in poverty in 1998 increases from 12.7% under the current measure to as much as 15.4% under the experimental measures (depending on what version is used). Using the experimental measures resulted in a lower rate of poverty for children and a higher rate of poverty for the elderly when compared to the current measure (Short et al. 1999). The measures also found lower poverty rates for female-headed households and a higher poverty rate for married couples. Further research found a higher rate of poverty among families in which adult members work full-time compared to the current poverty measure (Iceland 2000).

Is the federal government going to change the way it calculates the federal poverty thresholds?
The federal government has no immediate plans to change the way it calculates the poverty thresholds or the way it measures poverty. In response to press speculation that the Census Bureau was close to issuing new poverty measures, the Census Bureau issued a press release stating that, while it has issued new "experimental poverty measures," none of these experimental measures have been officially adopted.

Unfortunately, adopting a new measure is bound to be a long process. The current effort started in 1992, when Congress assigned the National Research Council the task of recommending revisions to the measure. By the Census Bureau's own admission, it will take several more years for government researchers to work out the methodological issues surrounding poverty measurement (see the above question regarding other criticisms).

Even if researchers reach a consensus about poverty measurement, the new measures will have to gain political approval. The Census Bureau cannot change the poverty measure without approval from the Office of Management and Budget. In the past, presidential administrations have been reluctant to change the way poverty is measured, since revising the poverty measures means increasing the number of people who are counted as poor.

How do basic family budgets improve on the current method of measuring poverty?
Basic family budgets improve on the official poverty measure in several key areas:

• Family budgets calculate the costs for every major budget item, including housing, childcare, healthcare, food, transportation, and taxes. As noted above, the official poverty measure is based solely on food costs and the share of income families spent on food in the 1950s. The official measure is unrealistic because it does not take into account the increased share of income families are spending on costs such as housing and child care relative to food.
• Many of the items in family budgets are geographically specific. For some budget items, such as food, there is no reliable data source for local costs. But for the items that vary the most by region, such as housing and child care, the use of locally specific data in family budgets make them a more accurate reflection of the costs faced by families than the official poverty measure.
• Family budgets are a more realistic reflection of the amount of money families have to spend on basic needs because they include taxes, including the Earned Income Tax Credit. Family budgets can also be used to determine the effects of work supports (e.g., child care subsidies) on a family's ability to meet its basic needs.
• Most poverty measures are viewed as a measure of deprivation, or the minimum necessary for survival. Family budgets set a higher standard; these budgets seek to reflect the income a family needs for a safe and decent standard of living. For example, family budgets include the cost of good quality child care, because families shouldn't be expected to skimp in this area. However, most family budgets use a set of conservative assumptions, by not including items such as restaurant meals, vacations, movies, or savings for education or retirement.

Basic family budgets do, however, share some weaknesses common to traditional poverty measures:

• Counting the number of people in poverty requires having a poverty threshold for every family type. To do this, most researchers derive a threshold for a "baseline" family, such as a family with two parents and two children, and then use a formula to generate the threshold for other family types. Although there are drawbacks to adjusting a baseline family to reflect the costs of other family types (for instance, small differences in the formula can cause very different results), it does allow for creating budgets for every family type. Family budgets, in contrast, usually illustrate the needs of a limited number of family types. In order for family budgets to be used to measure poverty, a budget for every conceivable family type would need to be created.
• As noted above in the critique of the official poverty measures, different families face different costs for the budget items based on their life circumstances. Some families are able to get services outside of the marketplace. For example, some families must pay for child care services, while other families have relatives who can provide these services for free. Some families face costs that other families don't have. For example, a family with an ill or disabled family member will have higher health care costs than a family with healthy family members. Family budgets are created for a model family, and they are not able to account for the variation of costs faced by different families. This is a shortcoming they share with traditional poverty measures.

How have basic family budgets been used to influence public policy?
Family budget studies are used to influence decisions regarding government programs that support working families and labor market policies. Some key areas where they have been used include welfare reform, living wage and minimum wage increases, and workforce development programs.

• Welfare reform and family budgets.The recent attention to family budgets has been in large part a reaction to welfare reform. The foremost goal of welfare reform is for single-mother families to reduce their dependence on government benefits and increase their economic self-reliance through paid employment. This has led many poverty analysts and advocates to ask whether the earnings of former welfare recipients in the low-wage labor market will be high enough to provide for their family's minimum consumption needs, including child care, transportation, and other costs associated with working. Recent "job gap" studies try to answer this question by seeking to determine the number of jobs that pay enough for families to support themselves above the income levels set by family budgets.
• Living wage and minimum wage laws. In most cases, the wage floor set by living wage ordinances is based not on a family budget measure but on a percentage of the federal poverty guideline for a particular family size (for example, 130% of the poverty guideline). However, some local living wage campaigns have commissioned researchers to develop basic family budget measures in order to have a more accurate measure of the cost of living in their communities.
• The Workforce Investment Act (WIA).The WIA, a companion policy to the broader welfare reform legislation, was established to provide a more efficient and comprehensive system of worker training. Eligibility for "intensive" and "training" services, such as those necessary for someone trying to move from welfare to work, are subject to an income eligibility cap. While the act recommends using the Lower Living Standard Income Level budgets (which were last issued by Bureau of Labor Statistics in the early 1980s) to determine eligibility, communities have advocated for the adoption a family budget measure as a standard for eligibility and a target for the earnings of program participants.

References:
Bernstein, Jared, Chauna Brocht, and Maggie Spade-Aguilar. 2000. How Much is Enough: Basic Family Budgets for Working Families. Washington, D.C.: Economic Policy Institute.

Citro, Constance and Robert Michael. 1995. Measuring Poverty: A New Approach. National Academy Press.

Jobs for the Future. 2000. Survey prepared for the Low-Wage Workers in the New Economy Conference. http://www.jff.org/programs/cluster3/projects/careeradvstrat.html

Iceland, John. 2000. "Poverty Among Working Families: Findings from Experimental Poverty Measures." P23-203. Washington, D.C.: U.S. Census Bureau.

Porter, Kathryn. 1999. "Proposed Changes in the Official Measure of Poverty." Washington, D.C.: Center on Budget and Policy Priorities,

Short, Kathleen, et al. 1999. "Experimental Poverty Measures." P60-205. Washington, D.C.: U.S. Census Bureau.

U.S. Census Bureau. 1999. "Correction: Today's New York Times Story on Poverty in Error." Press Release, October 18.

Download the entire Issue Guide in PDF format