The poverty line is a stock figure of American statistical and political culture. Most of us are at least dimly aware that it is an official level of income used to separate the poor from the non-poor, that the government somehow sets it, that it changes from time to time, and that it is the topic of periodic political fights. Even those who know its arcane details, however, usually forget that our official measure of poverty was once crafted by a real person with a passion and a pencil.
The person was Mollie Orshansky. In 1963, she published an article about poverty that set the U.S. government on a new path. "Children of the Poor" ran in the Social Security Bulletin, an otherwise rather dry publication of the Social Security Admini- stration's Division of Research and Statistics. By the time the decade was out, Lyndon Johnson had included her poverty estimates in his 1964 Economic Report of the President, the Office of Economic Opportunity had adopted her methods for determining eligibility to its programs, and the Bureau of the Budget had mandated that the entire federal statistical establishment use her definitions and methods in reporting poverty.
The story of Mollie Orshansky and poverty measurement is a case study in what makes political arithmetic persuasive. "I wanted to show what it was like not to have enough money," she told me in an interview last summer. "It's not just that the poor had less money--they didn't have enough. I knew you couldn't spend for one necessity without taking away from another."
Orshansky knew whereof she spoke. The child of Russian Jewish immigrants, she grew up in Brooklyn in the 1920s. Sometimes her father couldn't support the family of six daughters, even when he had work. Sometimes she went with her mother to apply for relief. ("If they gave you pork, you took it even though you couldn't eat it, because you didn't want to seem ungrateful.") She learned easily to forgive the other women in line from her neighborhood who lowered their eyes and pretended not to see her and her mother.
The core of Orshansky's idea was to use the cost of a nutritionally adequate diet as the basis for a cost-of-living estimate and to calculate a cost of living for families of different sizes and composition. Because she had come to Social Security from the Bureau of Home Economics and Human Nutrition in the Department of Agriculture, Orshansky was eminently familiar with the department's food plans. These plans set forth quantities of various food groups--meat, bread, potatoes, fats, fruits and vegetables, and so forth--necessary to sustain adults and children, and using current prices, found the cost of food necessary to feed families on these rations. The department calculated several levels of food plans, which were more or less generous but whose names were as opaque as canned-olive sizes: instead of jumbo, colossal, and extra-large, the Department of Agriculture called its food plans liberal, moderate, low-cost, and economy (in descending order), the latter renamed thrifty in 1975.
Orshansky came up with 62 family types based on the number and ages of children, and for each she created a farm and non-farm version, making 124 categories. For each of these, she calculated a food budget based on the low-cost plan and one based on the economy plan, giving 248 categories in all. Here's where the pencil came in.
Food is not the only thing a family needs to live, but Orshansky chose it as the basis for her measure because there were relatively good studies of diets and their cost. She still needed a way to relate the cost of food to the cost of a family's total needs. A 1955 Department of Agriculture survey had shown that the average family spent one-third of its income on food. Orshansky assumed that families who must spend more than a third of their income on food were probably giving up other necessities--or food. So she made the elegant simplification of multiplying the cost of a family's necessary food basket by three, to arrive at an imputed total income that would put that family on the threshold of poverty.
For a non-farm family of two adults and two children, the low-cost food plan (multiplied by three) yielded an annual income of $3,995, enough to keep the family above poverty, and the economy plan yielded a figure of $3,165. Though Orshansky was dogged in creating 124 family types and income requirements, the media latched on to the archetypal family of four (minus cat or dog) as the magic number that would be called "the poverty line." In fact, the poverty line is the entire set of threshold incomes for different types of families.
Now that Orshansky had a way to measure poverty--to decide whether to count each household as poor--she needed a way to find out how many households were poor. For that, she needed to apply her definition to the whole population. With a $2,500 budget from the Social Security Administration, she went to the Census Bureau to get data on family and household income. The bureau had regularly calculated income for families of different sizes, but had never looked at income in relation to family composition: the number of children and adults and the ages of children, and the sex of the head of household. Orshansky used her $2,500 to buy these calculations, and when all was said and done, she concluded that somewhere between 17 million and 23 million children were poor.
"Children of the Poor" did for Washington bureaucrats what Michael Harrington's The Other America did for the broad reading public. Harrington had a journalist's flair for revealing the poverty Americans had ignored and for evoking empathy and outrage. Orshansky had the economist's capacity to give quasi-scientific backing to political will.
Preprints of "Children of the Poor" were circulated widely before publication and reached Sargent Shriver and others in the Office of Economic Opportunity. Legal advocates for the poor found inspiration in her measures as well. In 1964, the Justice Department called on her to help fight the poll tax; she fashioned the argument that even a $2 poll tax could deprive a poor family of a day's worth of food. In 1965, the OEO adopted her measures; in 1969, after a slight revision in the method of setting poverty thresholds, the Bureau of the Budget directed all federal agencies to use the Orshansky measures. In six years, her methods had become the indispensable tools of the poverty agencies and the standardized procedures of the statistical agencies.
To be sure, Mollie Orshansky was not the first to relate poverty measurement to the cost of food. The English town of Speenhamland in 1795 instituted a relief system that made up the difference between a worker's wage and the cost of bread sufficient to feed him and his family. In the early 1900s, several American private charities and public agencies had calculated family budgets and used them to determine relief amounts. But in the early 1960s, everyone else was counting poverty by seeking a single income level for families of every size and description. The 1964 Report of the Council of Economic Advisers, for example, used a figure of $3,000 as a threshold for all families.
Asked why her work was so influential, Orshansky usually cites three reasons. First, there was an accident of timing. "It appeared when needed," she wrote in 1988. She credits Lyndon Johnson for being ready to do something about poverty. Second, by linking poverty measures to dietary requirements and household consumption surveys, she provided a seemingly scientific rationale for poverty thresholds. And last, her estimate of $3,195 for a family of four gained plausibility by being very close to the Council of Economic Advisers' "one-size-fits-all" figure of $3,000.
There are no doubt additional reasons why her work was so influential. It was innovative, argumentative, and aggressively political. It was arithmetic from the heart. It was meant to tug at our consciences, and it gained its leverage from having its feet planted in poor people's lives.
Orshansky insisted that family size and composition made a difference in people's living standard. She refused to think in abstract aggregates of households or families. She knew she "needed to prove" that it cost more to feed more kids, and that it cost more to feed teenagers than infants. "This was the female side of research," she said.
Others looked at data showing that large families spent less per person on food than smaller families and called it economies of scale. Orshansky looked at what food these large families actually bought and saw that it was often nutritionally inadequate. "It appears that what passes for 'economy of scale' in the large family may in part reflect a lowering of dietary standards enforced by insufficient funds," she noted in "Counting the Poor."
Many economists said that if a family has insufficient funds, the housewife will just stretch the food dollar. But "economizing inside the family doesn't go even-Steven," Orshansky insisted. She explained this comment to me by spinning a hypothetical tale that is only thinly disguised autobiography: Father brings a couple of unexpected guests home to dinner, confident that Mother will just stretch the food, as she is so good at doing. Mother pulls her daughters aside and tells them, "When the meat comes around, you take one small piece and pass the dish on. Same with the potatoes." And when the meat dish comes to Mother, she takes nothing.
Orshansky refused to act as if a family is a family is a family. She looked at families by sex of the head of household, because she thought it would make a difference. It did. Mother-child families had more children on average than what she called husband-wife families. Mother-child families of every size also had dramatically lower incomes than husband-wife families with the same number of children. For example, mother-child families with two children had a median income of $2,390 while husband-wife families with two children had a median income of $6,615.
One of Orshansky's "Aha!" moments came when she found that in 1961, the median income for a non-farm family headed by a woman was about $2,340, and 40 percent of these families had less than $2,000 a year to live on. She remembers rushing into her boss's office, interrupting a meeting to proclaim her horror: "These women have to live a whole year on no more than I paid for one statistical table!" One of the women at the meeting chided her not to get so excited: "If you didn't use the money for your tables, Mollie, those women still wouldn't get any of it."
Orshansky understood that having a plausible measure was more important than pinpoint accuracy. She dedicated herself to making poverty measures reflect the reality of poor people's lives, but for her, the purpose of research was action, not "producing an intellectual treatise." "I don't think we should sit around debating whether we should do something for 32 million children or 26 million children when we haven't even done anything for 10 million."
Orshansky made her measure something people could feel. She related it to food and peppered her studies with images of what it's like to be poor. Washington was talking in terms of a minimum annual income for a family of four--Johnson's 1964 Economic Report of the President cited her $3,165 figure. Orshansky broke down that annual income into the 23 cents per meal per person it allowed, and even some of the food budget-makers in Agriculture were stunned. No one could fix a meal for that, they protested.
Orshansky's poverty thresholds were often criticized by the Left for being too stringent, and especially for using the Department of Agriculture's economy plan instead of its low-cost plan. Orshansky, though, was more cognizant than anybody of just how stringent the economy plan was. She publicized the Agriculture Department's own caveat that the economy plan was sufficient only in emergencies and on a temporary basis. She didn't remain abstract in her criticism, either. In "Children of the Poor," she invited middle-class readers to ponder the assumptions of the economy plan budgets: "They assume the housewife will prepare all the family's meals at home. There is no additional allowance for snacks or the higher cost of meals away from home, or meals served to guests. Nor is there extra allowance for the ice-cream vendor or the soda pop so often a part of our children's daily diet."
In "Counting the Poor," Orshansky put a day's food allowance into the figurative housewife's pocket and asked readers to share her plight: "The poverty line would allow a housewife with a husband and two kids about 70 cents a day per person for food. For a meal all four of them ate together, she could spend only 95 cents, and to stay within her budget, she must allow no more a day than a pound of meat, poultry, or fish altogether, barely enough for one small serving for each family member at one of the three meals."
I asked Orshansky if she had always wanted to do something for children or against poverty. But my question came from a different generation and a different economic experience. "I never thought of what I wanted to be. . . . You didn't plan to be anything--you planned to get a job."
She was the only one of the "girls in the neighborhood" to go to high school because when she graduated from eighth grade at 14, she was not old enough to work. Since she had younger sisters at home to support, she would never have gone to college but for two scholarships from Hunter College. She graduated with a major in mathematics and statistics and a minor in physics. Her first position was as a statistical clerk in the New York Department of Health. After a year, she was recruited to the Children's Bureau in Washington, D.C. Then she worked in the Department of Agriculture, and eventually came to the Social Security Administration in 1958. In all these jobs, she did what she was assigned. She was not a crusader.
Perhaps she was as much fired up by the times as she herself fired up the poverty warriors around her. She came to work on children in poverty serendipitously, after being dropped from another project that she found pretty boring anyway. Looking at children was her idea. Her first excursions into that $2,500 worth of data resonated deeply. "I don't need a good imagination when I write about the poor," she told me. "I have a good memory."
Talking about the difficulties of lifting kids out of poverty, she reflected: "You're asking someone to go alone, to break away in a sense, from the family, and to move to a group where they're not necessarily accepted. If you're 13 or 14 or 15 . . . ," her voice trails off for a moment. "It's difficult to alienate children from what they know. The things parents want for their children--the things they do out of love--make a gap."
Perhaps because she knows what it's like to have lived in that gap, she looks back today and says, "I wasn't doing a poverty line--I was doing what was necessary."
Orshansky's necessity is still the foundation of today's poverty measurement, though her measures have been much distorted from what she originally intended. Orshansky shares the main criticisms that have been leveled. According to Gordon Fisher of the Social Security Administration, people in the SSA, the Census Bureau, the Council of Economic Advisers, and the OEO all knew by 1967 that food had gone from one-third to one-fourth of the typical family's expenditures. Orshansky proposed using a multiplier of four, but by then, the OEO was already using the old thresholds based on the one-third figure, and it opposed any change. Similarly, Orshansky would not have chosen the economy plan as the basis for an official measure, but the OEO did, and OEO's poverty definition set the terms for the future.
Food now constitutes only one-fifth of non-poor families' budgets, but no agencies have been willing to use a multiplier of five. To do so would send the official poverty rate way up and create political pressure for dramatic increases in poverty spending. In The Forgotten Americans, John Schwartz and Tom Volgy show that if the official poverty measure were adjusted to stick more closely to Orshansky's original principles, the poverty rate for 1989 would have been 25.6 percent instead of 12.8 percent.
The poverty measure has not kept pace with rising real income standards, even though, since 1969, the thresholds have been adjusted by the Consumer Price Index. As Schwartz and Volgy show, any realistic measure of the minimum income necessary to cover basic necessities is now about 50 percent higher than the official poverty thresholds. In the opening paragraphs of "Children of the Poor," Orshansky foresaw the problem: "As the general level of living moves upward and expands beyond necessities, the standards of what constitutes an irreducible minimum also change."
The poverty line illustrates how a measure can take on a life of its own. Agencies become invested in protecting it from change. The current poverty lines based on Orshansky's measures are certainly too low and therefore fail to acknowledge millions of people who lack a decent standard of living. Yet despite numerous attempts during the Reagan and Bush years to manipulate the measures so as to make poverty seem to disappear, Orshansky's concepts protected a large core of the poor from statistical and political invisibility. Her thresholds were there not only when they were needed, but also when they were distinctly unwanted.
For all its independence, the poverty line also reflects the life of its creator. Perhaps every social measurement should be designed by someone for whom the job is necessary.
"Children of the Poor," Social Security Bulletin, July 7, 1963
"Counting the Poor: Another Look at the Poverty Profile," Social Security Bulletin, January 1965
"Who's Who Among the Poor: A Demographic View of Poverty," Social Security Bulletin, July 1965
"Recounting the Poor," Social Security Bulletin, April 1966
"How Poverty Is Measured," Monthly Labor Review, 1969
"Measuring Poverty: A Debate," Public Welfare, Spring 1978
Gordon Fisher, "The Development and History of the Poverty Thresholds," Social Security Bulletin, Winter 1992