Gordon M. Fisher
(202)690-6143
March 1994
(NOTE The 75-page paper which is summarized here was presented October 28, 1993, at the Fifteenth Annual Research Conference of the Association for Public Policy Analysis and Management in Washington, D.C. [The paper was revised in August 1997 and is now 98 pages. The revised paper is available on the Census Bureau's Poverty Measurement Web site at the specific address http://www.census.gov/hhes/www/povmeas/papers/hstorsp4.html. The views expressed in both this summary and the full paper are those of the author, and do not represent the position of the U.S. Department of Health and Human Services.])
After quasi-official use beginning in 1965, Mollie Orshansky's poverty thresholds were adopted as the federal government's official definition of poverty in 1969. Perhaps because of the prominence and official status of Orshansky's thresholds, some people think that they were the first American poverty lines. However, there were a number of unofficial poverty lines in the U.S. before 1965, although their history is not widely known. I believe that the "ancient history" of 1904-1965 American poverty lines is relevant to poverty definition and measurement today because I look at the drawing of poverty lines as a social process not merely a technical economic exercise.
Early U.S. poverty lines grew out of a context of conflict during the late nineteenth and early twentieth centuries between two social groups with sharply opposed interests urban industrial workers (many of them immigrants from southern and eastern Europe) and their families, and factory owners and industrialists who often violently resisted paying the workers a living wage. Early poverty lines and family budgets were developed not by the workers themselves, but by reformist social workers and others who were trying to improve the wretched living conditions of the industrial workers and their families. After a period of "prehistory" of working-class family expenditure studies during the last third of the nineteenth century, poverty or minimum subsistence lines during a period from roughly 1899 to roughly 1946 were generally derived from the study of "standard budgets" lists of goods and services that a family of a specified size and composition would need to live at a designated level of well-being (together with the estimated monthly or annual costs of those goods and services). Some of the standard budgets of this period were developed to represent levels of living higher than poverty or minimum subsistence a fact partially obscured by the wide variation in terms used to characterize various budgets.
The first(1) American I found who used the word "poor" together with a specific dollar figure was W.E.B. DuBois, who became one of the most prominent Afro-American scholars and leaders of the twentieth century. In an 1899 book, The Philadelphia Negro, DuBois included an implicit poverty line for the Philadelphia Negro community. He did not apply this poverty line to the nation as a whole or even to the city as a whole.
The first national poverty line that I have found was set by social worker Robert Hunter in his 1904 book, Poverty $460 per year for the then-average family of five in the industrial states of the North, and $300 for such a family in the South. (After adjustment for inflation, the $460 figure would be equal to about 43 percent of Orshansky's nonfarm poverty threshold for a family of five.)
Although many poverty lines during the first third of the twentieth century were derived from standard budgets, three prominent book-length studies published in 1907-1910 did not use standard budgets. Instead, these studies assessed income adequacy for specific groups on the basis of families' actual consumption patterns rather than hypothetical budgets that families "ought" to be able to live up to. Margaret Byington, the author of one of these books, had a keen awareness of the unrealistically high expectations that the standard budget method imposes on the families for whom such budgets are developed, as indicated by a statement in a 1913 article: "Now that the theorists have effectively laid at rest the ghost of the economic man, there seems to be danger that the cost-of-living statisticians will create a new bogey, that of the economic woman; the woman who, without waste or extravagance, can on 22 cents per man per day for food, and 400 cubic feet of air space per adult, create a real home life and preserve the physical efficiency of her family."
As early as 1913 less than a decade after the first budget-based minimum income standards social worker John Shillady published an article in which he denounced private charities (run by fellow social workers) for making relief payments at levels well below these minimum income standards. This shows the deep roots in American social policy history of the current phenomenon of public assistance (Aid to Families with Dependent Children and Supplemental Security Income) payment levels being significantly lower than the official poverty level.
In 1917, a book on American labor conditions by a labor advocate and a U.S. Public Health Service statistician noted a "definite line of adequate subsistence" which was based not on standard budget studies but on recent studies of infant mortality by the U.S. Children's Bureau.
Most standard budget studies during this period were done by private or municipal researchers. During the New Deal, however, several low-income lines were developed in studies by federal government agencies, even though none of these lines was adopted as an official measure of poverty. In 1937, the Works Progress Administration issued standard budgets at two relatively low levels of living. In 1938, a report on the recent Study of Consumer Purchases included a low-income figure which was essentially an operationalization of President Roosevelt's "I see one-third of a nation ill-housed, ill-clad, ill-nourished."
After World War II there was a major break in the tradition of American poverty line studies. Poverty lines generally were derived not from standard budgets but by simply setting dollar figures, with greater or lesser amounts of supporting details and rationales. Poverty lines before about 1958 were relatively primitive; most of them did not have separate figures for different family sizes, and a number of them did not even adjust their figures for year-to-year price changes. After about 1958, poverty lines were less rudimentary; about half of them had separate figures for different family sizes. The post-1958 poverty lines were also generally higher in real terms than the pre-1958 poverty lines.
In 1949, the Congressional Joint Committee on the Economic Report [now the Joint Economic Committee] appointed a Subcommittee on Low-Income Families [SLIF] to do a study on that subject. A staff report cited figures of $2,000 (in 1948 dollars) for nonfarm families and $1,000 for farm families as designating the income group that the SLIF would study. In the relatively sparse poverty literature of the 1949-1958 period, the SLIF's $2,000 figure was adopted by a large majority of those articles which used someone else's poverty line rather than setting their own. [The best brief description of the status of the $2,000 figure would perhaps be "semi-unofficial."]
In December 1949 hearings, a representative of the Congress of Industrial Organizations (CIO)(2) rejected the $2,000 figure as too low; in 1953, CIO President Walter Reuther began using a $3,000 low-income line (the first person that I have found who did so).
In the 1954 and 1955 Economic Reports, President Eisenhower's Council of Economic Advisers quietly tried to replace the SLIF's $2,000 figure with figures of $1,500 and then $1,000. [The Council simply used the lower figures in brief discussions, without explicitly labelling them as poverty or low-income lines.] This attempt to push down the level of the poverty/low-income line was not successful.
In a chapter on poverty in The Affluent Society (1958), John Kenneth Galbraith used a low poverty line which must be classed with similar pre-1958 poverty lines. His eloquent conceptual description of poverty, however, may be classified as belonging with the post-1958 period of poverty studies: "People are poverty-stricken when their income, even if adequate for survival, falls markedly behind that of the community. Then they cannot have what the larger community regards as the minimum necessary for decency; and they cannot wholly escape, therefore, the judgment of the larger community that they are indecent. They are degraded for, in the literal sense, they live outside the grades or categories which the community regards as acceptable."
Poverty line studies were more frequent after 1958 than before 1958, reflecting and also contributing to an increasing concern about poverty in the U.S. Among those setting poverty lines in the post-1958 period were Robert Lampman (1959), Michael Harrington (The Other America, 1962), Leon Keyserling's Conference on Economic Progress (1962), and Gabriel Kolko (1962). In January 1964, when the War on Poverty was announced, President Johnson's Council of Economic Advisers (CEA) set a poverty line of $3,000 for families and $1,500 for unrelated individuals; this remained the quasi-official U.S. poverty line for a little over a year.
In 1960, Mollie Orshansky of the Social Security Administration had developed her first measures of income inadequacy in two publications which remain almost totally unknown to this day. In July 1963, Orshansky had published "Children of the Poor," describing an initial version of her poverty thresholds for families with children only; this article was cryptically referred to in the Council of Economic Advisers' January 1964 discussion of its $3,000/$1,500 poverty line. Disturbed by the CEA's failure to adjust its poverty line for family size (which had the effect of underestimating the number of poor children), Orshansky published "Counting the Poor: Another Look at the Poverty Profile" in January 1965. This article included a refined and extended version of her poverty thresholds(3) for all household types families without children and unrelated individuals as well as families with children.
Two interesting developments closely followed Orshansky's January 1965 article. In February 1965, the American Enterprise Institute published a pamphlet by Rose Friedman containing a set of poverty lines roughly 30 percent lower than Orshansky's poverty thresholds; Friedman's poverty lines represented an effort to push the poverty line down below the Johnson CEA's $3,000 figure. In the spring of 1965, Victor Fuchs published a paper which was the first paper in the United States to propose a relative (half-of-median-income) definition of poverty(4); he noted that "The fact that these [absolute] standards [of poverty] have varied enormously with time and place indicates that the search for an absolute standard is like the pursuit of a will-o'-the-wisp....attempts to define poverty in absolute terms are doomed to failure because they run contrary to man's nature as a social animal."
In May 1965, the poverty thresholds in Orshansky's January 1965 article were adopted by the Office of Economic Opportunity as a working definition of poverty, marking the end of the pre-Orshansky era in U.S. poverty definition and measurement.
This overview of American poverty lines from 1904 to 1965 has led me to the following preliminary conclusions:
One noteworthy exception to this generalization about the role of advocates is that a few poverty lines were put forward by people whom one might call anti-advocates, since their goal was to push the level of the poverty line below a currently accepted level. Examples include the Eisenhower Council of Economic Advisers in 1954-1955 and Rose Friedman in 1965. (Presciently, Robert Lampman in 1965 noted another way in which such persons might operate: instead of proposing a lower dollar figure, they might change the definition of income used.)
1. [Since writing the original paper, I have found one earlier instance of the use of the word "poverty" together with a specific dollar figure in the 1870-1871 report of the Massachusetts Bureau of Statistics of Labor. At that time, the Bureau was controlled by labor reformers who were advocating that the standard workday be reduced to only ten hours.]
2. [Before the 1955 merger of the American Federation of Labor and the CIO, the CIO was the more liberal of the two labor federations.]
3. In both the July 1963 article and the January 1965 article, Orshansky actually described two sets of poverty thresholds, one derived from the Agriculture Department's low-cost food plan and one derived from the cheaper economy food plan. It was the lower of these the set of thresholds based on the economy food plan that was adopted by the Office of Economic Opportunity in May 1965.
4. However, note that Peter Townsend (the dean of post-World-War-II British poverty studies) had proposed this relative poverty definition in a paper published in a British journal three years earlier.
[Return to the page of Further Resources on Poverty Measurement, Poverty Lines, and Their History]