Reasons for Measuring Poverty in the United States in the Context of Public Policy — A Historical Review, 1916-1995. The War on Poverty

06/01/2000

In his January 1964 State of the Union address, President Lyndon Johnson announced a War on Poverty. In connection with this announcement, the (January) 1964 Report of the Council of Economic Advisers (CEA) contained a chapter on "The Problem of Poverty in America."53 The chapter began by noting President Johnson's declaration of "all-out war on poverty in America," and setting forth the reasons for making the elimination of poverty a national goal. In presenting an analysis of the poverty population in recent years, the chapter said that "[a]lthough the analysis is statistical, the major concern is with the human problems that the numbers reflect." The chapter said that the analysis "provides a valid benchmark for assessing the dimensions of the task of eliminating poverty, setting the broad goals of policy, and measuring our past and future progress toward their achievement.... To mount an attack on poverty we must know how to select our targets. Are the poor concentrated in any single geographical area? Are they confined to a few easily identifiable groups in society?"54 The chapter presented figures on the number of families with incomes below $3,000 (in 1962 dollars) during 1947 and the years from 1950 to 1962, and on characteristics of such families in 1947, 1959, and 1962.55 The chapter concluded with a section on "Strategy Against Poverty."56

In January 1965, Mollie Orshansky published "Counting the Poor: Another Look at the Poverty Profile,"57 an analysis of the poverty population using a refined and extended version of her poverty thresholds (including thresholds for unrelated individuals and families without children, as well as for families with children). To a considerable degree, this article was a reaction to the fact that the Johnson CEA's family poverty line58 of $3,000 was not adjusted by family size, resulting in an underestimate of children in poverty and an overestimate of aged persons in poverty.59 Orshansky mentioned the problem of the CEA's size-unadjusted $3,000 figure in the first sentence of this article, and went on to write, "...the Nation is committed to a battle against poverty. And as part of planning the how, there is the task of identifying the whom. The initiation of corrective measures need not wait upon final determination of the most suitable criterion of poverty, but the interim standard adopted and the characteristics of the population thus described will be important in evaluating the effectiveness of the steps taken."60 After describing how she had developed her poverty thresholds, she compared the poverty populations under her thresholds and the Johnson CEA's poverty line. She noted a statement in the 1964 CEA report that even under a different poverty measure, "the analysis of the sources of poverty, and of the programs needed to cope with it, would remain substantially unchanged"; she went on to disagree with this statement, commenting, "...at least the relative importance of various phases of the poverty question does depend on the [poverty line] criterion used.... The resulting count of the poor [under her poverty thresholds] therefore includes fewer small families and more large ones, many of them with children.... Clearly a profile of the poor [under the Johnson CEA's poverty line] that includes large numbers of farm families and aged couples may raise different questions and evoke different answers than when the group is characterized [under her thresholds] by relatively more young nonfarm families — many of them with several children."61(Indeed, Orshansky herself subsequently noted that the significantly higher child poverty count under her measure "strengthened the rationale for the Head Start program...."62) After analyzing the characteristics of the poverty population under her thresholds, Orshansky stated in her conclusion that "[t]he causes of poverty are many and varied.... Neither the present circumstances nor the reasons for them are alike for all our impoverished millions, and the measures that can help reduce their number must likewise be many and varied. No single program, placing its major emphasis on the needs of one special group alone, will succeed. Any complex of programs that does not allow for the diversity of the many groups among the poor will to that degree leave the task undone."63

In May 1965, the U.S. Office of Economic Opportunity (OEO) adopted the SSA (Orshansky) poverty thresholds as a working definition of poverty.64 A May 10 internal OEO briefing memorandum on the new poverty definition noted that it "surfaces segments of the massive over-all [poverty] problem in a manner that focuses attention on developing a solution to the 'right' problems.... [and] provides the basis for the establishment of a 'poverty index' against which progress can be measured."65

In June 1965, the Census Bureau issued its first report66 in the Current Population Reports series that provided information on the number of persons in poverty under a quasi-official definition. However, this report did not use the SSA (Orshansky) poverty thresholds, which (as just noted) had been adopted by OEO as a working definition of poverty only the previous month. Instead, the report used a version of the Johnson CEA's poverty line; it applied the $3,00067 and $1,500 figures to 1963 income data without adjusting them for the change in prices between 1962 and 1963. The report presented information on what it called "low-income" (rather than "poor") families and unrelated individuals in 1963, as well as historical data on the number of families (but not unrelated individuals) with incomes less than $3,000 (also in 1963 dollars rather than 1962 dollars) during years between 1947 and 1963. Interestingly, the report had no comments about the purpose for preparing and presenting the figures, or uses to which they might be put; it did not even mention the War on Poverty or the fact that the $3,000 and $1,500 figures had been adopted as a (quasi-official) definition of poverty by the Council of Economic Advisers. It seems reasonable to assume that this sharp contrast with other analyses discussed here was related to a Census Bureau view of itself as a statistical agency rather than an agency advocating the adoption of specific policies.

In July 1965, Mollie Orshansky published "Who's Who Among the Poor: A Demographic View of Poverty,"68 a further analysis of the poverty population under her thresholds in 1963 together with initial figures on the poverty population in 1964. In the opening sections of the article, she wrote, "...the eyes of the world are trained on the Nation as for the first time in modern history an affluent state declares it has both the power and the persuasion to extend its bounty to all its citizens. To translate the national concern into effective action, attention must now be directed to the different groups represented among the poor and so to set the target for particular types of programs.... Much of the recent discussion of the war on poverty and the possibility of winning it centers on the number assumed as the target.... It is perhaps more difficult to define poverty as a public issue than in some other context because in a sense such a procedure implies how much of its public funds and civic energy the Nation wishes to commit to the task.... Even if we assume some consensus as to the number who merit and will receive attention, it must be determined for how many the best solution is likely to be a job, for how many it will rather be preparation for a better one, and for how many the best help will be in the form of increased financial assistance or special services."69

In January 1966, the CEA issued its annual report for 1966 — the first such report to use the SSA (Orshansky) poverty thresholds as the federal government's (quasi-official) definition of poverty. (The CEA's (January) 1965 report had used the $3,000 family poverty line as its main poverty definition, while briefly discussing the SSA thresholds as an alternative definition.70) The 1966 report presented figures on the number of persons in poverty by the SSA definition in 1959-1964. The report commented that "[f]ive years of prosperity and continued economic expansion have contributed significantly to reducing the number of people who live in poverty.... As a result both of further economic growth and of the new antipoverty programs, the data for 1965 will undoubtedly show a further drop in the number of poor." After providing some analysis of the poverty statistics, the report noted that despite "[e]ncouraging progress...the dimensions of poverty in America are still disturbing. Expanded investment in human resources and the eradication of racial discrimination are vital parts of the total antipoverty program. However, for the aged and for families headed by females, continued improvement of income-maintenance programs remains the major route out of poverty, since most of them are not — and cannot be — active members of the labor force."71

In March 1966, the Census Bureau published its first report that contained poverty statistics using the SSA (Orshansky) poverty concept. The report was not a tabulation of statistics on the national poverty population; instead, it was an advance report of results of an OEO-financed survey of Watts and neighboring areas of Los Angeles in the wake of the riots there during the summer of 1965.72 "The survey was conducted on the recommendation of Dr. Andrew F. Brimmer...in connection with his work on the task force appointed by President Johnson to investigate the riots in Watts and adjoining areas of Los Angeles in the summer of 1965." After describing the decrease in real median family income in these areas from 1959 to 1965, the report briefly discussed 1965 poverty rates in these areas.

In a December 1966 article, Orshansky commented that the SSA poverty measure (her poverty thresholds) "is now being used as a working definition of poverty to suggest the numbers and kinds of households to whom antipoverty programs might be directed."73 In a March 1968 article, she wrote that the existence of an "official working definition" of poverty "makes it possible to evaluate progress and pinpoint specific areas of concern in a way not feasible before."74

In May 1968, the Census Bureau issued its first report75 devoted entirely to the subject of the poverty population under the SSA (Orshansky) definition of poverty. The report covered trends in the poverty population between 1959 and 1966 and patterns in the incidence of poverty (particularly in 1966). It began by discussing the "pronounced decline in the extent of poverty in the United States" during the period covered. It identified the source of the poverty measure that it was using (as the Social Security Administration, not Orshansky), but did not mention the War on Poverty. The report contained passing references to "the social welfare legislation enacted in the 1960's, with its strong emphasis on education and training and on job creation and job placement," and "rapid economic growth...the inauguration of manpower development and antipoverty programs, and...the steady rise in transfer payments of all kinds...."76 However, these passing references were in the context of discussions of specific income or poverty patterns; they did not represent policy recommendations presented as stemming from the poverty trends and patterns discussed in the report. As with the June 1965 Census report discussed above, the content of the report seems to reflect a Census Bureau view of itself as a statistical agency rather than an agency advocating the adoption of specific policies.

In a February 1969 article on how poverty is measured (based on a paper presented to a 1968 seminar), Orshansky wrote, "The concept [of poverty] has to be limited by the purpose which is to be served by the definition. There is no particular reason to count the poor unless you are going to do something about them.... In the Social Security Administration, poverty was first defined in terms of the public or policy issue: To how many people, and to which ones, did we wish to direct policy concern.... A concept which can help influence public thinking must be socially and politically credible. We need benchmarks to distinguish the population group that we want to worry about. A benchmark should neither select a group so small, in relation to all the population, that it hardly seems to deserve a general program, nor so large that a solution to the problem appears impossible. For example, in the 1930's, President Roosevelt said, 'I see before me one-third of a Nation ill-clothed, ill-housed, and ill-fed.' This fraction is now part of our history. No matter how we get our number today, if more than a third of the population is called poor, it will lose value as a public reference point."77

In 1969, a federal interagency Poverty Level Review Committee which had been appointed in September 1968 decided to make two changes (relating to annual adjustment and the calculation of farm thresholds) in the SSA (Orshansky) poverty thresholds.78 On August 12, 1969, the Census Bureau issued a report79 describing and explaining the revision of the poverty thresholds. This Census Bureau report confined itself to the specific changes in the poverty thresholds and statistics based on the unrevised and revised thresholds; it did not discuss reasons for measuring poverty or uses of the poverty statistics.

On August 29, 1969, the U.S. Bureau of the Budget (predecessor of the Office of Management and Budget) issued a document80 which directed all federal Executive-Branch agencies to use the revised poverty statistics and thresholds (as issued by the Census Bureau) for statistical purposes. (It was this action that made the revised thresholds the federal government's official statistical poverty thresholds.) This document did not discuss reasons for measuring poverty, but did note that federal agencies had been using poverty statistics "in analytical and program planning work." It alluded to use of the poverty levels by federal programs for administrative purposes [to determine program eligibility] by stating that "nothing in this Circular should be construed as requiring that [the poverty levels] should be applied for such a purpose."

In December 1969, the Census Bureau issued a detailed report81 on the poverty population under the revised definition. The report covered trends in the poverty population between 1959 and 1968, and characteristics of the poverty population in 1968. The report began by noting the decline in the poverty population from 1959 to 1968 (from 39 million to 25 million persons), but did not note until later that the 1968 poverty rate was "sharply reduced" from the 1959 level. The report noted that the original poverty measure had been developed by the Social Security Administration, and cited half a dozen of Orshansky's articles.82 The report did not mention either the War on Poverty or any social programs of the 1960's. As with earlier Census Bureau reports discussed above, the content of this report seems to reflect a Census Bureau view of itself as a statistical agency rather than an agency advocating the adoption of specific policies.


53.  Chapter 2, pp. 55-84 in Economic Report of the President Transmitted to the Congress January 1964 Together With the Annual Report of the Council of Economic Advisers, Washington, United States Government Printing Office, 1964. For further information about the poverty line that the Johnson CEA used in this chapter, see Fisher, "From Hunter to Orshansky..." (1997 revision), pp. 54-56.

54.  "The Problem of Poverty...", pp. 55-58 and 61.

55.  "The Problem of Poverty...", pp. 56-57, 59-66, 71-73, and 80-84.

56.  "The Problem of Poverty...", pp. 73-78.

57.  Mollie Orshansky, "Counting the Poor: Another Look at the Poverty Profile," Social Security Bulletin, Vol. 28, No. 1, January 1965, pp. 3-29.

58.  The $3,000 figure was applied to all families of two or more persons. The January 1964 CEA report implied that $1,500 might be used as a poverty line for unrelated individuals, but did not make an explicit choice (see "The Problem of Poverty...", pp. 59 and 79). By March 1964, the CEA had made the selection of the $1,500 figure explicit; see Walter W. Heller, "Statement of...[the] Chairman of the Council of Economic Advisers" (March 17, 1964), p. 26 in U.S. House of Representatives, Committee on Education and Labor, Economic Opportunity Act of 1964[:] Hearings Before the Subcommittee on the War on Poverty Program...on H.R. 10440[,] A Bill to Mobilize the Human and Financial Resources of the Nation to Combat Poverty in the United States[ — ]Part 1.

59.  For a discussion of the genesis of the January 1965 article, see Fisher, "The Development of the Orshansky Poverty Thresholds..." (1997 revision), pp. 19-21. For a detailed discussion of the methodology that Orshansky used in this article to develop the poverty thresholds, see pp. 3-9 of the same paper.

60.  Orshansky, "Counting the Poor: Another Look...", p. 3.

61.  Orshansky, "Counting the Poor: Another Look...", pp. 11 and 13.

62.  Mollie Orshansky, "Commentary: The Poverty Measure," Social Security Bulletin, Vol. 51, No. 10, October 1988, p. 23.

63.  Orshansky, "Counting the Poor: Another Look...", p. 26.

64.  Fisher, "The Development of the Orshansky Poverty Thresholds..." (1997 revision), p. 23. Specifically, OEO adopted the lower of Orshansky's two sets of poverty thresholds — the set based on the economy food plan.

65.  P. 2 of Attachment to May 10, 1965, memorandum from Leon Gilgoff to [OEO] Senior Staff — Subject: Second Generation Definition of Poverty.

66.  U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 45, Low-Income Families and Unrelated Individuals in the United States: 1963, Washington, D.C., U.S. Government Printing Office, June 18, 1965.

67.  In April 1969, the Census Bureau issued a report on workers with earnings below $3,000 (U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 58, Year-Round Workers with Low Earnings in 1966, Washington, D.C., U.S. Government Printing Office, April 4, 1969). However, the $3,000 figure used in this report was derived as an approximation of the annualized minimum wage during the latter part of the 1960's (p. 1), not as a version of the Johnson CEA's family poverty line.

68.  Mollie Orshansky, "Who's Who Among the Poor: A Demographic View of Poverty," Social Security Bulletin, Vol. 28, No. 7, July 1965, pp. 3-32.

69.  Orshansky, "Who's Who Among the Poor...", pp. 3 and 7.

70.  Economic Report of the President Transmitted to the Congress January 1965 Together With the Annual Report of the Council of Economic Advisers, Washington, United States Government Printing Office, 1965, pp. 162-163.

71.  Economic Report of the President Transmitted to the Congress January 1966 Together With the Annual Report of the Council of Economic Advisers, Washington, United States Government Printing Office, 1966, pp. 110-114. For a few further details on the use of these poverty statistics by the CEA, see Fisher, "The Development of the Orshansky Poverty Thresholds..." (1997 revision), p. 23.

72.  U.S. Bureau of the Census, Current Population Reports, Series P-23, No. 17, Special Census Survey of the South and East Los Angeles Areas: November 1965, Washington, D.C., U.S. Government Printing Office, March 23, 1966. For the final report on this survey, see U.S. Bureau of the Census, Current Population Reports, Series P-23, No. 18, Characteristics of the South and East Los Angeles Areas: November 1965, Washington, D.C., U.S. Government Printing Office, June 28, 1966.

73.  Mollie Orshansky, "The Poor in City and Suburb, 1964," Social Security Bulletin, Vol. 29, No. 12, December 1966, p. 26.

74.  Mollie Orshansky, "The Shape of Poverty in 1966," Social Security Bulletin, Vol. 31, No. 3, March 1968, p. 3.

75.  U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 54, The Extent of Poverty in the United States: 1959 to 1966, Washington, D.C., U.S. Government Printing Office, May 31, 1968.

76.  The Extent of Poverty...1959 to 1966, pp. 1, 2, 4, and 8. On p. 8, the report referred to "the poverty index developed by the Social Security Administration," but did cite two of Orshansky's articles as sources for a more detailed description.

77.  Mollie Orshansky, "How poverty is measured," Monthly Labor Review, Vol. 92, No. 2, February 1969, pp. 37-38.

78.  For a discussion of the 1969 revision of the poverty thresholds, the events that led up to it, and the 1969 adoption of the revised thresholds as the official federal statistical definition of poverty, see Fisher, "The Development of the Orshansky Poverty Thresholds..." (1997 revision), pp. 24-33.

79.  U.S. Bureau of the Census, Current Population Reports, Series P-23, No. 28, Revision in Poverty Statistics, 1959 to 1968, Washington, D.C., U.S. Government Printing Office, August 12, 1969.

80.  Bureau of the Budget Circular No. A-46, Transmittal Memorandum No. 9, August 29, 1969, transmitting Exhibit L, "Definition of Poverty for Statistical Purposes"; see also Lawrence N. Bloomberg, "Definition of Poverty for Statistical Purposes," Statistical Reporter, No. 70-3, Office of Statistical Policy, U.S. Bureau of the Budget, September 1969, p. 37. (For the present version of the material in Exhibit L, seeStatistical Policy Directive No. 14 ("Definition of Poverty for Statistical Purposes") on p. 35 of the Commerce Department's Statistical Policy Handbook (May 1978). The text of Statistical Policy Directive No. 14 was also published in the Federal Register, Vol. 43, No. 87, May 4, 1978, p. 19269.)

81.  U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 68, Poverty in the United States: 1959 to 1968, Washington, D.C., U.S. Government Printing Office, December 31, 1969.

82.  Poverty in the United States: 1959 to 1968, pp. 1, 3, 9, and 11.