Skip Main Navigation Skip To Navigation Content

Poverty - Experimental Measures

Skip top of page navigation
Census.govPeople and Households Poverty - Experimental Measures Main Experimental Poverty Measures PublicationsWorking Papers and Conference Presentations: Arranged by TopicPoverty Thresholds › Is There Such a Thing as an Absolute Poverty Line Over Time? Evidence from the United States, Britain, Canada, and Australia on the Income Elasticity of the Poverty Line

Is There Such a Thing as an Absolute Poverty Line Over Time? Evidence from the United States, Britain, Canada, and Australia on the Income Elasticity of the Poverty Line

by
Gordon M. Fisher

(An earlier version of this paper was presented October 28, 1994, at the Sixteenth Annual Research Conference of the Association for Public Policy Analysis and Management in Chicago, Illinois. A 9-page summary of this paper is available on the Department of Health and Human Services Poverty Guidelines Web site at http://aspe.os.dhhs.gov/poverty/papers/elassmiv.htm.) For a brief summary of the U.S. evidence in this paper, see Gordon M. Fisher, "Relative or Absolute--New Light on the Behavior of Poverty Lines Over Time," GSS/SSS Newsletter [Joint Newsletter of the Government Statistics Section and the Social Statistics Section of the American Statistical Association], Summer 1996, pp. 10-12. (This article is available on the Department of Health and Human Services Poverty Guidelines Web site at http://aspe.os.dhhs.gov/poverty/papers/relabs.htm.))

The views expressed in this paper are those of the author, and do not represent the position of the U.S. Department of Health and Human Services.

August 1995 (202)690-6143 [elastap4]

CONTENTS

Introduction....................................................1

Evidence from the United States--General Comment................2

Evidence from the United States--Quantitative Studies of Standard Budgets...........................................3

Evidence from the United States--Additional Quantitative Data on Standard Budgets and Poverty Lines.........................8

Evidence from the United States--Quantitative Studies of Gallup Poll Responses.....................................12

Evidence from the United States--the Common Knowledge of Informed Analysts.........................................16

Evidence from Britain--Quantitative Comparisons of Poverty Lines.............................................28

Evidence from Britain--the Common Knowledge of Informed Analysts.........................................38

Evidence from Ireland--the Common Knowledge of Informed Analysts (One Quotation).........................48

Evidence from Canada--Quantitative Comparisons of Standard Budgets and Poverty/Low Income Lines..............48

Evidence from Canada--A Quantitative Study of Public Opinion Poll Responses.............................61

Evidence from Canada--the Common Knowledge of Informed Analysts.........................................62

Evidence from Australia--Quantitative Comparisons of Standard Budgets and Poverty Lines........................64

Evidence from Australia--A Quantitative Study of Gallup Poll Responses.....................................67

Evidence from Australia--the Common Knowledge of Informed Analysts.........................................68

The Social Processes Underlying the Income Elasticity of the Poverty Line..........................................69

Why Was the Official U.S. Poverty Line Not Raised in Real Terms in 1968-1969 to Reflect Rises in the General Standard of Living?...........................70

Why Is the Income Elasticity of the Poverty Line So Little Known in the U.S. Today?...........................76

"...the opinion as to what constitutes poverty has changed in fifty years, and it is desirable to draw up a new definition for future use."(1)

Introduction

One of the central questions in discussions and debates about poverty definition and measurement is whether poverty lines should be "absolute" or "relative."(2) A recent Canadian paper which reviews a number of poverty definition and measurement issues perceptively notes that the term "absolute" is used about poverty lines in two different ways: about the way in which poverty lines are constructed or developed, and about the way in which poverty lines are updated.(3) (The same is true of the term "relative.") In the context of development, an absolute poverty line is one which is constructed as an estimate of families' minimum consumption needs; this is done without reference to the income or consumption levels of the general population. In the same context, a relative poverty line is one which is set as a fraction of the median or mean income or consumption of the population as a whole (generally with appropriate adjustments for family size); poverty lines are set in this way to demarcate a group of families whose income or consumption is deemed too low in comparison to that of the general population. In terms of updating, an absolute poverty line is one which is updated for price changes only, while a relative poverty line is one which is updated for changes in the median or mean income or consumption of the general population. Except where otherwise indicated, the terms "absolute" and "relative" will be used with reference to updating in this paper.

In a theoretical context in which time is viewed as an abstract variable rather than the setting for specific historical events, arguments can be and have been made in support of both absolute and relative updating. (In the U.S. policy context, opinions generally tend to favor absolute updating, while researchers in Europe and other developed countries generally support relative updating.(4)) However, when one turns to the context of actual history, there is an impressive body of empirical evidence from the United States, Britain, Canada, and Australia showing that successive poverty lines developed as absolute poverty lines show a pattern of getting higher in real terms as the real income of the general population rises. (A technical term that has been used for this phenomenon in the U.S. is "the income elasticity of the poverty line."(5)) In the U.S., this evidence includes both "expert"-devised minimum budgets prepared over almost six decades (1905-1960) and "subjective" low-income figures in the form of national responses to a Gallup Poll question over four decades (from the late 1940's to the late 1980's). It also includes the recorded common knowledge of a number of experts on poverty lines and minimum budgets from about 1900 to 1970. Similar although somewhat less extensive evidence is available from Britain, Canada, and Australia, showing that the phenomenon is by no means confined to one country.

This paper summarizes the extensive body of evidence for the income elasticity of the poverty line in these four countries. It also includes a brief description of the social processes that result in successive poverty lines rising in real terms as the real income of the general population rises. It makes a preliminary examination of the reasons that the U.S. poverty line was not raised in 1968-1969. Finally, it briefly discusses some institutional and sociological factors that seem to account for the extensive body of evidence about the income elasticity of the poverty line being so little known in current academic U.S. discussions about poverty lines.

Evidence from the United States--General Comment

Since about 1900, as the real income of the general U.S. population has risen, successive "absolute" poverty/subsistence lines developed by U.S. analysts of poverty have also risen in real terms. The same has held true of people's responses to a Gallup Poll question (the so-called "get-along" question, asked in surveys since the 1940's) about the amount of money needed by a family of four to "get along in this community." The evidence of this phenomenon from both "expert" budgets and "subjective" survey-derived low-income figures is particularly impressive because it seems to reflect a process that goes on at an almost unconscious level. People answering the Gallup Poll question have no way of knowing what other people answered five or ten years ago--let alone what that answer amounted to in today's dollars. Some family budget experts may have known the level of a budget that a colleague established six or eight years earlier--but even they probably didn't know what figures their more remote colleagues had come up with twenty or thirty years earlier. Accordingly, budget experts and poll respondents cannot (in most cases) be asking themselves, "Should we raise the low-income figure this year or not?" Nevertheless, the resulting dollar figures (after adjustment for inflation) look as if they were responses to a question such as "What is an appropriate low-income figure in the light of today's standard of living?" (even though the actual question--in the case of the Gallup Poll--contains no reference to "today's standard of living").

Evidence from the United States--Quantitative Studies of Standard Budgets

Since the beginning of the War on Poverty, there have been a number of quantitative studies(6) of the levels of "expert" budgets and the Gallup "get-along" amount over time. The studies on "expert" budgets are summarized in this section, while the studies on the "get-along" amount are summarized in the next section but one.

To understand the studies summarized in this section, one needs to be familiar with the concept of the "standard budget." A standard budget is a list of goods and services that a family of a specified size and composition--and sometimes of a specified social class or occupational group--would need to live at a designated level of well-being, together with the estimated monthly or annual costs of those goods and services.(7) (The phrase "market basket" is often used nowadays to refer to this concept, but that phrase is sometimes used so loosely that it is preferable to use the more precise technical term "standard budget.")

To understand the studies summarized here, one also needs to realize that standard budgets can represent standards of living higher than poverty or minimum subsistence. This can be seen in a useful classification scheme for budgets which was presented by Oscar Ornati in a 1966 book(8) which included an analysis of the costs of about 60 standard budgets prepared during the 1905-1960 period; he classified these budgets as being at "minimum subsistence," "minimum adequacy," and "minimum comfort" levels.

  • While he equated his "minimum subsistence" category with eligibility standards for "relief," a comparison of his figures with Mollie Orshansky's original 1963 poverty thresholds and figures in 1949 and 1959 low-income studies indicates that this category corresponds to what we think of as poverty.
  • His "minimum adequacy" category represented levels used in the 1960's by charitable organizations to determine eligibility for free family counseling and child guidance services.
  • His "minimum comfort" category represented "the threshold of 'comfort, ' as...viewed in contemporary terms."

Ornati provided dollar figures for each budget (recalculating the figures on the basis of a four-person family when necessary); for all years during the 1905-1960 period for which no budget was developed in a given category, he provided dollar figures by interpolation. Ornati's classification and his three series of year-by-year dollar figures will be used as a framework in which to view the U.S. budgets discussed in this paper.

The major findings of the quantitative studies of standard budgets are as follows:

  • In a 1965 paper(9), Eugene Smolensky analyzed eight figures for "minimum comfort" budgets for New York City between 1903 and 1959. (He did not make any distinction between "minimum comfort" budgets and poverty lines.) He found that the New York City "minimum comfort" budget per capita rose in real terms over time in such a way that it remained roughly equal to a constant proportion (one half) of gross national product per capita.
  • In a 1973 article(10), Robert Kilpatrick reanalyzed Smolensky's "minimum comfort" budget figures. After adjusting the data for inflation, Kilpatrick found that the New York City "minimum comfort" level rose about 0.84 percent in real terms for each 1.0 percent increase in real disposable income per capita. (The 0.84 figure was significantly different from zero, but he could not say that it was significantly different from 1.0.)
  • Kilpatrick also analyzed the year-by-year figures for the 1905-1960 period that Ornati(11) had published for his three budget levels, again adjusting all data for inflation. Kilpatrick found that Ornati's "minimum subsistence" level rose 0.75 percent in real terms for each 1.0 percent increase in real disposable income per capita. For Ornati's "minimum adequacy" level, the corresponding figure was 0.88 percent. For Ornati's "minimum comfort" level, the corresponding figure was 0.998 percent. All three coefficients were significantly different from zero. The "minimum subsistence" and "minimum adequacy" coefficients were significantly different from 1.0 at the 95 percent confidence level, but there was clearly no significant difference between the "minimum comfort" coefficient and 1.0. (In other words, the "minimum comfort" level rose in real terms at essentially the same rate as real average income for the whole population.)
  • Ornati himself had not explicitly analyzed the quantitative relationship between the levels of his budgets and the level of the real income of the general population, but had written, "As the general U.S. income level has risen, living standards have risen. Expectations have risen. Concepts of who is poor have risen. Thus, through the years, standards of sufficiency, however defined, have generally risen." He calculated and graphed the constant-dollar values of his three budget levels, showing clearly how they had risen in real terms from 1905 to 1960; the minimum comfort and minimum adequacy levels more than doubled in constant dollars, while the minimum subsistence level almost doubled. He also found that his minimum subsistence level remained roughly equal during this period to an index of men's low wages (wages of unskilled male labor) that an associate had constructed, although the minimum subsistence level fell slightly below the index of men's low wages after the Depression.(12)
  • Ornati also provided some useful tabular comparisons of some "commodity" features of budgets at all three levels in 1908, 1935, and 1960.(13) In a 1974 book, Lee Rainwater used Ornati's tabular material to make some interesting comparisons(14) between the 1908 and 1960 minimum subsistence budgets and also between the 1908 minimum comfort budget and the 1960 minimum subsistence budget(15).
    • In the subsistence budgets, the amount of high-protein foods (e.g., meat and milk) for a family of four per week rose from 21.7 pounds in 1908 to 32.2 pounds in 1935 and 54.7 pounds in 1960. The number of rooms allowed for the family rose from four in 1908 to five in 1960. The 1908 budget did not allow for any electricity, but electricity for lights, a refrigerator, and an iron was allowed for in the 1960 budget. No transportation was allowed for in the 1908 budget, but the 1960 budget allowed for 520 "carfares." (In 1960 dollars, as deflated by the personal consumption expenditures deflator, the total costs of these budgets were $1,360 for 1908, $2,060 for 1935, and $2,660 for 1960.)
    • Minimum subsistence in 1960 also represented a higher physical standard of living than minimum comfort in 1908. The amount of high-protein foods for a family of four was 25.9 pounds in 1908 but 54.7 pounds in 1960. The 1908 comfort budget allowed for three "decent" and "sanitary" rooms for the family, while the 1960 subsistence budget allowed for five rooms. The 1908 comfort budget did not allow for any electricity or transportation, while the 1960 subsistence budget did contain the modest allowances for those categories noted in the previous paragraph. (In 1960 dollars, the 1908 minimum comfort budget cost $2,320, while the 1960 minimum subsistence budget cost $2,660.)
  • In a 1977 article(16), Diana Karter Appelbaum studied all minimum budget studies that she could locate which were made in major American cities between 1906 and 1929, and compared them with several minimum standards dating from 1960 or later. (Because of the difficulty of interpreting dollar prices in different areas in different years, she only studied pre-1929 budgets that specified quantities of goods and services.) She found that the modern Economy and Thrifty Food Plans "allow much more of almost every foodstuff" than the food components of the pre-1929 budgets. Persons-per-room housing standards in budgets before 1924 (no more than 1.5 to 1.7 persons per room) were "markedly less generous" than the standard of adequate housing used in the 1960 Decennial Census (no more than 1.0 persons per room); furthermore, almost none of the pre-1929 budgets required a bathtub for a family, whereas the 1960 Census standard did require a "bath" for the exclusive use of the occupant(s) of the housing unit. Appelbaum concluded that the poverty/minimum budget standard has risen in real terms over the twentieth century just as the general American standard of living (the average income) has risen in real terms.
  • In addition, Appelbaum tried to compare the general standards of living represented by the pre-1929 budgets and the Orshansky poverty line by comparing the shares of the total budgets going for food (making use of one aspect of Engel's Law--that a lower proportion of total expenditures going for food means a higher standard of living). This procedure is incorrect in this particular case, since the food share of one third used to calculate the Orshansky poverty line actually reflects the food/total-income share of the average American family in 1955, rather than stemming from a detailed item-by-item budget constructed for a 1963 poor family.(17) However, a valid comparison of this type can be made over a shorter period: by Engel's Law, Margaret Stecker's "emergency" budget for 1935 (with a food/total-expenditures share of 37.6 percent)(18) represents a higher standard of living than the 1906-1929 budgets (with a median food/total- expenditures share of 42 percent(19), including insurance and health care). This modified comparison supports Appelbaum's overall conclusion.

Evidence from the United States--Additional Quantitative Data on Standard Budgets and Poverty Lines

In 1993 I wrote a paper (cited in footnote 18) on unofficial poverty lines in the U.S. between 1904 and 1965. Since it added "derivative" poverty lines (20), early non-budget-based poverty/subsistence estimates, and several standard budgets to budgets cited by Ornati and budgets studied by Appelbaum, that paper(21) provided further evidence supporting the income elasticity of the poverty line. As with the Ornati and Appelbaum budgets, the budgets and poverty lines discussed in that paper were all conceptually derived as "absolute" poverty lines, without any reference to the income distribution as a whole. Yet over time, such "absolute" poverty lines rose in real terms as the income of the general population rose. Poverty lines and minimum subsistence budgets before World War I were, in constant dollars, generally in the range of 43 percent to 54 percent of Orshansky's poverty threshold. By 1923, Dorothy Douglas' "minimum of subsistence level" was equal to 53 percent to 68 percent of Orshansky's threshold. Margaret Stecker's "emergency" budget for the Depression year of 1935 (published by the U.S. Works Progress Administration in 1937) was equal to 65 percent of Orshansky's poverty threshold. Robert Lampman's low-income line for 1957 (published in 1959) was equal to 88 percent of Orshansky's poverty threshold.(22)

The early poverty lines and subsistence budgets mentioned in the preceding paragraph were either implicitly or explicitly applicable to a particular family size, and thus could be easily compared with the appropriate Orshansky poverty threshold. Poverty lines not adjusted for family size are more difficult to compare with the Orshansky poverty thresholds, but can be compared with each other. For instance, the 1949 family low-income line of the Subcommittee on Low-Income Families (SLIF) was equal in constant dollars to 84 percent of the family poverty line that President Johnson's Council of Economic Advisers (CEA) set in 1964. Or, to put it the other way around, the Johnson CEA's poverty line was 19 percent higher in real terms than the figure put forward by the SLIF only fifteen years earlier.(23)

As will be shown below, this historical pattern of poverty lines and subsistence budgets rising in real terms over time is found not only in the U.S. but also in Britain and Canada. In the U.S., as implicitly indicated above, the poverty/subsistence line (for a family of five) as measured by the Douglas minimum subsistence level and the Orshansky poverty threshold was between 1.5 and 1.9 times as high in real terms for 1963 as it was for 1923. Over approximately the same period of time, the U.S. poverty/subsistence line (for a family of four) as measured by Ornati's minimum subsistence budgets was 1.4 times as high in real terms for 1960 as it was for 1923.(24) In Britain, the P1 poverty line (for a family of five) was 1.7 times as high in real terms for 1963 as it was for 1924. (See p. 31 below.) And in Canada, as measured by the Canadian-priced Douglas minimum subsistence budget and the Montreal minimum budget (both for 1926, but with different dollar amounts) and Podoluk's original low income cut-offs, the low income/subsistence line (for a family of five) was either 1.7 or 2.1 times as high for 1961 as it was for 1926. (See p. 56 below.) Conceptually, the amount by which a poverty line rises in real terms over time is a function of two variables--the amount by which the real average income (however measured) of the general population rises, and the elasticity of the poverty line with respect to the measure of real average income. Since the values of these two variables are likely to differ from country to country for any given period of time, one should not necessarily expect the amount by which the poverty line rises to be identical in several countries over the same period of time. However, it is of interest that poverty/subsistence lines in these three countries rose in real terms by about the same order of magnitude over (approximately) the same four-decade period.

The examples given above demonstrate the income elasticity of the poverty line by showing that poverty lines or minimum subsistence budgets during later periods (periods of higher real income among the general population) are higher in real terms than poverty lines or minimum subsistence budgets during earlier periods (periods of lower real income among the general population). The income elasticity of the poverty line can also be demonstrated by taking a given constant-dollar level which is considered to be poverty during a later period (of higher general real income) and showing that the same level was considered to be well above poverty during an earlier period (of lower general real income). My 1993 paper did precisely that by finding five U.S. budgets or other estimates during the 1901-1929 period that were approximately equal to Orshansky's poverty threshold in constant-dollar terms and showing that they were all at Ornati's minimum comfort level, which was two levels above Ornati's minimum subsistence level. To consider just two of these early figures at Ornati's minimum comfort level:

  • Margaret Byington described families in 1907 with constant-dollar incomes equivalent to 92 percent of Orshansky's poverty threshold as "liv[ing] well" and "satisfy[ing]...the reasonable ambitions of an American who puts his life into his work"; in 1965--and even more so in the 1990's--no reasonable person would describe a family with that real income as "living well."
  • Similarly, Dorothy Douglas in 1923 described a "comfort level"--equal in constant-dollar terms to 102 percent of Orshansky's poverty threshold--which represented "the attainment of the highest class of wage-earners and the cynosure of the rest." In 1965--and even more so in the 1990's--no one would describe an income of 102 percent of the poverty threshold as "the attainment of the highest class of wage-earners and the cynosure of the rest."

In December 1919, the Bureau of Labor Statistics (BLS) published a "Tentative Quantity-Cost Budget" showing the cost required to maintain a five-person family headed by a federal government employee in Washington, D.C., "at a level of health and decency."(25) In June 1920, BLS published a version of this budget (a "Minimum Quantity Budget...") revised to apply to a workingman's family rather than to a government clerical worker's family.(26) (I did not mention these budgets in my 1993 paper because that paper was focused on poverty lines, while these budgets were for a standard of living significantly higher than poverty. I mention them here because of their connection with some Canadian and Australian evidence discussed below.) Ornati classed these two budgets at his minimum comfort level--two levels above his minimum subsistence [=poverty] level. As of August 1919, the annual cost of the "Tentative Quantity-Cost Budget" was $2,262.47. This figure would be equal to about $3,910 in 1963 dollars--106 percent of Orshansky's 1963 average nonfarm poverty threshold of $3,685 for a family of five. This represents yet one more example of a constant-dollar level which was considered well above poverty during an earlier period but was considered (approximately) equal to poverty during a later period.

This relationship found in the U.S. between the real levels of the 1919 BLS above-poverty budget and the 1960's poverty line is also found in two other countries. In both Canada and Australia (see pp. 52-53 and 65-66 below), the real value of a 1920's above-poverty budget more or less closely related to the 1919 U.S. BLS budget was roughly equal to the real value of a 1960's poverty line. (In Canada, the 1920's budget was equal to 95 percent of the 1960's poverty line in real terms. In Australia, the 1920's budget was equal to 82 percent of the 1960's poverty line in real terms. All figures in both countries were for five-person families.) As with the earlier cross-national temporal comparison of poverty lines, it is of interest that the quantitative relationships between the post-World-War-I above-poverty budgets and the 1960's poverty lines are of the same order of magnitude in all three countries.

Evidence from the United States--Quantitative Studies of Gallup Poll Responses

Since January 1946(27), the American Institute of Public Opinion's Gallup Poll has repeatedly asked the following question: "What is the smallest amount of money a family of four (husband, wife, and two children) needs each week to get along in this community?"(28) (This question is sometimes referred to as the "get-along" question. Versions of the question asked between 1937 and 1944 (four times) and in 1963 asked what a family needed "for health and comfort," "health and decency," or "comfort and decency" rather than "to get along.") The average response to the question has been higher than the Orshansky poverty line; indeed, Lee Rainwater has found that the average "get-along" amount is quite close to Ornati's minimum adequacy level. However, it seems reasonable to assume that the way the "get-along" amount varies over time in relation to family income would be a good indicator of the way that the public's perception of the poverty line would vary over time in relation to family income (if such a question had been asked).(29)

As noted earlier, there have been a number of studies of the level of the Gallup "get-along" amount over time. The major findings of these studies are as follows:

  • In his 1973 article (see footnote 10), Kilpatrick analyzed not only Ornati's and Smolensky's budgets but also the responses to the Gallup "get-along" question; indeed, the "get-along" amount was the primary focus of his article. He examined data for the period 1957-1971, during which the question was asked ten times. (He omitted figures for 1946-1954 from his final analysis because of data quality problems.) After adjusting all data for inflation (using the deflator for consumption expenditures), he ran regressions involving the "get-along" amount and two different measures of average income for the total population: per capita disposable income and median family income. For both measures, using several different versions of equations, he found that the "get-along" amount rises about 0.6 percent in real terms for each 1.0 percent increase in realaverage income. The differences between the 0.6 figure and zero and between the 0.6 figure and 1.0 were both statistically significant; in other words, he could state with confidence that the "get-along" amount rises in real terms as real average income rises, and that the inflation-adjusted "get-along" amount rises at a slower percentage rate than real average income rises.
  • In 1973 and 1974, Rainwater(30) analyzed the Gallup "get-along" responses, looking at figures for dates between 1946 and 1969. He found that the "get-along" amount rises so that it is a roughly constant percentage of overall average income. The "get-along" amount showed a small tendency to drop as a percentage of median family income [which would be consistent with Kilpatrick's results], but it remained roughly constant as a percentage of "per family disposable personal income" (a measure which Rainwater constructed as four times per capita disposable income).
  • In his 1973 and 1974 analyses, Rainwater also compared the "get-along" amount with figures for average weekly spendable earnings for a worker with three dependents(31). He found that the "get-along" amount rose so that it remained roughly equal to average weekly spendable earnings throughout the 1947-1969 period. (More precisely, the "get-along" amount averaged 106 percent of spendable earnings during this period. The comparison omitted 1946 because the spendable earnings series was not available for that year.)
  • In a 1978 article(32), Irving Leveson analyzed the Gallup "get-along" responses, looking at data for the period 1947-1973. He found that the "get-along" amount rises 0.65 to 0.67 percent in real terms for each 1.0 percent increase in real median family income --a result quite similar to Kilpatrick's. (The 0.65-0.67 figures were significantly different from 1.0; from the way he phrased another finding, the 0.65-0.67 figures may be presumed to be also significantly different from zero, although he did not say this in so many words.)
  • Leveson also found that the inflation rate, the unemployment rate, and defense expenditures had no statistically significant effect on the level of the inflation-adjusted "get-along" amount during this period (in equations that included these variables along with real median family income).
  • Leveson also noted the interesting fact that the 0.65-0.67 figure was almost identical to the 0.64 proportion of 1929-1969 growth in per capita output (potential national income per person potentially employed) which was attributable to general sources of economic growth (advances in knowledge and economies of scale) rather than to "those sources of growth which are associated with gains accruing specifically to individual workers, owners of factories, land, and so on as a result of their own efforts...." It is almost as if the society had unconsciously decided that the poor (more precisely, those with incomes below the "get-along" amount) would be allowed to share in the proportion of economic growth due to general sources, but not the proportion of growth attributable to the specific efforts (or resources) of individuals.
  • In a 1984 book on economic inequality in the U.S.(33), Canadian economist Lars Osberg examined the (U.S.) Gallup "get-along" responses, presenting figures for dates between 1947 and 1977 in a table together with "Median 14 [years of age] and over male full-year earnings, all industries." His table showed that the "get-along" amount remained roughly equal to median male earnings throughout the period in question. When one examines his figures closely, one finds that the "get-along" amount was equal to between 96 percent and 114 percent of median male earnings during the 1947-1961 subperiod, and between 89 percent and 102 percent of median male earnings during the 1962-1977 subperiod; this suggests the possibility that an earnings elasticity figure calculated for the whole period might be somewhat less than one.
  • In 1990 Lee Rainwater(34) reanalyzed the Gallup "get-along" responses, looking at data for the period 1950-1986. He used means calculated from the original survey microdata rather than the published medians. He found that the "get-along" amount rises so that it is a roughly constant percentage of overall average income--in other words, the income elasticity that he found was 1.0 rather than the lower figures found by Kilpatrick and Leveson. Because he found no time trend in the ratio of the current-dollar figures, he did not analyze any constant- dollar figures. The measures of average income that he constructed and used were "per couple consumption" (personal consumption expenditures divided by the number of persons 18 or over and then multiplied by two) and mean "household" income (which he actually calculated as an average of mean family income and mean unrelated individual income from the Current Population Survey). Other studies have found that the income elasticity of Gallup "get-along" and "health/decency" amounts can be affected by the use of means rather than medians (see the next paragraph) and by the use of different variables for average income (see the Saunders/Bradbury study summarized on pp. 67-68 below); both of these factors may well have contributed to the difference between Rainwater's elasticity figure and Kilpatrick's and Leveson's elasticity figures.
  • In 1993 Denton Vaughan published an article(35) in which he used the Gallup "get-along" responses for the period 1947-1989 plus the answer to a 1989 Gallup Poll "poverty" question to construct a social poverty line series covering that period. In this article he also examined the income elasticity of the (constant-dollar) "get-along" amount with respect to the real median family income of four-person families, on both a before-tax basis and an (estimated) after-tax basis. On the after-tax basis he found that his result varied noticeably (between 0.68 and 0.80) depending on whether he used a mix of published medians and Rainwater's microdata-derived means, or only Rainwater's means. Using only Rainwater's means, he found that the income elasticity of the "get-along" amount was 0.65 with respect to the before-tax income figures and 0.80 with respect to the after-tax income figures.

Evidence from the United States--the Common Knowledge of Informed Analysts

Another significant source of evidence about the income elasticity of the poverty line--although it has generally been neglected in previous income elasticity literature--is the common knowledge of experts on poverty lines, budgets, and living standards, as documented in quotations from their writings. These experts were quite familiar with the social processes in our society which result in the income elasticity of the poverty line, and felt no qualms about recording their knowledge without (in most cases) needing to rely on formal quantitative studies. Most of the quotations that I will cite come from the period between 1900 and 1970, although one goes back as far as 1841. (In some cases the discussion is in terms of a living standard somewhat higher than poverty or subsistence; however, it is clear that the same general social processes are being described.)

In an essay originally published in 1841, Unitarian preacher and social reformer Theodore Parker wrote, "The world no doubt grows better; comfort is increased from age to age. What is a luxury in one generation, scarce attainable by the wealthy, becomes at last the possession of most men....As society advances, the standard of poverty rises. A man in New England is called poor at this day, who would have been rich a hundred and fifty years ago....The world has grown rich and refined, but chiefly by the efforts of those who themselves continue poor and ignorant."(36) While Parker's discussion of the rise in "the standard of poverty" did not explicitly tie it to rising real incomes of the general population, he commented in the same context that "comfort is increased....society advances....The world has grown rich and refined...."; accordingly, it seems fair to credit him with at least a general knowledge of the phenomenon now known as the income elasticity of the poverty line.

In an April 1902 article on the concept of a living wage, Father John A. Ryan, a Catholic priest and economist, wrote, "Because of the development of new wants[,] a decent livelihood now may be less than a decent livelihood ten years hence. To ignore the new wants, then, would be as harmful as to ignore existing wants now. A true decent livelihood, therefore, is relative, not only to particular communities, but to different stages of progress in each community."(37)

In a 1903 book on organized labor, John Mitchell, president of the United Mine Workers, wrote, "...I believe that what should now be considered as the American standard and as a minimum wage, will, in the course of ten or twenty years, cease to be so considered, since it is reasonable to anticipate that the earnings of workingmen and their requirements for comfort, will, with the progress of the age, increase in the future as they have in the past."(38)

In 1917, W. Jett Lauck (an economist and labor advocate) and Edgar Sydenstricker (a Public Health Statistician in the U.S. Public Health Service) published a book on American labor conditions. After discussing family budget investigations as well as Children's Bureau infant mortality studies that showed a strong negative correlation between infant mortality and family income, they proceeded to point out that "...the social standard of minimum subsistence has become more costly [since 1900]. New desires and new wants have been created, and it is impossible to assume that the wage-working family has not been affected in much the same way as the family of the business man, the banker, the office worker, or even the farmer....there are...changes...which have been brought about in response to those wants whose creation has been the mark of advancing civilization. Good or bad, changes in the customs and manner of living can not be overlooked in considering the question of adequacy of wages and family income. They are social products for which we can blame the wage-working population least of all. The fact which is of distinct pertinence here is that even if the levels of prices and wages had remained without change since 1900, the cost of living would have increased, because the social standard of living has become more expensive. To live adequately to-day costs more than it did even ten or fifteen years ago, not simply because prices have gone up, but because our standards of health, comfort, and efficiency are more exacting...."(39)

In a 1919 article on budgets, living costs, and wages, William Ogburn, a University of Washington professor who had done a good deal of work on family budgets, wrote that "The minimum of subsistence will...change over a period of time, irrespective of the level of prices. What was the minimum of subsistence a number of years ago is certainly not a minimum of subsistence now."(40)

In a 1934 social sciences encyclopedia article on poverty, social insurance expert Isaac Rubinow wrote, "In addition to the discrepancy between existing needs and available resources as a whole[,] one must take into account [in measuring poverty] such factors as the influence of prevailing social standards, the inevitable if sometimes semiconscious comparison of differences of standards within the community and its resultant stimulation of desires....the concept of poverty as based upon scales and standards of living is comparatively recent and has arisen in dynamic industrial societies....Luxuries become comforts, comforts become necessaries....The recent efforts in industrial countries to measure the extent of poverty have usually been based on data as to the distribution of wealth and particularly of income. The standard of income is necessarily reduced to terms of monetary values which fluctuate rapidly even within short periods of time. In the United States, for instance, the estimate of the normal annual income required to maintain a standard of health and decency for a family of five has jumped from $600 at the beginning of the twentieth century to nearly $2000 at the end of the third decade [of the twentieth century]. A similar rise is to be noted in English estimates. The difference is due not entirely to increase in monetary costs [i.e., prices] but also to a large extent to changing standards. Presumably all families whose total income falls beneath the standard are in the poverty group."(41)

In July 1937, the Works Progress Administration (WPA) published a research monograph by Margaret Stecker presenting and pricing quantity budgets of goods and services at two levels of living. Although her language was terse, Stecker does seem to have recognized the concept of the income elasticity of the poverty line; discussing her low-cost budgets, she wrote, "...as consumption itself is raised to higher levels, budgets must take account of the improved standards of living thus manifested."(42)

In 1938, Carroll Daugherty, an economics professor, published a revised edition of a book on American labor problems which included a section on standards of economic well-being. In that section, he noted that "The [standard family] budget itself must also be changed occasionally, whenever there are significant shifts in the nature of the items concerned or whenever people's objectives and standards change. A standard budget worked out in the [1890's], for example, would have no place for electric appliances, automobiles, spinach, radios, and many other things which found a place on the 1938 comfort model. The budget of 1950 will undoubtedly make the present one look as antiquated as the hobble skirt."(43)

At the April 1949 meeting of the Conference on Research on Income and Wealth, economist Dorothy Brady presented a paper on income distribution research which included a section on poverty. In that section, she wrote, "The methods of determining the poverty line in different situations have, however, usually yielded a measure that varies with the general level of living."(44)

Early in 1955, the Congressional Joint Committee on the Economic Report [subsequently renamed the Joint Economic Committee] reconstituted its Subcommittee on Low-Income Families to do further studies of the problems of low-income families. (The first incarnation of the Subcommittee on Low-Income Families had been in 1949 and 1950, when--as in 1955--it held hearings and issued several reports.)(45) In hearings held by the Subcommittee on Low-Income Families in November 1955, two expert witnesses referred to the phenomenon now known as the income elasticity of the poverty line. On November 18, economist Robert Nathan testified and also presented a written statement which included the following comments: "...the term 'low income' is relative in time and location....What we term 'a low income' today was, in real terms, a middle income only a generation or two ago. And today's middle income will be a low income before the end of the century. Poverty or low incomes must therefore be appraised in the light of current and national circumstances. In such terms, we do have an important and challenging problem."(46) On November 23, economics professor Margaret Reid stated in the course of her testimony that "We have a certain notion as to what is poverty, but it doesn't stay fixed. As we get more real income, people are looked upon as poor even though they have improved their lot. We call them poor because they still are in the lower-income levels....I was thinking in terms of 20 years. I wasn't thinking in terms of a continual shifting [of the standard of poverty] every year. I don't think we shift our standard that way. We do establish certain kinds of benchmarks and we hold them for quite a while, but I am talking about the upsurge in productivity that comes over the longer period of time. If you trace statements on the extent of poverty in the United States from the [1930's] to date the shift in yardstick is very apparent. Even after allowance is made for price change it is apparent that the poverty point of $2,000 to which you refer was looked upon as a fairly good income throughout much of the [1930's]."(47)

In December 1959, the Conference on Economic Progress published a study entitled The Federal Budget and "The General Welfare"[--]We Can Afford To Serve Our Human Needs. (The Conference on Economic Progress was a non-profit group which favored increased public spending to stimulate economic growth, full employment, and full production, and to meet important human needs.) In a section on "American Poverty And The General Welfare," the study noted that "The standard of living does and should rise as a Nation's technology rises. A level of family living which was regarded as being above 'poverty' decades ago may not be above 'poverty' today."(48)

In December 1960, a (then) obscure Social Security Administration employee named Mollie Orshansky wrote an article about the Bureau of Labor Statistics' recent revision of a "modest but adequate" Budget for an Elderly Couple. Discussing the increase in the budget's cost over about a decade, she wrote, "The higher cost estimate for the current version of the budget [the Budget for an Elderly Couple] includes more than a simple response to changing prices. Some [of the] increase [from 1950 to 1959] came about, no doubt, from the general rise in real income during recent years and the corresponding shifts in consumer values and ideas of what is necessary....the standard of adequacy changes with time...."(49)

In March 1962, David Hamilton published an insightful article, "Drawing the Poverty Line at a Cultural Subsistence Level," which seems to have never received the attention that it deserved. Although he did not use the term "the income elasticity of the poverty line," Hamilton provided a perceptive description of the sociological processes underlying that phenomenon. His description is worth excerpting at some length: "As the state of the industrial arts advances, so does the minimum or subsistence standard of living essential to successful participation within any culture. New items which are a product of the advance of the industrial arts are diffused through the culture. They become a part of the standards of living. But in being adopted as a part of the standard of living they alter the way of life of the whole society and eventually become essential. In other words, items which were looked upon as luxuries initially force such changes in the way of life of a people that they eventually become necessities. The automobile is an excellent example of this effect. In American culture the automobile, shortly after its initial introduction, was viewed as a plaything for sons of the rich. But as the diffusion of the automobile proceeded it worked a change in the American way of life so that eventually what was a luxury became a technological necessity. This fact was recognized clearly during World War II when allowances of gasoline were made to owners of automobiles in order that they might get to work....All of these considerations would seem to indicate that the poverty line, that level of living which demarcates absolute poverty, is a moving one. With the advance of the industrial arts the minimum level of living essential to effective participation in a culture rises."(50) He revisited the subject in a 1968 book on poverty, noting that "The budget of items constituting the minimum level is subject to upward adjustment over time as the general level of living rises. For example, today electricity is such an essential item to our whole way of life that provision for electricity would be included in the budget. This would not be true if one were drawing up a budget demarcating poverty in 1900. Similarly, a minimum budget today would provide for a telephone, whereas in 1920 this would not have been the case. Thus, a budget defining the poverty line is not fixed for eternity, and what a budget takes as minimal in one period in our society might well be affluence in an earlier period...."(51)

In the opening paragraphs of the July 1963 article in which she presented the initial version of her poverty thresholds, Mollie Orshansky briefly alluded to the concept of the income elasticity of the poverty line: "Creature comforts once the hallmark of luxury have descended to the realm of the commonplace, and the marvels of modern industry find their way into the home of the American worker as well as that of his boss....As the general level of living moves upward and expands beyond necessities, the standards of what constitutes an irreducible minimum also change."(52)

The following two quotations from 1964 and 1965 are of particular interest because they show that during the mid-1960's, conservatives as well as liberals knew of and publicly recognized the income elasticity of the poverty line:

  • In March 1964, the Joint Economic Committee of Congress published its response to the January 1964 Economic Report of the President, which had included a chapter on poverty in connection with the Johnson Administration's announcement of its War on Poverty. In the Joint Economic Committee report, the Republican (minority) views included the following comments: "No objective definition of poverty exists. The definition varies from place to place and time to time. In America as our standard of living rises, so does our idea of what is substandard."(53)
  • In February 1965, the American Enterprise Institute published a pamphlet by Rose Friedman on defining poverty. Friedman criticized the Johnson Council of Economic Advisers' poverty definition for several reasons, and put forward a lower set of poverty lines of her own. However, it is interesting to note that despite her effort to push down the level of the poverty line, Friedman accepted the principle of the income elasticity of the poverty line, as indicated in the following quotation from her pamphlet: "If the trend in growth of real income of the past 35 years were to continue....one of its manifestations will be a rise in what is regarded as the standard of poverty....All groups will continue to share in economic progress and the people then [at the end of the twentieth century] labeled poor will have a higher standard of living than many labeled not poor today."(54)
  • In the spring of 1965, the Chamber of Commerce of the United States published a report on poverty that included a paper by Victor Fuchs entitled "Toward a Theory of Poverty"; in this paper, Fuchs became the first person in the United States to propose a relative (half-of-median-income) definition of poverty. Opening his discussion of problems of defining poverty, he wrote, "Ever since poverty has been an object of study and a focus for policy, there have been attempts to set absolute standards to identify the poor. The fact that these standards 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. Official and unofficial bodies investigating the extent of poverty have repeatedly proclaimed that they discovered what constitutes a 'minimum' or 'subsistence' or 'poverty line' budget for individuals and families of different sizes. The more candid among them have been frank to add 'relative to contemporary standards,' but this in effect means the abandonment of an absolute standard....It would appear that attempts to define poverty in absolute terms are doomed to failure because they run contrary to man's nature as a social animal. Even if 'subsistence' were defined in purely physiological terms such as 'food adequate for survival,' the definition would founder on the fact that the very concept of food is culturally and socially determined....As our nation prospers, our judgment as to what constitutes an 'insufficiency' of goods and services will inevitably change. Today's comfort or convenience is yesterday's luxury and tomorrow's necessity. In a dynamic democratic society how could it be otherwise?"(55)

In a November 1965 memorandum, Robert Ball (the head of the Social Security Administration) wrote, "Measures of income adequacy (or of poverty) change over time with the rise in general levels of living....one of the most difficult methodological questions we will have to face in the next few years is when and how to adjust the definition of poverty."(56)

In a January 1967 draft, Ida Merriam of the Social Security Administration wrote, "It is easy to observe that poverty in the U.S. today cannot meaningfully be defined in the same way as in the U.S. of 1900 or in India today. It is more difficult to project forward when and by what amounts the measure of poverty will need to be changed in the future. Yet obviously today's measure, even if corrected year by year for changes in the price level--the purchasing power of money--should not be acceptable twenty, ten or perhaps even five years hence."(57)

In a 1967 article, economist James Tobin wrote, "...it would scarcely be to the credit of this affluent society if it still adhered to Victorian norms of appropriate alms for the poor. If as our productivity has advanced we have lengthened the list of privations we think no member of our society should undergo, if we have raised our notions of minimal decent standards of life, surely this is only natural and proper. Similar escalation will happen in the future too."(58)

A section on measuring poverty in the 1967 report of the President's National Advisory Commission on Rural Poverty included the following comment: "Opinions as to where the poverty line really is, or should be, have changed as America has become more prosperous and more highly urbanized. Our standard of what is an adequate income for the poor will probably rise. Just as the poverty budgets of the 1920's by today's standards appear grossly inadequate, Americans in the year 1980 may have the same opinion of today's poverty lines."(59)

In a December 1967 professional paper, Ida Merriam of the Social Security Administration wrote, "An acceptable social minimum is obviously related to the general level of affluence of a society. In a dynamic economy it must therefore change over time. It is easy to reach agreement that what was an appropriate poverty measure in 1900 or 1933 is no longer relevant. It is also possible to get agreement that an acceptable social minimum in 1985 will be higher than today....Thus far the poverty index has been adjusted only for price changes. The need for a more substantial upward adjustment of the index level has been noted by many commentators."(60)

In May 1968, the Census Bureau published a report on poverty in the U.S. from 1959 to 1966; it was the first Census Bureau report on the subject of poverty.(61) The opening section of this report noted that the poverty line, "while adjusted each year for changes in the cost of food, is not adjusted for changes in society's standards as to the level of living that should be equated with poverty. That social standards for various levels of living do change more rapidly than the cost of a fixed amount of food is illustrated by the following comparison for 1959 and 1966 between the income required for a modest but adequate standard of living as determined by the Bureau of Labor Statistics (BLS) and the poverty income thresholds. In 1966, the poverty line for a nonfarm family of four was $3,335[;] however, according to the most recent BLS expenditure survey, it would require an income of $9,200 for a family of four to achieve a modest but adequate standard of living in most of our large cities. In 1959, the poverty line for families of this type was $3,100 while a modest but adequate budget cost approximately $7,000 (in 1966 dollars). Thus, the poverty line, as determined by the official criteria, has not maintained its earlier relationship to the modest but adequate level of living for U.S. urban families."(62) To my knowledge, this is the only explicit discussion of the income elasticity of the poverty line (even though it did not use that term as such) in any official Census Bureau publication on poverty. The discussion was probably written by Mollie Orshansky; she had commented in a 1960 article about the cost of the "modest but adequate" standard for an elderly couple rising in real terms during the 1950's (see p. 21 above), and in a March 1968 article she made a very similar comparison between the level of the poverty threshold and 1959 and 1966 median incomes for four-person families.(63)

In January 1969, President Johnson's Council of Economic Advisers published its last annual report. In a chapter on "Combating Poverty in a Prosperous Economy," the Council included the following comment: "As average incomes rise, society amends its assessment of basic needs....Consequently, an absolute standard that seems appropriate today will inevitably be rejected tomorrow, just as we now reject poverty definitions appropriate a century ago."(64)

In an early 1969 memorandum on the poverty line to the new Nixon Administration's incoming Secretary of Health, Education, and Welfare, Robert Ball (the head of the Social Security Administration) wrote, "Probably the most controversial question regarding a poverty or low-income index is the frequency and manner in which it should be changed over time. No one would argue that a poverty definition that was appropriate for the U.S. in 1900 or even in 1940 would be appropriate today. It is less clear whether a definition that was used for 1959 is still appropriate."(65)

In a 1969 article, sociologist Lee Rainwater wrote, "The emphasis on minimum standards, on 'poverty lines,' which informs so much of our thinking today, is subject to a very real embarrassment. The embarrassment is that as the nation becomes more affluent the poverty line seems to creep up. That is, each decade, the 'subsistence package' seems to involve a little bit more in the way of goods and services than it did the decade before. The implicit or explicit model used in most definitions of minimum standards revolves around the problem of inadequate diet and is therefore supposed to reflect a line that separates life and death, or at least sickness and health. The standard in a very real sense is concerned with man as an animal and with keeping him alive. Yet, even so, this standard seems to creep up imperceptibly with the decades."(66)

In January 1968, President Johnson had appointed a Commission on Income Maintenance Programs. In its November 1969 report, the Commission had several chapters on poverty and the poor. One of those chapters included the following comments: "Solely as a result of growing affluence, a society will elevate its notions of what constitutes poverty. Many factors account for this. All people tend to measure their well-being against some norm....But there are objective as well as subjective reasons for requiring poverty standards to rise with increases in the general level of prosperity. As affluence increases, community standards will constantly raise the level of income needed by the poor in order to exist. City housing codes will be upgraded, and the poor will have to improve their homes or pay more rent for their better buildings. When most of the community owns automobiles or moves to the suburbs, public transportation will probably deteriorate, leaving the poor with either inadequate or more expensive transportation....As economic advancement both allows and demands increasing education, children in families unable to provide money for higher education will fall further behind. Clearly, as a society's general standard of living rises, increasingly expensive consumption patterns are forced on the poor, not in order to catch up, but in order to remain a part of that society. Moreover, as society's normal standard of living rises[,] the poor will seek to emulate it--since they are part of society--and feel increasingly deprived if they cannot."(67)

Evidence from Britain--Quantitative Comparisons of Poverty Lines(68)

I have seen two analyses showing that absolute poverty lines in Britain have risen in real terms over time during the twentieth century; one of them explicitly stated that "there has been a tendency to raise the minimum standard as general living standards improve."(69) The poverty lines examined in these analyses all excluded any allowance for rent, following a tradition in British poverty measurement that goes back to B. Seebohm Rowntree's 1899 study of poverty in York.(70) The pre-1945 poverty lines examined (as well as Rowntree's poverty line for his 1950 study of York) were all "expert"-devised poverty budgets. For the post-World-War-II period (excluding Rowntree's 1950 study), both analyses used the National Assistance/Supplementary Benefits scale (benefit payment) rate as the unofficial poverty line, following the 1965 example of Brian Abel-Smith and Peter Townsend's The Poor and the Poorest.(71)

However, recent conceptual work on British poverty lines has shown that not all of them represent the same standard of living, as discussed below. This means that some of the specific comparisons made in the Fiegehen et al. analysis cannot be used to demonstrate the income elasticity of the poverty line. Furthermore, the "primary" poverty line of Rowntree's 1899 survey of York was used only to identify one relatively small subdivision (amounting to about one third of the total) of his total poverty population(72), whereas later British poverty lines were generally used to identify the total poverty population in the surveys in which they were used; accordingly, it would probably be inappropriate to compare Rowntree's 1899 poverty line with later poverty lines as part of a demonstration of the income elasticity of the poverty line in Britain.

In recent work on British poverty lines, John Veit-Wilson has found that these poverty lines represent not one but two different standards of living, which he designates as the "primary poverty" (P1) standard and the "Human Needs of Labour" (HNOL) standard.(73) (The HNOL standard is the higher of the two.)

  • The P1 standard is named after Rowntree's "primary poverty" measure used in his 1899 survey of York. P1 represents a standard of living at a level providing only for "merely physical subsistence"; it was recognized by Rowntree as being inadequate for even minimal normal participation in society. It included budget allowances only for food, rent (implicitly, as described in footnote 70), clothing, fuel and lighting, and other household sundries (e.g., cleaning materials). With the two exceptions noted below, all the British standard budget poverty lines were at this standard, as were the subsistence standard presented in the 1942 Beveridge report and the National Assistance/Supplementary Benefits scale rate adapted from the Beveridge subsistence standard.
  • The HNOL standard is named after Rowntree's book, The Human Needs of Labour(1918; revised edition, 1937), in which he presented a standard budget to be used in determining a proposed minimum wage for workers. HNOL represents a standard of living at a level which would allow at least minimal normal participation in society. In addition to budget allowances for food, rent (implicitly), clothing, fuel and lighting, and other household sundries, it also included an allowance for personal sundries--for such items as unemployment and health insurance contributions, travel to work, a trade union membership fee, stamps, writing-paper, a newspaper, a radio, and "all else." Rowntree adopted this standard as a poverty line for his 1936 and 1950 surveys of York.

A comparison between an earlier and a later P1 poverty line would demonstrate the income elasticity of the poverty line if the later figure were higher in real terms than the earlier one; the same would be the case for a comparison of two HNOL poverty lines. Since the HNOL standard is a higher standard of living than the P1 standard, a comparison of an earlier HNOL poverty line with a later P1 poverty line would also demonstrate the income elasticity of the poverty line if the later figure were higher in real terms than the earlier one, as discussed above on p. 10; that would be a case of a given constant-pound-sterling level being above the P1 standard during an earlier period but at or below the P1 standard during a later period. However, a comparison of an earlier P1 poverty line with a later HNOL poverty line (i.e., Rowntree's poverty line for either the 1936 or the 1950 survey of York) would not demonstrate the income elasticity of the poverty line.

The comparisons of P1 poverty lines given below are excerpted from the Fiegehen et al. analysis.(74) All figures are for a family of five persons (two adults and three children)--the "conventional" or "typical" family size figure used in poverty and family budget studies in both Britain and the U.S. during the first third of the twentieth century.(75)


Amount per week Amount as percentage (in constant (1971) Amount-- of average male Source and pounds sterling index manual earnings base year (decimalized)) (1924=100) (gross)

Bowley, 1924 7.35 100 61.0% Beveridge, 1938 9.69 132 62.8%

Natl. Assistance:

1948 8.66 118 48.6% 1953 9.80 133 52.6% 1963 12.85 175 51.8%

Supplem. Benefits:

1971 15.95 217 51.6% 1973 17.28 235 49.4%

As can be seen, the P1 poverty lines shown here rose in real terms during the period covered except between 1938 and 1948.(76) They rose about as rapidly as real male manual earnings between 1924 and 1938, failed to keep pace with earnings between 1938 and 1948, and then rose about as rapidly as real earnings again between 1948 and 1973. For the years in question, real gross domestic product per capita figures--another approximate measure of the standard of living of the general population--were as follows (in pounds sterling at 1900 prices)(77):


1924 47 1938 60 1948 66 1953 74 1963 93 1971 112 1973 121

An examination of the figures shows that (again except for the 1938-1948 period) the P1 poverty lines rose in real terms at roughly the same rate as real gross domestic product per capita.

The following additional comparisons can be made between different P1 poverty lines:

  • In 1937, R.F. George published "a new calculation of the poverty line" in 1936 prices. For a four-person family (a married couple with two children aged 10 and 4), his poverty line was 28 percent higher in real terms than the poverty line used for such a family by Arthur L. Bowley and his colleague Margaret Hogg in a 1924 study of poverty in five towns. (Poverty lines for such a family used in four other British studies between 1928 and 1931 were fairly close to Bowley and Hogg's poverty line in real terms--actually between 0.4 percent and 6 percent lower.) Most of the difference between George's poverty line and the earlier poverty lines was due to his adoption of recently released recommendations about the content and cost of minimum adequate diets, although his figure for fuel was also somewhat higher than that component in most of the earlier poverty lines.(78) George's higher poverty line or a standard essentially equal to it was used in three British poverty studies during the late 1930's.(79) The real gross domestic product per capita figure for 1936 (57 pounds, in 1900 prices) was 21 percent higher than the corresponding figure for 1924 (47 pounds).(80)
  • According to the Atkinson analysis, the unofficial poverty line (the Supplementary Benefits scale rate) for a five-person family (two adults and three children) in 1981 was between 2.3 and 4.1 times as high in real terms as the poverty lines used by Percy Ford for such families in a 1931 study in Southampton, and between 1.5 and 3.0 times as high in real terms as the poverty lines used by Herbert Tout for such families in a 1937 study in Bristol.(81) The real gross domestic product per capita figure for 1981 (129 pounds, in 1900 prices) was 2.6 times as high as the corresponding figure for 1931 (49 pounds) and 2.2 times as high as the corresponding figure for 1937 (58 pounds).(82)

The comparison of HNOL poverty lines given below is excerpted from the Fiegehen et al. analysis.(83) The figures are for a family of five persons (two adults and three children).



Amount per week Amount as percentage (in constant (1971) Amount-- of average male Source and pounds sterling index manual earnings base year (decimalized)) (1936=100) (gross) Rowntree, 1936 10.26 100 67.0% Rowntree, 1950 12.17 119 66.6%

As can be seen, the HNOL poverty line rose by 19 percent in real terms from 1936 to 1950--almost exactly as rapidly as real male manual earnings. The real gross domestic product per capita figure for 1950 (69 pounds, in 1900 prices) was 21 percent higher than the corresponding figure for 1936 (57 pounds).(84)

The following additional comparison can be made between HNOL poverty lines:

  • In 1993, Sean Stitt and Diane Grant published a book in which they presented a standard-budget-based poverty line designed to be a Rowntree-style poverty line for 1992 conditions, "constructed upon the approaches and ethos of...Rowntree." For a four-person family (two adults and two children), in 1992 prices, their poverty line was 129.78 pounds sterling a week (excluding rent, as with the other poverty lines discussed in this section).(85) They were not entirely clear about whether their poverty line was at the P1 or the HNOL standard, but since it included budgeted amounts for personal sundries as well as household sundries, it must be considered an HNOL poverty line. Rowntree and Lavers' 1950 HNOL poverty line for the same type of family (with an employed head) was 4 pounds 6 shillings 4 pence (in October 1950 prices)(86) -- or 4.32 pounds in decimalized terms. Using the official retail price index(87), 129.78 pounds in 1992 prices would be equal to 7.8 pounds in 1950 prices. This means that the Stitt/Grant 1992 [HNOL] poverty line was 1.8 times as high in real terms as the Rowntree/Lavers 1950 HNOL poverty line. The real gross domestic product per capita figure for 1992 (162 pounds, in 1900 prices) was 2.3 times as high as the corresponding figure for 1950 (69 pounds).(88)

In the context of the Stitt/Grant standard-budget-based poverty line for 1992, one should probably mention that Jonathan Bradshaw of the Family Budget Unit at the University of York published a book(89) in 1993 which presented two "budget standards" (standard budgets) developed by the Family Budget Unit--a "modest-but-adequate" budget and a "low cost" budget. However, the two budgets were described in terms of American budget-related standards, rather than in terms of earlier British budget standards. (The book noted that the "budget standards" method has been very little used in Britain since World War II.) Accordingly, it would be inappropriate to try to equate the low cost budget with the HNOL standard or to make a quantitative comparison between the real values of the low cost budget and an HNOL poverty line in an effort to provide further evidence of the income elasticity of the poverty line.


The comparison of earlier HNOL poverty lines and later P1 poverty lines given below is excerpted from the Fiegehen et al. analysis.(90) The figures are for a family of five persons (two adults and three children). The two HNOL poverty lines are identified by asterisks.



Amount per week | Real gross domestic (in constant (1971) Amount-- | product per capita Source and pounds sterling index | (in constant (1900) base year (decimalized)) (1936=100) | pounds sterling)(91) | *Rowntree, 1936 10.26 100 | 57 Beveridge, 1938 9.69 94 | 60 Natl. Assistance: | 1948 8.66 84 | 66 *Rowntree, 1950 12.17 119 | 69 Natl. Assistance: | 1953 9.80 96 | 74 1963 12.85 125 | 93 Supplem. Benefits: | 1971 15.95 155 | 112 1973 17.28 168 | 121

As can be seen, of the years shown, 1963 was the first in which the real value of the P1 poverty line exceeded the real value of the 1936 HNOL poverty line; coincidentally, 1963 was also the first of the years shown in which the real value of the P1 poverty line exceeded the real value of the 1950 HNOL poverty line. Consistent with the discussion on p. 10 above, the general standard of living in 1963 (as measured by real gross domestic product per capita) was significantly higher than in 1936 and in 1950.

The following additional comparison can be made between an earlier HNOL poverty lineand a later P1 poverty line:

  • According to the Atkinson analysis, the unofficial P1 poverty line (the Supplementary Benefits scale rate) for a five-person family (two adults and three children) in 1981 was between 1.5 and 2.3 times as high in real terms as the HNOL poverty line used by Rowntree for such families in his 1936 study in York.(92) The real gross domestic product per capita figure (in 1900 prices) was 57 pounds in 1936 and 129 pounds in 1981(93); in other words, consistent with the discussion on p. 10 above, the general standard of living in 1981 was significantly higher than in 1936.

Taken together, these various comparisons--P1-to-P1, HNOL-to-HNOL, and earlier-HNOL-to-later-P1--provide an extensive and overlapping network of evidence solidly demonstrating the income elasticity of the poverty line (or rather of both standards) in Britain over the 1924-1992 period. If the comparisons were adjusted to be more comparable in detail (e.g., by deciding on a single price index to use for the 1938-1955 period, or by using a price index excluding the housing component in all cases), details of the picture might change somewhat, but the overall conclusion would remain unchanged. An examination of the various comparisons suggests that the income elasticity of the two standards with respect to real gross domestic product per capita was relatively close to 1.0 over most of the period--but this is one of the specific details that might shift somewhat if the comparisons were adjusted for greater comparability.

To the best of my knowledge, there is no time series of poll results in which British respondents were asked a question resembling the (U.S.) Gallup Poll "get-along" question.

Evidence from Britain--the Common Knowledge of Informed Analysts

In a much quoted(94) (and sometimes misquoted) passage in The Wealth of Nations (given here from the text of the 1789 edition), Adam Smith wrote, "By necessaries I understand, not only the commodities which are indispensably necessary for the support of life, but whatever the custom of the country renders it indecent for creditable people, even of the lowest order, to be without. A linen shirt, for example, is, strictly speaking, not a necessary of life. The Greeks and Romans lived, I suppose, very comfortably, though they had no linen. But in the present times, through the greater part of Europe, a creditable day-labourer would be ashamed to appear in public without a linen shirt, the want of which would be supposed to denote that disgraceful degree of poverty, which, it is presumed, no body can well fall into without extreme bad conduct....Under necessaries therefore, I comprehend, not only those things which nature, but those things which the established rules of decency have rendered necessary to the lowest rank of people."(95) Strictly speaking, of course, Smith was defining "necessaries," not "poverty." However, his concept of necessaries at least implies a definition of "poverty" that would be based not on an unchanging biological concept of subsistence ("the commodities...indispensably necessary for the support of life"--cf. also the later reference to "nature") but on whatever "the custom of the country" or "the established rules of decency" considers necessary. The discussion of the linen shirt is a specific example of the "custom of the country" changing on the basis of the introduction of a product of technological/industrial progress. Although the pace of change involved was glacial (covering a period of roughly sixteen centuries, according to a footnote), it is nevertheless a genuine example of the sociological processes underlying the phenomenon of the income elasticity of the poverty line. It is an irony of history that those today in Britain and America who aggressively identify themselves as disciples of Adam Smith are generally opposed to definitions of poverty that are consistent with their master's definition of "necessaries."

In the 1870's--even before Charles Booth had developed the concept of "the line of poverty"--two economists presented and discussed evidence that seems to indicate that the economic "standard of comfort" of "the poorest of the poor" had risen in real terms during the nineteenth century along with the standard of comfort of the general population in England. In a lecture published in 1876, J.T. Danson wrote, "A recent parliamentary paper affords a striking illustration of the change which has occurred in what is termed 'the standard of comfort' as regards the lowest classes in this country....The pauper does not, as you know, fix for himself the style of his living. It is fixed for him by others; and the common rule is that he shall not live materially better, nor much worse, than he would do if he worked for his living as a labourer of the lowest class. Now, we find that the cost of maintaining the 100,000 paupers of London in 1875 was about five times as great as the cost of a similar number in 1815. And, on examining the accounts, it appears that the excess arises almost entirely from the difference between the ideas prevailing, at the two periods, as to what is absolutely, or humanly, necessary for the decent maintenance of even the poorest of the poor....there can be no doubt that the standard of comfort, in all classes, has risen greatly, in England, during the last sixty years...."(96) Two years later, Bonamy Price quoted much of this passage of Danson's, adding, "And the rise is not in nominal wages only...for the articles purchased by the working classes have not increased in price. Bread is cheaper than it was sixty years ago; so are sugar and tea; clothing is better and not dearer [i.e., not more expensive]; and numerous small comforts and conveniences, which materially smooth and civilize life, may be procured at trifling cost."(97) (Figures available to me do not go back as far as 1815, but do show that real gross domestic product per capita was 1.8 times as high in 1875 as in 1830.(98))

In a chapter on "Necessaries" in the first (1890) edition of his Principles of Economics, the eminent economist Alfred Marshall wrote, "Adam Smith and the more careful of his followers observed indeed variations in the standard of comfort at different times and places....every estimate of necessaries must be relative to a given place and time...."(99) (The latter sentence is often quoted or paraphrased, frequently without giving credit to Marshall.) While this statement is not an explicit recognition of the income elasticity of the poverty line, it does clearly rule out absolute poverty lines, since absolute poverty lines are based on an assumption that estimates of "necessaries" are not relative to any given time.

During the 1880's and 1890's, Charles Booth (the originator of the concept of "the line of poverty") conducted extensive studies of poverty and industrial and religious conditions in London which were published in what ultimately became a 17-volume series. The final volume (1903) included the following sentence: "Side by side with these improvements the whole level of poverty has been pressed upwards by increasing demands on life--demands which were unthought of forty, thirty, or even twenty years ago."(100) While this sentence could be interpreted in several ways, it suggests at least the possibility that Booth had some awareness of the income elasticity of the poverty line.(101)

In 1925, statistician Arthur L. Bowley and his colleague Margaret Hogg published a study(102) in which they compared poverty rates in five towns in 1923-1924 with poverty rates for the same towns in 1912-1914 as determined by an earlier study(103) published in 1915 by Bowley and another colleague. In the 1925 study (the second "Five Towns" study) Bowley and Hogg used the same poverty line as in the 1915 study, adjusting it only for price changes.(104) Bowley also developed the poverty line used in the 1928-1930 New Survey of London Life and Labour(105) (see next paragraph); this New Survey poverty line was intended to be equal--after adjustment for price changes only--to the poverty line used by Booth in his original survey four decades earlier.(106) To the best of my knowledge, Bowley and Hogg did not explicitly discuss in the second "Five Towns" study the issue of whether the poverty line should be adjusted over time for anything other than price changes.(107),(108) In connection with the new London survey, by contrast, Bowley made several comments suggesting that a poverty line adjusted only for price changes over an extended period of time would be an unsatisfactory measure of the social concept of poverty at the end of the time period. In a discussion of a 1929 paper by Sir Hubert Llewellyn Smith (see next paragraph), Bowley said that he wanted "to emphasize what had already been said about the rising [poverty] standard, in the sense that people now regard as poor those whom in a former enquiry would have been regarded as comfortable. No statistical investigation on this point was being made, and he [Bowley] was merely collecting opinions; but it would be found that there was very definite evidence of the rise in the standard of living and diminution in poverty, simply from the fact that the popular conception of what constituted poverty had changed."(109) In one of the volumes of the report on the New Survey, Bowley wrote, "...with the general rise in the standard of living in the generation before the war [World War I], and the further rise at least in unskilled labour ranks since 1914, many people would set the minimum [standard--i.e., the poverty line] higher."(110),(111) And in a brief 1936 note about the New Survey poverty line, Bowley wrote, "...the opinion as to what constitutes poverty has changed in fifty years [since Charles Booth's survey], and it is desirable to draw up a new definition for future use."(112)

In 1928-1930, the London School of Economics conducted a New Survey of London Life and Labour, one of the goals of which was to compare the level of poverty in London with the results of Charles Booth's survey four decades earlier. As indicated in the previous paragraph, the New Survey used a poverty line developed by Bowley which was intended to be equal (after adjustment for price changes only) to Booth's forty-year-old poverty line. Sir Hubert Llewellyn Smith, director of the New Survey, stated even more clearly than Bowley that an old poverty line adjusted only for price changes would fall below the level of the contemporary social concept of poverty. In a 1929 paper giving a preliminary overview of the New Survey, Llewellyn Smith wrote, "...to obtain results comparable with those of the Booth Survey, we find ourselves engaged in an uphill fight against a strong natural tendency, on the part of the persons interviewed, to replace Booth's poverty standards by those which accord more closely with the changed ideas of the present day, as to the necessary minimum requirements of a civilized existence. There is abundance of indication that current ideas as to the meaning of 'poverty' and the position of the 'poverty line' have been considerably modified during the past forty years. Probably it would be not far from the truth to say that, at any given time, most people visualize 'poverty' as arising when family purchasing power falls materially below that resulting from the current income habitually obtained by unskilled labour."(113) In the third volume of the report on the New Survey, discussing the decision to use a fixed-constant-shilling poverty line, Llewellyn Smith wrote, "...in proportion as there is a change in current conceptions of what it means to be 'poor'[,] the standard of poverty originally chosen will diverge from the standard which corresponds at any given moment with these changing conceptions. Since the 'real' purchasing power of the average London workman has risen in forty years by a third, it is unlikely that a standard of poverty based on the conditions and conventions of forty years ago would commend itself to the public opinion of to-day....It cannot...be too clearly emphasized that there has been no attempt to fix a level of present-day 'poverty' [i.e., a poverty line] according to present-day ideas. The sole aim has been to apply Charles Booth's standard to present economic conditions."(114) And in the sixth volume of the report on the New Survey, Llewellyn Smith wrote, "During the past forty years concurrently with the rise in the standard of life, there has been a change in the prevailing view as to what constitute the minimum necessaries of a civilised existence."(115)

In a 1929 book on the economics of inheritance, Josiah Wedgwood wrote, "...in the search for material welfare, our modern civilisation under conditions of industrial progress is continually manufacturing new and previously unwanted sources of pleasure, so that the old luxuries become the new necessities, alike for those who can and those who cannot afford them....Thus a continually increasing amount of income becomes necessary in order to produce the same degree of material welfare. Though the amount of goods and services enjoyed by the poor man in 1924 may be enormously greater than those enjoyed by his predecessor in 1824, the former's poverty is probably little less tedious and unpleasant to him than an actually more grinding poverty was to the latter."(116)

In 1929-1932, the University of Liverpool conducted a Social Survey of Merseyside [greater Liverpool]. In a discussion of a 1931 paper presenting some of the results of the survey, two people made comments about the poverty line. Henry Clay said, "The fact that all the Poverty Lines seem to bear a close relation to the wage of unskilled labour in the country in which they are made makes one doubt the scientific value of the dietetic data on which the line is based." Sir Josiah Stamp (the President of the Royal Statistical Society) commented, "What was called a subsistence level was not absolute at all, but relative to one's own ideas....If one delved into the literature of [William] Paley [1743-1805] and his school, many illustrations would be found of things which he regarded as above subsistence level which to-day were considered to be below it. The subsistence level changed and would continue to change...."(117) In the 1934 final report on the Merseyside survey, the section on the poverty line included the following discussion: "If we are to estimate the number of families living in poverty it is evident that we must first agree as to what is to be meant by poverty....[I]n order to make the desired estimate it is clearly essential to decide what are the 'necessaries' of life, a decision which depends entirely upon the accepted standard of living. The standard adopted is bound to be arbitrary, but it should bear a reasonable relation to facts. The relevant facts, however, are not static. They vary--and with them the standard should vary--from country to country, from time to time within the same country, and from class to class at the same time in the same country. In other words, the poverty line will be drawn at different levels according to the standard of living prevailing at different places, at different times, and in different classes. The minimum sum required to feed an Indian peasant is considerably less than that required to feed an English agricultural labourer; the British workman's standard of living to-day is appreciably higher than it was one hundred years ago....All we can say is that a certain sum of money, if properly spent, would purchase a bare sufficiency of such necessaries as food, clothing, fuel, light and cleaning materials, according to present-day standards of consumption, for a family of a defined class and constitution."(118)

As noted above (p. 30), a comparison of the real levels of Rowntree's 1899 P1 poverty line and his 1936 HNOL poverty line cannot be used as evidence of the income elasticity of the poverty line. However, the following comments that Rowntree made in connection with his survey of York in 1936 (actually in 1935-1936) do serve to illustrate the changes in social standards over time that underlie the income elasticity of the poverty line: "...ideas of what constitutes 'obvious want and squalor' have changed profoundly since [1899]. There is no doubt that in 1899 investigators would not have regarded as 'obvious want and squalor' conditions which would have been so regarded in 1936, and on the other hand a large proportion of the families living below the 1936 poverty line would not in 1899 have been regarded as 'showing signs of poverty'."(119) Discussing the same topic in more specific terms, one of his research workers later wrote, "the second-hand boots and jacket which were adequate in 1899 would not have been entertained by the worker of 1935; a bath and a garden, absent in almost every working class dwelling in 1899, were by now [by the 1930's] considered essential."(120)

In a 1938 report on a social survey of Bristol, Herbert Tout included the following comment: "Ideas as to what constitutes poverty vary from generation to generation. What we consider squalor to-day might have passed as something less fifty years ago."(121)

In his famous 1942 Report, in a discussion of provisional rates of benefit proposed for social insurance, Sir William Beveridge wrote, "...determination of what is required for reasonable human subsistence is to some extent a matter of judgment; estimates on this point change with time, and generally, in a progressive community, change upwards."(122)

In a brief 1954 article on the poverty line--one of a six-part series on poverty and social welfare in industrial Lancashire--Norman MacKenzie wrote, "It is, of course, much easier to say that there should be a 'national minimum' than to say what it should be. In the first place, any definition of subsistence is bound to be relative. Tastes and resources change...." Perhaps with some degree of hyperbole, he continued, "...the growth of the national income and the development of the social services provide amenities for a contemporary pauper that, a generation ago, even a millionaire could not buy."(123)

In a 1963 article, sociologist and poverty researcher Peter Townsend wrote that "although there are certain universal human needs, such as for food, shelter, warmth and care in illness, they differ enormously in degree and character according to time and place. The minimum standards grudgingly conceded by one generation can be quickly outgrown and become unacceptable to the next."(124)

In a 1968 social sciences encyclopedia article on poverty, British economic and social historian Eric Hobsbawm wrote, "...in economically developing and richer countries a conventional minimum, tending to rise with time, must and always does determine poverty."(125)

In a 1978 report on "lower" incomes, the Royal Commission on the Distribution of Income and Wealth included the following comments: "A low income standard, which in normal parlance would be generally equated with poverty, is now generally accepted as a standard which changes with the general standard of living of society; and therefore the fixing of minimum needs cannot be absolute or immutable....People who today are regarded--or who regard themselves--as poor are not necessarily so by the standards of earlier periods...."(126)

In February 1983, Market and Opinion Research International (MORI) conducted a survey of a nationwide sample of 1,174 persons in Britain on behalf of London Weekly Television. The survey may be accurately described as an effort to identify "necessaries" according to "the custom of the country." Respondents were given a list of 35 items--not only consumption items but also common social activities--and were asked if each item was one "which you think [is] necessary, and which all adults [or families with children] should be able to afford and which they should not have to do without...." An August 1983 article briefly reporting on the results of the survey stated that "Few of the items [that] a majority of people [in the survey] now consider 'necessary' would have appeared in a list of subsistence needs drawn up by Charles Booth [for his 1880's study of poverty in London]."(127) A 1985 book reporting on the survey included the following comments: "This widespread consensus [in the survey] on what are necessities clearly reflects the standards of today and not those of the past....The importance of viewing minimum standards in terms of contemporary living conditions is highlighted most forcefully by the impact of labour-saving household goods. A large majority of people think that a refrigerator and a washing machine are necessities - items that were unknown to the Victorians and even twenty years ago would have been seen as a luxury. In part, this reflects shifting standards and expectations; but it also reflects the fact that, in a practical sense, items that become customary also become necessary because other aspects of life are planned and built on the very fact that these items are customary. For example, many single elderly people have commented to us that, whereas once they could manage without a [refrigerator], it is now so difficult to buy perishable food in small quantities that they find they need [a refrigerator]."(128)

In 1989, Conservative Party politician John Moore gave a lecture denouncing present-day unofficial relative definitions of poverty in Britain. In the course of this lecture, perhaps not focusing on the subtle inconsistency, he commented that "...there is bound to be disagreement over which items are essential to a decent standard of living. The judgement changes over time. The diet underlying the work carried out by [B. Seebohm] Rowntree at the turn of the century would certainly now be regarded as inadequate."(129)

Evidence from Ireland--the Common Knowledge of Informed Analysts (One Quotation)

In a 1989 article, four prominent Irish poverty researchers wrote, "It is widely accepted that poverty has to be seen in the context of the actual society in question, and the standard of living generally considered adequate there. What is considered adequate will thus change over time and differ across countries--we do not apply the standards of the 1880s to the 1980s, or of Switzerland to Greece. Even what many currently conceive as 'absolute' needs, when closely examined, turn out to reflect prevailing standards--or perhaps those of 10-20 years ago--rather than invariant requirements for subsistence."(130)

Evidence from Canada--Quantitative Comparisons of Standard Budgets and Poverty/Low Income Lines

I do not know of any published study which examines changes in the real value of Canadian subsistence budgets over time and compares them with changes in the real average income or general standard of living of Canadian families. However, the analysis given below of data on several Canadian standard budgets and poverty/low income lines does provide some confirmation of the income elasticity of the poverty line in Canada since 1896.

(The real values of the pre-1965 budgets and poverty lines will be compared with the original low income cut-offs developed by Jenny Podoluk of the [Canadian] Dominion Bureau of Statistics (now Statistics Canada) in 1965. Revisions of these low income cut-offs continue to be used by Statistics Canada on an unofficial basis to identify the low income population in annual reports.)

In 1897, Herbert Brown Ames published a sociological study of conditions in certain working-class areas of Montreal.(131) The primary social reform that he advocated was model working-class dwellings funded by private investors at a limited rate of profit ("Philanthropy and 5 per cent"). He classified the population in his study area into three groups: the "well-to-do," the "real industrial class," and the "poor" or the "submerged tenth." The poor were identified as those with incomes (in 1896) below $C5.00(132) a week or $C260.00 a year. (Ames referred to this amount at one point as "the limit of decent subsistence.") The well-to-do were identified as those with incomes above $C20.00 a week or $C1,000 a year.(133),(134) There is no indication that either of these income lines was derived from a standard budget; neither of them was adjusted for family size. (In the summary article in the Bulletin of the Department of Labor, Ames divided the population into six income classes, with the well-to-do being class A and the poor being class F.(135)) It seems reasonably safe to conclude that Ames' poverty line was essentially Charles Booth's London poverty line converted into Canadian dollars(136), even though Ames did not mention Booth by name in his pamphlet.(137) The income range most commonly cited in connection with Booth's "line of poverty"--"18s [shillings] to 21s per week for a moderate family"(138)--was equivalent to a range of from $C4.38 to $C5.11 at the then-official exchange rate of $C4.8667 to 1 pound(139); Ames' $C5.00 was the only integral dollar figure in the $C4.38-5.11 range. (Ames' six income classes designated by letter are an additional point of similarity with Booth, who divided his population into eight letter-designated classes on the basis of income level, regularity of wages, and social class.(140)) Ames' $C260 poverty line (in 1896 dollars) would be very roughly(141) equal to about $C970 in 1961 dollars--24 percent of Jenny Podoluk's (1959-based)(142) low income cut-off of $C4,000 for a family of five or more persons.(143) Ames' $C1,000 "well-to-do" income line would be very roughly equal to about $C3,700 in 1961 dollars--about 92 percent of Podoluk's low income cut-off for a family of five or more.

As noted above, there is no indication that either of Ames' income lines was derived from a standard budget. Indeed, according to 1926 testimony by a Canadian labor union researcher, hardly any standard budgets were developed in Canada during the first quarter of the twentieth century.(144) The only Canadian "commodity budget" which the researcher's labor union could find was one published by Arthur Martel in 1922.(145) By developing the budget, Martel, vice-president of the Trades and Labor Congress of Canada, hoped to help "eliminate strikes and lockouts, bring about a better understanding between employer and employe, and thus establish a fair return on invested capital, and a decent living wage...." Martel described the budget as "a minimum standard of living for Canadian citizenship," but he essentially equated it with a contemporary Philadelphia workingman's budget which Americans Dorothy and Paul Douglas classified at their "subsistence plus" or "minimum health and decency" level(146)--one level above their "minimum of subsistence" [=poverty] level. Furthermore, the dollar cost of this Philadelphia budget was slightly above Ornati's minimum adequacy level for 1921(147)--one level above his minimum subsistence [=poverty] level. Accordingly, following the reasoning given in the next paragraph, it seems reasonable to consider the Martel budget as a 1920's Canadian living standard definitely above minimum subsistence [=poverty]. The cost of the Martel budget for a five-person family at March 1921 prices was $C1,744.38. This figure would be equal to about $C2,740 in 1961 dollars--68 percent of Podoluk's (1959-based) low income cut-off of $C4,000 for a family of five or more persons.

In 1926 Canadian Parliamentary hearings on a minimum wage resolution, three American standard budgets repriced in Canadian terms (and with minor modifications to reflect Canadian conditions) were submitted as evidence. Two of the budgets represented Canadian repricings of budget levels developed by Americans Dorothy and Paul Douglas; they were prepared by a [Canadian] Labour Department statistician in response to a request from members of the Parliamentary committee holding the hearings. The other budget was the U.S. Bureau of Labor Statistics' 1920 Minimum Quantity Budget for a workingman's family of five (see p. 11 above), priced in Canadian terms and with minor modifications for Canadian conditions; it had been used by a Canadian labor union in a wage arbitration, and a researcher for the union discussed it in the context of budget studies. (In both the contemporary Douglas classification and the 1966 Ornati classification mentioned above, the 1920 BLS budget was two levels above the minimum subsistence [=poverty] level.) Both the Parliamentary committee members and the Canadian labor union accepted the American budgets as being generally applicable to Canadian conditions.(148) Accordingly, the Canadian cost of the U.S. BLS budget can be accepted as a reasonable approximation of a 1920's Canadian living standard considerably above minimum subsistence [poverty], while the Canadian cost of the Douglas minimum subsistence budget can be accepted as a reasonable first approximation of a 1920's Canadian minimum subsistence [poverty] living standard. The Canadian cost (annual) of the minimum subsistence budget for a five-person family was $C1,396.92, presumably in 1926 dollars. The Canadian cost of the 1920 BLS quantity budget for a five-person family was $C2,202.37, in 1925 dollars.(149) The Canadian cost of the minimum subsistence budget would be equal to about $C2,370 in 1961 dollars--59 percent of Podoluk's (1959-based) low income cut-off of $C4,000 for a family of five or more persons. The Canadian cost of the 1920 BLS budget would be equal to about $C3,790 in 1961 dollars--95 percent of Podoluk's low income cut-off for a family of five or more.

In 1928 Canadian Parliamentary hearings on the question of unemployment insurance, the Secretary of the Financial Federation of the Montreal Council of Social Agencies presented a two-year-old standard budget for the "absolute minimum necessary" expenditures of a working-class family of five persons. The budget was prepared by a committee of four women and one man (the chairman). The budget study was initiated because "we questioned...whether all the problems of sickness, poverty, and delinquency were not in part, at least, attributable to insufficiency of income." The budget was described as "ultra-conservative...represent[ing] a scale below which no family could maintain its industrial efficiency or social normality." It excluded all expenditures for medical and dental care, life insurance, Christmas or birthday presents, union dues, church and charity, books, magazines,postage, stationery, amusements, tobacco, candy, and the replacement of household china, tinware, towels, or bedding. From the way in which this budget was described (and because its cost was even lower than the cost of the Douglas minimum subsistence budget priced in Canadian terms as discussed in the previous paragraph) it seems fair to characterize this budget as a poverty or minimum subsistence budget. In 1926 prices, the cost of this budget for a family of five was $C1,101.76.(150) This figure would be equal to about $C1,870 in 1961 dollars--47 percent of Podoluk's (1959-based) low income cut-off of $C4,000 for a family of five or more persons.

In 1939 the Welfare Council of Toronto published a standard budget embodying a standard of living which "should maintain"--or "cover the bare essentials for"--" health and self-respect." The Council undertook the study to provide social agencies with a standard to measure the adequacy of income, and analyze how it was spent; to give "[p]rogressive employers" an assessment of what money wages would provide; and to provide information to the general public to help them "judge the wisdom and effect" of proposed legislation to set a minimum wage for men.(151) Social scientist Leonard Marsh used this budget in a discussion of minimum standards in his 1943 Report on Social Security for Canada. He described it as a "'living-wage' budget" and a "Desirable Living Minimum," and further characterized it as a "line at which there would be certainty over a long period for better than subsistence standards...." He also assessed the standard of living of the budget by comparing the cost of its food component with the costs of the U.S. Bureau of Home Economics' four food plans(152), repriced in Canadian terms; he found the Toronto budget's food component closest in cost to the U.S. "adequate diet at minimum cost." (153) Note that the adequate diet at minimum cost was used as the food component of Margaret Stecker's [U.S.] maintenance budget, and that the cost of that maintenance budget was close to the 1935 figure for Oscar Ornati's minimum adequacy level--one level above his minimum subsistence [=poverty] level.(154) Because of the preceding descriptions of the Toronto budget and the food component comparison (and because Marsh cut the budget back to get a lower standard for which he specifically used the term "subsistence"--see next paragraph), it seems fair to characterize the Toronto budget as representing a standard of living higher than minimum subsistence or poverty. In 1939 prices, the cost of this budget for a family of five was $C28.35 a week (155), or $C1,474.20 a year.(156) This figure would be equal to about $C3,010 in 1961 dollars--75 percent of Podoluk's (1959-based) low income cut-off of $C4,000 for a family of five or more persons.

Besides using the 1939 Toronto "health and self-respect" budget in his 1943 report, Marsh also used a lower budget or standard which he developed by cutting back the "health and self-respect" budget.(157) He described it as an "Assistance Minimum" standard and as "restricted," and as having been developed "on a subsistence basis...." While indicating that the Desirable Living Minimum ought to be used "if either wages or [social security] benefit payments are not to raise any deficiency problems in the course of a long period," he developed the Assistance Minimum for situations which called for "concessions...to the possibilities of restricted [social security] provision in emergency periods, to the areas of sub-marginal incomes already existent, or to the exigency of keeping down the gross contributions for a comprehensive national [social security] scheme....The lower standard must be regarded as conceded rather than recommended: a level which it is desirable to raise." He did not describe the components of the lower standard except to say that, compared with the Desirable Living Minimum, "restrictions...include more crowded housing accommodation and the absence of any allowance for advancement expenditures or savings at all...."(158) In view of the preceding descriptions of the Assistance Minimum standard, it seems fair to characterize it as a poverty or minimum subsistence standard. In 1939 prices, the cost of the Assistance Minimum standard for a family of five was $C88.36 a month(159), or $C1,060.32 a year. This figure would be equal to about $C2,160 in 1961 dollars--54 percent of Podoluk's (1959-based) low income cut-off of $C4,000 for a family of five or more persons.

The various Canadian poverty/low income or subsistence standards reviewed in this section can be summarized as shown below. (All figures are for five-person families.)



                                                     Real consumer

Amount per year Amount-- expenditure per capita Source and (in constant index (in constant (1957) base year (1961) dollars) (1961=100) dollars)(160) Ames, 1896 $C 970 24 NA (Douglas), 1926 2,370 59 $C 716 Montreal, 1926 1,870 47 716 Marsh, 1939 2,160 54 780 Podoluk, 1961 4,000 100 1,253

It seems reasonable to assume that real consumer expenditure per capita in 1896 was lower than in 1926, although I was unable to find figures to verify that. The picture given by these figures is somewhat complicated by the presence of two budgets at different levels for one year, 1926. (Ornati dealt with the same problem with U.S. budgets by either selecting only one budget for a given year or averaging two budgets to get a single figure for a given year.(161)) Despite these problems, the figures do show a general picture of poverty or subsistence standards rising in real terms over periods during which the real standard of living of the general population rose.

The various Canadian above-poverty standards discussed in this section do not necessarily all represent the same standard of living, but they do represent standards that were considered to be noticeably above poverty or subsistence at the time they were developed. In constant dollars, Ames' "well-to-do" income line for 1896 was roughly equal to 92 percent of Podoluk's original low income cut-off. Martel's budget for 1921 at the "health and decency" level was equal to 68 percent of Podoluk's low income cut-off. The 1925 Canadian-priced version of the U.S. Bureau of Labor Statistics Minimum Quantity Budget was equal to 95 percent of Podoluk's low income cut-off. The Toronto "health and self-respect" budget for 1939 was equal to 75 percent of Podoluk's low income cut-off. Although I was unable to find real consumer expenditure per capita figures for 1896, 1921, and 1925, it seems reasonable to assume that if they did exist, they would be lower than the figure for 1961. On this assumption, the budget figures in this paragraph provide examples of constant-dollar levels that were considered well above poverty/subsistence during an earlier period (of lower general real incomes) but were considered roughly equal to or even below poverty/low income during a later period. As discussed above on p. 10, this constitutes further evidence of the income elasticity of the poverty line.


Jenny Podoluk of the Dominion Bureau of Statistics (now Statistics Canada) presented her original low income cut-offs in an unpublished paper in 1965 and in a Census Monograph published in 1968.(162) Statistics Canada has subsequently revised the Low Income Cut-Offs (LICOs)(163) several times based on data from successive Surveys of Family Expenditures (FAMEX).(164) It should be noted that because of these revisions, the LICOs themselves exhibit income elasticity over time, as indicated below:

  • In 1973, when revised LICOs calculated on the basis of consumption data from the 1969 FAMEX were introduced, these 1969-based LICOs were higher in real terms than the original 1959-based LICOs, with the average(165) figure for a family of four being 18 percent higher; this reflects the rise in the real income of the general population (and consequent changes in consumption patterns) from 1959 to 1969. (The real average [mean] income of Canadian families (in 1993 dollars) rose from $C27,677 in 1959 to $C38,834 in 1969(166)--an increase of 40 percent.)
  • When 1978-based LICOs were introduced in 1982, they were higher in real terms than the 1969-based LICOs, with the average figure for a family of four being 13 percent higher; this reflects the rise in the real income of the general population (and consequent changes in consumption patterns) from 1969 to 1978. (Real average family income (in 1993 dollars) rose from $C38,834 in 1969 to $C52,340 in 1978--a net increase of 35 percent.)
  • When it was decided not to revise the LICOs in 1986 (based on the fact that tentative 1982-based LICOs would have been very little changed from the 1978-based LICOs), the small amount of change in the LICOs reflected the fact that the real income of the general population was about the same in 1982 as in 1978, due to recession. (Real average family income (in 1993 dollars) was $C52,340 in 1978 and $C52,173 in 1982--a net decrease of 0.3 percent.)
  • When 1986-based LICOs were introduced in 1991(167), they were higher in real terms than the 1978-based LICOs, with the average figure for a family of four being 6 percent higher(168); this reflects the net rise in the real income of the general population (and consequent changes in consumption patterns) from 1978 to 1986. (Real average family income (in 1993 dollars) was $C52,340 in 1978 and $C53,792 in 1986--a net increase of 3 percent.)
  • In 1993, 1992-based LICOs were introduced.(169) An average figure for a family of four for these new LICOs is not available to me. However, it appears that if such a figure were available, it would probably show relatively little change in real terms from the corresponding 1986 figure; for the five size-of-area-of-residence categories, in real terms, the 1992-based four-person LICO figure was slightly higher (by between 0.4 percent and 2.7 percent) for four categories and slightly lower (by 1.2 percent) for one category.(170) The small amount of change in the LICOs reflected the fact that the real income of the general population was only slightly higher in 1992 than in 1986, due to recession. (Real average family income (in 1993 dollars) was $C53,792 in 1986 and $C54,657 in 1992--a net increase of 2 percent.)

In 1992, the conservative Fraser Institute (located in Vancouver, Canada) published a book(171) by Christopher Sarlo that put forward a poverty line considerably lower than Statistics Canada's (1978-based) LICOs. Sarlo strongly denounced the relative approach (which he also pejoratively labeled the "social amenities approach") to measuring poverty, instead favoring an absolute approach, and argued that poverty should be defined as the lack of basic physical necessities--items whose absence is likely to compromise long-term physical well-being. He used a standard budget approach to develop his poverty line, including food, shelter, clothing, transportation, personal hygiene items, household cleaning/maintenance items and furnishings, and basic telephone service. He did not include a budget item for health care on the grounds that health care needs of the poor would be met by Canadian provincial government health plans, plus charity care for two specific services. He did not include a budget item for secondary school non-textbook supplies--not on the grounds that they were not a necessity, but on the basis of the somewhat anomalous argument that the "relatively minor costs of such items is assumed to be offset by part-time or summer earnings." He also did not include budget items for radios, televisions, VCRs, newspapers, magazines, alcohol, tobacco, children's toys, books, and writing materials. His overall poverty line for a family of four for 1988 was $C13,140.(172)

Sarlo's 1988 four-person poverty line would be equal to about $C12,100 in 1986 dollars--58 percent of Statistics Canada's overall weighted average 1986-based LICO of $C20,724 for a family of four.(173) Furthermore, Sarlo's 1988 four-person poverty line would be equal to about $C2,890 in 1961 dollars--83 percent of Podoluk's (1959-based) low income cut-off of $C3,500 for a family of four. Because Sarlo set a poverty line significantly lower (in real terms) than any of Statistics Canada's LICOs, he may be classified as what I have elsewhere(174) termed an "anti-advocate"--someone whose "goal [is] to push the level of the poverty line (or budget) below a currently accepted level." In light of this, it is quite interesting that the person who wrote the foreword to Sarlo's book (pp. xiii-xiv) was Rose Friedman(175)--one of the Americans whom I specifically classified as an "anti-advocate" in my earlier paper because of her 1965 pamphlet putting forward a low set of poverty lines.

While Sarlo's poverty line is lower in real terms than even the earliest of Statistics Canada's LICOs, it is higher in real terms than pre-World-War-II Canadian poverty/subsistence standards. This is best demonstrated by using Sarlo's poverty line for a five-person family (the family size taken as typical during the pre-World-War-II period). Sarlo gave five-person poverty lines for individual provinces, but not an overall Canada-wide average figure. However, based on an examination of his four- and five-person provincial poverty lines and his overall four-person poverty line(176), it seems safe to conclude that his overall 1988 poverty line for a five-person family would be very close to $C14,700--an amount equal to about $C3,230 in 1961 dollars (81 percent of Podoluk's original low income cut-off for a family of five or more). Sarlo's five-person poverty line would thus be 50 percent higher in real terms than Marsh's Assistance Minimum standard in his 1943 report ($C2,160 in 1961 dollars, as noted above), 73 percent higher than the Montreal minimum budget for 1926 ($C1,870 in 1961 dollars), 36 percent higher than the Canadian-priced Douglas minimum subsistence budget for 1926 ($C2,370 in 1961 dollars), and about 233 percent higher than Ames' poverty line for 1896 (about $C970 in 1961 dollars). Over a long enough period of time, then, even the extremely low poverty line of an "anti-advocate" can serve as an example of the income elasticity of the poverty line.

Evidence from Canada--A Quantitative Study of Public Opinion Poll Responses

There is some evidence suggesting that Canadian responses to a public opinion poll "get-along" question tend to rise in real terms when the real income of the general population rises. In a 1989 article(177), Alex Michalos presented responses to the Canadian Institute of Public Opinion poll question, "Generally speaking, what do you think is the least amount of money a family of four--husband, wife and two children--needs each week to get along in this community?"(178) He presented data for the period 1973-1985, during which the question was asked twelve times. He compared the annualized "get-along" amounts with median family income data from Statistics Canada; his (current-dollar) figures indicated that the "get-along" amount showed some tendency to drop as a proportion of median family income (a finding similar to that of Rainwater during the 1970's from U.S. data--see p. 13 above). He also presented data from Statistics Canada showing that during this period, real median family income rose from 1973 to 1980, dropped from 1980 to 1984, and rose from 1984 to 1985. He did not convert the "get-along" amounts to constant dollars. However, it is possible to use Canadian Consumer Price Index figures to perform this conversion. The resulting real "get-along" amounts fluctuate during the 1973-1980 period but reach a peak in 1980; they then drop from 1980 to 1983 and rise from 1983 to 1985. This suggests that the real "get-along" amount has some tendency to rise when real income rises, and to fall when real income falls.(179) An analysis regressing real "get-along" amounts against real median (or mean) family income would be helpful--especially if "get-along" data for years before 1973(180) exist and could be included.

Evidence from Canada--the Common Knowledge of Informed Analysts

In the unpublished 1965 paper in which she first presented her low income cut-offs, Jenny Podoluk included the following comments: "....The decisions as to what the components of such budgets [standard family budgets of various types] should be, are then, to a considerable extent, subjective decisions, not objective ones, and are very much influenced by the current overall level of living of the rest of the population. As a result, opinions as to what constitutes need change as real incomes in the country rise....The subjective element is evident today in some of our present views as to what are symptoms of poverty [in Canada] in 1965. For example, the lack of bathrooms, running water and central heating in rural housing have been cited as evidence of poor housing. But in 1941, half of Canadian homes had no installed baths or showers; this ratio was nearly 25 per cent in urban areas; 45 per cent of homes had no inside toilets. I am only referring to these statistics to indicate that as recently as 25 years ago amenities which we now consider to be needs were not available to nearly half of Canadian households...."(181)

In November 1968, the Senate of Canada constituted a Special Senate Committee on Poverty to investigate and report on the problem of poverty in Canada and to make recommendations for more effective remedial measures. The opening chapter of the Committee's 1971 report discussed the relativity of poverty as follows: "Poverty is always relative to a given time and place. The differences between Canadian and Asian poverty do not make the former any more tolerable. The poor in Canada are judged, and judge themselves, relative to the general situation in their own country, at any given point in time. They are not comforted or helped by reminders of their apparent affluence compared with the abysmally low living standards of the Asiatic or Latin American poor. Furthermore, the Canadian poverty level of 1971 is not the same as that of 1961, because the general standard of living has continued to rise over the past ten years. What we define as 'poverty' must change constantly in relation to general living standards...."(182)

In 1982, as noted above, Statistics Canada replaced Low Income Cut-Offs based on the 1969 FAMEX with a new set of Low Income Cut-Offs based on the 1978 FAMEX. A description of this revision in the annual income data report included the following comment: "...so long as levels of living increase in general, one expects that the number of families earning below the fixed (in real terms) [low income] cut-offs will decrease. In fact, over a long period of time the number must almost necessarily decrease to zero (in the extreme, consider the number of families [today] below almost any unrevised low income cut-off established in the days of Charles Dickens)...."(183)

In 1991, John Evans (co-author of the Statistics Canada discussion paper cited in footnote 3) presented a paper at a poverty measurement conference in Poland in which he summarized much of the material in the discussion paper and described the Low Income Review for which the discussion paper had been written. In his conference paper, Evans wrote, "Historically, it has been found that [low income] measures designed to be at [the] subsistence level always depend on...contemporary standards of living."(184)

Evidence from Australia--Quantitative Comparisons of Standard Budgets and Poverty Lines

I have seen descriptions of only two standard budgets constructed for Australian families:

I do not know whether any consistent series of Australian subsistence-level budgets exists--let alone whether anyone might have studied changes in the real value of such budgets over time and compared them with changes in the real income of Australian families. However, using the two standard budgets mentioned above and the unofficial but widely used Australian poverty line (known as the Henderson poverty line(188)), it is possible to make two comparisons involving a given constant-pound/dollar(189) level which was considered to be well above poverty during an earlier period but was considered to be below poverty during a later period.

The Royal Commission on the Basic Wage was appointed by Australian Prime Minister W.M. Hughes in December 1919 "to inquire into the cost of living in relation to the minimum or basic wage....to ascertain what is a fair basic wage," and to develop a means for adjusting the amount of the basic wage in response to price changes. The concept of the basic wage (sometimes referred to as a living wage or minimum wage) for all male adult workers had been part of Australian federal and state wages policy since the opening years of the twentieth century; the concept was not very clearly defined at first, but in 1907 Justice H.B. Higgins of the Commonwealth Arbitration Court ruled that the standard for "fair and reasonable remuneration" for employees should be "the normal needs of the average employee, regarded as a human being living in a civilized community." (It was assumed that the average employee's needs would include supporting a family of "about five persons.")(190) To carry out its charge, the Royal Commission developed a budget for a family of five persons--a [working]man, his wife, a son aged 10 1/2, a daughter aged 7, and a son aged 3 1/2. The budget was priced as of November 1, 1920. The Commission referred to its budget as being "according to reasonable standards of comfort"--a phrase quoted from its instructions. It developed its own budget rather than adapting another budget; however, the Commission did explicitly state that the standard of living represented by its budget was essentially the same as the standard represented by the U.S. Bureau of Labor Statistics' 1919 Tentative Quantity and Cost Budget and 1920 Minimum Quantity Budget for a workingman's family (see p. 11 above). The Commission also refused an employers' request to develop its budget specifically for "the humblest worker" or the "lowest-paid employee," instead developing a budget applicable to all wage-earners.(191) Because of the direct comparison with the two American budgets which were two levels above the minimum subsistence [=poverty] level and because of the refusal to develop a budget for the "humblest" or "lowest-paid" worker, it seems safe to assume that the Commission's budget represented a standard of living noticeably above the poverty/subsistence level.

The 1919-1920 Commission's (weekly) budget cost for a family in Melbourne was 5 pounds, 16 shillings, and 6 pence. (Melbourne is chosen for this comparison because the Henderson poverty line was developed in connection with a 1966 survey of that city.) That figure would be equal to about $A31.20(192) in 1966 dollars--82 percent of the 1966 Henderson poverty line of $A37.95(193) for a family of the same composition.

The 1919-1920 Commission's (weekly) budget cost for a family in Sydney was 5 pounds, 17 shillings. (Sydney is chosen for this comparison because Halladay's poverty budget (see p. 64 above) was developed for a study in that city.) That figure would be equal to about $A34.30 in 1969 dollars--74 percent of what the Halladay poverty budget would be ($A46.30)(194) for a family of the same composition.

The preceding two examples demonstrate the income elasticity of the poverty line (as discussed above on p. 10) by showing that a given constant-pound/dollar level was considered to be well above poverty during an earlier period (when general real incomes were lower) but was considered to be below poverty during a later period (when general real incomes were higher).(195)

Evidence from Australia--A Quantitative Study of Gallup Poll Responses

There is evidence indicating that Australian responses to a Gallup Poll [moderate-income] question have risen in real terms since 1950 as the real income of the general population has risen. In a 1989 paper(196), Peter Saunders and Bruce Bradbury analyzed responses to the [Australian] Morgan Gallup Poll question, "In your opinion, what is the smallest amount a family of four - two parents and two children - need a week to keep in health and live decently - the smallest amount for all expenses including rent?"(197),(198) They examined data for the period 1950-1988, during which the question was asked 58 times.(199) After adjusting all data for inflation, they ran regressions involving the "health/decency" amounts and three different measures of average income for the total population: private final consumption expenditure per capita (PFCEPC), an index of male average weekly earnings (MAWE), and household disposable income per capita (HHDIPC). (The last of these measures is the one used to update the Henderson poverty line since 1981.(200)) They found that the "health/decency" amount rises in real terms as the three measures of real average income rise, but at a slower percentage rate; there was a shift in the relationships in 1970. "These results...imply an income elasticity of the ["health/decency" amount] with respect to HHDIPC of 0.75 over the whole period, elasticities with respect to PFCEPC of 0.85 and 0.53 in the pre-1970 and post-1970 periods, and an elasticity with respect to MAWE of 0.63 over the whole period."

Evidence from Australia--the Common Knowledge of Informed Analysts

The Henderson poverty line was developed in connection with a 1966 survey of income and needs in Melbourne. In a 1970 book reporting the results of the survey, a section on the poverty line included the following comment: "...through time, the [poverty] line must move upwards: we recognise that an income that would have kept a[n Australian] family quite comfortably at the turn of the century would be regarded as quite inadequate today."(201)

In a 1982 article on the Henderson poverty line, Ian Manning wrote, "Historically[,] poverty lines have often been set by compiling typical budgets for low income families of different size....as economic growth proceeds[,] the budgets become progressively more out of date, and in due course a new generation of altruists points this out; the debate starts again and perhaps ushers in a new, higher standard and a new set of hypothetical budgets. In other words, poverty line household budgets at any time are specific to that time, and will change as the budgets of households on higher incomes change...."(202)

A 1988 paper on income support policies by the [Australian] Economic Planning Advisory Council included the following comment: "It is probably true that most people think about poverty in absolute terms - insufficient food, the absence of decent housing and clothing, and so on. Even so, most people's ideas about the minimum acceptable standards of food and clothing are affected by prevailing community standards. Thus even poverty standards established by prescribing a minimum acceptable standard of living tend to move upwards with average incomes."(203)

The Social Processes Underlying the Income Elasticity of the Poverty Line

The social processes underlying the phenomenon of the income elasticity of the poverty line can be gleaned fairly well from some of the quotations cited in this paper--particularly Hamilton (the 1962 quotation given on pp. 21-22 above), the [U.S.] President's Commission on Income Maintenance Programs (see p. 27 above), and Mack and Lansley (the 1985 quotation given on p. 47 above).(204) As technology progresses and the general standard of living rises, new products or consumption items are introduced. They may at first be purchased and used only by upper-income families; however, they gradually diffuse to middle- and lower-income levels. Things that were originally viewed as luxuries--for instance, indoor plumbing, telephones, radios, and automobiles(205)--come to be seen as conveniences and then as necessities. In addition, changes in the ways in which society is organized (sometimes in response to new "necessities") may make it more expensive for the poor to accomplish a given goal--as when widespread car ownership and increasing suburbanization lead to a deterioration in public transportation, and the poor are forced to buy cars or hire taxis in order to get to places where public transit used to take them. Finally, the general upgrading of social standards can make things more expensive for the poor--for instance, housing code requirements that all houses have indoor plumbing add to the cost of housing.

In the light of these social processes, the only kind of American society in which it would be sociologically justified to have had the same fixed-constant-dollar poverty line since the mid-1960's would be a society in which there had been essentially no technological change or innovation since about 1960. Accountants and economists in such a society would still perform involved numerical analyses using mechanical calculators and slide rules, and secretaries would type up the results using manual typewriters. On certain nights they would all drive home from work hurriedly in their Edsels in order to catch favorite television programs, knowing that if they missed the program, they would never have a chance to see it again. The lack of technological innovation would mean that no new consumer products had been introduced since 1960 to work their way down from the upper to the middle and lower levels of society; the list of "necessities" for a minimal standard of living would thus be the same in 1995 as it was in 1963 (the base year for Orshansky's original poverty thresholds) or 1955 (the year of the food survey from which Orshansky derived the multiplier for her thresholds). If this were an accurate description of our society since 1960, then there would be no need to worry about changing the poverty thresholds from the level at which they were set in the 1960's.

Why Was the Official U.S. Poverty Line Not Raised in Real Terms in 1968-1969 to Reflect Rises in the General Standard of Living?

From the U.S. material cited above it is clear that the income elasticity of the poverty line was reasonably well known among Americans who developed and studied poverty lines during the first seven decades of the twentieth century. However, despite that knowledge,

  • the official U.S. poverty line was never raised in real terms to reflect rises in the general standard of living, and
  • in the 1990's, the extensive body of evidence about the income elasticity of the poverty line is not very well known at all in U.S. academic discussions about poverty lines.

The second of these anomalies will be addressed in the next section. The first will be addressed in this section.

It would probably require a book to explore in full the social and political reasons for the fact that the official U.S. poverty line has never been raised in real terms--but a treatment in such depth is beyond the scope of this paper. Instead, I will briefly review the (obscure) historical record of the primary occasion(206) when the official poverty line was not raised, and compare it with events only half a decade earlier when higher poverty lines were accepted.

The poverty thresholds developed by Mollie Orshansky of the Social Security Administration (SSA) were adopted as the U.S. government's poverty measure on a quasi-official basis in 1965 and on an official basis in 1969. The first of these events occurred in May 1965, when the Office of Economic Opportunity (OEO) adopted the lower of Orshansky's two sets of poverty thresholds as a working definition of poverty for statistical, planning, and budget purposes. In November 1965--only six months after the quasi-official adoption of the thresholds--Robert Ball, the Commissioner of Social Security (head of SSA), wrote a memo in which he referred to the fact of the income elasticity of the poverty line and stated that "one of the most difficult methodological questions we will have to face in the next few years is when and how to adjust the definition of poverty" to reflect the rise in general levels of living. (See p. 24 above. Orshansky had in effect set the scene for this concern by the comment in her July 1963 article quoted on p. 22 above.) Comments in unpublished and published papers by Orshansky and her bosses Ida Merriam and Lenore Epstein between April 1966 and March 1968 (see pp. 24-25 above for two of these comments) show an ongoing concern in SSA about the need to raise the poverty thresholds to reflect rising general standards of living and about how this should be done.(207)

In the spring of 1968, SSA actually made a decision to raise the poverty thresholds modestly to bring them more in line with the higher general level of living. The specific step proposed to do this was to use the recently revised version of the Agriculture Department's economy food plan--updated on the basis of consumption data from the Agriculture Department's 1965 Household Food Consumption Survey--to recalculate the poverty thresholds; since the revised economy food plan cost 8 percent more than the old plan (which had been based on data from the 1955 Household Food Consumption Survey), the thresholds would have been raised in real terms by that percentage. There was also informal discussion at this time about the possibility of replacing the poverty threshold multiplier of 3 (derived from the 1955 Household Food Consumption Survey) with a higher multiplier derived from the 1965 survey, but this idea was not part of the formal SSA proposal for revising the thresholds. (Together with the use of the revised food plan, this higher multiplier would have resulted in poverty thresholds 25 to 30 percent higher than the existing thresholds.)(208)

In April 1968, SSA convened a meeting to seek advice from technical staff from the federal agencies with an interest in poverty. SSA personnel presented to the group their proposal to recalculate the poverty thresholds using the revised economy food plan, and the group agreed with this proposal. SSA started to implement its decision to revise the thresholds, and by about mid-June had poverty population figures for calendar years 1966 and 1967 on both the old and the revised basis. For 1967, the revised poverty population figure was higher than the unrevised figure by about 2.8 million persons. In connection with the review of a report on the American Negro which was to be released by the White House and which was to use the revised poverty figures, several agencies, particularly OEO and the Council of Economic Advisers, argued that the revised poverty figures should not be published, thus repudiating the position taken by their representatives at the April advisory meeting. The issue became entangled with several conceptually unrelated problems; a further problem (mentioned--in the material that I have--only in a later memorandum) was that OEO was using the poverty line to determine eligibility in a number of its programs, and a change would have seriously affected budgets and/or regulations. In July the head of the Office of Statistical Standards in the Bureau of the Budget (BoB) sent a letter to the head of the Census Bureau directing him to use the old (unrevised) poverty thresholds for tabulations of the 1967 poverty population. The letter also said that BoB would appoint a task force to develop concepts and technical information required to re-evaluate the poverty thresholds.(209)

An interagency Poverty Level Review Committee was appointed, and held its first meeting in October 1968. Ida Merriam of SSA addressed that meeting, mentioning the issue of how to adjust the poverty thresholds for price changes, but stating that that "is a very minor question. The real issue is how often adjustments should be made to reflect changes in the level of living. I stated our reason for recommending that the 1967 levels [thresholds] reflect the new food expenditure patterns shown in the Department of Agriculture's '65 [Household Food Consumption] survey...." Additional issues were discussed. A representative of the Council of Economic Advisers commented that "from the political point of view one simply could not increase the poverty level in a year when employment was good."(210) Later in the meeting "[i]t was emphasized that with respect to overall government policy it was essential that means be found for 'score-keeping' without sudden shifts in the level [poverty threshold]. A consistent measure, adjusted basically for price movement, is necessary in order to measure the effect of government policies and changes in the economy upon the number of families and persons below a 'poverty' income level."(211) In all the records of subsequent meetings of the Committee that I have, there was essentially no discussion of the issue of raising the poverty thresholds in real terms to reflect changes in the general standard of living.(212) In early 1969 the Committee made recommendations (which were adopted) about the indexing of the thresholds for price changes and the differential between farm and nonfarm thresholds(213), but it said nothing about the issue of raising the poverty thresholds to reflect changes in the general standard of living.

In the sequence of events reviewed here, the primary objection to a higher poverty line seems to have been the fact that it would have resulted in a higher number of people being counted as poor. The Johnson Administration had proclaimed a War on Poverty, and was in a position of being able to boast of a three-year drop in the poverty population of 5.6 million persons(214). In that context it would have been politically embarrassing to have reported a 2.8 million "increase" in the poverty population; too many people might have misunderstood or misinterpreted the "increase" as being the result of failed Administration anti-poverty policies or failed Administration management of the economy, rather than as the statistical result of a redefinition of poverty. (The concern about "score-keeping" is related to this objection, as the 2.8 million "increase" would not have been the result of either government policies or changes in the economy.) The concern about possible budgetary effects of a higher poverty line on OEO programs appears to have been less important than the objection to a higher poverty count. (Indeed, any budgetary effects on OEO programs would have been only indirect, since the programs in question were fixed-appropriation programs, not open-ended ("entitlement") programs.(215))

Further insight can be gained by comparing these 1968-1969 events with events only half a decade earlier, when quasi-official poverty lines were first adopted in the initial stages of the War on Poverty. In its January 1964 report, President Johnson's Council of Economic Advisers (CEA) put forward a poverty line of $3,000 (in 1962 dollars) for families of all sizes (and $1,500 for unrelated individuals) which became the federal government's quasi-official poverty line for a little over a year. As noted above on p. 9, this poverty line was 19 percent higher in real terms than the low-income line of $2,000 (in 1948 dollars) for families of all sizes put forward in 1949--only fifteen years earlier--by the Congressional Subcommittee on Low-Income Families (SLIF). Robert Lampman (the person primarily responsible for the chapter on poverty in the 1964 Economic Report) must have realized that the $3,000 CEA poverty line was higher in real terms than the $2,000 SLIF low-income line, since he had become familiar with the latter while doing a 1959 study on the low-income population for the Joint Economic Committee. In addition, as also noted above (footnote 23), at least two other analysts at the time explicitly noted that the CEA $3,000 figure was higher in real terms than the 1949 SLIF $2,000 figure. However, in all the material on poverty definition and measurement from this period that I have read, I have never found anyone who objected to the adoption of the CEA poverty line on the grounds that poverty is absolute, that a satisfactory estimate of minimum absolute needs had been developed fifteen years ago, and that it would be wrong and/or confusing to adopt a new poverty line that was higher in real terms than the SLIF figure. Just over a year later, in May 1965, Orshansky's poverty thresholds were adopted by the Office of Economic Opportunity as the federal government's [new] quasi-official poverty line. The adoption of Orshansky's thresholds was another example of the income elasticity of the poverty line in action; her 1963 nonfarm poverty threshold of $3,128 for a family of four was 14 percent higher in real terms than Lampman's low-income line of $2,516 (in 1957 dollars) for a family of four from his 1959 study--published only six years earlier. Lampman and a number of other people must have been aware of this, since Lampman included a version of his low-income lines expressed in 1963 dollars in a paper that he presented at the West Virginia University Conference on Poverty amid Affluence in May 1965. But I have never found any record of someone objecting to the adoption of the Orshansky poverty thresholds in place of the Lampman low-income lines on the grounds that I have just outlined above.(216) The events of 1964-1965 show that American policymakers and analysts did not have a long-standing,principled objection to having poverty lines that reflected contemporary standards. The difference between the 1964-1965 events and the 1968-1969 events was due not to matters of principle but to matters of expediency. (Note the comment cited in the next paragraph about "political, not conceptual, constraints.") In 1968-1969, a government agency was issuing widely publicized annually updated statistics on the poverty population using the Orshansky thresholds. (By contrast, the poverty population statistics based on the 1949 SLIF and 1959 Lampman low-income lines were not widely known and were not updated annually.) The decision not to adopt a more up-to-date poverty line was made because it was deemed to be too politically embarrassing and difficult to explain the apparent "increase" in the poverty population due to the adoption of a more up-to-date poverty line.

As early as 1967, activist scholars S.M. Miller, Martin Rein, Pamela Roby, and Bertram Gross had anticipated an outcome such as occurred in 1968-1969, writing, "We believe that poverty lines based upon budget-oriented approaches will continue to be inadequate because of the deep political implications which each upward adjustment involves. Although a budget-based poverty line may be rapidly falling farther behind the rising standard of living enjoyed by the rest of the population, it will not be adjusted upward until that change appears politically feasible."(217) After the decision not to raise the poverty line had been made, Miller and Roby wrote, "In the 1960's, unlike earlier periods, the budget-oriented estimates [of the poverty line] have not changed to keep up with changes in average styles of life. This break with previous practice is because of political, not conceptual, constraints."(218) Expressing similar insights, Canadian economist Lars Osberg wrote in 1984 (in a book on economic inequality in the U.S.), "Unfortunately, once three times the subsistence food budget became enshrined in 1963 as the poverty line, it has become very difficult to make the periodic revisions to the subsistence budget which were normal before 1963. As a result, 'subsistence' as officially defined has fallen further and further below the average American's standard of living."(219)

The effects of the decision not to raise the poverty line in 1968-1969 (and the similar decision in 1973) may be summed up in the following comparison:

  • Over the 28-year period from 1937 to 1965, the level that was considered to be poverty or minimum subsistence for a family of four rose in real terms by 55 percent.(220)
  • Over the 28-year period from 1965 to 1993, by contrast, there was no change in real terms in the level of the U.S. poverty line.

Presumably Mollie Orshansky had some comparison of this type in mind when she commented (in a 1970 newspaper interview) that the 1969 decision "tends to freeze the poverty line despite changes in buying habits and changes in acceptable living standards."(221)

Why Is the Income Elasticity of the Poverty Line So Little Known in the U.S. Today?

As noted above, it is clear that the income elasticity of the poverty line was reasonably well known among Americans who developed and studied poverty lines during the first seven decades of the twentieth century. In the 1990's, however, this extensive body of evidence is not very well known at all in U.S. academic discussions about poverty lines. What could be the reason for this change?

My studies of the history of U.S. poverty lines during the twentieth century have led me to believe that this change in knowledge is connected with a change in the identity of the groups that develop and/or study poverty lines in the U.S. Before about 1965, the people who developed (and studied) poverty lines were to a significant extent advocates of the disadvantaged rather than theoretical social scientists elucidating abstract concepts about minimum consumption. In terms of professions, the people involved included social workers (especially during the Progressive Era, when social work included many activist social reformers), employees of state bureaus of labor statistics, labor union representatives, home economists, and employees of federal social agencies. Some economists did participate, but they were only one of several elements in the mix, and relatively little of the pre-1965 poverty line literature that I have found was in traditional economic publications.(222) In addition, there was some tendency for economists involved with poverty lines to be more likely to be women interested in distributional or household economics, rather than men interested in macroeconomics.(223)

That situation changed with the beginning of the War on Poverty in 1964. Poverty studies became a distinct field as such, and (macro)economists began to get involved in poverty line studies in large numbers.(224) People who had been involved in poverty line studies during the earlier period gradually retired and/or died during the 1970's and 1980's. By 1980, Walter Korpi of Sweden was noting that sectorially oriented poverty research in the United States was "clearly dominated" by economists (in contrast to the situation in Britain and Scandinavia, where there was a "relatively even balance" between sociologists and economists).(225) Today if one mentions (in an American context) that one works on poverty, the common response is, "You must be an economist, then."

Knowledge of the income elasticity of the poverty line was part of the history and traditions that had been passed down among the earlier groups of people who developed and studied poverty lines. As these groups were gradually replaced by economists who got involved in studying poverty and poverty lines beginning in the 1960's, I believe that what happened was that the history and traditions of the earlier groups tended not to be transmitted to the newcomers; the newcomers did not perceive them as part of "our" history. As a result, much of the knowledge about the income elasticity of the poverty line was lost to those who are now studying poverty lines.

In addition, the fact that proposals to raise the poverty thresholds in real terms were rejected--leaving in place an official poverty line that had been made absolute--tended to serve as a further barrier to retention of the knowledge of the income elasticity of the poverty line. When the current official poverty line--the only poverty line that more and more people had ever been aware of--was made absolute, and remained so, it became more difficult for many people to realize (and less likely that they would investigate history to find out) that the basic pattern both in this country and in other countries is for poverty lines to rise in real terms as the real income of the general population rises.


ENDNOTE

1. A[rthur] L. Bowley, "The New Survey of London Life and Labour. Effect of Modifying the Poverty Line," Journal of the Royal Statistical Society, Vol. 99, Part II, 1936, p. 364.

2. See, for instance, U.S. Department of Health, Education, and Welfare, The Measure of Poverty[:] A Report to Congress as Mandated by The Education Amendments of 1974, Washington, D.C., U.S. Government Printing Office, April 1976, pp. 20-21; Isabel V. Sawhill, "Poverty in the U.S.: Why Is It So Persistent?", Journal of Economic Literature, Vol. 26, No. 3, September 1988, pp. 1076-1077; and A[nthony] B. Atkinson, "How should we measure poverty? Some conceptual issues," pp. 15-17 in Atkinson, Poverty and Social Security, London, Harvester Wheatsheaf, 1989.

3. M[ichael] C. Wolfson and J[ohn] M. Evans, Statistics Canada's Low Income Cut-Offs[:] Methodological Concerns and Possibilities[--]A Discussion Paper [December 1989], Statistics Canada, Analytical Studies Branch, Research Paper Series, pp. 4-5; cf. also p. 47.

4. See, for instance, Walter Korpi, "Approaches to the Study of Poverty in the United States: Critical Notes from a European Perspective," pp. 290-294 in Vincent T. Covello (editor), Poverty and Public Policy[:] An Evaluation of Social Science Research, Boston and Cambridge, Massachusetts, G.K. Hall & Co. and Schenkman Publishing Co., 1980.

5. Although this phenomenon has been known for over three quarters of a century, the first application of the term "income elasticity" to it that I have found was in Theodore W. Schultz, "Investing in Poor People: An Economist's View," American Economic Review, Vol. 55, No. 2, May 1965, pp. 511-512.

6. Of the studies discussed in this section and in the "get-along" section, I found the citations for four (Appelbaum, Kilpatrick, the 1978 Leveson article, and the 1974 Rainwater book) in the bibliography (pp. 9-10) of James H. Hauver, John A. Goodman, and Marc A. Grainer, "The Federal Poverty Thresholds: Appearance and Reality," Journal of Consumer Research, Vol. 8, June 1981, pp. 1-10.

7. Judith Innes de Neufville, Social Indicators and Public Policy[:] Interactive Processes of Design and Application, New York, Elsevier Scientific Publishing Company, 1975, p. 86; and Mollie Orshansky, "Family Budgets and Fee Schedules of Voluntary Agencies," Social Security Bulletin, Vol. 22, No. 4, April 1959, p. 10.

8. Oscar Ornati (with the editorial assistance of J. Stouder Sweet), Poverty Amid Affluence[:] A Report on a Research Project carried out at the New School for Social Research, New York, Twentieth Century Fund, 1966. For discussions and tabular material on the sixty budgets and Ornati's three categories, see pp. 10-14, 141-145, and 147-150.

9. Eugene Smolensky, "The Past and Present Poor," pp. 40 and 45 in U.S. Chamber of Commerce, Task Force on Economic Growth and Opportunity, The Concept of Poverty, Washington, D.C., U.S. Chamber of Commerce, 1965. (These data can also be found in a shorter reprinted version of this paper--see Smolensky, "The Past and Present Poor," pp. 87-88 in Robert William Fogel and Stanley L. Engerman (editors), The Reinterpretation of American Economic History, New York, Harper & Row, 1971.)

10. Robert W. Kilpatrick, "The Income Elasticity of the Poverty Line," Review of Economics and Statistics, Vol. 55, No. 3, August 1973, pp. 327-332.

11. Ornati, pp. 10-11 and 149-150.

12. Ornati, pp. 12-15; the direct quotation is on p. 13, while the constant-dollar values of the budgets and the index of men's low wages can be found in tabular form on pp. 149-150. (For the construction of the wage index, see pp. viii and 146-147.)

13. Ornati, pp. 151-154, with a brief discussion on pp. 20-22.

14. Lee Rainwater, What Money Buys[:] Inequality and the Social Meanings of Income, New York, Basic Books, 1974, p. 48.

15. Cf. the discussion on p. 10 below about how comparisons between the constant-dollar values of early minimum comfort standards and later minimum subsistence standards over a long enough period of time can be used to demonstrate the income elasticity of the poverty line.

16. Diana Karter Appelbaum, "The Level of the Poverty Line: A Historical Survey," Social Service Review, Vol. 51, No. 3, September 1977, pp. 514-523. This otherwise excellent article does contain some misunderstandings (pp. 516-517) about the use of Agriculture Department food plans in the Orshansky poverty line. (This is by no means the only account to contain such misunderstandings.) The "multiplier" that Orshansky used to calculate her poverty line from food plan costs was derived from the 1955 average food-to-total-income ratio for all families, not for low-income families. It is not true that a poverty line based on the Low-Cost Food Plan was in use until 1964, or that such a poverty line was replaced by a lower poverty line based on the Economy Food Plan. The Economy Food Plan was developed by the Agriculture Department in 1961, not 1964; what happened in 1964 was a revision of all four food plans. The Thrifty Food Plan was never used to revise or update the poverty line. The poverty line is not recalculated each year by multiplying current food plan costs by three; instead, it is updated each year from the 1963 base-year (nonfarm) values by means of the Consumer Price Index. (For an accurate account of the use of the food plans in the Orshansky poverty line, based on an intensive review of primary sources, see Gordon M. Fisher, "The Development of the Orshansky Poverty Thresholds and Their Subsequent History as the Official U.S. Poverty Measure" (unpublished paper--76 pp.), May 1992, pp. 6-13 and 48-49--summarized in Fisher, "The Development and History of the Poverty Thresholds," Social Security Bulletin, Vol. 55, No. 4, Winter 1992, pp. 4-5 and 8.)

17. Fisher, "The Development of the Orshansky Poverty Thresholds..." (unpublished), pp. 4-6 and 9-13--summarized in Fisher, "The Development and History of the Poverty Thresholds," pp. 4-5.

18. For an account of Stecker's "emergency" budget, see Gordon M. Fisher, "From Hunter to Orshansky: An Overview of (Unofficial) Poverty Lines in the United States from 1904 to 1965" (a paper presented at the Fifteenth Annual Research Conference of the Association for Public Policy Analysis and Management in Washington, D.C. (75 pp.)), October 1993, pp. 38-39, and the primary sources cited there. As indicated on p. 39, the emergency budget may be considered to be a minimum subsistence budget. For the food/total-expenditures share of the emergency budget, see Margaret Loomis Stecker, Intercity Differences in Costs of Living in March 1935, 59 Cities, Research Monograph XII, Works Progress Administration, Division of Social Research, Washington, U.S. Government Printing Office, 1937, p. 6.

19. It is of interest to note that in 1919--approximately the midpoint of the period studied by Appelbaum--William Ogburn wrote, "In minimum-of-subsistence budgets food usually costs about 44 per cent of the total..."--William F. Ogburn, "Measurement of the Cost of Living and Wages," Annals of the American Academy of Political and Social Science, Vol. 81, January 1919, p. 116.

20. From about 1909 on, a number of analysts did reviews of individual-city standard budgets and estimated a more broadly based standard of living (or several such standards) based on their review, along the general line of "Based on budget A for city 1, budget B for city 2, and budget C for city 3, a good estimate of the minimum subsistence level for a family of five for year 19xx would be an annual income of Y dollars." To distinguish them from individual-city poverty-level budgets, I referred to these more broadly based standards of living as "derivative" poverty (or minimum subsistence) lines.

21. In particular, see Fisher, "From Hunter to Orshansky...", pp. 13-15, 18-23, 27-28, 32-33, 38-39, 48-49, and 56-57.

22. During the post-World-War-II period, it is true that more liberal poverty lines were already sometimes higher than the family poverty line set by President Johnson's Council of Economic Advisers (CEA), while conservative poverty lines were set much lower than the Johnson CEA or Orshansky figures.

23. At least two contemporary analysts made explicit note of this fact. See Herman P. Miller, "Statistics and Reality," New Leader, Vol. 47, No. 7, March 30, 1964, p. 15; and Burton A. Weisbrod, "The Economics of Poverty: An American Paradox," pp. 13-14 in Burton A. Weisbrod (editor), The Economics of Poverty, Englewood Cliffs, New Jersey, Prentice-Hall, Inc., 1965.

24. Ornati, pp. 149-150.

25. [no author given], "Tentative Quantity-Cost Budget Necessary to Maintain Family of Five in Washington, D.C.," Monthly Labor Review, Vol. 9, No. 6, December 1919, pp. 22-29; also published separately as U.S. Department of Labor, Bureau of Labor Statistics, Tentative Quantity and Cost Budget Necessary to Maintain a Family of Five in Washington, D.C., at a Level of Health and Decency, Washington, D.C., [U.S.] Government Printing Office, 1919. For the historical background and context of this budget, see Daniel Horowitz, The Morality of Spending[:] Attitudes toward the Consumer Society in America, 1875-1940, Baltimore, Johns Hopkins University Press, 1985, pp. 120-129; and Naomi Aronson, "The Making of the U.S. Bureau of Labor Statistics Family Budget Series: Relativism and the Rhetoric of Subsistence" (June 1984 version of a paper presented at the American Sociological Association Meetings, San Antonio, August 1984), especially pp. 2 and 20-25.

26. [no author given], "Minimum Quantity Budget Necessary to Maintain a Worker's Family of Five in Health and Decency," Monthly Labor Review, Vol. 10, No. 6, June 1920, pp. 1-18.

27. Rainwater, What Money Buys..., p. 53.

28. Kilpatrick, p. 327.

29. Rainwater, What Money Buys..., pp. 50-54; and Kilpatrick, pp. 327 and 329 (footnote 12).

30. Lee Rainwater, "Poverty, Living Standards, and Family Well-Being," pp. 207-254 in U.S. Congress, Joint Economic Committee, Subcommittee on Fiscal Policy, Studies in Public Welfare, Washington, D.C., U.S. Government Printing Office: Paper No. 12 (Part II), "The Family, Poverty, and Welfare Programs: Household Patterns and Government Policies," December 3, 1973; and Rainwater, What Money Buys..., Chapter 3, "Changing Standards of Living," pp. 41-63.

31. This (subsequently discontinued) Bureau of Labor Statistics series was derived from the series on average weekly earnings for private nonsupervisory nonagricultural workers (regardless of number of dependents) by attributing the overall average figure to a hypothetical married worker with a total of three dependents and deducting estimated amounts for federal income tax and Social Security payroll tax.

32. Irving Leveson, "Updating Poverty Standards and Program Benefits," Growth and Change, Vol. 9, No. 1, January 1978, pp. 14-22. (For earlier versions of this material, see Irving Leveson, Poverty and Public Policy (HI-2307-RR--a study conducted as part of a project of U.S. social policy studies conducted under a grant from the Office of Economic Opportunity), Chapter III, "Revision of Poverty Standards Over Time," pp. III-1 through III-31, Croton-on-Hudson, New York, Hudson Institute, Inc., July 14, 1975; and Irving Leveson, "Updating Poverty Standards and Program Benefits," American Statistical Association[:] Proceedings of the Social Statistics Section[,] 1975, pp. 523-528.) Leveson's interesting and useful article does contain chronological errors about the base year and (quasi-)official adoption of the Orshansky poverty thresholds. The base year for Orshansky's thresholds was 1963, not 1960. (Similarly, the base year for the Council of Economic Advisers' poverty line was 1962, not 1960.) Orshansky's poverty thresholds for 1963 were first published in January 1965, and they were adopted on a quasi-official basis by the Office of Economic Opportunity in May 1965, not in 1960. (See Fisher, "The Development of the Orshansky Poverty Thresholds..." (unpublished), pp. 2-3, 17, and 28--summarized in Fisher, "The Development and History of the Poverty Thresholds," pp. 3, 4, and 6.)

33. Lars Osberg, Economic Inequality in the United States, Chapter 4, "Poverty and Inequality," pp. 61-84, Armonk, New York, M.E. Sharpe, Inc., 1984.

34. Lee Rainwater, "Poverty and Equivalence as Social Constructions" (Luxembourg Income Study Working Paper No. 55), August 1990.

35. Denton R. Vaughan, "Exploring the Use of the Public's Views to Set Income Poverty Thresholds and Adjust Them Over Time," Social Security Bulletin, Vol. 56, No. 2, Summer 1993, pp. 22-46. For his discussion of the income elasticity of the "get-along" amount, see p. 42 (footnote 10) and p. 43 (footnote 14).

36. Theodore Parker, "Thoughts on Labor," The Dial, Vol. 1, No. 4, April 1841, p. 514--reprinted on pp. 144-145 of Parker, The Critical and Miscellaneous Writings of Theodore Parker... (second edition), Boston, Little, Brown and Company, 1856--reprinted (under the title "The Laboring Classes") on pp. 64-65 of Parker (edited with notes by Samuel A. Eliot), Social Classes in a Republic [Volume 10, The Works of Theodore Parker, Centenary Edition], Boston, American Unitarian Association, [1907-1911].

37. John A. Ryan, "What Wage is a Living Wage?", Catholic World, Vol. 75, No. 445, April 1902, p. 5 (emphasis in original). The quotation was reprinted almost verbatim in Father Ryan's 1906 book on the living wage; see Ryan, A Living Wage[:] Its Ethical and Economic Aspects, New York, Macmillan, 1906, p. 127. (For brief comments on Father Ryan's 1902 article and for a brief discussion of his 1906 book, see Fisher, "From Hunter to Orshansky...", pp. 13-14 and 15-16.)

38. John Mitchell, Organized Labor[:] Its Problems, Purposes and Ideals and the Present and Future of American Wage Earners, Philadelphia, American Book and Bible House, 1903, p. 116.

39. W. Jett Lauck and Edgar Sydenstricker, Conditions of Labor in American Industries[:] A Summarization of the Results of Recent Investigations, New York, Funk & Wagnalls Company, 1917, pp. 378-379.

40. William F. Ogburn, "Measurement of the Cost of Living and Wages," Annals of the American Academy of Political and Social Science, Vol. 81, January 1919, p. 114.

41. I[saac] M. Rubinow, "Poverty," pp. 284-285 in Encyclopaedia of the Social Sciences, Volume Twelve, New York, Macmillan, 1934.

42. Margaret Loomis Stecker, Intercity Differences in Costs of Living in March 1935, 59 Cities, Research Monograph XII, Works Progress Administration, Division of Social Research, Washington, U.S. Government Printing Office, 1937, p. xiv. (For a brief discussion of Stecker's budgets, see Fisher, "From Hunter to Orshansky...", pp. 38-39.)

43. Carroll R. Daugherty, Labor Problems in American Industry (fourth edition), Cambridge, Massachusetts, Houghton Mifflin, 1938, p. 137. (For further discussion of Daugherty's section on standards of economic well-being, see Fisher, "From Hunter to Orshansky...", pp. 41-42.)

44. Dorothy S. Brady, "Research on the Size Distribution of Income," p. 31 in Studies in Income and Wealth[,] Volume Thirteen, New York, National Bureau of Economic Research, Inc., 1951.

45. For the 1949 and 1955 Subcommittees on Low-Income Families, see Fisher, "From Hunter to Orshansky...", pp. 48-50 and 52-53.

46. 46Robert R. Nathan, [written] "Statement...", p. 64 in U.S. Congress, Joint Committee on the Economic Report, Low-Income Families[:] Hearings Before the Subcommittee on Low-Income Families...Pursuant to Sec. 5(a) of Public Law [79-]304..., Washington, D.C., U.S. Government Printing Office, 1955.

47. Margaret G. Reid, "Statement...", pp. 725 and 728 in U.S. Congress, Joint Committee on the Economic Report, Low-Income Families[:] Hearings....

48. Conference on Economic Progress, The Federal Budget and "The General Welfare"[--]We Can Afford To Serve Our Human Needs, Washington, D.C., December 1959, p. 59 (emphasis in original). (For brief comments on the Conference on Economic Progress and its poverty lines during the 1950's and 1960's, see Fisher, "From Hunter to Orshansky...", pp. 51-52, 53, 61-62, and 66-67.)

49. Mollie Orshansky, "Budget for an Elderly Couple: Interim Revision by the Bureau of Labor Statistics," Social Security Bulletin, Vol. 23, No. 12, December 1960, pp. 31-32 and 33.

50. David Hamilton, "Drawing the Poverty Line at a Cultural Subsistence Level," Southwestern Social Science Quarterly, Vol. 42, No. 4, March 1962, pp. 343 and 345.

51. David Hamilton, A Primer on the Economics of Poverty, New York, Random House, 1968, p. 27.

52. Mollie Orshansky, "Children of the Poor," Social Security Bulletin, Vol. 26, No. 7, July 1963, p. 3.

53. Minority [Republican] views, p. 46 in U.S. Congress, 1964 Joint Economic Report[--]Report of the Joint Economic Committee...on the January 1964 Economic Report of the President With Minority and Additional Views, Washington, D.C., U.S. Government Printing Office, 1964.

54. Rose D. Friedman, Poverty[:] Definition and Perspective, Washington, D.C., American Enterprise Institute for Public Policy Research, February 1965, p. 12. (For a brief discussion of Friedman's poverty lines, see Fisher, "From Hunter to Orshansky...", pp. 68-69.)

55. Victor R. Fuchs, "Toward a Theory of Poverty," pp. 72-74 in Task Force on Economic Growth and Opportunity, The Concept of Poverty, Washington, D.C., Chamber of Commerce of the United States, 1965. (For a brief discussion of Fuchs' relative poverty line, see Fisher, "From Hunter to Orshansky...", pp. 70-71.)

56. P. 3 of November 4, 1965, memorandum from Robert M. Ball, Commissioner of Social Security, to Wilbur J. Cohen, Under Secretary [of Health, Education, and Welfare]--Subject: Poverty Research - your memorandum of October 29.

57. I[da ]C[. ]Merriam, "The Meaning of Poverty-Effectiveness" (draft), January 4, 1967, p. 2.

58. James Tobin, "It Can Be Done!", New Republic, Vol. 156, No. 22, June 3, 1967, p. 15.

59. [U.S.] President's National Advisory Commission on Rural Poverty, The People Left Behind..., Washington, D.C., U.S. Government Printing Office, September 1967, p. 7.

60. Ida C. Merriam, "Concepts and Measures of Welfare," p. 182 in American Statistical Association[:] Proceedings of the Social Statistics Section[,] 1967. (She went on to discuss a possible basis for "a decision in 1970 as to a new starting level [for the poverty index] and perhaps a different basis for adjustment over the following decade.") Merriam used this paper as the nucleus for a chapter in a 1968 book on indicators of social change. For a reprinting of the first part of the above quotation (with minor changes in the last sentence), see Merriam, "Welfare and its Measurement" (Chapter 14), p. 762 in Eleanor Bernert Sheldon and Wilbert E. Moore (editors), Indicators of Social Change[:] Concepts and Measurements, New York, Russell Sage Foundation, 1968.

61. Statistics on poverty had been included in Census Bureau publications on other subjects in 1966 and 1967; see Fisher, "The Development of the Orshansky Poverty Thresholds..." (unpublished), p. 38.

62. 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, pp. 1-2. (After the 1969 revision in the poverty definition, poverty thresholds for 1959-1967 were retroactively revised; accordingly, the poverty thresholds given in this quotation for 1959 and 1966 do not match the thresholds shown for those years in current historical tables.)

63. Mollie Orshansky, "The Shape of Poverty in 1966," Social Security Bulletin, Vol. 31, No. 3, March 1968, p. 6.

64. Economic Report of the President Transmitted to the Congress January 1969 Together With the Annual Report of the Council of Economic Advisers, Washington, D.C., U.S. Government Printing Office, 1969, p. 151.

65. P. 3 of "Considerations Relating to an Official Poverty Index," a February 5, 1969, attachment to a March 6, 1969, memorandum from Robert M. Ball, Commissioner of Social Security, to the Secretary [of Health, Education, and Welfare--Robert H. Finch]--Subject: Responsibility for Development of Poverty Index.

66. Lee Rainwater, "The Problem of Lower-Class Culture and Poverty-War Strategy," p. 253 in Daniel P. Moynihan (editor--with the assistance of Corinne Saposs Schelling), On Understanding Poverty[:] Perspectives from the Social Sciences, New York, Basic Books, Inc., 1969.

67. [U.S.] President's Commission on Income Maintenance Programs, Poverty Amid Plenty[:] The American Paradox[--]The Report of The President's Commission..., Washington, D.C., U.S. Government Printing Office, November 1969, pp. 37-38. Some of these comments were paraphrased in Harvey D. Shapiro, "Making Poverty Relative," New Leader, Vol. 53, No. 5, March 2, 1970, pp. 8-9. (Shapiro had served as a consultant to the Commission, as noted on p. vi of its report.)

68. References to the poverty surveys done by B. Seebohm Rowntree and by Arthur L. Bowley and colleagues can be found in a number of sources. However, for references to other surveys of poverty in various British towns during the 1930's, I am indebted to Peter Townsend, "Poverty: Ten Years After Beveridge," Planning (published by Political and Economic Planning), Vol. 19, No. 344, August 4, 1952, pp. 22-23; and Townsend, "Measuring Poverty," British Journal of Sociology, Vol. 5, No. 2, June 1954, pp. 136-137. (I have referred to some but not all of these surveys in this section and the following section.) A number of these surveys are also cited in A[nthony] B. Atkinson, "Poverty in Britain from the 1930s to the 1980s," pp. 40-61 in Atkinson, Poverty and Social Security, London, Harvester Wheatsheaf, 1989; and in John H. Veit-Wilson, "Muddle or Mendacity? The Beveridge Committee and the Poverty Line," Journal of Social Policy, Vol. 21, Part 3, July 1992, pp. 269-301.

69. G.C. Fiegehen, P.S. Lansley, and A.D. Smith (with a contribution by N.C. Garganas), Poverty and Progress in Britain[,] 1953-73[--]A statistical study of low income households: their numbers, types and expenditure patterns (The National Institute of Economic and Social Research[--]Occasional Papers[--]XXIX), Cambridge, Great Britain, Cambridge University Press, 1977, Appendix II, "Changes in Poverty Standards," pp. 129-135 (see also pp. 11-13 and 15-16); and Atkinson, "Poverty in Britain from the 1930s...", pp. 48-52 and 55-60. (The direct quotation is from Fiegehen et al., p. 15.)

70. Fiegehen et al., pp. 130-131; and Atkinson, "Poverty in Britain from the 1930s...", pp. 55 and 57-58. Rowntree assumed that the amount of rent that a family was actually paying represented its minimum necessary expenditure for housing ("...rent being almost the first thing in which a poor family will try to economise."--B. Seebohm Rowntree, Poverty[,] A Study of Town Life, London, Macmillan and Co., Ltd., 1901, p. 106); accordingly, he deducted rent paid from a family's income and compared the result with a minimum standard which did not include a budgeted amount for rent. For a brief discussion of how this practice was continued by Sir William Beveridge and in post-1960 British poverty studies, see Atkinson and John Micklewright, Economic transformation in Eastern Europe and the distribution of income, Cambridge, Great Britain, Cambridge University Press, 1992, p. 190.

71. Fiegehen et al., pp. 16 and 131; Atkinson, "Poverty in Britain from the 1930s...", pp. 45, 49-50, 52, 56, and 59. For a discussion of the decision to use "the level of living of National Assistance Board applicants" as the principal (but not the only) measure of poverty in the 1965 study, see Brian Abel-Smith and Peter Townsend, The Poor and the Poorest[:] A New Analysis of the Ministry of Labour's Family Expenditure Surveys of 1953-54 and 1960 (Occasional Papers on Social Administration No. 17), London, G. Bell & Sons Ltd., 1965, pp. 15-20. For an earlier discussion of the decision to use National Assistance rates as a poverty measure in this study, see Peter Townsend, "The Meaning of Poverty," British Journal of Sociology, Vol. 13, No. 3, September 1962, pp. 212 and 214-215.

72. J[ohn] H. Veit-Wilson, "Paradigms of Poverty: A Rehabilitation of B.S. Rowntree," Journal of Social Policy, Vol. 15, Part 1, January 1986, pp. 70 and 72-78.

73. Veit-Wilson, "Muddle or Mendacity?..."; and personal (electronic) communications from Veit-Wilson, October 18, 1994. As noted in Veit-Wilson's article, some analysts during the 1940's were aware that Rowntree's HNOL standard represented a higher standard of living than the P1 standard used in most British poverty surveys; the same point had also been recognized in A[lexander] M. Carr-Saunders and D[avid] Caradog Jones, A Survey of the Social Structure of England & Wales as Illustrated by Statistics, London, Oxford University Press, 1927 [first edition], pp. 188-189. In a 1941 pamphlet, Caradog Jones suggested that the "Poverty Line standard" was appropriate for a family when the head was unemployed or receiving "public support," while the "Human Needs standard" was appropriate "[f]or families which are self-supporting" (University of Liverpool, Social Science Department, Statistics Division, [D. Caradog Jones,] Cost of Living of Representative Working Class Families, University Press of Liverpool, 1941, pp. 10, 11-12, and 15-16); this is somewhat reminiscent of some American discussions of different standards of living during the first third of the twentieth century.

74. Fiegehen et al., pp. 130-131. Figures were adjusted for price changes using official retail price indices, except for the period 1939-1947; for that period, because of weighting problems with the official index, an index prepared by R.G.D. Allen was used. The housing component was not excluded from the indices used. The 1971 Supplementary Benefits figure shown includes the long-term addition of 50 pence; the 1973 figure is the "long-term" rate. I calculated the index figures in the third column directly from the figures in the second column (replacing the index figures in Fiegehen et al., which were calculated on the basis 1899=100).

75. Concerning the "conventional" or "typical" family size figure, see, for instance, Carr-Saunders and Caradog Jones (1927), pp. 182, 186, and 188; P[ercy] Ford, Work and Wealth in a Modern Port[:] An Economic Survey of Southampton, London, George Allen & Unwin Ltd., 1934, p. 124; and (for the U.S.) Fisher, "From Hunter to Orshansky...", pp. 33-34.

76. For families with children, the National Assistance scale rate set by the National Assistance Board in 1948 was lower in real terms than Beveridge's 1938-base-year subsistence standard; in other words, the National Assistance scale rate for these families did not account for the full amount of 1938-1948 price increases when compared with the Beveridge standard (T[ony] Lynes, National Assistance and National Prosperity, Welwyn, Codicote Press, 1962, cited on pp. 6-7 of Jonathan Bradshaw and Tony Lynes, Benefit Uprating Policy and Living Standards (Social Policy Reports, Number 1), Social Policy Research Unit, University of York, 1995; and Fiegehen et al., p. 134).

77. Peter Flora, Franz Kraus, and Winfried Pfenning, State, Economy, and Society in Western Europe[,] 1815-1975[:] A Data Handbook in two Volumes[--]Volume II[--]The Growth of Industrial Societies and Capitalist Economies, Frankfurt, Campus Verlag, 1987, pp. 397-400. Figures for 1963-1973 were converted from 1970 prices to 1900 prices using the ratio for the 1960 figures.

78. R.F. George, "A New Calculation of the Poverty Line," Journal of the Royal Statistical Society, Vol. 100, Part I, 1937, pp. 74-95. George adjusted the earlier poverty lines to 1936 prices using the official retail price index; he appears to have done this separately for each component of the poverty lines (food, clothing, cleaning and lighting, and fuel), which would mean that he did not use the housing component of the retail price index, since (as noted above on p. 28) these poverty lines did not include a budget allowance for rent.

79. [no author given], Men Without Work[:] A Report made to the Pilgrim Trust, Cambridge, Great Britain, Cambridge University Press, 1938, pp. 104-106 and 420 (using George's poverty line with two modifications); Herbert Tout, The Standard of Living in Bristol[:] A Preliminary Report of the Work of the University of Bristol Social Survey, Bristol, Great Britain, J.W. Arrowsmith Ltd., 1938, p. 16 (using a minimum standard "very nearly equal to" George's poverty line); and M[ary] S[tewart] Soutar, E[dgar] H[enry] Wilkins, and P[hilip] Sargant Florence, Nutrition and size of Family[:] Report on a New Housing Estate--1939, London, George Allen & Unwin Ltd., 1942, p. 15 (using a minimum standard which "closely followed" Tout's minimum standard for the University of Bristol Survey--and thus will also have been "very nearly equal to" George's poverty line).

80. Flora et al., pp. 397-398.

81. Atkinson, "Poverty in Britain from the 1930s...", pp. 49, 50, 52, and 56-60. Atkinson adjusted figures for price changes using official retail price indices, except for the 1938-1955 period. For that period, because of weighting problems with the official index, he made two computations, one using the official index and one using an index prepared by Tony Lynes. (These alternative computations were one reason for his having a range of ratios rather than a single ratio for each comparison.) He excluded the housing component from both sets of indices that he used [to be consistent with the fact that the poverty lines in question did not include a budget allowance for rent, as noted above on p. 28]. The other reason for his having a range of ratios rather than a single ratio for each comparison was that the various poverty lines for a five-person family varied depending on the ages of the children. (As Atkinson pointed out, the ratios would have been different for a different family size because the various poverty lines had different implicit equivalence scales.)

82. For the 1931 and 1937 figures (and for the 1900-price figure for 1960 referred to below), see Flora et al., pp. 397 and 400. I calculated 1960 and 1981 real gross domestic product per capita figures in 1990 prices from figures in Organisation for Economic Co-operation and Development, Statistics Directorate, National Accounts[--]Main Aggregates[--]Volume I...1960-1993, 1995, pp. 80-81 and 158-159. I converted the 1981 figure to 1900 prices by using the ratio of the 1900-price and 1990-price figures for 1960.

83. Fiegehen et al., pp. 130-131. (For technical details concerning the price indices used, see footnote 74.) I calculated the index figures in the third column directly from the figures in the second column (replacing the index figures in Fiegehen et al., which were calculated on the basis 1899=100).

84. Flora et al., p. 397.

85. Sean Stitt and Diane Grant, Poverty: Rowntree Revisited, Aldershot, England, Avebury, 1993; see especially pp. ix-x, 1-7, 10, 13-14, 46, 58-59, 63, 84-96, 100-101, 110-111, 113, and 116. (The direct quotation is from p. 110.) The poverty line amount is given as 129.78 pounds on pp. 110 and 113, but as 129.31 pounds on p. 100. The discrepancy seems to be in the figures for water rates, given as 2.15 pounds in the summary table on p. 100 but as 2.62 pounds in the detailed discussion on p. 89. I am assuming that the 2.62 and 129.78 pounds figures are the correct ones. (Because of the small magnitude of the discrepancy, my conclusion about the 1950-1992 income elasticity of the poverty line would not be changed even if the lower figures were the correct ones.)

86. B. Seebohm Rowntree and G.R. Lavers, Poverty and the Welfare State[:] A third social survey of York dealing only with economic questions, London, Longmans, Green and Co., 1951, pp. vi and 28.

87. I used the official retail price index for calendar year 1950; theoretically, I should have used the monthly figure for October 1950, but that was not feasible. I used retail price index figures that included the housing component, since a series recalculated to exclude the housing component was not available to me. Since the 1950 retail price index figure that I used was given to only two significant digits, I calculated the 1950-price value of the Stitt/Grant poverty line to only two significant digits.

88. For the 1950 figure, see Flora et al., p. 397. I calculated the 1900-price figure for 1992 in the same way as the corresponding figure for 1981 discussed in footnote 82.

89. Jonathan Bradshaw (editor), Budget Standards for the United Kingdom (Cash & Care Series), Aldershot, England, Avebury, 1993. For the "modest-but-adequate" budget, see especially pp. 1-5 and 172-176. For the "low cost" budget, see Autumn C.S. Yu, "The low cost budget" (Chapter 13), pp. 196-215 in this book.

90. Fiegehen et al., pp. 130-131. (For technical details concerning the price indices used, see footnote 74.) I calculated the index figures in the third column directly from the figures in the second column (replacing the index figures in Fiegehen et al., which were calculated on the basis 1899=100).

91. See footnote 77.

92. Atkinson, "Poverty in Britain from the 1930s...", pp. 49, 50, 52, and 56-60. (For technical details concerning Atkinson's calculations, see footnote 81. In this case, the Supplementary Benefits scale rate for a five-person family varied depending on the ages of the children, but Rowntree's 1936 HNOL poverty line did not.)

93. For the 1936 figure, see Flora et al., p. 397. For the 1981 figure, see footnote 82.

94. For instance, in Peter Townsend, "The Meaning of Poverty," British Journal of Sociology, Vol. 13, No. 3, September 1962, p. 219; and in Victor R. Fuchs, "Comment" [on Lenore A. Epstein, "Measuring the Size of the Low-Income Population"], p. 198 in Lee Soltow (editor), Six Papers on the Size Distribution of Wealth and Income (Studies in Income and Wealth, Volume Thirty-Three, by the Conference on Research in Income and Wealth), New York, National Bureau of Economic Research, 1969.

95. Adam Smith, An Inquiry into the Nature and Causes of the Wealth of Nations (edited by Edwin Cannan), New York, Modern Library, 1937, pp. 821-822 (Book V, Chapter II, Part II, Article 4th). (The text in this edition is that of the fifth edition, of 1789.)

96. J[ohn] T[owne] Danson, Thirteen Short Lectures on the Political Economy of Daily Life...., Liverpool, Henry Young, [1876], p. 47. (I did not try to track down the "recent parliamentary paper" to which Danson referred.)

97. Bonamy Price, Chapters on Practical Political Economy[,] Being the Substance of Lectures Delivered in the University of Oxford, London, C. Kegan Paul & Co., 1878, p. 238. (Price did not make things easy for later readers; he cited his source simply as "Lectures," by "Mr Danson," and cited the page incorrectly.) Price's discussion, in turn, was summarized by an American, Joseph Cook, in a lecture published in 1880; see Joseph Cook, Boston Monday Lectures. Labor, With Preludes on Current Events., Boston, Houghton, Mifflin and Company, 1880, pp. 198-199.

98. Flora et al., pp. 399-400. The figure for 1875 was 35 pounds (in 1900 prices), while the figure for 1830 was 19 pounds. (The 1830 figure was based on National Product rather than Domestic Product.) The 1815-1875 national increase was presumably somewhat greater than the 1830-1875 increase. Moreover, the standard of living in London may have risen more rapidly over this period than the standard of living for the country as a whole.

99. Alfred Marshall, Principles of Economics (first edition), London, Macmillan, 1890, Vol. I, pp. 121 and 122.

100. Charles Booth, Life and Labour of the People in London[:] Final Volume[--]Notes on Social Influences and Conclusion, New York, AMS Press, Inc., 1970 [originally published in London in 1903], p. 201.

101. Note also that in an earlier description of his "poor" and "very poor" groups, Booth described the latter as "those whose means are insufficient for [decent independent life] according to the usual standard of life in this country" (Charles Booth et al., Labour and Life of the People. Volume I: East London., London, Williams and Norgate, 1889, p. 33 (emphasis added)). The underlined phrase suggests at least the possibility that if "the usual standard of life in this country" increased significantly, Booth would have agreed that an increase in "the line of poverty" (the standard of living at which people were considered poor) should be made.

102. A[rthur] L. Bowley and Margaret H. Hogg, Has Poverty Diminished? A Sequel to "Livelihood and Poverty", London, P.S. King & Son, Ltd., 1925. For the dates of the earlier and later studies in the individual towns, see p. 17.

103. Bowley and A[lexander] R. Burnett-Hurst, Livelihood and Poverty[:] A Study in the Economic Conditions of Working-Class Households in Northampton, Warrington, Stanley and Reading, London, G. Bell and Sons, Ltd., 1915. Bowley and Burnett-Hurst did an economic investigation of working-class households in a fifth town, Bolton, which was incorporated in a later edition of the book; see Bowley and Hogg, p. vii.

104. Bowley and Hogg, pp. 12, 14, and 35-36.

105. D[avid] Caradog Jones, Social Surveys, London, Hutchinson's University Library, 1949, pp. 89-90. The table of contents of the third volume of the New Survey of London Life and Labour (see next footnote) indicates (pp. x-xi) that Bowley was the author of the chapter on the poverty line.

106. London School of Economics and Political Science, The New Survey of London Life & Labour[:] Volume III[--]Survey of Social Conditions[--](1) The Eastern Area (Text), London, P.S. King & Son, Ltd., 1932, pp. 10-11, 71-73, and 129-130. For comments on the degree to which Bowley was successful in setting the New Survey poverty line equal in constant shillings to Booth's poverty line, see E.P. Hennock, "Concepts of poverty in the British social surveys from Charles Booth to Arthur Bowley," pp. 211-212 in Martin Bulmer, Kevin Bales, and Kathryn Kish Sklar (editors), The social survey in historical perspective, 1880-1940, Cambridge, Great Britain, Cambridge University Press, 1991; and Colin A. Linsley and Christine L. Linsley, "Booth, Rowntree, and Llewel[l]yn Smith: a reassessment of interwar poverty," Economic History Review, Vol. 46, No. 1, February 1993, pp. 88-92.

107. They did include a possibly suggestive half-sentence ("Whether or no the standard so defined is the best that can be devised for any one date...."--Bowley and Hogg, p. 14), but did not expand on it. In another book published in 1915, Bowley had stated that the housing component of a theoretical standard of minimum subsistence "is, of course, rising, as ideas of housing change and local authorities enforce more drastic regulations," and that the non-food components of such a minimum standard "are really conventional...and are dominated by the standard of living otherwise determined at the time and place" (see Bowley, The Nature and Purpose of the Measurement of Social Phenomena, London, P.S. King & Son, Ltd., 1915, pp. 173 and 174); however, Bowley did not generalize these comments to the minimum standard or poverty line as a whole.

108. Due to the post-World-War-I slump, real gross domestic product per capita during the years of the second study was the same as or slightly lower than during the years of the first study. (In 1900 prices, the figure was 47 pounds for 1912, 49 pounds for 1913 and 1914, 46 pounds for 1923, and 47 pounds for 1924. See Flora et al., p. 398.) For these particular years, then, a constant-price poverty line would not have been inconsistent with the concept of the income elasticity of the poverty line.

109. "Discussion on Sir H. Llewellyn-Smith's Paper," Journal of the Royal Statistical Society, Vol. 92, Part IV, 1929, p. 555.

110. Bowley, "The Poverty Line," p. 72 in The New Survey of London Life & Labour[:] Volume III.... In the previous paragraph (p. 71), he had written, "We have brought the London 1890 [Booth] standard up to date by a study of price changes...."; this and the intervening sentences show that when he wrote of a "higher" minimum standard, he meant a minimum standard that was higher in real terms.

111. Concerning "the general rise in the standard of living...before the war, and...since 1914," note that the real gross domestic product per capita figure for 1929 (52 pounds, in 1900 prices) was 30 percent higher than the corresponding figure for 1890 (40 pounds). (See Flora et al., p. 398.) Cf. Llewellyn Smith's statement (see next paragraph) that the real purchasing power of the average London workman rose by one third over this period.

112. See footnote 1. In the preceding sentence, Bowley had referred to "a careful assessment of [Booth's] Poverty Line at the prices of the year 1929" for the New Survey--indicating that this was the "definition" of poverty that needed to be replaced. However, despite his statement, Bowley did not actually draw up a new definition of poverty, noting that doing that "was found to be impracticable in the absence of Budgets of expenditure [from the New Survey] and of any accepted definition." Instead, he made a modest revision in the New Survey poverty line which only reflected changes in the amount of milk and fresh fruit and vegetables required for children in the light of recent dietary studies.

113. H[ubert] Llewellyn-Smith [sic], "The New Survey of London Life and Labour," Journal of the Royal Statistical Society, Vol. 92, Part IV, 1929, p. 537.

114. Llewellyn Smith, "Introduction," pp. 11-12 in The New Survey of London Life & Labour[:] Volume III....

115. Llewellyn Smith, "Introduction," p. 2 in London School of Economics and Political Science, The New Survey of London Life & Labour[:] Volume VI[--]Survey of Social Conditions[--](2) The Western Area (Text), London, P.S. King & Son, Ltd., 1934.

116. Josiah Wedgwood, The Economics of Inheritance, Port Washington, New York, Kennikat Press, 1971, pp. 39-40. (The original edition was published in London in 1929.)

117. "Discussion on Mr. [D. Caradog] Jones's Paper ['The Social Survey of Merseyside...']," Journal of the Royal Statistical Society, Vol. 94, Part II, 1931, pp. 256 and 258-259.

118. D[avid] Caradog Jones (editor), The Social Survey of Merseyside... Volume One, University Press of Liverpool, Hodder & Stoughton Ltd., London, 1934, pp. 148-149 (emphasis in original).

119. Rowntree, Poverty and Progress[:] A Second Social Survey of York, London, Longmans, Green and Co., 1942 (second edition [slightly revised]), p. 461. (Concerning the precise date of the survey, see p. vii, footnote 1.)

120. D. Chapman, "York - laboratory of social welfare" (manuscript), 1951, p. 7--quoted in Atkinson, "Poverty in Britain from the 1930s...", p. 41.

121. Herbert Tout, The Standard of Living in Bristol[:] A Preliminary Report of the Work of the University of Bristol Social Survey, Bristol, J.W. Arrowsmith Ltd., 1938, p. 14.

122. William Beveridge, Social Insurance and Allied Services [American edition], New York, Macmillan, December 1942, paragraph 27 (p. 14). (For an account of how Beveridge actually developed his "subsistence" benefit amounts, see Veit-Wilson, "Muddle or Mendacity?...".)

123. Norman MacKenzie, "Poverty-Line," New Statesman and Nation, Vol. 47, No. 1208, May 1, 1954, p. 553.

124. Peter Townsend, "The Double Edge of Socialism," New Statesman, Vol. 65, No. 1675, April 19, 1963, p. 568.

125. E[ric] J. Hobsbawm, "Poverty," p. 402 in David L. Sills (ed.), International Encyclopedia of the Social Sciences, Volume 12, Macmillan Company and The Free Press, 1968 (emphasis in original).

126. Royal Commission on the Distribution of Income and Wealth, Lower Incomes (Report No. 6, Cmnd 7175), London, Her Majesty's Stationery Office, 1978, p. 3--quoted in David Piachaud, "Poverty in Britain 1899 to 1983," Journal of Social Policy, Vol. 17, Part 3, July 1988, p. 338.

127. Stewart Lansley and Stuart Weir, "Towards a popular view of poverty," New Society, Vol. 65, No. 1084, August 25, 1983, p. 283. The authors were aware that Booth did not, in fact, develop his "line of poverty" by drawing up and costing out a list of subsistence needs. For how Booth actually operationalized his concept of poverty, see T[homas] S. Simey and M[argaret] B. Simey, Charles Booth[,] Social Scientist [original edition published in 1960], Westport, Connecticut, Greenwood Press, 1980, pp. 95, 184, 187, and 278-279. (See also Llewellyn-Smith, "The New Survey of London Life and Labour," p. 532; and Paul Spicker, "Charles Booth: the examination of poverty," Social Policy & Administration, Vol. 24, No. 1, April/May 1990, pp. 21-38.)

128. Joanna Mack and Stewart Lansley, Poor Britain, London, George Allen & Unwin Ltd., 1985, pp. 55-56.

129. John Moore, The End of the Line for Poverty, London, Conservative Political Centre, May 1989, pp. 7-8.

130. T[im] Callan, D.F. Hannan, B[rian] Nolan, and B[rendan] J. Whelan, "Measuring Poverty in Ireland: A Reply," Economic and Social Review, Vol. 20, No. 4, July 1989, pp. 361-362.

131. Herbert Brown Ames, "The City Below the Hill"[--]A Sociological Study of a portion of the City of Montreal, Canada., Montreal, Canada, Bishop Engraving and Printing Company, 1897. (Ames described this study as a "pamphlet" (p. 5).) For a modern reprint, see Ames, The city below the hill (with an introduction by P.F.W. Rutherford) [the Social History of Canada series], Toronto, Canada, and Buffalo, New York, University of Toronto Press, 1972. (The 1972 edition also includes the text of a November 1897 lecture by Ames, "The housing of the working classes.") For a summary of the study, see Ames, "Incomes, wages, and rents in Montreal," Bulletin of the [U.S.] Department of Labor, Vol. 3, No. 14, January 1898, pp. 39-51. The copy of the 1897 pamphlet in the library of the U.S. Department of Labor is inscribed with the name of Oren W. Weaver (identified in the 1898 Bulletin of the Department of Labor as the Chief Clerk of the Department [actually the Bureau] of Labor); this copy includes a brief note signed "O.W.W." stating that "[t]his little volume came to me from the author" and suggesting that it be summarized as an article in the Bulletin.

132. I am using "$C" to refer to Canadian dollars in order to avoid possible confusion between Canadian dollars and U.S. dollars.

133. Strictly speaking, of course, the annualized equivalent of $C20.00 a week should have been given as $C1,040, using the same 52-week year implicit in the income figures given for the poor.

134. Ames, "The City..."... (1897 edition), pp. 4-5, 7, 14, 22-23, and 49-50. (In the 1972 edition, see pp. 6-7, 9, 22, 32-33, 36, and 68.) See also Ames, "The housing of the working classes," pp. 102, 107, 110, and 112-115 in the 1972 edition.

135. Ames, "Incomes...", pp. 42-43.

136. Note that just two years later an American, W.E.B. DuBois (one of the most prominent Afro-American scholars and leaders of the twentieth century), used a poverty line that was clearly Booth's London poverty line converted into U.S. dollars. (See Fisher, "From Hunter to Orshansky...", pp. 12-13.) In The Philadelphia Negro[:] A Social Study (1899), DuBois used a poverty line of $5 (U.S.) a week. His presentation not only mentioned Booth by name but also compared some of Booth's income/wage-class data with Philadelphia income distribution data. It is of interest that DuBois had published an earlier study--on Negroes in rural Virginia--in the same 1898 issue of the Bulletin of the Department of Labor as Ames' summary article; see W.E. Burghardt Du Bois [sic], "The Negroes of Farmville, Virginia: A social study," Bulletin of the Department of Labor, Vol. 3, No. 14, January 1898, pp. 1-38. DuBois' 1898 study did not use a poverty line as such, but did describe (pp. 27-29) a hypothetical rural family of five persons with an annual income of $174 as "in poor circumstances," and a similar family with an annual income of $221.40 as "in moderate circumstances."

137. Brief references by Ames to "careful observers and honest thinkers... those who study city life" and "the slums of London" were probably allusions to Booth. (See Ames, "The City..."... (1897 edition), p. 4--in the 1972 edition, p. 7.) In the introduction to the 1972 edition, P.F.W. Rutherford wrote (p. xiii) that "It seems that The City below the Hill was modelled upon some of the more ambitious of these [American and British] works, notably Charles Booth's mammoth analysis of life in London"; however, Rutherford did not discuss the relationship between Ames' and Booth's poverty lines.

138. Charles Booth et al., Labour and Life of the People. Volume I: East London., London, Williams and Norgate, 1889, p. 33. (The series title used for this edition varied slightly from the one used later--Life and Labour of the People in London.)

139. Canada, Dominion Bureau of Statistics, General Statistics Branch, The Canada Year Book[,] 1925..., Ottawa, Canada, F.A. Acland (Printer to the King's Most Excellent Majesty), 1926, pp. 818-819; and Statistics Canada, 1972 Canada Year Book..., Ottawa, Canada, 1972, p. 1252.

140. Booth et al., Labour and Life...Volume I..., pp. 33-62.

141. Canadian Consumer Price Index data are available for years from 1913 to the present, while a Canadian family budget price index is available for selected years during the 1900-1913 period. (For the latter, see B.R. Mitchell, International Historical Statistics[--]The Americas and Australasia, Detroit, Michigan, Gale Research Company, 1983.) I made an assumption that there was no change in Canadian retail prices from 1896 to 1900--partly because no data on that period were available to me, but also partly because a price index series for the closest major neighboring country (the Rees Cost-of-Living Index for the United States) showed no net change from 1896 to 1900. However, note that no conceivable 1896-1900 change in Canadian retail prices could have made Ames' poverty line higher in real terms than Podoluk's low income cut-off, and even a sizable change during that short period would at most have made Ames' "well-to-do" income line slightly higher rather than somewhat lower than Podoluk's low income cut-off.

142. Podoluk's original low income cut-offs were developed on the basis of consumption data from the 1959 Survey of Family Expenditures, and thus are referred to as "1959-based." However, they were first applied to income data for 1961, and were thus originally expressed in 1961 dollars.

143. The average family size in Ames' study area, excluding lodgers, was 4.6 persons. (See Ames, "The City..."... (1897 edition), p. 17--in the 1972 edition, pp. 25 and 28.) Since the nearest integral equivalent of that family size is five persons, Ames' poverty line was compared with the Podoluk low income cut-off applicable to a family of five persons.

144. Margaret S. Gould, pp. 44-45 [verbal testimony] and p. 82 [written statement], in [Canadian] House of Commons, Select Standing Committee on Industrial and International Relations, Minutes of Proceedings and Evidence [on a minimum wage resolution] (Session 1926), Ottawa, Canada, F.A. Acland (Printer to the King's Most Excellent Majesty), 1926.

145. For the union's reference to the Martel budget, see Gould, p. 82 [written statement], in [Canadian] House of Commons, Minutes... [on a minimum wage resolution]; the budget was also briefly referred to in verbal testimony by Charles William Bolton on p. 25 of the same Parliamentary document. Neither witness gave a specific citation for the budget, but I was finally able to track it down through a reference in the Canadian Department of Labour's Labour Gazette. See Arthur Martel, "A Minimum Standard of Living in Canada," Canadian Congress Journal [official magazine of the Trades and Labor Congress of Canada], Vol. 1, No. 2, February 1922, pp. 85-91.

146. Dorothy W. Douglas, "A Description of Standards of Living," p. 286 in Paul H. Douglas, Curtice N. Hitchcock, and Willard E. Atkins, The Worker in Modern Economic Society, Chicago, University of Chicago Press, 1923. (For the classification of budget levels or standards of living developed by the Douglases (including the equivalence of their "minimum of subsistence" level with our concept of poverty), see Fisher, "From Hunter to Orshansky...", pp. 10 and 32-33.) On p. 25 of the Parliamentary document cited in footnote 144, Charles William Bolton referred to the Martel budget as being at the Douglases' "Health and Decency" level.

147. Ornati, p. 147. The figure there for a four-person family was converted to a figure for a five-person family by multiplying by the factor of 1.14 cited in Ornati, p. 154.

148. The union explicitly stated, "The items of goods and services and the quantity of each required (which is all that a commodity budget sets forth) vary little as between many parts of the United States and Canada...."--see Gould, p. 82 [written statement], in [Canadian] House of Commons, Minutes... [on a minimum wage resolution].

149. Charles William Bolton, pp. 11, 17, and 22-26 [verbal testimony] and pp. 27-36 ([written] "Appendix to Evidence..."), and Gould, pp. 38-39, 44-45, and 49-50 [verbal testimony] and pp. 82-84 [written statement], in [Canadian] House of Commons, Minutes... [on a minimum wage resolution].

150. Howard T. Falk, pp. 122, 126, and 133 [verbal testimony] and pp. 127-130 [written statement] in [Canadian] House of Commons, Select Standing Committee on Industrial and International Relations, Report, Proceedings and Evidence...upon the question of Insurance against Unemployment, Sickness and Invalidity... (Second Session of the Fourteenth Parliament of Canada), Ottawa, Canada, F.A. Acland (Printer to the King's Most Excellent Majesty), 1928. (The two extended quotations were both found on p. 128.) The Montreal budget was also briefly discussed on pp. 161-162 of Leonard C. Marsh, in collaboration with A. Grant Fleming and C.F. Blackler, Health and Unemployment[:] Some Studies of Their Relationships (McGill Social Research Series: No. 7), Toronto, Canada, Oxford University Press (for McGill University), 1938; the monthly cost of the budget was given for 1926 and 1936.

151. The Cost of Living[:] A Study of the Cost of a Standard of Living in Toronto which Should Maintain Health and Self-respect, Toronto, Canada, The Welfare Council [of Toronto], 1939. The quotations given were found on the cover and on pp. 43 and 1, and the purpose of the study was discussed on p. 1. Associated with this budget study was a study of the adequacy of the diets of one hundred non-relief low-income families in Toronto which found that all but three of those families did not secure nutritionally adequate diets. For an illuminating and insightful article which describes the historical background and context of these studies, see James Struthers, "How Much is Enough? Creating a Social Minimum in Ontario, 1930-44," Canadian Historical Review, Vol. 72, No. 1, March 1991, pp. 39-83--reprinted (with revisions) as Chapter 3, pp. 77-116 in Struthers, The Limits of Affluence: Welfare in Ontario, 1920-1970, Toronto, Canada, University of Toronto Press, 1994. Struthers found that the people who performed and supported these two studies were predominantly advocates of the poor; they were mainly women, and many of them either were social workers or worked in related professions. (For somewhat similar findings concerning people who developed poverty lines and family budgets in the U.S. during the 1904-1965 period, see Fisher, "From Hunter to Orshansky...", especially pp. 11-12 and 73-74.)

152. For a brief discussion of these food plans, see Fisher, "From Hunter to Orshansky...", pp. 35-36.

153. L[eonard] C. Marsh, Report on Social Security for Canada (Prepared... for the Advisory Committee on Reconstruction), Ottawa, Canada, Edmond Cloutier (Printer to the King's Most Excellent Majesty), 1943, pp. 19-21 and 23. For a modern reprint of this report, see Leonard Marsh, Report on social security for Canada (with a new introduction by the author and a preface by Michael Bliss) [the Social History of Canada series], Toronto, Canada, University of Toronto Press, 1975; for the discussion of the 1939 Toronto budget in this edition, see pp. 32-34, 36-37, 39, and 44. I am grateful to Lars Osberg for another reference which indirectly led me to Marsh's 1943 report.

154. See Fisher, "From Hunter to Orshansky...", pp. 38-39.

155. The Cost of Living..., p. 42. This cost is misprinted twice on p. 43 as $[C]29.65; in the copy of this pamphlet in the U.S. Department of Labor's library, both instances have been corrected in ink to $[C]28.35. (Marsh, who mentioned having collaborated with the Executive Director of the Toronto Welfare Council, used the $C28.35 figure. See Marsh, Report... (1943 edition), pp. 5 and 20-21 (pp. 4, 37, and 39 in the 1975 edition).) The cost of the budget is specified as being for a family comprising a man, a woman, a boy aged 12, a girl aged 10, and a boy aged 6 (see The Cost of Living..., pp. 42 and 43).

156. The Toronto Welfare Council pamphlet gave only a weekly cost for the budget, while Marsh's 1943 report gave 1939 costs for this budget in both weekly and monthly terms. I multiplied Marsh's monthly figure by twelve to get an annual figure; the result turned out to be exactly 52 times the weekly cost.

157. Marsh may well have made use of his familiarity with the Montreal minimum budget (see footnote 150) in developing his Assistance Minimum standard. As noted in Marsh et al., Health and Unemployment..., p. 161, the 1936 monthly cost of the Montreal minimum budget for a five-person family was $C83.33. This amount was equal to about $C85.80 in 1939 dollars, while the 1939 monthly cost of Marsh's Assistance Minimum standard for a five-person family (as noted in the text) was $C88.36.

158. Marsh, Report... (1943 edition), pp. 21 and 23 (pp. 37-39 and 41 in the 1975 edition).

159. Marsh, Report... (1943 edition), p. 21 (p. 39 in the 1975 edition).

160. [Canada,] Royal Commission on Health Services, Canadian Economic Growth (by T.M. Brown), Ottawa, Canada, Roger Duhamel, Queen's Printer and Controller of Stationery, 1965, pp. 204-205.

161. Ornati, pp. 141 and 145.

162. See J[enny] R. Podoluk (Central Research and Development Staff, Dominion Bureau of Statistics), "Characteristics of Low Income Families" [an unpublished paper prepared for the Federal-Provincial Conference on Poverty and Opportunity held in Ottawa], November 1965; and Jenny R. Podoluk, Incomes of Canadians (one of a series of studies in the 1961 Census Monograph Programme), Ottawa, Dominion Bureau of Statistics, 1968, Chapter 8.

163. Since at least 1974, Statistics Canada has made information about the Low Income Cut-Offs and the low income population available in a bilingual (English and French) format. (See Statistics Canada, Consumer Income and Expenditure Division, Statistics on Low Income in Canada[/]Faibles revenus au Canada[,] 1969 [Catalogue 13-553 occasional], Ottawa, Canada, Information Canada, January 1974.) The French equivalent of the English "Low Income Cut-Offs (LICOs)" is "seuils de faible revenu (SFR)." I have not capitalized the term "Low Income Cut-Offs" when referring to Podoluk's original work because there was still some variation in Statistics Canada's terminology for this concept as late as 1974 (see p. 7 of the publication just cited); Podoluk herself did not use this specific term in her original 1965 paper or her 1968 Census Monograph, although she did refer once in the former (p. 6) to "an income cut-off."

164. Wolfson and Evans, pp. 12-15 and 31-33.

165. Beginning with this revision, Statistics Canada calculated LICOs that varied not only by family size but also by size of area of residence (rural areas, and urban areas with population less than 30,000, 30,000-99,999, 100,000-499,999, and 500,000 and over). On that basis, Statistics Canada generally publishes five separate LICO figures for each family size. For their 1989 paper, Wolfson and Evans calculated and published (p. 15) LICO figures averaged across the five size-of-area-of-residence categories for the 1969-based through 1986-based LICOs (in constant (1986) dollars), providing a single figure for each family size for each revision to facilitate comparisons of the revisions.

166. For the real average family income figures cited in this and subsequent paragraphs, see Statistics Canada, Household Surveys Division, Income distributions by size in Canada[,] 1993 [bilingual] (Catalogue No. 13-207), Ottawa, Canada, December 1994, p. 17. (The figure for 1959 excludes self-employment farm income.)

167. "Appendix[:] Results of the Low Income Review," pp. 167-174 in Statistics Canada, Household Surveys Division, Income distributions by size in Canada[,] 1990 [bilingual] (Catalogue 13-207), Ottawa, Canada, December 1991.

168. Wolfson and Evans, p. 15.

169. Statistics Canada, Household Surveys Division, Income distributions by size in Canada[,] 1992 [bilingual] (Catalogue No. 13-207), Ottawa, Canada, December 1993, pp. 42-43.

170. These comparisons were made on the basis of the 1992-based and 1986-based LICOs for 1993 shown on p. 45 of the report for 1993 cited in footnote 166. The corresponding figures for 1992 (on p. 43 of the report for 1992 cited in footnote 169) were not used because an errata notice indicated that many of those figures were incorrect.

171. Christopher A. Sarlo, Poverty in Canada, Vancouver, British Columbia, Canada, The Fraser Institute, 1992. See also Fred Langan, "Who Qualifies as Poor in Canada?", Christian Science Monitor, Vol. 84, No. 192, August 26, 1992, p. 7; and Lars Osberg, [review of Poverty in Canada by Christopher A. Sarlo], Canadian Public Policy/Analyse de Politiques, Vol. 19, No. 2, June 1993, pp. 222-223.

172. Sarlo, pp. 3, 19, 25-27, 29-32, 45-47, 49-51, 98-107, and 196-199.

173. For this weighted average LICO, see Wolfson and Evans, p. 15. As noted above (footnote 165), Statistics Canada generally presents LICOs by five size-of-area-of-residence categories, ranging from rural areas to urban areas of 500,000 and over, without giving a single weighted average LICO for each family size. Sarlo makes a somewhat different comparison in his book (p. 3), comparing his 1988 poverty line with the 1978-based four-person LICO for 1988 (i.e., in 1988 dollars) for urban areas of between 100,000 and 499,999 persons; his poverty line can be shown to be equal to 59 percent of this particular LICO.

174. Fisher, "From Hunter to Orshansky...", p. 74.

175. In her foreword, Friedman supported "a stable measure of the 'basic needs' of families..."--an "absolute standard" of poverty. While she quoted several sentences from her 1965 pamphlet, she did not quote the sentences in that pamphlet in which she recognized that the "standard of poverty" rises in real terms over time as the real income of the general population rises. (For those sentences, see p. 23 of this paper.)

176. Sarlo, pp. 3 and 115. For four-person families, Sarlo's overall poverty line was between the poverty lines for Prince Edward Island and Newfoundland, being somewhat closer to the latter.

177. Alex C. Michalos, "Discrepancies Between Perceived Income Needs and Actual Incomes," Social Indicators Research, Vol. 21, No. 3, June 1989, pp. 293-296.

178. Note that the wording is almost identical to that of the U.S. Gallup Poll question as quoted in Kilpatrick (see p. 12 above).

179. Interpretation of the trends is complicated by the fact that the weekly "get-along" amounts (described by Michalos as "average[s]"--i.e., presumably means) were "sticky," in several cases remaining at the same current-dollar round figure ($C200 or $C300) for several years in succession before moving up to a higher figure. Rainwater reported a similar phenomenon for U.S. Gallup "get-along" data, although the figures he was discussing were medians rather than means--see Rainwater, "Poverty and Equivalence...", p. 5, footnote 3.

180. In Canada, much as in the U.S., real family income rose significantly before 1973 but has generally fluctuated up and down within a relatively narrow range since that date. The real average income of Canadian families (in 1993 dollars) rose from $C21,882 in 1951 to $C46,136 in 1973 and $C49,044 in 1975. From 1976 through 1993, it fluctuated between $C51,126 and $C57,281. (See the source cited in footnote 166. The figure for 1951 excludes self- employment farm income.) If "get-along" data for years before 1973 exist and could be added to the data to be regressed against real median or mean family income, it would probably yield stronger evidence of the relationship between the two variables. (Ideally, one might use real income data for four-person families, as Vaughan did in his article using U.S. data--see p. 16 above.)

181. Podoluk, "Characteristics of Low Income Families," pp. 3-4. For a published version of these comments (with minor changes), see Podoluk, Incomes of Canadians, pp. 183-184. This published version included an additional comment: "In the postwar years it is obvious that an automobile has joined the list of family 'needs'; in pre-war years it would still have been considered a luxury item." (Cf. the American comment quoted in footnote 205 below.)

182. Special [Canadian] Senate Committee on Poverty, Poverty in Canada[:] Report..., Ottawa, Information Canada, 1971, p. 2.

183. "Revision of Low Income Cut-offs," p. 122 in Statistics Canada, Income distributions by size in Canada[,] 1980 [bilingual] (Catalogue 13-207), Ottawa, June 1982.

184. John M. Evans, "Measuring low incomes: The experience of Statistics Canada," p. 231 in Polish Statistical Association and [Polish] Central Statistical Office, Poverty Measurement for Economies in Transition in Eastern European Countries[--Papers Presented at an] International Scientific Conference, Warsaw, 7-9 October 1991, Warsaw, Poland, 1992.

185. [Australia,] Royal Commission on the Basic Wage, Report..., Albert J. Mullett (Government Printer for the State of Victoria), 1920. For the historical background and context of this report, see Paul H. Douglas, "Wages Regulation and Children's Maintenance in Australia," Quarterly Journal of Economics, Vol. 37, No. 4, August 1923, pp. 643-686; see also H. Heaton, "The Basic Wage Principle in Australian Wages Regulation," Economic Journal, Vol. 31, No. 123, September 1921, pp. 309-319.

186. I am using "$A" to refer to Australian dollars in order to avoid possible confusion between Australian dollars and U.S. dollars.

187. The budget was presented in A[llan] Halladay, Poverty and The Large Family in Sydney, 1968/1969. A Study of Living Conditions and Social Policy (unpublished Ph.D. thesis), Canberra, Australia, Australian National University, 1971, pp. 148 and 440-442; I found the citation for it in Halladay, "The Extent of Poverty Among Large Families in the Heart of Sydney," Economic Record, Vol. 48, No. 124, December 1972, p. 488.

188. For the Henderson poverty line, see Ronald F. Henderson, Alison Harcourt, and R.J.A. Harper, People in Poverty: A Melbourne Survey, Melbourne, Australia, F.W. Cheshire Publishing (for the Institute of Applied Economic and Social Research, University of Melbourne), 1970, pp. 1, 24-31, and 208-209; David Johnson, "The Calculation and Use of Poverty Lines in Australia," Australian Economic Review, No. 80, 4th Quarter (Summer) 1987, pp. 45-55; and Anthony King, "Poverty measurement in Australia: towards a new consensus," Australian Quarterly, Vol. 63, No. 3, Spring 1991, pp. 235-246.

189. Australian currency was decimalized in February 1966, with the conversion rate being two Australian dollars to one Australian pound.

190. Royal Commission, Report..., pp. 4, 7-8, 10, 12, and 15-16; Heaton, pp. 309-310; and Paul Douglas, "Wages Regulation...", pp. 646-649.

191. Royal Commission, Report..., pp. 4, 13-14, 17-18, 20, 25-26, 43-44, and 57-58; Royal Commission, Supplementary Report, Albert J. Mullett (Government Printer for the State of Victoria), 1921, pp. 106-111.

192. I made this computation using annual retail price index figures for 1920 and 1966. (I am grateful to Todd Godbout of the U.S. Bureau of Labor Statistics for supplying me with both these Australian price index figures and the British retail price index figures that I used to compare the real values of the Rowntree/Lavers 1950 poverty line and the Stitt/Grant poverty line (see pp. 34-35 above).) Strictly speaking, I ought to have used a monthly retail price figure for November 1920 in this computation, particularly since 1920 was the peak year for post-World-War-I inflation in Australia; however, monthly figures were not available to me.

193. The 1966 Henderson poverty line for a standard family of four was $A33.00 per week. I computed the 1966 poverty line for a five-person family of the composition assumed by the 1919-1920 Commission by using equivalence scale information on pp. 27-29 and 208-209 of Henderson, Harcourt, and Harper. (I assumed that both parents of the five-person family--like those of the Henderson poverty line's standard family--were aged under 40.)

194. Halladay's 1972 article gave the Henderson poverty line for 1969 both for the six-person family that Halladay was using as a standard family and for the four-person family used as a standard family in connection with the Henderson poverty line. (See Halladay, "The Extent of Poverty...", p. 487.) I used that information and the equivalence scale information referenced in the immediately preceding footnote to calculate what Halladay's poverty budget would have been for the standard family of five used by the 1919-1920 Commission.

195. I think that it is safe to assume that Australian living standards were higher in the late 1960's than in 1920, even though I have not yet been able to find figures to verify that assumption.

196. Peter Saunders and Bruce Bradbury, "Some Australian Evidence on the Consensual Approach to Poverty Measurement" (Social Welfare Research Centre Discussion Paper No. 14), Kensington, New South Wales, Australia, Social Welfare Research Centre, University of New South Wales, July 1989. (I am grateful to Denton Vaughan for sending me a copy of this paper and also for giving me the reference to the Michalos article cited in footnote 177.)

197. Note that the descriptive phrase in the Australian poll question--"to keep in health and live decently"--differs from the descriptive phrase in the U.S. and Canadian poll questions--"to get along" (see pp. 12 and 61 above). However, four times during the 1937-1944 period (as noted above on p. 12), the U.S. Gallup Poll had asked similar questions in which the operational phrases were "for health and comfort," "health and decency," or "comfort and decency." A comparison of dates discloses that the Australian question, with its wording reminiscent of the 1937-1944 U.S. questions, was first asked in February 1945, while the "get along" wording was not used in the U.S. until January 1946.

198. Saunders and Bradbury described this question as a "minimum income question" and suggested that responses to it represent "income levels seen as necessary for minimum subsistence"; they used responses to the question to develop a "consensual" ["subjective" or Leyden-style] poverty line, although they noted that "the terms 'health' and 'decency' are possibly more ambiguous for [this purpose] than an evaluation in terms of 'making ends meet'." (Note that in his 1972 article, Halladay used the 1969 response to this question to develop an alternative poverty line; see Halladay, "The Extent of Poverty...", pp. 487-488.) In the context of the U.S. Gallup Poll, Rainwater (What Money Buys..., pp. 50-54) found that the "health/comfort/decency" response was higher than the "get-along" response, while the "get-along" response was close to Ornati's minimum adequacy level [one level above minimum subsistence [=poverty]]. Vaughan reported (p. 25) that the "get-along" response was higher than the response to a Gallup Poll question specifically using the term "poverty line." While one should not assume that what is true for the U.S. is automatically also true for another country such as Australia, these U.S. findings suggest at least the possibility that an Australian poll question using the words "to get along" or "poverty line" might yield a response lower than the "health/decency" response.

199. The question was asked an additional four times during the 1945-1949 period. However, Saunders and Bradbury did not analyze these four observations, as the income variables that they used were not available for that period.

200. Johnson, "The Calculation and Use of Poverty Lines in Australia," p. 47.

201. Henderson, Harcourt, and Harper, p. 29. The comparison made above of the real values of the 1919-1920 Commission's "reasonable standards of comfort" budget and the 1966 Henderson poverty line provides a concrete illustration of the quoted statement, with the one exception that the figure for the "comfortabl[e]" family dates from 1920 rather than 1900.

202. Ian Manning, "The Henderson poverty line in review," Social Security Journal [a publication of the Australian Department of Social Security], June 1982, pp. 8-9.

203. Office of the [Australian] Economic Planning Advisory Council, Income Support Policies, Taxation and Incentives (Council Paper No. 35), Canberra, Australia, Australian Government Publishing Service, October 1988, p. 13--quoted in Saunders and Bradbury, p. 6, footnote 2.

204. See also Peter Townsend, Poverty in the United Kingdom[:] A Survey of Household Resources and Standards of Living, Berkeley and Los Angeles, University of California Press, 1979, pp. 50-53.

205. "Quite obviously, a family which can secure from its income not more than a bare subsistence or even a minimum for health and efficiency [the standard of living immediately above the "bare subsistence level" in Nystrom's classification] cannot afford the luxury of a car"--Paul H. Nystrom, Economic Principles of Consumption, New York, The Ronald Press Company, 1929, p. 303.

206. A second such occasion occurred in 1973, when an interagency Subcommittee on Updating the Poverty Threshold, recognizing that nutritional standards and consumption patterns change over time, recommended that the food plan and the multiplier used to calculate the poverty line be updated every ten years. No action was ever taken on this recommendation. (See Fisher, "The Development of the Orshansky Poverty Thresholds..." (unpublished), pp. 54-55 and 57-58--summarized in Fisher, "The Development and History of the Poverty Thresholds," pp. 9-10.) This recommendation would have resulted in the poverty line being raised in real terms when the general standard of living rose. I am not considering this occasion further in this paper only because I do not have any documentary evidence indicating or suggesting specific reasons why this recommendation was ignored.

207. Fisher, "The Development of the Orshansky Poverty Thresholds..." (unpublished), pp. 2-3, 35-36, and 38-41--summarized in Fisher, "The Development and History of the Poverty Thresholds," pp. 4 and 7.

208. Fisher, "The Development of the Orshansky Poverty Thresholds..." (unpublished), pp. 42-44--summarized in Fisher, "The Development and History of the Poverty Thresholds," p. 7.

209. Fisher, "The Development of the Orshansky Poverty Thresholds..." (unpublished), pp. 42-43 and 45-46--summarized in Fisher, "The Development and History of the Poverty Thresholds," p. 7. For the report on the American Negro which was to be released by the White House, see U.S. Bureau of the Census, Current Population Reports, Series P-23, No. 26/B[ureau of ]L[abor ] S[tatistics] Report No. 347, Recent Trends In Social and Economic Conditions of Negroes in the United States, U.S. Government Printing Office, July 1968.

210. Fisher, "The Development of the Orshansky Poverty Thresholds..." (unpublished), p. 47.

211. October 9, 1968, memorandum from Lawrence N. Bloomberg, Secretary, to Members, Poverty Level Review Committee--Subject: Minutes, Meeting of October 2, 1968.

212. The minutes of the March 9, 1969, meeting made the following indirect reference to this issue: "There was some feeling expressed that there should be no changes in the present procedure for determining poverty levels until either (a) all or part of the data from the 1965 [Department of] Agriculture [Household] food [Consumption] survey are used to update the 1955 data currently used as a base, or (b) a different concept or procedure [for determining poverty levels] is developed" (March 19, 1969, memorandum from Lawrence N. Bloomberg, Secretary, to Members, Poverty Level Review Committee-- Subject: Minutes, Meeting of March 7, 1969). With that one exception, the issue was not mentioned at all in any of the meeting records that I have.

213. Fisher, "The Development of the Orshansky Poverty Thresholds..." (unpublished), pp. 47-49--summarized in Fisher, "The Development and History of the Poverty Thresholds," p. 8.

214. According to a Census Bureau report published in May 1968--the month before the dispute over the report on the American Negro to be released by the White House--the number of persons in poverty had dropped from 35.3 million in 1963 to 29.7 million in 1966; see 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, Table 1, p. 12. (After the 1969 revision in the poverty definition, figures for 1959-1966 were retabulated under the revised definition, so figures shown for 1963 and 1966 in Census Bureau reports printed after 1969 do not match the figures given here.)

215. At that point, neither the food stamp program nor the school lunch program was using the poverty guidelines in determining eligibility; see Congressional Quarterly Inc., Future of Social Programs, Washington, D.C., August 1973, p. 22.

216. Fisher, "From Hunter to Orshansky...", pp. 56-57 and 64-66.

217. S.M. Miller, Martin Rein, Pamela Roby, and Bertram M. Gross, "Poverty, Inequality, and Conflict," Annals of the American Academy of Political and Social Science, Vol. 373, September 1967, p. 20--reprinted on p. 288 in Bertram M. Gross (editor), Social Intelligence for America's Future[:] Explorations in Societal Problems, Boston, Allyn and Bacon, Inc., 1969. Note that in April 1995--almost 28 years after the original article--the United States had still not reached a point at which an upward adjustment in the real level of the poverty line appeared "politically feasible."

218. S.M. Miller and Pamela A. Roby, The Future of Inequality, New York, Basic Books, Inc., 1970, p. 35.

219. Lars Osberg, Economic Inequality in the United States, Armonk, New York, M.E. Sharpe, Inc., 1984, p. 69 [emphasis in original].

220. Margaret Stecker's "emergency" budget for a family of four, published in 1937, was equal to $2,020 in 1963 dollars. (See Fisher, "From Hunter to Orshansky...", pp. 38-39.) Orshansky's poverty threshold for a nonfarm family of four, published and quasi-officially adopted in 1965, was $3,128 in 1963 dollars.

221. William J. Eaton, "The Poverty Line," New York Post, April 4, 1970, p. 24 [magazine page four]. (The comment is an indirect rather than a direct quotation.) Orshansky made a very similar comment in a 1971 Congressional hearing when she referred to "the decisions...made in 1969 to formalize and, in a sense, freeze the index, except for price change." (See Mollie Orshansky, "Statement..." [March 23, 1971], p. 71 in U.S. Senate, Economic Opportunity Amendments of 1971[:] Hearings Before the Subcommittee on Employment, Manpower, and Poverty of the Committee on Labor and Public Welfare...on S. 1290...S. 397...S. 1305...[and] S. 2007...Part 1, Washington, D.C., U.S. Government Printing Office, 1971.)

222. Fisher, "From Hunter to Orshansky...", pp. 1 and 73-75.

223. Mollie Orshansky (personal communications, June 14, 1988, and February 14, 1989) has described a pattern in the field of economics during her professional career in which men tended to be interested in macroeconomics (an area with more professional prestige), while women tended to be interested in distributional or household economics, looking at such subjects as how many families have unmet needs and working with household surveys and family studies--empirical studies with individual observations. See also Fisher, "From Hunter to Orshansky...", pp. 47-48.

224. Fisher, "From Hunter to Orshansky...", pp. 74-75.

225. Walter Korpi, "Approaches to the Study of Poverty in the United States..." (cited in footnote 4), pp. 289-290.


Source: U.S. Census Bureau | Poverty - Experimental Measures |  Last Revised: July 02, 2012