Printer Friendly

A profile of the working poor.

A profile of the working poor

More than 6 million persons who spent at least half of 1987 in the labor force were poor; among families with workers, those headed by unmarried women with children have the highest poverty rates

Policymakers and researchers have been interested for a long time in the link between the labor market problems of workers and the economic status of their families. When workers are either unemployed or limited to part-time work, or when they have to work at very low wages, their personal finances suffer. But the impact on their families varies. In many cases, it may not be severe. At the extreme, however, several million families live below the official poverty level, even when some of their members are active in the work force. This situation is most common when only one family member works and earns low wages.

This article focuses on persons who are labor force participants but live in poor families. The labor market experiences and family circumstances of these persons are contrasted with those of persons in the work force who are not poor. Some general findings are as follows: * The working poor made up about one-third of

all persons age 16 and over who were in

poverty. These are the more than 6 million

persons whose family income was below the

official poverty level in 1987, even though

they worked or looked for work at least half of

the year. * Labor market problems such as

unemployment or the inability to find full-time work are

most likely to cause poverty when they occur

in conjuction with low wages. Two-thirds of

the working poor who usually work full time

have weekly earnings that are below a "low-earnings"

threshold used in this analysis. * The presence of more than one worker in a

family dramatically lowers the probability of

poverty. In particular, poverty is rare in

husband-and-wife families where both are

employed. * Unmarried women maintaining families are

the workers with the greatest risk of living in

poverty. Their earnings are rarely

supplemented by those of other family members,

and their wages, like those of women in

general, are substantially lower than men's.

Almost one-fourth of single-earner families

maintained by women are poor. * Because education has a strong influence on

earnings, individuals with low levels of

schooling are overrepresented among the

working poor. Most at risk are black workers

and women, because, at every level of

education, they have lower earnings than white


Background and definitions

A number of substantial efforts have been made to study issues which have been broadly labeled "economic hardship."(1) The Bureau of Labor Statistics issued annual reports between 1982 and 1987 entitled Linking Employment Problems to Economic Status. In these reports, Current Population Survey (CPS) data were used to provide estimates of the number of workers who had encountered any of a list of labor market problems during a given year, however slight they might have been. The number of such persons totaled 33 million in 1985.(2)

This article focuses on a much smaller universe. It first identifies families living in poverty and then examines the labor market characteristics and problems of the workers in these families.(3) The approach emphasizes the "working poor," a term often used, although with a wide range of meanings. Here, the working poor are defined as persons who have devoted at least half the year to labor market efforts, being either employed or in search of a job during that period, but who still lived in poor families. While the 6-month cutoff is somewhat arbitrary, it is meant to exclude not only nonparticipants in the labor force, but also marginal participants. Such persons may also live in poverty, but their economic problems are not likely to have stemmed primarily from their failures in the workplace, or the failure of the workplace to provide jobs. Among those falling into this category are students who look for work for 1 or 2 weeks before finding summer employment and persons who are ill or disabled most of the year.

The most complicated aspect of the analysis presented here is that the working poor, as a group, owe their poverty status to two sets of circumstances: (1) low earnings, resulting from a range of labor market problems, including unemployment, inability to find full-time work, and low wage rates; and (2) a family structure that is conducive to poverty, such as the presence of dependent children and only one earner. Because the poverty threshold--that is, the amount of money needed to stay out of poverty--is a function of family size, it is actually possible for a "poor" worker to have earned much more than a worker who, because of different family circumstances, is classified as nonpoor. The following are some hypothetical examples of persons whose employment and family characteristics leave them in poverty: * Bob is married, is the father of two children,

worked as a construction laborer, and earned

$5.25 an hour. His wife did not work. Bob

usually worked full time, but because of bad

weather and temporary layoffs, he lost

several weeks of work. He earned $9,555 in

1987, before taxes. * Barbara is a single mother with two children,

worked in a cleaning store, and earned $3.35

an hour, the minimum wage. She worked all

year, except for 2 weeks when the children

were sick, and earned $6,700. * Jane lives alone, worked in a cafeteria 4 hours

a day, and earned $3.75 an hour. She would

have preferred to work full time, but was

unable to find another job. In 1987, she earned

$3,750 before taxes.

Families of four, like Bob's, required at least $11,611 in 1987 to be considered above poverty by the Federal Government's definition. Families of three, like Barbara's, required at least $9,056 to be considered out of poverty, while a person living alone, such as Jane, needed an annual income of $5,909 or more. Thus, assuming that their families had no other sources of income, Bob, Barbara, and Jane would each have been a member of the "working poor."

The working poor

Of the 112 million persons who spent at least half a year in the labor force in 1987, 6.4 million were members of poor families. Thus, the poverty rate among workers was 5.6 percent. Exhibit 1 profiles poor and nonpoor workers, comparing their personal traits, labor market performances, and family situations. Table 1 provides additional detail on the demographic and personal characteristics of workers who are poor and those who are not.

While persons from every age, race, sex, and educational group are found among the working poor, the key variables that relate to poverty among workers are family relationship and education. Family structure largely determines the number of potential wage earners, and education is the best predictor of earnings.

Despite the fact that men's earnings are generally much higher than women's, a working husband had a higher probability of his family's being poor in 1987 (4.2 percent) than did a working wife (2.5 percent). The reason is that husbands are more likely than wives to be the sole support of their families. Black workers have very high poverty rates (13.2 percent, compared to 4.7 percent for whites), largely because they tend to live in family arrangements that are most conducive to poverty. Black men are disproportionately in the group of unrelated individuals (those not living with other relatives), and black women are far more likely than women of other racial or ethnic groups to maintain families themselves. Women who maintain families had the highest poverty rates of any of the major groups shown in table 1--nearly 18 percent. The poverty situation for black workers is exacerbated by their relatively low levels of educational attainment (employed blacks are almost 50 percent more likely than whites not to have completed high school) and the resultant low earnings.

Nonagricultural wage and salary workers made up the bulk of the working poor, although 10 percent had been employed in agriculture, an industry with a poverty rate four times that of the nonagricultural sector. Twelve percent were self-employed, possibly reflecting work in some very small-scale enterprises.(4)

The labor market problems that poor workers experience are quite different from those of nonpoor workers. Nearly half of the working poor experienced unemployment at some time during 1987, while only 1 in 8 of the nonpoor did so. (See table 2.) And the median number of weeks of unemployment was much higher for the poor than the nonpoor workers--26 versus 13 weeks. Also, relative to nonpoor workers, poor workers were nearly four times as likely to have been limited, for at least part of the year, to working part time when they would have preferred full-time work.

Nevertheless, these labor market problems, by themselves, generally did not make workers poor. Among both the unemployed and involuntary part-time workers, nonpoor persons out-numbered poor persons by 4 1/2 to 1. In fact, the nonpoor even predominate among those unemployed for half a year or more.

Aside from experiencing unemployment or being limited to involuntary part-time work, the working poor have a strong tendency to work in jobs that pay low wages. Previous research supports the contention that low pay may be the primary cause of poverty among workers. In a 1976 study, Frank Levy addressed the effect of unemployment, measured in terms of work-hour losses, on the earning levels and poverty status of workers.(5) Levy found that merely increasing the number of hours for which poor workers were paid (at their usual wage rate) would have removed few of their families from poverty status. Many of the workers who had lost work due to unemployment or were unable to get full-time work were in poverty primarily because their jobs paid low wage rates.

The situation reflects the workings of low-wage labor markets. Unemployment and involuntary part-time work do not occur randomly across the earnings spectrum. Unemployment, particularly, is most common among workers who have low-wage jobs, as is evident from data on the rates of joblessness in individual occupations.

The significance of low earnings

Determining the prevalence of low earnings among poor and nonpoor workers involves defining exactly what is meant by low earnings, establishing a cutoff line, and then applying the cutoff to the available data on earnings. Information is collected in the March CPS supplement on annual earnings in the prior calendar year, along with weeks worked and usual hours worked. Using these data, past BLS analyses on labor market hardship have focused on the earnings of full-time, year-round workers, identifying them as "low-wage" workers if their yearly earnings fell below the Federal minimum hourly wage multiplied by 2,000.

The methodology was changed for this article for two reasons. First, we wanted to examine the earnings of workers who worked less than year round, particularly because so many of the working poor experienced unemployment. Second, the number of workers employed at the Federal minimum wage, set at $3.35 since 1981, has been gradually declining as nominal wages have increased. The resulting drop in the number and proportion of minimum-wage workers does not necessarily mean that low earnings are any less of an issue as an employment problem.

The major goal in defining a more relevant low-earnings level was to choose a method that accepted the minimum wage as an important indicator of society's view of low wages, but also allowed analytically meaningful comparisons to be made over time. There is no one method which lends itself to this end, and, certainly, the choice of methodologies largely determines the number of low earners that the analysis will identify. (See the appendix for a discussion of the sensitivity of the number of low earners to several low-wage options.)

The low-wage level chosen for this analysis is an average of the minimum-wage levels in effect from 1967 to 1987, calculated from each year's value, expressed in 1987 dollars.(6) The average minimum-wage value for the entire 21-year period, in 1987 dollars, was $4.18 per hour. Assuming a 40-hour week, this would translate to weekly earnings of $167.20. This figure was then compared with the weekly earnings for each full-time wage and salary worker to determine whether actual 1987 earnings were above or below the "low-earnings" threshold.

About 2.1 million poor full-time wage and salary workers who were in the labor force at least half the year earned the low-earnings level of $167.20 per week or less. To place this measure in perspective, 1.6 million earned the prevailing minimum wage of $3.35 or less, while 2.6 million earned 150 percent of the minimum or less ($5.03). (See the appendix for details on determining the low-earnings figure.)

The data for 1987 indicate that fully two-thirds of the poor who usually worked in full-time wage and salary jobs had earnings at or below the low-earnings threshold. Three-quarters of these low earners had average weekly earnings of $134 or less, which would be the equivalent of earning the minimum wage of $3.35 for a 40-hour week.

Analysis shows that there is considerable evidence that a strong relationship exists between low earnings and poverty status. Two-thirds of poor full-time workers experienced low earnings. Furthermore, even among the poor full-time wage and salary workers who also experienced either unemployment or involuntary part-time work, most had low earnings. (See table 3.) By contrast, the poverty rate was quite low--only 7 percent--among those who had been unemployed but did not also experience low earnings. For those unemployed persons who also had low earnings, the rate was dramatically higher--37 percent. Similarly, those who were forced to work part time at least some of the year even though they would have preferred full-time work had only about a 2-percent chance of being poor if they experienced no other labor market problem. Those who also had low earnings, though, had a poverty rate of 26 percent. Thus, among fulltime workers, low earnings alone are an important contributor to poverty, and they greatly increase the probability of poverty among those with other labor market problems.

While low weekly earnings (stemming from low hourly wage rates) were the most common problem for those working poor who usually worked full time, it should be noted that, as with the unemployed, most low-wage earners were not in poverty. In fact, for each low-wage worker in a poor family, three were in families that were not poor. The poor families were most often those in which no one other than the low earner had worked.

As an illustration, table 4 indicates that, among persons earning low wages, husbands in married-couple families, persons who maintain families without a spouse, and unrelated individuals had the highest probability of being poor. These individuals are more likely than others to be the sole support of their families or households. In contrast, when wives or other persons related to a householder work for low pay, their earnings are generally supplemented by others. Hence, their poverty rates are relatively low.

Overall, a full-time wage and salary worker with low earnings had a 25-percent probability of being poor. By comparison, full-time workers who earned more than the low-earnings level had only a 2-percent chance of being below the poverty level.

Low-earning levels were only estimated for full-time workers because past research has found that weekly earnings calculated from annual data for part-time workers are quite unreliable.(7) Although part-time workers work fewer hours, the hourly wages of poor part-time workers, could, in theory, be higher than those earned by full-time workers. Still, given the relatively low wages paid part-time workers in general, it is reasonable to infer that a large proportion of poor part-time workers also earned wage rates below the $4.18 "low-wage" level.

The group most affected by low wages was women heading families containing children (not shown in table 4). Three-fourths of these women who worked full time at low wages were living below the poverty level. More will be said about these workers later.

Of all readily observable personal characteristics, researchers have repeatedly found education to have the most consistently powerful effect on earnings. Aside from education, only one's family background seems to influence earnings and poverty to a large extent, and it does so primarily indirectly, through its influence on education. Other factors, such as years of work experience and test scores, have been found to have less impact in and of themselves.(8) The following tabulation demonstrates the marked difference in the poverty profiles of workers in 1987 in terms of their levels of education:
 Numbers Percent
 (in millions) poor

Total, in the labor force

27 weeks or more ........ 113.5 5.6

Fewer than 4 years of high school ....... 18.5 13.3

4 years of high
school only ....... 46.0 5.7
1-3 years of college .... 23.1 3.8

4 years of college or more ....... 25.9 1.7

Numerous explanations are available for the relatively low earnings profiles of blacks and women. To begin with, blacks have lower educational levels than whites. This, by itself, tends to lower blacks' earnings (relative to white men's) and, hence, raise their poverty rates. In addition, both blacks and women have lower earnings than white men at all levels of educational attainment, which also contributes to the relatively high poverty rates of blacks, as shown in table 5. (Women's rates are not affected as much as blacks' because women's earnings are so often supplemented by those of a working husband.) The earning differentials between white men and blacks and women have often been attributed to discrimination, and a variety of theories have been proposed by economists and sociologists which seek to explain the effect of discrimination on differences in earnings.(9)

One factor explaining why blacks and women have lower earnings is that they tend to be in jobs that provide less on-the-job training. Saul Hoffman, in a 1981 article entitled "On-the-job training: difference by race and sex," suggests that this is one reason why blacks and women have lower earnings than white men even when educational levels are similar.(10) Hoffman found that blacks and women were seldom in jobs in which they were currently receiving training. Similarly, jobs held by blacks and women required a relatively short period before workers felt that they were "fully trained and qualified," the assumption being that such time is spent acquiring skills. By virtue of their increased skills over time, which are generally associated with job or career advancement, the earnings of white men tend to rise at a faster rate than the earnings of blacks and women. Hence, Hoffman concludes that white men are less likely to be trapped at very low wages.

Meeting family needs with earnings

As noted earlier, poverty is a condition closely tied to one's family situation, that is, to the amount of family resources and to the number of people who need to share those resources. For example, the family of a worker with minimum-wage earnings and extensive unemployment will not be poor if another member earns enough to keep the family's income above the poverty threshold. Indeed, most individuals who experienced the labor market problems identified in this analysis were not members of poor families in 1987.

Researchers have also noted how closely a family's economic status is linked to the size or composition of the family. Divorce, the death of a spouse, marriage, birth, or the departure of a child from the home can radically alter both family composition and earnings and thus have as profound an effect on poverty status as unemployment or a decline in wage rates.(11) Families of the working poor. Of the 7 million families that were in poverty in 1987, 3.4 million were there despite the fact that at least one member was in the labor force most of the year. This represents about 6 percent of all families with a working member. The median income of these families that were below the poverty level was $6,805, compared with $36,716 for nonpoor families with workers. The fact that 83 percent of the families of the working poor had children, compared with only 55 percent of nonpoor families, shows the strong impact of family composition on poverty.

The most dramatic difference between poor and nonpoor families is the percent with only one earner. As shown in the following tabulation, 76 percent of poor families had only one working member, while the majority of nonpoor families had two or more earners.
 Nonpoor Poor
 families families

All families with a member

in the labor force 27 weeks

or more (thousands) ....... 50,012 3,382

Percent with:

One member in the labor

force 27 weeks or more ... 39.3 75.9

Two or more members in

the labor force 27 weeks

or more ....... 60.7 24.1

Almost 40 percent of poor families were maintained by women, compared with just 12 percent of nonpoor families. The fact that so many poor families are maintained by women reflects several of the influences on poverty already discussed: these women's relatively low level of education and their resultant low wages; less career advancement among both women and blacks (female family heads are disproportionately black); the lack of other earners in those families; and the interrelationship between family size and the poverty threshold. As the next tabulation shows, when only the family head worked, the poverty rate in such families was 24 percent. Among married-couple families, in contrast, even when only one spouse worked, the poverty rate was only 8 percent. This is because, in the latter case, the one earner is most often a man, and few married men earn the low wages that result in poverty.
 below the
 poverty level

All families with at least one

person in the labor force 27 weeks
 or more 6.3
Married-couple familes, total 4.4

Only one member in the labor

force 27 weeks or more 8.2

Two or more members in the

labor force 27 weeks or more 2.3

Families maintained by

women, total 17.2

Householder is the only person

in the labor force 27 weeks
 or more 24.2
Families maintained by men, total 7.8

Householder is the only person

in the labor force 27 weeks

or more 11.6

When a woman maintains a family and her earning potential is at or near the minimum wage, she generally cannot keep out of poverty. Researchers have found that, to many in this circumstance, welfare, generally in the form of Aid to Families with Dependent Children (AFDC), becomes a more attractive alternative.(12) Even when the welfare benefit is less than the potential earnings, receiving welfare may still be attractive because it does not involve such employment-related costs as child care, transportation, clothing, taxes, and Social Security withholding.

In the study mentioned earlier, Levy found that poverty among working women was not primarily a function of unemployment, or of voluntarily working less than year-round, full-time hours, or even of earning lower wage rates than other women.(13) In fact, even if poor female heads of household earned a "normal" wage--that is, a wage equal to that of nonpoor women with similar characteristics, such as age and education--for a full-time work year, few would rise above poverty. This reflects the concentration of women, both poor and nonpoor, at low wage rates relative to men and underscores the fact that those women who are the sole earners in their families often have a difficult time staying out of poverty.

Women who maintain families actually have median average weekly earnings for full-time work that are nearly identical to those of married women. The latter, however, almost always have a working husband. Another, perhaps more meaningful, comparison is that families maintained by women have only half the median earnings of married-couple families.(14) Yet their financial requirements are not much less, because their average family size is little different from that of married-couple families.(15)

Families headed by black women are overrepresented among the poor, not because black women's earnings are that much lower than white women's (they are not), but because such a large proportion of these women are the sole earners in their families. The proportion of all black families headed by women (no spouse present) has risen dramatically over the past several decades--from less than 20 percent in 1950 to more than 40 percent in the 1980's.(16) Part of the rise stems from a dramatic increase over this period in the proportion of never-married black women who head families. Also, black women have much higher separation and divorce rates than white women, and the differences are exaggerated by the very low remarriage rates among blacks.(17)

Much of the literature related to the increase in the proportion of black female-headed families focuses on the role of various welfare programs, particularly AFDC, in encouraging such a family structure. In particular, William Julius Wilson and Kathryn M. Neckerman have suggested that the relatively poor economic status of young black men, as evidenced by their low labor force participation rates, has reduced the pool of "marriageable" black men.(18)

Poverty is also relatively common among workers living alone or with unrelated individuals. Three out of 10 poor workers fall into this category. They are younger than most workers, a large proportion being 16 to 24 years of age. They generally work at low wages, and, while they have no family to support, neither can they depend on the earnings of other family members to keep them out of poverty. Of course, many such persons live with others and may share housing costs and possibly other expenses. If their household units were treated as families, it is possible that the combined financial contributions of all members would result in higher-than-poverty incomes. But regardless of living arrangement, each unrelated individual is held to a poverty standard for a one-person economic unit.

The dynamics of poverty

The view presented here, which also appears in many earlier reports on economic hardship, relies primarily on cross-sectional data that provide a snapshot of the working poor. This type of data, however, cannot be used to study the long-term status of the working poor. Most importantly, it cannot be used to determine the extent to which families with workers are persistently poor or the extent to which their poverty is transitory.

Longitudinal surveys, that is, those which interview the same people over a period of years, are the best source of information on the dynamics of poverty. Some research using data from the Panel Survey of Income Dynamics (PSID), a study of 5,000 households which began in 1968, has been conducted on issues related to labor market problems and poverty. For example, in addition to Levy's study mentioned earlier, Mary Corcoran and Martha S. Hill have investigated the effect of unemployment on poverty status.(19) These researchers' findings indicate that 10 percent fewer persons would be living in "poor" families if family householders had experienced no unemployment. In their study, they defined poverty as the condition of having an average income, during the 9-year period 1967-75, below 125 percent of the average poverty level. Their findings might have been quite different using cross-sectional data; perhaps unemployment has a greater influence on a family's income and poverty status in any given year than it does over an extended period of time. While the Corcoran and Hill study had a narrow focus--unemployment--it does demonstrate the different view that longitudinal data provide when examining poverty issues.

Research on the entire poverty population using data from the PSID indicates that poverty is rarely a permanent state for a family over a long period of time. Many individuals enter poverty because of a major change in their circumstances, such as divorce, the death of a spouse, illness, or unemployment. The PSID results show that, while fully a quarter of all U.S. families (including families with and without workers) spent at least 1 year in poverty between 1969 and 1978, fewer than 3 percent were "persistently poor"--that is, below the poverty level in at least 8 of the 10 years studied.(20) Those who were poor only temporarily had characteristics quite similar to those of the general population, supporting the notion that poverty was the temporary result of sudden changes in family or economic status. The persistently poor were even more concentrated in two overlapping groups--blacks and women who head families--than they are found to be in the annual CPS-derived poverty data. While blacks made up only 12 percent of the U.S. population, they were found to constitute 62 percent of the persons who were persistently poor, a far larger proportion than are poor in any one year. Black women made up a third of the total.

Whether poverty among families with workers is in fact mostly transitory is difficult to infer from these data representing all poor families. Perhaps poverty is more permanent among families of the working poor, because this group is less affected by events such as divorce or death than are the nonworking poor. On the other hand, some poor workers may have enough upward mobility in their jobs or careers that they are able to earn their way out of poverty, an option that is unlikely for a poor disabled person or an elderly woman living alone. In any event, it is fairly clear that, as in all poor families, the persistently poor among families with workers are overrepresented by blacks and, particularly, black women.

The policy implications of the differences between the persistently poor and the temporarily poor are important, because measures to deal with temporary poverty would necessarily be quite different from those designed to deal with long-term poverty.


While unfortunate circumstances can leave many families temporarily below poverty, among workers poverty is chiefly a feature of those with a particular profile. This article suggests that 6.4 million persons in 3.4 million poor families were either employed or in search of a job during at least half of 1987. Poor workers tend to have low levels of education and, often as a result, to be employed at very low wages. Workers most likely to be poor are those who have children and are the only earners in their families. Thus, among workers, women who head families are in the greatest jeopardy of living in poverty. [Tabular Data 1 to 5 Omitted] [Exhibit 1 Omitted]


(1)For studies on labor market-related hardship, see: U.S. Department of Labor, Manpower Report of the President, 1967, pp. 74-76; William Spring, Bennett Harrison, and Thomas Vietorisz, "Crisis of the Underemployed," The New York Times Magazine, Nov. 5, 1972; Herman P. Miller, "Subemployment in poverty areas of large U.S. cities," Monthly Labor Review, October 1973, pp. 10-17; Sar A. Levitan and Robert Taggart, Employment and Earnings Inadequacy: A New Social Indicator (Baltimore, The Johns Hopkins University Press, 1974); Thomas Vietorisz, Robert Mier, and John Giblin, "Subemployment: exclusion and inadequacy indexes," Monthly Labor Review, May 1975, pp. 3-12; Francis Horvath and Janet Scholl, "Measurement of Labor Market Related Economic Hardship," unpublished paper prepared for the National Commission on Employment and Unemployment Statistics, 1977; National Commission on Employment and Unemployment Statistics, Counting the Labor Force (Washington, U.S. Government Printing Office, 1979), pp. 57-81; Bruce W. Klein, "The Adequacy of the Earnings Capacity of the Subemployed and Its Policy Implications," Ph.D. diss., The George Washington University, 1981; Robert Taggart, Hardship--The Welfare Consequences of Labor Market Problems: A Policy Discussion Paper (Kalamazoo, MI, The W.E. Upjohn Institute for Employment Research, 1982); and Bruce W. Klein, "Measuring Labor Market Related Hardship Using SIPP Data," American Statistical Association: 1986 Proceedings of the Social Statistics Section (Washington, American Statistical Association, 1986). (2)Updated tables 1-19 from the aforementioned reports are available upon request from the Bureau of Labor Statistics, Office of Employment and Unemployment Statistics, Division of Labor Force Statistics, Washington, DC 20212. (3)Like past BLS efforts, this one utilizes the March work experience and income supplements to the Current Population Survey. These supplements have questions on individuals' work activity during the entire previous calendar year, such as weeks worked, weeks spent in search of a job, and weeks spent out of the labor force, and about income and earnings over the period. The data reported were collected in March 1988 and refer to calendar year 1987. (4)Data on the self-employed relate to nonagricultural workers only. For this article, we did not investigate the characteristics of the poor self-employed, although an analysis of their detailed occupational characteristics would probably help to understand the group. (5)See Frank Levy, "How Big Is the American Underclass?" Working Paper 0090-01 (Washington, The Urban Institute, 1977). (6)This measure is meant for research use only and is in no way intended as an endorsement for indexing of the minimum wage. (7)Shirley J. Smith and Nancy F. Rytina, "Testing a New Measure of Annual Hours of Work," presented at the annual meeting of the American Statistical Association, Toronto, Canada, August 1983, p. 12. (8)Christopher Jenks and others, Who Gets Ahead? (New York, Basic Books, Inc., 1979), pp. 229-30; and Greg J. Duncan and others, Years of Poverty, Years of Plenty (Ann Arbor, MI, Institute for Social Research, University of Michigan, 1984), pp. 111-14. (9)See, for example, Glen G. Cain, "The Economic Analysis of Labor Market Discrimination: A Survey," in Orley Ashenfelter and Richard Layard, eds., Handbook of Labor Economics, Volume I (New York, North-Holland, 1986); and the classic work by Gary S. Becker, The Economics of Discrimination (Chicago, Chicago University Press, 1957). (10)Saul D. Hoffman, "On-the-job training: difference by race and sex," Monthly Labor Review, July 1981, p. 34. (11)Duncan and others, Years of Poverty, Years of Plenty, p. 10. (12)Rosemary Kern and Jack A. Meyer, "Reforming Welfare: Basic Tenets and Fundamental Choices," in Jack A. Meyer, ed., Ladders out of Poverty (Washington, American Horizon Foundation, 1986), p. 15. Aid to Families with Dependent Children is the cash assistance program for children with no father, or a disabled or unemployed father, in their home. (13)Levy, "How Big Is the American Underclass?" Although the earnings of women have increased relative to men's since this study was conducted, the basic finding that full-time low-wage workers cannot escape poverty by increasing their hours worked seems to remain valid. (14)Bureau of Labor Statistics, Employment and Earnings, January 1988, p. 214. (15)Household and Family Characteristics: Current Population Reports, Series P-20, No. 419 (Bureau of the Census, March 1986), p. 13. (16)William Julius Wilson and Kathryn M. Neckerman, "Poverty and Family Structure: The Widening Gap between Evidence and Public Policy Issues," in Sheldon H. Danziger and Daniel H. Weinberg, eds., Fighting Poverty: What Works and What Doesn't (Cambridge, MA, Harvard University Press, 1986), p. 235. (17)Wilson and Neckerman, in Danziger and Weinberg, eds., Fighting Poverty, p. 237. (18)Ibid., pp. 244-45 and p. 259. (19)Mary Corcoran and Martha S. Hill, "Unemployment and Poverty," Social Service Review, September 1980, pp. 407-13. (20)Duncan and others, Years of Poverty, Years of Plenty, p. 41.

APPENDIX: Measurement of low earnings

Past analyses of economic hardship focused on those workers whose hourly wage rates were at or below the Federal minimum wage. However, because, as of this writing, the legislated minimum wage has not been changed since 1981, its real value declines every year, making comparisons over time of limited analytical value. If the actual minimum is used as a demarcation line, the number of low-wage workers would decline almost every year, as nominal wages rise. For that reason, we have drawn a "low-earnings" line that controls for changes in the real value of the Federal minimum wage. Our low-wage measure is equivalent to the average minimum wage during the period 1967-87, expressed in 1987 dollars. (See table A-1.) That wage came out to be $4.18 an hour.

The Consumer Price Index for All Urban Consumers, Experimental Measure 1 (REBASED)--referred to as CPI-U-X1--was used to convert minimum wages prior to 1983 to the 1987 dollar level. Before 1983, the measurement of homeownership costs in the official CPI included changes based on the asset value of homes. Recognizing that this method failed to distinguish between the investment and consumption aspects of homeownership, the BLS began a program of research in the early 1970's, and the rental equivalence method was introduced in 1983. The BLS also developed, for research purposes, an index which links the rental equivalence method to years before 1983 and provides a series which treats homeownership consistently over time.(1)

The calculation begins with the year 1967 because that was the first year in which those covered under minimum-wage legislation comprised the same broad group of workers that are currently covered.(2) This low-earnings measure, updated each year, will permit more meaningful year-to-year comparisons than would be possible using the actual minimum-wage level in effect at any particular time.

A weekly "low-earnings" value was determined by multiplying $4.18 by 40 hours, yielding $167.20. This figure was then compared with the average weekly earnings for full-time wage and salary workers, which was obtained by dividing annual earnings by the number of weeks worked.(3) The minimum-wage level and 150 percent of this level, used to demonstrate the sensitivity of the number of low-wage workers to the choice of wage level used, were determined in a similar way. The minimum wage of $3.35 was multiplied by 40 hours (= $134.00) and 150 percent of the minimum ($5.03) by 40 hours (= $201.20). These levels were then compared with the computed average weekly earnings of the full-time wage and salary workers. Alternatively, the estimated low-earnings level using the official CPI is $4.33 times 40 (= $173.20). A comparison of these four estimates is presented in table A-2. No values are given for part-time workers, because past research has shown the unreliability of weekly earnings estimates for that group.

The low-wage level of $4.18 was not applied to hourly earnings directly, because (1) many workers are not paid at hourly rates; (2) earnings such as tips and commissions are generally not reported as part of the worker's hourly rate; and most importantly, (3) the reference period for the other data used in this analysis is a year. No hourly wage data are collected that would apply to such a reference period. [Tabular Data A-1 to A-2 Omitted]

Footnotes to the appendix

(1)For more information on price indexes using a rental equivalence approach, see the following Monthly Labor Review articles: Janet L. Norwood, "Two Consumer Price Index issues: weighting and homeownership," March 1981, pp. 58-59; "Indexing Federal programs: the CPI and other indexes," March 1981, pp. 60-65; and "The effect of rental equivalence on the Consumer Price Index, 1967-82," February 1985, pp. 53-55. See also "Changing the Homeownership Component of the Consumer Price Index to Rental Equivalence," CPI Detailed Report, January 1983, pp. 7-13. For a prior use of the CPI-U-X1, see Michael W. Horrigan and Steven E. Haugen, "The declining middleclass thesis: a sensitivity analysis," Monthly Labor Review, May 1988, pp. 3-13. (2)The 1966 Amendments to the Fair Labor Standards Act extended private employee coverage to about the level which currently exists. Government employees became covered in 1985 as the result of a court decision which rescinded their previous exclusion. (3)The technique implicitly assumes that full-time work occurred in each week, although it is applied to individuals who usually worked full time. Such workers may have worked less than a full-time workweek during some of their weeks of employment.

Bruce W. Klein and Philip L. Rones are economists in the Division of Labor Force Statistics, Bureau of Labor Statistics.
COPYRIGHT 1989 U.S. Bureau of Labor Statistics
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Klein, Bruce W.; Rones, Philip L.
Publication:Monthly Labor Review
Date:Oct 1, 1989
Previous Article:The Hispanic Population of the United States.
Next Article:Developing statistics to meet society's needs.

Related Articles
Working and poor in 1990.
Labor Month in Review.
Working women and poverty.
Education and the working Poor.
Occupations and poverty.
Working poor in 2003.
Working poor.
A profile of the working poor.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters