The impact of employer payments for health insurance and social security on the premium for education and earnings inequality

The trend toward increased wage and income inequality that emerged in the 1980s - with "the rich getting richer and the poor poorer" - has attracted a great deal of attention and concern. Journalists have brought the growing gap between workers on the top and bottom rungs of the earnings ladder to the public's attention,(1) while academics have sought the reasons for the change. One aspect of this phenomenon has been the growing premium for education, with the disparity between the wage and salary earnings of the least and best educated rising since 1979. Explanations generally focus on the slowing growth in the supply of college-trained workers entering the labor force as the baby boom generation has matured and, on the demand side, the widespread adoption of new technologies requiring skilled workers. However, many economists find that these explanations are not fully satisfactory. A related observation, also not fully understood, involves the increased earnings inequality among similar workers - young, male, high school graduates working full-time, year-round, for example. This rising within-group inequality occurred in the 1970s as well as the 1980s, unlike the increased between-group inequality seen only in the 1980s.

This exploratory article seeks to broaden the discussion by asking whether the rising cost of another element of compensation - employer-provided health insurance and employer payments for FICA taxes - has contributed to the growth in observed and actual inequality among workers over this period. The cost of these two benefits increased from 11 percent of total compensation in 1970 to 17 percent in 1990.(2) Because the cost of these fringe benefits looms large in comparison to the lowest wages, these employer obligations would be expected to take a relatively large bite out of the wages of workers on the bottom rung.(3) Since workers with few years of work experience or schooling also tend to earn low wages, this country's job-based system for financing health insurance and Social Security may have exaggerated the premium for education (and the growth in that premium) found when compensation is measured by wages alone. The rising cost of these benefits might also help to explain the observed increase in within-group inequality for groups defined in a variety of ways, including at the industry or plant level.

On the other hand, the share of the population covered by employer-provided health insurance has shrunk over this period as health insurance costs have soared and as the structure of employment has shifted from manufacturing, mining, and transportation (where health insurance benefits are common) to services (where they are less so). It seems quite likely, moreover, that the declining availability of employer-provided health insurance may have hit less-skilled workers particularly hard. If so, measures of the growth in between-group and within-group inequality based on wage trends would understate the real growth in compensation inequities.

Although several studies have explored the impact of including health insurance benefits and worker payments for FICA on family or household income inequality,(4) discussions of the premium for education are generally based on relative wages rather than on the theoretically preferable concept of total compensation. Compensation is the preferred measure because that total is the value set by supply and demand conditions. If some component of total compensation - employer payments for health insurance, for example - rises as a share of the total, then, other things equal, real wages or other fringe benefits should fall. For this reason, using wage behavior as an indicator of changing supply and demand conditions could be misleading. Moreover, in addition to signaling imbalances in the supply of and demand for specific types of labor, an index of inequality can also serve as a gauge of economic or social equity. From this second perspective too, adding the value of health insurance and Social Security benefits to wages results in a better measure of inequalities in all forms of remuneration than wages alone.

Admittedly - and particularly from the workers' perspective - this modification improves our measure of economic equity only partially because the value of these benefits is not equal for each employee. For example, health care costs vary considerably across states, and some workers might obtain insurance through a spouse's employer. Similarly, while employers view their legally required FICA tax payments as "benefits," the amount paid for each employee has little correlation with the value of the Social Security benefits that the employee eventually receives. Moreover, all face different tax situations. Thus, individual workers might prefer higher wages in lieu of certain benefits.

Researchers have been forced to use wages in studying the premium for education because other components of total compensation have not been available in a data base linking pay with personal characteristics like age (experience) and years at school. However, starting with the March 1980 Current Population Survey, the U.S. Bureau of the Census has collected information on employer-provided insurance and employee and (because they are equal) employer contributions for Social Security. The Bureau of the Census has been publishing this information, including estimates of the value of employer-provided insurance for Census respondents, starting with the March 1988 survey.

This study uses a small part of the published and unpublished data to explore the impact of employer payments for health insurance and FICA on the observed and actual premium for education from 1979 to 1992 for males(5) working full-time and year-round.(6) It concludes that adding the cost of health insurance to wages boosts the rise in the premium for education for all males working full-time, year-round, by as much as 25 percentage points, because over this 13-year period men with relatively little education lost access to jobs with health benefits to a disproportionate extent. For all full-time male workers, moreover, adding the cost of health insurance to wages also increases the growth in inequality within narrowly defined groups.(7)

By exception, for the more limited group of men with employer-provided health insurance, adding the cost of these benefits to wages reduces the observed premium for education in any given year and the growth in that premium during the 1980s,(8) as expected. Adding insurance payments to wages also tends to moderate the rise in within-group inequality for men with job-based health benefits.

In sum, then, the decreased availability of jobs with health insurance has had a particularly severe impact on less-skilled workers. As a result, the compensation of full-time male workers has actually become substantially more unequal since 1979 than the traditional measure based on wages alone indicates. The article ends with a brief consideration of the policy implications of these results.

I. The Premium for Education: The Picture to Date

According to Frank Levy and Richard Murnane's extensive review of trends in U.S. earnings levels and inequality and proposed explanations for these developments (Levy and Murnane 1992), inequality between groups of men defined by age and education declined slightly in the 1970s and grew in the 1980s. By contrast, inequality within groups defined by age and education grew steadily through both decades. Very important in explaining the decline in the education premium in the 1970s was the entry of large numbers of the relatively well-educated members of the baby boom generation into the labor force, starting in the late 1960s.(9) Between 1971 and 1979, the number of 25- to 34-year-old male college graduates in the labor force increased by 85 percent while the number of young male high school graduates rose just 13 percent. As a consequence of this influx, the premium for experience rose and the premium for a college education fell. During the 1980s, by contrast, the number of young college-educated males in the labor force grew slightly more slowly than the number of young high school graduates - perhaps because of the decline in the education premium during the 1970s and the move to a volunteer army.(10)

Table 1 shows the distribution by age and level of education for men working full-time, year-round in the two years used in this study, 1979 and 1992. In 1979 the first baby boomers were 33 years old. By 1992, the leading edge of the baby boom was just entering the 45-54 years of age category.

It is generally agreed, however, that the change in the relative supply of young college graduates does not fully explain the dramatic rise in the premium for education seen in the 1980s. Changes in the demand for skilled labor must also have been at work. Indeed, as Olivier Blanchard (1995) put it, the situation is best [TABULAR DATA FOR TABLE 1 OMITTED] described as "a race, over the last twenty years, between increases in relative demand for skills and increases in relative supply. In the 1970s, relative supply won; in the 1980s, relative demand won. But in both decades, the race has been fast on both sides."

Two frequently cited explanations for the rise in the demand for skilled workers include the growth in world trade with a consequent decrease in the demand for unskilled workers in the United States and the other industrialized countries, and the spread of new technologies that increase the productivity of, and thus the relative demand for, highly skilled workers. Economists still have not entirely sorted out the relative contributions of these and other explanations (like the declining importance of trade unionism) to the increased premium for education, although the majority tend to give most weight to the technology-based explanations.(11) Whatever their relative importance, however, these developments together do not appear to account for the entire change.(12)

Another important unsolved piece of the puzzle, according to Levy and Murnane, concerns the 20-year trend of rising earnings inequality within narrowly defined groups. For example, analysts have noted an increase in inequality among workers of similar age and skills even within a given industry or firm. Because "skill" is usually measured by years at school,(13) one suggested cause of this rise in within-group inequality (with the groups defined by age and years of education) is an increase in the demand for specific vocational skills or for abilities not necessarily associated with years of formal education - interactive skills like mentoring, negotiating, or supervising, for example.

Cross-country comparisons point to still another aspect of the wage gap phenomenon requiring explanation. While many OECD countries experienced a growing premium for education during the 1980s, wage dispersion increased more dramatically in the United States than in most other industrialized countries examined so far (Higgins 1994). This observation raises the role of institutional differences and their contribution to growing inequality in labor compensation.

Altogether, recent reviews of the literature(14) suggest that much work remains to be done in untangling the many intertwined developments that have contributed to growing labor income inequality in recent years. This article does not attempt to assess the relative merits of the explanations already posited. Rather it explores the impact of expanding our measure of compensation to include health benefits and payroll taxes for Social Security.

II. The Impact of Including Health and Social Security Benefits

To broaden the discussion of earnings inequality, this article examines the impact of shifting from wages towards compensation by adding employer contributions for health insurance and FICA taxes to pre-tax wage and salary earnings(15) in measuring the premium for education and within-group inequality. The first step involved identifying in the Current Population Surveys conducted in 1980 and 1993 all males working full-time, year-round. These men were then classified by age (as a proxy for experience) and by years of education; a subgroup included men with employer-provided health insurance.(16)

The Premium for Education

Tables 2 and 3 show the ratios of the median annual wage and salary earnings for the members of each education group to the median for high school dropouts of the same age/experience; Table 2 provides data for males with health insurance benefits, while Table 3 covers all males. The columns labeled "W" measure the education premium in the traditional way, using annual wage and salary earnings. In the columns labeled "W+H," compensation includes employer contributions for health insurance. The tables show the results for insured men (Table 2) and for all men (Table 3) separately in order to distinguish the impact of employer payments for health insurance on wages and compensation from the impact of insurance availability.

In both tables, the ratios based on wages exhibit the premium for education noted in previous studies. Referring to Table 2, for example, in 1992, the median wage and salary earnings for 25- to 34-year-old males with a college degree and employment-based insurance were 73 percent greater than the median earnings for young men without a high school diploma. For all males, with or without employer-provided health insurance, the comparable premium was 106 percent.

The wage-based data also show the increase in the premium for education over time found in other studies. In 1979, for example, the average young man who had completed high school earned 23 percent more than the average high school dropout (see Table 3), while young men with postgraduate education earned an average 54 percent more than men without a high school diploma. By 1992, young high school graduates earned 35 percent more and young men with some years in graduate school earned 135 percent more than young high school dropouts. As the tables show, much of the total increase in the premium for education occurred at the graduate level. In 1979, the bulk of the premium accrued to college graduates; men who studied beyond college made limited additional gains. By 1992, however, the relative reward for postgraduate study had risen considerably. The shift undoubtedly relects a growing demand for men with professional degrees combined with a declining share of men holding such degrees, as shown in Table 1.(17)

Returning to Table 2, the figures in the columns labeled "W+H," with employer contributions for health insurance added to wages and salaries, also show a comparable premium for education and a [TABULAR DATA FOR TABLE 2 OMITTED] large increase in that premium between 1979 and 1992. However, as the third column in each year's panel indicates, measuring the premium using wages plus health benefits generally reduces the premium for education by as much as 15 percent. By exception, in 1992, measuring the education premium with wages plus health insurance raises the reward for finishing high school. This exception may reflect that men with a high school education and a job with health insurance benefits may be disproportionately employed at unionized facilities in the northeast quadrant of the country, where health benefits and medical costs may both be above average.

Turning to the growth in the premium for education between 1979 and 1992, the final three columns in Table 2 show that for most groups with job-based insurance this premium has risen by less over the 13 years covered when compensation includes health insurance than when wages are the only criterion. The difference reflects the fact that health insurance benefits had grown as a share of compensation between 1979 and 1992, especially for low-paid, less-skilled workers.(18)

The data in Table 3, which cover all men working full-time, whether or not they receive health insurance benefits, present a very different picture from that shown in Table 2. When the sample contains all full-time male workers, including the employer cost [TABULAR DATA FOR TABLE 3 OMITTED] of health benefits generally increases the growth in the premium for education by as much as 25 percentage points. As before, adding the cost of insurance benefits to wages reduces the premium for education in any given year; however, the reduction shrinks from as much as 17 percent in 1979 to a maximum of 10 percent in 1992. Table 4 provides the explanation: for all male workers, access to jobs with health benefits declined sharply over this period, but the drop was particularly severe for men with limited education. While the share of full-time male workers with health insurance benefits fell from 87 to 70 percent between 1979 and 1992, the share of full-time male workers with less than a high school education and job-related health benefits fell from 88 percent to 54 percent. In 1979, men with no more than a high school education suffered little disadvantage in terms of access to health insurance; in 1992, they did. Clearly, the decreased availability of employer-provided health insurance has hit the least educated particularly hard. This development almost surely reflects declining employment opportunities for males with little formal education in manufacturing and in highly unionized nonmanufacturing industries, like mining and trucking, where health insurance benefits have been an important part of the compensation package. By contrast, the sectors where many unskilled workers now find jobs - retailing, personal services, and construction - have below-average insurance coverage.

Of course, men employed in jobs without health benefits are not necessarily uninsured. They may, for instance, be covered through their wife's health plan, or they may pay the entire cost of insurance themselves. Alternatively, some workers - particularly the young and healthy - may choose to take their chances and go uninsured. Appendix Table 1 shows the health insurance status of full-time male workers with no job-related health insurance benefits in 1993. The table indicates that, even taking other sources of insurance coverage into account, working men with limited education were much more likely to be uninsured in 1993 than were men with a college degree or more. Among workers ages 35 to 44, for instance, 71 percent of those with less than a high school diploma had no insurance; the comparable number for those with some post-graduate education was 23 percent. Relatively well-educated, and presumably well-paid, men were much more likely to have insurance coverage through a wife's health plan or to have other private insurance coverage purchased out-of-pocket than were men with a high school education or less.

Table 4
Share a Full-Time, Year-Round Male Workers Ages 25 to 54 with
Employer-Provided Health Insurance, by Age and Years of Education,
1979 and 1992

Accordingly, one approach to ameliorating the recent increase in worker inequality might be to rethink the way this country finances health insurance and Social Security. For example, if citizens want to keep the U.S. system of job-based health insurance, total employer and employee premium payments could be allocated on a sliding scale by earnings rather than as a flat premium that weighs most heavily on the wages or job prospects of the lowest-paid workers. If needed, the government could cover any shortfall by subsidizing low-wage workers and their employers. Alternatively, policymakers might prefer to consider financing health care through a tax on income. Such an approach would neither aggravate income inequality nor encourage employers to substitute capital for labor. In other words, health reform could contribute to reducing earnings inequality and to raising labor force participation, an issue likely to become increasingly important as the population ages.

A recent World Bank study (International Bank for Reconstruction and Development 1994) makes somewhat similar recommendations concerning social security. It suggests that public social security should be financed out of general revenues rather than through a capped tax on wages. Again, the ceiling on taxable earnings aggravates inequality while the taxation of wages distorts the choice between capital and labor. Likewise, Olivier Blanchard hints that a subsidy for low-skilled workers, possibly in the form of a reduction in the payroll tax, might be desirable, but he finds such a step to be highly unlikely politically.

Whether or not U.S. policymakers decide to change this country's approach to financing its health insurance and Social Security systems, U.S. voters need to recognize that our current financing methods contribute to growing compensation inequality and, most probably, to reduced labor force participation. Similarly, researchers may want to pay more attention to institutional issues in exploring earnings trends across various sectors and times.

IV. Conclusions

In sum, then, this article indicates that the U.S. system of employer-based health insurance and a capped payroll tax to fund Social Security has distorted our traditional wage-based measures of between-group and within-group inequality and their growth. In fact, the study points out, when compensation is measured to include employer costs for health and Social Security benefits, inequality among male workers has increased more than generally recognized. Accordingly, these findings underscore the need for U.S. citizens to rethink this country's approach to financing its health insurance and Social Security systems.

1 Indeed, a recent article on the front page of The New York Times (Bradsher 1995) presented recent and forthcoming research indicating that among the industrial countries the United States has the most unequal distribution of income and wealth.

2 As a share of benefits, moreover, employer payments for group health insurance and for Social Security - Old Age, Survivors, and Disability Insurance (OASDI) and Medicare hospital insurance (HI) - jumped from 46 percent to 63 percent over the same period.

3 Federal Insurance Contributions Act (FICA) taxes are proportional to wages only up to the maximum taxable earnings or wage base. This ceiling on taxable earnings results in the FICA taxes having a disproportionately small impact on the wages of workers whose earnings exceed the cutoff. The wage base for OASDI has risen from $22,900 in 1979 to $55,500 in 1992; the base for HI has risen from $22,900 to $130,200 over the same period.

4 Data in a study titled Measuring the Effect of Benefits and Taxes on Income and Poverty: 1979 to 1991 (U.S. Bureau of the Census 1992) show that including health insurance benefits reduces family income inequality slightly, while including worker payments for FICA increases income inequality in any given year. However, these data also indicate that including health insurance benefits in income increases the growth in income inequality (measured by the Gini index) between 1979 and 1992.

5 This exploratory study focuses on males because the increase in the education premium during the 1980s noted by other studies for both men and women was particularly pronounced for men. In addition, the issue of the changing availability of employer-provided health insurance is undoubtedly clarified by examining men since they have held a disproportionate number of jobs in the industries, like manufacturing and transportation, that have traditionally offered employer-paid health insurance.

6 While it clearly would be preferable to have data for other components of total compensation, the employer cost of many of the excluded benefits, like vacation and sick pay, tends to be proportional to the individual worker's earnings; thus, these excluded benefits are less likely to have a disproportionate impact on the lowest wages.

7 Including FICA payments (where coverage has increased rather than contracted) tends to reduce the growth in the premium for education slightly but increases the growth in within-group inequality.

8 Except in the case of young, less educated workers, as will be discussed more fully below.

9 The first cohort of the baby boom (born in 1946) graduated from high school in 1964 and from college in 1968; thus, baby boom high school graduates began entering Census tabulations of the work force in 1965 while those with college degrees typically were included as full-time workers beginning in 1969.

10 Particularly during the Vietnam War, the existence of the draft may have given some young men an added incentive to continue their education in order to postpone or avoid military service. With the shift to a volunteer army, this added incentive to stay in school disappeared.

11 Although often treated as contending, these two explanations are not necessarily incompatible. For example, increased foreign competition may have prompted some U.S. firms' efforts to improve productivity. In addition, while some analysts argue that the rise in the premium for education in many countries in different stages of development indicates that trade pressures cannot be an important reason for the rise in inequality here in the United States, U.S. firms often respond to rising foreign competition by establishing affiliates overseas. They then export U.S. capital equipment and production methods to these affiliates. As foreign workers using these new methods become more productive, their rising wages could exhibit a growing premium for education similar to that seen in the United States. Accordingly, increased income inequality in foreign countries, including the LDCs, could be a corollary of increased inequality caused by trade pressures experienced here in this country. Finally, many observers have deemphasized the trade-based explanations by suggesting that the rise in the U.S. merchandise trade deficit has simply been too small to account for much of the growing wage gap. However, this country's rising surplus in services trade may have augmented trade's role, since increased foreign demand for U.S. business and professional services may have boosted the demand for skilled workers within the service sector at the same time that the growing merchandise deficit has helped reduce demand for low-skilled production workers in manufacturing.

12 Brauer and Hickok (1995), for example, find that technology, trade, and shifts in product demand across industries account for just 35 to 40 percent of the widening hourly wage gap between college graduates and high school graduates or dropouts between 1979 and 1989.

13 Of course, years at school may not be a very good measure of skill, especially since the quality of those years at school could vary considerably.

14 In addition to Levy and Murnane, see the January 1995 issue of the Federal Reserve Bank of New York's Economic Policy Review, which is devoted to that Bank's November 4, 1994 colloquium on U.S. wage trends in the 1980s and early 1990s.

15 Annual pre-tax wage and salary earnings or, for the self-employed, net earnings.

16 The sample used in this study includes all males, aged 25 to 54, working 50 to 52 weeks and at least 35 hours every week (including the self-employed). The subsample received health insurance paid in part or in full by their own employer. (In a very small number of cases, a union may have contributed to the health insurance. For most unionized workers, however, health benefits are largely paid by the employer. Self-employed workers who are incorporated and buy health insurance are treated as employees with employer-provided health benefits; unincorporated self-employed workers who buy health insurance are considered to have other private insurance, not provided by the worker's employer.) In an attempt to isolate the impact of health benefit costs on annual wage and salary earnings, men with health insurance were restricted to those working for just one employer in the year before the survey. In other words, the restriction reflects an effort to avoid including workers who were covered by job-related health insurance for just a fraction of a year. On the other hand, some men classified as not covered by employer-provided health insurance may have been working for a firm with health benefits but may have chosen not to participate in that firm's health plan.

17 Explanations for this decline in the share of men obtaining graduate degrees could include the impact of the draft and the Vietnam War on graduate enrollment in the 1960s. Alternatively, the data may simply reflect a growing tendency for individuals to attend graduate school after obtaining several years of work experience.

18 By exception, the growth in the premium for education is somewhat greater for young and prime age men with little education when the cost of insurance is included. This result may again reflect the concentration of men with a high school education and jobs with health benefits in unionized industries and in regions with relatively high health care costs.

19 The coefficients of variation shown in Tables 5 and 6 are ratios of the standard deviation to the mean multiplied by 100. The coefficient of variation has the attribute of being scale invariant (the degree of inequality does not change when all observations are multiplied by a constant). In addition, this measure has the desirable property that it always registers a decrease in inequality when income is shifted from the higher- to the lower-income person, regardless of where in the distribution the transfer occurs.

20 of course, in response to rising premium costs, a growing number of workers, particularly young workers, may have chosen to self-insure. In addition, as women's labor force participation has increased, a growing number of men may be obtaining health insurance through their wives.

21 Consistent with the hypothesis that rising health insurance costs have depressed wage growth for workers with health benefits is the fact that the median annual wage and salary earnings for full-time, year-round male workers with health benefits rose 36 percent between 1979 and 1992, while the median for similar workers with no health benefits rose 71 percent.

22 The data shown in Table 1 are clearly consistent with the suggestion that, as unskilled men mature, they have been dropping out of the full-time work force at an above-average pace.

References

Blanchard, Olivier. 1995. "Macroeconomic Implications of Shifts in the Relative Demand for Skills." Federal Reserve Bank of New York Economic Policy Review, January, pp. 48-53.

Borjas, George J. and Valery A. Ramey. 1993. Foreign Competition, Market Power, and Wage Inequality: Theory and Evidence. Cambridge, MA., National Bureau of Economic Research Working Paper No. 4556.

Bound, John and George Johnson. 1995. "What Are the Causes of Rising Wage Inequality in the United States?" Federal Reserve Bank of New York Economic Policy Review, January, pp. 9-17.

Bradsher, Keith. 1995. "Gap in Wealth in U.S. Called Widest in West." The New York Times, April 17, p. A1.

Brauer, David and Susan Hickok. 1995. "Explaining the Growing Inequality in Wages Across Skill Levels." Federal Reserve Bank of New York Economic Policy Review, January, pp. 61-75.

Federal Reserve Bank of New York. 1995. "Colloquium on U.S. Wage Trends in the 1980s." Frederic S. Mishkin, ed. Economic Policy Review, January, entire issue.

Fieleke, Norman S. 1994. "Is Global Competition Making the Poor Even Poorer?" New England Economic Review, November/December, pp. 3-16.

Green, Gordon, John Coder and Paul Ryscavage. 1992. "International Comparisons of Earnings Inequality for Men in the 1980s." In U.S. Bureau of the Census, Studies in the Distribution of Income, Current Population Reports, Consumer Income, P60-183, pp. 51-60.

Grubb, W. Norton and Robert H. Wilson. 1992. "Trends in Wage and Salary Inequality, 1967-88." Monthly Labor Review, June, pp. 23-39.

Higgins, Matthew. 1994. "Wage Trends Abroad: The Role of Trade, Technology and Labor Market Institutions." Federal Reserve Bank of New York, mimeo, November.

International Bank for Reconstruction and Development. 1994. Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth (A World Bank Policy Research Report), New York: Oxford University Press.

Lawrence, Robert Z. 1995. "U.S. Wage Trends in the 1980s: The Role of International Factors." Federal Reserve Bank of New York Economic Policy Review, January, pp. 18-25.

Levy, Frank and Richard J. Murnane. 1992. "U.S. Earnings Levels and Earnings Inequality: a Review of Recent Trends and Proposed Explanations." Journal of Economic Literature, Vol. XXX (September), pp. 1333-81.

Nelson, Charles T. 1994. "Levels of and Changes in the Distribution of U.S. Income." In The Changing Distribution of Income in an Open U.S. Economy, edited by Jeffrey H. Bergstrand, Thomas F. Cosimano, John W. Houck and Richard G. Sheehan. New York: North-Holland, pp. 40-63.

Pierce, Brooks and Finis Welch. 1994. "Dimensions of Inequality in Labor Income." In The Changing Distribution of Income in an Open U.S. Economy, edited by Jeffrey H. Bergstrand, Thomas F. Cosimano, John W. Houck and Richard G. Sheehan. New York: North-Holland, pp. 145-79.

Ryscavage, Paul Gordon Green and Edward Welniak. 1992. "The Impact of Demographic, Social, and Economic Change on the Distribution of Income." In U.S. Bureau of the Census, Studies in the Distribution of Income, Current Population Reports, Consumer Income, P60-183, pp. 11-30.

U.S. Bureau of the Census. 1992. Measuring the Effect of Benefits and Taxes on Income and Poverty: 1979 to 1991. Current Population Reports, Consumer Income, Series P-60, No. 182-RD.

Jane Sneddon Little is Senior Economist, Federal Reserve Bank of Boston. The author is most grateful to Katharine Bradbury, Catherine Jew, and Stephen McNees for helpful suggestions and to Rachel Deyette for valuable research assistance.

COPYRIGHT 1995 Federal Reserve Bank of Boston
COPYRIGHT 2004 Gale Group