This paper reviews recent evidence on the level and distribution of wages, income, and wealth in the United States. The principal focus is on how these three measures of economic well-being have changed over the current business-cycle recovery.

The principal findings for the 1990s are:

• Wage inequality increased between 1989 and 1997, but at a slower pace than it did between 1979 and 1989.
• Increases in the federal minimum wage in 1990, 1991, 1996, and 1997, and low unemployment, especially in 1997, seem to have played an important role in retarding and eventually reversing the long-standing decline in wages for workers at bottom of the wage distribution.
• At the median and at the 90th percentile, wage growth was slower in 1989-97 than it was in 1979-89.
• Over the period 1989-97, even those workers at the 90th percentile and those with a four-year college degree saw their wages, on average, lag behind growth in average economy-wide productivity.
• Family income inequality grew sharply between 1989 and 1996, with the Gini coefficient rising from 0.401 to 0.425.
• Over the 1989-96 period, family incomes were flat or falling in real terms for the bottom four quintiles of the family-income distribution; the share of the bottom four quintiles in total family income fell 2.2 percentage points.
• Between 1989 and 1996, the share of persons in poverty increased from 12.8% to 13.7%.
• Household wealth inequality rose significantly between 1989 and 1997, driven in part by the bouyant U.S. stock market.
• While direct and indirect stock holdings have become much more widely distributed, the value of stock holdings for the vast majority is relatively small. The middle quintile of the wealth distribution, for example, owned about $7,500 in stock in all forms in 1997; the bottom 40% held an average of about $1,500

We have converted all nominal values to real 1997 dollars using the CPI-U-X1 series. The CPI-U-X1 corrects the official CPI for problems in house prices prior to 1982. Since the Bureau of Labor Statistics has not produced a CPI-U-X1 series prior to 1967, we deflate earlier years using a series created by chaining the pre-1967 CPI to CPI-U-X1.

Tables 1A, 1B, and 1C summarize developments in the hourly wage distribution between 1973 and 1997. Table 1A shows the real hourly wage in 1997 dollars for the 10th, 50th, and 90th percentile of all wage and salary workers, ages 18 to 64. The data cover the years 1973, 1979, 1989 (all cyclical peaks, defined by low-points in the national unemployment rate) and each year in the 1990s through 1997, a year with a national unemployment rate that is probably close to the lowest level to be achieved in the current business cycle. For the 1980s, the table tells the now-familiar story of declining real wages for workers at the middle and the bottom of the wage distribution. Between 1979 and 1989, real wages of workers at the 10th percentile fell at a 1.6% annualized rate between; at the median, real wages declined an average 0.2% per year over the same period. (The table also demonstrates another, less-widely commented, aspect of the 1980s wage distribution: even at the 90th percentile real wages grew little over the period --from $22.46 per hour in 1979 to $23.46 per hour in 1989, an annualized growth rate of just 0.4% per year.)

Table 1A also shows the year-to-year changes in wages over the decade of the 1990s. The most striking development over the current business-cycle is the reversal of fortune of low-wage workers, with real wages at the 10^th percentile rising on average 0.2% per year between 1989 and 1997, up considerably from the 1.6% per year decline for 1979-89. A closer look at the pattern of wage changes suggests that a combination of increases in the federal minimum wage and low unemployment may have helped to boost wages at the bottom. Real wages at the 10th percentile rose in only three of the eight years in 1990s: 1990, 1991, and 1997. These are all years when the federal minimum wage increased (in 1990 from $3.35 per hour to $3.80; in 1991, to $4.25; in 1997, to $.4.75). While the federal minimum wage lies below the 10th percentile wage in all three years, even modest spillover effects could explain the small increases at the 10th percentile. ² The data for 1997 suggest an independent role for low unemployment in explaining real wage growth at the bottom. In 1990 and 1991, as unemployment increased, real wages at the 10th percentile grew about 1.1% per year; in 1997, as unemployment fell 0.5% to below 5.0%, real wages at the 10^th percentile increased at a much faster 3.2% annual rate. ³

A second feature of the 1990s is the worsening performance of real wages at the middle and the top. Between 1989 and 1997, the median real wage fell at twice the rate it did in the 1980s (-0.4% compared to -0.2%). Over the same period, real wage growth at the 90th percentile decelerated from 0.4% to 0.2% per year. At the median, real wages were generally flat through the recession in 1992 and then, perversely, fell continuously over the recovery from 1993 through 1996. In 1997, however, real median wages grew 2.2%. At the 90th percentile, no clear pattern emerges, with real wages rising and falling over much of the period, ending with a 1.5% increase between 1996 and 1997. The deterioration in performance at the middle and the top together with the rise in real wages at the bottom mean that the rise in overall wage inequality has decelerated during the current business cycle, with wage inequality (defined by the 90-10 differential) actually declining between 1994 and 1997.

A final feature of the data in Table 1A is the failure of the vast majority of workers to keep pace with average productivity after 1973. The median hourly wage fell 0.2-0.4% per year after 1973, while average productivity grew at about a 1.0% annual rate (see Appendix Table 1). Even at the 90th percentile, annual wage growth ranged between 0.2-0.4%, still well below the average gain in labor productivity.

Tables 1B and 1C show the same real wage data separately for men and women. These data demonstrate another well-documented pattern: while women’s wages remain below men’s wages at comparable positions in the respective wage distributions, over the last 25 years or so, women’s wages have generally grown more rapidly than those of men. Table 1B demonstrates that men’s wages at the 10^th percentile have been following continuously over the business cycles of the 1970s, 1980s, and 1990s. The decline in male wages at the bottom was strongest in the 1980s (-1.3% per year), but the deterioration has continued, albeit at a slower pace, in the 1990s (-0.5% per year). At the median, male wages fell rapidly through both the 1980s (-0.9% per year) and the 1990s (-0.8% per year). Male wages at the 90^th percentile were virtually stagnant over the entire 1973-97 period, growing at annualized rate little more than 0.1% per year. For women (see Table 1C), wage developments were generally better. Women at the 10^th percentile saw rapid real wage increases in the 1970s (3.6% per year); steep declines (worse than for men at the 10^th percentile) in the 1980s (-1.8% per year); and a considerable recovery over the 1990s (0.3% per year). Women’s median wages have risen over the three business cycles, most rapidly in the 1980s (0.6% per year) and much less rapidly in the 1990s (0.1% per year), far outperforming men at the median over the same time periods. Women at the 90^th percentile posted the highest wage increases of any group, with a 1.6% per year increase in the 1980s and a 0.8% increase per year in the 1990s.

B. Real Wages By Education Level

Since much of the discussion of rising wage inequality during the 1980s and 1990s has focused on the role of formal education as a proxy for workers’ broader "skill" levels, we present wage data by education level in Tables 2A, 2B, and 2C. Again, much of the story is familiar. Wages for less-educated workers (see Table 2A), whether male (see Table 2B) or female (see Table 2C) have all declined in the 1980s and the 1990s. Discussions of wage trends over the last two decades, however, have not fully addressed another important feature of the data in the tables. The real wage declines extend far beyond those workers with the lowest level of formal education (less-than-high-school-educated workers), who account for about 15% of the total workforce. Real wages have also fallen for the 40% or so of the U.S. workforce with a terminal high school degree and the 20% or so of the workforce with one-to-three years towards a four-year college degree. Even the real wages of workers with a four-year college degree (but no further education) –about a 15% of the workforce— have seen real wage increases at (in the case of women) or below (in the case of men) the average increase in productivity over the period. The only education group that has done consistently well over the last two decades are those with advanced degrees (just over 5% of the workforce). ⁴

Using the education breakdown, the real wage pattern in the year-to-year data in the 1990s differs somewhat from that in the overall distribution. In the overall distribution, wage increases during the 1990s were largest at the bottom. By education level, where all the average wage levels are well above the 10^th percentile wage, wages grew slowest at the bottom of the educational distribution and most quickly at the middle and the top. Even in the boom year of 1997, the real wages of less-than-high-school-educated workers, for example, did not change. Over the same period, real wages increases rose for all better-educated workers (high-school, 1.7%; some college, 2.1%; college, 2.9%; and, college plus, 1.1%).

C. Poverty-Level Wages

Widening wage inequality with stagnant and declining wages, on the one hand, and apparently rapid rates of job creation, ⁵ on the other, gave rise in the 1990s to the view that the jobs being created in the U.S. economy were of poor quality. In the spring of 1996, the Council of Economic Advisors (1996) entered the debate with an influential study of the quality of job creation between February 1994 and February 1996. Since the U.S. government does not collect data on "new" jobs, ⁶ the CEA divided respondents to the CPS into about 400 industry-occupation categories. The CEA then calculated the median wage in each industry-occupation category in 1994 as well as the total employment in each category in both 1994 and 1996. The CEA used changes in the estimated total employment in each of the categories over the two-year period to identify where the "new jobs" were being created. The CEA finally compared the 1994-96 job creation data with the 1994 wage data to see whether the net job creation was occuring in high-wage or low-wage industry-occupation categories. The CEA concluded that most of the jobs were being created in industry-occupation categories where the median wage was above the median for the economy as a whole. Another recent study, by Ilg (1997), uses a similar methodology with more recent data to arrive at a similar conclusion.

We have several objections to this approach. First, neither the CEA nor Ilg examined how wages in their categories changed over the same periods that they examine employment changes. An analysis of Ilg’s data, for example, shows that while job growth was indeed strongest in those categories where the median wage was above the national median, the median real wage also fell in every job category, including those that paid above the national median. Second, the approach is at best indirect, since we can never observe new jobs, but only infer, where, on net they are being created. That the median real wage is falling across categories suggests that the new jobs may be below the median for the category and are therefore not necessarily themselves above the national median. Third, the approach focuses too much attention on net job creation, which has averaged only 1.2% per year in the 1990s, rather than on gross job creation, which has probably been about 20 times larger on an annual basis (just slightly higher than gross job destruction). ⁷ Even if the CEA analysis could clearly demonstrate that the net new jobs being created in the economy were "good jobs", ⁸ the vast majority of the gross new jobs created could still be of poorer quality than the gross old jobs destroyed. This final point can be taken one step further. In slightly imprecise terms, it is difficult to explain large changes in the wage distribution by concentrating on the 1% or so of new jobs created each year, while ignoring what is happening to the "other 99%" of jobs. Our own view is that the deterioration in the U.S. wage structure since the end of the 1970s is not primarily a function of the quality of new jobs created, on net, since that time. Rather, the principal problem stems from the institutional changes in the labor market that have undermined the bargaining power of workers in the vast majority of "existing jobs".

Given the problems with approaches based on the wage levels of new jobs, and our own concern with what is happening across the entire stock of jobs, we prefer to examine "job quality" using the approach outlined in Table 3. We have converted the wage distributions for 1973, 1979, 1989, and 1997 into real 1997 dollars and then grouped workers in each year into categories based on ratios of the 1997 "poverty-level wage" (the hourly wage rate necessary to earn the federal poverty-level for a family of four if a worker works 40 hours per week, 52 weeks per year). The data for men and women together (see panel (a)) show that jobs paying "poverty-level" wages or less jumped markedly between the 1970s (about 23.6%) and 1989 (28.5%). The number of poverty level jobs, however, had not changed much by 1997 (28.6%). Between 1989 and 1997, jobs paying just above the poverty-level wage (100-125%) grew slightly as did those paying 300% or more of the poverty-level wage. This cut of the data suggests that the overall wage distribution is shifting simultaneously toward both lower-paying and higher-paying jobs.

The patterns for men (panel (b)) and women (panel (c)) separately show important differences. Poverty-level jobs for men increase across every business cycle peak from 1973 though 1997. In the 1990s, the share of men holding a job paying below the poverty-level wage grew from 21.2% to 22.5% of male employment. The share earning near-poverty-level wages (100-125%) also increased. At the same time, the share of men earning 200-300% and 300% or more of the poverty-level wage fell a combined 3.6 percentage points.

While many more women than men earned poverty level wages in 1997 (35.3% of women compared to 22.5% of men), the share of women earning low-wages fell over the 1990s business cycle, as did the share earning near-poverty-level wages. Over the same period, women also saw a 1.8 percentage-point rise in the share earning 300% or more of the poverty-level wage. The gains for women in the 1990s, however, are mixed compared to the 1980s. The increase in the share of women earning high wages was much more rapid in the 1980s than in the 1990s. At the same time, the 1990s saw some improvement in the composition within the lowest-paid categories, after a dramatic increase in the 1980s in the share of women in the lowest wage category (0-75% of the poverty-level wage).

The distribution of family income in the United States grew more equal between 1947 and 1973. Table 4 demonstrates that between 1947 and 1967, the share of family income going to each quintile grew for the bottom four quintiles and declined for the top quintile. Meanwhile, the Gini coefficient fell from 0.376 to 0.358. The data in Table 5, which looks at the annual income of families at the 20^th, 40^th, 60^th, 80^th, and 95^th percentiles, show that, across the board, family-income growth was strong, but skewed slightly in favor of less well-off families in the 1947-67 period. Annual growth rates in family income, for example varied from 2.8-2.9% for the 20^th and 40^th percentile families to 2.5% for the 95^th percentile family. Between 1967 and 1973, family income continued to grow rapidly (between 2.2% and 3.0% per year), but families at the bottom experienced the slowest growth.

After 1973, growth in family income decelerated and grew more unequal. The share of the bottom two quintiles in the total family income fell over the 1973-79, 1979-89, and 1989-96 cycles; the share of the bottom four quintiles declined in 1979-89 and 1989-96. The 1990s produced what is arguably the worst family income figures for the postwar period. Between 1989 and 1996, real family incomes fell at the 20^th, 40^th, and 60^th percentile; were flat at the 80^th percentile; and grew at just a 0.3% annual rate at the 95^th percentile. Over the same period, the annual growth rate in average income (0.3%) was the slowest in any postwar business cycle. The income share of each of the bottom four fifths of families declined, while the share of the top fifth in total income grew 2.2 percentage points. The year-to-year numbers for the 1990s show the typical cyclical pattern in family income. At the 20^th, 40^th, and 60^th percentiles, family incomes fall through 1993 and then recover continuously through 1996. Growth rates at these same percentiles, however, are actually slower in 1996 than they were in 1995. At the 80^th and 95^th percentiles, real income hits bottom earlier (1991 for the 95^th percentile, 1992 for the 80^th) and recovers more strongly through the rest of the cycle.

B. Poverty Rate

One important aspect of the income distribution is the share of persons living in poverty. Table 6 reports the poverty rates for selected years between 1967 and 1996 according to two definitions of poverty. The first definition is the official U.S. government poverty rate, which is based on a minimum consumption basket established in the 1960s. The second definition is based on a relative poverty measure that is more typical for international comparisons. According to the official definition, poverty dropped sharply between 1967 and 1973 and then grew continuously over the business cycle peaks in 1979, 1989, and 1996. The cyclical pattern of poverty rates is obvious in the yearly data for the 1990s. Poverty grows uninterruptedly between the cyclical peak in 1989 through the cyclical trough in 1992 and even into the "jobless recovery" in 1994. Poverty then fell in 1995 and by a smaller amount in 1996.

Table 6 also allows us to compare the official poverty rate with a common alternative definition based on relative incomes. Using the broadest measure of relative poverty --the share in families receiving less than half of the median income-- the poverty rate increases substantially between 1973 and 1979 and again between 1979 and 1989, but changes little over the decade of the 1990s. The broad definition puts about 22% of the population in poverty in the 1990s, above the 13-15% official rate over the same years. A narrower definition --those receiving less than one-fourth of the national median increases significantly between 1979 and 1989 (from 6.7% to 8.3%), and then by a smaller amount between 1989 and 1996 (from 8.3% to 8.6%).

Unfortunately, the Survey of Consumer Finance, which is the underlying source of the wealth data in Tables 8 and 9, is not an annual survey. We therefore cannot report the distributions for the peak years as in previous tables. Instead, we will concentrate on general characteristics of the distribution and how the distribution changed between 1989 and 1997 (using Wolff’s projections for the latter year).

Table 7 demonstrates that the wealth distribution is more unequal than either the wage or income distribution. The share of wealth held by the top quintile has been above 80% in every survey year since 1962, with the share increasing slightly in each survey. Even within the top quintile, wealth is heavily concentrated in the top one percent of the full distribution, which controls more than one-third of all wealth.

Table 8, panel (d), attaches dollar figures to the share figures in Table 7. In 1997, the estimated average net worth (assets minus liabilities) of the top one percent of households was about $9.5 million compared to about $53,400 in the middle quintile, and to just $3,100 among the least wealthy 40% of households. Between 1989 and 1997, net worth for the top one percent grew by almost $1 million; declined about $1,600 in the middle quintile; and grew about $6,900 for the bottom 40% (from a negative net worth of -$3,800 to a positive net worth of $3,100).

B. Stock Market Boom

Table 8 also demonstrates the impact of the recent stock market boom on the overall wealth distribution. Panel (a) shows the gross value of direct and indirect stock holdings across the household wealth distribution. Between 1989 and 1997, the value of stock held by the top one percent of households rose $1.2 million, from about $1.1 million to about $2.3 million. In dollar terms, the level and growth in stock holdings were much smaller for the bottom 80% of households. The average stock holdings in the fourth quintile (from the 61^st to the 80^th percentile), for example, grew from $8,300 in 1989 to $18,600 in 1997. For households in the middle quintile, average stock holdings grew from $3,500 to $7,500. Households in the bottom 40% of the wealth distribution saw their holdings more than double – from about $600 to $1,500.

Nevertheless, large problems remain and some have even grown worse in the 1990s. Wages in the middle and top of the wage distribution have grown more slowly over the current cycle than they did in the 1970s or 1980s. The distribution of family income, at least through 1996, has grown more unequal, with the bottom four fifths of the distribution experiencing stagnant or declining real incomes over the period. Not surprisingly, given the family income trends, the poverty rate rose from 12.8% to 13.7% between 1989 and 1996. Finally, the already highly unequal distribution of wealth grew even more unequal between 1989 and 1997, in part, as a function of a stock market boom, whose benefits are still rather narrowly confined.

¹ For a complete description of the CPS wage data, see Webster (1997).  

² The combined 1990-91 increases, for example, raised the minimum wage about 24% in nominal terms. Real wages at the 10th percentile grew about 1.1% in both years. The decline in 10th percentile real wages, despite a rise in the federal minimum wage in 1996, however, argues against the importance of the minimum wage.  

³ A preliminary analysis of CPS ORG wage data for 1998, a year in which unemployment continued to fall, also suggests an important role for unemployment in the determination of wages at the 10th percentile.  

⁴ Education shares are approximate and refer to 1989. For a more detailed breakdown, see Mishel, Bernstein, and Schmitt (1997), Tables 3.18, 3.19, and 3.20.  

⁵ Job creation rates over the current business cycle (about 1.2% per year between 1989 and 1997) are slow even by recent historical standards. Job creation rates averaged 1.7% between 1979 and 1989, and 2.5% between 1973 and 1979.  

⁶ U.S. government employment numbers generally come from two sources: the Current Population Survey, which interviews individuals in noninstitutional households; and the Bureau of Labor Statistics' Establishment Survey, which asks establishments about overall staffing levels.  

⁷ See Davis, Haltiwanger, and Schuh (1996).  

⁸ This would involve answering the sticky question about which of the "gross new jobs" were "net new" ones.

APPENDIX TABLE 1
Macroeconomic Indicators, 1947-97

Year	Unemployment Rate %	Real GDP (1992$ billions)	Nonfarm Labor Productivity (1992=100)
1947	3.9	1,463.4	36.6
1967	3.8	3,147.2	69.2
1973	4.9	3,916.3	80.7
1979	5.8	4,630.6	86.3
1989	5.3	6,062.0	95.7
1990	5.6	6,136.3	96.2
1991	6.8	6,079.4	96.9
1992	7.5	6,244.4	100.0
1993	6.9	6,389.6	100.1
1994	6.1	6,610.7	100.5
1995	5.6	6,742.1	100.7
1996	5.4	6,928.4	102.0
1997*	4.9	7,191.4	103.5
Change	Percentage-Point	Annualized Percent Change
1947-67	-0.1	3.9	3.2
1967-73	1.1	3.7	2.6
1973-79	0.9	2.8	1.1
1979-89	-0.5	2.7	1.0
1989-97	-0.4	2.2	1.0

*Productivity figure is for the first 3 quarters.
Source: Economic Report of the President (1985, 1998).  

TABLE 1A
Real Hourly Wages, All Workers 18-64, 1973-97
(1997 Dollars)

Year	Percentile
	10th	50th	90th
1973	6.07	11.61	22.22
1979	6.42	11.46	22.46
1989	5.39	11.18	23.46
1990	5.45	11.12	23.48
1991	5.52	11.12	23.49
1992	5.49	11.16	23.07
1993	5.44	11.03	23.33
1994	5.36	10.79	23.64
1995	5.33	10.67	23.54
1996	5.29	10.59	23.54
1997	5.46	10.82	23.90
Annualized Percent Change
1973-79	1.0	-0.2	0.2
1979-89	-1.6	-0.2	0.4
1989-97	0.2	-0.4	0.2

Source: Authors' analysis of CPS ORG data; inflation-adjusted using CPI-U-X1.  

TABLE 1B
Real Hourly Wages, Men 18-64, 1973-97
(1997 Dollars)

Year	Percentile
	10th	50th	90th
1973	7.16	14.08	25.74
1979	7.07	14.39	25.93
1989	6.17	13.07	26.12
1990	6.05	12.82	26.52
1991	5.91	12.71	26.28
1992	5.79	12.54	26.16
1993	5.72	12.37	26.34
1994	5.67	12.11	26.19
1995	5.78	12.24	26.20
1996	5.81	12.12	25.85
1997	5.92	12.19	26.44
Annualized Percent Change
1973-79	-0.2	0.4	0.1
1979-89	-1.3	-0.9	0.1
1989-97	-0.5	-0.8	0.2

Source: Authors' analysis of CPS ORG data; inflation-adjusted using CPI-U-X1.  

TABLE 1C
Real Hourly Wages, Women 18-64, 1973-97
(1997 Dollars)

Year	Percentile
	10th	50th	90th
1973	5.05	8.89	16.07
1979	6.13	9.03	16.62
1989	5.02	9.55	19.33
1990	5.05	9.54	19.35
1991	5.13	9.52	19.66
1992	5.20	9.55	20.00
1993	5.20	9.61	20.15
1994	5.14	9.49	20.46
1995	5.10	9.39	20.19
1996	5.07	9.41	20.36
1997	5.15	9.63	20.61
Annualized Percent Change
1973-79	3.6	0.3	0.6
1979-89	-1.8	0.6	1.6
1989-97	0.3	0.1	0.8

Source: Authors' analysis of CPS ORG data; inflation-adjusted using CPI-U-X1.  

TABLE 2A
Real Hourly Wages, By Education* Level, All Workers 18-64, 1973-97
(1997 Dollars)

Year	Less Than High School	High School	Some College	College	College Plus	College/  High School
1973	11.22	12.82	14.16	18.60	22.66	1.45
1979	11.15	12.49	13.61	17.43	21.42	1.40
1989	9.38	11.36	13.20	17.88	23.24	1.57
1990
1991
1992	8.86	11.07	12.52	18.04	23.03	1.63
1993	8.72	11.02	12.47	17.97	23.22	1.63
1994	8.52	11.10	12.36	18.14	24.17	1.63
1995	8.25	10.90	12.20	18.13	23.90	1.66
1996	8.21	10.84	12.18	17.86	23.80	1.65
1997	8.22	11.02	12.43	18.38	24.07	1.67
Annualized Percent Change
1973-79	-0.1	-0.4	-0.7	-1.0	-0.9
1979-89	-1.6	-0.9	-0.3	0.3	0.9
1989-97	-1.5	-0.4	-0.7	0.3	0.4

*Education levels from 1992-1997 are estimates designed to be consistent with pre-1992 educational coding.

Source: Authors' analysis of CPS ORG data; inflation-adjusted using CPI-U-X1.  

TABLE 2B
Real Hourly Wages, By Education* Level, Men 18-64, 1973-97
(1997 Dollars)

Year	Less Than High School	High School	Some College	College	College Plus	College/  High School
1973	13.11	15.43	16.25	21.40	23.82	1.39
1979	12.87	15.01	15.94	20.46	23.33	1.36
1989	10.63	13.17	15.03	20.58	25.73	1.56
1990
1991
1992	9.80	12.58	14.04	20.45	25.61	1.63
1993	9.60	12.45	13.93	20.33	25.77	1.63
1994	9.31	12.51	13.85	20.45	26.47	1.63
1995	9.07	12.31	13.74	20.43	26.45	1.66
1996	9.04	12.22	13.68	20.21	26.21	1.65
1997	9.01	12.48	14.00	20.78	26.62	1.67
Annualized Percent Change
1973-79	-0.3	-0.5	-0.3	-0.7	-0.3
1979-89	-1.7	-1.2	-0.6	0.1	1.0
1989-97	-1.9	-0.7	-0.9	0.1	0.4

*Education levels from 1992-1997 are estimates designed to be consistent with pre-1992 educational coding.

Source: Authors' analysis of CPS ORG data; inflation-adjusted using CPI-U-X1.  

TABLE 2C
Real Hourly Wages, By Education* Level, Women 18-64, 1973-97
(1997 Dollars)

Year	Less Than High School	High School	Some College	College	College Plus	College/  High School
1973	7.91	9.75	10.86	14.43	19.40	1.48
1979	8.26	9.77	10.70	13.08	17.05	1.34
1989	7.37	9.50	11.36	14.77	19.49	1.55
1990
1991
1992	7.38	9.49	11.05	15.40	19.55	1.62
1993	7.31	9.50	11.08	15.41	19.84	1.62
1994	7.22	9.57	10.95	15.59	21.18	1.63
1995	6.92	9.36	10.73	15.61	20.55	1.67
1996	6.85	9.32	10.76	15.41	20.69	1.65
1997	6.94	9.44	10.92	15.89	20.87	1.68
Annualized Percent Change.
1973-79	0.7	0.0	-0.2	-1.6	-2.0
1979-89	-1.1	-0.3	0.6	1.3	1.4
1989-97	-0.7	-0.1	-0.5	0.9	0.9

*Education levels from 1992-1997 are estimates designed to be consistent with pre-1992 educational coding.

Source: Authors' analysis of CPS ORG data; inflation-adjusted using CPI-U-X1.  

TABLE 3
Distribution of Total Employment by Real Wage Level, 1973-97
(Share of Employment in Percent)

Year	Poverty-Level Wages			100-125	125-200	200-300	300+	Total
	0-75	75-100	Total
(a) All
1973	8.0	15.6	23.5	13.3	34.7	19.9	8.6	100.0
1979	4.2	19.4	23.7	15.3	31.4	21.0	8.7	100.0
1989	13.4	15.1	28.5	13.5	29.2	19.1	9.7	100.0
1997	12.1	16.5	28.6	14.1	29.6	17.3	10.3	100.0
Percentage-Point Change
1973-79	-3.7	3.9	0.1	2.0	-3.3	1.1	0.0
1979-89	9.2	-4.3	4.9	-1.8	-2.2	-1.9	1.0
1989-97	-1.3	1.3	0.1	0.6	0.4	-1.7	0.7
(b) Men
1973	3.8	9.0	12.8	9.7	36.7	27.9	12.9	100.0
1979	2.4	11.0	13.4	11.3	32.1	29.6	13.6	100.0
1989	9.1	12.1	21.2	11.3	29.7	23.7	14.1	100.0
1997	8.8	13.7	22.5	12.6	30.6	20.5	13.8	100.0
Percentage-Point Change
1973-79	-1.5	2.1	0.6	1.6	-4.6	1.7	0.6
1979-89	6.8	1.1	7.8	0.0	-2.5	-5.9	0.5
1989-97	-0.3	1.6	1.3	1.3	0.9	-3.2	-0.3
(c) Women
1973	14.0	25.1	39.1	18.6	31.7	8.3	2.3	100.0
1979	6.7	30.4	37.0	20.5	30.4	9.8	2.3	100.0
1989	18.2	18.6	36.8	16.0	28.6	13.8	4.7	100.0
1997	15.7	19.5	35.3	15.8	28.5	13.9	6.6	100.0
Percentage-Point Change
1973-79	-7.3	5.3	-2.1	1.9	-1.2	1.5	0.0
1979-89	11.6	-11.8	-0.3	-4.5	-1.8	4.1	2.5
1989-97	-2.5	1.0	-1.5	-0.3	-0.1	0.0	1.8

Source: Authors' analysis of CPS ORG data; inflation-adjusted using CPI-U-X1.  

TABLE 4
Distribution of Family Income, 1947-96
(Percent of All Families)

Year	Lowest Fifth	Second Fifth	Middle Fifth	Fourth Fifth	Top Fifth	Top 5%	Gini Coefficient
1947	5.0	11.9	17.0	23.1	43.0	17.5	0.376
1967	5.4	12.2	17.5	23.5	41.4	16.4	0.358
1973	5.5	11.9	17.5	24.0	41.1	15.5	0.356
1979	5.4	11.6	17.5	24.1	41.4	15.3	0.365
1989	4.6	10.6	16.5	23.7	44.6	17.9	0.401
1990	4.6	10.8	16.6	23.8	44.3	17.4
1991	4.5	10.7	16.6	24.1	44.2	17.1
1992	4.3	10.5	16.5	24.0	44.7	17.6
1993	4.1	9.9	15.7	23.3	47.0	20.3
1994	4.2	10.0	15.7	23.3	46.9	20.1
1995	4.4	10.1	15.8	23.2	46.5	20.0
1996*	4.2	10.0	15.8	23.1	46.8	20.3	0.425
Percentage-point Change
1947-67	0.4	0.3	0.5	0.4	-1.6	-1.1
1967-73	0.1	-0.3	0.0	0.5	-0.3	-0.9
1973-79	-0.1	-0.3	0.0	0.1	0.3	-0.2
1979-89	-0.8	-1.0	-1.0	-0.4	3.2	2.6
1989-96	-0.4	-0.6	-0.7	-0.6	2.2	2.4

*These shares allow the top-code to increase and thus show greater inequality.

Source: U.S. Bureau of the Census.  

TABLE 5
Upper Limit of Annual Family Income, by Income Group, 1947-96
(1997 Dollars)

Year	Lowest Fifth	Second Fifth	Middle Fifth	Fourth Fifth	Lower Limit
					Top 5%	Average
1947	10,506	16,952	22,988	32,618	53,536	23,517
1967	18,167	29,823	39,992	54,827	88,094	38,914
1973	20,678	34,120	47,606	65,468	102,063	46,321
1979	21,388	35,169	49,825	68,607	110,064	48,402
1989	20,714	36,242	52,810	77,079	128,093	53,723
1990	20,687	35,666	51,625	75,510	125,696
1991	20,033	34,304	50,672	74,229	121,169
1992	19,119	33,946	50,335	73,271	121,275
1993	18,849	33,321	50,016	74,190	125,714
1994	19,429	33,897	50,901	75,807	130,006
1995	20,083	34,738	51,588	76,101	130,228
1996	20,132	35,102	52,258	77,044	130,937	54,908
Annual Percent Change
1947-67	2.8	2.9	2.8	2.6	2.5	2.5
1967-73	2.2	2.3	2.9	3.0	2.5	2.9
1973-79	0.6	0.5	0.8	0.8	1.3	0.7
1979-89	-0.3	0.3	0.6	1.2	1.5	1.0
1989-96	-0.4	-0.5	-0.1	0.0	0.3	0.3

Source: U.S. Bureau of the Census.  

TABLE 6
Persons in Poverty, 1967-96
(Percent of Population)

Year	Persons with Income:			Official Rate
	Less than 1/4 the Median	1/4 to 1/2 of the Median	Less than 1/2 of the Median
1967				14.2
1973			18.9	11.1
1979	6.7	13.3	20.1	11.7
1989	8.3	13.7	22.1	12.8
1990			21.8	13.5
1991			22.3	14.2
1992	9.1	13.6	22.8	14.8
1993			22.8	15.1
1994			22.6	14.5
1995	8.6	13.6	22.2	13.8
1996	8.6	13.7	22.3	13.7
Percentage-point Change
1967-73				-3.1
1973-79			1.2	0.6
1979-89	1.6	0.4	2.0	1.1
1989-96	0.3	0.0	0.2	0.9

Source: U.S. Bureau of the Census, published and unpublished data.  

TABLE 7
Changes in the Distribution of Household Wealth*, 1962-97**

Wealth Class	Share of Wealth*						Change
	1962	1983	1989	1995	1997**		1962-83	1983-89	1989-95	1989-97**
Top Fifth	81.0	81.3	83.5	83.7	84.3		0.4	2.2	0.2	0.8
Top 1%	33.4	33.8	37.4	37.6	39.1		0.3	3.6	0.2	1.8
Next 4%	21.2	22.3	21.6	22.3	22.3		1.2	-0.8	0.7	0.7
Next 5%	12.4	12.1	11.6	11.7	11.4		-0.2	-0.5	0.1	-0.2
Next 10%	14.0	13.1	13.0	12.1	11.5		-0.9	-0.1	-0.8	-1.5
Bottom
Four-fifths	19.0	18.7	16.5	16.3	15.7		-0.4	-2.2	-0.2	-0.8
Fourth	13.4	12.6	12.3	11.5	10.8		-0.8	-0.2	-0.8	-1.5
Middle	5.4	5.2	4.8	4.5	4.4		-0.2	-0.4	-0.3	-0.4
Second	0.9	1.2	0.8	0.9	1.0		0.2	-0.3	0.0	0.1
Lowest	-0.7	-0.3	-1.5	-0.7	-0.5		0.4	-1.2	0.8	1.0
Total	100.0	100.0	100.0	100.0	100.0

*Net worth.
**1997 figures are estimates.

Source: Authors' analysis of Wolff (1998).  

TABLE 8
Household Assets and Liabilities by Wealth Class, 1962-97*
(Thousands of 1995 Dollars)

Assets and   Liabilities	Top 1.0%	Next 9%	Next 10%	Next 20%	Middle 20%	Bottom 40%	Average
A. Stocks**
1962	2,252.4	115.2	12.8	4.1	1.1	0.3	35.8
1983	1,462.4	94.4	11.3	4.3	1.5	0.4	25.9
1989	1,103.9	121.3	23.8	8.3	3.5	0.6	27.3
1995	1,656.9	185.0	34.8	13.2	5.3	1.1	41.1
1997*	2,346.2	262.0	32.9	18.6	7.5	1.5	58.1
1962-83	-790.0	-20.8	-1.5	0.2	0.4	0.1	-9.9
1983-89	-358.5	26.9	12.5	4.0	2.0	0.3	1.4
1989-95	553.0	63.7	11.0	4.9	1.8	0.5	13.8
1989-97	1,242.3	140.7	9.1	10.3	4.0	0.9	30.8
B. All Other Assets
1962	2,450.3	423.1	201.0	111.8	60.5	14.4	122.2
1983	5,628.6	730.6	295.3	152.0	74.8	15.7	202.9
1989	7,823.0	803.1	317.5	173.4	83.3	18.1	240.3
1995	6,313.0	663.8	257.3	143.3	82.3	19.3	203.2
1997*	7,471.7	721.5	271.6	149.8	85.5	19.9	224.4
1962-83	3,178.3	307.5	94.3	40.2	14.3	1.4	80.7
1983-89	2,194.4	72.5	22.2	21.4	8.5	2.4	37.4
1989-95	-1,510.0	-139.3	-60.2	-30.1	-1.0	1.2	-37.1
1989-97	-351.3	-81.6	-45.9	-23.6	2.2	1.9	-15.9
C. Total Debt
1962	166.3	32.5	24.1	24.9	24.7	13.9	22.3
1983	382.5	63.6	46.0	31.3	24.4	11.7	30.0
1989	417.1	85.0	45.9	41.5	31.8	22.5	39.9
1995	361.6	84.3	44.4	39.7	41.7	19.4	39.8
1997*	343.5	77.5	40.9	37.7	39.6	18.4	37.8
1962-83	216.2	31.1	21.9	6.4	-0.3	-2.2	7.7
1983-89	34.6	21.3	-0.1	10.2	7.4	10.8	9.9
1989-95	-55.5	-0.7	-1.5	-1.8	9.9	-3.1	-0.1
1989-97	-73.6	-7.4	-5.0	-3.8	7.8	-4.1	-2.1
D. Net Worth
1962	4,536.3	505.7	189.7	90.9	36.8	0.7	135.7
1983	6,708.5	761.3	260.6	124.9	51.9	4.4	198.8
1989	8,509.8	839.6	295.4	140.3	55.0	-3.8	227.7
1995	7,608.3	764.5	245.8	116.7	45.9	1.0	204.5
1997*	9,474.3	905.9	277.3	130.7	53.4	3.1	244.7
1962-83	2,172.2	255.6	70.9	34.0	15.1	3.7	63.1
1983-89	1,801.3	78.3	34.8	15.4	3.1	-8.2	28.9
1989-95	-901.5	-75.1	-49.6	-23.6	-9.1	4.8	-23.2
1989-97	964.5	66.4	-18.1	-9.6	-1.6	6.9	17.0

*Projections.
**All direct and indirect stock holdings.

Source: Authors' analysis of Wolff (1998).