Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 52
America Becoming: Racial Trends and Their Consequences - Volume II 4 Race and Ethnicity in the Labor Market: Trends Over the Short and Long Term James P.Smith Of the many disturbing labor-market trends in recent years, the stagnated wage gap between races may be the most disheartening. Race continues to be America’s most persistent area of social and economic disparity. Many Americans were encouraged by the steady and significant economic progress Blacks made after World War II. The recent stagnation, however, challenges that optimism. In addition, the average economic status of Hispanics appears to be deteriorating at an even more alarming rate than that of Blacks. This paper describes major, long-term trends that have had an impact on the economic status of Blacks and Hispanics, including long-term trends that appear to be influenced mostly by skill-related factors. Also addressed are alternative explanations for the 1960s-to-1990s stagnation in the economic position of minority households; explanations include changes in schooling, quality of students, affirmative action, and rising wage inequality. In addition, the role of immigration in altering the labor-market position of Hispanic workers is analyzed. LONG-TERM WAGE TRENDS Since 1940, the American economy has enjoyed substantial economic growth; inflation-adjusted incomes of all its citizens have risen dramatically. For example, real incomes of White men expanded almost threefold between 1940 and 1990, but this improvement was surpassed by even more rapid earnings growth among Black men, whose real incomes more
OCR for page 53
America Becoming: Racial Trends and Their Consequences - Volume II than quadrupled during the same 50-year period (for detailed comparisons, see Smith and Welch, 1989). Not only did the standard of living of Black men improve as measured against earlier Black generations, it rose relative to their White contemporaries. Table 4–1 shows comparative increases in the relative economic status of Black and Hispanic men from 1940 to 1990. In 1940, the average Black male worker earned only 43 percent as much as his White counterpart; by 1990, it was 75 percent as much. The pace at which Blacks were able to narrow the wage gap, however, was far from uniform. The largest improvement occurred during the 1940s; during the 1950s, advances slowed considerably; during the 1960s and 1970s, the rise in Black men’s wages was more than 10 percent higher than the rise in White men’s wages; but after 1980, the pace of relative labor-market progress for Blacks slowed considerably. Although the improvement in the relative economic status of Blacks from 1940 to 1990 was impressive, by 1990 incomes of Black males were still significantly lower than those of White males. The description of the last half century’s racial income differences has two messages: (1) considerable progress has been made in eradicating the wage gap between the races; (2) but progress has not eliminated race as an important predictor of an individual’s income. Table 4–1 also shows a remarkably constant wage gap for Hispanics from 1940 to 1990. In 1990, Hispanics earned 67 percent as much as U.S.-born White men, only slightly higher than the Hispanic-White wage gap of 1940. This aggregate stability, however, hides important changes over time. For example, from 1970 to 1990, there was a steady deterioration in the relative economic status of Hispanic men, as their wages decreased by 16 percent compared to White men. The lack of Hispanic economic progress is most apparent when com- TABLE 4–1 Minority Male Wages as a Percent of White Male Wages Minority Group 1940 1950 1960 1970 1980 1990 Blacks 43.3 55.2 57.5 64.4 72.6 74.5 All Hispanics 64.2 73.8 70.2 73.7 70.7 67.3 Mexicans 55.6 71.3 70.0 70.1 68.0 63.0 Puerto Ricans 82.9 71.5 61.3 66.7 66.1 74.5 Cubans n.a. n.a. n.a. 75.6 82.8 86.6 Other Hispanics 82.1 85.4 82.3 82.7 77.6 71.9 Blacks as a percent of Hispanics 67.4 74.8 81.9 87.4 1.03 1.07
OCR for page 54
America Becoming: Racial Trends and Their Consequences - Volume II pared to Blacks. In 1940, the average Black male worker earned two-thirds as much as the average Hispanic male worker, but in 1990, Black men earned approximately 7 percent more than Hispanic men, on average. One tale of these two minorities is the significant progress achieved by Black men over the last half century. No such story of progress is possible for Hispanics, who seem over the long term to have stagnated and, in recent years, to have deterioriated. This sharp contrast between Hispanics and Blacks suggests that, for Hispanics, forces related to immigration, such as immigration status and English language proficiency, may have played a central role. Although often viewed as an aggregate, there has always been considerable economic heterogeneity within the Hispanic population. Among the major Hispanic subgroups, Mexicans have always fared the worst economically. In 1940, when the other Hispanic ethnic groups were being paid more than 80 percent of wages paid to U.S.-born White men, Mexican men were earning only 56 percent. By 1950, however, Mexicans were earning 71 percent of what White men earned, a ratio that stayed constant for the next 20 years. After 1970, however, Mexican relative wages declined steadily, expanding their wage disparity to its highest level in more than 40 years. THE POOR, THE MIDDLE CLASS, AND THE AFFLUENT Trends in Economic Status The distribution of this long-term labor-market progress is addressed in Table 4–2. Building on the simplicity of the poverty line, all workers are divided into three wage classes—poor, middle class, and affluent.1 Coming out of the Depression era in 1940, 31 percent of working White men had jobs that placed them in the ranks of the poor. The 1 Since the first attempts to measure poverty, debate has continued as to whether poverty is an absolute or relative concept. To determine the percent of the population to be categorized as poor, I used elements of both the absolute and the relative definitions. It turns out that my definition corresponds more closely to most people’s notions of what poverty means. Over time, when asked in surveys about the amount of income required not to be poor, the poverty threshold has increased roughly 50 cents for every $1 increase in real income. Based on that observation, my definition of poverty increases the poverty threshold income by 0.5 percent for every 1 percent growth in real income. For 1979, I selected as the initial criterion an income level such that 11 percent of average White male earnings equaled “poor.” This poverty threshold was then adjusted for any real income growth or contraction relative to that year. My definition of affluent is asymmetric—i.e., to be affluent, one must have an income equal to 1.33 percent of the White median income for that year (Smith and Welch, 1989).
OCR for page 55
America Becoming: Racial Trends and Their Consequences - Volume II TABLE 4–2 Income Group Status of Male Workers (percent) 1940 1950 1960 1970 1980 1990 White Men Poor 31 18 13 9 11 12 Middle Class 38 59 63 65 61 60 Affluent 31 23 24 26 28 28 Black Men Poor 74 44 37 25 20 23 Middle Class 22 51 58 68 67 64 Affluent 4 5 5 7 13 13 Hispanic Men Poor 57 32 27 17 23 27 Middle Class 34 61 66 75 66 62 Affluent 9 7 7 8 11 11 Mexican Men Poor 63 37 29 21 24 30 Middle Class 31 57 64 70 65 60 Affluent 6 6 7 9 11 10 Puerto Rican Men Poor 33 23 27 15 21 18 Middle Class 49 71 70 80 70 69 Affluent 18 6 3 5 9 13 Other Hispanics Poor 45 19 18 13 21 24 Middle Class 40 68 70 73 64 63 Affluent 15 13 12 14 15 13 situation for Blacks and Hispanics was far worse. In 1940, the overwhelming majority of Blacks were poor; three-quarters were destitute, with little hope that their lot or even that of their children would improve. The Black middle class then comprised only one-fifth of all Blacks. On the other extreme, the econmic elite resembled an exclusive White club. Similarly, more than one-half of all Hispanic men in 1940 worked in jobs that confined them and their families within the ranks of the poor; only one in three earned middle-class wages; and the Hispanic affluent comprised one-eleventh of that population. Among Hispanics, Mexicans fared the worst; almost two-thirds of working Mexican men earned wages below the poverty threshold. The subsequent changes have been dramatic. Driven by economic growth and improvements in the skills of the workforce, poverty rates
OCR for page 56
America Becoming: Racial Trends and Their Consequences - Volume II declined rapidly for the White male majority. Between 1940 and 1970, median White wages grew by 3.2 percent per year, a growth that was fairly uniform across the wage distribution. By 1970, only one in every 11 White male workers earned wages below the poverty threshold, and almost two-thirds earned middle-class incomes. Unfortunately, this historic trend reversed during the 1970s and 1980s. The stagnant economic conditions of those decades, combined with expanding wage inequality, led to an increase in both the percentage of White men who were poor and the percentage who were affluent. The real story of the 1940s–1990s was the emergence of the Black and Hispanic middle class, whose income gains were real and substantial. Unfortunately, as was true for the White majority, these gains in poverty reduction reversed, but at a more rapid rate for Blacks and Hispanics in the 1970s and 1980s. Since 1980, there has been a more than 10 percent increase in the relative numbers of Black and Hispanic working men who were poor. Notwithstanding these downturns, the growth of the Black middle class was so spectacular that in 1990 it outnumbered the Black poor. In 1990, about two-thirds of Blacks and Hispanics had incomes that met the criteria for middle class. In addition, the odds of a Black man penetrating the ranks of the affluent tripled. Trends in Education A basic index of the skill workers bring with them to the labor market is the number of years of schooling completed. Because, on average, Blacks and Hispanics complete fewer years of schooling than Whites, education should play a central role in explaining both levels and trends in their wage gaps. It does. Table 4–3 lists mean years of schooling completed for White, Black, and Hispanic males. To highlight differences, education deficits of Blacks and Hispanics, compared to Whites, are also shown. Not surprising, among all groups, education levels of each new generation increased from 1940 to 1990. Although this secular improvement exists for men of both races, data in Table 4–3 indicate that improvement was much sharper among Black men. Educational differences still persist between Blacks and Whites, but to a lesser extent in 1990 than at any other time in American history. In 1990, the average Black male had completed 1.1 fewer years of schooling than the average White male, representing a steady and continuous decrease from the 3.7-years difference in 1940. Between 1940 and 1990, almost three-quarters of the education gap between Blacks and Whites was eliminated (Smith and Welch, 1989). The rate of secular improvement in Hispanic schooling by 1990 was slower and more uneven. Although Black men erased three-quarters of
OCR for page 57
America Becoming: Racial Trends and Their Consequences - Volume II TABLE 4–3 Education Levels of Males Year White Black Hispanic Mexican Puerto Rican Cuban Other Hispanic A. Average Education Levels of Males 1990 13.30 12.19 10.57 9.97 11.37 12.28 11.45 1980 12.76 11.37 10.18 9.57 10.09 11.72 11.37 1970 11.84 9.82 9.52 8.76 8.92 10.75 10.82 1960 10.37 7.54 7.88 7.34 7.77 n.a. 9.94 1950 10.23 6.71 6.61 6.08 7.73 n.a. 8.07 1940 9.48 5.74 5.95 5.34 7.95 n.a. 7.15 B. Education Deficits Compared to White Men 1990 -0- 1.11 2.73 3.33 1.93 1.02 1.85 1980 -0- 1.39 2.58 3.19 2.67 1.04 1.39 1970 -0- 2.02 2.32 3.08 2.92 1.35 1.02 1960 -0- 2.83 2.49 3.03 2.60 n.a. 0.43 1950 -0- 3.52 3.62 4.15 2.50 n.a. 2.16 1940 -0- 3.74 3.53 4.14 1.53 n.a. 2.33
OCR for page 58
America Becoming: Racial Trends and Their Consequences - Volume II their educational disparity with White men, Hispanics were able to eliminate less than one-third of their initial 1940 deficit. In the process, their education ranking was reversed. Hispanics had a 0.2-year lead over Black men in 1940; by 1990, Black men had more than a 1.5-year schooling advantage. The 1990 Hispanic education gap with White men was nearly two and a half times as large as the schooling gap between Blacks and Whites in that year. To understand reasons for these disparities, it is necessary to distinguish among Hispanic immigrants and U.S.-born Hispanics. Because immigrants tend to have much less schooling than do U.S.-born Hispanics, secular trends in schooling can be quite sensitive to swings in the size of immigration flows. Some insight into the central role of immigration is suggested by the data in Table 4–4 where Hispanic education levels are listed (and their deficits compared with Whites) by U.S. birth or foreign birth and whether they were recent immigrants—i.e., arrived within the past five years. Data in Table 4–4 indicate that the changing composition of recent immigration and the increasing percentage of immigrants within the Hispanic population are two dominant underlying trends. Given the better educational opportunities available in the United States, compared to those in their home countries, it is not surprising that U.S.-born Hispanic men have more schooling than their foreign-born counterparts; however, the different secular trends for the U.S.-born and foreign-born Hispanics are more surprising. From 1940 to 1990, the education disparity between TABLE 4–4 Male Hispanic Years of Schooling Completed, by Nativity 1990 1980 1970 1960 1950 1940 A. Average Education Levels of Males All Hispanics 10.57 10.18 9.37 7.88 6.61 5.95 U.S. born 11.98 10.93 9.80 8.18 7.04 6.18 Foreign born 9.36 9.24 9.26 7.17 5.24 5.27 1–5 years in U.S. 8.96 8.36 9.13 8.47 n.a. 8.21 6 or more years in U.S. 9.50 9.56 9.33 6.75 n.a. 5.23 B. Education Deficits Compared to White Men All Hispanics 2.66 2.49 2.47 2.49 3.62 3.53 U.S. born 1.25 1.74 2.04 2.19 3.19 3.30 Foreign born 3.77 3.43 2.58 3.20 4.99 4.21 1–5 years in U.S. 4.27 4.31 2.71 1.90 n.a. 1.27 6 or more years in U.S. 3.73 3.11 2.51 3.62 n.a. 4.25
OCR for page 59
America Becoming: Racial Trends and Their Consequences - Volume II U.S.-born Hispanics and U.S.-born Whites steadily narrowed. In 1940, U.S.-born White men had a 3.3-year schooling advantage over U.S.-born Hispanic men. By 1990, 60 percent of this deficit had been eliminated, and U.S.-born Hispanics had a deficit of 1.3 years. Similar trends are found for all Hispanic groups, especially for the numerically important Mexican subpopulation. A far different picture emerges among the foreign-born. Not only are their disparities with White men considerably larger, there is no longer a trend of uniform progress. In particular, from 1970 to 1990, the era of reversal in the aggregate data, the education gap for foreign-born men increased significantly. Indeed, the mean education of foreign-born Hispanic men is little higher now than it was in 1970. Compared to U.S.-born Whites, the education deficit of foreign-born Hispanic workers rose from 2.58 years in 1970 to 3.77 years in 1990. The force of these changes is most apparent among recent immigrants, who represent a better index of the education of newly arriving immigrants. From 1940 to 1990, there was a steady deterioration in the relative education levels of new Hispanic immigrants. In 1940, compared to U.S.-born White men, new Hispanic immigrants had a deficit of 1.3 years; by 1990, the deficit had risen to 4.3 years. Education deficits of recent immigrants accelerated after 1970. In sum, the slow rate of Hispanic educational progress largely reflects a changing composition of the Hispanic immigrant workforce. The rising percentage of immigrants in the Hispanic male workforce in recent decades slowed aggregate gains in Hispanic schooling. Increasing numbers of poorly educated Mexicans among Hispanic immigrants also served to lower the mean schooling advances achieved. The aggregate education data for all Hispanics raised an important dilemma best highlighted by comparing limited aggregate Hispanic education gains with the substantial gains achieved by Blacks from 1940 to 1990. If we limit our comparison to Blacks and U.S.-born Hispanics, the dilemma is resolved. Both groups show, in 1990, a substantial narrowing of their education deficits with White men. Hispanics born in the United States seem no less able than Blacks to improve their educational position over time. RECENT LABOR-MARKET WAGE TRENDS In a number of important ways, the long-term historical trends begun in 1940 did not continue from the 1960s to the 1990s. In this section, trends in weekly wages by year, race, ethnicity, and gender from the 1960s to 1990 are examined using the yearly March Current Population Surveys (CPS), starting in 1962. To set an overall context, Figure 4–1 shows yearly trends in mean
OCR for page 60
America Becoming: Racial Trends and Their Consequences - Volume II FIGURE 4–1 Yearly trends (1962 to 1997) in mean inflation-adjusted weekly wages for (A) males and (B) females.
OCR for page 61
America Becoming: Racial Trends and Their Consequences - Volume II inflation-adjusted weekly wages for working Black, Hispanic, and White men and women. Data for White men can serve as an index of what was happening with average wages for all groups. As an approximation, secular trends can be separated into three periods. From 1962 to 1973, real wages grew (1.6 percent per year). Then followed a sharp decline until 1981, when the average real wage of White men fell 14 percent from the 1973 high. This decrease was so steep that real wages in 1981 were only 3 percent higher than in 1962. Fortunately, the years since 1981 were ones of recovery, with real wages of White men in 1996 15 percent higher than at the 1981 trough (1 percent per year growth). Time-series trends for other groups primarily mimic trends for White men; however, of interest here are those instances of departure from the White male series. These departures are captured in Figure 4–2, which measures yearly wage gaps, as percentages, for Hispanic and Black men and women, relative to White male wages. Consider, first, working Black men. As was true for White men, real wages among Black men increased from 1962 to 1973, fell from 1973 to 1981, and then gradually rose. Their trends, however, were far from identical. In particular, from 1962 to 1976 the wage gap between Black and White males decreased sharply, from more than 50 percent at the beginning of the period, to about 32 percent at the end. Then, relative progress ceased and actually reversed, as the gap increased to 41 percent by 1986. Given the noise in the data, it is difficult to know with complete confidence what happened since, but a reasonable characterization would be a modest but steady narrowing of the male racial wage gap. CPS categorization of data based on Hispanic ethnicity began in 1971. Since then, trends in the wage gap among Hispanic men could not be more different. Throughout the 1970s, the Hispanic male wage deficit, using White male data as an index, held steady at about 30 percent. Since the 1980s, however, this wage gap grew, and reached about 45 percent by the mid-1990s. If we examine data for Mexicans alone, the trends are similar, except that the wage gap is about 5 to 8 percent greater than that observed for all Hispanics. A different pattern again emerges by gender. White female wage gaps expanded from the early 1960s to the mid-1970s—the same years when Black men enjoyed their largest gains. Since the mid-1960s, however, there has been a long, sustained improvement in the wage position of working women. By 1997, the White female wage deficit was about 50 percent, compared to a peak wage deficit of 73 percent in 1973. Black and Hispanic women exhibit their own unique patterns. By far, the largest relative wage gains were made by Black women. The pace of early improvements was staggering. From 1962 through 1973, when real wages among White men increased by 17 percent, real wages of Black women
OCR for page 62
America Becoming: Racial Trends and Their Consequences - Volume II FIGURE 4–2 Percentage wage deficits, relative to White males, for Black and Hispanic (A) males and (B) females.
OCR for page 87
America Becoming: Racial Trends and Their Consequences - Volume II FIGURE 4–10 Career wage growth of new female immigrant cohorts. Each group consists of 25–34 year-old women. (A) All new immigrants. (B) New immigrants from Mexico. implies that the true relative life-cycle wage progression of Mexican immigrants may actually be even more negative. Why would this be so? To date, there is no convincing answer to this question, which should receive high priority in the research agenda. One can speculate about the role of language or the implications of the geo-
OCR for page 88
America Becoming: Racial Trends and Their Consequences - Volume II graphic closeness to country-of-origin, but there is little concrete evidence to document any compelling explanation. Other immigrant groups— Asians would be a good example—arrive without complete English language proficiency, and many immigrants have frequent trips back and forth to their home countries. Part of the problem lies in the inherent ambiguity in using tracking of cohorts across census or CPS files to evaluate life-cycle progress of immigrants. Although cohort tracking has become the standard technique for evaluating economic assimilation, it is problematic as immigrant cohorts are not closed. An initial immigrant cohort can be depleted as some immigrants return home. If, as seems likely, those immigrants who left the country were highly selective, the wage trajectories obtained from cohort tracking will be biased. For example, if high-skill/high-wage immigrants left, average wages of the remaining members of the cohort would decrease even if the wage of every remaining immigrant stayed the same. This problem caused by out-migration from an initial entering immigrant cohort is especially severe among immigrants from Mexico. In the aggregate, roughly one-third of the 1970 Mexican immigrants emigrated by 1980. An even smaller percentage of the 1970 cohort remained by 1990. Until this problem of the nature of the selectivity of emigration of previous Mexican immigrant cohorts can be resolved, one should be cautious about reaching any strong conclusions about the nature of life-cycle labor-market careers of Mexican immigrants. Generational Assimilation Regarding the issue of generational assimilation, the conventional wisdom for Hispanics—whom some argue have not enjoyed the same level of success of earlier European immigrants—leans toward the pessimistic side. The reasons for pessimism vary, but one theme is that Hispanic immigrants and their children may be less committed to assimilation than the Europeans were. The data supporting this concern are often derived using cross-sectional comparisons between first-, second-, and third-generation Hispanics of their income and schooling levels. Such comparisons universally show a narrowing of the schooling and income gap between the first and second generation, but either retrogression or little progress between the second and third generation (Smith, 1999). Although conclusions about generational assimilation are often drawn from such data, these inferences are not appropriate. In any cross section, members of the second generation are not sons and daughters of current immigrants, and members of the current third generation in a cross section are not direct descendants of current, second-generation persons. Fortunately, the conventional wisdom appears to be in error. In Smith
OCR for page 89
America Becoming: Racial Trends and Their Consequences - Volume II (1999) the data are arrayed in a way that more directly tracks progress made across generations. The schooling deficits of Hispanics are uniformly smaller in the second generation than in the first and smaller still in the third generation. For example, the mean education disparity among all first-generation Mexicans was 4.94 years and decreased to 2.95 years among second-generation Mexicans. The youngest third-generation cohorts had a schooling gap of less than 1 year compared with White men— half as large as their fathers’ education deficit. At least based on the historical record, fears about Hispanic generational assimilation appear to be unwarranted as second- and third-generation Hispanic men have made considerable strides in narrowing their economic disparities with U.S.-born White men, as schooling gains across the generations were translated into generational progress in incomes. Each new Hispanic generation not only had higher incomes than their predecessors, but their economic status converged relative to White men with whom they had to compete. HOUSEHOLD WEALTH Until recently, data limitations forced most comparisons of racial economic status to rely only on income differences, but improvements in measuring wealth have made contrasts of household wealth levels feasible. Table 4–10 lists mean and median household wealth levels by race derived from the Panel Study of Income Dynamics (PSID).3 Racial wealth gaps are extremely large, especially compared to the already sizable household income differences by race. In 1984, mean wealth of non-White households was 22 percent of wealth of White households, and the income ratio was 0.58:1. Wealth differentials are even larger if medians are used as the yardstick; using medians, in 1984 non-White households had less than 10 percent of the wealth of White households. The glimmer of hope is that the relative wealth differentials narrowed over the 10-year period covered in this table. By 1994, mean non-White wealth rose to 31 percent of that of Whites. Racial and ethnic disparities are even greater in financial assets; these 3 PSID is a longitudinal survey of a representative sample of U.S. individuals (men, women, and children) and the families in which they reside. The study has been conducted at the Survey Research Center, University of Michigan, since its beginning in 1968, with the Inter-University Consortium for Political and Social Research (ICPSR). Wealth modules were included in the 1984, 1989, and 1994 PSIDs. See Juster et al. (1999) for a detailed discussion, and see Browning and Lusardi (1996) for an excellent review of the micro savings literature.
OCR for page 90
America Becoming: Racial Trends and Their Consequences - Volume II TABLE 4–10 Wealth and Income Levels by Race Wealth Income Mean Median Mean Median A. Total Household Wealth White 1984 169.0 59.4 48.9 39.1 1989 181.1 59.6 59.6 39.5 1994 178.5 64.7 52.9 41.0 Non-White 1984 37.1 5.3 28.5 21.7 1989 53.5 6.8 29.9 22.3 1994 54.5 10.4 31.9 23.8 B. Financial Assets White 1984 48.0 6.0 1989 48.6 7.4 1994 60.9 13.7 Non-White 1984 7.2 0.0 1989 8.6 0.0 1994 13.7 0.0 SOURCE: PSID—1996 dollars. Calculations by author. more-liquid assets may be a better index of resources a household has on hand to meet emergencies. In 1984, mean financial assets for non-Whites were one-seventh of those of White households. Not only is the ratio of financial assets by race low, the financial assets held by non-White households are meager. In 1984, non-White households had a little more than $7,000 per household in financial assets. But even this number exaggerates their holdings due to the extreme skew in the distribution. In all three years, the median non-White household had no financial assets at all. A more complete description of racial wealth differences is given in Figure 4–11, which plots, for Whites and non-Whites, household wealth at percentiles of the wealth distribution. These data illustrate the extreme skew in wealth holdings—the top 5 percent of White households have 50 percent more wealth than White households at the 90th percentile, while those at the 90th percentile have more than five times as much as the median White household. This nonlinearity prevails within the lower half of the wealth distribution as well, as the median White household has
OCR for page 91
America Becoming: Racial Trends and Their Consequences - Volume II FIGURE 4–11 White and non-White wealth distribution in 1984 (in 1996 dollars). (A) Percentiles 1 to 50. (B) Percentiles 51 to 98.
OCR for page 92
America Becoming: Racial Trends and Their Consequences - Volume II 10 times as much wealth as those at the 20th percentile. A similar extreme skew characterizes the non-White wealth distribution. Why are racial wealth differentials at least twice as large as house-hold-income differences in the same years? One possibility can easily be dismissed; it is not a consequence of financial wealth being transmitted across generations with the poor unable to give and the affluent ensuring their heirs remain at the top through financial inheritances. Although plausible, this possibility is quantitatively unimportant as the vast majority of households—both White and non-White—do not receive any financial inheritances.4 Using PSID, mean inheritances (in 1996 dollars) for Black households were about $1,000; for White households about $10,000. Even if these inheritances were completely saved so that they show up in current household wealth, they would account for a small percentage of racial wealth differences documented in Table 4–11. Similarly, two-thirds of all White households and 90 percent of all minority households received no financial inheritances by the time the householders were in their mid-50s. Racial disparities in wealth would be almost the same if we subtracted that part of current wealth derived from past financial inheritances (Smith, 1995).5 If not financial inheritances, then all we have left is people saving at different rates from their income and/or experiencing different ex-post rates of return on their savings. The question then becomes, Why do Black and Hispanic households save so much less than White households? This is a much under-researched question partly because of the lack of adequate data. Because wealth disparities far exceed income disparities, there has been some thought that the reasons for the lack of savings behavior must lie in some unique historical events specific to the Black or Hispanic experience. For Blacks, it is sometimes argued that a culture promoting savings was not encouraged or was too difficult to develop. However, the data presented in Figure 4–11 suggest that it is premature to jump to race- and ethnic-specific explanations. 4 This point about the relative unimportance of past bequests in creating wealth differences in the current generation is fundamentally different than the debate about the importance of the bequest motive in accounting for savings behavior of the current generation. The current generation’s savings behaviors are forward looking, so that any savings for bequests by the current generation are meant for the subsequent generation. Given the secular rise in bequests, savings for bequests may be a large part of a current generations’ savings, while their receipt of past financial inheritances are inconsequential to their own present wealth holdings. 5 The argument here deals only with bequests and does not speak to the issue of the role of inter-vivos transfers in creating racial and ethnic differences in wealth. See Gale and Scholz (1994) for a good discussion of inter-vivos transfers.
OCR for page 93
America Becoming: Racial Trends and Their Consequences - Volume II TABLE 4–11 Wealth by Source (000 of dollars) Household Wealth Social Security Pensions Total White $264 $124 $109 $503 Black 72 94 65 231 Hispanic 80 94 39 218 SOURCE: Smith (1995). Once again, the reason lies in a concave relation between savings or wealth and household income. This nonlinear relation between savings and income explains a good deal of the racial wealth discrepancies. Although vastly less than average White household wealth, wealth of the median Black household is actually quite similar to Whites with the same income. In 1994, White households with incomes equal to the median Black household lie at the 25th percentile of the White income distribution. If we compare wealth of the median Black to the 25th percentile White, their wealth levels are quite similar. When the nonlinearity is taken into account, income explains a good deal of racial differences in wealth. If low savings behavior is not a racial or ethnic issue, the unanswered question is, Why do low- and middle-income people save so little, no matter what their race or ethnic background? Different racial groups may also experience different ex-post rates of return to their past savings, which may expand or contract wealth differences between them. For example, the surge in the stock market during the last 15 years increased the wealth of those households with greater amounts of stock market holdings. Because rates of stock ownership and holdings were larger among White households, wealth of White households increased more than wealth of Black households. This dismissal of financial inheritance as an important source of racial differences in household wealth does not imply that all forms of intergenerational transmission are unimportant. For example, the inheritance of human capital is another source of intergenerational transmission that clearly creates racial and ethnic differences in income. Indeed, one important form in which these differences in inheritances of human capital show up are the education differences discussed earlier. A second factor distorting racial wealth comparisons is that household wealth represents only part of the wealth households have at their disposal. Despite its widespread use, household wealth ignores large components of wealth that are critical to many households. For example, a household’s future expected Social Security benefits are a lifetime annu-
OCR for page 94
America Becoming: Racial Trends and Their Consequences - Volume II ity that can be discounted to give a present value of Social Security wealth. In a similar vein, private pensions, either directly in defined-contribution plans or indirectly for defined-benefit plans, are an important source of wealth for many households, especially in their pre- and postretirement years. Virtually all households in their 50s anticipate Social Security benefits when they retire, and more than half are counting on income from their pensions. When discounted to the present, these expected income flows translate into considerable wealth. Table 4–11 demonstrates how large they actually are for families with one member aged 51 to 61. Mean household, Social Security, and pension wealth are listed in the table, providing a better measure of wealth than the conventional concept, which counts only household wealth. For Whites, total wealth is half a million dollars, rather than the quarter of a million in conventional household wealth. More important, the distortion caused by conventional wealth is much larger among minority families. Among Blacks and Hispanics, conventional household wealth is less than one-third of total wealth and Social Security represents the largest part of their wealth. If the enlarged total wealth concept is used, Black households have 46 percent as much as White households do compared to 30 percent for household wealth alone. CONCLUSIONS This paper has covered some wide territory in describing the major trends that have impacted on the economic position of Blacks and Hispanics. In addition to long-term trends that appear to be influenced mostly by skill-related factors, I have also evaluated alternative explanations for the recent stagnation in the economic position of minority households. These explanations included changing schooling, quality of students, affirmative action, and rising wage inequality. In addition, the role of immigration in changing the labor-market position of Hispanic workers was evaluated. Long-term trends in the relative economic status of Blacks and Hispanics appear mainly to reflect long-term trends in their relative skills. For example, relative income differences and education deficits of Blacks compared to Whites are quite closely related. For Hispanics, it is also necessary to distinguish between immigrants and the U.S.-born. The slow rate of Hispanic educational and economic progress largely reflects a changing composition of the immigrant workforce. The rising percentage of immigrants in the Hispanic male workforce in recent decades slowed the aggregate gains in Hispanic schooling. Until the mid-1970s, schooling continued to assume its historical role as the primary determinant of the male racial wage gap; however, male
OCR for page 95
America Becoming: Racial Trends and Their Consequences - Volume II education differences by race cannot account for the timing and magnitude of the male wage stagnation of the last 25 years. Moreover, there is little evidence that the quality of Black or Hispanic students entering the labor market during the past few decades has declined. Nor can schooling account for the impressive narrowing of the gender wage gap of the last few decades. However, the stagnation and decrease in Hispanic wages relative to U.S.-born Whites is consistent with the apparent lack of relative education progress of the average Hispanic worker. In addition, affirmative action led to changes in the location of minority employment and produced significant early jumps in the wages of Black men; however, these wages gains proved to be temporary. The bulk of the remaining stagnation in minority group wages since the mid-1970s is the result, principally, of the rising wage inequality in the labor market. Because minority workers’ skills place them in the lower part of the wage distribution, increasing wage dispersion across skill levels will decrease their wages more than those of majority workers. The last 20 years were actually a time during which slowly evolving historical forces continued to close the wage gap of Black and White male workers. These forces were simply overwhelmed by the structural shift of rising wage dispersion. Because of the central role immigration plays in the Hispanic population, some additional factors are relevant when discussing their changing economic status. The well-documented decrease in wages of new Hispanic immigrants appears to reflect three forces. First, a growing skill gap reinforced by an expanding wage gap (conditional on a given skill gap), and possibly an increasing percentage of undocumented Mexican immigrants among all recent immigrants in recent census and CPS surveys. Second, across their careers in the United States, wages of Hispanic immigrants appear to hold steady relative to the White U.S.-born majority. However, Mexican immigrants appear not to do as well over their careers as immigrants from other ethnic groups. There is no consensus explanation as to why this is so. Third, at least based on the historical record, fears about Hispanic generational assimilation appear to be unwarranted as second- and third-generation Hispanic men have made considerable strides in narrowing their education and economic disparities with U.S.-born White men. Finally, I document in this paper that racial differences in household wealth are extremely large; much larger in fact than racial differences in income. However, in spite of these large racial disparities, the reasons for these large wealth disparities are unlikely to have been produced by factors that are specific to individual racial or ethnic groups. Instead, the reason appears to arise from the more general tendency of low-income households—of either race—to engage in little savings behavior. Because
OCR for page 96
America Becoming: Racial Trends and Their Consequences - Volume II there are more Black and Hispanic than White households in the low-income group, racial and ethnic differences in household wealth will be large. ACKNOWLEDGMENTS I thank David Rumpel of RAND and Joseph Lupton of the University of Michigan for expert programming assistance. Funds were provided by a grant from NICHD. Useful comments were made by Gerald Jaynes and an anonymous referee. REFERENCES Borjas, G. 1985 Assimilation, changes in cohort quality, and the earnings of immigrants. Journal of Labor Economics 3(4):35–52. Browning, M, and A.Lusardi 1996 Household savings: Micro theories and micro facts. Journal of Economic Literature XXXIV(4):1797–1855. Chay, K. 1998 The impact of federal civil rights policy on Black economic progress: Evidence from the Equal Employment Opportunity Act of 1972. Industrial and Labor Relations Review 51(4):608–632. Chiswick, B. 1978 The effect of Americanization on the earnings of foreign-born men. Journal of Political Economy 86(5):897–921. Donahue, J., and J.Heckman 1991 Continuous vs. episodic change: The impact of affirmative action and civil rights policy on the economic status of Blacks. Journal of Economic Literature 29(4):1603– 1644. Gale, W., and J.Scholz 1994 Inter-generational transfers and the accumulation of wealth. Journal of Economic Perspectives 8(4):145–160. Heckman, J., and B.Payner 1989 Determining the impact of federal antidiscrimination policy on the economic status of Blacks: A study of South Carolina. American Economic Review 79:138–177. Holzer, H., and D.Neumark 1999 Assessing affirmative action. Unpublished paper. Jasso, G., M.Rosenzweig, and J.Smith 2000 The changing skills of new immigrants, recent trends and their determinants. Pp. 185–225 in Issues in the Economics of Immigration. Chicago: University of Chicago Press. Juhn, C, K.Murphy, and P.Brooks 1991 Accounting for the slowdown in Black-White wage convergence. Pp. 107–143 in Workers and their Wages: Changing Patterns in the United States, M.Kosters, ed. American Enterprise Institute Press. Juster, F., F.Stafford, and J.Smith 1999 The measurement and structure of wealth. Labour Economics 6(2):253–275.
OCR for page 97
America Becoming: Racial Trends and Their Consequences - Volume II National Center for Education Statistics: Education Testing Service 1991 Trends in Academic Progress. Passel, J. 1999 Undocumented immigration to the United States: Numbers, trends, and characteristics. In Illegal Immigration in America, D.Haines and K.Rosenblum, eds. Greenwood Publishing. Smith J. 1993 Affirmative action and the racial wage gap. American Economic Review 83(2):79–84. 1995 Racial and ethnic differences in wealth. Journal of Human Resources 30:S156–S183. 1999 Progress across the generations. Unpublished paper. Smith, J., and B.Edmonston 1997 The New Americans: Economic, Demographic, and Fiscal Effects of Immigration. Washington, D.C.: National Academy Press. Smith, J., and M.Ward 1989 Women in the labor market and the family. Journal of Economic Perspectives 3(1):9– 24. Smith, J., and F.Welch 1984 Affirmative action and labor markets. Journal of Labor Economics 2(2):269–302. 1989 Black economic progress after Myrdal. Journal of Economic Literature 27:519–564.
Representative terms from entire chapter: