Report | Wages, Incomes, and Wealth

Recent Wage Trends


The most dramatic economic change over the last two decades has been the large increase in wage inequality in the context of stagnant growth in average wages. The consequence has been reduced earnings opportunities for a broad spectrum of the workforce. Perhaps most important, however, is that changes in the wage structure have led to a large growth of workers with low and very low earnings. Not surprisingly, these wage developments are strongly linked to the overall growth of family income inequality and to recent high levels of poverty.

This memo describes these trends and attempts to both explain them and delineate their importance for policy discussions related to ameliorating poverty, the underclass and inequality. Part I reviews the most significant wage and employment trends. Part II examines a variety of explanations of these labor market development. Part III discusses what labor market developments in the future might be expected. Last, Part IV examines the policy implications of these adverse wage and employment trends.

I. The Growth of Low Wage Work

This section presents data which describe the dramatic change in the wage structure over the last twenty years, the significant growth in the proportion of workers earning low wages, and the eroded earnings power of the vast majority of the workforce. These adverse wage developments are the primary factor behind the equally dramatic recent growth in overall income inequality, high poverty rates, and the expanded number of working poor.

Since most families primarily rely on wage income, or earnings, for their income (rather than interest, rent or government cash assistance) the starting point of our analysis is trends in the annual earning of workers in low, middle and high income families. Table 1 presents the change in annual earnings among men and women ranked by their taxable income. Data are also presented on the trends in the components of annual earnings – hourly wages and the annual hours of work – and the employment rate (the proportion of the population employed).

Earnings inequality clearly increased in the 1979-1989 period, as men in the highest income families saw their annual earnings grow by 9% while men in the bottom sixty percent of families saw their annual earnings fall by 6% to 11%. -Among women, the higher the family income, the greater was the growth in annual earnings.

The divergence in male annual earnings after 1979 is due mainly to differences in hourly wage trends rather than work hour trends. For instance, there was very little change in either the amount of time employed men worked or in the employment rate (ranging from 2% higher to 2% lower). In contrast, hourly wage trends varies widely as the wages of men in the lowest income groups fell from 7% to 11% while wages for the highest income men grew 7%. This pattern indicates that any explanation of growing male earnings inequality must lie Primarily in factors affecting hourly wages rather than shifts in men’s ability or willingness to work.

Essentially, the same pattern holds true for women as well. Women in the best off two-fifths of families saw-wages rise 17% and 28% in the 1979-1989 period, vastly different than the 3% decline among women in the bottom two-fifths of families. Although change in employment rates and annual hours were more important for women, the differences between income groups were not as marked as they were with wages.

Table 2 and Table 3 directly examine the proportion of the workforce earning poverty level wages, a proportion that grew from 21.4% in 1973 to 31.2% in 1991. There has been a general shift of the workforce out of the middle earning categories towards low-wage work except for a small growth in high earners in the 1979-1989 period. The largest growth was among workers earning very low wages – less than 75% of the poverty level wage, or $4.88 in 1991 dollars. In the 1979-1989 period, for instance, the incidence of very low wages grew from 4.1% to 13.2%.

The increased incidence of low wage work affected both men and women ( Table 2). Women are much more likely than men to earn poverty level wages (37.2% versus 25.8% in 1991), but the incidence of low wage work grew faster among men in the post-1979 period, growing to 25.8% in 1991 from 16.4% in 1979.

Since 1979, the largest growth in low wage work has occurred among minority men. While 25.1% of black men earned poverty level wages in 1979, 38.6% earned poverty wages in 1991, with most of the growth in those earning very low wages (less than 75% of the poverty level wage). Similarly, among Hispanic men the incidence of poverty level wages grew from 26.6% in 1979 to 43.7% in 1991. There was a smaller growth of low wage work among minority women, although there was a shift towards earning wages in the lowest category. Minority women are still more likely than minority men to earn low wages.

There was a lesser, but still sizeable growth in low wage work among white men and women after 1973. White women, however, were the only demographic group not to experience a higher incidence of low wage work in the 1979-1989 period, although there was a shift towards very low wages.

One of the most significant shifts in the wage structure in recent years has been the divergent trend in wages of workers with different level of wages. In economics terminology, there were significant changes in the returns to education, or the ratio of the wages of “more educated” workers (e.g., high school graduates). The data in Table 4 shows the pattern of wage by education level. In the 1979- 1989 period wages grew more rapidly, or fell less, the more educated the group. Generally, those with college degrees experienced growing real wages while those with less than a college degree experienced real wage reductions.

These trends are sometimes said to show that wages grew for “more educated” or “more skilled” workers relative to the wages of “less educated” or “unskilled” workers. A common illustration is to note that the ratio of the college to the high school wage grew from 137.7% to 155.7% from 1979 to 1989. This is said to reflect an increase in employers, relative demand that has bid up the wages of “skilled workers”. This characterization of wage trends can be somewhat misleading. First, the economics rhetoric wrongly implies that those without a college degree are “unskilled”. Moreover, it is odd to characterize the group with falling wages as “less educated” or “unskilled”. After all, wages for those with a few years of college fell
6% and the noncollege educated workforce that experienced wage reductions comprises roughly 75% of the workforce. Second, particularly for men, these wage patterns more accurately reflect the driving down of the wages of the noncollege educated workforce (e.g., a 12.7% fall in high school wages from 1979 to 1989) rather than a bidding up of college wages (e.g., the college wage only rose 0.2%). Last, since the wages of both white-collar workers and college graduates have been falling since 1987 — through the last years of the 1980s recovery, into the last downturn and continuing through the first two years of the current recovery -there has been little “bidding up” of the real wages of “educated” or “skilled” male workers (Mishel and Bernstein, 1993b; Teixeira and Mishel, 1993). The most appropriate description of recent wage trends is that workers with less than a college degree have seen their real wages “driven down”.

The adverse wage trends affecting the noncollege educated workforce have not been offset by a growth of employment prospects. As the data in Table 5 show, the groups with the greatest fall in wages also tended to have the least employment growth, the greatest growth (or least fall) in their unemployment rate and no greater growth in their employment/population rate. There have been adverse labor market trends for the noncollege educated workforce in all the available measures, including both wages and employment.

The data in Table 6 address another important dimension of recent changes in the wage structure – the tremendous drop in wages for young workers. These data show that the wages earned by recent young high school graduates are far less than those earned by comparable workers in either 1979 or 1973. For instance, the wage earned by young male high school graduates in 1991 was roughly 26% less than that earned by similar workers in 1979. This drop in wages occurred among whites, blacks, and Hispanics. The fall in wages among young women high school graduates was less, but still a significant 16% for whites and blacks. The fact is that young high school graduates command a far lower wage now than in the past.

II. Explaining Earnings Trends

When earnings inequality increases, it is natural to ask which sources of inequality explain the increase. Some sources of extra inequality are matters of great social concern. if, for example, the gap between the earnings of black and white workers widened, most observers would be concerned that workplace discrimination had worsened. Some other sources of higher inequality should not be matters of much concern, however. Inequality might grow if the job market were suddenly flooded with new workers with little practical work experience. Inexperienced workers can expect to receive very low wages. As they obtain additional job experience, their earnings should rise and overall inequality decrease.

Inequality in wages (or in family income for that matter) occurs for a variety of reasons. Young workers typically earn lower wages than older ones. People in occupations that require lengthy or costly training earn more than people in occupations requiring few credentials and little experience. Full-time workers earn more than part-time workers. Among people in the same industries and occupations, women often earn less than men, and black and Hispanic workers frequently earn less than whites. In addition, Census statistics show that earnings differ greatly even among workers who share the same demographic characteristics and job qualifications and who are employed in the same industry and occupation. In fact, most of the variability of earnings in the United States arises from this last source. Even after suitable adjustments for age, work experience, educational attainment, occupation, and other characteristics, a great deal of unexplained and apparently random variation remains.

The statistics presented in the previous section derive from two main sources: anemic productivity growth and rising inequality in the distribution of wages. Productivity growth began to fall in the early 1970s, slowed to a crawl in the late 1970s, and then recovered slightly in the 1980s. The productivity slowdown had a dramatic and immediate effect in slowing the growth in average worker compensation. The trend in earnings inequality has followed a different pattern. Overall wage inequality among men increased gradually during the 1970s and began to surge during the early 1980s. Female wage inequality did not begin to rise until the 1980s. Apparently, earnings inequality among both men and women has continued to grow in the 1990s (Mishel and Bernstein, 1993b).

In the remainder of this section, we discuss four kinds of explanation for the recent rise in earnings inequality and the drop in real wages received by poorly paid workers. The first broad explanation stresses changes in the level or composition of the supply of U.S. workers. We then examine three different demand-side explanations: the changing geographical or spatial distribution of jobs across the United States; structural and institutional changes in the kinds of industries and firms offering employment; and changes in technology or methods of production that affect the demand for different classes of worker.

Our discussion draws on the burgeoning literature on recent wage structure developments (footnote reviews). Our goal here is not to present a comprehensive literature review since others have already done that (Levy and Murnane, 1992; Moss and Tilly, 1991; Mincy, 1993). Rather, we aim to present in broad strokes the themes of this research as it applies to explaining the growth of low wage work, particularly since 1979. We are necessarily hampered by the limitations of the existing literature. Much of it relates only to men’s earnings. Hardly any of it directly attempts to explain increased low wage work. Rather, the economics literature tends to focus on the growth in aggregate measures of wage inequality (which do not correspond directly to low wages) or on particular wage differentials such as between race, occupations and education groups.

A. Labor Supply

One set of possible explanations for growing wage inequality and shrinking earnings among poorly paid workers focuses on changes in the size and composition of the U.S. labor force (i.e., labor supply) since the 1960s. An obvious compositional change occurred as a result of the entry of the huge baby boom generation into the work force starting in the late 1960s and continuing through the early 1980s. In addition, there has been an unprecedented rise in paid employment among women. Other compositional shifts have occurred because of immigration, changes in educational attainment and quality among new labor market entrants, and gradual evolution in the racial makeup of the working- age population.

Changes in the supply of labor can produce two kinds of effects on the distribution of earnings. A rise in the number of less skilled or less experienced workers will raise the proportion of all workers who receive low wages. In addition, an increase in the percentage of workers with limited skill or experience can drive up the competition for jobs requiring limited skill, driving down the wages paid in those jobs.

Even though supply-side trends in the labor market can help explain some of the changing wage structure, scholars studying these trends believe they cannot account for the most important trends. Women at all educational levels have managed to reduce the gap between their wages and those of men, especially during the past decade, in spite of a continued and fairly steady increase in the proportion of women in the labor force. Moreover, wage disparities have increased within many groups in the labor force. Increased disparities were not restricted to those age, skill, and work- experience groups that had the l
argest gains in relative supply. Generational crowding cannot explain all the rise in inequality among men, for example, because the trend is apparent even in the oldest age groups, where no effect would be anticipated.

Demographic shifts since 1979 should in fact have contributed to growing equalization of earnings among both men and women. Most changes in the distribution by age, education, and annual work experience occurring in the 1980s raised the proportion of male and female workers who would be expected to be near the middle of the earnings distribution. Workers with age, education, or annual experience levels that would place them near the bottom became relatively less abundant.

It is possible, of course, that the decline in wages received by less educated younger workers stems from a deterioration in the quality of American secondary schools. The drop in the quality of U.S. schools is reflected in a long-term decline in SAT scores received by high school graduate and declines in achievement test scores received by 17- and 18-year olds. The decline ended in the early 1980s, however. Test scores from the National Assessment of Educational Progress confirm that there was no deterioration in test scores in 1980s.

There is some evidence to support the idea that deteriorating school quality may have affected the wage structure: among less educated men, the largest losses were sustained by younger men, precisely the ones who would be most affected if the quality Of schools had recently dropped, reducing the comparative quality of recent high school graduates. The school quality explanation can only go so far since less educated men have suffered earnings losses relative to better educated workers, regardless of their age. Also, another explanation for a more severe relative wage decline among young workers is that older high school-educated workers were better able to maintain their wages and good jobs because of their seniority rather than their better high school education. Overall, it seems unlikely that deteriorating school quality can explain a large part of recent wage trends. Further evidence against a school quality explanation is that even though the secondary schooling received by young men and young women is similar, the trend of wages in the two groups is very dissimilar. Likewise, test scores among black students have risen markedly relative to whites; yet, the wages of blacks have fallen relative to those of whites.

One last labor supply consideration is the recent large changes in the flow and composition of immigrant workers. The annual rate of immigration has nearly doubled in recent decades. The skills and educational attainment of recent immigrants are also less impressive than they once were. Immigrants to the United States were once drawn mainly from Europe or from well-educated groups in less developed countries. They are now increasingly drawn from less skilled populations in the third world. One-quarter of U.S. workers with less than a high school education are now immigrants. Perhaps one-fifth of the recent deterioration in the relative wages of high school dropouts may be due to increased immigration of less skilled workers.

B. Spatial Mismatches

Labor market problems for low wage, urban workers are frequently ascribed to a growing spatial mismatch between the types of jobs being created in the urban core and the lack of necessary skills needed among resident workers. In this view, as manufacturing jobs decline and/or move to the suburbs and as central city jobs expand primarily in the financial, business services and professional sectors there is an associated shrinkage in job opportunities for unskilled workers. This “spatial mismatch” is said to particularly disadvantage the urban black workforce who are becoming increasingly isolated from the current pool of unskilled jobs (Kasaid, 1989; Holzer, 1991).

It is difficult to isolate the effect of urban spatial mismatches from racial issues such as job and housing discrimination, i.e., to decide whether the issue is “race not space.” Racial issues are likely to be involved in business decisions on where to locate work and will definitely affect the ability of black workers to move to where jobs are available.

There seems to be little doubt that spatial distance from jobs has hurt the employment and earnings of central city black workers. This is reflected in the fact that jobs for unskilled workers tend to pay more in the suburbs. Nevertheless, it is unlikely that greater spatial mismatches led to the growth of low wage work in the 1980s. For one, the growth of low wage work has occurred among all racial groups and in the city, suburban, and rural areas. For instance, Table 6, discussed in Section 1, showed that since 1979 the entry level wages of young, high school graduates fell dramatically (about 25%) among both white and black men and fell significantly (about 15%) among white and black women. Moreover, annual earnings have fallen among the lowest two-fifths of men in both central city and non-central city areas. The explanations for the growth of low wage work seem to be forces affecting all noncollege educated workers rather than those in central city areas. It is also unclear whether low wage workers become increasingly spatially mismatched in the 1980s.

The evidence that job dislocation and the negative consequences of a shrinking manufacturing sector have affected blacks to a greater degree than whites suggests that spatial considerations play a part in the overall story of low wage work. It is likely that particular demand side factors such as deindustrialization have interacted with both spatial and racial factors to produce a greater degree of economic distress among blacks. At the same time, given the severe wage problems experienced by noncollege educated whites, the growth of spatial mismatches must be a considered relatively minor factor in the overall growth of low wage employment.

C. Trade, Structural and Institutional Factors

A variety of demand side factors have been significant forces in raising wage inequality and in the growth of low wage work. In this section we discuss the role of the shift of employment towards lower paying industries, increasing trade and the trade deficit, declining unionization and the low value of the minimum wage. other demand side factors such as technological change and the spatial distribution of jobs are considered below.

It is well known that there are now fewer manufacturing jobs than there were in 1979, a factor that in the popular press is sometimes said to be the cause of our wage problems. It is easy to either understate or overstate the role of manufacturing employment’s decline. on the one hand, one can point to the growth of low wage work and wage inequality within manufacturing and within nonmanufacturing, suggesting that the shift of employment can only explain part of rising inequality. It is also possible to point out that manufacturing’s share of employment has been shrinking for decades and that both high wage and low wage jobs are provided in the service sector. On the other hand, there was a larger shift of employment out of manufacturing in the 1980s than in prior decades. Plus, there was a general proportionate shift out of higher-paying industries (e.g., construction, mining, communications, transportation and manufacturing) to the lowest paying industries in the service sector (e.g., retail trade and personal and business services).

A rising share of employment in low-wage industries and a declining share of employment in high- wage industries causes average wages to fall and a growth of relatively low wage employment. Several studies suggest that industry shifts in the 1980s may have lowered compensation by as much as 3.5% overall and by 6.9% among blue-collar workers (Mishel and Bernstein, 1993). Industry shifts can also account for 20% to 30% of the higher education premiums that developed over the 1980s (Blackburn et al, 1990). Computations of the effect of industr
y shifts on the share of workers earning low wages show that industry shifts can explain about 20% of the growth of low wage work. Empirical research thus suggests that industry employment shifts have had a significant impact, but cannot be a sole explanation for the recent growth in wage inequality. It can be argued that industry shifts explain only a small (i.e., 20%) part of the growth of wage inequality. Equally true is that in most studies, industry shifts are the single largest identifiable factor leading to growing wage inequality.

The large growth in the trade deficit in the early 1980s occurred at the same time as the largest surge in wage inequality. The similar timing of these events has led economists to investigate the connection between trade and wages. A higher trade deficit can be expected to adversely affect the wages of noncollege educated workers if the domestic industries affected by imports disproportionately employ such workers. There has not been extensive study of this issues. The available research suggests that the groups experiencing the greatest wage declines are the ones that were also adversely affected by trade. Some of the effect of trade is an indirect effect due to displaced manufacturing workers competing for jobs in the service sector and thus lowering the wages of comparable nonmanufacturing workers. Also, the wage growth in manufacturing industries has probably been restrained by foreign competition. The most thorough studies in this area suggest that trade pressures can account for 10% to 20% of the growth in inequality (Borjas et al, 1991).

Collective bargaining has been a force for lower wage inequality since unions raise the wages of lower paid workers relative to better paid workers and union policies favor small wage differentials among comparable workers. It should not be surprising then that the decrease in unionization over the last two decades has been associated with growing wage inequality. Several studies suggest that about one- fourth of the growth in wage inequality has been due to decreased union representation. Because the minimum wage was not raised between 1981 and 1990 its inflation-adjusted value declined a great deal. Even after the increase to $4.25 in 1991, the minimum wages, value was a dollar less than it was in 1967 or 1979. This lower wage floor obviously affects the earnings of workers at or near the minimum wage. Since those working at or below the new minimum wage when the change was made in 1990 comprise about 10% of the workforce then at least that many workers can be said to have been affected by the decline. However, in 1991, there was about 17% of the workforce earning at or below the 1979 value of the minimum wage. If we consider that those earning up to a dollar more than the minimum wage would have lower wages if the minimum wage were lowered, then roughly 20% of the workforce has their wage tied to the minimum wage. Thus, the value of the minimum wage clearly plays a role in the earnings of low wage workers. The degree to which the fall in the minimum wage raised wage inequality has not been firmly established by researchers. Spriggs (1993) for instance, concludes that the erosion of the minimum wage explains a large part of the growth of low wage work since 1979, especially among entry level jobs for high school graduates and dropouts. Blackburn et al (1991), however, find a smaller role for the minimum wage. The difference between these two results might be because Spriggs accounts for a spillover effect of the minimum wage on other low wage earners while Blackburn et al do not.

A variety of other factors probably played a role in rising wage inequality, but have not been studied. The growth of parttime and contingent (e.g., temporary) work was probably a factor. The growth of employment among small businesses most likely played a significant role, as did the deregulation of several industries.

The total impact of these unstudied factors as well as those that have been researched is hard to establish because it is difficult to establish indirect effects. When workers are displaced (e.g., trade) or are forced to take lower wages (e.g., deregulation and deunionization) the labor market works to spread their impact to other comparable workers. These “spillover effects” are hard to quantify. It is clear, however, that the various structural and institutional factors we have discussed have had a substantial effect on wage inequality.

D. Technology, Productivity and Other Factors

A partial explanation for widening wage disparities may be a demand-side shift linked to the introduction of new kinds of production techniques. This kind of technological change has occurred in both the goods-producing and the service-producing industries as employers have changed their production methods in a way that requires a more able and educated work force. Employers have persisted in this strategy in the face of growing wage premiums for more skilled workers, a development that makes it more costly to hire a skilled work force than it was in the 1960s or 1970s.

Direct evidence about the effects of technological change on wage inequality is very difficult to obtain. The government does not survey employers to determine whether and how much their skill requirements have changed over time. Analysts must deduce the effects of technology shifts based on trends in the relative price of different kinds of labor and the relative supply of those same kinds of labor. It is important to understand what kind of evidence economists use in inferring the effects of technological change. Much of the evidence is negative. The shift in the industrial composition of demand can explain only part of the change in the structure of wages. As we have seen, supply-side changes also explain only a small percentage of the growth in inequality. This leaves technological change as one possible explanation.

Several strands of evidence suggest that recent technological change has favored the more skilled over the less skilled. Bureau of Labor Statistics projections show that the fast-growing occupations within most industries are those requiring high levels of education and skill (Teixeira and Mishel, 1993). Economists have also found evidence of a positive association between simple measures of technological progress and the rate of change within industries in the pay premium received by skilled workers. Faster technological change, as indicated by R&D spending, is associated with increasing pay differentials between highly educated and less educated workers (Bartel and Lichtenberg, 1987).

Alan Krueger (1993) has recently established that workers who use computers in their jobs enjoyed higher wages in the 1980s than workers who did not use or know how to use them. Although the number of computer-literate workers is comparatively small, these workers received a disproportionate share of all wage gains in the middle and late 1980s. Use of computers is particularly common among well-educated workers in skilled occupations; it is increasingly common among all types of workers. Krueger’s research suggests that the introduction of microcomputers has conferred a significant advantage on workers who possess the skills and background necessary to learn how to use them. Unfortunately, unskilled workers do not fit this description.

Although this kind of evidence indicates that technological progress has favored the more skilled, especially since 1979, it is unclear that it can account for all the wage inequality not explained by shifts in the industrial structure and the relative supply of different kinds of labor. More than half the increase in wage and earnings inequality derives from widening wage disparities among workers who have the same amount of education and previous work experience. Inequality has grown among workers within the same industry and among workers in the same occupation. Moreover, there is no concrete evidence that workplace technological trends accelerated in the 1980s, paralleling the surge in wage inequality.

p>Several researchers have suggested that the sharp increase in inequality within groups of workers defined by age, education, occupation, industry, and work experience signals larger wage premiums for unmeasured job skills. Much of the rise in inequality we have been discussing so far has been associated with the growing inequality between groups of workers who differ in education, occupation, and work experience. If we could observe the effects of other dimensions of skill or ability, such as specific job training or general aptitude, we might be able to obtain a clearer understanding of the ultimate sources of increased wage inequality. If these dimensions could be measured, most economists believe we would find that inequality across skill categories has risen in recent years. The increases in skill premiums reflect the fact that employers, demand for highly skilled workers has outpaced supply of these workers. To explain growing wage inequality since the early 1970s, economists point to steady increases in demand for highly skilled workers combined with fluctuations in the growing supply of more educated and more skilled workers. So far, however, the evidence is suggestive rather than conclusive.

III. The Future

If future trends in the earnings distribution were easy to forecast, economists and policy makers would not have been surprised by widening wage inequality in the 1980s. In this section we discuss some of the forces at work and the likely qualitative outcome: Wage growth in the 1990s will be sluggish, at best, and wage inequality will probably continue to grow. If our intuition is correct, then there will be a further growth in the low wage workforce in the 1990s and no reversal of past trends.

Economists and demographers can confidently predict only one of the determinants of relative wage changes over the next decade — movements in the size and demographic composition of labor supply. The number of 18-year-olds entering the job market, which has been falling for more than a decade, will once again begin to rise in 1997. For the foreseeable future, however, the number of new entrants will remain below the average levels reached in the 1970s. At the same time, the number of workers reaching retirement age (65) will remain comparatively low through the remainder of this century. Thus, the aging of the active work force means that an increasing percentage of workers will be in their peak earnings years, ages 40-55. The effect of a growing supply of experienced and highly educated workers is to reverse or slow the increase of the wage premium that highly experienced and educated workers currently command. However, this demographic effect may be small–the effect of the workforce substantially aging between 1979 and 1993 was swamped by other forces at work.

The large pay premiums received by college graduates during the 1980s has stimulated a rise in college attendance, both among new high school graduates and among adults in their late 20s and early 30s. Even though the number of Americans aged 18-22 fell throughout the 1980s, the number of Americans attending college climbed by 1.6 million — 13 percent — during the same decade. Assuming that immigration of unskilled workers is limited, the increase in college attendance should boost the percentage of workers with college degrees and reduce the supply of non-college educated workers. These trends should reduce or restrain the pay premium that workers receive from an advanced degree and be a force for lesser wage inequality.

The increase in the supply of college graduates may not have a large effect, however. The size of the entering cohort of workers is small relative to the total size of the workforce so there must be a very large increase in college enrollment to make a difference. Higher tuition, state budget rises and income stagnation may impede the growth in college enrollment. Plus, the workforce near retirement age has a fairly high level of educational attainment, so the natural evolution of young workers replacing retiring workers will not raise average education levels as much as was the case twenty years ago.

It is much more difficult to forecast developments in labor demand and technology than it is to predict trends in the composition of labor supply. One development that contributed to the loss of well-paying jobs for the less skilled during the 1980s was the loss of manufacturing employment, especially in industries where international trade is significant. Some trade economists believe that the U.S. merchandise trade position will resume its trend toward balance after the economies of Western Europe and Japan begin to recover. If this happens, the overall demand for U.S. manufactured goods will grow faster than demand in the United States would allow on its own. A declining trade imbalance could thus offset some of the employment decline due to manufacturing productivity growth. On the other hand, even with balanced trade, increased globalization may lead to greater wage inequality, especially as trade with developing countries increases. Also, it may be optimistic to believe that our several decades, decline in international competitiveness will reverse itself.

It should be emphasized, however, that foreign demand for manufactured goods is not the main determinant of demand. If the U.S. economy should continue to grow slowly, the demand for workers in well paying occupations could easily fall short of the moderate growth in supply of American workers. The shrinking size of defense industries, for example, will almost certainly reduce the availability of well-paid industrial jobs. The distribution of demand between well-paying, highly- skilled occupations and poorly paid or less skilled occupations is also very difficult to forecast. This will depend on the structure of labor demand across different kinds of industries and trends in technology within different kinds of industries. Economists have little reason to be confident they can accurately forecast these trends.

However, it is likely that many of the technological and trade factors that depressed the wages of the noncollege educated workforce will persist. It also is doubtful whether the value of the minimum wage will rise appreciably or that trade unions will be able to expand their membership and bargaining power. Last, the small business sector and various types of labor contingency (part-time, contract, or temporary work) will probably continue to expand, putting further downward pressure on wages. Thus, one can expect a variety of demand side and structural factors to lead to continued growth in wage inequality and low wage work despite some offsetting pressure from a greater relative availability of more skilled and educated workers.

Perhaps the recent past is the best indicator of the future. Wage data over the last business cycle since 1989, including the first two years of the recovery, suggest that wage inequality has continued to grow, especially for men (Mishel and Bernstein, 1993b). It is likely, but not inevitable, that overall wage inequality will grow somewhat more slowly in the 1990s than in the 1980s. What appears most unlikely is that labor markets in the 1990s will reverse what occurred in the 1980s and restore the earnings power of the noncollege educated workforce to where it was in 1979.

IV. Policy Implications

There are a variety of issues where policy formulation should take into account the recent and expected expansion of low wage work and the employment difficulties faced by noncollege educated workers, especially those with no more than a high school degree.

A. Welfare Reform

The current discussion of welfare reform combines two dimensions. One aspect is to provide the training and social support (e.g., child care, health care) that would enable current welfare recipients to transition to paid employment. The other dimension is to limit the use of welfare to a fixed period of time, thus forcing long term recipients onto the labor market. A major issue is
whether the private sector labor market can absorb this new workforce at wages that would allow current welfare recipients to at least maintain current consumption levels. A first problem is that the extra competition from several million more workers in the low wage labor market might cause wages to decline among currently employed high school dropouts and graduates. These new workers are also likely to have a difficult time finding steady work given the overall sluggish pace of job growth and the lessened opportunities for less skilled workers. Moreover, the wages paid to workers with skills comparable to that of most long term welfare recipients are already low and are likely to fall further over the decade. It seems apparent that additional skills and training coupled with better job prospects are required to ensure that those forced off welfare will be able to sustain their current low standard of living. If the private sector can not provide the necessary jobs at the needed wages then public sector alternatives might have to be considered.

B. Minimum Wage

The value of the minimum wage in 1993 was more than $1.30 less, in inflation-adjusted terms, than it was in 1979. Restoring the minimum wage to its 1979 level could substantially lower the proportion of the workforce earning low wages. A higher minimum would affect those earning at the minimum now but would also affect those earning wages above but near the new minimum (say within a dollar). The downside would be somewhat lower employment. However, the evidence is ambiguous as to whether and how much employment contracts when a new minimum wage is legislated (Spriggs, 1993). A possible upside would be a reduction in the Federal expenditures for the Earned Income Tax Credit as low wage workers earn more and require less subsidy. The resulting savings could be redirected to training and other services for those making the transition from welfare to paid employment.

C. Demand Side and Institutional Policies

The biggest challenge is to create jobs with good wages for those without a college degree. Policies which rebuild our manufacturing competitiveness, reorient our production methods towards less hierarchical methods, and expand training opportunities for employed workers, will move us in the right direction. Policies which expand union representation will lead to lesser wage inequality and enhanced job opportunities for noncollege educated workers. The requirement that all employers must provide health insurance to their workforce can lead to a better mix of jobs by removing employer financial incentives to create part-time and temporary jobs.

Return to Idea Central’s Welfare and Families Page Bibliography

Bartel, Ann P. and Frank R. Lichtenberg. “The Comparative Advantage of Educated Workers in Implementing New Technology.” Review of Economics and Statistics, Vol. 69, February 1987, pp. 1- 11.

Bartik, Timothy. Economic Development and Black Economic Success. Upjohn Institute Technical Report No. 93-001. Kalamazoo, MI: W. E. Upjohn Institute for Employment Research, 1993.

Blackburn, McKinley L. David E. Bloom, and Richard B. Freeman. “The Declining Economic Position of Less Skilled American Men.” In Gary Burtless, ed., A Future of Lousy Jobs? Washington, DC: Brookings, 1990.

Bluestone, Barry, Mary Huff Stevenson, and Chris Tilly. An Assessment of the Impact of “Deindustrialization” and Spatial Mismatch on the Labor Market Outcomes of Young White, Black, and Latino Men and Women Who Have Limited Schooling. Boston, MA: The John W. McCormack Institute of Public Affairs, University of Massachusetts at Boston, 1992.

Borjas, George L., Richard B. Freeman, and Lawrence F. Katz. “On the Labor Market Effects of Immigration and Trade.” National Bureau of Economic Research Working Paper No. 3761. Cambridge, MA: NBER, 1991.

Burtless, Gary. “Economic Inequality over the Business and Demographic Cycles,” in Gary Burtless, ed., A Future of Lousy Jobs? The Changing Structure of U.S. Wages. Washington, DC: Brookings Institution, 1990.

Burtless, Gary. “The Contribution of Employment and Hours Changes to Family Income Inequality.” American Economic Review, Vol. 83, May 1993, p. 134.

Holzer, Harry J. “The Spatial Mismatch Hypothesis: What Has the Evidence Shown?” Urban Studies, 28;4 (February 1991):104-22.

Ihlanfeldt, Keith R. “Housing Segregation and the Wages and Commutes of Urban Blacks: The Case of Atlanta Fast-Food Restaurant Workers.” Policy Research Center Paper No. 30. Atlanta, GA: Policy Research Center, Georgia State University, 1992.

Ihlanfeldt, Keith R. and David L. Sjoquist. “The Impact of Job Decentralization on the Economic Welfare of Central City Blacks.” Journal of Urban Economic, Vol. 26, 1989, pp. 110-130.

Ihlanfeldt, Keith R. and Madelyn Young. “The Spatial Distribution of Black Employment Between the Central City and the Suburbs.” Policy Research Center Paper No. 36. Atlanta, GA: Policy Research Center, Georgia State University, 1993.

Kasarda, John D. “Urban Industrial Transition and the Underclass.” The Annals of the American Academy of Political and Social Science, Vol. 501, January 1989, pp. 24-47.

Katz, Lawrence F. and Kevin M. Murphy. “Changes in Relative Wages, 1963-87: Supply and Demand Factors.” The Quarterly Journal of Economics, Vol. 107, February 1992, pp. 35- 78.

Krueger, Alan B. “How Computers Have Changed the Wage Structure: Evidence from Microdata, 1984-89.” The Quarterly Journal of Economics, Vol. 108, February 1993, pp. 33- 60.

Levy, Frank, and Richard Murnane. “U.S. Earnings Levels and Earnings Inequality: A Review of Recent Trends and Proposed Explanations.” Journal of Economic Literature, No. 30, Vol. 3, September 1992, pp. 1333-1381.

Mincy, Ronald B. The Underclass: Concept, Controversy and Evidence. Washington, DC: The Urban Institute, 1993.

Mishel, Lawrence and Jared Bernstein. The State of Working America: 1992-93 Edition. Armonk, NY: M.E. Sharpe Inc., 1993a.

Mishel, Lawrence and Jared Bernstein. The Joyless Recovery: Deteriorating Wages and Job Quality in the 1990s. Washington, DC: Economic Policy Institute, 1993b.

Moss, Phillip and Chris Tilly. “Why Black Men are Doing Worse in the Labor Market: A Review of Supply- Side and Demand-Side Explanations.” mimeo, Social Science Research Council, August, 1991.

Spriggs, William E. and Bruce W. Klein. “The Minimum Wage: A New Assessment of its Effects”. Washington, DC: Economic Policy Institute, 1993.

Teixeira, Ruy A. and Lawrence Mishel. “Whose Skills Shortage-Workers or Management?” Issues in Science and Technology, Summer 1993, pp. 69-74.

Return to Idea Central’s Welfare and Families Page The Electronic Policy Network / Send us a message at © 1995 New Prospect, Inc.

See related work on Income and wages | Wages, Incomes, and Wealth

See more work by Lawrence Mishel