Read table of contents, executive summary and introduction (2006/2007 edition)
[return to main excerpts page]

Download PDF version of Table of Contents, Executive Summary and Introduction

jump to:
[table of contents] [executive summary] [introduction] [documentation and methodology]

Introduction

Starting in 1995, a new and important change occurred in the U.S. economy: productivity—the output of goods and services per hour worked—began to grow more quickly. After growing 1.4% per year since the mid-1970s, productivity accelerated to 2.5% a year from 1995 to 2000 and then jumped to 3.3% a year from 2000 to 2005. The post-1995 shift in productivity growth, partly attributed to the diffusion and more efficient use of information technology, has sometimes been labeled the “new economy.” Because productivity growth provides the basis for rising living standards for everyone, its acceleration is an unequivocally positive development for the economy.

In addition, the productivity acceleration has produced ancillary benefits. One key example of such a benefit is the fact that Federal Reserve officials view faster productivity growth as forestalling inflationary pressures, thereby enabling them to pursue more accommodating monetary policy than would otherwise be the case. Research suggests that these dynamics themselves will lead to longer and less volatile business cycles, an obviously positive development.

Yet, despite these clear economic advances, many Americans report dissatisfaction with where the economy seems to be headed, and many worry about their own and their children’s well-being. These concerns have led some policy makers and economists to ask: why aren’t people getting it? Why are people pessimistic when the economy is growing strongly? The question seems reasonable to those who follow the top-line numbers of the economy such as the growth of the total economy (e.g., gross domestic product), the stock market, or corporate profits. The question is a strange and dissonant one, however, for those who follow trends in living standards and pay attention to the distribution of growth, since the answer is so apparent.

The economic recovery that began in November 2001 set a record not for growth but for hosting the longest “jobless recovery” on record (the government began tracking employment in the 1940s). By 2006, the employment rate (the share of the population at work) was still below its 2000 peak. The weak job market and the lack of policies and institutions that help ensure that growth is broadly shared meant that the living standards of most working families have been stagnant in recent years. In fact, by 2004 (the latest data), the real income of the median or typical family was lower than in 2000, and inflation-adjusted wages, whether for high school or college-educated workers, grew hardly at all since 2000. The number of poor persons grew by 5.4 million between 2000 and 2004, while six million were added to the ranks of the uninsured.

If the nation is indeed wealthier in 2006 than at the peak of the last business cycle in 2000, but many families’ incomes are lower and the share in poverty has grown, where is all the money going? This answer is fairly obvious as well: wages, income, and wealth are being drawn to the very top earners and families. This redistribution is a continuation of a historic trend that began in the late 1970s, paused for a few years when the financial bubble burst in 2000, and has most recently returned.

In other words, the economist’s mantra that faster productivity growth leads to higher living standards needs updating. Such growth creates the potential for widely shared prosperity. For that potential to be realized, a number of other factors—labor market institutions such as strong collective bargaining and a minimum wage with some bite, and, importantly, a truly tight labor market—have to be in place to ensure that the benefits of growth reach everyone, not just those at the top of the wealth scale.

A second key ingredient of the new economy, globalization, also has the potential to lift living standards by lowering prices and providing much greater supplies of the products and inputs that help keep the economy humming. But its fingerprints are all over the diminished bargaining clout of blue-and white-collar workers who now compete directly with workers from abroad, many of whom are highly skilled but from low-wage countries.

As these disparate economic forces have interacted, we have seen some of the best and worst of what the new economy has to offer. With the late-1990s boom, the U.S. achieved the first full employment economy in three decades; rising real wages throughout the wage scale; steeply falling poverty rates, especially for the least advantaged; and the narrowing of gaps between most income classes and racial groups (for example, the gap between African American and white median family incomes fell to its lowest level on record).

The 2000s have not been nearly so beneficial. In the absence of the institutional mechanisms that convert overall growth to broadly shared prosperity, and once the recession and jobless recovery did away with the full employment conditions of the latter 1990s, the benefits of productivity growth stopped flowing to the majority of working families.

In this introduction, we highlight current trends as well as some longer-term trends in order to address this fundamental question: to what extent is the economy’s growth reaching those largely responsible for its creation, America’s working families? The chapters that follow elaborate this story in greater detail by examining trends in incomes, mobility, wages, jobs, wealth, and poverty, and by placing recent developments in their historical, regional, and international context.

The discussion that follows breaks our usual rule of making historical comparisons at similar points in a business cycle because it is important to examine the unique period that started in the middle of the last business cycle. The expansion that began in 1991 and peaked in 2000 includes two distinct periods, one from 1991 to 1995 that had the same sluggish productivity growth that had prevailed since 1973, and one from 1995 to 2000 that was characterized by faster productivity growth. After the recession of 2001 productivity accelerated even further but, unlike in the 1995-2000 period, the high productivity was accompanied by a jobless recovery and then modest employment growth. An interesting question is, how have families fared during the productivity-rich period between 1995 to 2005?

From the perspective of working families, the comparison yields stark and disappointing results. The wage and income growth of middle- and low-income families reversed course sharply, severing the latter-1990s link between productivity growth and living standards. Poverty rose, as did the number of persons without health coverage. Growth in inequality, which slowed markedly following the bust in the capital markets in late 2000, began climbing again by the mid-2000s, as the largest income gains began to accrue to those at the very top of the wealth scale.

Table 1 shows the change in a comprehensive set of indicators of living standards between 1995 and 2005 (for some the latest data are 2004). The reversal of fortune between 1995-2000 and 2000-05 is striking.

Productivity grew 13.2% in the second half of the 1990s, and the typical family’s in-come rose at a similar rate of 11.3%. Even more impressive is the fact that the real median income of minority and single-mother families outpaced productivity, a powerful example of who reaps the greatest benefits from a full employment economy.

Productivity rose 4.5% faster in the 2000s, yet income growth has been uniformly negative, with especially large reversals for less-advantaged families. Real median income for African Americans fell by 4.8% in 2000-04, for a swing, or deceleration, of -20.4% between the two periods (meaning their incomes grew about 20% less quickly in the 2000s compared to the latter 1990s). For Hispanic families, the reversal was even larger, from 24.9% to -6.3%. Young families saw a large switch as well.

The number of poor people fell 13.3% from 1995 to 2000, for a decline of 4.8 million people from the poverty rolls, and the number of uninsured fell slightly (-1.9%). These trends reversed in the 2000-04 period, as the number of poor grew by 17.1%, or 5.4 million people, and the number without health coverage grew by 15.1%, adding six million to the ranks of the uninsured.

Net worth per household grew by almost 27% from 1995 to 2000, led by housing wealth and financial asset appreciation. Though homeownership kept growing in the 2000s, financial losses were large, and net worth grew 24.7% more slowly in the 2000s than in the latter 1990s.

The main factors behind the slide in family incomes are slower hourly wage growth and fewer available hours of work in the paid job market. As jobs grew 11.1% more slowly in the 2000s, growth in real median hourly wages decelerated as well for both high school and college graduates. The latter experienced a larger reversal: the real hourly wage for persons with a bachelor’s degree rose 11.3% in the 1995-2000 period but only 1.3% over the next five years. The 10% deceleration in wage growth is indicative of a job market that was particularly tough for college-educated workers, due in part to the bursting of bubbles in industries like information technology and financial services that use a lot of these work-ers. The 2000s have also been a period of greater offshoring of white-collar jobs, and this too puts downward pressure on college wages.

Annual hours worked by both individuals and middle-income families (this latter mea-sure sums hours worked across all family members) rose in the late-1990s but fell after 2000. The share of the workforce with employer-provided health and pension coverage also contracted in the 2000s. In addition, as shown in Chapter 3, employers have been shifting more of the cost of these fringe benefits onto workers.

Given the different points in the business cycle that we are comparing—the 1995-2000 period had five years of expansion behind it, whereas the 2000-05 period began with a recession—we would expect wages and incomes to perform less well over the early 2000s relative to the latter 1990s. Yet the reversals shown in the last column of the table are economically large and disconcerting, especially given the significant acceleration in productivity growth. Furthermore, yearly trends of some of the key living standards indictors shown in the table reveal movements that have actually worsened as the 2000s business cycle has progressed.

Figure A plots productivity and real hourly wages for the median, high-school-educated, and college-educated worker from 1995 through the first half of 2006, indexed to 100 in 1995. During the latter 1990s, as the job market moved toward full employment, these values all grew together and helped link real wages to overall growth.

After 2000, even as the recession and jobless recovery occurred, the momentum of the latter 1990s boom fueled continued real wage growth through 2003. Since then, real wages have been flat for college-educated workers and falling for median and high-school-educated workers, leading to a historically large gap between wage and productivity growth. In other words, real wages stopped trending along with the business cycle, which was clearly improving when real wages reversed course. By 2006, the fifth year of a uniquely productivity-rich expansion, with solid GDP growth, one would expect to see a different trend in real wages.

A tight labor market is an essential element in ensuring that the benefits of growth flow broadly throughout the income scale. Unfortunately, the recession of 2001 and the jobless recovery erased the full employment conditions responsible for the healthy wage growth of the late 1990s. Even by mid-2006 the job market was still failing to generate the number and quality of jobs needed to ameliorate the negative trends that have prevailed thus far in the expansion.

Figure B plots the loss and then growth of jobs over this business cycle compared to the last cycle (the early 1990s) and to the average of the three other previous cycles that lasted at least as long as this one (63 months as of this writing; note that we also performed this analysis including all cycles, and the results were the same). The lines plot the percent change in job growth since the previous business cycle peak.

The unique weakness of the current business cycle in terms of job growth is clear in the figure, even compared to the early 1990s recovery, which was also labeled jobless.By month 63 in the current cycle, employment was up by just 2%, compared to 7% in the 1990s and 12% in the average of the three prior cycles. If this cycle were comparable to the last one at this point, the U.S. economy would have seven million more jobs, or about 100,000 more per month.

To put this monthly number in context, consider that, since net job growth turned positive in mid-2003, we have added about 170,000 jobs per month, on average (leaving out the impact of the Gulf Coast hurricanes, which lowers the average). Over the comparable period in the last recovery, the one that ultimately led to the full employment job market and the commensurate gains in wages, the economy was adding about 250,000 jobs per month. In other words, another 100,000 jobs per month would have put the economy back on a track similar to the last recovery, helping to rid the post-2000 period of the job market slack that has precluded growth from reaching many working families.

Figure C shows a related dimension of the jobs problem. While it took 21 months for payrolls to regain their peak in the average post-war business cycle, in this case it took more than twice that long: 46 months for total payrolls and even longer for the private sector. These long lags delay the arrival of any job-market-related bargaining power workers need in order to claim a larger share of productivity growth.

A look at current unemployment rates—they have fallen as low as 4.6% in recent months—might suggest that the economy is now at or near full employment. But the unemployment rate was 4.0% in 2000, and there is a significant difference between these values for those groups with the least bargaining power. A sustained unemployment rate of 4.0% is a necessary ingredient for generating broad-based wage and income gains, and the current weak trend in median wages certainly belies any claim that the economy is at full employment. Furthermore, the unemployment rate is artificially reduced if job seekers give up their search. Only those actively seeking work are counted as unemployed, and if, due to a perceived lack of opportunity, enough people give up the job search—they put off entering or they leave the job market—they will not be counted as jobless. This dynamic can serve to lower reported unemployment. In such periods, it is essential to examine labor force participation rates and employment rates to see if they reveal more slack than does the unemployment rate.

In fact, even by mid-2006, the share of the population at work—the employment rate—was 1.6 percentage points below its 2000 peak. This drop implies that literally mil-lions of potential workers have not been looking for work, and their exit from the labor force lowers the jobless rate.

Some analysts object to this comparison of today’s employment rates with those of the last cycle’s peak, claiming that the 2000 job market was inflated by a market bubble. But the gap between today’s employment rate and that of other, pre-bubble years is still large. For example, employment rates in 1997, 1998, and 1999, a period of strong but not obviously speculative demand, were still about one point higher than today’s.

One of the key lessons of our analysis of the full employment economy is that we need to aim high when setting employment benchmarks. That is, since it takes a truly tight labor market (“chock-full employment,” as late Nobel Laureate economist William Vickery once put it) to ensure widespread gains, we diminish the chance of reaching full employment if we set benchmarks to periods when income and wage growth were stagnant. The speculative demand of the late 1990s may not have been good for the economy overall, but that does not mean one should ignore the success story in the job market. Besides, there are other, non-speculative sources of strong demand, and good economic policy can push back against demand-killers, like the large and growing trade deficit.

A strain of the bubble argument says that demographic change (an older work-force), not a shortage of jobs, is responsible for the decline in both labor force participation and employment rates since 2000. This is a tough case to make, given that job creation was weak over the period when these rates fell, so a decline in labor force participation due to insufficient demand for workers seems to offer a more straightforward interpretation. Moreover, demographic changes usually don’t have much impact over the short run.

In fact, as Figure D shows, holding employment rates constant for each demographic group at their 1998 level and simply allowing their population shares (by age and gender) to change leads to a simulated rate that falls much less than the actual trend; this suggests a very minor role for demographic change (to stave off the 2000 bubble critique, this simulation starts earlier, in 1998). While the actual employment rate falls 1.4 percentage points over this period, the simulated rate driven solely by the aging of the workforce falls only 0.3 points, indicating that forces other than demography lead to a 1.1 percentage-point drop in employment.

Given the size of the U.S. working-age population in 2005 (226 million), the lower employment rate implies that over three million workers were missing from the labor force in that year, and of these, about 675,000 might be explained by the aging of the workforce since 1998. That leaves about 2.4 million persons missing from the job market. It is likely that many of these workers, were they to enter the job market, would be added to the seven million unemployed in mid-2006, leading to a significantly higher unemployment rate.

The uptick in the actual employment rate trend at the end of the figure coincides with the period when job growth turned positive, and this is another indication that cyclical forces are at work. We pursue this analysis further in Chapter 4 by tracking the employment rates of groups like young college graduates and minorities, who might be expected to be particularly responsive to the availability of jobs. Like this figure, these analyses reveal a clear cyclical response to the deterioration and then improvement in the job market, all of which suggests that the low unemployment rate is, at least for now, an inadequate measure of the true degree of labor market tightness.

Middle-income families derive about three-quarters of their income from the labor market. Thus, weakening labor market trends can explain the income reversals experienced since 2000. Table 2 summarizes some of the results from Chapter 1, where we examine changes in family income and its components for families by income fifth. Here we focus on middle-income families, whose real income fell 2.1% from 2000 to 2004 after growing almost 12% in the five years prior.

The data clearly suggest that the income loss was a function of labor market contraction. Annual hours worked by all family members fell 4.3%, after rising 5.2% during the 1990s boom. Hourly wages continued to rise (though yearly trends show a real decline in 2004), but by 8.0% less in the 2000s compared to the latter 1990s. Losses in hours worked explain almost 60% of the reversal in income growth, and much slower hourly wage growth explained the rest.

The gap between productivity, wages, and incomes suggests that inequality is on the rise. The growth embodied in these macro-indicators has to be going somewhere, and if only trace amounts are finding their way into the paychecks of many working families, it stands to reason that the gains are flowing up the wealth scale.

We provide two pieces of supportive evidence. First, Table 3 examines recent research on income movements among the richest households, including the value of realized capital gains. Between 1995 and 2000, real income grew 12.5% for the bottom 90%, 15.0% for the next highest 5% (the 90-95th percentile), 25.1% for the next highest 4%, and so on, up to the more than doubling of income—up 156.0%—at the very top sliver of the top 1% (the 99.99-100th percentile). So, even as the job market progressed toward full employment, income growth was still skewed toward the top over this period.

The bursting of the financial and investment bubbles led to sharp real declines in the incomes of wealthy households in 2001 and 2002, with the biggest losses among the rich-est. By 2003, however, the earlier pattern was returning, and in 2004, a year when the economy expanded by 4.2% and productivity grew by 3.4%, the income of the bottom 90% got a mere 1.4% boost. The rest accrued to the top 10%, and the gains were largest at the top of the income scale. The step pattern repeats, with larger percentage gains as we move up to the very top of the income scale. The average income gain at the very top was 27.5% in 2004 alone.

The growth of capital incomes (from corporate profits or, as received directly by individuals, from dividends, interest, and capital gains) has also driven up inequality. First, most capital income is received by the best-off families, so the growth in capital incomes reinforces income inequalities (the decline in capital incomes in 2000-02 actually reduced inequality). For 2006, for instance, it is estimated that 84.2% of all capital income will have been received by the upper fifth, with 55.3% received by the top 1% and 36.6% by the top 0.1% alone. In contrast, 3.6% of such income is held by middle-income families. This income landscape implies that fast growth for capital income will disproportionately benefit the best-off income groups.

The rise in inequality caused by the shift toward greater capital income has been compounded by the growing concentration of capital income among the very highest income groups, particularly the top 1%. According to the Congressional Budget Office, the share of capital income received by the top 1% grew by 14.3 percentage points from 1995 to 2003 (the latest data), with proportionately less capital income received by the remaining households in the top fifth (down 7.5 percentage points) and a lesser share received by the bottom 80%.

The shift to greater capital income in the economy can be illustrated by the recent movements in the division of corporate sector income (which excludes proprietorships and other unincorporated businesses but includes more than 75% of the entire private sector) and in the growth of profitability (returns on capital investments). As Figure E shows, capital’s share of income tumbled in the late 1990s in the context of a tight labor market, but it more than recovered by 2005. This helped boost profitability in 2005 to the highest level reached in 36 years (excepting 1997).

The growth in profitability has left less room for (or, rather, was caused by the muting of) wage growth, and might be considered the consequence of businesses successfully restraining wage growth as sales and profits grew, even in years of seemingly low unemployment. If the pre-tax return to capital in 2005 (11.9%) had been at the 1979 level (9.6%), then hourly compensation would have been 5.0% higher in the corporate sector. This difference is equivalent to an annual transfer of $235 billion from labor to capital (measured in 2005 dollars), or roughly $1,760 per wage-and-salary employee or $3,032 if this redistribution came from the bottom 80% of workers, who earn just 58% of all hourly earnings.

Even this analysis misses an important part of the picture, since the definition of labor’s share includes the pay of chief executive officers and thereby overstates the income share going to typical workers and understates profits, since the bonuses and stock options given CEOs are more akin to profits than wages. The amount of CEO pay relative to corporate profits has grown: in the mid-1990s CEO pay was equivalent to roughly 5% of corporate profits, but that amount rose to roughly 10% of profits by 2003 (unfortunately, more recent data are not available).

The challenges to living standards that working people and their families have faced over the last few years have been even more pronounced for young workers and their families. When times are good, they are particularly good for young workers, but when times are not so good, they tend to be awful for young workers. Since young workers and their families are at the beginning of their economic life cycle, these differences are important. Those starting off in a down period often suffer losses that will reverberate throughout their careers.

Not surprisingly, the recession, jobless recovery, and then weak job growth of the last five years have sharply reversed the progress that young workers and their families were making in better times. Table 4 presents indicators of labor market trends (wages, benefit coverage, and employment) and family income and living arrangements that illustrate the hit on young people.

One way to illustrate the difficulties faced by young workers in the labor market is to examine the wages and benefit coverage of entry-level workers, both those with a high school degree (and no further education) and those with a four-year college degree. The table presents the trends of the late 1990s and compares them to the latest five-year period in order to illustrate the change in fortunes that has occurred. The wages earned by entry-level high-school-educated workers (those out of school for one to five years) rose over 9% from 1995 to 2000, then deteriorated over the 2000-05 period. The difference between these two trends was a 10.9% turnaround for men and a 14.1% turnaround for women. Entry-level high school graduates had low rates of benefit coverage from their employers, with only 38.2% receiving health coverage and 20.6% receiving pension coverage in 1995; these rates are substantially down from their levels in 1979 (the health coverage rate was 63.3% that year). Given the declining coverage before 1995, the good news about entry-level high-school-graduate jobs is that employer-provided health and pension benefits fell only a small amount over the late 1990s. However, a sharp erosion of benefits returned from 2000 to 2005, with health coverage falling so that only a third now receives it, roughly half of the coverage in 1979. This leaves young workers joining the growing ranks of the uninsured, unless they are lucky enough to have coverage through a spouse’s employer.

The change in fortune was even starker for young college graduates. This group benefited during the late 1990s a great deal, as those graduating then (reflected in the entry-level wages and benefits in 2000) enjoyed far greater pay than those who had graduated in the early 1990s (and measured for 1995). Entry-level wages for male college graduates ballooned 20.9%, and those for women rose 11.7%. Benefit coverage, particularly pensions, improved in the late 1990s for these new workers, but the poor labor market conditions of the 2000-05 period hit college graduates hard, as entry-level wages fell and benefit coverage dropped significantly (for example, health coverage fell from 70.6% to 63.5%). Had the late 1990s trends continued over the 2000-05 period, the entry-level wages of college graduates would have been 24.8% higher for men and 13.6% higher for women, and a much larger share would have been covered by employer-provided health and pensions (a 14.9% coverage rise for pensions and an 8.4% rise for health).

Another reflection of the dramatic turnaround in labor market trends is that employment rates for young (in this case 25-34-year-old) high school graduates grew in the late 1990s (up 2.0%) but then fell 4.2 percentage points from 2000 to 2005, a 6.2 percent-age-point turnaround. Even college graduates saw their fortunes turn around as the share employed fell from 2000 to 2005.

Given these faltering labor market trends, it should come as no surprise that families headed by someone age 25-34 had 5.8% lower incomes in 2005 than in 2000. In contrast, this type of family enjoyed incomes 12.3% higher in 2000 than in 1995, amounting to a remarkable 18.1% reversal. Another reflection of these trends is that more young men age 25-34 are living with their parents. The share declined in the late 1990s, from 15.4% in 1995 to 12.9% in 2000, but rose to 13.7% in 2005. Young women’s living arrangements have been relatively stable in this period.

Thus far, we have examined the productivity gap by comparing, for example, the hourly wages or income of the median worker or the median family in a recent year to that of an earlier year. This is a useful and conventional way of tracking group trends where the median is representative of a group. However, this method does not directly portray how individuals or individual families actually experience the economy. Both dimensions of economic reality are important. Analysts need to track how people experience the economy as they and their cohort age, as well as how groups such as communities, the nation, the working class, or the middle-income family are faring in the economy. The difference arises because the comparisons we have shown thus far inevitably compare different people—for instance, the median worker in 1995 is not likely to be the median worker in 2005.

An illustration reveals the differences between these approaches. How can it be true that the average high-school-educated or non-college-educated worker’s wage (inflation-adjusted) can fall over a 10-year period while most individual workers in this group saw their inflation-adjusted wage rise over that same period? As individuals gain experience, change jobs, or get promotions, they usually obtain higher pay, and so it makes sense that the vast majority of workers earn more when they are in their thirties than in their twenties and make further gains into their forties and fifties. Yet it still can be true that the average wage of a group—say, non-college-educated workers—can decline over a decade or two. This is because the average includes workers of all ages and the process by which a group’s average wage has declined arises because younger members of the group start out at lower wages and progress more slowly than did their predecessors.

Figure F illustrates this phenomenon. Consider the annual earnings of the group of workers with some college—a group with schooling beyond high school, possibly including an associate’s degree, but without a degree from a four-year college. We can illustrate the process of workers raising their wages as they age by following a cohort of “some college” workers over time. For instance, in 1970, when these workers were 20 to 24 years old, they earned an average annual wage of $23,071 (in 2005 dollars). By 1980 this group’s earnings had grown to $32,498, a 41% improvement. By ages 40-44 in 1990 their earnings had grown 22% further to $39,488, and by 2000, at ages 50-54, they had grown to $43,809. Overall then, this group saw its earnings grow 90% as it progressed from 1970 to 2000.

Figure G puts this experience of the 1970 cohort into a broader historical context by presenting the annual earnings of particular age groups at each 10-year point from 1950 to 2000. This exercise allows a comparison of how particular age groups in 2000 fared relative to similarly aged workers in the earlier decades. As the figure shows, the $43,809 earnings level that the 1970 cohort attained in 2000 at ages 50-54 was roughly 10% less than the earnings of $48,511 attained by similarly educated workers at the same age 30 years earlier in 1970. Compared to this earlier cohort, the members of the 1970s cohort made far less progress over their working lives. In fact, the earnings of every age group of workers with some college in 2000 were less than what that same age group of some-college workers earned in 1970. So, although the cohort climbed the ladder successfully, the rungs on the ladder were lowered. That is why the average earnings of prime-age earners (ages 25-54) with some college were 11% lower in 2000 than in 1970.

Another way of viewing this development is to compare the starting earnings of the same age/education group over the decades. As the economy becomes more productive and workers become more educated, we expect new cohorts to start out with higher earnings than their predecessors. Table 5 illustrates that this process prevailed from 1950 to 1970, as earnings for 25-29-year-olds for both high-school-educated workers and those with some college grew each decade. In this period of fast productivity growth and broadly shared income growth, their starting earnings grew appreciably (about two-thirds more) as each successive cohort started out on a higher rung of the ladder.

However, in the ensuing years, including the rapid growth period in the late 1990s, the starting wage remained below that obtained in 1970. So, despite an economy that was two-thirds more productive in 2000 than in 1970, the beginning earnings of high school workers and workers with some college were actually lower.

Every era’s economy seems new to contemporary observers, and our discussion of the historically large gains in productivity is not meant to oversell the notion that today’s economy is vastly different from yesterday’s. As has always been the case, the majority of families work hard to make ends meet, improve their living standards, and create better opportunities for their children. This remains the case today much as it was a century ago.

Yet there are clearly aspects of today’s economy that make it historically unique. Some of these tilt against the bargaining power of American workers: increased global trade, fewer unions, and more low-skilled and high-skilled immigration. There are fewer favor-able social norms that guide employer behavior and support public and employer policies that provide adequate safety nets, pensions, and health care arrangements.

Other new forces in play have the potential to lift the living standards of working families in ways hardly seen in this country for 30 years. Most important of these is a new, stronger productivity growth regime and a brief encounter with full employment which showed that, once workers’ bargaining power gets a boost, the benefits of this regime shift in productivity growth can be broadly shared.

In other words, the biggest challenge in what many have called the new economy is not growth per se; it’s how growth is distributed. Of course, economists and policy makers should be concerned with whether the economy is growing as fast and efficiently as it can, and they might turn to greater investments in public and private capital stock, more research and development, monetary policy that stresses full employment, and the educational upgrading of the workforce.

Yet, if the findings in the hundreds of tables and figures that follow can be reduced to one observation, it would be that, when it comes to an economy that is working for working families, growth in and of itself is a necessary but not a sufficient condition. The growth has to reach the people: the bakers need to benefit from bread they create each day of their working lives.

The benchmarks by which we judge the economy must reflect these distributional concerns, and we must construct policies and institutions to address them. If we do not—if our enhanced productive capacity continues to benefit mostly the wealthiest Americans—we risk sacrificing bedrock principles that have historically defined the American economic experience. Such principles include basic fairness—the notion that those who work hard can truly get ahead—as well as the belief that the economy will provide growing opportunities for the least-advantaged while sustaining a large and flourishing middle class.

In the previous version of this book, we wondered whether the distributional problems of the then-young post-2000 expansion would work themselves out. By now, in the fifth year of the recovery, it is clear that they have not. Some commentators urge working Americans to be patient and wait for the lags in the economy to play out. Eventually, we are told, wages and incomes will start rising with productivity.

Except if they don’t. In an era with severely weakened distributional institutions, where labor market slack is the norm and full employment is the exception, working families can-not afford to wait for growth to catch up with them. That argument might have some saliency in the second or third quarter of an expansion. But by year five, it is an unacceptable position for policy makers to take. The time for economic growth to be a “spectator sport” for the majority of American families is past.

And even if those urging patience are correct, what about the lost jobs, income, and wages that have already been experienced? What about the years spent in poverty that might have been prevented or at least diminished if more of the growth were flowing widely? These losses can never be made up.

Thus, there is a role for policy makers in reconnecting growth and prosperity. The goal of the chapters that follow is to provide readers with the empirical analysis needed to judge the extent to which the economy is working for working families. In the interest of suggesting ways to reconnect growth and living standards, we have also created an accompanying policy document, available at www.epi.org, to go with this book.

America’s working families continue to work harder and smarter. But, while the economy provides them with the potential for prosperity, that potential has yet to be consistently realized. Until that occurs—until living standards can once again be counted on to regularly reflect the benefits of growth—the state of working America will continue to be challenged by inequities that undermine our basic values.

[top of page]

jump to:
[table of contents] [executive summary] [introduction] [documentation and methodology]

[return to main excerpts page]

Download print-friendly PDF version of Table of Contents, Executive Summary and Introduction