Commentary | Wages, Incomes, and Wealth

Challenges Facing American Workers

Opinion pieces and speeches by EPI staff and associates.


Challenges Facing American Workers

By Jared Bernstein

Introduction: The Increasingly Insecure American Worker

Chairman McDermott, ranking member Weller, I thank you for inviting me to testify and applaud this committee for taking up this issue of great concern to working families across America.
Many of these families are facing uniquely tough times.  Most recently, a recession has gripped the labor market.  Payrolls are down by over 600,000, unemployment is up sharply, and compensation is consistently lagging inflation. 

But the difficulties facing American workers predated the recession.  There may be no more telling statistic of this point than the fact that the real wage for the median male was lower in 2007 than in 1973.  In the same spirit, it has been widely recognized that the current business cycle of the 2000s is the first on record where the income of the median family gain no ground in real terms, despite strong productivity growth over these years. 

In other words, for many in the workforce, income, wages, and compensation have failed to keep pace with their contribution to their firm’s output, violating both a fundamental principle of economics and a basic American value.  For the last few decades, they have been losing employer-provided health coverage, or paying more out-of-pocket for premiums, health services, or medications.  Their pensions are less secure, and have flipped from majority guaranteed benefit to guaranteed contribution, shifting the risk of an adequate retirement benefit from their employer to themselves and their family.

Some aspects of jobs have also become less secure.  Over the longer term, job tenure has declined, especially for men.  More recently, job creation was particularly weak, and this had led to numerous problems in the job market.  The share of persons stuck in long-term unemployment—at least six months—was much higher on average in the 2000s than in earlier periods.  For the first time on record for a business cycle, the share of adult population at work never regained its prior peak, meaning employment rates were lower at the end of the 2000s business cycle than at the beginning (this analysis assumes that the recession, or at least a labor market recession, began around January of this year).

This testimony briefly outlines some of these points, focusing first on current recessionary conditions, then on recent trends over the 2000s, and finally on longer term trends in compensation, inequality, and other factors contributing to worker insecurity.  The testimony concludes with some explanations for why these trends persist and some policy suggestions.

The Current Job Market

As noted, employment has contracted consistently this year, down 605,000 overall and over 700,000 in the private sector (government job creation is less cyclically sensitive).  As shown in Figure 1, Unemployment has risen almost two percentage points since its most recent low in early 2007, and underemployment, a more comprehensive measure of the extent to which workers and potential workers are underutilized, is already higher than at any point in the last recession or jobless recovery that followed.

This last point regarding underemployment bares more examination.  Along with the growth in the number of the unemployed, the largest contributor to its growth in recent months is the increase in so-called “involuntary part-time workers,” persons working part-time who would prefer full-time jobs.  In August, there were 5.7 million of these underemployed persons, up 1.2 million from one year ago.

One symptom of this weakening job market, in tandem with the recent, commodity-driven acceleration in inflation, is reduced earnings.  Figure 2 shows annual changes in inflation-adjusted earnings, including hourly and weekly earnings for production and non-supervisory workers, as well as an average, economy wide measure of total compensation: wages plus benefits.  All three measures have been falling for the last few quarters, with the broadest measure, average hourly compensation for all workers, falling most quickly.  Note also that weekly earnings are falling faster than hourly earnings, due to the decline in average hours of work.

Though inflation may grow more slowly in coming months, as gas prices have come down off their recent peaks, these labor market conditions are not expected to improve in the short term.  Most forecasts are for unemployment to continue to increase and remain elevated in recessionary territory through next year.

These shorter-term difficulties are characteristic of recession.  What is more unusual is that the job market of the 2000s, i.e., over the expansion, was characterized by many of these same trends: job and wage growth was historically slow, and incomes of middle-income, ended up significantly lower at the end of the cycle than at the beginning.

The 2000s: Weak Job, Wage, and Income Growth Amidst Strong Productivity

As noted, though these difficulties have deepened in the downturn, the business cycle that appears to have ended late last year was uniquely unrewarding to working families, especially considering their contributions to productivity growth.  For example, Figure 3 shows the trend in the real median income of working-age households—those headed by someone less than 65— 1989-2007.   Their median income, after adjusting for inflation, fell $2,000 between 2000 and 2007, from about $58,500 to $56,500 (2007 dollars).

The trend was very different in the 1990s.  After declining in the recession (and the jobless recovery that followed), the median income of working-age households reversed course and rose consistently through 2000.  Over the 1990s (1989-2000), it was up almost 10%, or about $5,200.  Had this growth rate prevailed in the 2000s, the median income of working age households would have gone up $3,600 instead of falling $2,000.

One key factor behind this result, and it is an important source of worker insecurity, is the historically weak job growth over the 2000s business cycle, the weakest on record going back to the 1940s.  When employment growth is weak, there tends to be less pressure in the job market such that employers need to bid wage offers up to get and keep the workers they need.  This lack of worker bargaining power shows up as weak wage and income growth for working families, even amidst strong productivity growth and relatively low unemployment.

Figure 4 plots the annual growth in jobs in this versus past cycles and shows the rate in the 2000s was less than one-third the average rate.  In terms of numbers of jobs, compared to the 1990s, payrolls expanded by about 23 million; in the 2000s, payrolls grew by less than 6 million. 

A symptom of weak job growth is that once workers lose their jobs, their unemployment spells can be quite long, another factor contributing to weak income growth and increased worker insecurity.  This tendency has also been exacerbated by the aging of the workforce, since older workers tend to be choosier about their job offers and thus have longer spells of unemployment.  The result, as shown in Figure 5, is an historically large g
ap between unemployment and the share of the “long-term” unemployed: persons who have been jobless for at least six months.

This increase in the share of the long-term unemployed is especially notable given the secular decline in the unemployment rate.  A lower jobless rate suggests a tighter job market, which might lead us to expect that unemployment spells would be diminished.  Figure 6 shows that this is not the case: though the average unemployment rate was slightly lower in the 2000s cycle relative to prior cycles, spells of unemployment were considerably longer.  About seven percentage points more of the unemployed were long-termers over the 2000s cycle, and they were, on average, unemployed for about five more weeks.

The insecurity bred by this longer-term unemployment was something experienced not solely by marginal, less educated, or younger workers.  Older and college-educated persons increased as a share of the labor force over these years, but as shown in the Table 1, also as a share of both the unemployed and the long-term unemployed.  

Part of this is a composition effect of the aging labor force, but the tendency to experience long-term unemployment also increase within these groups. Table 2 shows the increased likelihood of long-term unemployment by education and age.  As time has progressed, more highly educated and older workers are more likely to experience longer spells of joblessness.

Longer Term Evidence

Earlier, it was argued that productivity and earnings diverged significantly in recent years, but this is not a recent phenomenon. Figure 7 plots the average compensation—wages plus benefits—of non-managers in services and blue-collar workers in manufacturing, against productivity growth.1 Between the mid-1940s and the mid-1970s, the real compensation of these workers and the productivity of the American workforce grew in lockstep, both doubling. 

Since the latter 1970s, however, the two trends in the figure diverge.  Real compensation grew only 7%, 1979-2007, while productivity grew 70%.  More recently, the gap has grown particularly wide, as productivity grew more quickly in the 2000s business cycle than in either that of the 1980s or 1990s, while average compensation of these workers was flat.

This split between economic growth and the labor market earnings of working class persons is at the heart of today’s economic insecurity.  Of course, in polling and the popular debate, that insecurity often is associated with the difficulty that working families have making their budgets: the “middle-class squeeze.”  But the squeeze itself is intimately related to the previous figure, wherein too few workers can count on their contribution to the economy’s growth to boost their living standards. 

Figures 8
and 9 show a broader set of wage trends which underscore these points by shows the disparate paths of real wages for men and women since 1973 and different wage percentiles.  For both genders, wages “fan out” significantly in an unequal pattern.  For men, median (50th percentile) wages are essentially unchanged over these years, while lower wage men lost ground.  Women’s wages grew for each group, though much faster at the higher end of the wage scale.

Why has this split between wages or compensation and productivity occurred, leading the inequality patterns in the previous two figures?  Various explanations have been offered and I offer only short summaries here.

— Higher returns to education: This argument maintains that since the latter 1970s, the benefits of growth have flowed disproportionately to those with higher levels of education.  There is some evidence to support this, but it is by no means a complete explanation.  Advocates have failed to show that employers’ skill demands accelerated over this period, and more recently, the college wage premium has been relatively flat, implying that inequality is being driven by other forces.  Finally, we note that the real wages of college-educated workers have been quite flat in recent years, up only 2.5%, 2000-07.
— Diminished bargaining power: Less collectively bargaining has contributed to inequality.  Research on unionization’s impact on wages is quite clear on the point that less union density in the workforce has contributed to the growth in inequality.

— Increased trade: The increase in traded goods, in tandem with large and persistent trade deficits has been identified as another source of increased inequality, stemming partly from the loss of manufacturing employment.  This is especially the case in trade between our economy and developing economies, like China, that are heavily endowed, relative to us, with large, low-wage workforces.

–Absence of full employment: Periods of very tight job markets have been associated with a more equitable distribution of earnings, as such periods boost the bargaining power of less-advantaged workers who would otherwise be in excess supply, face discrimination, or simply have less leverage than other groups of workers.  As an example of this effect, note the period of the latter 1990s, when unemployment ultimately fell below 4% for the first time in 30 years, in the prior two figures.

Other longer term trends contributing to worker insecurity are:2

–the long-term shift from pensions which guarantee a fixed payout versus variable pensions (i.e., defined benefit to defined contribution), a clear shift in the locus of risk from the firm to the worker;

–the secular erosion of employer-provided health care coverage;

–the long term decline in men’s job tenure, down by 1½ years for men age 34-44, 1973-2006, and two years, for men age 45-54; the share of men with 10 years on the job fell 10 percentage points over these years for men in these age groups; the share with 20 years on the job fell about the same amount.

Policy Actions to Enhance Worker Security

First, Do No Harm: It is important not to exacerbate the problems documented above with policies that exacerbate weak job and wage growth, and increasing inequality.  For example, changes since 2001 to the Federal tax code have worsened the distributional outcomes, by disproportionately lowering the tax liabilities of the wealthiest families.

Such regressive tax policies hurt most families both directly and indirectly.  Directly, they exacerbate the already excessive inequalities in market outcomes (i.e., the pretax distribution).  Indirectly, they diminish revenues such that the Federal government is less able to perform needed functions (without borrowing), many of which, like safety net policies, disproportionately benefit the least well off.  While the direct impact of the regressive tax cuts has been extensively measured and is well-appreciated, this indirect effect—the defunding of public services that boost economic security of the least advantaged—is also important and problematic.

Beyond tax policy, other policy “sins of omission” have contributed to higher inequality.  We have failed to strengthen workers’ legal ability to organize, gutted investments in their skills and training, under-invested in our public infrastructure, or stood by as the employer-based systems of health coverage and pensions slowly unravel. 

Bargaining Power: As noted, the diminished bargaining power of many workers is a key factor in the wage/productivity split and the insecurity problem.  Historically, a b
road set of policies and norms, including unions, minimum wages, defined-benefits pensions, and health care provisions, helped to lift workers’ ability to bargain, and were thus associated with more broadly shared prosperity. 

Unions play a key role in precisely this area.  There decline has been partly a mechanical function of the loss of jobs in unionized industries, like manufacturing, but the more important explanation is the very unbalanced playing field on which unions try to gain a foothold.  In fact, Freeman (2007) argues that slightly more than half of the non-union workforce would like some type of union representation, a finding that is not particularly surprising given the divergence of incomes and productivity shown above. 
The problem here is that the legal and institutional forces that have historically tried to balance the power of anti-union employers and their proxies have significantly deteriorated in recent decades, as described by Shaiken (2007).   One legislative solution is the Employee Free Choice Act (EFCA), a bill that helps to restore the right to organize in the workplace.  A central component of EFCA is so-called majority sign-up or “card-check,” which gives the members of a workplace the ability to certify a union once a majority sign authorizations in favor the union.  The law also puts much needed teeth back into labor law by ratcheting up the penalties for those who violate the rights of workers trying to organize or negotiate a contract.

Macro-Economic Conditions: Full employment—a tight match between labor supply and labor demand—is another important criterion for reducing the gap between overall growth and living standards of working families.  Historically, very low unemployment rates have also been a key contributor to workers’ bargaining power, ensuring that employers needed to bid compensation up to get and keep the workers they needed in order to meet the demand for their goods and services.

The policy levers here, at least in normal times, i.e., outside of recessions, rest mainly with the Federal Reserve, but Congress can also play an important role which I discuss below under the rubric of investment policy.

Safety Nets:  Historically, working families in our country have depended on employers to provide health care and pensions, but as this system unravels, a vibrant debate regarding reform of our health care system is underway.  The details of the debate are beyond my scope here, but it is especially urgent given the realization that the rate of increase in health spending in both the public and private sector is unsustainable. 

I will, however, note that the achievement of guaranteed, affordable health care would play a major role in offsetting the insecurity that working families face around this issue.  For details, interested parties should consult EPI’s Agenda for Shared Propserity, an initiative by our institute to elaborate in some detail the best plans for meeting the challenge of both health care and pension reform.

Another safety net issue in need of attention is Unemployment Insurance.  Given the changes in the structure of work and the demography of the workforce, our nation’s UI system is also in need of reform and modernization.  The Unemployment Insurance Modernization Act, already passed by this chamber, would make such changes, including providing benefits to both part-time workers and those who leave their jobs for compelling family reasons.  The bill also accounts for shorter job tenures by considering a worker’s most recent work history when determining eligibility for UI benefits.

Finally, lower wage workers would benefit from an expansion of so-called work supports: programs that enhance or subsidize the incomes of low-income working families, by either subsidizing the wages (the Earned Income Tax Credit), offsetting their expenses (child, health care, and housing subsidies, for example), or supporting their income (e.g., the child tax credit).

Investments in Human and Physical Capital: Economists widely agree that it is critical to invest in the skills of the workforce of both today and tomorrow.  Unfortunately, our budgetary priorities have been moving in the opposite direction, as federal budgets over the past few decades have shortchanged training programs.  Eisenbrey (2007), for example, shows that Federal investment in employment services and training is down about $1 billion in real terms since 1986 (from about $6 to $5 billion, 2006 dollars) even while the workforce has grown in size considerably over those years.  The result is a decline in the budget for worker training and services from $63 to $35 per worker, in 2006 dollars. 

According to the Coalition for Human Needs (2008) analysis of Congressional appropriations for a number of training programs, real declines have occurred in a number of job training programs between FY05 and FY08.  Spending on both adult (-12%) and youth training (-14%) through the Workforce Investment Act are down, as are dislocated worker training (-9%) and adult basic education (-12%).

Along with human capital, investments in public physical capital should also be considered, and particularly given today’s weak labor market, such investments should be considered in the context of macroeconomic stimulus.

Three facts motivate this contention.  First, American households are highly leveraged, and may well be poised for a period of enhanced savings and diminished consumption.  In this context, public investment should be viewed as an important source of macro-economic stimulus and labor demand—the creation of new, and often high quality jobs—which is clearly lacking from our current labor market. 

Second, there are deep needs for productivity-enhancing investments in public goods that will not be not made by any private entities, who by definition cannot capture the returns on public investments in roads, bridges, waste systems, water systems, schools, libraries, parks, etc.  Three, climate change heightens the urgency to make these investments with an eye towards the reduction of greenhouse gases and the conservation of energy resources.

These are admittedly brief outlines of only a few steps that could help to accomplish the critical goal of reconnecting the living standards of working families to the growth in the economy, especially given that they themselves are responsible for generating much of that growth.  At the same time, policy makers can help to significantly reduce the insecurity generated by unfavorable developments in the economy by strengthening safety nets and social insurance, especially in the areas of health care and pension coverage.  I again applaud this committee for taking up these issues.  Without your attention, and that of your colleagues, these insecurities are much more likely to deepen than disappear.

I thank Tobin Marcus for excellent research assistance, and Heidi Shierholz and Larry Mishel, my co-authors of State of Working America, 2008/09, from which much of the data in this testimony are drawn.

1. This series is derived by scaling up the BLS production, non-supervisory wage series by the ratio of compensation to wages from the NIPA accounts.  It implicitly assigns the average compensation to these lower-wage work force—these workers roughly omit the top 20% of the workforce—a “conservative” assumption in the sense that it is likely an overestimate of their average benefits package.
2. These facts are all taken from Mishel et al, 2008.

Jared Bernstein is a senior economist at the Economic Policy Institute in Washington, D.C.


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

See more work by Jared Bernstein