The State of American Labor

Deindustrialization and the "gig economy" can't explain the weakness of the American working class.

We all know that there’s something different about today’s working class. One obvious difference is that today’s working class produces fewer things “you can drop on your toe,” as the Economist famously put it, and more that you can’t. What’s actually changing in capitalist production in the United States?

While Marx mostly spoke of industrial workers who extracted or made goods, he did not, in fact, define the proletariat by what commodities it produced. As he wrote in Capital, “capital is indifferent to the particular nature of every sphere of production.”

For Marx social classes were defined by their relations to capital. Capitalist production produces not only “commodities, not only surplus value, but it also produces and reproduces the capital-relation itself, on the one hand the capitalist, on the other the wage-laborer.”

From this comes three basic conditions that define the working class: the need to sell one’s labor power “for a definite period of time” to live; exploitation that produces surplus value; and the nature of the labor process that “is purely despotic.”

Whether employed full-time, part-time, precariously, or for years at a time, about two-thirds of the workforce fit into these conditions.

As a class, of course, the proletariat composes more like three-quarters of society. Not all of the employed workers directly produce surplus value, but as they are “capitalistically employed” and work longer hours than it takes to produce their own income, their “socially necessary labor time,” they are exploited and working class.

In a sense, the current debate over just how much employment is or isn’t “precarious” misses the bigger change in US working-class life over the past three decades or more: the decline in living standards experienced by the vast majority of this class.

One measure of this is the fall in both hourly and weekly real wages, which despite some ups and downs remain below their 1972 level. So stagnant has been the income of the working-class majority that 30 percent of the workforce now relies on public assistance to get by.

Furthermore, the labor share of income has declined in relation to capital, whose piece of the pie climbed from 18.8 percent in 1979 to 26.2 percent in 2010.

If the US Bureau of Labor Statistics (BLS) employment projections for 2014–2024 are any guide, over 80 percent of all new non-managerial, non-professional jobs will fall in the official low-income range below $32,390 a year, over a third of those in the very low range below $21,590.

There are fewer workers making goods in the United States and more providing services, or at least what the BLS calls services. The blows to industrial employment that accelerated in the 1970s and 1980s, particularly in the “Rust Belt” of the Midwest, still resonate in working-class culture.

But there appears to be more to it than that. Employment everywhere, it seems, is not what it used to be: full-time, regular, reasonably well-paid as in the Keynesian-Fordist era.

Of course, it wasn’t exactly like that even then. As one study of precarious work across the developed industrial economies reminds us, “The assumption that the principal norms regulating work were those of full-time permanency has never reflected the full variety of working relationships in industrial economies.”

Still, something has changed, and there have been no shortage of attempts to explain what is new. By way of explanation in the last few years we have been offered the gig economy, the sharing economy, the servant economy, the “age of self-employment,” and the precariat.

More people are moving from job to job, more are self-employed, and more jobs are insecure. Or so these stories go.

Speaking of the alleged rise in the self-employed, economist Robert Solow summed up the problem with many of these propositions when he said, “you can see the age of self-employment everywhere except in the self-employment statistics.”

In fact, if we look past the neologisms, we will see that many of these characterizations of post-Keynesian or post-Fordist capitalism do not hold up well.

The Gigariat?

Let’s take the gig economy. To situate ourselves in the parade of economic neologisms we will call these nomadic workers the “gigariat.” (The term gig, of course, was coined by jazz musicians who had to go from job to job to make a living.)

The gig economy is usually said to have emerged in the wake of the Great Recession of 2008. In this emerging economy, job-seekers cobbled together a living by taking and letting go of two or more jobs, sometimes through internet outfits like TaskRabbit, Fiverr, or Uber.

Although Uber denies it is an employer, while most of the others are simply digital “platforms” that provide a link between employers and gigsters. (Some accounts throw in outfits like Airbnb, eBay, or Etsy because they can bring in income, but these are digital marketplaces, not labor markets.)

A JPMorgan survey found when it looked at what it called “capital platforms” such as Etsy, Ebay, and Airbnb, versus “labor platforms” such as Uber and TaskRabbit, that it was the “capital platforms” that captured the lion’s share of the 1 percent of adults who used any income-generating digital platform. By mid-2015, those who used “labor platforms” accounted for .04 percent of adults surveyed.

Not surprisingly, the BLS does not enumerate the gigariat. But assuming these peripatetic workers do actually work someplace for someone — even if themselves — they must show up somewhere in the BLS’s statistical series.

An obvious place to start is with the category of multiple jobholders, since gigsters famously hold multiple jobs. Yet while there are millions of multiple jobholders, their numbers haven’t grown since the early 1990s.

Since the alleged acceleration of the gigariat is usually placed in the wake of the Great Recession when the number of multiple jobholders fell to 6,878,000 by 2010, the subsequent increase in these workers to 7,262,000 by 2015 appears as a trend.

The problem is that in 2007, just before the recession, there were 7,655,000 multiple jobholders, while in 1994 there were already 7,260,000.

In other words, in over twenty years the number of these workers has not grown much and their percentage of the workforce has actually fallen from an average of 6 percent in the 1990s to 4.9 percent from 2010–15.

Furthermore, 55 percent to 58 percent of these workers are full-time workers who hold one or more additional jobs. Far from emulating the jazz-club gigging of Charlie Parker or Thelonious Monk, these overworked employees are engaged in an age-old and (despite its name) un-romantic practice of “moonlighting” to make ends meet.

Perhaps the rest of the multiple jobholders who string together part-time and variable-hours work include some genuine gigsters, such as the singers and musicians whose ranks rose from about 55,000 in 2001 to 75,000 by 2013, but overall their numbers are down since the 1990s as well.

In fact, the percentage of multiple jobholders in the employed workforce, with some ups and downs, has not changed significantly since the 1970s when they averaged 4.9 percent just as they do today.

Self-employment is often mentioned as part of the gigariat. The number of un-incorporated self-employed workers, who form about two-thirds of the self-employed, however, has fallen for some time.

The greatest increase in these workers came in the 1980s, leveled off and then declined. In fact, they have declined in numbers since 1990 and as a proportion of the workforce since 1967!

The incorporated self-employed (i.e., who register as “corporate” entities for business and tax purposes), on the other hand, have grown since the late 1980s both in numbers and as a percentage of the employed labor force, although they remain only about 4 percent or less of the workforce.

These are mostly small-business entrepreneurs, however, and don’t fit into our picture of the working class even if some conform to the gig image.

It seems that the enthusiasts of the gig economy have looked at job trends for a few years since the Great Recession and generalized the future.

What they saw, in reality, was not a trend but a long-standing cyclical feature of multiple jobholding, which happened to be accompanied by an as-yet-marginal rise in those seeking work through the internet rather than newspaper want-ads, old-fashioned employment agencies, or vanishing state employment services.

Even if this latter trend grows to be a norm among job seekers, it will not by itself define the nature or stability of future employment, a power that still resides with capital, which is the shaper of the labor market and the jobs in it and the actual employer of more and more workers.

There are, to be sure, millions of workers mis-classified as self-employed “independent contractors,” such as construction workers, taxi drivers, some truckers, etc.

While some of these might fit the gig profile, such as the 150,000 or so Uber “driver-partners,” many are simply in industries long characterized by intermittent or seasonal employment and dominated by contractors who are large or small capitalist employers.

The “independent contractors” who work for them or others who employ this dodge are not really self-employed at all. This is a legal fiction exploited by employers to evade paying benefits, employment taxes, insurance, the cost of layoffs, etc.

Overall, there isn’t much evidence of a burgeoning gigariat.

What there is for many — particularly for younger workers relatively new to a labor market still characterized by high levels of unemployment and a growing proportion of low-paid jobs, whether or not they are precarious — is a bleak future.

Or Precariat?

This neologism has a greater claim on reality than the “gig economy.” One result of the “flexibility” of the workforce demanded by lean production has been the increase in precarious employment such as temporary agency work, short-term contracts, on-call work, independent contracting (i.e., bogus self-employment), involuntary part-time work (for economic reasons, where work hours often actually amount to full-time), all counted by the BLS.

The BLS categories for “contingent and alternative employment arrangements,” however, do not include those who regularly work part-time. I believe this is a correct exclusion because these workers are employed in industries where employment traditionally averages about thirty hours a week.

The growth of “usual part-time” workers by 7.2 million from 1990 to 2015 is fully explained by the growth of 20 million employees in retail trade, health care, administration, and waste services, and leisure and hospitality, who work from twenty-five to thirty-four hours a week and always have.

By 2005 BLS estimates (the most recent available, adjusted for some undercounts of temporary and involuntary part-time workers and overlap of categories), some 21.6 million employees worked in precarious situations, compared to 18.7 million in 1995.

That’s an increase of nearly three million precarious jobs, certainly a significant growth in precarious work over one decade. Yet surprisingly, the proportion of precarious workers hardly rose at all, from 15.2 percent in 1995 to 15.5 percent in 2005.

The 1995 survey was the first of its kind, so there are no comparable figures for earlier years. Nevertheless, we can get an idea of how much precarious employment grew before 1995 by looking at a couple of available statistics.

In 1980, all personnel supply services (which includes temp agencies, but is broader) supplied only 543,000 workers. By 1990 those working for temp agencies alone rose to 1,288,000 and then to 2,189,000 in 1995.

Similarly, the number of un-incorporated self-employed, most of whom would be “independent contractors,” grew by 1.6 million between 1980 and 1995 before their numbers declined.

Thus, it is most likely that the biggest jump in precarious work came with the initial spread of lean production in the 1980s and early 1990s.

It is also possible that precarious jobs have increased since 2005. The data available on this, however, don’t point clearly in this direction.

Temporary agency employment dropped from its high point of 2,637,000 in 2006 to 1,823,000 in 2009, as employers dumped temps in the wake of the recession, and rose to 2,869,000 in 2015, indicating that employers still turn to temp agencies when business improves.

Those who usually work full-time, but were forced into part-time work for economic reasons, did rise from 1,556,000 to 2,245,000, almost 700,000 higher. The un-incorporated self-employed, i.e, “independent contractors,” however, fell by almost 1.5 million people from 2006 to 2015.

Temps did rise as a percent of the non-farm workforce, reaching a new high of 2 percent by 2015, but the un-incorporated self-employed fell as a percent from 7.4 percent in 2006 to 6.3 percent in 2015.

So while there might have been some net gain in total precarious jobs, it seems unlikely the overall proportion of precarious workers in the total workforce could have risen much since 2005 or 1995.

Give or take a couple of percentage points, it is hard to avoid the conclusion that at least since the mid-1990s precarious work in the United States has not grown as much as many impressionistic accounts claim, and that the large majority of workers, about 85 percent, are still in “traditional” employment arrangements; though these like the precarious workers have seen their incomes and conditions change significantly for the worse.

This limited growth in precarious work is further attested by the fact that job tenure has not changed much in the United States since the introduction of lean production norms and neoliberalism in general.

For those of age twenty-five to thirty-four the average length of job tenure fell from 3.8 years in 1979 to 3.5 in 2006, while those in the thirty-five to forty-four age range saw it fall from 7.1 years to 6.6, and those ages forty-five to fifty-four from 11.3 to 10.3.

These figures, of course, don’t include the crucial eighteen to twenty-four age cohort, precisely when new entrants to the workforce will experience the most precarity.

BLS figures covering wage and salary workers of all age groups and industries, using median years of job tenure, actually show an increase for all groups. For all those sixteen and older the median years on the job at the time of the survey rose, with business cycle ups and downs, from 3.5 years in 1983 to 3.8 years in 1996 to 4.6 year in 2014.

To be sure, nearly twenty-two million workers in precarious employment is no small number. Nor is the influence of so much insecure work on the conditions of those in traditional employment situations likely to be negligible.

Indeed, more and more jobs of all kinds are “dead end” in that they don’t offer a clear path to higher earnings as wages remain low over time and benefits become rarer.

Nevertheless, on average workers still hold jobs for a number of years: the longer one is in wage labor, the longer the job lasts on average. The idea that workers change jobs all the time, making organizing impossible, is misleading.

A Changing Proletariat

An important change in the composition of the employed working class is the proportion of workers of color, which grew from about 15–16 percent of workers in production, transportation, and material-moving occupations as well as in service occupations in 1981 to 40 percent of each of these broad occupational groups by 2010.

Immigration has played a major role in this increase. Along with African Americans and women, immigrant workers will fill many of the low-wage jobs projected by the BLS. Some will also join the labor movement.

Reflecting this, 200,000 Latinos joined unions between 2011 and 2014, while 96,000 Asian and Pacific Islanders joined in one year between 2013 and 2014.

Probably most commented on, however, is the decline of manufacturing employment from 27 percent of private employees in 1980 to 11 percent in 2010. It is this change that most often leads to speculation about the declining significance and power of the working class.

While manufacturing has been a declining source of employment for a long time, the dramatic loss of nearly five million manufacturing, production, and non-supervisory jobs since 1980 calls for an explanation.

Many, particularly in the labor movement, argue that the culprit was trade. Clearly some industries like basic steel, textiles, garments, etc. saw big losses to imports. But these losses account for only about 20 percent of the five million.

Nor does “offshoring,” which grew over much of this period but recently slowed down, account for massive losses, as domestic content in US manufacturing still averages about 85–90 percent, well above the global average of 72 percent.

As the United Nations observed, “Large economies, such as the United States or Japan, tend to have significant internal value chains and rely less on foreign imports.”

The problem with trade-based explanations is that manufacturing output hadn’t shown a decline, but had grown in real terms by 131 percent from 1982 to 2007 just before the Great Recession reduced output.

At an annual average of 5 percent this is only slightly less than the 6 percent annual growth of the 1960s. The mystery behind this massive loss of jobs lies in both the destruction of capital, on the one hand, and its increased application in the last thirty years, on the other.

The disappearance of manufacturing jobs hasn’t followed the more or less steady upward trajectory of imports since the mid-1980s.

Rather, massive job destruction has occurred during the four major recessions of this period as capital itself has been destroyed: in 1980–82 2.5 million manufacturing production jobs lost; 1990–92 725,000; 2000–3 about 678,000, and during the Great Recession another two million jobs gone.

Between the recessions of 1980–82, 1990–92, and 2000–03 output increased by 6 percent a year, but employment remained flat due primarily to the large productivity gains, averaging over 3 percent a year achieved by capital through the application of new technology and lean production methods often supplemented or even supplanted by biometric and electronic monitoring, measuring and enforcing of labor standardization, and intensification.

One measure of the intensification of labor over these years has been the decrease in break time from 13 percent of the work day in the 1980s to 8 percent in the 2000s for those in routine goods and service-producing jobs.

Both growing investment and work intensification are behind this rising productivity. Real private fixed assets in manufacturing doubled between 1979 and 2014, while manufacturing employment fell by over 40 percent.

While fixed investment for the whole economy, like GDP, grew more slowly since the early 1980s than in US capital’s heydays following World War II, the proportion of non-residential fixed investment in GDP has actually been larger than back then: 11–12 percent of GDP compared to 10 percent during the 1960s.

As a result, the capital-labor ratio for the economy as a whole, which was basically flat in the 1970s, rose in the 1980s and accelerated during the 1990s, increasing by almost two-thirds up to the Great Recession.

The increase in “service” employment, on the other hand, is explained by the shorter hours worked in many of these occupations; the statistical shift of outsourced services from the manufacturing column to services such as food services, accounting, data processing, security, etc.; and above all huge increases in the commodified labor of social reproduction.

As millions of women entered the workforce from the 1950s onward, and their hours of work increased from 925 a year in 1979 to 1,664 in 2012, with those of women with children growing from 600 to 1560 over that period, capital stepped in and organized the commodification of many aspect of social reproduction formerly done in the home such as health care, elder care, food services, etc.

This trend created some eight million new service jobs from 1990 to 2010. Millions more were created to maintain capital’s expanding facilities and buildings and clean up its accumulating mess.

Few of these jobs are “white collar,” most are physical, their pay is low, and many are filled by women, workers of color, and immigrants.


One of capitalism’s features is that the structures of production and distribution change over time under the pressures of competition. Thus, in the first phase of the era of globalization, neoliberalism, and lean production, roughly the early 1980s through the mid-1990s, the owners of such capital as had remained in the “Rust Belt” sought to escape the huge urban concentrations that had fostered unionism and high labor costs.

Relocation from urban to semi-rural areas, mostly still in the Midwest but some in the South, and the vertical disintegration of production through outsourcing, subcontracting, and offshoring, swept not only goods production but the organization of services and distribution as well.

The experience for workers and unions was one of disruption and fragmentation. To a large extent this is still the dominant narrative on the Left. But the very competition that brought on this fragmentation eventually created two major counter-tendencies.

Under increased competitive pressures at home and abroad, relocation and vertical disintegration called forth the reorganization of supply chains along just-in-time metrics, a reduced number of suppliers, and technologically linked and guided systems of commodity movement, all organized around giant logistics “clusters” that employed tens of thousands of workers in relatively finite geographic areas.

85 percent of the nearly 3.5 million workers employed in logistics in the United States are located in large metropolitan areas — inadvertently recreating huge concentrations of workers in many of those areas that were supposed to be “emptied” of industrial workers.

There are about sixty such “clusters” in the United States, but it is the major sites in Los Angeles, Chicago, and New York-New Jersey, each of which employ at least one hundred thousand workers and others such as UPS’s Louisville “Worldport” and FedEx’s Memphis cluster that exemplify the trend.

These are home to new types of warehouses a growing number using cross-docking in which “Same-day receipt and dispatch is the target.”

In addition, many of the workers in these supply chains perform final steps in actual production, including on imports.

Most of the workers in these supply chain gangs fit Marx’s definition of value-producing transport workers as analysed in the Grundrisse and Capital Volume II, and are, in fact, an integral part of the overall production process even of imports.

From the spatial disintegration of production has come its temporal, just-in-time, electronically guided re-integration — “the annihilation of space by time” — and with it the tightening of the links between concentrations of workers.

And as a consequence, comes also the heightened vulnerability of supply chains, transportation “nodes,” production sites, and retail outlets to worker actions.

Industrial Consolidation

Another consequence of increased global and domestic competition has been the largest corporate merger movement in the history of US capitalism. In 1980 mergers and acquisitions numbered 1,560 at a value of $32.9 billion, rising to 4,239 worth $205.6 billion in 1990 and 11,169 valued at $3.4 trillion in 2000.

After that they leveled off at about seven thousand a year until the Great Recession. Since then they have grown again. Unlike the conglomerate mergers of the 1960s and 1970s, these waves of mergers have been along rational industrial lines.

Companies are returning to their “core competencies” and seeking to increase profits via expanded market share by devouring the competition.

This, together with the rising capital-labor ratio, has meant both the concentration and centralization of capital in industry after industry, resulting not in “monopoly” but in even more intense competition by larger firms.

A few examples of the results will suffice to make the point. By 2009, due to a massive merger movement and consolidation of the auto-parts industry, the top ten parts-suppliers controlled a third of the US original equipment market.

The top four meatpacking firms controlled 75 to 81 percent of all US meat production. UPS and FedEx now employ 40 percent of all trucking and express delivery workers. After years of mergers, five rail freight carriers now employ 80 percent of freight rail workers.

Following the breakup of AT&T in 1984, the industry re-consolidated so that today four companies control 90 percent of both the wired and wireless telecommunications’ market. Since 2000 the number of major airlines has fallen from ten to four.

Today, almost three-quarters of the nation’s formerly independent (non-publicly-owned) community hospitals are in urban-based corporate systems. Walmart and e-commerce outfits like Amazon have brought about consolidations in both general retail and retail grocery.

By 2009 the top eight food retailers controlled almost half the national market. The examples could go on and on, but what is important here is that capital has consolidated meaning that more and more workers are employed by relatively larger firms along industrial lines.

At the same time, more and more workers are employed in workplaces that are both more capital intensive and employ more workers on average. The vast majority of workers in the United States, about 80 percent, have always worked in relatively small workplaces of fewer than five hundred workers and this proportion hasn’t changed much.

Nevertheless, while the number of manufacturing workers in “establishments,” i.e. single facilities or workplaces not firms, has fallen, the number who work in facilities employing five hundred or one thousand or more workers in the total private economy has grown.

Excluding those in FIRE (finance, insurance, and real estate) employment, from 1986 to 2008 some 8.2 million more workers were employed in workplaces with five hundred or more workers, while 5.7 million more work in those with a thousand or more workers.

One of the big gains has come in health and social care, where 4.4 million workers are employed in facilities of a thousand or more employees with hospitals accounting for much of this.

In addition more workers in service and retail occupations work in large concentrations in big-box retailers, hospitals, hotels and call centers that along with those in logistics clusters are the “factories” of today.

Taking the Reins

By the turn of the twenty-first century these trends — lean production norms, increased capital intensity at work, the “logistics revolution,” and industrial consolidation — had converged to alter the terrain of class conflict as well as the structure of industry and of the working class itself.

The shadow of “deindustrialization” still hung over large parts of the country and continued to imbue much of the industrial working class with a sense of lost power.

While there would be no “re-industrialization,” no new towering steel mills or inner-city auto assembly plants employing thousands, the production of goods and service has been restructured, concentrated, and linked in ways that could be to the advantage of working class organization and action.

For one thing, both goods- and service-producing industries today more nearly approximate the more or less clear lines of production that made industrial unionism possible in the 1930s. For another, increased capital intensity could allow for greater gains by workers.

Finally, the just-in-time supply systems that tie together the production of most good and services today makes the entire system more vulnerable to worker action.

The huge concentration of labor in the logistics clusters within or near large metropolitan areas provides an organizing target on a scale that could reverse the decline of unions.

While some workers have lost power over the years, many more have gained new sources of potential power.

These, of course, are objective conditions, not guarantees of success. There are, however, signs that the passivity and resignation that lingered so long is passing, particularly among the young.

From teamsters to teachers there are rank-and-file rebellions in many unions that reject the norms of bureaucratic business unionism. Workers once thought of as having little social power have injected action and even successful organizing into hotels, building services, hospitals, and to some extent fast food chains.

Resistance and organization in some of the major sections of the working class have produced a new activism through immigrants’ rights groups, the growing number of workers’ centers, the Black Lives Matter Movement, the local and national fights for $15, unofficial strikes in warehouses and Walmart stores, and even Bernie Sanders’s run for the presidency, particularly the Labor for Bernie campaign that brought in thousands of union members.

Then, there is the fact that by 2014, 56 percent of “millennials” or Generation Y-ers characterized themselves as “working class” rather than “middle class,” followed by 50 percent of Gen X-ers and 44 percent of baby boomers.

Perhaps even more surprising are the recent polls that show large percentages of adults with a “positive image” of socialism or a willingness to vote for a socialist.

This revival of grassroots activism along with new forms of rudimentary class consciousness and political openness could spill over into permanent worker organization if, and it is a big “if,” either unions take on the task and embrace this activism or if the workers in some of the key newer areas take on the job themselves.