In a televised speech to the nation in February 1981, then-President Ronald Reagan warned that 7 million Americans were caught up “in the personal indignity and human tragedy of unemployment.” “If they stood in a line, allowing three feet for each person,” he explained, “the line would reach from the coast of Maine to California.” The unemployed were not only former assembly-line workers—they were secretaries, accountants, and cashiers, among other professions, too.

The 1980 recession had bludgeoned the labor market, and the 1981-82 recession was about to bludgeon it even more. Manufacturing plants across the country were shutting down or relocating. IBM was about to introduce the personal computer, rendering skills such as shorthand all but useless. A crack epidemic was ravaging cities. The unemployment rate, which had been hovering around 7 percent from 1980 to 1981, was rising; it would be close to 11 percent by 1982. The federal government’s main source of funding for job training was due to expire a year later.

From one vantage point, today’s workers seem to be facing a vastly different landscape: The U.S. has somewhat recovered from a historic recession, unemployment is low at 4 percent, and the manufacturing sector is growing again, albeit slowly. But many aspects of today’s climate echo that of the 1980s: The global economy is expanding, new technology is entering the workplace at every turn, and an opioid epidemic is ravaging communities. Automation will create new jobs, but analyses show that they’ll require digital skills that only the most educated and experienced people will have. Researchers today are predicting that white- and blue-collar workers will be in predicaments similar to their counterparts in the 1980s and ’90s, as soon as 2020, if they aren’t already there.

As a response to these challenges, many policymakers and business leaders turn to a federal policy solution that has stuck around since before Reagan’s day: job-training programs. These types of policies remain popular—at least in name, if not in investment—despite little evidence that they succeed on a large scale. In 2014, then-President Barack Obama signed the Workforce Innovation and Opportunity Act. And over the summer, Senator Robert Menendez and Representative Albio Sires, both New Jersey Democrats, introduced the Better Education and Skills Training for America’s Workforce Act.

Reagan’s version of this was the Job Training Partnership Act (JTPA) of 1982, which spent nearly $3 billion yearly between 1984 and 1998 and made some adjustments to the previous job-training formula by removing provisions for subsidized jobs in local and state governments and giving more control to the private sector. Although JTPA accounted for a relatively small share of the president’s response to unemployment, it was arguably Reagan’s most direct response to workers. And it worked about as poorly as any other program has.

Its flaw wasn’t Reagan’s approach. Rather, studies suggest it suffered from the challenges typical of federal job-training programs even before Reagan’s time. Such programs have historically been unable to change an economy in which low-wage workers suffer from both low pay and a lack of autonomy. The initiatives have struggled to achieve their mission for several reasons. One, those who need the training typically don’t know about—or are excluded from—them. Two, course material tends to be disjointed from the needs of employers. Three, and perhaps most importantly, job-training programs don’t force employers to pay skilled people decent wages. It’s worth taking a closer look at why programs like this go awry.

I spoke with experts who have studied job-training programs in the United States and abroad. They say Reagan’s JPTA epitomizes various job-training programs that have been spun up over the years in America; indeed, its track record echoes what has happened in training programs before and since. As Americans prepare for the next wave of innovation, history shows that new technologies have affected white- and blue-collar workers differently. In turn, although policymakers often talk about a single workforce, when recessions and automation strike, the fortunes of white- and blue-collar workers are likely to diverge, and training is largely an ineffective approach to the needs of the latter. History also shows that any successful response to automation's impact will have to recognize that training can never be enough to shepherd Americans into a new economy.

For the country’s white-collar workers (those who worked in office and administrative settings) automation wasn’t a job killer but rather a job changer. White-collar workers had to contend with the advent of personal computers and word-processing systems, but rarely did they lose their jobs because of the new technology. For them, the changes amounted to little more than a new chapter, as companies tended to offer some in-house training to help them keep abreast of the shifts.

In the middle of the 1990s, a decade during which office automation was rampant, employees in establishments with 50 or more workers received an average of 10.7 hours of formal training per year, according to the Bureau of Labor Statistics. Some companies took extra steps to ensure their employees weren’t left behind. The Nashville, Arkansas, accounting firm Woods & Woods is one example. Its owners traveled from IBM seminar to seminar, learning how to use the new personal computer so that they could train their staff. Meanwhile, white-collar workers whose companies didn’t offer training went back to school, learned new skills on their own, or waited for a new job that fit their needs. Workers could wait if they had employers in their social networks who could connect them to jobs or had healthy savings accounts on which they could depend.

For the country’s blue-collar workers, on the other hand, automation was a job killer, as was globalization. Blue-collar workers employed in factory or construction-type settings found themselves grappling primarily with the latter phenomenon. Some companies partnered with unions, such as the United Automobile Workers and the Communications Workers of America, to train workers. But although unions were eligible for JTPA funding, a very small percentage of it went to these programs. And regardless of the funding source, training for blue-collar workers was not as effective as was training for their white-collar counterparts: These laborers weren’t merely losing jobs because they didn’t know how to use the latest technology—they lost jobs because people in places such as Japan and China could do the work cheaper and faster.

What’s alarming today is that companies are less loyal to their workers than they were then. For example, companies in the 1970s reinvested the majority of their profits into research and development, retraining their workers, and raising salaries. But now, of the nearly 500 large companies in the S&P 500 index, half of their profits were spent on buybacks and roughly 40 percent were spent on dividends, instead of on workers. Many job vacancies are caused not by a lack of skilled workers, some labor economists argue, but rather by employers’ inability to create good working conditions for workers. What’s more, unions have lost influence—20 percent of employed workers were union members in the mid 1980s, compared to around 11 percent in 2014—and with that the power to force companies to protect workers. This reality makes it all the more important to get job-training right.

Federal policy has consistently failed at training, focusing for decades on the worker and the question of what makes that worker well-trained; it’s paid little attention to employers and the question of how they can change to better recruit and retain employees. But what if many of the 6 million jobs that are currently vacant are open not because people aren’t qualified but rather because employers aren’t hiring? What happens when a company is more interested in employing someone who’s willing to work in dreadful conditions for less than a decent wage than in someone who has computer literacy? And what’s the upshot in a society that lacks unions that would otherwise force better conditions? It’s critical that policymakers and business leaders consider these questions as they help design future job-training and workforce-development initiatives.

* * *

After studying job-training programs for 20 years, Gordon Lafer learned that they don’t work. Lafer, the author of The Job Training Charade, wasn’t always disenchanted with federal job-training programs. Back in the mid ’80s, when JTPA was first implemented, Lafer was working as an economic-policy analyst in the office of then-New York City Mayor Ed Koch. Filling job vacancies and addressing unemployment was one of the office’s top priorities. Lafer, like most of his colleagues, believed that job training was the best solution; the challenge, he assumed, was that programs needed more funding.

Lafer was convinced otherwise four years later. He recalled that Koch had been meeting with executives at sheet-metal firms, and one of the executives allegedly told the mayor something along the lines of: Even for 10 bucks an hour, you can’t get people who want to work. The sheet-metal executive didn’t say that he had open positions paying that wage, but Koch apparently took it that way and in speeches around the city told New Yorkers that $10-an-hour jobs were waiting for them. Letters from struggling single parents and the homeless soon filled the office’s mailbox. And those letter-writers all asked one question: Where are those jobs?  

It turned out that many of the open jobs actually paid below minimum wage, and Lafer drafted responses apologizing for the mayor’s error. But the envelopes never made it out of the office. You can’t possibly tell people that the mayor was wrong, his supervisor allegedly instructed him. Tell those people to ask their local job-training centers where those $10-an-hour jobs are instead. Lafer was dumbfounded.

Those job centers were funded through JTPA—and if the experience of the policy’s beneficiaries elsewhere in the country is any indication, those job centers would’ve done little to help land those struggling New Yorkers decent-paying jobs.

Several studies have explored whether Lafer’s experience was the norm. JTPA would be deemed successful by the federal government’s standards if participants got jobs, earned better wages than they did previously, and were less likely to depend on welfare relative to their time before training. It’s worth noting that a relatively low percentage of unemployed Americans participated, and most of those who did participate came from manufacturing. In the most extensive evaluation of a federal job-training program at the time, the Labor Department scrutinized enrollment info and employment outcomes from programs in 16 regions across the country representing around 20,000 Americans; it also collected additional data, such as food-stamp records, from some of the regions and consulted with vocational schools to assess the costs of their programs as a comparison, among other analyses. Another prominent study published in Cornell University’s Industrial and Labor Relations Review looked at participation in JTPA in Tennessee, comparing enrollee information to data from a monthly survey of U.S. households conducted by the U.S. Census Bureau and Bureau of Labor Statistics. A working paper published by the National Bureau of Economic Research also analyzed participation at four sites.

The studies all found that JTPA was largely unsuccessful. It had a “modest positive impact” on disadvantaged adults’ earnings, hovering around an increase of $500, according to the National JTPA analysis. The IRL Review study found that programs were “cream-skimming,” or only accepting into their programs those who were bound to get a job; those who were most in need, it found, were driven away to ensure the programs had high success rates. The NBER paper found that people self-selected into the programs, and so those who would get jobs were likely the ones who were proactive in finding training.

Andrea Chronopoulos

Harry Holzer, a fellow at the Brookings Institution, emphasized that analyses have shown mixed results; JTPA’s limited success, he argued, has been overstated. (Conducting more follow-up evaluations is one way to get a clearer picture of what exactly the programs achieve, Sires, the New Jersey representative who co-introduced a job-training bill this past summer, told me.) But other scholars disagree with Holzer, offering various reasons for JTPA’s minimal success over the years: They contend that the programs are too divorced from employers’ needs, too unrelated to workers’ interests, too light-touch, and too limited in their reach, among other flaws. And these same flaws explain why earlier federal job-training programs failed, too.

So why have federal programs, with their focus on changing the worker rather than the job, remained the solution of choice? The answer is simple: It’s much harder to force employers to change people’s wages and work conditions—especially when those companies can offshore to foreign countries and employ foreign workers—than it is to convince them to place money and responsibility in the hands of government officials.

It’s important to stress that many people affected by automation and globalization had much more positive experiences than those New Yorkers who could not find decent-paying jobs in the 1980s. Cecilia Mejia’s trajectory is one example. Having participated in 1988 in a youth program in Arizona that was funded by JTPA, Mejia is now the director of Santa Cruz County’s public fiduciary office, and she credits her career climb to the digital skills such as making spreadsheets on Excel that she learned in the youth program.

Citing examples like Mejia and the increased availability of intricate, real-time data on the needs of local economies, some policymakers are convinced that effective federal job-training is possible.

Still, as Lafer, who now works at the left-leaning Economic Policy Institute, stressed, anecdotes are not enough to inform policy. After 20 years of studies, he believes there is enough data to show that the programs (though perhaps grounded in a genuine interest in expanding social supports) cannot work on a large scale. According to Lafer, that’s because they’re rooted in a superficial understanding of unemployment—in a misguided conviction that people don’t have jobs simply because they don’t have the skills. These programs won’t work, he emphasized, until they acknowledge the possibility that job vacancies are also, perhaps primarily, caused by employers that are disloyal and inadvertently create poor working conditions for their workers. Until then, the programs will remain purely symbolic.

* * *

The story of Roberta Gainer, who was both a worker and a job-trainer, shows how ineffective these programs can be. Six days a week, Roberta Gainer commuted every morning from Niagara Falls to Buffalo, New York, to work. Starting at dawn, she’d spend nine hours daily putting rocker arms on engines, a difficult job she’d had since graduating from high school. Modern-day engines will wait for a laborer to put the pieces on, Gainer told me, but back then, workers had to run down the line chasing the engine—or worse, stop the line, which affected everyone on the floor. Automation and globalization didn’t spare her—even though she’d been trained to become a trainer for her fellow workers through a fund sponsored by her union and Ford, Chrysler, and General Motors. Gainer got laid off after two years on the job, but according to her, it was not because she was unable to complete her tasks.

The training she received was intensive. Academics, mainly from Cornell’s labor-relations school and Erie Community College, stood at the front of classrooms alongside factory-workers like Gainer teaching laborers about the economy, the history of the labor movement, and how to use and repair computers. Workers also got help with resume-writing and the job search. Gainer learned so much that she was able to train her colleagues for jobs in construction and entry-level positions in the medical and mechanical fields. “I was always nervous to teach, before every class,” Gainer said, but she managed the pressure by reminding herself that no one knew or cared if she was making a mistake. Indeed the training was as impractical as it was intensive: Workers weren’t interested in classroom training—they wanted education that was more hands-on. Given that even Gainer’s qualifications ultimately didn’t guarantee her job security, perhaps those workers were onto something.

Marc Perrone’s experience further supports conclusions that job-training programs are ineffective. His experience as a blue-collar worker was so tumultuous, it inspired him to become a leader in the labor movement. In blue-collar work, the proliferation of technology means that virtually no skill is evergreen— even newly adopted skills will become less valuable over time. Tasks that once took hours now take minutes; machines become easier to understand. Perrone was a victim of this reality and, like Gainer, ultimately couldn’t retain a full-time job despite being well-trained.

Perrone’s first job, as a butcher’s assistant, required him to work with large slabs of hanging beef in the meat departments of three different grocery stores in Pine Bluff, Arkansas. He learned how to cut the meat, package it, and price it for sale. He was eventually moved over to the retail side (those under 18 weren’t supposed to operate the sharp machinery in the meat department), where he was responsible for stocking on the shelves and ringing up grocery items. The machines didn’t do mathematical calculations at the time, so he had to add up the prices in his head; UPC barcodes didn’t exist, so he had to memorize the cost of each item. Companies like Kroger’s, the grocery stores at which he worked as a stocker and cashier, had substantial training programs to help employees adapt to new technology such as computerized cash registers, and he took advantage of those opportunities and within two years had advanced to a journeyman, the ranking just below master.

But according to Perrone, no amount of training would fully guarantee job security. It wasn’t like the new technology required that many new skills; in fact, it often “deskilled” their work, requiring less knowledge than employees needed before the digitization took over.  

In the ’80s, the U.S. was beginning to move from an economy in which many employers had strong, long-term relationships with their employees to one in which companies and workers often had shorter-term relationships. In that new universe, technology often and quickly rendered skills obsolete. Training helped workers up to a certain point, and then they were on their own—especially if their employer was more interested in productivity and shareholder value than in the workers themselves. Perrone’s belief that workers needed more benefits and supports led him to the helm of the United Food and Commercial Workers Union, where he still serves as president.

Perrone and Gainer ended up doing okay for themselves. After being laid off from her rocker-arms job, Gainer, who was displaced from the auto industry for a total of nearly five years, worked at another manufacturing plant for six months making sandpaper. She subsequently traded in her assembly-line jumpsuit for business-casual attire, landing a job as a customer-service representative at Goldome, a mortgage-bank company, where she helped people try to save their homes. General Motors eventually called her back, as was common in the auto-industry after dry spells, and she returned without hesitation. Perrone, on the other hand, eventually left his job in the grocery stores because he could not put together a full-time schedule and reentered the formal education system in his 20s. Still, it wasn’t their training that immediately protected them from broader economic phenomena such as automation and globalization. It was the seesawing tendency of the economy; it was patience; it was luck.

* * *

White-collar workers were luckier. Compared to their blue-collar counterparts, they were more buffered from the impact of the economic trends. Take the story of Nan DeMars. As a young child in Minnesota in the 1940s, DeMars spent hours in her home’s attic, which her architect father had fashioned into a small office. There was a desk and a chair. Pencils and papers rested on the desk, flanking the black Smith Corona typewriter that belonged to DeMars’ mother. This is where she practiced to become a secretary.

She took two years of shorthand courses in high school, after which she went on to get her bachelor’s degree in business from the University of Minnesota. She then started working at Ringer Corporation, a lawn and garden company. Walking into the office one morning, a huge opened box greeted DeMars. As she edged closer to her desk, she peeked inside. A word processor, with a pink ribbon tied around it, looked up back at her. “I immediately gave it a girl’s name, Elizabeth, and sat down to see what in the world it would do,” DeMars told me. It was her responsibility to learn the new technology, which was not unusual for workers facing office automation during this time. She managed, in part, because of the studying skills she developed in college. Having successfully adapted to the automation, she kept her job.

Other secretaries weren’t as fortunate. While president of the National Secretaries Association (now known as the International Association of Administrative Professionals), DeMars noticed that those who were unwilling to engage with the new computers, perhaps out of fear, did not keep up with technology and were demoted to lower-paying jobs such as file clerks. Indeed, Beverly Peterson, a secretary who worked at General Mills at the time, noted that an open mind was essential to job security in the field. Peterson’s boss encouraged her to get a degree to protect herself from the ramifications of a changing workplace. He emphasized that “credentials really mean something in this world,” said Peterson, and she “really took it to heart.” Both DeMars and Peterson were able to hold onto their jobs, in part because their college educations prepared them to learn and keep up with new technologies, but also because they’d developed bonds with their devoted employers. Those who did not attend college, or who were marginalized in other ways, namely women of color, suffered.

These lessons about blue- and white-collar workers still apply. Amid a rapidly growing and evolving economy, the country faces a similar divergence between the fortunes of the two kinds of workers. Short-term job opportunities will become even shorter-term as newly gained skills deteriorate in value more quickly than they did in the ’80s. The methods for obtaining new skills are more expensive than they were then. White-collar workers may be able to adapt, but the blue-collar folks trapped in a job landscape where all skills have an expiration date will be hit hard.

Training is undeniably crucial: When the economy settles down from the turmoil to come, workers will need it to be ready to reenter the workforce. The problem is: Federal programs have since the 1930s consistently been the default training choice for alleviating the burdens of innovation. As soon as one federal job-training program expires, another replaces it—despite history showing that large-scale efforts are all but impossible to pull off.

Related Video