NAE Logo
In This Issue
Engineering, Technology, and the Future of Work
Autonomous Vehicles Implications for Employment Demand
Creating New Value with Performance-Based Industrial Systems Design and Operations Management An Engineering Opportunity
Education Transformation to Support Lifelong Learning
Beyond Schooling Educating for the Unknowable Future
Download PDF
Fall Bridge Issue on Engineering, Technology, and the Future of Work
September 15, 2015 Volume 45 Issue 3

Beyond Schooling Educating for the Unknowable Future

Tuesday, September 15, 2015

Author: John A. Alic

One day she learned something and the next day was left to do it alone. . . . Out here she saw it all the time, . . . nurses and soldiers and doctors and drivers and engineers alike, no one did what they’d been trained to do, they did what needed doing. She learned more in six weeks than she had in six months of concerted practice at home.

– Barrett (2013, p. 210)

A century ago, the majority of young people left US schools after seven or eight years, having learned enough, in the eyes of most, to exercise the rights of citizenship and earn an acceptable living. Subsequent decades brought world wars, the 1930s depression, and a great deal of technological and economic change. During three decades or so following World War II robust productivity growth fed steady increases in wages and benefits—a period my coauthors and I called the Wonder Years (Herzenberg et al. 1998). It did not last. Since the 1970s pay has stagnated for many millions of workers, with widening gaps between those in the topmost tiers of the distribution—mostly white male college graduates—and the rest (figure 1).

  Figure 1

Debate about the causes has focused mostly on education and in so doing misses at least half of the problem and the possible solution set. In some accounts it can seem as if an individual’s working life is determined at the point of school-leaving, her prospects and earnings following a ballistic trajectory (with leaving college short of a degree amounting to something like failure to launch). In fact, many of the actual skills used in actual workplaces are only loosely connected with formal education, and the many ways in which people learn outside of school get far too little attention.

There are two main reasons why formal education gets so much attention and informal learning so little. First, many skills are hard to measure or even to observe, so that employers hire in large part based on educational credentials: in the absence of better indicators, education serves as a signal. Second, as Peter Capelli (2014) has argued, even as employers complain that workers lack needed skills, they have pulled back from support for education and training, pushing more of the costs onto schools and society at large. In so doing, businesses have left educational institutions—secondary schools, community colleges, universities—no real choice but to try to prepare young people not just to enter the labor market but to enter with employer-ready skills.

It is wrong to expect schools to produce “job-ready” workers. Indeed, attempts to do so aggravate the underlying problem, which is one of adaptation to the always unknowable future. No one can know what sort of skills young people in school today might need 30 or 50 years from now. New systems—new institutional structures—are needed, designed to support ongoing learning for everyone.

Trends

There is no debating the evidence of wage stagnation and wage dispersion, only the causes. During the 1970s productivity growth slowed and, as figure 1 shows, wage growth for those in the lower tiers of the distribution stopped. Benefits also declined (CEA 2015, pp. 149–151).

Some observers pointed to the productivity slowdown itself, arguably associated with lagging innovation, as a cause. Others blamed globalization and imports from low-wage countries. Neither explanation persuades: there are too many confounding variables and too much left out of account. For example, also beginning in the 1970s a growing fraction of the rewards from increasing productivity flowed to capital (e.g., business profits) (Fleck et al. 2011). For decades, compensation—wages plus benefits—had risen in lockstep with productivity, the labor share of output remaining roughly constant at around 70 percent. The labor share has now fallen to about 60 percent, a decline that would appear still more dramatic if the compensation of earners at the very top (say, the 99th percentile) were left aside.

There is more to understand than the widening gaps between deciles. Economists analyze such trends in terms of human capital—the knowledge, skills, and competencies that employees bring to the workplace. Unlike physical capital—blast furnaces, computer systems, and so on—human capital cannot be measured directly, only inferred. Census data on educational attainment—years of schooling—have been the usual proxy, sometimes supplemented by years of experience. Here too things have changed, as figure 2 shows, with widening gaps in pay separating those with more schooling from those with less—another shift that demands explanation, and one that underlies much of the attention focused recently on education.

  Figure 2

The retreat of business firms from efforts to enhance the human capital on which their own operations depend—through apprenticeships, internal labor markets providing training and well-marked career ladders, and workforce preparation more generally—is more difficult to document quantitatively. But it is consistent with other evidence of growing separation between the interests of capital and those of society as a whole, such as the much criticized tendency of businesses to underinvest for the long term in favor of immediate profits.

Explanations

Partial explanations for the patterns shown in figures 1 and 2 abound. The manufacturing share of US employment peaked in the mid-1950s at nearly 30 percent. A source of stable, well-paying positions for those with a high school education or less, unionized industries such as steel and automobiles provided pathways to the middle class and collective bargaining agreements pushed up wages elsewhere in the economy (Western and Rosenfeld 2011). But manufacturing employment now stands at just over 11 percent, and union coverage has declined in all sectors.

Declining Wages and Benefits

With high wages taking part of the blame for the competitive difficulties of US manufacturing, employers resurrected earlier antiunion strategies such as “right to work” laws (Griffin et al. 1986). Textile and apparel firms had already moved from New England southward. From the 1960s on, the steel industry restructured on the basis of nonunion minimills, many of them also in the South. When foreign-owned automakers began to invest in the United States they, too, often chose border and southern states. Two-tier wage structures now leave many people in auto and auto parts plants doing the same work for less pay and fewer benefits (Welch 2015).

Long common in service industries, contingent employment, a term that covers both part-time and temporary employees with no expectation of a permanent job, has now spread to manufacturing (Dey et al. 2012). Contingent workers usually earn less on an hourly basis than those with “regular” jobs in the same firm; they also lose out on benefits such as cost-shared retirement and health insurance plans, even sick days. Although some people prefer to work part time, for others it is the only job they can get. While daily and seasonal sales fluctuations in industries such as retailing help justify part-time and temporary arrangements, a growing number of employers have come to view contingent workers simply as another way to hold down costs (Stone 2015).

With freelance work expanding, facilitated by digital ubiquity, it may soon be appropriate to think in terms not just of contingent employment but of irregular employment. To the extent that growing numbers of Americans sell specialized skills, abilities, and assets over the Internet and in some sort of “sharing economy,” this type of work, while it may benefit many people, could also come to seem a high-technology version of the casual employment associated with poor countries, leaving behind those unable to carve out a secure niche.

Human Capital

Figures 1 and 2 cover all US employment—nearly 160 million people working in jobs the Labor Department classifies into some 1100 occupations. The labor market is big, and the dynamics messy. Economists, perhaps needless to say, approach these dynamics in terms of supply and demand: When pay rises disproportionately in the upper tiers of the wage distribution and for workers with more education (i.e., more human capital), this must reflect in some way a rise in demand relative to supply.

But what is the cause of this rise? Early explanations cited technical change, notably the spread of digital systems (see, e.g., Bresnahan et al. 2002). Auto mechanics, after all, now diagnose faults in microprocessor-controlled powertrains, Wall Street traders make their millions with the help of algorithms developed by PhDs, and farmers will soon rely on drones to monitor crops. The so-called knowledge economy seemed to raise wage premia for workers able to learn and adapt, and many analysts went on to assume that education could serve as a proxy for these abilities.

Recent studies go somewhat beyond simple human capital models based on educational attainment. The OECD’s Programme for the International Assessment of Adult Competencies (PIAAC), notably, assesses literacy, numeracy, and computer-based problem solving (Paccagnella 2015; also see Autor and Handel 2013). Yet these too are crude measures; job skills and ability are hard even to observe. Who can tell whether their child’s fourth grade teacher, or a local auto mechanic, or the family physician is above or below average? Any careful look at workplace practices—whether those of warehouse employees who plan in their heads the shortest routes through labyrinthine settings or the vastly different skills of creative engineers and scientists—will show the differences between what people learn in school and what they do in their jobs (Alic 2008).

At the same time, it seems plain that wages rise for workers with more years of schooling not so much because the jobs they fill require particular skills (excepting more specialized occupations such as nursing or engineering or professional sports) but because employers act as if educational attainment signals potential (Spence 2002). Employers, in other words, weigh job candidates on the assumption (perhaps implicit) that those with more education will be safer bets, likely to have “basic employability skills (attendance, timeliness, etc.) and the ability to work well in a team environment” (Deloitte and the Manufacturing Institute 2015, p. 6). Another survey states, quite bluntly, “employers have frequently come to rely on a bachelor’s degree as an employment screen, even if it may not be related to actual job duties” (Accenture et al. 2014, p. 18). The signaling function of educational credentials extends from low-wage jobs in, say, the front offices of retail banks to highly coveted positions in prestigious law firms, investment banks, and consulting firms (Rivera 2012).

Technological Change, Deregulation, and Globalization

A look at three broad forces affecting earnings—technological change, deregulation, and globalization—gives further insight into the trends illustrated in figures 1 and 2. Taken together, these forces spur economic dynamism, boosting productivity and creating much new wealth and many new jobs as businesses expand and hire more people. At the same time, less efficient older firms, if unable to adapt, decline and some die off.

The direct impacts of technological change on workers have probably been greatest in offices. During the first half of the 20th century US high school enrollments exploded in response to demand for clerks, typists, bookkeepers, secretaries, and administrative employees. Some of these jobs began to vanish in the 1960s as stand-alone computing systems took over tasks such as payroll processing, and the pace picked up in the 1980s with PCs. Digital systems now do work once performed by millions of modestly educated office workers (Alic 2004)—with more displacement to come, as suggested by rapid advances in powerful computing systems implementing forms of artificial intelligence and apps for everything.

Technological change also forced deregulation of industries such as telecommunications, and helped open the way for competition from low-wage countries, not just in goods-producing industries such as steel, electronics, and automobiles but in many services as well. With low-cost international voice and data transmission, call center personnel in the Philippines or India could compete with those in Des Moines or Sioux Falls, and some better-paid work such as software coding and radiological interpretation has also migrated overseas, again with more to follow.

While the ramifications of technological change include job creation associated with expanding industries such as mobile telephony, other opportunities have disappeared. Before deregulation and the waves of disruptive innovation that have swept through the economy, an entry-level position might open the way to career progression along well-marked job ladders in the internal labor market of a bank, downtown department store, or telephone company, with on-the-job learning and advancement leading to a position as loan officer, buyer, or maintenance supervisor. Few such career paths remain and more people must acquire new skills on their own, outside the workplace.

Whose Interests?

In one way or another, all wealthy countries seek a balance of some sort between the positives and negatives of Schumpeterian creative destruction. In this respect, as in many others, the United States stands out as exceptional, arriving at what can only be termed an imbalance, with labor market policies that cater unabashedly to employer interests at the expense of employees (Estlund 2002; Robertson 1988).

Each year tens of millions of Americans enter or exit the workforce or move from one job to another; in 2014 employers hired some 59 million Americans, about 3 million more than the number who quit a job or were laid off or fired (BLS 2015). Labor market churn both drives competition and innovation and is a consequence of these forces. Engineers, scientists, and man-agers leave their jobs and take their know-how and ideas to other companies or perhaps start their own. Other workers, though, are threatened. People with unimpressive educational credentials may have an abundance of skills that employers seek but no good way to signal them. For similar reasons, experienced and productive employees with generic skills and less education may be laid off first when organizations downsize, and if they can find a new job will likely earn less than before (Couch and Placzek 2010).

Neither public nor private sectors in the United States invest much in active labor market measures such as training and job placement assistance to help people prepare for new work (as opposed to passive measures such as unemployment insurance) (Martin 2000; Nie and Struby 2011). With some exceptions for fast-tracked technical and managerial employees, employers do relatively little to train their workers. Only about 20 percent of employees surveyed by Accenture (2011), more than half of them with four-year or graduate degrees, reported gaining skills through employer-provided training over the five-year period 2006–2011. Other studies find that both on-the-job and employer-sponsored off-the-job training have declined since the mid-1990s (e.g., CEA 2015, p. 147).

Remedies

Limits of Formal Education

The near universal prescription for the problems summarized above has become more and better education for all. A simple thought experiment suggests the fallacy.

Suppose that everyone in the labor force had a four-year college degree. What, exactly, would change? Would employers raise wages? Alter their hiring and promotion practices? Not likely. They would still respond to educational credentials as a signal. Overeducation and underemployment, already common, would rise. The economy might adjust to take advantage of a workforce with higher average levels of education, but no one can say what this might mean for wages.

The starting point in seeking remedies is to accept that formal education is no panacea. All learning is in some sense experiential, something everybody at some level understands (think about how much children learn before they enter kindergarten).

Beyond basic skills—reading, writing, arithmetic—and some capacity to learn, employers place considerable value, nearly regardless of the position, on ability to work with others, communicate effectively, accept authority, and follow directions. Some jobs, of course, demand specialized skills. As Google chairman Eric Schmidt put it, “You cannot innovate and build new products without engineers in your field” (quoted in Huey et al. 2013, p. 80). Firms in such industries, well aware that a handful of highly creative individuals or even a single extraordinary “talent” can create a billion-dollar revenue stream, arguably seek a larger STEM (science, technology, engineering, and mathematics) pipeline to maximize the chances of finding superstars (Lev-Ram 2015).

Yet it should be clear that correlations between what people learn in school and actual skills used in actual workplaces can be elusive, even for occupations such as engineering. Anyone who has earned an undergraduate degree and then taken an entry-level engineering job will recall, as I do, how much there was to learn in the first few weeks and months—and in all the years that followed.

Even in the 21st century, for example, large numbers of people find their way into STEM occupations without academic credentials in STEM fields. Recent US Census surveys continue to show that many Americans working in occupations classed as science and engineering hold degrees in non-STEM fields and about one quarter of workers in these occupations have no four-year degree at all (NSB 2014, p. 3-15).

Entry into engineering and related technical fields has always been a source of strength for the US economy, the world’s most innovative. More than twice as many Americans have found jobs associated (sometimes loosely) with computer and information technology than have graduated from four-year programs since colleges and universities first began to offer computer-related coursework in the 1960s (Alic 2010). They learned in the workplace, from their own experience and that of their colleagues; from handbooks, manuals, and other codified literature; from formal and informal instruction provided by employers, hardware and software vendors, consultants, and trade and professional societies; and from short courses offered by colleges and universities. This sort of learning is essential and indeed inevitable; it should be encouraged and supported.

New Approaches to Skill Acquisition

A good deal is known, in part based on advances in cognitive science and related disciplines, about how people develop competence and expertise and about how the sort of learning that takes place after people have left school could be enhanced (see, e.g., Bransford et al. 2000). As yet, little of this has found its way into the larger debate.

A dozen years ago when I argued for new approaches to skill acquisition, a common response was something like, “Well, the Internet will take care of it.” Now I think people recognize the pitfalls that come with easy access to huge amounts of undigested information, some of it wrong or contradictory or simply of unknown reliability. Improvements have been coming, albeit in pieces—MOOCs (massive open online courses), online encyclopedias and handbooks that can be updated constantly, technology-focused chat rooms (e.g., for trading work-related problem-solving lessons), and ventures into personalized, computer-based learning.

These and other new means of facilitating skill acquisition coexist with the heterogeneous offerings of colleges and universities, trade associations, and professional societies directed at those already in the workforce. So far, little is known about the effectiveness of such offerings, with the exception of continuing medical education, which has been extensively evaluated because it is a common (and costly) requirement for board certification. The evaluations have shown it to be almost totally ineffective in changing what physicians do in everyday practice (Grimshaw et al. 2001) (perhaps because, unlike a number of other high-skill occupations, weak market forces—namely the inability of typical patients to judge physician competence—reduce incentives for learning once medical school has been left behind). As this example suggests, experimentation and evaluation will be necessary to fill the vacuum that now exists between formal education and unmediated experiential learning, which, again as most people at some level understand, tends to be painfully slow and error-prone.

On a societywide basis, flexible institutional structures should be designed for everyone, not just those who are motivated and capable of self-directed study—they should support, in effect, “just in time” learning. Without more structure and guidance too many people will flounder—a recipe for still more inequality. Feedback and adaptability—self-management and self--correction—should be watchwords.

Conclusion

The personal experiences of many Bridge readers can serve as representations of a major task that American institutions concerned with education and training face as the 21st century unfolds: learning, in all its facets, to deal more effectively with a radically uncertain future. Adaptation has lagged because the steamroller force of post–World War II technological and structural changes has been apparent mostly in hindsight. Social learning will be essential and many of the lessons of the past will have to be modified or discarded.

In this light, the nation’s institutions for continuing education and training can only be seen as inadequate and anachronistic. The current nonsystem leaves the burden of managing learning—navigating through a vast maze of possible opportunities and potential cul-de-sacs—to individuals. Once they leave school, support structures fall away, and so, for many, do prospects for satisfying careers.

Responses to current and future skill gaps require greater attention to learning that takes place outside of school and after labor force entry. Networks to provide such learning will have to address the varying needs of an expanding labor force that already numbers nearly 160 million people. Demand for skills will change over time, perhaps quite dramatically—as illustrated by the past half-century of digitization. Flexibility will be essential, with multiple delivery modes that are at once tailored to individual learning styles and skill sets and designed to exploit advances in cognitive science and artificial intelligence. Incentives will have to be put in place to encourage both employees and employers to participate.

Whatever directions this evolving learning structure or system takes, it cannot be defined by what employers want. Beyond the simple truth that employers’ interests do not align with those of individuals or society at large, businesses (and most other employers) emphasize today’s jobs, knowledge, and skills, not tomorrow’s. Employers should have a voice, just not a determining one.

References

Accenture. 2011. Accenture Skills Gap Study. New York.

Accenture, Burning Glass, and Harvard Business School. 2014. Bridge the Gap: Rebuilding America’s Middle Skills.

Alic JA. 2004. Technology and labor in the new US economy. Technology in Society 26(2/3):327–341.

Alic JA. 2008. Technical knowledge and experiential learning: What people know and can do. Technology Analysis and Strategic Management 20(4):427–442.

Alic JA. 2010. Everyone an innovator. In: State of Innovation: The US Government’s Role in Technology Development, pp. 236–260. Block F, Keller MR, eds. Boulder, CO: Paradigm.

Autor DH, Handel MJ. 2013. Putting tasks to the test: Human capital, job tasks, and wages. Journal of Labor Economics 31(2):S59–S96.

Barrett A. 2013. Archangel. New York: Norton.

BLS [Bureau of Labor Statistics]. 2015. Job openings and labor turnover: February 2015. USDL-15-0562. Washington: Department of Labor.

Bransford JD, Brown AL, Cocking RR, eds. 2000. How People Learn: Brain, Mind, Experience, and School, expanded edition. Washington: National Academy Press.

Bresnahan TF, Brynjolfsson E, Hitt LM. 2002. Information technology, workplace organization, and the demand for skilled labor: Firm-level evidence. Quarterly Journal of Economics 117(1):339–376.

Capelli P. 2014. Skill gaps, skill shortages and skill mismatches: Evidence for the US. Working Paper 20382. Cambridge, MA: National Bureau of Economic Research.

CEA [Council of Economic Advisors]. 2015. Economic Report of the President 2015. Washington.

Couch KA, Placzek DW. 2010. Earnings losses of displaced workers revisited. American Economic Review 100(1):572–589.

Dey M, Houseman SE, Polivka AE. 2012. Manufacturers’ outsourcing to staffing services. ILR Review 65(3):533–559.

Estlund CL. 2002. The ossification of American labor law. Columbia Law Review 102:1527–1612.

EPI [Economic Policy Institute]. 2012. The State of Working America, 12th ed. Ithaca, NY: Cornell University Press.

EPI. 2015. 2014 Continues 35-Year Trend of Broad-Based Wage Stagnation. Issue Brief #393. Washington.

Fleck S, Glaser J, Sprague S. 2011. The compensation--productivity gap: A visual essay. Monthly Labor Review (January):57–69.

Griffin LJ, Wallace ME, Rubin BA. 1986. Capitalist resistance to the organization of labor before the New Deal: Why? How? Success? American Sociological Review 51(2):147–167.

Grimshaw JM, Shirran L, Thomas R, Mowatt G, Fraser C, Bero L, Grill R, Harvey E, Oxman A, O’Brien MA. 2001. Changing provider behavior: An overview of systematic reviews of interventions. Medical Care 39 (8, Supplement 2): II-2–II-45.

Herzenberg SA, Alic JA, Wial H. 1998. New Rules for a New Economy: Employment and Opportunity in Postindustrial America. Ithaca, NY: Cornell University Press.

Huey J, Nisenholtz M, Sagan P. 2013. Brainstorm tech: The new news business. Fortune, July 22.

Lev-Ram M. 2015. The business case for STEM education. Fortune, January 22.

Martin JP. 2000. What works among active labour market policies: Evidence from OECD countries’ experiences. OECD Economic Studies 30(2000/I):79–113.

Nie J, Struby E. 2011. Would active labor market policies help combat high US unemployment? Federal Reserve Bank of Kansas City Economic Review 96(3):35–69.

NSB [National Science Board]. 2014. Science & Engineering Indicators 2014. NSB-14-01. Washington.

Paccagnella M. 2015. Skills and Wage Inequality: Evidence from PIAAC. OECD Education Working Papers, No. 114. Paris: Organization for Economic Cooperation and Development.

Rivera LA. 2012. Hiring as cultural matching: The case of elite professional service firms. American Sociological Review 77(6):999–1022.

Robertson DB. 1988. Governing and jobs: America’s business-centered labor market policy. Polity 20(3):426–456.

Spence M. 2002. Signaling in retrospect and the informational structure of markets. American Economic Review 92(3):434–459.

Stone KVW. 2015. Green shoots in the labor market: A cornucopia of social experiments. Comparative Labor Law and Policy Journal 36(2):293–321.

Welch D. 2015. Can you spot the difference? Bloomberg Businessweek, March 9–March 15:18–19.

Western B, Rosenfeld J. 2011. Unions, norms, and the rise of US wage inequality. American Sociological Review 76(4):513–537.

About the Author:John Alic is a consultant on technology policy.