Manufacturing jobs aren’t coming back. Neither are construction jobs. America’s workers need to learn some new skills to stay ahead.

While manufacturing jobs have long since departed for China and India, the U.S. economy continues to grow and even manufacture products that the world wants to buy—we export more in dollar terms than we did a decade ago. But what we’re sending (and how it’s made) is drastically different today. As Enrico Moretti documents in compelling detail in a recently released book, The New Geography of Jobs, even if we don’t assemble iPhones or sneakers in America, we supply their designs to those who do. And we do still make things—things like precision scientific instruments and jetliners. But the way we’re producing them has changed as well: Even in sectors that have expanded production over the last decade, there are fewer jobs to be had—the so-called productivity paradox. The reason? Production is increasingly automated, requiring more computers and fewer human beings.

All this adds up to an economy that generates just as much income, but with profits flowing into far fewer pockets than they did in the previous century. Moretti suggests that the prognosis for the average American worker need not be so gloomy if, as he predicts, America continues to thrive as a hub of knowledge generation and innovation. While the idea creators—those who design iPhones and develop new drugs—will continue to be the drivers of prosperity, more than a few crumbs may fall to the workers who support them. For example, Moretti estimates that Microsoft alone is responsible for adding 120,000 low-skill jobs to the Seattle area, where the company is based. This is because of the support workers required to style the hair, cut the grass, and yes, build the houses, of all those Microsoft engineers and computer scientists. And they earn more doing it—a barber in San Francisco earns about 40 percent more than his counterpart in Detroit or Riverside, Calif. So one way of boosting incomes of the bottom quintile would be to provide incentives for them to pick up and move from the rust belt to innovation hubs like Austin, San Francisco, and Boston.

Of course, if people actually start moving in significant numbers, the benefits of cutting hair or grass in Austin rather than Detroit will quickly evaporate—the price of low-income housing will be bid up, and the salaries of barbers bid down. In the longer run, the bottom 20 percent—indeed the bottom 99 percent—will need to be retrained and re-educated to get a larger share of U.S. GDP. Eminent Harvard labor economist Larry Katz sees a future where many lower-skilled workers are employed in the service sector supporting America’s innovative class. But he sees it as an open question as to whether these service jobs will be as sales clerks and lawn hands, or fashion consultants and landscape designers. Katz refers to these would-be consultants, designers, and other skilled service providers as forming the foundation of the New Artisan Economy.

If jobs are being lost to low-wage Indians and computer programs, then what today’s worker needs is a set of skills that offers the personal touch and judgment that can’t be provided by a machine or someone 12 time zones away. Katz argues that this will be crucial for those with only high school educations, who will need to learn a “high touch” trade—like personal trainers, kitchen designers, and home health aides—where personal interaction is critical. He makes a similar argument for the college educated as well: With many clerical and lower-level management jobs made obsolete by advances in information technology or lost to off-shoring, they’ll have to reinvent themselves as, say, IT support professionals or consultants. (In making the argument that college graduates will also need to be retrained for the job market of the future, Katz points out that middle-income earners have gotten hammered the hardest in the past decade—also part of a longer trend going back decades—particularly in IT-intensive sectors.) Katz’s hope for the new economy is a workforce whose skills make their services sufficiently desirable to Moretti’s idea-creators that the bottom 99 percent do better than single-digit hourly wages in the job market.

An artisan economy can’t be built overnight. We’ve spent the last decade funneling too many workers into construction jobs that may never come back. These workers now lack the skills required in Katz’s economy of the future. And perhaps the most depressing statistic that Hurst points to in describing the plight of low-skill Americans is that, after falling steadily for 15 years, the fraction of men who stopped their educations by the end of high school went up by a few percent between 1997 and 2006, before resuming its decline. Why? Presumably more school looks less attractive to an 18-year-old if he can get a decent job doing construction. Not exactly a lost generation, but these are yet more young males who will need retraining to get decent jobs or even stay in the workforce.

Is America up to the task of retraining its workforce to be artisans rather than burger flippers? In recent congressional testimony, Katz is critical of current government training schemes like those funded under the Workforce Investment Act. He calls them “fragmented and difficult for many workers to navigate.”

At the same time, professor Katz maintains a glimmer of hope. As he observed in his testimony, there is emerging evidence that job retraining can be effective in teaching older workers new skills when done right. Amid the many failures, Katz points to some exceptional retraining programs that have demonstrated promise in helping workers to cost-effectively upgrade their earnings. For example, Per Scholas, a 15-week program in New York that provides training for installing and repairing computer networks, increased participants’ annual incomes by nearly $5,000 within two years of beginning their training. Programs like Per Scholas produce these large effects through a combination of in-class instruction and on-the-job training, often with a post-training middleman to help with placement in a job where their skills are well-utilized. Katz calls for evidence-based funding that rewards programs that do well by their clients.

Realistically, it’s going to be hard to transform an illiterate and innumerate burger flipper into an IT support specialist overnight—Per Scholas, for example, will only take applicants with a high school diploma or GED who also test at the 10^th grade level or higher in math and English. So Katz also sees improving basic education—upgrading school quality and graduation rates, and channeling more graduates into post-secondary training—as essential to building a new artisan economy.

Whether a gridlocked and partisan government can come together to develop a sensible agenda for skills development and job creation is an open question. But the future of the American worker depends on it.