Looking ahead 25 years is risky. Neither economists nor journalists are good at it. Science fiction writers are better, though only with hindsight is it possible to tell which ones were accurate. The Bureau of Labor Statistics makes long-run occupation-by-occupation projections every couple of years. That’s brave. But I’m more impressed that the Monthly Labor Review publishes an evaluation of the projections a decade or so later. From one of those evaluations, we know that in 1997 BLS predicted a decline of 5,580 (9.2 percent) in the number of crossing guards over the ensuing decade whereas, in fact, their number grew by 8,777 (14.5 percent).1
Mindful that someone will mock any prediction I make about 2040, I won’t make any. I know what I don’t know.2 Instead, I’ll list a few questions preoccupying us today that we will be able to answer fully only with the passage of time.
How many human workers were displaced by robots and artificially intelligent computers?
Nineteenth-century farmers fretted that mechanization would leave their children without livelihoods. Factory workers have feared automation for more than half a century.3 In 1964, big-name, left-leaning academics warned of mass unemployment because “potentially unlimited output can be achieved by systems of machines which will require little cooperation from human beings.” The doomsayers were wrong. The factory replaced the farm. Service jobs once unimagined replaced the assembly line. New wants led to new jobs.
Today, those warnings are heard again. Machines long ago replaced muscle; now computers seem to be replacing the human brain: the driverless car, face-recognition software, computers that learn. My initial reaction is to be reassured by history. But maybe this time is different.
Mass unemployment seems unlikely, although the fact that in the United States 15 percent of prime-age men—those between 25 and 54—were not working in the spring of 2015 is alarming. But technology changes faster than humans can, and the open question is whether the wages of those with middle-skill, middle-wage, once-middle-class jobs will continue to sag.
Did the gap between winners and losers in the labor market continue to widen?
For the past three decades, there has been an inexorable increase in almost every measure of inequality. Educated workers earn a lot more than the less educated. The top 1 percent have pulled away from the rest of us. It hasn’t always been this way: the decades after World War II saw shrinking wage inequality.4 I was among those who predicted that this trend would be disrupted as the lure of higher wages increased the supply of college graduates and decreased the relative wage premium; I was wrong.5
One can easily catalog the forces propelling inequality. Unions are weaker. Globalization has strengthened the hand of capital (which can move abroad) versus labor (which can’t move so easily). Technology and globalization amplify the benefits of being the best or the most popular CEO or lawyer or football player or novelist, the winner-take-all phenomenon. Social mores have changed: it is more acceptable for those at the top (of law firms, of universities, of baseball teams) to take a larger share. Greater wealth at the top today means greater income to the already well off in the future. Not so easy to forecast is which of these force will persist and which will be restrained by governments.
Were we stuck in an era of slow growth, of “secular stagnation”?
The years after the Great Recession saw disappointing growth and investment despite interest rates so low that some borrowed at negative rates. Slow growth in labor forces, persistently low productivity growth, a surplus of savings, and a dearth of attractive investment opportunities, the gloomy forecasters argued, would yield “secular stagnation” unless fiscal tools were deployed more aggressively than governments were willing or able to do.6
Some argued that everything worth inventing had been invented and we were doomed to marginal improvements, an argument at odds with fears that computers were becoming smart enough to challenge livelihoods of the white-collar elite.7 Others argued that technology allowed companies to create enormous value with very few workers and very little investment, portending a persistent savings glut.8 Still others saw the savings glut as a temporary condition that would ease as the economy continued to return to normal.9
Income data and headlines offer contradictory data. Unfortunately, getting the diagnosis correct is essential to prescribing the right policies. Policymakers can’t wait until 2040, and the rest of us won’t know until then, but that’s what it’ll take to truly resolve the debate.
I look forward to readers’ answers to these questions in the 125th anniversary edition of the Monthly Labor Review.
United States Department of Labor
