A: Productivity is a measure of how much income (or output) is generated in an average hour of work in the U.S. economy. Over the long-run, the only way economies can generate higher standards of living, at least on average, is if productivity grows. In the near-term, productivity puts a limit on how fast wages can rise without putting so much upward pressure on prices that we start seeing price inflation rise above the Fed’s comfort zone. At the moment, price inflation is still below the Fed’s 2 percent target and has been for a long time now. But wage growth has picked up a bit over the past year. Whether it’s running fast enough to push price growth above 2 percent or not crucially depends on whether the underlying rate of productivity growth is 0.5 percent or 1.5 percent. In recent years, it’s been close to 0.5 percent, but my paper argues that this should be seen as temporary so long as we continue to try to push the economy hard toward full employment.
Q: You find that slower growth in investment in productive capital (e.g., machines, structures, R&D) is a key factor in the productivity slowdown. Can you explain how that works? And given such low borrowing costs in recent years, wouldn’t we expect a lot more investment?
A: Productivity grows for essentially three reasons: workers get better-educated and experienced, technological advance leads to new ways of producing things, or workers are given more and better machines and equipment to do their jobs. Investment — spending more on plant and equipment — has grown very slowly in this recovery relative to previous ones. This has radically decreased the contribution that such investment usually makes to productivity growth.
This is pretty mechanical — invest in less capital and capital contributes less to productivity growth. Less mechanical but likely at-play is slow growth in investment in research and development accompanied by slow growth in technological advances. It makes sense that if firms are deciding to not invest much in physical capital that they’re probably also pulling back on research. And, it also makes sense that spending less on R&D could slow the pace of technological advance.
Q: I’d have to say that my favorite sentence in your paper is: “A ‘high-pressure economy’ that eliminates the remaining demand shortfall in the U.S. economy and leads to low rates of unemployment and rapid wage growth would likely induce faster productivity growth.” Unpack that for our readers.
A: One spur to firms to invest in more capital goods is more expensive workers. Capital investment is often undertaken to partially replace workers. And workers are more valuable to replace when they’re expensive. So, if we managed to drive unemployment low enough (create a “high-pressure” labor market) to spark some rapid wage-growth as workers now have real bargaining power vis-à-vis employers scrambling to recruit and retain employees, then firms might also realize that investing in labor-saving capital is a good deal.
Over the last seven years, there has been no wage pressure like this — workers have been cheap and have mostly stayed cheap (again, the last year has seen some wage pickup). This has likely reduced the urgency to invest in productivity-enhancing capital. Often, the threat of investing in labor-saving technology is used as a cudgel against efforts to demand higher wages (like when the first Trump labor secretary nominee and fast-food chief executive Andy Puzder argued that boosting minimum wages would just have workers replaced with kiosks in his restaurants). But from a macroeconomic perspective this isn’t a threat, this is just another benefit of the higher wages!
Q: Hold up, dude! It sounds like you’re advocating for robots taking jobs. That’s got a lot of people spooked these days. Is that really what you’re suggesting?
A: Actually, yes! We need more robots in the short-run, not fewer. Yes, particular workers displaced by the automation will have to find new jobs, but in a high-pressure economy, they’ll be able to. And just to be really clear, anybody who is genuinely worried about the effect of automation on jobs should be screaming their head off against interest rate increases in the near-term. The point of these increases is to slow the pace the job-growth. If you think robots are taking jobs and leading to a labor glut, you really need to be on our side in arguing to run the economy even hotter.
Q: Figure J in your paper shows a pretty tight correlation between lagged wage growth and investment growth. Why should wage growth yesterday lead to investment growth today?
A: The process described above where wage-pressure spurs firms to spend money on labor-saving technology likely comes with a lag. How do firms know that it’d be good to have more machines tomorrow to economize on labor costs? They know this when labor costs are rising rapidly today.
Also, causality between wage-growth and productivity growth is clearly two-way; in the near-term when the economy has slack, faster wage growth leads to more investment. But in the long-run, to the degree that more investment leads to higher productivity, this can lead to higher potential wage-gains. This figure isolates the causality a bit, since it would be slightly odd for investment tomorrow to affect wage-growth today.
Q: I take it you’d like Chair Janet L. Yellen of the Federal Reserve to read this paper and reconsider tapping the interest rate brakes, right?
A: Yeah. It’s a bit analogous to the “missing workers” issue. When labor-force participation (even of non-retirement age people) tanked early in the recovery, policymakers had a choice: write this off as a permanent change, or continue to try to boost economic growth and hope some of these potential workers came back in response to a better labor market. The last year has vindicated the strategy of keeping monetary policy accommodative — participation has greatly firmed up. They should apply the same logic to productivity — experiment aggressively with trying to fix some of the damage done by running the economy too cold for a long time with a period of running the economy hot. Given the stakes, it’s definitely worth trying.
Q: EPI was the first to document the now well-known split between productivity growth and middle-class wages and incomes. While we’re all for faster productivity growth, don’t we have to worry that it won’t reach most working families?
A: For sure. Productivity growth is a necessary but not sufficient condition for boosting wages across the board. In the case of spurring this growth by running a high-pressure economy, however, you really are killing two birds with one stone: you’re boosting productivity and boosting the bargaining power of those low- and moderate-wage workers who really need low rates of unemployment to get any traction in looking for raises.