What’s the link between robots, artificial intelligence and tax reform? We’ve been debating whether new technologies can ignite a productivity resurgence or whether tech has lost its potency; whether increased productivity will benefit workers or eliminate jobs altogether.
Understanding these relationships can help show why tax reform might boost all three — technology, productivity, and pay.
One of the most serious anti-tax reform claims is that it won’t help the average worker. Investment, productivity and growth, this argument says, are accruing mostly to the fortunate few. So, even if we could boost those top-line metrics, we may not be doing much for the typical American.
The Economic Policy Institute helped launch this debate back in 2012. In a blockbuster chart, EPI showed that for much of history, productivity and wages rose together, almost one to one, until the early 1970s.
Then, in a clean break with the past, U.S. productivity soared over the next 40 years, while wages hardly rose at all. It looked as if something had changed to rig the system against workers.
Several smart analysts, however, showed the chart was highly misleading. If you corrected the price indexes, looked not at wages but total compensation (including benefits like health insurance), and compared mean productivity with mean (not median) pay, the chart’s massive productivity-pay gap dramatically shrunk, though not to zero. Productivity and pay were still linked, although perhaps less tightly than before.
In a new, preliminary presentation, Harvard’s Larry Summers and Anna Stansbury find much the same. Holding other factors equal, “If productivity accelerates, the likely impact will be increased pay growth for middle income workers,” they concluded.
The pay-gap pessimists are skeptical. The median worker hasn’t participated equally in recent years, so why should we believe faster growth will help going forward?
Summers and Stansbury say that if technology and innovation were the central factor in the remaining mean-median pay gap, they would have expected the mean-median gap to widen during productivity booms. But there is “little co-movement between productivity growth and widening inequality.”
Productivity growth has been very slow over the last dozen years, so how can productivity explain the growing pay gap? “This tends to imply,” they concluded, “that technology is not the key driver of changes in the labor share, or in the mean/median compensation gap.”
Looking at overall productivity growth, however, may miss a central point: the gigantic innovation differentials between firms and industries. For example, in a recent paper, Michael Mandel and I found a yawning productivity gap between firms that use lots of information technology and those that don’t.
Productivity in the “digital industries” grew at a robust 2.7-percent annual rate over the last 15 years. But productivity growth in the “physical industries” plummeted, to just 0.7 percent per year.
We think a significant portion of this differential can be explained by the industries’ huge differential investments in — and creative use of — information technology. The physical industries account for around 70 percent of U.S. output and 74 percent of employment but make just 30 percent of the investments in infotech.
This “information gap” gap may help explain the large pay divergence. At the end of 2016, average total compensation for the 90.5 million workers in the physical industries was $55,600. But total compensation for the 32.6 million workers in the digital industries averaged $92,000.
Digital employees thus account for 26 percent of the private sector workforce but earn 37 percent of total pay. It’s not just because of a few high-paid executives in Silicon Valley and Wall Street.
In separate work, Mandel found that wage premiums for mid-skill occupations in digital versus physical industries range from 15 percent to 38 percent. For example, sales representatives in digital firms earn 15-percent more than non-digital sales reps.
Workers in advanced distribution centers, such as Amazon’s high-tech warehouses, earn 31-percent more than retail workers in the same geographic area. This tends to support the conclusions that productivity and pay are still linked, and technology differentials are a big part of the explanation.
So how does all this fit into tax reform? The productivity and wage slowdowns have been concentrated in the physical industries, which have also suffered an investment drought.
Where many tech startups could rely on deep venture capital markets for growth financing and Silicon Valley giants for exit opportunities, physical economy firms enjoy no such advantages.
Bank lending, public equity and initial public offering markets have been weak or nonexistent, especially for lower-margin businesses in the physical industries. As foreign nations dramatically reduced their corporate tax rates, the 35-percent U.S. rate soared on a relative basis and became yet another barrier to marginal investments in America.
Between the fourth quarter of 2000 and the first quarter of 2008 (a period of 7.25 years), nonresidential fixed investment in the U.S. rose 19.5 percent. But over the next 9.5 years, through the end of 2016, investment rose just 11.3 percent.
Full expensing of equipment and a new 21-percent rate could help reverse this investment drought. The goal is not to bring back old industries and old jobs with old capital goods. The goal is to help traditional industries — manufacturing, transportation, retail, wholesale, health care, food, education, energy, etc. — to transform and create new jobs, using a mix of physical capital and cutting edge technology.
We’re beginning to see this with the coming of autonomous vehicles, drones, personalized digital health and genomic medicine, the shale petroleum boom and robot-filled distribution centers. The Tax Foundation estimates the current tax reform bills would boost wages by around 2.8 percent over the next decade.
No doubt, the digital industries will benefit from tax reform, too, and the migration of workers into those firms will be part of the productivity and pay growth stories. But more than ever, the physical industries are poised to leverage information.
Tax reform will help us build more cloud computing and new 5G wireless networks, for example, and will, in turn, empower the overhaul of transportation,manufacturing, pharmaceutical reserach and development and the Internet of Things.
Faster productivity growth in the physical industries doesn’t guarantee closure of the mean-median gap, but it almost certainly means better pay for millions of Americans who’ve yet to fully benefit from the information revolution.
Bret Swanson is president of the technology research firm Entropy Economics LLC, a visiting fellow at the American Enterprise Institute and a fellow at the U.S. Chamber of Commerce Foundation.