Skip to main content
The Avenue

Is the nation’s 10-year productivity slump almost over? The digitalization of the US workforce suggests an answer

Riddles remain, but findings in the Metro Program’s new analysis of the changing digital content of 545 occupations suggest that the 10-year U.S. productivity slump might soon have run its course and that we know some of the things that will help end it.


The new research shows not only that increases in the digital character of jobs in industries is strongly associated with improved productivity but also that many industries are surprisingly under-digitalized, leaving ample room for future productivity gains.

A look at some of the industry analysis in the report is instructive on these points, and aligns with recent research by the McKinsey Global Institute and Michael Mandel and Bret Swanson.

Built out of detailed occupational data from the U.S. Department of Labor’s O*NET surveys covering the years 2002 to 2016, our analysis shows that during the 2000s productivity and output growth were heavily associated with industries’ mean digital scores, a measure of the average level of the digital skills required by workers in industries.

Most notably, highly digital service industries like professional, scientific, and technical services; the information-communications-technology (ICT) sector itself; and media have all turned in some of the economy’s fastest output, productivity, and wage growth in the 2010 to 2016 period while other less-digital industries have lagged. See, for a sense of this, the industry data displayed here:

metro_20171218_Table 5

These data comport with academic evidence here compiled by the Information Technology and Innovation Foundation that shows that digital technologies have been a major enabler of productivity growth in the U.S. At the top of the chart and the digital spectrum, the ICT and media sectors (with digital scores of 44 and 52) increased their annual productivity growth by 3.9 percent and 2.6 percent. By contrast, productivity in medium-digital sectors like health care and advanced manufacturing grew by just 0.7 and 0.3 percent a year, while the low-tech basic goods manufacturing, construction, administration, and accommodation and food industries all saw negative productivity growth.

The upshot: the rapid adoption of technology by the most highly digital service and other industries with digital scores over 43 or so has driven superior output, productivity, and wage growth in recent years as leading-edge companies adopted digital enterprise management processes, data-driven decisionmaking, and the organizational changes needed to optimize all of that. By contrast, the slower adoption of such technologies in an array of larger, far less tech-oriented sectors with lower scores has led to anemic growth that has depressed those sectors’ own efficiency as well as that of the nation as a whole.

That lagging industries with digital scores lower than 43 account for half of the nation’s economy in terms of employment means their slow adoption of digital technology is a huge problem.  The size of those low-digital, low-productivity sectors—combined with increasing evidence of vast productivity lags within industries between “frontier” firms and most other firms—means that a lot of potential productivity gains are being left on the table.

In that sense, the so-called productivity “riddle” may not be as confounding as many experts say. The nation’s anemic productivity likely owes at least in part to the insufficient adoption of digital technologies in many of its largest industries, which are operating far less efficiently than they might otherwise.

And here both the pace of change and the sheer size of the lagging sectors license an optimistic forecast about the future.

Increases in the digital character of jobs in industries is strongly associated with improved productivity.

Rising digital scores across the board align with Mandel’s and Swanson’s prediction that the adoption of the current and next waves of the information revolution—especially through the arrival of IT in sleepy physical sectors—is soon going to bring a “productivity boom.” Mandel and Swanson argue (and it is hard to disagree) that the increased use of mobile technologies, cloud services, artificial intelligence, big data, inexpensive and ubiquitous sensors, virtual reality, robotics, additive manufacturing, and a new generation of 5G wireless is “on the verge of transforming” traditional physical industries, such as health care, transportation, energy, education, manufacturing, agriculture, retail.

Likewise, that half of the economy has barely started on its digital journey warrants confidence that significant productivity gains should be possible in the next decade. With so many large sectors operating well below their digital potential we really do know where to go and what to do to get productivity gains that will increase growth and improve living standards. Work to better diffuse digital technologies and support the organizational changes needed to fully adopt them both among leading companies and the mainstream should make a difference.

In view of that, several productivity initiatives aimed at speeding tech adoption by lagging firms and industries appear relevant.

For one thing, government should actively encourage digital innovation and transformation, including by providing tax incentives for ICT adoption (say for the immediate expensing of capital expenditures), as a timely strategy for accelerating digitalization and boosting productivity. With investment levels low in large swaths of the economy, incentives to invest in IT gear and process change would likely spur the tech update that brings productivity gains. Similarly, industry, training and education organizations, and policymakers need to develop skills initiatives to expand the IT workforce that will be needed to support broader digitalization.

On this front, our report goes into more detail, but suffice it to say that governments, firms, industry associations, and educational institutions must work urgently with workers and students to expand the high-skill IT talent pipeline even as they work to broaden basic digital literacy, especially among underrepresented groups, to connect all workers to the digital opportunities ahead.

In sum, while the productivity slump is frustrating, it isn’t entirely mysterious, either in its causes or the way forward. Too little of the economy has truly gotten with the digital program, so we need to change that. The U.S. has the potential to raise productivity significantly if it moves to more broadly adopt digital technologies, and should make that a top priority, not just in elite “high-tech” industries, but across the entire economy.

Get daily updates from Brookings