Equity markets have hit multiyear highs and consumer sentiment is buoyant. Yet economic productivity remains lackluster. The Labor Department announced Thursday that worker productivity fell 0.6% since January, a much bigger drop than expected. This is neither a statistical illusion nor a hangover from the Great Recession.
The productivity slowdown began long before the financial crisis, and it has worsened markedly in the past six years. The drop-off extends to wholesale and retail trade, manufacturing, construction, utilities and a host of private and public services. Industries that consume and produce information technology and communications are not immune to the slowdown.
From 1950 to 1970, U.S. productivity grew on average by 2.6% annually. From 1970 to 1990 it fell to 1.5%. The information technology boom of the ’90s interrupted the slide, but since 2010 U.S. productivity growth has been in free fall. It is now roughly 0.6% a year. No wonder Federal Reserve Chair Janet Yellen recently called low productivity a “significant problem.” Various estimates suggest that had U.S. productivity growth not slowed, GDP would be about $3 trillion higher than it is today.
How is this happening during a technological revolution? Some think the data are wrong. Economist Joel Mokyr explained in 2014 that metrics devised for a “steel-and-wheat economy” fail to capture adequately transformative advances in information technology, communications and the biosciences. Technology has reduced the cost of information, expanded consumer choice, and provided customization and better price comparison. This progress has been mostly missed in current statistics.
Gross domestic product also does not fully capture metrics like time saved from shopping online. Nor does it include the value of leisure and the well-being that technology provides its users. Many economists contend that properly counting free digital services from companies like Google and Facebook would substantially boost productivity and GDP growth. One of the highest estimates, calculated by economists Austan Goolsbee and Peter Klenow, stands at $800 billion. That’s a big number, but not big enough to fill a $3 trillion hole.
On the other hand, recent studies by Brookings, the Organization for Economic Cooperation and Development and the Bureau of Economic Analysis claim that so-called free services are actually accounted for in GDP through the money spent by firms on internet ads. Still more calculations attempt to put a dollar figure on the privacy and personal data consumers give up in exchange for “free” services, as well as the time gobbled up by employees surfing the web.
Similar debates about data abound in the biomedical sector. Technology costs account for almost half the recent rise in U.S. health-care costs, but some economists argue that official statistics don’t count the intangible benefits of drugs like statins, which prevent death from heart disease. Skeptics contend that while the full benefits of such drugs may not show up in GDP today, the benefits of penicillin and other antibiotics weren’t properly counted in the past either.
Debates about the shortcomings of official data can be useful. But pinning the productivity slowdown on a statistical deficiency ignores other factors. In his 2016 book, “The Rise and Fall of American Growth,” Northwestern University economist Robert Gordon contends that the current economy fails to measure up to the great inventions of the past, and that innovation today is more incremental than transformative. He has argued vigorously that the transformative effects of technologies like electric lighting, indoor plumbing, elevators, autos, air travel and television are unlikely to be repeated. Technological innovation, he argues, will not be sufficiently robust to counter the headwinds of slowing population growth, rising inequality and exploding sovereign debt.
Former Treasury Secretary Larry Summers has resurrected Alvin Hansen’s 1938 theory of secular stagnation. Morgan Stanley economist Ruchir Sharma has argued that a 2% economy is the new normal. Former Fed Chairman Alan Greenspan has repeatedly said that the growing share of social benefits and entitlements in GDP crowds out national savings and reduces investments required to boost productivity growth.
The growth dividends from disruptive technology often require time before they are widely diffused and used. To Mr. Gordon’s point, economic historians respond that the Industrial Revolution did not improve British living standards for almost a century. Likewise the productivity boost spurred by the transformative innovations of the early 20th century took decades to kick in.
In the short term, as companies try to develop online capabilities while maintaining a physical presence, some costs are duplicated. At the same time, the economy is becoming increasingly bifurcated, with top firms generating substantial economic rents and leaving the rest behind. In 1990 the top 10% of U.S. firms saw returns on invested capital three times that of the median firm. Today those returns are eight times the median, suggesting economic rents far above the cost of capital for a small group of superstar companies.
It’s possible that economic dynamism and entrepreneurship are no longer driving the U.S. economy. Startups are being created at a slower pace. From 1996 to 2007 the ratio of new firms to the total number of firms oscillated between 9.6 and 11.2. Today it has dropped to 7.8. Existing firms do innovate and contribute to improved productivity, but the declining share of young firms suggests a less dynamic economy.
Concurrently, the most recent numbers from the Bureau of Labor Statistics confirm that churn in the U.S. labor market remains weak across industries, regions and age groups. People are simply not moving or changing jobs for better alternatives.
The U.S. economy is reaping some of the deleterious effects sowed by decades of weak capital spending. Since the 1970s, the compound annual growth in nonresidential fixed investment has plummeted from 12% to 3.3%. In the past decade the rate at which money turns over in the economy has also declined. The otherwise dismal first-quarter 2017 GDP report provided a slight glimmer of hope — posting a 9.4% increase in business spending. How long will it last?
The real debate is about policies that favor productivity and GDP growth. Predicting future innovation is hazardous, but deregulation and streamlined licensing requirements will facilitate job mobility. Tax reform that encourages and rewards investment should stimulate capital investment. Some form of fiscal stimulus and diverse infrastructure spending should buoy demand and efficiency. Leveraging technology to improve education and job training will help match skills and jobs across the economy. Last but not least, entitlement reform must be on the policy agenda.
These necessary policy changes provide options for improving productivity and GDP growth. Waiting for the data debate to resolve itself gets us nowhere.