Foreign countries are sapping US productivity, study shows

• Why do official figures show slowing productivity growth when innovation by U.S. firms seems as strong as ever?
• Legal profit shifting by U.S. companies to overseas may hold the answer, according to a new paper.
• U.S. productivity growth is being reduced by almost 0.1 percent each year — a small but not insignificant change for rates averaging less than 2 percent a year.

Where did America's productivity growth go? Some of the missing gains can be found overseas, along with the diverted profits of U.S. corporations, according to recent research.

American companies have become adept at stashing their earnings in foreign countries. Legal tax avoidance practices have not only allowed companies to shortchange the IRS, but they have also depressed official labor productivity growth statistics, according to a National Bureau of Economic Research working paper.

That lost economic output has cut almost 0.1 percent each year from productivity growth rates — a small but not insignificant change for rates averaging less than 2 percent a year.

"The current international accounting system allows a lot of flexibility — you might say too much flexibility — which allows companies to shift their profits to low-tax jurisdictions," said Fatih Guvenen of the University of Minnesota, one of four authors on the working paper. "A lot has been written about that in the context of taxation, but the same problem also matters for how we measure GDP in official statistics."

The study is another piece of evidence to help explain one of the biggest economic mysteries of our time: Why do official figures show slowing productivity growth when innovation by U.S. firms seems as strong as ever? Profit shifting has ramped up over the last two decades, the same period over which productivity growth has declined.

"The twist is that most of the slowdown has happened in technology," said Guvenen. "This is surprising to anyone who follows the world — just look at Uber, AirBnb, the iPhone and Facebook and so on."

The causes of the productivity growth slump are contentious. It could be that we really are producing less economic output for every hour that we work, or it could be — as some economists have argued — that there has been a breakdown in how we measure that output. If it is a mismeasurement issue, the problem would have to be getting worse over time to cause the decline in the growth rate.

One popular explanation is that the new business models used by tech companies — many of which give away their products for free — mean that users are getting better products, but aren't paying more for them. Those improvements may not be properly reflected in official statistics. But that's a difficult question to study empirically, and previous research hasn't found much evidence for a strong mismeasurement effect on productivity.

Guvenen and three economists at Pennsylvania State University and the Bureau of Economic Analysis used confidential labor compensation and sales survey data from the BEA to reclaim U.S. economic activity that is being counted toward the GDPs of tax havens like Ireland, Bermuda and the Netherlands. They determined how much output should rightly be counted in the U.S. and compared those figures to the ones calculated using profit and expenditure data, and the difference was substantial.

While their adjustments didn't reverse the productivity growth slowdown that occurred between 2004 and 2014, they did find that accounting for the production shifted offshore would slow the growth rate decline.

American output counted abroad

The adjustments start making a difference around 1994, when American multinational corporations began rapidly expanding their foreign operations. From 2004 to 2008, including that foreign production would have increased aggregate productivity growth rates by 0.25 percent each year. That seems small, and it doesn't erase the downward trend, but it is much larger than other measurement effects, and even small changes in productivity can have an outsized effect on how policymakers like the Federal Reserve see our economy.

Offshoring fits in well with the pattern in the productivity growth rate because it's a change that has been ramping up over time as companies move abroad. The data also show the biggest changes in exactly the right areas.

"The largest adjustments we find are precisely in the industries where official statistics show the biggest slowdowns," said Guvenen. "These are the big multinationals, the tech sector, pharmaceuticals, oil and gas — big companies with a lot of this profit shifting happening."

Economic value appears to be hidden outside the country more for industries with high levels of research and development, like publishers, electronic products manufacturing and chemical manufacturing, according to the paper. Those companies can easily shift intellectual property to subsidiaries in other countries, pushing U.S. productivity figures far below where they should be.

Consider the classic example — the iPhone. Much of the value of the device is arguably created by Apple employees in California. But intellectual property is transferable, so Apple can move that value to, say, its Irish headquarters. Such a switch, especially if it is undervalued, removes production value that would otherwise be correctly counted in the United States. Similar transfers are common among big tech companies like Microsoft and Alphabet, as well as drug companies and other types of companies that can place a lot of value in intangible assets like brands and patents.

More than 90 percent of the missing output is being counted in 10 other countries, according to the paper. The Netherlands and Ireland were among the biggest recipients, which isn't surprising to anyone familiar with the "double Irish with a Dutch sandwich" corporate tax strategy. Those countries have seen their GDPs swell at an unusual rate. Other major tax havens on the list included Bermuda, Luxembourg, Singapore and Switzerland. Like the U.S., high-tax countries such as Japan, France and Italy also appear to be losing some output to lower-tax countries.

The actual impact of mismeasurement due to offshoring may be even greater than the conservative estimates presented in the paper, said Guvenen. Most importantly, the research suggests that while productivity growth is declining, measurement problems may in fact be responsible for at least some of the change.

"This doesn't overturn the conclusion that there is a productivity slowdown, but this is not a negligible number" he said. "There is a mismeasurement, so it's not as bad as the official statistics make it look."