guvenenfig1.png
VoxEU Column

Offshore profits and domestic productivity

US multinational enterprises may legally shift profits from high tax countries (like the US) to low tax countries (like Bermuda). This column uses unpublished survey data to show how this causes part of the US economic activity generated by these multinational enterprises to be attributed to their overseas affiliates, leading to an understatement of measured US GDP and, in turn, productivity. Profit-shifting activity has increased significantly since the mid-1990s, resulting in an understatement of the growth rate of US productivity.

What is behind the decade-long slowdown in the US’ productivity growth? Despite a steady stream of technological innovations, such as the smartphone and cloud computing, productivity growth rates since 2004 are half those from 1994 to 2004. This apparent contradiction has spurred studies — and debate — among economists, researchers, and government statisticians seeking to understand the productivity slowdown. The stakes are high – the US economy’s long-term growth, as measured by GDP, hinges on future productivity gains.  

Some economists, including Brynjolfsson and McAfee (2011), argue that the potential growth from recent technological innovations will be realised once businesses and consumers learn how to harness them. These economists believe the US may very well be on the cusp of a surge in economic growth. Another line of inquiry considers the measurement of modern economies. Feldstein (2017) and Varian (Aeppel 2015), for example, posit that official statistics on productivity and economic growth may not fully capture the impact of technological innovations. While Americans are reaping the gains from these new technologies, they argue, their impact is not completely captured in gross domestic product.

Against this backdrop, our research shows that legal offshore profit-shifting by multinational enterprises operating in the US is causing the slowdown in US productivity growth to appear larger than it actually is (Guvenen et al. 2017). Shifting profits from the US to other countries causes part of the US economic activity generated by these multinational enterprises to be attributed to their overseas affiliates, leading to an understatement of measured US GDP and, in turn, productivity. Furthermore, profit-shifting activity has increased significantly since the mid-1990s, resulting in an understatement of the growth rate of US productivity.

Profit-shifting in multinational enterprises

Multinational enterprises operate in countries that vary widely in their corporate tax rates, which enables them to legally shift profits from high-tax countries, such as the US, to low-tax countries, such as Bermuda, through transfer pricing and complex global structuring. This profit-shifting is facilitated by the mobility of intangible capital (e.g. intellectual property, brand names, trademarks) owned by multinational enterprises. While intangible capital is difficult to measure, R&D expenditures provide an indication of a firm’s intangible capital intensity. Among US private firms, 77% of the $302 billion in R&D expenditures in 2012 were concentrated in multinational enterprises.

In the framework for calculating GDP, payments made from one country to another for the use or purchase of intangible capital are counted as exports in the recipient country’s expenditure-based GDP. These are explicit returns for the use of intangible capital. A multinational firm shifts profits when an entity in a low-tax country pays only part of the total return to intangible capital owed to an affiliated entity in a high-tax country. The part of the return that is not paid is part of the earnings (i.e. revenues minus expenditures) of the low-tax entity. In this case, some of the return on intangible capital is treated as earnings in the low-tax country instead of earnings in the high-tax country.

Adjusting for profit-shifting

We provide a measure of US productivity growth by adjusting for the understatement in value-added introduced by multinational enterprises’ offshore profit shifting. For this purpose, we used unpublished survey data collected by the US Bureau of Economic Analysis that cover the worldwide operations (e.g. employment, equipment, sales, and profits) of US multinationals. These data allow us to examine the effects of profit shifting by US multinationals. For some of the same tax reasons, foreign multinationals are also likely to shift profits out of the US, biasing measured US production further downward. Data availability keeps us from measuring the effects of foreign multinational profit-shifting, but preliminary research suggests that foreign multinationals are shifting profits out of the US.

Using the data described above, we reattribute the worldwide earnings of each multinational enterprise to its entities (the US parent and its affiliates) based on factors meant to capture the true economic activity at each location — this methodology is called formulary apportionment. We use the labour compensation and the sales to unaffiliated parties at each entity in the firm as factors, as they are likely to reflect real production taking place in each location. Any additional earnings attributed to US parents (and away from foreign affiliates) imply an increase in measured US GDP and measured productivity.

We use the results of formulary apportionment to compute an adjusted measure of value-added and labour productivity growth for the US business sector for 1973–2014. Since earnings by multinationals are disproportionately booked to low-tax countries with little real economic activity, our adjustments reattribute earnings toward the US, thereby increasing measured US GDP and labour productivity growth.

Adjusted productivity

In Figure 1, we report the aggregate adjustment to US business-sector value added. The adjustment is small until the late 1990s, when income on US direct investment abroad explodes. From 2005 onward, the adjustment adds about $250 billion per year to US value added — about 2.5% of business-sector value added per year. From 2004 to 2014, the adjustment adds $2.6 trillion to US value added.

Figure 1 Aggregate adjustments for business-sector value added in labour productivity measurement

Figure 2a presents unadjusted and adjusted cumulative labour productivity growth for 1994–2014. Figure 2b plots the difference between the two series. The cumulative impact reaches its peak in 2008, adding two log percentage points to cumulative growth.

The overall effect of our adjustments on aggregate productivity growth remains modest.  The adjustments have their largest impact during the productivity slowdown period.  Our adjustments raise productivity growth in the US by 0.1 log percentage points annually for 1994-2004, and by 0.25 log percentage points annually for 2004-2008. The adjusted series grows a bit slower than the unadjusted series for 2008-2014.

Figure 2a Cumulative labour productivity growth    

Figure 2b Increase in cumulative growth

Our adjustments are most noticeable in industries that are R&D intensive, which is consistent with the importance of intangible capital in profit shifting. This is particularly interesting given that these types of industries are the ones that Fernald (2015) finds to be most responsible for the productivity slowdown.

In Figure 3a, we plot the adjustments to R&D-intensive and non-R&D-intensive industries. The adjustments for R&D-intensive industries are as high as 8.0% of the group’s value-added in 2008. We plot the increase in cumulative labour productivity growth from our adjustments in the right panel of Figure 3b. The effects of the adjustment are substantial:  the annual productivity growth rate from 2000–2008 is 0.6 log percentage points higher in R&D-intensive industries after the adjustment. By 2008, the adjustment has added almost 6 log percentage points to cumulative labour productivity growth in the R&D-intensive industries.

Figure 3a Adjustments by R&D intensity   

              

Figure 3b Increase in cumulative growth by R&D intensity

Authors’ note: The views expressed here are solely those of the authors and not necessarily those of the U.S. Commerce Department’s Bureau of Economic Analysis or the Federal Reserve Bank of Minneapolis.

References

Aeppel, T (2015), “Silicon Valley Doesn’t Believe U.S. Productivity is Down,” Wall Street Journal, July 17.

Brynjolfsson, E and A McAfee (2011),  Race against the Machine:  How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy, Lexington, MA:  Digital Frontier Press.

Feldstein, M S (2017), “Underestimating the Real Growth of GDP, Personal Income and Productivity.” NBER Working Paper No. 23306.

Fernald, J G (2015), “Productivity and Potential Output Before, During, and After the Great Recession,” in J A Parker and M Woodford (eds), NBER Macroeconomics Annual 2014, 29, University of Chicago Press: 1-51.

Guvenen, F, R J Mataloni Jr, D G Rassier and K J Ruhl (2017),  “Offshore Profit Shifting and Domestic Productivity Measurement.” NBER Working Paper No. 23324.

2,730 Reads