The use of low-tax centers by multinational companies as part of tax minimization strategies distorts measures of capital flows and direct investment, which are integral to globalization. Last year, for example, Microsoft shifted $52.8 bln of assets from Singapore to Ireland in a few pen and computer strokes. The Dublin-based entity, the company’s European headquarters, took control of the Asian and Pacific operations. This also led to a 3.5 bln dividend payment to the US parent, a ten-fold increase from 2017.
It is easy to be cynical and reduce globalization largely to trade and tax arbitrage, but that would be a mistake. There was a small hint in Microsoft’s announcement. Its Dublin-based staff increased from 895 to nearly 1410 employees. To be sure, it is not proportionate to the accounting changes. Still, the actual hiring of foreign employees is another essential element of globalization, which reflects deeper economic integration than import/export and is less likely to be distorted by tax minimization strategies.
Nor can it be written off as merely labor-cost arbitrage and the search for cheaper labor inputs. Employment by the majority-owned US affiliates (MOFA) accounted for a third of the multinationals’ workforce in 2017 (14.4 mln abroad vs. 28 mln domestic employment). Roughly 40% of the MOFA employment is in high-income countries, including Canada, the EU, Japan, and Australia.
In Central America, MOFAs’ employ 1.6 mln people, and a full 85% are accounted for by the highest wage country, Mexico. In fairness, Mexico, like Canada, is a special case as US businesses have organized their production for a continental market. Then, consider Africa. South Africa alone accounts for almost half MOFA employment (112.5K) of the entire continent (252k).
Access to cheap labor does not explain the hiring of European’s by majority-owned affiliates of US companies. It if did, the UK would not have accounted for a third (1.47 mln) of the MOFA employment in the EU (4.4 mln). The UK’s share of MOFA employment in Europe is greater than German and France put together. These three of the 28 EU countries accounted for nearly two-thirds of their EU workforce.
The UK appears determined to leave the EU, and this will alter the accounting a bit. The UK’s role as a regional hub and a platform to service the continent may change, depending on the outcome of next year’s negotiations, assuming Brexit takes place at the end of January. It could be significantly disruptive, but one must assume, and anecdotes suggest, businesses who used the UK as a conduit into Europe have made the necessary adjustments in the three and half years since the referendum.
The MOFA employ 130k in Ireland, and finance and insurance account for 14%. However, they hire almost three-times as many in manufacturing (nearly 40%). Switzerland is not part of the EU and only accounts for 2% (98k) of MOFA employment in Europe. Yet it accounts for 14% of the MOFA research and development in Europe and 14% of their value-added. Poland stands out as a favorite home for MOFA employment. With 205k employees, it accounts for more MOFA jobs than Hungary (68k) and the Czech Republic (87k) combined. A full two-thirds of the Polish workers are in manufacturing, of which about 40% help produce transportation equipment.
Asia is interesting. MOFA employment in 2017 was near 5.14 mln. China and India account for nearly 60% (1.73 mln and 1.30 mln, respectively). Japan and Australia roughly equally split another 13% (357k employees in Japan and 323k in Australia). There is little doubt that labor costs play a role here, though it is not the whole story.
Among these four countries, China accounts for about 40% of the employment but less than 30% of the compensation paid by the MOFA. Japan and Australia account for 10% and 9% of the employment by the MOFA, but those employees account for 24% and 25% of the compensation. The higher compensation is justified in broad terms as the value-added is greater. Among the four countries, Australia accounts for 24% of the value-added and Japan 23%. China’s share is 35% of the value-added, but the real outlier was India. It accounts for 35% of the MOFA employment in these four countries but only 17% of the value-added.
In turn, the value-added reflects, at least in part, a global division of labor by American multinationals. The externalities generated may spur self-perpetuating forces. For example, about 44% (763k) of the MOFA employment in China is in manufacturing and a quarter (202k) of which are manufacturing computers and electronic equipment. In contrast, professional, scientific, and technical services account for the 44% (580k) MOFA employment in India. Less than 5% (79k) of the employees of the MOFA in China provide such services. On the other hand, less than 3% (36k) of the Indian employees of MOFA manufacture computers or electronic goods.
US multinationals appear to have more sectoral diversification in Japan and Australia than they do in China or India. Manufacturing, finance and insurance, and professional services accounted for a higher percentage of MOFA employment in China (50%) and India (62%), but only 44% of the jobs in Japan and Australia.
One sector dominated MOFA employment in China and India, but in Australia and Japan, MOFA employment was less concentrated. Manufacturing is the largest sector of MOFA employment in both countries (23% in Japan or 82.5k and 25% in Australia or 88k). Another quarter of MOFA employment in Australia is in professional services, finance and insurance, and information. In Japan, half of those employees are manufacturing chemicals or computers and electronics. Retail and wholesale trade accounted for more MOFA employment in Japan (96k or 26%) than the manufacturing sector (82.5k or 23%). The MOFA employs practically no Indians in the retail trade and a relatively small amount in wholesale trade (87k or 6%).
The competition between nations and the divergence of tax rates and regimes provide powerful incentives for shareholder value maximizer to arbitrage between them and to seek the lowest tax burden. For some countries some of the time, such tax-inspired flows distort capital flows and foreign direct investment. The sketch here of where majority-owned foreign affiliates of US business hire workers reveals another dimension of globalization that is more stable and transparent than capital flows per se. Contrary to conventional wisdom, cheap labor does not appear to offer a compelling explanation of the employment of these affiliates. The pockets of specialization reflected in concentrations of employment in certain sectors may be sticky and point to the contours of the global division of labor, a touchstone of globalization itself. Indeed, at a time when protectionist risks are elevated, the trade slowing, build locally and sell locally may be increasingly appealing.