The globalization of R&D is not new; companies have been locating research and development facilities abroad for decades. IBM founded its first overseas research center, in Zurich, Switzerland, in 1956, and Japanese auto companies have had design studios in the U.S. since the 1960s. The share of R&D facilities located outside the home markets of multinational corporations, however, has been rising steadily — from 45 percent in 1975 to 66 percent in 2005, according to a 2006 study conducted jointly by Booz & Company and INSEAD. That share continues to increase: Between 2004 and 2007, global multinationals increased their total R&D sites by 6 percent, and of those new sites, 83 percent were in China and India. They also increased R&D staff by 22 percent; 91 percent of that increase was in China and India.
To gain further insight, we closely examined the global innovation footprints of the top 100 R&D spenders, along with the top 50 companies in each of the three highest-spending industries: auto, health care, and computing and electronics. The 184 companies in this group support more than 3,400 facilities in 47 countries around the world. Together, they spent more than $350 billion on R&D in 2007. That amounts to 71 percent of the total spend of the Global Innovation 1000 and 57 percent of all private-sector R&D spending. On average, just 45 percent of these companies’ total R&D spending occurred inside their home countries.
Judging by our sample, companies in the U.S. spent the largest amount on R&D in other countries, making the U.S. the top “net exporter” of R&D spending in 2007, followed by Japan and Switzerland. When the numbers are adjusted for the overall size of the countries’ economies, however, the U.S. falls to sixth place in the ranking of net exporters of R&D investment. The top “net importer,” by a long shot, is China, where $24.7 billion in R&D spending was accounted for by foreign companies, with product development activities now following the numerous manufacturing sites established there. India was the second-largest, with $12.9 billion in net imports; its large, English-speaking talent pool and fast-growing auto, computing and electronics, and pharmaceutical markets will certainly stimulate further growth. Other top net importers included Canada, Israel, and the United Kingdom.
Although we use the terms exporting and importing as convenient shorthand for describing the flow of R&D money around the world, these terms have limited usefulness and can easily be misunderstood or misused. When economic nationalists and opponents of globalization look at innovation flows, for example — particularly in Europe and the United States — they see only the “export” of high-paying science and engineering jobs to low-cost countries (LCCs) in the developing world. Our analysis, however, suggests that cost reduction is not the most important of the several reasons that multinationals are moving their R&D facilities abroad. Furthermore, it is receding in significance. The underlying reasons are more complex and multifaceted. They include:
• Lower costs. To be sure, the initial impetus for conducting research and development overseas was often to save money, in part by replacing higher-paid “home-country” engineers with lower-paid replacements in LCCs. But our analysis shows that lower engineering labor rates explain only one-third of the move to site R&D facilities overseas. Labor costs are rising rapidly in many LCCs as demand for skilled engineering and other talent grows. In India, for example, the wage rate for high-end service workers was 53 percent of the equivalent rate for U.S. workers in 2005. In 2008, the percentage had risen to 65 percent, and it is projected to rise to 77 percent in 2012 and 90 percent in 2020.