The Indian software industry, which employed more than 500,000 people and exported $17 billion in 2005, has grown at an annual rate of 30 to 40 percent over the past decade (Dossani, this volume). At the same time, the proportion of software exports accounted for by foreign (predominantly U.S.-based) companies has increased, as has the sophistication of the product mix. Thus the Indian software industry is tightly integrated with U.S.-based software development.
In the semiconductor industry, some steps (e.g., assembly, packaging, and testing) in the manufacturing process have long been globalized. In recent years, more sophisticated steps, such as wafer fabrication, have followed suit. Offshoring of semiconductor design is also increasing rapidly. In fact, 18 of the top 20 U.S.-based companies have opened design centers in India, nine of them since 2004 (Brown and Linden, this volume).
The manufacture of PCs and many PC components was moved from the United States to Taiwan more than a decade ago. Since then, it has been moved again, almost exclusively to China (Dedrick and Kraemer, this volume). In addition, much of the product design and engineering for PCs is now done by original design manufacturers based mainly in Taiwan.
Some engineering activities in the automotive industry and construction engineering and services industry have long been internationalized. U.S.-based auto companies have traditionally followed an imperative of manufacturing where they sell, and they often design and develop vehicles for specific markets (Moavenzadeh, this volume). Thus the employment of significant numbers of engineers abroad by U.S.-based companies in the auto industry is nothing new. Similarly, construction engineering and services firms that operate globally have always required engineering help in the countries where projects are located (Messner, this volume).
Nevertheless, the offshoring of less complex engineering work is increasing in both of these industry sectors. In the auto industry, some companies are trying to boost the productivity of their global engineering workforces by organizing distributed teams around global tasks. For example, global engineering leadership for a certain category of vehicle may be located in a specific country (e.g., full-size trucks in the United States, compact cars in Korea). Engineering teams in several countries contribute to the design of specific models.
Finally, the trend toward globalization of R&D in a range of other industries, including pharmaceuticals, is almost certain to gain momentum in coming years. For example, well over half of the more than 200 U.S.- and Europe-based companies that responded to a recent survey anticipate increasing technical employment in China, India, and other locations in Asia in the next three years (Thursby and Thursby, 2006).
FINDING 2. More and better data on offshoring and other issues discussed in this report, such as the effects on the engineering workforce and engineering education, are necessary for discerning overall trends. As has been pointed out in other recent reports, better U.S. and international statistics on trade in services and employment would give us a much better grasp of basic trends.
Although various surveys, projections, and analyses by consulting companies, academics, and others can shed some light on the situation, significant data gaps have kept policy makers and the public from getting an accurate read on what is actually occurring in the international trade in services and offshoring. Several recent reports (GAO, 2005a,b; NAPA, 2006; Sturgeon, 2006, etc.) have pointed out deficiencies in U.S. government statistics. For example, trade statistics track many fewer categories of service products than manufactured goods, even though services now constitute a much larger share of the U.S. economy than manufacturing. In addition, current employment statistics make it impossible to track employment by occupation over time.
Statistics on the science, technology, engineering, and mathematics workforce could also be improved (Ellis et al., 2007). One improvement would be for agencies that collect and publish these data to adjust the classifications and coding so that occupations are easier to identify and track. An example of the problem, cited by Ellis et al. (2007), is the difficulty of tracking postsecondary teachers, who are usually subsumed in the general category of educators. Thus tracking jobs in engineering is difficult because, in some fields, academics make up a large percentage of the total workforce. In addition, more information on citizenship and the migration of engineers would make it easier to understand offshoring and discern other trends in the engineering workforce.
A study of offshoring in specific industries is no doubt valuable, but we must remain cognizant of the lack of timely, comprehensive data. We must also keep in mind that, even if we had all relevant information, it would represent only a snapshot in time. Thus all estimates or projections include considerable uncertainties, as offshoring continues to change!
Although many basic questions about offshoring, particularly questions specific to engineering, cannot be answered definitively, a review of the literature on offshoring, and trade in services generally, reveals several points of rough consensus. The combination of technological advances, innovations in management techniques, and the accessibility of overseas talent has made a growing number of services jobs vulnerable to offshoring. Estimates of the number of vulnerable U.S. jobs vary considerably, from the most common estimates of around 10 percent of the current workforce (NAPA, 2006)1 up to 40 million (Blinder, 2006).