National Academies Press: OpenBook
« Previous: 3 The Changing Landscape of the U.S. Information Technology R&D Ecosystem: 1995-2007
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 106
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 107
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 108
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 109
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 110
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 111
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 112
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 113
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 114
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 115
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 116
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 117
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 118
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 119
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 120
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 121
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 122
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 123
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 124
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 125
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 126
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 127
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 128
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 129
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 130
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 131
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 132
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 133
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 134
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 135
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 136
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 137
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 138
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 139
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 140
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 141
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 142
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 143
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 144
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 145
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 146
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 147
Suggested Citation:"4 A Globalized, Dynamic Information Technology R&D Ecosystem." National Research Council. 2009. Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment. Washington, DC: The National Academies Press. doi: 10.17226/12174.
×
Page 148

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

4 A Globalized, Dynamic Information Technology R&D Ecosystem Profound changes have altered the U.S. national information technol- ogy (IT) research and development (R&D) ecosystem during the 1995- 2007 period that is the focus on this report. The forces of globalization have shaken the foundations of the product, labor, and financial markets of the IT industry. They have created tremendous opportunity, but they also mean that the United States will have to work even harder to remain the global leader in IT R&D. R&D funding models, in both academic and industrial environments, have also evolved. Finally, the nature of the employer-employee relationship has continued to change across most sec- tors, but perhaps in a deeper way in the IT industry than anywhere else. The Globalization of PRODUCT AND LABOR Markets As world markets such as those of India, China, and Eastern Europe open, competition for information technology workers has become global, with many U.S. companies looking the world over for the best talent, in the right place, at the right price. Most U.S.-based technology companies are now global from birth, driving innovation through collaborations with foreign technologists. For example, Figure 4.1 shows the significant increase from 1990 to 2005 in joint patenting by Silicon Valley inventors working with global teams. Figure 4.1 and Box 4.1 illustrate the global nature of IT innovation and sourcing. Fueling the trend toward global sourcing are significant advances in telecommunications and networking technologies, as well as the evolu- 106

a globalized, dynamic information technology R&D ecosystem 107 700 600 2000-2005 Number of Patents 500 1995-1999 400 300 1990-1994 200 100 0 Taiwan Israel Japan Singapore South Germany China India Finland Korea Figure 4.1 Foreign co-inventors listed on patents with Silicon Valley inventors, 1990-2005. Source: AnnaLee Saxenian, University of California, Berkeley, pre- Figure 4-1.eps sentation to the committee, Mountain View, Calif., February 23, 2007. Based on data needs shades for shapes analysis conducted by Collaborative Economics, Inc., Palo Alto, Calif., 2007. tion of work and business processes. One powerful trend is for firms to consider what work they should retain internally and what they should purchase from outside vendors. The decision to purchase from an outside vendor work that was formerly done internally is termed outsourcing. The other powerful trend is to scan the globe to decide where specific work processes should be undertaken. Often, firms are deciding that work can be done more efficiently and effectively in nations outside the United States. Of course, multinational firms have a long history of establish- ing subsidiaries abroad. What has changed in the past four decades is the increasing movement of work to developing nations. This practice is referred to as offshoring. More recently, there has been an upsurge in offshore outsourcing. Finally, this offshoring initially was for the manufac- ture of goods, but recently it has extended to the production of software and IT services. The Offshoring of U.S. IT Jobs According to a recent study by the McKinsey Global Institute, the offshoring of work is more prevalent in the IT sector than it is in any of the other U.S. industry sectors studied. Published in 2005, the report esti- mated that by 2008, U.S. firms would offshore 18 percent of their demand for high-wage workers in the packaged-software sector and 13 percent Ron Hira and Anil Hira, Outsourcing America: The True Cost of Shipping Jobs Overseas and What Can Be Done About It, AMACOM, New York, N.Y., 2005.

108 assessing the impacts of changes in the it R&D ecosystem BOX 4.1 iPod and iTunes: Internationalization of Design and Implementation The Apple iPod is a digital music player with a highly stylized industrial design and an easy-to-use click-wheel user interface. It was not the first media player, but it is certainly the most commercially successful. The first model was announced on October 2001. By April 2007, over 100 million had been sold. As an example of the rapid design cycle of modern consumer products, five generations have been launched in only 6 years: the iPod, iPod mini, iPod shuffle, iPod nano, and video iPod. The iPod plays audio and video media in standard formats, such as the open standards MP3 (MPEG-1 Audio Layer 3) and Apple proprietary formats. A key element of Apple’s success is the platform which it developed for digital media that encompassed its online store, iTunes. iTunes was introduced in April 2003 to sell individual songs at the price of $0.99 each. iTunes Media is encoded using Apple’s AAC format with additional levels of encryption. The representation and its associated digital rights management system make it possible to authorize up to five computers and an unlimited number of iPods to play the files. An unlim- ited number of audio compact disks can be produced from the digital representa- tion, but at a loss in quality. The iPod offers an interesting case study in the internationalization of prod- uct design and implementation.1 For the fifth-generation video iPod, among the most costly components are those contributed by companies headquartered in Japan (Toshiba, which supplies the hard drive), Korea (Samsung Electronics, which supplies the flash memory), and the United States (Broadcom Corpora- tion, which supplies the multimedia processor). These components are in turn manufactured around the globe—in China (hard drive), in Taiwan or Singapore (media processor), and in Korea (the memory). The device is assembled by the Taiwanese firm Inventec Corporation in Mainland China. The analysis by Linden, Kraemer, and Dedrick indicates that out of a suggested retail price of $299, the cost of all components of the iPod is $144. Of the $155 price difference, $80 ac- crues to Apple and $75 to the distributor and retailer. Apple’s value is the single largest component and is larger than that associated with the most expensive physical component, Toshiba’s hard drive. This value of Apple represents the company’s considerable competitive advantage in product conception, design, and marketing. Apple is amply compensated for the innovation that the firm has embedded in the product. The portion captured by U.S. firms—design, distribu- tion, and sales—exceeds the value of its manufactured components. Although to a large extent the iPod is not manufactured in the United States, it is designed and sold here, and U.S. firms do quite well in the bargain. 1See G. Linden, K. Kraemer, and J. Dedrick, “Who Captures Value in a Global Innovation System? The Case of Apple’s iPod,” Personal Computing Industry Center, University of Cali- fornia, Irvine, June 2007.

a globalized, dynamic information technology R&D ecosystem 109 in IT services. The McKinsey study used data from the U.S. Department of Labor’s Bureau of Labor Statistics, as well as global company-level data, to derive a microeconomic picture of the extent of offshoring that had occurred and was expected to occur. The McKinsey study reported that the theoretical maximum global resourcing for packaged software and IT services represents from 44 to 48 percent of the industry’s total employment. However, it is estimated that only 13 to 18 percent would be offshored, owing to a number of barriers ranging from management attitudes, to business process suitability, to lack of sufficient scale, to intel- lectual property protection. Another study, by Alan S. Blinder, also uses the Bureau of Labor Sta- tistics data, devising a model that categorizes jobs into two groups—those that can be personally delivered (e.g., medical care, child care, and so forth) and those that can be “impersonally” delivered—that is, the job can be delivered to the end user electronically over long distances with little or no degradation in quality (for example, by call center operators).  Blinder’s study places all IT jobs in offshorable categories and concludes that the percentage of offshorable IT jobs is roughly twice that estimated by the McKinsey study. To fully understand the real impact of offshoring in IT, however, one must match up the demand for workers with the supply of workers in the countries to which work is being outsourced. The McKinsey report concludes that although the potential talent pool in low-wage companies is large and growing rapidly, only 17 percent of the potential engineering talent supply is suited for work with international companies. The report explains the reasons for its conclusion, which was based on interviews with 83 human resource managers in multinational companies: the ­reasons are McKinsey Global Institute, The Emerging Global Labor Market, June 2005, available at http://www.mckinsey.com/mgi/publications/emerginggloballabormarket/index.asp; ac- cessed August 27, 2007. As this report was being prepared for publication, a continued weakening of the U.S. currency had increased the cost of goods and services sourced from abroad. Such a trend decreases the benefits of outsourcing and offshoring for U.S. firms. Alan S. Blinder, How Many U.S. Jobs Might Be Offshorable?, Center for Economic Policy Studies [CEPS] Working Paper No. 142, CEPS, Princeton University, March 2007, available at www.princeton.edu/~blinder/papers/07ceps142.pdf. The notion that all IT jobs can be done remotely from the consumer and/or the core business was questioned by the committee. Many IT jobs are critical to the successful imple- mentation of a business and/or are central to a firm’s competitive differentiation. Further, IT R&D also involves work by large teams, who collaborate to create new platforms or services. Notice that the McKinsey study’s conclusion is a point estimate. It is likely, even ex- tremely likely, that nations and workers will try to improve their education and capabilities so that they can participate in the global economy, because in many of these nations this will ensure higher incomes.

110 assessing the impacts of changes in the it R&D ecosystem dispersion of the labor force, domestic competition for talent, and indi- vidual limitations (e.g., inadequate language skills, limited practical skills, lack of cultural fit, inability to work on teams, and lower educational attain- ment) as well as considerable scarcity of middle-management skills. This shortage of qualified staff is becoming a headache, according to a recent survey conducted by The Economist. According to the 600 executives of multinational companies that were surveyed, the shortage of qualified staff ranked as their biggest concern in China, second in Japan (after cul- tural differences), and fourth in India (after problems with infrastructure, bureaucracy, and wage inflation). The Economist goes on to say: Technical skills, particularly in information technology, are lacking in many parts of the region, even India. One of the main concerns is that there are not enough skilled graduates to fill all the jobs being created in a vibrant sector. Nasscom, which represents India’s software companies, has estimated that there could be a shortfall of 500,000 IT professionals by 2010. This means companies recruiting at job fairs in India are having to make lucrative offers to capture the most promising students. Even a junior software engineer can expect to take home $45,000/year.  A high turnover rate also helps to drive up wage costs. The same article in The Economist reminds readers how supply and demand in labor markets must equalize through wages, and that the transfer of IT jobs from countries such as the United States to countries such as India and China, while politically and socially alarming, tends to be an overstated and self-regulating phenomenon. According to the M ­ cKinsey study, for IT and engineering-based services, if the United States and the United Kingdom continue at their current rate to concen- trate their activities in India, China, and the Philippines, the U.S. and U.K. demand for engineers will fully absorb the supply of suitable engineers in India, China, and the Philippines by 2011. The U.S. Workforce and the Global IT Industry Information technology professionals play an important role beyond the research and development organizations of technology vendors. “Capturing Talent,” The Economist, August 16, 2007, available at http://www.economist. com/business/displaystory.cfm?story_id=9645045; accessed August 27, 2007. See Rafiq D ­ ossani, India Arriving: How This Economic Powerhouse Is Redefining Global Business, Ameri- can Management Association, New York, N.Y., 2007, for a discussion of how institutions of higher education in India are responding to this shortage. McKinsey Global Institute, The Emerging Global Labor Market, June 2005, available at http://www.mckinsey.com/mgi/publications/emerginggloballabormarket/index.asp; ac- cessed August 27, 2007.

a globalized, dynamic information technology R&D ecosystem 111 Because IT is transforming the foundational business processes of all corporations, IT professionals are increasingly critical to corporations in diverse sectors, such as retailing, hospitality, finance, and pharma- ceuticals. Outside corporate walls, IT professionals are also at work in entrepreneurial and small businesses, as well as creating the next wave of software services through technologies such as the Web 2.0 infrastructure. Many are independent consultants. IT professionals work on a wide variety of challenging technical proj- ects, ranging from research into new scientific frontiers such as high performance computing, speech recognition technology, sensors or radio- frequency identification to new computing platform development and corporate business re-engineering (integrating technology to improve productivity significantly). A recent report from market research firm Forrester Research points to the sophistication of the IT professional job in today’s enterprise. According to Forrester, IT professionals can follow a variety of career paths—sourcing, management, innovation, architec- ture—each of which requires a combination of relationship-management, not just project-management, skills and activities. Continued Strong Demand for IT Workers According to data collected by the U.S. Department of Commerce, there are more professional IT workers in the United States today than ever before; “IT professional workers” in this case are defined as computer support specialists; computer programmers; computer systems analysts; computer software engineers; applications, computer, and information systems managers; computer software engineers; systems software, net- work, and computer systems administrators; all other computer spe- cialists; network systems and data communications analysts; database administrators; computer hardware engineers; computer and information scientists; and computing researchers. In fact, a recent report on globaliza- tion and the offshoring of software states:10 According to the U.S. Bureau of Labor Statistics reports, despite a sig- nificant increase in offshoring over the past five years, more IT jobs are available today in the US than at the height of the dot.com boom. More- over, IT jobs are predicted to be among the fastest-growing occupations over the next decade. Lorie M. Orlov, Samuel Bright, and Lauren Sessions, Is There a Career Future in Enterprise IT? Forrester Research, Cambridge, Mass., August 10, 2006. 10Association for Computing Machinery Job Migration Task Force, Globalization and Off- shoring of Software: A Report of the ACM Job Migration Task Force, W. Aspray, F. Mayadas, and M. Vardi, eds., Association for Computing Machinery, New York, N.Y., 2006.

112 assessing the impacts of changes in the it R&D ecosystem A recent report from the Bureau of Labor Statistics that contains occu- pational employment projections through 2016 states: Computer and mathematical science occupations are projected to add 822,000 jobs—at 24.8 percent, the fastest growth among the eight profes- sional subgroups. The demand for computer-related occupations will increase in almost all industries as organizations continue to adopt and integrate increasingly sophisticated and complex technologies. Growth will not be as rapid as during the previous decade, however, as the software industry begins to mature and as routine work is outsourced overseas. About 291,000—or 35 percent—of all new computer and math- ematical science jobs are anticipated to be in the computer systems design and related services industry. The management, scientific, and technical consulting services industry is projected to add another 86,000 computer and mathematical science jobs. This expected 93-percent increase is due to the growing need for consultants to handle issues such as computer network security.11 The report also states that among all fields of science and engineering, “computer specialist” is projected to account for 77 percent of all job growth and 66 percent of all available jobs (which includes both growth and positions available due to retirement). Data from the National Science Foundation (NSF) reinforce the picture of a relatively strong job market for science and engineering graduates, particularly for computer and information science graduates. According to NSF’s Scientists and Engineers Statistical Data System, the overall unemployment rate of scientists and engineers in the United States was 2.5 percent in 2006, compared with 3.2 percent in 2003; 2.5 percent is the lowest rate since the early 1990s. For computer/information scientists, the overall unemployment rates were 2.5 percent in 2006 (down from 4.0 per- cent in 2003).12 Also, according to a 2006 survey from NSF, the median sal- ary level for computer and information science graduates with bachelor’s degrees was $45,000 (the median for all science and engineering fields was $39,000); at the master’s level, the median salary was $65,000 (the median for all science and engineering fields was $56,000).13 11Arlene Dohm and Lynn Shniper, “Occupational Employment Projections to 2016,” Monthly Labor Review, Bureau of Labor Statistics, Washington, D.C., November 2007, pp. 86-125. 12Nirmala Kannankutty, Unemployment Rate of U.S. Scientists and Engineers Drops to Record Low 2.5% in 2006, Science Resources Statistics InfoBrief, NSF 08-305, National Science Foun- dation, Washington, D.C., March 2008. For electrical/computer hardware engineers, overall unemployment rates for 2006 were higher than for computer/information scientists, but still improved: 3.3 percent (down from 5.5 percent in 2003). 13Steven Proudfoot, An Overview of Science, Engineering, and Health Graduates: 2006, NSF- 08-34 (revised), March 2008, available at http://www.nsf.gov/statistics/infbrief/nsf08304/,

a globalized, dynamic information technology R&D ecosystem 113 Strong Concerns About Sustaining a Strong IT Workforce Despite the demand, the number of students specifying an intention to major in computing and information sciences has dropped significantly in the past 6 years. For example, according to College Board data for 2006, the number of students indicating on their SAT test a desire to major in computing and information sciences has dropped by almost 50 percent since 2001.14 Also according to the College Board, in 2006 the SAT math- ematics scores (an indicator for success in IT) of those intending to major in computing and information sciences averaged 478, far lower than the mathematics scores for those intending to major in other scientific and mathematical disciplines. These statistics not only point to a sharp decline in the number of students entering the IT educational pipeline, 15 but also raise a concern about the skill sets of those attracted to the discipline. The problem of declining enrollments in the computing disciplines (as compared with the projected demand) is compounded by the severe lack of participation of underrepresented groups in IT. Although the participation of women, minorities, and people with disabilities in other science, technology, engineering, and mathematics fields is rising overall, their participation is especially low, and even declining, in computing. In 2006, women received 59 percent of all bachelor’s degrees, but only 21 percent of computer science degrees.16 African-American and Hispanic graduates received only 10 percent and 6 percent of 2004 computer sci- ence degrees, respectively. Women and minorities are even more severely underrepresented in positions requiring a doctoral degree. Of the 1,189 Ph.D. graduates in computer science or computer engineering in 2005, only 18 percent were women, and only 38 of the total 1,189 (3 percent) accessed April 9, 2008. See also Jay Vegoso, “Employment and Salaries of Recent CS Gradu- ates,” CRA Bulletin, March 25, 2008, available at http://www.cra.org/wp/index.php?p=141; accessed April 9, 2008. 14College Board, 2006 College Bound Seniors: Total Group Profile Report, 2006, available at http://www.collegeboard.com/prod_downloads/about/news_info/cbsenior/yr2006/ national-report.pdf; accessed February 20, 2007. 15Although the following facts are not necessarily a perfect surrogate for high school stu- dents’ interest in computer science, it is interesting to note that about 12,000 students in the class of 2007 took the Computer Science A Advanced Placement (AP) test; about 4,000 took the more rigorous Computer Science AB test. For comparison, about 14,000 students took the Art History and French tests; almost 50,000 took the Economics Macro test, and about 28,000 took the Economics Micro test. See College Board, The 4th Annual AP Report to the Na- tion, Appendix B, 2008, available at http://professionals.collegeboard.com/profdownload/ ap-report-to-the-nation-2008.pdf; accessed April 4, 2008. 16National Center for Education Statistics, Integrated Postsecondary Educational Data System (2005-06), U.S. Department of Education, Washington, D.C., May 1, 2007.

114 assessing the impacts of changes in the it R&D ecosystem were members of underrepresented minorities (African-American, Native American, or Hispanic).17 The picture is also bleak in the workforce. In 2006, the percentage of women in management, professional, and related occupations was 50.6 percent, whereas the percentage of women in computer and mathemati- cal occupations was only 25.6 percent.18 Such low participation has implications beyond the nation’s ability to create and sustain a sufficiently large IT workforce. Women and minori- ties can bring different life experiences and perspectives to innovation, which lead to the design of products and services that benefit a broader range of people. Such perspectives are especially important, considering the changing demographics of the U.S. population19 as well as the global market for IT products and services. If U.S. companies intend to maintain their competitive advantage both at home and abroad, they must seek the input of a broader segment of the population to achieve innovation. For example, a recent analysis of innovation and diversity with respect to IT patenting revealed that within the United States, mixed-gender invention teams produced the most frequently cited IT patents—with citation rates that were 26 to 42 percent higher than the norm.20 How can young people be encouraged to enter computing fields? One essential ingredient is to ensure a strong national IT educational pipeline that prepares and encourages all qualified students regardless of race, gender, or ethnicity to enter the discipline. Without sustained attention and additional measures to attract and retain all qualified stu- dents, it will be especially difficult to reverse the negative trends. 21 17S. Zweben, “Record PhD. Production Continues; Undergraduate Enrollments Turning the Corner,” Computing Research News 19(3):7-22, 2007. 18Bureau of Labor Statistics (BLS), Current Population Survey: Household Data: Annual Aver- ages: 2007, BLS, Washington, D.C., Table 11: Employment by detailed occupation, sex, race, and Hispanic ethnicity, p. 212. 19Council of Economic Advisors for the President’s Initiative on Race, Changing America: Indicators of Social and Economic Well-Being by Race and Hispanic Origin, U.S. Government Printing Office, Washington, D.C., September 2007, available at http://www.access.gpo. gov/eop/ca/index.html. 20Catherine Ashcraft and Anthony Breitzman, Who Invents IT? An Analysis of Women’s Par- ticipation in Information Technology, National Center for Women and Information Technology, Boulder, Colo., March 2007. 21For examples of new measures to improve STEM education and strengthen educational opportunities for students in K-12 (such as ways to retain and reward the most effective teachers), see “Testimony of William H. Gates, Chairman, Microsoft Corporation and Co- Chair, Bill & Melinda Gates Foundation, Before the Committee on Science and Technology, United States House of Representatives, March 12, 2008,” available at http://democrats. science.house.gov/Media/File/Commdocs/hearings/2008/Full/12mar/gates_testimony_ 12mar08.pdf; accessed March 17, 2008.

a globalized, dynamic information technology R&D ecosystem 115 Concerns About K-12 IT/Computing Education and Talent Generation Concerns about talent generation are exacerbated by the state of the kindergarten-through-grade-12 (K-12) IT/computing education system in the United States. In its report The New Educational Imperative: Improving High School Computer Science Education, the Computer Science Teachers Association (CSTA) correctly assesses the situation as follows: Computers have infiltrated all areas of society, and there is now a clear link between technology, innovation, and economic survival. In light of this, one would expect a move within our society to support and stan- dardize computer science education. Yet, no national K–12 computer science curriculum exists. Lack of leadership on high school computer science education at the highest legislative and policy levels has resulted in insufficient funding for classroom instruction, resources, and profes- sional development for computer science teachers. In addition, complex and contradictory teacher certification requirements as well as salaries that cannot possibly compete with industry make it exceedingly difficult to ensure the availability of exemplary computer science teachers. In the face of confusing definitions of computer literacy, information fluency, and the various sub-branches of computer science itself, many schools have lost sight of the fact that computer science is a scientific discipline and not a “technology” that simply supports learning in other curricu- lum areas. Computer science is not about point and click skills. It is a discipline with a core set of scientific principles that can be applied to solve complex, real-world problems and promote higher-order thinking. In short, knowledge of computer science is now as essential to today’s educated student as any of the traditional sciences.22 In addition to resources, appropriate information technology fluency objectives for K-12 are needed.23 Recent research by the CSTA shows the following: • Only 26 percent of schools require students to take a computer sci- ence (CS) course; • Only 40 percent of schools even offer advanced placement (AP) CS; • Lack of time in the students’ schedules is the greatest impediment to students taking computing courses; 22Computer Science Teachers Association (CSTA), The New Educational Imperative: Im- proving High School Computer Science Education, available at http://csta.acm.org/; accessed August 27, 2007. 23For an early assessment of fluency issues, see National Research Council, Being Fluent with Information Technology, National Academy Press, Washington, D.C., 1999.

116 assessing the impacts of changes in the it R&D ecosystem • 89 percent of high school computer science teachers say that they experience a sense of isolation and a lack of collegial support in their schools and in their districts; • Most administrators do not understand that computing is a scien- tific discipline just like physics and biology; • There is no consistency in CS teacher certification requirements; • Computing teachers do not receive the professional development that they need to keep their teaching and technical skills current; • Administrators, legislators, and congressional committees do not understand the link between supporting K-12 computing education and economic and workplace issues.24 Such concerns about the professional IT pipeline and talent pool have arisen as the U.S. share of worldwide bachelor’s and doctoral degrees in science and engineering has decreased significantly. The relative decline in the U.S. global position in science and technology overall is also evi- dent in the falling U.S. share of global R&D investment, patents, scientific publications, and researchers (see Table 4.1). If it is to maintain its founda- tion for competitive strength, the United States faces a long-term need to attract qualified people to science and technology careers.25 The Globalization of Venture Capital Until the late 1980s, for all intents and purposes the United States was the only nation with a vibrant venture capital industry that supported technology-based start-ups.26 For this reason the United States was in a privileged position. For an entrepreneur seeking to build a global- class IT firm, it was necessary to come to the United States—and many entrepreneurs did. It was in the 1990s that venture capital industries in Taiwan and Israel began growing, with the Taiwanese venture capitalists funding manufacturing firms such as Quanta Computer Incorporated and AsusTek Computer; in Silicon Valley they funded start-ups particularly 24Computer Science Teachers Association (CSTA) Curriculum Improvement Task Force, The New Educational Imperative: Improving High School Computer Science Education, CSTA, As- sociation for Computing Machinery, New York, N.Y., February 2005. 25For a business-oriented discussion of the importance of maintaining the STEM pipeline, see, for example, Testimony of William H. Gates, Chairman, Microsoft Corporation and Co- Chair, Bill & Melinda Gates Foundation, Before the Committee on Science and Technology, United States House of Representatives, March 12, 2008, available at http://democrats. science.house.gov/Media/File/Commdocs/hearings/2008/Full/12mar/gates_testimony_ 12mar08.pdf; accessed March 17, 2008. 26The committee thanks Martin Haemmig, Martin Haemmig International, for providing much of the venture capital data used in this section.

a globalized, dynamic information technology R&D ecosystem 117 Table 4.1 Declining Relative U.S. Standing in Worldwide Share of Various Areas of Science and Technology: Share of Global Total (in percent) for Selected Years from 1985 to 2003 Area 1985 1986 1988 2002 2003 Investment in domestic research and 46 37   development New U.S. patents 54 52 Scientific publications 38 30 Scientific researchers 41 29 Bachelor’s degrees in science and 39 29   engineering New doctorates in science and engineering 52 22 NOTE: Data are only for select years, as provided in source. Source: Data from Council on Competitiveness, Competitiveness Index: Where America Stands, Washington, D.C., 2007, p. 67. in semiconductor design.27 The Israeli firms funded by venture capitalists were concentrated in IT (particularly in enterprise software, communica- tions technologies, and security), and many of them immediately opened U.S. offices and later went public on the NASDAQ. For both countries, strong relationships with the United States and U.S. venture capitalists were important for their growth.28 Changing Patterns in Global Venture Capital For at least the past three decades, the U.S. venture capital limited partnerships have been the beneficiaries of inflows of capital from around the world, particularly from European financial institutions. This capital was primarily invested in U.S. technology start-up firms. During that period, the United States was the destination of choice for investment funds, entrepreneurs, and venture capital firms. 27For example, for evidence from semiconductor design firms that went public on U.S. markets, see M. Kenney and D. Patton, “The Coevolution of Technologies and Institutions: Silicon Valley as the Iconic High-Technology Cluster,” in P. Braunerhjelm and M. Feldman, eds., Cluster Genesis: Technology-Based Industrial Development, Oxford University Press, Ox- ford, England, 2006, pp. 38-60. 28Martin Kenney, Martin Haemmig, and W. Richard Goe, “Venture Capital,” in Innova- tion in Global Industries: U.S Firms Competing in a New World, Jeffrey T. Macher and David C. Mowery, eds., The National Academies Press, Washington, D.C., 2008. On Israel, see Gil Avnimelech and Morris Teubal, “Venture Capital Start-up Co-evolution and the Emergence and Development of Israel’s New High Tech Cluster,” Economics of Innovation and New Tech- nology 13(1):33-60, 2004.

118 assessing the impacts of changes in the it R&D ecosystem As late as 1995, there were few globalized venture capital firms, such as Apax Partners, Advent, Sofinnova Ventures, Hambrecht and Quist Capital Management, and Walden International, but most of the elite U.S. IT-oriented venture capital firms invested in international deals rarely and idiosyncratically. During the rise of the dot-com era, start-up firms from Europe, Asia, and Latin America were able to secure funding through European and U.S. markets. This encouraged U.S. venture capitalists to expand their investments abroad. However, it was not until after the collapse of the stock market bubble in 2000 that U.S. venture capital firms decidedly expanded their global reach, with a particular focus on China, Israel, and India. U.S. venture capitalists began building link- ages with venture capital firms in other nations and, if and when there was a sufficient deal flow, contextual understanding, and relationships, they developed more permanent foreign operations abroad. According to Ernst & Young, over the period 2005 to 2006, 19 percent of venture capital investing was done across national borders and continents,29 an increase of 250 percent from 5 years earlier. Another 10 percent of the deals are intra-European or intra-Asian. Thus nearly 30 percent of venture capital is invested across borders. As Table 4.2 illustrates, North America (predominantly, the United States) clearly remains the most important venture capital location in the world. First, more venture capital is invested there than in the rest of the world combined. Second, it is at the center of the flows of venture capital, with more flowing into and out of North America than to and from any other region. Finally, with the exception of U.S. centricity and the resultant flows into and out of North America, there are only minimal interregional flows of capital. This situation is unlikely to change soon, although that is no reason for complacency: the sophistication of non-U.S.-based venture capitalists and the track record of their funds are beginning to rival those of U.S. venture capitalists and the success of their funds, thereby creating unavoidable competition for both deal flow and limited-partner capital over the long term. The United States continues to invest more venture capital in IT than do all other regions combined, despite a decline of approximately 8 percent in such U.S. investments from 2003 to 2007. The U.S. investment across all sectors was roughly flat over this period with the exception of communications, which continued to drop, suggesting a powerful hang- 29Ernst& Young, Acceleration: Global Venture Capital Insights Report 2007, 2007, available at http://www.indiavca.org/upload/library/29_E&Y_Global_VC_Insight_Report_2007.pdf; accessed November 2008.

a globalized, dynamic information technology R&D ecosystem 119 Table 4.2 Interregional Flows of Venture Capital Investment (in millions of U.S. dollars), by Location of Firm and Location of Investment, 2005 Location of Venture Location and Amount of Investments ($ million) Capital Firm (Origin North Rest of of Investor) America Europe Israel Asia World Total North 21,914 1,837 158 798 218 24,925   America Europe 840 3,486 35 163   59 4,583 Israel 139 30 208 0    0 377 Asia 502 118 4 502    3 1,129 Rest of 52 42 7 59 231 391   world    Total 23,447 5,513 412 1,522 511 31,405 Source: Compiled by Martin Haemmig, Martin Haemmig International, www.martin haemmig.com, from data provided by National Venture Capital Association/Venture Eco- nomics, European Venture Capital Association, Asian Venture Capital Journal, and Israeli Venture Capital Association. over effect from the collapse of the dot-com bubble. In Europe and Israel, the decline was greater and spread into nearly every sector.30 The greatest recent change in the location of venture capital investing is the emergence of China as a major focus of the investment of venture capital, in particular by foreign firms. In 2006, Chinese firms received $1.9 billion in venture capital investment, making China the second-largest national recipient of venture capital investment.31 In IT, much of the venture capital investment in China thus far has been in firms that are adapting Western Internet business models for China (for example, gam- ing, travel sites, job sites, portal, search, and so forth). One area of innova- tion has been in the mobile applications field. Such investments require relatively little technological innovation but can be very successful, as the Chinese market and online population are already very large and growing 30Martin Haemmig, Martin Haemmig International, presentation to the committee, data on cumulative capital invested in IT by region and sector, based on Ernst & Young data, Mountain View, Calif., February 23, 2007. 31Ernst & Young, Acceleration: Global Venture Capital Insights Report 2007, 2007, available at http://www.indiavca.org/upload/library/29_E&Y_Global_VC_Insight_Report_2007.pdf; accessed November 17, 2008.

120 assessing the impacts of changes in the it R&D ecosystem very rapidly. Thus far there have been very few impactful R&D-intensive Chinese start-ups funded by venture capitalists,32 and unlike Japan or Europe, no global IT brand except possibly Lenovo and Huawei has emerged out of China despite its economic prowess. However, there can be little doubt that the technological and scientific level of Chinese R&D is advancing rapidly, and given the size of the market there is the distinct possibility that global-class IT firms could emerge in the next 5 years. India differs from China in important respects. Indian start-up firms have access to talented engineers and benefit from a sizable local market. However, growth is hampered by a scarcity of management skills and a weak, though improving, business infrastructure. In terms of developing an IT R&D ecosystem in India capable of generating top-quality start-ups, the role of the R&D operations of Silicon Valley firms in India has to be considered. Already, firms such as Adobe Systems, Broadcom Corpora- tion, Cisco Systems, Google, Intel Corporation, Juniper Networks, Oracle, and many more are developing products in India. The results of this hands-on experience will be seasoned product-development specialists. It is almost certain some of these engineers will become entrepreneurs. Despite the scarcity of management skills, indigenous and foreign venture capitalists are already looking for opportunities in India. This is a likely indication that many Indian start-ups will begin to emerge and receive funding over the next several years. Prior to the late 1990s, the United States benefited from the inflow of capital from other nations. Today the flows of capital are bidirectional. U.S. venture capitalists are globalizing rapidly as IT entrepreneurship becomes more globally dispersed. U.S. venture capitalists are also mobi- lizing their networks and unique know-how to benefit their portfolio firms regardless of location. As long as the U.S. venture capitalists retain their edge in the soft skills required to build successful start-up compa- nies, they will remain at the center of gravity of the global start-up deal flow and will likely retain their centrality despite this globalization. If, however, portfolio financial returns degrade, if the number of successful initial public offerings (IPOs) fails to recover to historical levels, and if the weight of ecosystem frictions (discussed later in this chapter) proves impossible to overcome, the asset class will weaken, and the globalization of venture capital could cause non-U.S. firms to rise to prominence. 32The most R&D-intensive Chinese IT firms are probably Huawei Technologies and ZTE Corporation, neither of which was funded by venture capital. Lenovo is another important Chinese IT firm, which purchased IBM’s personal computer division, but Lenovo is not known for cutting-edge IT research or products.

a globalized, dynamic information technology R&D ecosystem 121 Table 4.3 Decline in Percentage of Venture Capital Invested in Information Technology Between 2001 and 2006 2001 2006 Total Total Invested IT Share Invested IT Share Capital of Total Capital of Total Venture Capital Investments ($ billion) (percent) ($ billion) (percent) In the United States 36.4 66 25.7 48 In Europe   9.8 60   5.2 51 In Israel   2.2 79   1.4 72 In China   2.9a 73   1.9 49   aChina venture investment for 2001 is skewed by $1 billion investments in each of two companies: Heijan Technology and Semiconductor Manufacturing International Corpora- tion (SMIC). Source: Martin Haemmig, Martin Haemmig International, “China’s and Israel’s Role in ‘IT’ Through Venture Capital,” presentation to the committee, Mountain View, Calif., Febru- ary 23, 2007. Based on data from Ernst & Young and VentureOne, Q1 2007. Venture Capital Investment in IT U.S. venture capitalists continue to invest more than those of any other nation in the IT fields: in the first half of 2006 they invested $7.15 billion in IT, with the software and communications segments receiving the most capital. In the first half of 2006, the rest of the world’s venture capital investment in IT firms was not even one-third that of the United States.33 Nevertheless, the percentage of venture capital investment in IT declined from 66 percent of the U.S. total to approximately 55 percent in the first half of 2006. As Table 4.3 indicates, total venture capital invest- ments declined in the United States, Europe, Israel, and China from 2001 to 2006 (after the collapse of the Internet bubble). This global decline was accompanied by a decline in the percentage of total venture capital invested in IT. There was also a slight decrease in the number of IT firms funded, from 3,420 in 2000 to 3,192 in 2007.34 Venture capital investment in the U.S. IT firms dropped from year 2000, but appears to have stabi- lized by 2007. From the perspective of this committee, there can be no doubt that with respect to venture capital funding of IT firms, the United States completely dominates other areas of the world. Moreover, with the exception of China and India, the other major locations of venture capital 33Martin Haemmig, “China’s and India’s Role in ‘IT’ Through Venture Capital,” presenta- tion to the committee, Mountain View, Calif., February 23, 2007. Based on data from Ernst & Young and Venture One. 34Ibid.

122 assessing the impacts of changes in the it R&D ecosystem investing suffered similar or even greater declines in IT-related invest- ment between 2001 and 2006. Frictions in the U.S. IT R&D Ecosystem It appears that a number of inefficiencies have been growing in the U.S. IT R&D ecosystem over the past several years. They do not seem to be concentrated in any single area. Together, they form a pattern of fric- tions that, over time, could hurt the health and competitiveness of the U.S. ecosystem—particularly given its increasingly global nature. Symptoms of these frictions can be found by examining the data on technology com- pany initial public offerings, technology company mergers and acquisi- tions (M&As), and overall venture capital activity during the 1995-2006 period. Although success stories like that of Google (see Box 3.2 in Chapter 3) leave many with the superficial impression that all is well with venture- funded innovation, closer examination suggests that Google is a unique case in scope and magnitude and that the field of play in recent years has lacked depth and breadth. Although reaching an IPO is not a guarantee of long-term future success, IT companies that do not have the opportunity to tap public equity markets will not have the capital required to grow into major industry players and to contribute meaningfully to the creation of high-quality jobs in this country. In 2006, there were only 40 U.S. technology IPOs; by contrast, in 1995, there were 195.35 The year 1995 was not yet caught in the distortion of the technology bubble of the late 1990s (the number of technology IPOs peaked during the bubble, at 381 in 1999), and the year 2006 is no longer held back by the post-technology bubble crash. Instead, these numbers represent a meaningful downward trend that has become even more pro- nounced in 200836 and is unlikely to reverse in the near term. This decline in IPOs is not due to a retrenchment of overall U.S. venture capital activity. The total amount of venture capital invested in the United States in 2006 was reported to be about $26 billion, compared with just over $8 billion in 1995 (see Table 4.4). Similarly, almost twice as 35Paul Deninger, Jeffries and Company, presentation to the committee, Boston, Mass., April 19, 2007, citing data from Thompson Financial and Jeffries Broadview IPO Database. The figure excludes telecommunications providers, Internet Protocol service providers, and transactions in which under $15 million were raised. 36See, e.g., Ernst & Young, “Global IPO Activity Fallen by More Than Half Since 2007: Lowest Number of Deals over an 11 Month Period Since 1995,” Ernst & Young, London, December 9, 2008. Available at http://www.ey.com/global/content.nsf/International/ Media_-_Press_Release_-_Global_IPO_activity_fallen_by_more_than_half_since_2007; ac- cessed December 11, 2008.

a globalized, dynamic information technology R&D ecosystem 123 Table 4.4 U.S. Venture Capital, Merger and Acquisition (M&A), and Technology Company Initial Public Offering Activity, 1995-2006 U.S. Total U.S. U.S. M&A Technology Venture Venture Transactions: Company Capital Capital IT, Media, Initial Public Deals Investment Telecommunications Offerings Year (no.) ($ billion) (no.) (no.) 1995 1,844 8.1 1,461 195 1996 2,573 11.3 1,956 243 1997 3,156 14.9 2.652 155 1998 3,647 21.1 2,847 116 1999 5,507 54.1 3,602 381 2000 7,911 105.2 3,704 264 2001 4,481 40.7 2,403   26 2002 3,091 21.9 2,452   22 2003 2,914 19.8 2,000   22 2004 3,069 22.5 2,294   52 2005 3,127 23.1 2,524   54 2006 3,533 26.3 2,584   40 Sources: Venture capital deal data from PricewaterhouseCoopers/National Venture Capi- tal Association MoneyTree data, available at https://www.pwcmoneytree.com/MTPublic/ ns/nav.jsp?page=notice&iden=B; accessed August 28, 2007. Data on initial public offering and merger and acquisition transactions from Paul Deninger, Jeffries and Company, presen- tation to the committee, Boston, Mass., April 19, 2007, citing data from Ernst & Young. many venture capital deals (companies financed) were reported for 2006 as for 1995.37 This increase in overall venture activity (amounts invested, numbers of deals) does not correspond to an increase in the number of venture capital firms. To the contrary, the number of firms has decreased. Fueled by the bubble, the number of U.S. venture capital firms making at least one investment in a given year reached 2,206 in the peak year 2000. As is often the case in such situations, the weaker players do not survive and the industry must consolidate: by 2005, this number had dropped to 960.38 What accounted for a much lower number of IPOs in 2006 as com- pared with 1995? Although the IT industry did traverse a rough period 37According to PricewaterhouseCoopers/National Venture Capital Association Money­Tree data, available at https://www.pwcmoneytree.com/MTPublic/ns/nav.jsp?page=notice& iden=B; accessed August 28, 2007. 38Paul Deninger, Jeffries and Company, presentation to the committee citing data from Ernst & Young, Boston, Mass., April 19, 2007.

124 assessing the impacts of changes in the it R&D ecosystem from 2001 to 2003, IT spending in the United States has bounced back, new technology platforms have emerged and attracted a new generation of IT start-up companies, and IT has become more widely deployed and economically and socially important. One factor in this changing scene has been a shift toward M&As: a greater percentage of young technol- ogy companies chose to merge with a larger strategic partner rather than becoming a publicly traded company on a U.S. exchange. As shown in Table 4.4, the number of M&A transactions in the IT, media, and telecom- munications sectors rose steadily from 1,461 in 1995 to a peak of 3,704 in 2000, then dropped to an average of about 2,400 transactions annually during the period 2001 to 2006.39 However, although M&As have become the preferred exit of U.S. IT companies, the number of M&A transactions has not grown in recent years. The juxtaposition of these two trends (a rapid decline in U.S. IT IPOs and a stable, but flat, M&A environment) suggests why the returns to venture funds from their IT investments have sharply declined over the period of this study’s scope. With this decline in returns from IT investments, venture investors will naturally rebalance their portfolios, making fewer IT investments in favor of investments in other sectors. The reasons that may explain the decline in IPOs are multiple and hard to quantify. It is instructive to note that the decline has been a U.S. phenomenon. Even as U.S. public equity markets such as NASDAQ and the New York Stock Exchange (NYSE) have experienced a dearth of IT IPOs in recent years, other markets outside the United States have been more successful and have managed to attract listings from companies that a decade ago would not have considered an IPO in other than a U.S. market. Clearly, the globalization of financial markets and the increased competitiveness of exchanges such as the London Stock Exchange’s Alter- native Investment Market (AIM) or the Hong Kong exchange have con- tributed to the weakness described in this section.40 Another factor may be Chinese government incentives for companies in China to use domestic exchanges for their public offerings. However, these factors do not appear to fully explain the decline. Over the years, new laws and regulations have been introduced that appear to have had negative and unanticipated side effects on the effec- tiveness of the U.S. IT R&D ecosystem. Moreover, there are indications that older laws and regulations have not been fully adapted to the chang- 39Paul Deninger, Jeffries and Company, presentation to the committee, citing data from Jeffries Broadview Global Mergers 7 Acquisitions database, Boston, Mass., April 19, 2007. M&A transactions dipped to only 2,000 transactions in 2003, and then recovered. There were 2,584 M&A transactions in 2006. 40Paul Deninger, Jeffries and Company, presentation to the committee, Boston, Mass., April 19, 2007.

a globalized, dynamic information technology R&D ecosystem 125 ing realities of a globalized IT environment that is based on new techno- logical platforms and new innovation methods. As one example, a major source of friction for young IT companies is the current U.S. patent system. Patents are being more actively acquired and vigorously enforced in recent years.41 Firms are facing dramatically increased hazards of litigation as plaintiffs and even more rapidly increas- ing hazards as defendants.42 The increase in litigation cannot be explained by the patenting rate, the level of R&D activity, firm value, or indus- try composition,43 leaving changes in patent system implementation (for example, increases in the number of patents being sought and imperfec- tions in patent issuance44) as the most likely explanations. Firms that spend more on R&D are more likely to be sued, and firms that acquire more patents are more likely to sue. The sharp increase in the probability of being sued per R&D dollar spent implies an increase in the “tax” that litigation imposes on innovation. Small firms face much higher marginal enforcement costs and marginal taxes on R&D. The number of patent lawsuits filed annually in the United States began to rise in the late 1980s and doubled (to almost 1,600 a year) dur- ing the 1990s.45 Simultaneously, the cost to try a patent case in the United States has also increased far more sharply than R&D budgets have. According to a 2001 economic survey conducted by the American Intel- lectual Property Law Association, the median cost to try a patent case with $1 million to $25 million at risk was almost $1.5 million.46 By 2003, this amount had increased to $2 million. Moreover, as the amount at risk increases, litigation becomes more expensive. In cases with more than $25 million at risk, the litigation costs were $3 million in 2001 through 2004, 41National Research Council, A Patent System for the 21st Century, The National Academies Press, Washington, D.C., 2004, p. 19. 42According to Bessen and Meurer, the number of patent lawsuits filed annually in the United States doubled during the 1990s, from almost 800 in 1990 to almost 1,600 in 1999; their research also “suggests that patent litigation can affect innovation incentives.” James Bessen and Michael Meurer, “The Patent Litigation Explosion,” paper presented at Ameri- can Law and Economics Association Annual Meeting, 2005, p. 1 and Figure 1, available at http://papers.ssrn.com/sol3/Papers.cfm?abstract_id=831685#PaperDownload; accessed March 6, 2008. For litigation hazard findings, see ibid., Table 2. 43Ibid., Abstract. 44For discussion of patent system implementation and issuance, see, for example, National Research Council, A Patent System for the 21st Century, The National Academies Press, Wash- ington, D.C., 2004, Chs. 3 and 4. 45James Bessen and Michael Meurer, “The Patent Litigation Explosion,” paper presented at the American Law and Economics Association Annual Meeting, 2005. These analyses are based on Derwent data from the United States Patent and Trademark Office. 46American Intellectual Property Law Association, 2001 Report of the Economic Survey, Arlington, Va., 2001.

126 assessing the impacts of changes in the it R&D ecosystem and $4 million in 2003 through 2005.47 Most patent litigation never reaches trial but is settled instead. The phenomenon dubbed by some as the “patent troll” has also been on the rise.48 Some in the legal profession have argued that entities that acquire ownership of patents with the intention of licensing them, rather than acquiring patents by developing new products, are not in themselves harmful or the root cause of excessive litigation. Instead, they consider poor-quality patents as the root cause.49 However, in the current patent environment, others do consider these activities to have adverse effects both on the patent system and on innovation.50 Furthermore, the choice of jurisdiction where a filing is made can dramatically influence the outcome of a patent lawsuit, introducing more risk and volatility into the litigation process for intellectual-property-intensive companies. This can give rise to “forum shopping,” where plaintiffs seek a jurisdiction thought to favor plaintiffs. The likelihood of a plaintiff’s verdict, for example, is substan- tially higher in courts in the Eastern District of Texas and in the Central District of California than anywhere else in the country. These trends suggest that the U.S. IT R&D ecosystem has become far more contentious than it was in the past. In summary, the cost of protect- ing and defending intellectual property is undergoing rapid inflation. The long-term effects of this phenomenon may be more pernicious: • The costs of protecting an invention go up. It takes more money to file a patent. It takes longer to be granted a patent. One must file in multiple jurisdictions as markets have become more global. • It costs more to defend oneself. It is possible for companies that never produce or commercialize a product to extract relatively high 47American Intellectual Property Law Association, 2003 Report of the Economic Survey, Arlington, Va., 2003; and American Intellectual Property Law Association, 2005 Report of the Economic Survey, Arlington, Va., 2005. For more on direct costs of and potential inefficiencies in the patent system, see National Research Council, A Patent System for the 21st Century, The National Academies Press, Washington, D.C., 2004. 48The term patent troll was reportedly coined in 1981 as a pejorative term to describe com- panies that game the patent system by snapping up critical bits of technology, then shopping for settlements from companies that might be infringing on the patents. Mike McNamee, ed., “Washington Outlook: A Patent War Is Breaking Out on the Hill,” Business Week, July 4, 2001, available at http://www.businessweek.com/magazine/content/05_27/c3941058_mz013. htm; accessed September 12, 2007. 49See, for example, James F. McDonough, “The Myth of the Patent Troll: An Alternative View of the Function of Patent Dealers in an Idea Economy,” Emory Law Journal 56:189-228, 2006, available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=959945; accessed March 6, 2008. 50See, for example, David G. Barker, “Troll or No Troll? Policing Patent Usage with an Open Post-Grant Review,” Duke Law and Technology Review, No. 9, 2005, available at http:// www.law.duke.edu/journals/dltr/articles/pdf/2005dltr0009.pdf; accessed March 6, 2008.

a globalized, dynamic information technology R&D ecosystem 127 license fees from companies that must then add these costs to those of building a market and bringing products to market. • When damages are awarded, the contribution of the infringed pat- ent can be attributed a disproportionate role. Taken together, these trends may have a stifling effect on young IT compa- nies, especially those just bringing products to market, that have limited funds and no patent portfolios for use in cross-licensing agreements or as the basis for countersuits. Such companies run a greater risk today of never acquiring sufficient intellectual property protection and mustering enough legal resources to withstand costly and lengthy litigation. TechNet, a preeminent bipartisan political network of chief executive officers and other senior executives of leading U.S. IT companies, views this matter as a fundamental issue for the IT industry. Key elements of a successful reform of the U.S. patent litigation system might include the following: • Clear standards for forum selection that curtail the ability of plaintiffs to file infringement actions in jurisdictions most likely to favor plaintiffs; • Reforms that direct courts to calculate the royalty or damages awards on the basis of a consideration of the proportionate value of the patentee’s contribution to the product in question rather than on the full value of the entire product; • Provisions of current law that have never been interpreted to per- mit the recovery of worldwide damages in U.S. courts; • Standards governing awards of multiple damages for willful in­- fringement; and • Additional reforms, as necessary, to curtail practices that are a drain on innovation. Another source of friction comes from the unexpected and unan- ticipated consequences of corporate-governance reform legislation on venture firms pursuing an IPO. The Sarbanes-Oxley Act of 2002 (Public Law 107-204), referred to as SOX, and in particular its Section 404, were passed to improve the quality of corporate governance among U.S. pub- licly traded companies and to reduce the risks of financial fraud. SOX was created and enacted to a large degree in response to the corporate scandals of such large companies as Enron, WorldCom, and Tyco. (See the discussion of financial scandals as shocks to the IT R&D eco­system in the Chapter 3 subsection entitled “Financial Scandals and Bankrupt- cies [December 2001]”). Thus, the intended firm for which SOX was designed was a multi-billion-dollar, multinational corporation listed on

128 assessing the impacts of changes in the it R&D ecosystem U.S. exchanges. It was not a typical high-growth, sub-$100 million tech- nology company led by creative entrepreneurs and technologists and funded by U.S. venture capitalists. Yet, these smaller companies have been subjected to the same regulations created for the large firms, and the costs of compliance are disproportionately more burdensome.51 Young technology companies lack the critical mass required to deploy the administrative staff, processes, and controls mandated by SOX in order to pursue an IPO in the United States. Their lifeblood is technology innovation. They often cannot afford to reallocate a large percentage of their resources away from research and development toward general and administrative costs in order to become compliant and seek a U.S. IPO. Therefore, when crafting corporate-governance legislation and regula- tions, it is important that policy makers take into consideration unin- tended consequences on smaller companies. Industrial Research: SHIFTING PATTERNS OF CORPORATE Information Technology R&D The large industrial research laboratories have traditionally been a significant institutional category in the U.S. IT R&D ecosystem. 52 How- ever, by the late 1980s the firms supporting major laboratories, such as AT&T Bell Laboratories, IBM, and Xerox Palo Alto Research Center, came under intensifying pressure to shift their research portfolios toward more applied research and development work. Some other firms, most notably Cisco, have pursued a corporate strategy of “research by acquisition,” rather than establishing and maintaining a central research infrastructure (see Box 4.2).53 51CRA International, “Sarbanes-Oxley Section 404 Costs and Implementation Issues: Spring 2006 Survey Updates,” Washington, D.C., April 17, 2006. The CRA survey was spon- sored by four large accounting firms. Self-reported costs for compliance with SOX Section 404 were as follows: smaller companies with market capitalizations of between $75 million and $700 million estimated that implementation costs (including audit fees) amounted to about $1.2 million the first year and $860,000 the second year. For larger companies with market capitalizations over $700 million, the first- and second-year costs were estimated to be about $8.5 million and $4.8 million, respectively. For the smaller companies, SOX Section 404 compliance costs were estimated to be about half of all audit fees the first year and about the same as non-404 audit costs the first year—in other words, SOX Section 404 compliance basically doubled the audit fees. 52R. Rosenbloom and W. Spencer, eds., Engines of Innovation, Harvard Business School Press, Boston, Mass., 1996. 53These changes are part of a more general trend toward what is sometimes called open innovation, whereby ideas flow both from and into corporations: in order to prosper in the face of new markets and competitors, incumbents must transform themselves from “closed” innovation models (with heavy corporate investment in internal R&D) that are no longer sustainable to more open models, without centralized control and where ideas transfer

a globalized, dynamic information technology R&D ecosystem 129 In the process of shifting toward more applied work, the traditional industry research laboratories underwent a traumatic downsizing. The firms within which they were housed experienced fundamental changes in their business environments owing to increased competition caused by deregulation as well as that from new industry entrants, which in at least some cases were leveraging new IT developments. The business utility of the large central R&D laboratory was called into question. Already in the early 1980s, the first of the major electronics-related laboratories to expe- rience downsizing was RCA’s Sarnoff Laboratories—which became too expensive to support as RCA lost its competitive position in the television industry as Japanese and European (i.e., Royal Philips Electronics N.V.) firms and competitors increasingly wrested control of the newest techno- logical developments from RCA.54 U.S. firms irrevocably lost control of image-display technologies to East Asian companies.55 The explanation for the demise of corporate IT R&D laboratories is complex. Ultimately, corporate executives, representing their sharehold- ers, judged that there was insufficient or too delayed return on investment in research. Firms that controlled their industry sector through monopo- lies or near monopolies often operated the leading laboratories. As these monopolies ended, their ability and commitment to maintain R&D spend- ing waned.56 Underinvestment yielded to a vicious downward spiral of fewer new technology products to bring to market, further eroding market share and profitability. For example, RCA was the pioneering U.S. television manufacturer, holding most of the basic TV patents. By the mid-1990s its market dominance had collapsed. Yet even as many firms in the IT industry reduced the size of their research laboratories, Microsoft and Intel established and greatly expanded their own industrial research activity. out into start-ups and enter by way of acquisition or merger. See Henry Chesbrough, Open Innovation: The New Imperative for Creating and Profiting from Technology, Harvard Business School Press, Boston, Mass., 2003. 54See, for example, Alfred D. Chandler, Jr., Inventing the Electronic Century: The Epic Story of the Consumer Electronics and Computer Industries, with a New Preface, Harvard University Press, Cambridge, Mass., 2005; Margaret Graham, RCA and the VideoDisc: The Business of Research, Cambridge University Press, Cambridge, England, 1986; and Michael Porter, Cases in Competitive Strategy, Free Press, New York, N.Y., 1983. For a further discussion of the col- lapse of the U.S. consumer electronics industry, see Martin Kenney and James Curry, The Globalization of the Television and Personal Computer Industries, Final Report to the Alfred P. Sloan Foundation, July 26, 1999, available from the authors upon request. 55For a discussion of this process, see T. Murtha, S. Lenway, and J. Hart, Managing New Industry Creation, Stanford University Press, Stanford, Calif., 2001. 56For a discussion of the impacts of the AT&T divestiture on Bell Laboratories and on telecommunications research more broadly, see National Research Council, Renewing U.S. Telecommunications Research, The National Academies Press, Washington, D.C., 2006.

130 assessing the impacts of changes in the it R&D ecosystem BOX 4.2 Acquisition and Development— A Substitute for Basic Research? The overwhelming success of U.S. firms that emerged from the venture capital-financed information technology (IT) research and development (R&D) ecosystem in the United States belongs to firms that were not disposed to basic research conducted in central laboratories. For example, Gordon Moore, one of the founders of Fairchild Semiconductor and Intel Corporation, believed that such basic research was unwise, and at Intel a research laboratory was never estab- lished.1 If Intel, which has grown to be one of the largest firms and certainly the richest semiconductor firm in Silicon Valley, did not establish a corporate labora- tory, then it is not surprising that other venture capital-financed semiconductor firms also have not established laboratories. The strategic question of whether not having a laboratory to develop new business opportunities places a firm in danger of being outflanked in the rapidly changing IT industries has not yet been satisfactorily answered. The wisdom in Silicon Valley prior to the 1990s was that purchasing start-ups was futile. Key personnel in the newly acquired firm would leave to establish a competitor, and the acquirer would be left with an empty shell. Cisco Systems dem- onstrated that it was possible to do acquisitions and, more important, to use the en- trepreneurial ecosystem as a substitute for laboratory-derived new technologies.2 Through a sophisticated, multifaceted monitoring of its environment, Cisco was able to uncover innovations before their widespread adoption and then to purchase a firm that had created such an innovation together with its technology—and, all importantly, to retain a sufficient number of the firm’s key personnel. In effect, Cisco is allowing the ecosystem to do its R&D. As the data on the location of Cisco’s acquisitions show, the company particularly depends on the entrepreneurial envi­ ronments such as Silicon Valley to provide its acquisitions. The great majority of the 60 firms acquired by Cisco as of 2007 were located in the San ­Francisco Bay Area.There are small clusters of Cisco acquisitions in ­Massachusetts (about 12 firms) and Texas (about 10 firms), but only about 20 firms from the rest of the United States and fewer than 20 from the rest of the world.3 These acquisitions came in addition to Cisco’s spending approximately 15 percent of its revenue on R&D, much of which is simply to keep existing product lines up to date. Cisco’s ability to evolve with the rapidly changing networking marketplace without a dedicated research laboratory, along with the movement from circuit switching to packet switching, combined to overwhelm the incumbent telecommunications equipment providers in the marketplace and sealed the fate of their research laboratories, not only in the United States but in most other na- tions as well. Acquisitions have become a routine aspect of Cisco’s business model and increasingly of the business model of other major firms in the IT R&D ecosystem. Established firms routinely monitor the environment for promising start-ups. Salient examples are the purchase by Microsoft Corporation of Vermeer Technologies, the producer of an early client-server Web publishing software firm, and of Hotmail,

a globalized, dynamic information technology R&D ecosystem 131 BOX 4.2 continued the e-mail start-up, and the purchase by Google of Keyhole for its satellite map- ping technologies, and then a variety of other small digital mapping start-ups to improve its map program. It is interesting to note that Intel has subsequently established a network of more conventional research laboratories and created a collection of small “lablets” close to major universities to capture innovations emerging from that element of the ecosystem.4 In 2006, Cisco appointed its first vice president of research, Douglas E. Comer, a computer science professor from Purdue University, and established the Cisco Research Center. At this time, the organization appears to play a coordination role rather than that of a research laboratory.5 1Gordon E. Moore, “The Accidental Entrepreneur,” 2001, available at http://nobelprize.org/ nobel_prizes/physics/articles/moore/index.html; accessed June 20, 2007; and Keith Naughton, “Outsourcing: Silicon Valley East,” Newsweek, March 6, 2006, available at http://www.news- week.com/id/46807; accessed November 18, 2008. 2David Mayer and Martin Kenney, “Economic Action Does Not Take Place in a Vacuum: Understanding Cisco’s Acquisition and Development Strategy,” Industry and Innovation 11(4):299-325, 2004. 3Data on Cisco acquisitions compiled by committee member Martin Kenney from information available on Cisco’s Web site, www.cisco.com. 4See “Intel Research Network of Labs,” available at http://techresearch.intel.com/articles/ None/1475.htm; accessed August 22, 2007; and “Network of Labs Home,” available at http:// www.intel-research.net/; accessed March 27, 2008.The U.S. lablets are in Berkeley, California; Pittsburgh, Pennsylvania; and Seattle, Washington. 5See “Cisco Research,” available at http://www.cisco.com/web/about/ac50/ac207/crc/ index1.html; accessed August 22, 2007. The affordability of research was also affected by the shift from large, technically sophisticated computers to personal computers (PCs) and sys- tems built as commodity boxes with standardized software. Once-well- established computer manufacturers such as Control Data Corporation, Prime Computer, Digital, Tandem Computers, and Compaq Computer Corporation have disappeared from the industry landscape, in part owing to the rise of commodity PCs. High-volume PC makers such as Dell, Acer Incorporated, and Hewlett-Packard Company’s personal computer division cannot justify basic research on the very slim profit margins that characterize this industry sector. What has happened in the case of the PC is that R&D has shifted up the supply chain to the component mak- ers, that is, Intel, and to software firms such as Microsoft, Adobe, Intuit, and others. As importantly, if the products of the component makers and

132 assessing the impacts of changes in the it R&D ecosystem software firms are affected by commoditization or a move toward open- source software, this supplier-based R&D could also be threatened. Among the new generation of Internet firms, the firm perhaps best known for hiring holders of advanced degrees in IT fields is Google, though it does not have a traditional industrial research organization; 57 rather, its researchers are developers who are part of product teams who happen to write technical papers. Yet the size and scale of the systems and applications that they are building place them at the research frontier in many areas. Even here, however, competitive pressures and time to mar- ket make it difficult to come up with the sustained investments necessary to tackle truly fundamental research problems.58 The Funding and Organization OF information technology r&d Federal Versus Industrial R&D The principal reason for the dramatic advances in information tech- nology and the subsequent increase in innovation and productivity is the “extraordinarily productive interplay of federally funded university research, federally and privately funded industrial research, and entre- preneurial companies founded and staffed by people who moved back and forth between universities and industry.”59 This flow of ideas and people, stimulated by investments in research, is a critical element of the IT R&D ecosystem. Looking across all fields of science and engineering, it can be seen the United States has significantly increased total R&D funding over the past 50 years. Particularly in the past two decades, most of that increase has been in the industrial rather than the federal portion (see Figure 4.2). However, the vast majority of industry R&D funding is for develop- ment, with limited funding devoted to applied research and a relatively small amount for basic research (see Figure 4.3). While not all indus- trial research is applied research and not all university research is basic 57See “About Google Research,” available at http://research.google.com/about.html; ac- cessed August 22, 2007. 58Historically, the locus of industrial research has tended to be in industries and companies that enjoy high growth and high margins. As these industries and companies mature, unless they find new high-growth/high-margin opportunities, their profit margins decrease and they often cut back on research. In this view, the demise of industrial research is not inevi- table; rather, industrial research investment flows from slowing industries and companies to others in emerging, high-growth sectors. 59National Research Council, Funding a Revolution: Government Support for Computing Re- search, National Academy Press, Washington, D.C., 1995, p. vii.

a globalized, dynamic information technology R&D ecosystem 133 Federal Share Industry Share Other 350,000 In millions of adjusted dollars (2005 constant dollars) 262,500 175,000 87,500 0 2000 2002 2004 1960 1964 1966 1968 1990 1980 1984 1996 1954 1962 1986 1988 1998 1956 1958 1982 1992 1994 1972 1970 1978 1976 1974 Year Figure 4.2 Total U.S. research and development funding across all fields of Figure 4-2.eps science and engineering, 1954-2004, by source. Source: Computing Research Association. Basic Research Applied Research Development 250,000 In millions of adjusted dollars (2005 constant dollars) 187,500 125,000 62,500 0 1994 1996 1998 2000 2002 2004 Year Figure 4.3 Industry research and development funding for basic research, ap- plied research, and development, 1994-2004. Source: Computing Research Association. Figure 4-3.eps

134 assessing the impacts of changes in the it R&D ecosystem research, to a large extent this is a valid characterization of activities by these two critical elements of the IT R&D ecosystem. The character of this development-heavy R&D is different from and complementary to feder- ally funded university-based research. The former tends to be focused on product and process development, areas that have more immediate impact on business profitability. However, it is basic research that “puts ideas in the larder” for later use in innovative products. Getting the fruits of university-based research into the marketplace presents a growing challenge for the people and the institutions involved. The speed and success with which university research results are trans- ferred to industry and commercialized depend critically on universities’ choices of technology-transfer mechanisms and incentives and the degree to which entrepreneurship is encouraged within the university commu- nity. As Litan, Mitchell, and Reedy have found, the type of technology- transfer organization (TTO) established by a university, as well as the metrics chosen for evaluating the TTO’s effectiveness and the incentives offered for entrepreneurial activity by the university community, can fos- ter or impede technology transfer to industry.60 One common arrangement is a centralized TTO (which receives all faculty invention disclosures and negotiates all licenses). Although TTOs focus on revenues, licensing, and commercialized inventions, they often have maximization of university revenue (looking for a “big hit”) as the central objective, rather than maximization of the numbers of commer- cialized inventions. Alternative mechanisms and incentive structures are in use, each with its own advantages and disadvantages. These include policies of “free agency,” whereby faculty members can choose whether or not to go through the university TTO, as long as they return some por- tion of their profits to the university; regional alliances among multiple universities; Internet-based “matchmaking” approaches that are built to maximize volume; and “loyalty” models in which universities relinquish all rights in hopes of faculty donating some of their gains back to the uni- versity. Litan, Mitchell, and Reedy conclude that it is preferable to move away from the “big hit” model of university technology transfer toward models (including open-source collaborations and non-exclusive licens- ing) that concentrate on the number of and speed with which university innovations are sent into the marketplace.61 60See Robert E. Litan, Lesa Mitchell, and E.J. Reedy, “The University as Innovator: Bumps in the Road,” Issues in Science and Technology, Summer 2007, available at http://www.issues. org/23.4/litan.html; accessed December 13, 2007. 61Ibid.

a globalized, dynamic information technology R&D ecosystem 135 Federal Funding of Information Technology R&D Unlike industrial R&D, federal R&D funding consists of a much greater proportion of the investment—and significant increases—in basic research. Note that the main increase in the federal investment in basic and applied research in the past 35 years, and particularly in the most recent decade, has been in the biomedical sciences (see Figure 4.4 for federal agency funding data compiled by the Computing Research Association). Concerns over the level of federal support for IT R&D are longstand- ing. In its 1999 report Information Technology Research: Investing in Our Future,62 the President’s Information Technology Advisory Committee (PITAC) argued in great detail for a doubling over a period of 5 years of the federal investment in IT R&D, noting that “critical problems are going unsolved, and we are endangering the flow of ideas that has fueled the information economy,” and describing the level of investment at that time as “dangerously inadequate.” Comparing the targets for annual increases set in the PITAC report (Table 4.5) and the actual budget levels shown in Figure 4.5 shows that although the federal IT R&D budget initially rose rapidly following pub- lication of the 1999 PITAC report, the lower rate of growth in subsequent years has meant that the budget level 9 years after the release of the PITAC report still has not reached the target set in that report. This pattern mirrors a broader underinvestment in the physical sci- ences and engineering highlighted in two recent studies: Engineering Research and America’s Future (2005) and Rising Above the Gathering Storm (2007).63 Rising Above the Gathering Storm had as a major focus the rela- tive lack of investment in engineering and the physical sciences (which include information technology). It is not that investments in the biomedi- cal sciences have been excessive but that investments in engineering and the physical sciences have been too small, placing U.S. technological and economic leadership at risk. Looked at in isolation rather than in comparison with the rapid growth in funding for the biomedical sciences, the federal investment 62President’s Information Technology Advisory Committee (PITAC), in “Executive Sum- mary,” Information Technology Research: Investing in Our Future, Report to the President, February 24, 1999, available at http://www.nitrd.gov/pitac/report/exec_summary.html; accessed June 27, 2007; see also from the same report: Section V, “Creating an Effective Management Structure for Federal IT R&D,” available at http://www.nitrd.gov/pitac/ report/section_5.html; accessed June 27, 2007. 63National Academy of Engineering, Engineering Research and America’s Future: Meeting the Challenges of a Global Economy, The National Academies Press, Washington, D.C., 2005. National Academy of Sciences, National Academy of Engineering, and Institute of Medicine, Rising Above the Gathering Storm: Energizing and Employing America for a Brighter Economic Future, The National Academies Press, Washington, D.C., 2007.

136 assessing the impacts of changes in the it R&D ecosystem 60 “Other” In billions of constant 2007 dollars Social Science Engineering 45 Math and Computer Science Environmental Science Physical Sciences 30 Psychology All Other Life Science 15 National Institutes of Health Biomedical Sciences 0 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 Year Figure 4.4 Federal funding for basic and applied research, by field, 1970-2006. Source: Computing Research Association. Figure 4-4.eps in IT R&D has also enjoyed a generous increase in the past two decades (see Figure 4.5). However, this increase must be calibrated by several factors, including the enormous and increasing importance of the field, the continued potential for high-impact breakthroughs, and the nation’s investment in other fields. As Figure 4.4 clearly shows, not only does the federal investment in IT R&D included in “Math and Computer Science” pale in comparison with the investment in “Biomedical Sciences,” but it is smaller than the investment in “All Other Life Science,” “Engineering,” “Physical Sciences,” and “Environmental Science”—exceeding only the investment in “Psychology” and “Social Science”! Table 4.5 Funding Increases for IT R&D Recommended by the President’s Information Technology Advisory Committee, FY 2000- FY 2004 ($ millions) Fiscal Year (FY) Recommended Increase FY 2000 472 FY 2001 733 FY 2002 976 FY 2003 1,192 FY 2004 1,370 SOURCE: President’s Information Technology Advisory Committee (PITAC), “Executive Summary,” in Information Technology Research: Investing in Our Future, Report to the Presi- dent, February 24, 1999, available at http://www.nitrd.gov/pitac/report/exec_summary. html; accessed June 27, 2007.

a globalized, dynamic information technology R&D ecosystem 137 3,500 In millions of adjusted FY 2000 dollars 2,625 1,750 875 0 2007 (req) 2008 (req) 2002 1978 1982 2000 2006 1976 1988 1980 1992 1988 2004 1986 1990 1996 1984 1994 Fiscal Year Figure 4.5 Federal funding for information technology research and develop- ment, fiscal years 1976-2008. NOTE: (req), requested. Source: Computing Research Association. Figure 4-5.eps Indeed, America’s public investment in civilian (nonmilitary) R&D in information technology communication lags other economies of inter- est in absolute dollar terms (see estimates developed by the Institute for Defense Analyses for the President’s Council of Advisors on Science and Technology in Table 4.6). Table 4.6 Comparisons of European Union, Japanese, and U.S. Estimated Public Funding of Civilian Information Technology and Communications Research and Development (in billions of dollars, purchasing power parity) Year European Union-15 Japan United States 1999 2.7 1.9 1.2 2000 2.9 2.1 1.3 2001 3.0 2.3 1.6 2002 3.3 2.5 1.5 2003 3.3 2.6 1.7 2004 3.4 2.7 1.9 2005 3.5 2.7 1.8 Source: Institute for Defense Analyses, Science and Technology Policy Institute, brief- ing to the President’s Council of Advisors on Science and Technology, January 9, 2007. Based on data from “Research and Development in Information Science and Technology in Large Industrialised Countries,” Commissioned by the Ministère délégué à l’Enseignement s ­ upé­rieur et à la Recherche, Summary Report, April 2006.

138 assessing the impacts of changes in the it R&D ecosystem According to the Bureau of Economic Analysis (BEA) of the U.S. Department of Commerce, U.S. gross domestic product (GDP) was about $13,247 billion in 2006. Of this, some $521 billion (almost 4 percent) is attributed to what BEA classifies as the “information-communications- technology (ICT) producing industries.”64 Furthermore, this sector— sparked and fueled by IT R&D—experienced double-digit real growth for the third consecutive year in 2006, increasing by 12.5 percent. 65 In 2006, these industries accounted for about 4 percent of the economy but contributed 14.2 percent of real GDP growth.66 Table 4.7 shows a differ- ent measure of the sector’s economic contribution: its contribution to real value added (real value added captures the contribution of an industry’s labor and capital to real GDP). These contributions, although substantial, reflect only a portion of the overall long-term benefits from IT research investments. Organization of University Research Federal funding for university research in information technology has traditionally followed a model of a three-legged stool (see Box 4.3 for a quick view into one university’s funding sources and patterns). One leg consisted of modest grants provided by the National Science Foundation67 and the Defense Science Offices (Office of Naval Research, Army Research Office, and Air Force Office of Scientific Research) to single investigators to work primarily on fundamental research problems. These were either peer-reviewed or evaluated by a panel drawn from technical experts within the government. The grants were sufficient to fund one to two students to work on a research problem. 64According to the BEA, the ICT-producing industries consist of the following: com- puter and electronic products within durable-goods manufacturing; publishing industries (including software) and information and data processing services within information- producing industries; and computer systems design and related services within profes- sional, scientific, and technical services. See Thomas F. Howells III and Kevin B. Barefoot, “Annual Industry Accounts—Advance Estimates for 2006,” Survey of Current Business, Table 1, May 2007, Bureau of Economic Analysis, Washington, D.C., available at http://bea. gov/scb/pdf/2007/05%20May/0507_annual_industry_accounts.pdf; accessed August 28, 2007. 65See ibid., Table B. 66See ibid., Table A. 67NSF’s Directorate for Computer and Information Science and Engineering (CISE) sup- ports research in three broad areas: computing and communication foundations, computer and network systems, and information and intelligent systems. Other IT-relevant funding sources include the NSF’s Office of Cyberinfrastructure and initiatives within the Engineer- ing Directorate.

a globalized, dynamic information technology R&D ecosystem 139 Table 4.7 Percentage Changes in and Real Value Added to U.S. Gross Domestic Product, by Industry Group, 2003-2006 2003 (%) 2004 (%) 2005 (%) 2006 (%) U.S. gross domestic product 2.5   3.9   3.2   3.3 Private industries (overall) 2.7   4.2   3.3   3.7 Information-communications- 7.2 13.7 13.3 12.5 technology (ICT)-producing private industries Source: Data from Thomas F. Howells III and Kevin B. Barefoot, “Annual Industry Accounts—Advance Estimates for 2006,” Survey of Current Business, Table B, May 2007, Bureau of Economic Analysis, Washington, D.C., available at http://bea.gov/scb/pdf/ 2007/05%20May/0507_annual_industry_accounts.pdf; accessed August 28, 2007. As a second leg of this model and at the opposite extreme from mod- est grants to single investigators, the NSF also funded larger-scale, theme- oriented research endeavors through such programs as Engineering Research Centers and Science and Technology Centers. The Department of Defense (DOD) developed the Multidisciplinary University Research Initiative (MURI) Program for similar purposes. These programs were intended to receive support for relatively long periods of time—5 to 10 years, rather than 2 or 3 years for single-investigator grants—and often involved further requirements in terms of industry or institutional match- ing support. Such centers could encompass the research activities of two dozen faculty members or more, with the result that funding was thinly spread and best used to support work at the intersection of individual investigators’ interests. Critical-mass research efforts necessary to achieve breakthroughs were difficult to achieve. The third leg, uniquely epitomized by Defense Advanced Research Projects Agency (DARPA) support, was critical-mass funding for small teams of faculty and their graduate students: 5 to 6 investigators plus 15 to 20 graduate students. The level of funding was comparable with and sometimes exceeded that of an NSF center, but it was focused on the research activity of a much smaller group. Furthermore, such efforts were not pursued in a vacuum but in the context of a program (see Box 4.4). These efforts consisted of perhaps a dozen similarly sized teams, span- ning universities and industry, developing competing technologies but also cooperating on developing a common underlying infrastructure— including, importantly, a research community in an area of strategic need. Examples include the DARPA VLSI Project and the High Performance

140 assessing the impacts of changes in the it R&D ecosystem Box 4.3 The Changing Sources of Information Technology R&D Funding The Computer Science and Artificial Intelligence Laboratory (CSAIL) at the Massachusetts Institute of Technology (MIT) is one of the premier information technology university laboratories in the world. With 93 principal investigators, 471 graduate students, 112 research staff, 46 other staff, and a $45 million per year research expenditure, it is without question a large research enterprise. The laboratory has long enjoyed high levels of research support from the Defense Advanced Research Projects Agency (DARPA). Table 4.3.1 shows the percentage breakdown of funding sources for the MIT laboratory’s activities between 2000 and 2008.1 The data show a dramatic decrease in the percentage of DARPA funding, matched by a similarly large in- crease in funding from the National Science Foundation (NSF). By the end of the period, the laboratory’s funding base is more balanced than in 2000, with roughly equal portions from nongovernment sources (mostly industry), NSF, and the Department of Defense (DOD). In 2000, DOD provided almost two-thirds of the laboratory’s funding. TABLE 4.3.1 Percentage of Funding for MIT’s Computer Science and Artificial Intelligence Laboratory, 2000-2008, by Source 2000 2001 2002 2003 2004 2005 2006 2007 2008 Source (%) (%) (%) (%) (%) (%) (%) (%) (%) Nongovernment 28.3 33.0 43.2 46.7 39.5 33.1 32.8 32.1 30.8 Government 71.7 67.0 56.8 53.3 60.5 66.9 67.2 67.9 69.2 NSF 7.5 7.9 9.9 15.3 22.9 25.3 26.8 26.7 27.4 DOD Total 62.9 54.2 43.6 33.4 29.7 28.6 24.3 27.8 29.7 DARPA 51.6 47.9 37.9 26.6 25.6 25.6 19.6 23.1 24.2 Other U.S. 1.3 4.9 3.3 4.6 7.9 13.0 16.1 13.4 12.1 Government Source: Rodney Brooks, Massachusetts Institute of Technology, “IT Research Funding: An MIT CSAIL Perspective,” presentation to the committee, Boston, Mass., April 19, 2007. Table 4.3.1 in Box.eps Updated and corrected percentages provided to the committee by personal communication from Rodney Brooks, July 15, 2008. One dimension of the data not made obvious in this table is the increas- ing level of support from foreign firms for MIT’s research. Quanta Computer, a major manufacturer of personal computers based in Taiwan, has entered into a long-term, $20 million research agreement with MIT to investigate what will come “beyond the notebook computer.”2 Nokia, a major manufacturer of telecommunica-

a globalized, dynamic information technology R&D ecosystem 141 Box 4.3 continued tions equipment based in Finland, has established a research laboratory close to MIT to pursue collaborative activities.3 Clearly, even support for university research is becoming globalized. 1Rodney Brooks, Massachusetts Institute of Technology, “IT Research Funding: An MIT CSAIL Perspective,” presentation to the committee, Boston, Mass., April 19, 2007. 2See “Quanta Computer, Inc. and the Massachusetts Institute of Technology Announce TParty Project–CSAIL Spotlight,” http://www.csail.mit.edu/node/363; accessed December 11, 2008. 3See “Nokia and the Massachusetts Institute of Technology Celebrate the Opening of Nokia Research Center Cambridge,” April 21, 2006, available at http://press.nokia.com/ PR/200604/1046070_5.html; accessed August 24, 2007. Computing and Communications Program of the late 1980s through the 1990s. This program type of organization was essential in transitioning fundamental research to a size and scale of proof of concept that the rest of the ecosystem could then begin to commercialize.68 In 2000, NSF introduced the Information Technology Research (ITR) Program to provide a large-grant funding mechanism. The program did not, however, provide the same sort of programmatic context that DARPA has been able to provide. Thus, the research teams were not organized in a way that enabled them to achieve even better results through the process of competition, cooperation, shared infrastructure, and research community formation. With DARPA’s shift away from its traditional sup- port for university-based information technology research in this decade, this third leg of the stool, critical for the field’s success in the past, has largely been lost. In 2007, however, the NSF Directorate for Computer and Information Science and Engineering (CISE) made a small but positive step forward in this regard, with the new Expeditions in Computing Program, which is 68For more on DARPA’s early and continuing roles in IT, see “Happy Birthday, Sputnik! (Thanks for the Internet),” Computerworld, September 24, 2007, available at http://computer­ world.com/action/article.do?command=viewArticleBasic&articleId=9036482&pageNumber=1; accessed October 18, 2007.

142 assessing the impacts of changes in the it R&D ecosystem Box 4.4 The Role of the Defense Advanced Research Projects Agency in the Organization of Information Technology R&D In the early years of the information technology (IT) industry in the United States, the Department of Defense (DOD) played a crucial role, as a supporter of research and as a sophisticated procurer of IT systems. In the 1960s and 1970s, the DOD pulled forward such strategic IT sectors as integrated circuits, computer- aided design software, time-sharing systems, and packet switching networks (i.e., the Advanced Research Projects Agency network, or ARPAnet). The science offices of the military services—the Office of Naval Research (ONR), the Army Research Office (ARO), and the Air Force Office of Scientific Research (AFOSR)—have a long history of supporting fundamental research related to DOD missions. Further, the DOD maintains its own establishment of research laboratories, to develop specific prototype defense capabilities while also evaluating concepts from the defense contractor community. In terms of organizing research outside the DOD, the major funder of IT research and advanced devel- opment has traditionally been DARPA: the Defense Advanced Research Projects Agency (originally “ARPA”).1 Formed in the late 1950s in the wake of the Soviet launching of Sputnik and the American public furor that followed,2 the agency has acquired an almost magical reputation for establishing ambitious research goals, organizing research communi- ties, and executing programs that expand the technology base to demonstrate new military capabilities. The agency’s unofficial charter is to “avoid future technological surprise.” Its modus operandi is critical-mass funding to support project teams, or- ganized into cooperative and competitive multiteam programs, under the direction of an empowered program manager (PM) who stands as the mediator between the researchers on the one hand and the DOD customers on the other.3 Many within the IT research community point to the late 1980s and early 1990s as the high-water mark of DARPA support for the field. The mid-1980s saw the emergence of DARPA’s Strategic Computing Program (SCP) to apply artificial intelligence (AI) techniques to DOD applications in autonomous vehicles, in fleet battle management, and in a pilot’s associate. In addition to the demand pull that these stressing applications placed on speech understanding, computer vision, user interfaces, and planning systems, they also put stretch demands on the un- derlying networked hardware and software systems on which they would execute. That is, these applications’ requirements pushed the state of the art in these fields and also required more capabilities in the underlying systems. Therefore, SCP represented a very significant increment in defense funding for IT research.4 1For more on DARPA (originally ARPA) management style, see National Research Coun- cil, Funding a Revolution: Government Support for Computing Research, National Academy Press, Washington, D.C., 1999, pp. 98-105. 2Roger D. Launiusk, “Sputnik and the Origins of the Space Age,” available at http://history. nasa.gov/sputnik/sputorig.html; accessed March 27, 2008. 3See “Strategic Vision,” available at http://www.darpa.mil/stratvision.html; accessed January 7, 2009. 4Alex Roland and Philip Shiman, Strategic Computing: DARPA and the Quest for Machine Intelligence, 1983-1993, MIT Press, Cambridge, Mass., 2002.

a globalized, dynamic information technology R&D ecosystem 143 Box 4.4 continued The High Performance Computing Act of 1991 (Public Law 102-194) was motivated in part by a 1988 report of the National Research Council, Toward a National Research Network.5 Key outcomes of the act were the creation of the federal High Performance Computing and Communications (HPCC) Program 6 and the establishment of a mechanism to coordinate research in communications across the science and technology agencies of the government.7 Although some agencies saw their funding for HPCC-related research increase owing to the act, much of the work that came under the HPCC rubric was already being carried out by federal agencies. The multiagency focus combined with the HPCC Program’s high visibility and compelling stretch performance goals are credited with motivating a whole generation of researchers to enter the field and contribute to HPCC’s success. The size and diversity of the research program grew significantly, encompassing more universities and more firms. A perhaps less well known outcome was that the High Performance Computing Act has fostered collaborative work in which a small number of research administrators within these diverse and often competitive organizations have worked together to rationalize their research and development investments in order to maximize leverage and minimize overlap of effort and to promote and publicize their scientific and technical accomplishments. On the nega- tive side, some have observed that the size of the program attracted the attention of lobbyists, who sought to influence procurements, and of legislators, who sought to earmark funds for projects within their constituencies. Within the DOD, SCP evolved from an AI program with a modest comput- ing component to a major program in HPCC. The program laid the foundation for today’s scalar cluster-based processors and storage systems on which virtually every major Web site depends. By the mid-1990s, DARPA deemphasized its in- vestment in high performance computing, with the technical leadership shifting to the Department of Energy Accelerated Strategic Computing Initiative (ASCI) Program.8 The ASCI Program focused on developing very large scale parallel 5National Research Council, Toward a National Research Network, National Academy Press, Washington, D.C., 1988. This publication is sometimes referred to as the Kleinrock report, after the authoring committee’s chair, Leonard Kleinrock. 6See, for example, D.B. Nelson, “High Performance Computing and Communications Pro- gram,” Proceedings of the 1992 ACM/IEEE Conference on Supercomputing, Minneapolis, Minn., 1992. 7Membership in the program, now known as the Networking and Information Technology Research and Development Program, has expanded over the years. The current members are the Agency for Healthcare Research and Quality, the Defense Advanced Research Projects Agency, the National Nuclear Security Agency, the Office of Science of the Department of Energy, the Environmental Protection Agency, the National Archives and Records Administra- tion, the National Aeronautics and Space Administration, the National Institutes of Health, the National Institute of Standards and Technology, the National Oceanic and Atmospheric Administration, the National Security Agency, the National Science Foundation, and the Offices of the Deputy Under Secretary of Defense (Science and Technology) and Director of Defense Research and Engineering of the Department of Defense. 8Department of Energy, Defense Programs, Accelerated Strategic Computing Initiative (ASCI) Program Plan, DOE/DP-99-000010592, Washington, D.C., January 2000. continued

144 assessing the impacts of changes in the it R&D ecosystem BOX 4.4 continued machines targeted for the department’s nuclear weapons design needs (known as stockpile stewardship). Most of its funding was directed to its contractor- managed weapons laboratories—Lawrence Livermore National Laboratory, Los Alamos National Laboratory, and Sandia National Laboratories—and the machine vendor community. The university HPCC community found it increasingly difficult to receive critical funding to sustain the project teams that had been formed during the earlier stages of HPCC, particularly in areas of computer architecture, parallel software, and internetworking. In 1998 the Next Generation Internet Research Act (Public Law 105-305) was passed, broadening the scope and name of the pro- gram to the Networking and Information Technology Research and Development (NITRD) Program. Today, the NITRD Program and the National Coordination Office for NITRD are together the major coordinating umbrella for IT research within the federal government. Two major characteristics of DARPA-sponsored IT research between the 1960s and 1980s contributed to its success. The first was DARPA’s particular style of project-focused research, mentioned above, typically spanning teams of four to five faculty investigators and their students (although teams also included industrial participants), organized into programs in which the teams are driven to cooperate and/or compete through the oversight of the PM. The PM served as a critical intermediary between the researchers and the military customer, plac- ing the research results in the relevant military context while also expressing the military needs in a language that the researchers could understand. The second characteristic was the recognition that it is often just as strategic to build a research community, such as one skilled in developing software for new parallel architec- tures, as it is to develop the particular technologies that such a community might invent. These characteristics of DARPA successes suggest that simply increasing funding without such a programmatic structure will not yield an ecosystem that is as effective as it was during the past. intended to provide longer-term research support for teams. The program currently has a total budget of $30 million. Each expedition will be funded at up to $2 million per year for 5 years, and CISE estimates that it will provide three new awards each year.69 69According to CISE, “The intent is to provide the opportunity to pursue ambitious, fun- damental research agendas that promise to define the future of computing and information. In planning Expeditions, investigators are encouraged to come together within or across departments or institutions to combine their creative talents in the identification of compel- ling, transformative research agendas that promise disruptive innovations in computing and information for many years to come.” See “Expeditions in Computing,” September 13, 2007, available at http://www.nsf.gov/pubs/2007/nsf07592/nsf07592.txt; accessed October 23, 2007.

a globalized, dynamic information technology R&D ecosystem 145 Changes in the Relationship Between Employees and Employers The foundations of the American system of employment were con- ceptualized under the New Deal and institutionalized in law and by col- lective bargaining agreements. In return for employees’ loyalty and best efforts, employers agreed to fulfill both legally and culturally prescribed obligations: a reasonable expectation of job security and such benefits as health insurance and pension plans. Beginning in the mid-1980s, the cul- tural contract between worker and employer began to unravel as employ- ment practices and policies shifted toward a laissez-faire philosophy remi- niscent of the 19th and early 20th centuries.70 This context is important when considering the IT workforce issues and patterns of student enroll- ments discussed previously. A number of developments contributed to this unraveling. The first was “downsizing” or “rightsizing,” euphemisms for what had formerly been known as “layoffs.” Until the mid-1980s most layoffs occurred dur- ing recessions or when firms found themselves in financial trouble. Lay- offs were primarily confined to blue-collar and clerical workers, who often returned to work once the economy improved. In hard times, pro- fessionals and managers could assume that they were safe even from temporary layoffs. During the 1980s, the rules of the game changed. For the first time in history, firms began to shed professional, technical, and managerial workers in large numbers. In fact, by the mid-1990s corporate downsiz- ings were more likely to target managers and professionals than to dis- miss other white-collar or blue-collar workers.71 Moreover, downsizings, 70The unraveling of the New Deal employment system has been exetensively documented in Thomas A. Kochan, Harry C. Katz, and Robert B. McKersie, The Transformation of American Industrial Relations, Basic Books, New York, N.Y., 1986; Peter Cappelli, Laurie Bassi, Harry Katz, David Knoke, Paul Osterman, and Michael Useem, Change at Work, Oxford, New York, N.Y., 1997; Paul Osterman, Broken Ladders: Managerial Careers in the New Economy, Oxford University Press, New York, N.Y., 1996; Paul Osterman, Securing Prosperity: The American Labor Market: How It Has Changed and What to Do About It, Princeton University Press, Prince­ ton, N.J., 1999; and Paul Osterman, Thomas A. Kochan, M. Locke Richard, and Michael J. Piore, Working in America: Blueprint for the New Labor Market, MIT Press, Cambridge, Mass., 2001. 71On the nature and extent of downsizing, see Thomas S. Moore, The Disposable Work Force, Aldine, New York, N.Y., 1996; American Management Association, 1996 AMA Survey on Downsizing, Job Elimination and Job Creation, New York, N.Y., 1996; and Harry S. Farber, “The Changing Face of Job Loss in the United States: 1981-1995,” pp. 55-142 in Brookings Papers on Economic Activity: Microeconomics, The Brookings Institution, Washington, D.C., 1977. On the impact of downsizing on employee perceptions, attitudes, and lives, see Katherine Newman, Falling from Grace: The Experience of Downward Mobility in the American Middle Class, Vintage, New York, N.Y., 1989; Charles Heckscher, White-Collar Blues, Basic Books, New York, N.Y.,

146 assessing the impacts of changes in the it R&D ecosystem unlike layoffs of the past, seemed independent of economic cycles. Under pressure from global competition and their stockholders, firms had dis- covered that streamlining the workforce was necessary to achieve their bottom-line targets and boost their stock prices.72 By the late 1990s three practices were augmenting and exacerbating the downsizing: the outsourcing of work to external suppliers, the off- shoring of jobs, and the use of contingent labor. Contingent workers are individuals hired, often through staffing agencies, for a limited period of time to perform specific work. Although firms have long employed temporary workers for seasonal and short-term needs, during the late 1980s corporations began to view temporary labor as an extension of the broader strategy of outsourcing. The shift from permanent to contingent employment became particularly widespread in IT centers and among high-technology firms.73 The Bureau of Labor Statistics reported that, by 1995, 40 percent of all programmers and 29 percent of other IT work- ers were either contingently employed or worked through outsourcing firms.74 In Silicon Valley, contractors often comprise between 15 and 30 percent of the labor force.75 Data on employment turnover are consistent with the demise of employment security and stable relations between employers and employ- ees. Between 1983 and 2004, average tenure with one’s current employer fell by 2.1 years (from 7.3 to 5.2 years) among men between the ages of 35 and 44. Among men between 45 and 54 and between 55 and 65 years of age, the declines were greater: 3.2 years (from 12.8 to 9.6 years) and 5.5 years (from 15.3 to 9.8 years), respectively.76 The combination of downsizing, outsourcing, offshoring, and contin- gent work dramatically altered the tenor of the employment relationship. The first casualty was loyalty. Despite stable levels of job satisfaction and 1995; Denise Rousseau, Psychological Contracts in Organizations, Sage Publications, Thou- sand Oaks, Calif., 1995; and Denise Rousseau and R.J. Anton, “Fairness and Obligations in Termination Decisions: The Role of Contributions, Promises and Performance,” Journal of Organizational Behavior 12(4):287-299, 1991. 72Wayne F. Cascio, Clifford E. Young, and James R. Morris, “Financial Consequences of Employment-Change Decisions in Major U.S. Corporations,” Academy of Management Journal 40(5):1175-1189, 1997. 73Stephen R. Barley and Gideon Kunda, Gurus, Hired Guns and Warm Bodies: Itinerant Ex- perts in a Knowledge Economy, Princeton University Press, Princeton, N.J., 2004. 74Angela Clinton, “Flexible Labor: Restructuring the American Workforce,” Monthly Labor Review 120(8):3-27, 1997. 75Chris Benner, Work in the New Economy: Flexible Labor Markets in Silicon Valley, Blackwell, Malden, Mass., 2002. 76Bureau of Labor Statistics (BLS), Employee Tenure in 2004, USDL-04-1829, BLS, Washing- ton, D.C., 2004, available at http://www.bls.gov/news.release/History/tenure_09212004. txt.

a globalized, dynamic information technology R&D ecosystem 147 a vibrant economy, employees became increasingly distrustful of their employers and less sanguine about their future over the 1990s.77 Employ- ers, for their part, dropped the pretense of hiring with any expectation of a long-term relationship. Some openly cautioned new hires about the firm’s limited commitment to them. Apple’s human resource policy, which was reputedly given to each new hire, stated: Here’s the deal Apple will give you; here’s what we want from you. We’re going to give you a really neat trip while you’re here. We’re going to teach you stuff you couldn’t learn anywhere else. In return . . . we expect you to work like hell, buy the vision as long as you’re here. . . . We’re not interested in employing you for a lifetime, but that’s not the way we’re thinking about this. It’s a good opportunity for both of us that this is probably finite.78 The second and more important casualty of the altered employment relationship has been the integrity of America’s system for insuring the health and welfare of the workforce. During the New Deal the govern- ment and industry reached an agreement on how to care for the sick and elderly: Rather than adopting national and universal health care coverage and pension funds, Americans would receive health insurance and pen- sions through their employers. The employment relationship thus became the cornerstone of America’s social safety net, but as the health care costs and pension obligations have risen and as job security has fallen, an increasing number of employers have ceased providing either benefit to workers. Between 1979 and 2004, the percentage of Americans with employer-provided health insurance fell from 69 percent to 56 percent. The rate of decline has been even steeper for Hispanic Americans.79 While the trends are not specific to the IT industry, not only is it not immune to them, but the fast-changing nature of IT businesses, their rapid globaliza- tion, and the need for maximal flexibility of operations has exacerbated these trends in the IT industry. Trends in pension funds are equally striking.80 Between 1983 and 2004, the percentage of American workers covered only by a defined- 77National Research Council, The Changing Nature of Work: Implications for Occupational Analysis, National Academy Press, Washington, D.C., 1999. 78Barbara Ettorre, “The Contingency Workforce Moves Mainstream,” Management Review 83(2):10-16, 1994, quoting from Apple Computer’s written employment contract with every full-time employee. 79Lawrence Mishel, Jared Bernstein, and Sylvia Allegretto, The State of Working America, An Economic Policy Institute Book, Cornell University Press, Ithaca, N.Y., 2007, Figure 3H. 80Data on participation in pension plans are from Alicia H. Munnell and Annika Sunden, “401K Plans Are Still Coming Up Short,” in Issues in Brief, Center for Retirement Research, Boston College, Boston, Mass., 2006.

148 assessing the impacts of changes in the it R&D ecosystem benefit plan fell from 62 percent to 20 percent. Conversely, during the same period the percentage of the workforce covered only by a defined- contribution plan grew from 12 percent to 63 percent. As of 2004, one-fifth of working Americans who were eligible to contribute to defined contri- butions made no contribution whatsoever. Less than 1 percent of work- ers earning less than $60,000 annually contribute the maximum. Among those earning between $60,000 to $80,000 annually, only 8.3 percent make maximum contributions. In fact, only 58 percent of Americans who make more than $100,000 a year contribute maximally. The situation among technical contractors is at least equally dire, if not more so. Although the well-educated and well-paid high-tech contractors whom Barley and Kunda81 interviewed were mostly in their 40s and 50s, 45 percent had no retirement account whatsoever. Another 20 percent had only an indi- vidual retirement account (IRA). Only 20 percent participated in a 401K or simplified employee pension (SEP) plan. 81Stephen R. Barley and Gideon Kunda, Gurus, Hired Guns and Warm Bodies: Itinerant Ex- perts in a Knowledge Economy, Princeton University Press, Princeton, N.J., 2004.

Next: 5 Findings and Recommendations »
Assessing the Impacts of Changes in the Information Technology R&D Ecosystem: Retaining Leadership in an Increasingly Global Environment Get This Book
×
Buy Paperback | $62.00 Buy Ebook | $49.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

The U.S. information technology (IT) research and development (R&D) ecosystem was the envy of the world in 1995. However, this position of leadership is not a birthright, and it is now under pressure. In recent years, the rapid globalization of markets, labor pools, and capital flows have encouraged many strong national competitors. During the same period, national policies have not sufficiently buttressed the ecosystem, or have generated side effects that have reduced its effectiveness. As a result, the U.S. position in IT leadership today has materially eroded compared with that of prior decades, and the nation risks ceding IT leadership to other nations within a generation.

Assessing the Impacts of Changes in the Information Technology R&D Ecosystem calls for a recommitment to providing the resources needed to fuel U.S. IT innovation, to removing important roadblocks that reduce the ecosystem's effectiveness in generating innovation and the fruits of innovation, and to becoming a lead innovator and user of IT. The book examines these issues and makes recommendations to strengthen the U.S. IT R&D ecosystem.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!