CRA Bulletin

The New York Times has an article about the challenge computer science has had bringing women into the field. It mentions CRA’s Taulbee Survey of doctorate-granting computing departments, which has reported that women have received a declining share of undergraduate degrees in CS since the mid-1980s, even as they have made gains in other science and engineering fields. This drop has gone hand-in-hand with results from HERI at UCLA’s survey of incoming freshmen at all undergraduate institutions, which had shown that women’s interest in CS as a major declined significantly between 2000 and 2004.

Data collection is underway for this year’s Taulbee Survey. Anecdotal evidence suggests that there has been a turnaround in undergraduate enrollments and it will be interesting to see what role women play in that.

What about beyond the undergraduate level? Between the early 1980s and late 1990s, women received a little under 30 percent of master’s degrees in CS. Since then it has edged up above that. At the doctorate level, the portion of degrees granted to women has increased slowly, and now is around 20 percent. Women also are making slow inroads among faculty. An encouraging sign is that the share of newly hired, tenure-track faculty that are women has increased steadily for the past several years. For more information, see http://www.cra.org/info/taulbee/women.html and http://www.cra.org/info/education/us/women.html

As a career choice, it is worth point out that the earnings gap for women in IT is narrower than in the overall workforce. And while the current economic downturn probably will cause the Bureau of Labor Statistics to revise its figures downward, its most recent workforce projections predicted strong increases among well-paying professional-level IT occupations between 2006 and 2016.

CRA itself works hard to increase the representation of women in CS education and the IT workforce. Its Committee on the Status of Women in Computing Research (CRA-Women) has had several successful programs, including

Distributed Mentoring Project and Collaborative Research Experiences for Undergraduates

Grad Cohort

Career Mentoring Workshops

The series of reports prepared for the NSF by Michael Finn at the Oak Ridge Institute for Science and Education is one of the best sources of information on the share of foreigners who stayed in the U.S. after they received their science and engineering (S&E) doctorates. In his latest report, Finn found that 68% of foreigners who received S&E doctorates in 2000 were still in the U.S. in 2005, including 73% of those who studied computer science (CS). The five-year stay rate was 92% for Chinese students and 85% for Indian students.

Turning to the two-year stay rates, 66% of foreigners on permanent or temporary visas who received S&E doctorates in 2003 were still in the U.S. in 2005, including 73% of those in CS. Of those on temporary visas (the majority of foreign students), 64% of the 2003 class (and 70% of those in CS) were still in the U.S. in 2005.

When you compare results from the past few years, the trend seems to be that stay rates declined slightly in the earlier part of the decade and are now stabilizing. The two-year stay rate for those on temporary visas dropped from 68% among the 1999, 2000 and 2003 classes, to 64% among those who graduated in 2005. However, the one-year stay rate was relatively consistent at between 67 and 69% for the classes that graduated in 2002, 2003 and 2004. Finn also points out that results from an NSF survey have shown a turnaround in the share of foreign doctorate recipients who said that they wished to stay in the U.S. after graduation and of those who had firm commitments that would allow them to stay. In addition, since more foreigners have been coming to the U.S. to study, even a slight drop in stay rates has not affected the overall number of those who have stayed after graduation.

Since foreigners receive over half of computer science (CS) and engineering doctorates, any sign of a decline in stay rates would be a cause for worry. Hopefully, future reports will confirm that they have stabilized.

Finn, Michael G., Stay Rates of Foreign Doctorate Recipients from U.S. Universities, 2005, Oak Ridge, TN: Oak Ridge Institute for Science and Education, 2005. Online at http://orise.orau.gov/sep/files/stayrate07.pdf (180 kb PDF).

Immigrants are a critical part of U.S. higher education, providing over 60% of doctorate recipients in computer sciences and engineering. They also play a critical role in engineering and technology startups, according to a report by researchers at Duke University and University of California, Berkeley. Among successful tech companies founded between 1995 and 2005, about 25% had at least one key founder who was foreign-born. Of these, almost 80% of the companies were in software and innovative/manufacturing-related services. Although many ethnic groups were represented,

Indians have founded more engineering and technology companies in the US in the past decade than immigrants from the U.K., China, Taiwan and Japan combined. Of all immigrant-founded companies, 26% have Indian founders.

The authors also found that the share of tech startups founded by foreigners varied by region and that there was ‘ethnic clustering’. In California, nearly 40% of tech startups had a key founder who was foreign, including 52% of those in Silicon Valley. Chinese founders also concentrated there: 49% of mainland Chinese and 81% of Taiwanese tech founders were located in California.

In all, the authors estimate that tech companies with foreign-born key founders “produced more than $52 billion dollars in 2005 sales and in 2005 had just under 450,000 employees.”

Wadhwa, Vivek, Saxenian, Annalee, Rissing, Ben and Gereffi, Gary, America’s New Immigrant Entrepreneurs: Part I (January 4, 2007); PDF at http://www.kauffman.org/pdf/entrep_immigrants_1_61207.pdf

According to a recent NSF InfoBrief, postdoctoral appointments have become more common among science, engineering and health (SEH) fields. At the same time, the duration of a postdoc has been relatively consistent since the 1970s.

Among SEH doctorates who received their degrees in the past several decades, 38% had held a postdoc appointment. This ranged from a low of 21% in computer/mathematical sciences (CMS) and engineering, to a high of 57% in life sciences. Over time, a postdoc has become increasingly common, with particularly strong increases in CMS and engineering. Among those who had received their CMS doctorate in the past 5 years, 31% had held a postdoc, compared to 16% of those who had received their degrees more than 20 years ago.

The median length of time for a postdoc has hovered at or a little below 2 years since the 1970s. The longest postdoc appointments tend to be in the life sciences (2.2 years). Among CMS PhDs, the median postdoc duration was 1.8 years.

Most SEH postdoc appointments (70+%) were in academic or university research settings, with an additional 12% each in for-profit/non-profit organizations and government. Compared to all SEH fields, a higher share of CMS postdocs were in education (between 80 and 85%) and a lower share (about 7%) were in government.

As the authors of the report point out, relatively few doctorate recipients (11%) claim that they took a postdoc because other employment was not available. Most (34%) cite ‘additional training in doctoral field.’ The majority of CMS postdocs (59%) mention ‘additional training’ and ‘work with a specific person or place.’ Only 10% cited a lack of available employment (compared to 23% of engineers).

National Science Foundation, Division of Science Resources Statistics, Postdoc Participation of Science, Engineering, and Health Doctorate Recipients, Arlington, VA (NSF 08-307), March 2008, http://www.nsf.gov/statistics/infbrief/nsf08307/

According to a recent report sponsored by the Kauffman Foundation, founders of technology companies rarely fit the stereotype of being young and having little formal education. After surveying 500 engineering and technology companies established between 1995 and 2005, the authors found that the median age of U.S.-born tech founders was 39 (with twice as many aged over 50 as were younger than 25) and that 92% of them held at least a bachelor’s degree.

The amount of time that passed between when U.S.-born tech founders received their terminal degrees and when they started their companies varied by the level of degree and field that they had studied. The shortest average period between degree completion and startup creation was held by those who graduated with an MBA (13 years) and the longest was among those with a PhD (21 years). Tech founders who had received their terminal degree in computer science or information technology had the shortest average period between degree completion and startup creation (14 years). Among science, technology, engineering and math (STEM) degree holders, the average interval was 18 years.

Of the 92% of tech founders who held at least a bachelor’s degree, 47% held advanced degrees, including 30% with a master’s and 10% with a PhD. The most common groups of fields for the terminal degrees were business, accounting and finance (33%) and engineering (28%). Overall, 55% of tech founders had received at least one degree in a STEM field. The 628 U.S.-born tech founders had received their terminal degrees from 287 different universities, though graduates of Ivy-League schools were both overrepresented and headed companies that earned more than their counterparts.

Source: Education and Tech Entrepreneurship by Vivek Wadhwa, Richard Freeman, and Ben Rissing, http://www.kauffman.org/pdf/Education_Tech_Ent_061108.pdf

PCAST makes recommendations on science and technology policy to the President. Last August, it issued a review of the Federal government’s $3.1 billion budget for Networking and Information Technology R&D (NITRD): http://ostp.gov/pdf/nitrd_review.pdf

Yesterday, Dan Reed, CRA’s Board Chair and a member of PCAST, discussed the report and other issues with the House Committee on Science and Technology. Information and analysis about the hearing can be found on CRA’s Computing Research Policy Blog, at http://www.cra.org/govaffairs/blog/archives/000694.html

In brief, the report described the networking and information technology (NIT) ‘ecosystem’, its weaknesses, where it is being challenged by other countries, and how to improve it in order to meet those challenges.

The argument made most forcefully in the report was that

Federal agencies, as well as universities and industry, must expand the ways in which NIT R&D is funded and conducted. The R&D will need to include more investigations that are long-term rather than short-term; large in scope and scale instead of focused and small; conducted by multidisciplinary teams that work together for many years and not just by single-discipline researchers or small groups; and/or visionary and innovative rather than solid but incremental. These changes are likely to increase the level of risk in Federal agency NIT R&D portfolios, but have the potential of significant payoffs.

In terms of a research agenda, the report pointed to four areas that should receive larger shares of new funding:

NIT Systems Connected with the Physical World (which are also called embedded, engineered, or cyberphysical systems). The NITRD Subcommittee should develop and implement a Federal Plan for high-confidence NIT systems connected with the physical world.

Software: The NITRD Subcommittee should facilitate efforts by leaders from academia, industry, and government to identify critical issues in software design and development to help guide NITRD planning on software R&D.

Digital Data: The Interagency Working Group on Digital Data, in cooperation with the NITRD Subcommittee, should develop a national strategy and develop and implement a plan to assure the longterm preservation, stewardship, and widespread availability of data important to science and technology.

Networking: The PCAST endorses the ongoing effort to produce a Federal Plan for Advanced Networking Research and Development, expected in 2008, which includes an R&D agenda for upgrading the Internet and R&D in mobile networking technologies, and addresses network security and reliability.

It also argued that four areas of research should continue as priorities: High-End Computing, Cyber Security and Information Assurance, Human-Computer Interaction, and NIT and the Social Sciences.

The authors also made a number of recommendations concerning the education and workforce pipeline. It pointed to continued demand for IT workers in the face of both falling enrollments and declining shares of degrees granted to women and minorities. To help encourage the supply of advanced degrees in NIT fields, it suggested increasing the number of fellowships for graduate study, as well as streamlining the visa process for foreign graduate students to study and then work in the U.S., and for for foreign R&D experts who visit the country.

There are few good sources of information about what happens to undergraduates after they receive their degrees. One is the NCES‘ Baccalaureate and Beyond Longitudinal Study (B&B).

B&B provides snapshots of work and life experiences in 1994, 1997 and 2003 for those who received undergraduate degrees in 1992-93. It divides majors into those that are ‘academic’ or ‘career-oriented’, with computer science (CS) included in the latter (along with business, education, health, and engineering). Sixty-five percent of the 1992-93 graduates had career-oriented majors. CS majors represented a little over 2% of the bachelor’s degree recipients tracked by the survey. B&B also reports results by ‘STEM’ fields, which include engineering, CS, biological sciences, mathematics, and physical sciences.

Here are a few interesting findings about CS majors in the report:

They tended to be older than their fellow undergraduates when they received their degrees.

About 22% of academic majors were aged 25 and older when they graduated, compared to 31% of career-oriented majors. Of all 11 majors or groups of majors reported in the survey, CS had the second highest share of students aged 25 and older (36%).

Few of them studied beyond the baccalaureate level.

When you combine results for each of the three survey years (1994, 1997 and 2003), CS majors tied with business and management majors in having the lowest share of those who had gone on to enroll in additional degree programs– 17%, compared to 25% of career-oriented majors and 39% of STEM majors.

They enjoyed high salaries.

In 2003, those who had received their undergraduate degree in CS earned $72,600. This trailed only engineers, who earned $74,600. CS also compared well to majors grouped into the STEM ($68,300) and non-STEM fields ($58,900). As a result, it is not surprising that 77% of CS majors reported that they were satisfied with their salaries in 2003, a higher share than any other major (though closely followed by engineering).

They tended to stay in their jobs.

Although IT workers are portrayed as moving between jobs frequently, CS majors in the B&B study reported being with their current employer the longest out of all groups tracked. In 2003, CS majors had been with their current employer an average of 6.2 years, compared to 4.3 years for academic majors, 5.6 for career-oriented majors, 5.1 for STEM and 5.2 for non-STEM majors. In light of this, it is odd that CS majors also reported the lowest satisfaction with their job security (followed by engineers).

Their skills were in demand among a variety of work sectors.

To quote the report: “Although 48 percent of computer science majors were employed in computer science occupations in 2003, graduates with this major were spread across many industries, reflecting the broad demand for their skills.”

Source: Choy, S.P., and Bradburn, E.M. (2008). Ten Years After College: Comparing the Employment Experiences of 1992–93 Bachelor’s Degree Recipients With Academic and Career-Oriented Majors (NCES 2008-155). National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. Washington, DC. http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2008155

The goal of the Computing Community Consortium (CCC) is to catalyze the computing research community to debate longer range, more audacious research challenges; to build consensus around research visions; to evolve the most promising visions toward clearly defined initiatives; and to work with the funding organizations to move challenges and visions toward funding initiatives.

Discussion of “longer range, audacious research challenges” is the job of the entire community, and so the CCC is seeking ways to encourage and facilitate broad community participation. Thus, the CCC will support and help to organize many venues for this: at workshops and conferences, in publications, and on its own web site at http://www.cra.org/ccc.

The purpose of the CCC blog (http://www.cccblog.org/) is to provide a more immediate, online mechanism for dissemination of visioning concepts and community discussion/debate about them. Initially, this will be run as a group weblog, with volunteer academic and industrial leaders providing commentary and their own opinions about new ideas and developments for future research. We invite anyone in the community to participate — the easiest way to start is simply to provide comments on the articles as they are posted. In time, we hope also to evolve the members of the contributing author group.

Please bookmark the site (http://www.cccblog.org/), or use your favorite RSS feed reader to keep up on the latest articles as they are posted. For those of you who are email-bound, you may also subscribe on the CCC blog site for email delivery of new articles.

This article reports on Ph.D. production and graduate enrollments among computer science (CS) departments in the United States. It draws on data collected by CRA’s Taulbee Survey of doctorate-granting departments.

In the early- and mid-1990s, the number of doctorates granted annually by US CS departments peaked at between 1,000 and 1,100. From the mid-1990s until a few years ago, annual doctorate production hovered at around 800 (Figure 1). At the same time, however, the number of new students entering doctorate programs increased rapidly– their number nearly tripled between 1995 and 2002 (Figure 2). In turn, this led to a near 90% increase in enrollment between 1995 and 2005. The result has been dramatic: doctorate production doubled between 2003 and 2007, to over 1,500.

What can be expected in the next few years? There are signs that the surge in degree production is close to peaking and that the number of doctorates granted will decline somewhat. The number of new students entering doctorate programs has declined in each of the past four years, and is now 17% lower than it was at its peak 2002. Likewise, total enrollments in doctorate programs have leveled off since 2004, and the number of students passing qualifying exams dropped nearly 30% between 2005 and 2007. Overall, it is too soon to tell if degree production will edge back to levels seen in the mid-1990s or if a new, higher level of production will become the norm.

This article reports on the proportion of women who either have received degrees from or are on the faculty of computer science and engineering ("CS/CE") departments in the United States and Canada. It draws on data compiled from CRA’s Taulbee Survey of doctorate-granting CS/CE departments.

Table 1 shows the percentage of bachelor’s, master’s and doctoral degrees granted to women since 1985. Although the gender of Ph.D. recipients has always been tracked by the Taulbee Survey, questions about the gender of bachelor’s and master’s degrees recipients have only been included since 1994. The results from Taulbee are compared with data from National Science Foundation surveys on science and engineering (S&E) degrees. The most recent results for NSF data are from 2005 (no data were reported at the bachelor’s and master’s degrees levels for 1999).[1]

Table 2 focuses on the percentage of bachelor’s and master’s degrees in CS/CE that have been granted to women. Although the Taulbee results combine CE and CS results, while NSF figures reflect CS degrees only, the inclusion of CE data has little impact on the ratio of men to women in Taulbee’s results. Table 2 shows that the Ph.D.-granting departments targeted by the Taulbee Survey grant a lower proportion of bachelor’s and master’s degrees to women than the much broader range of schools that are surveyed by NSF. Unfortunately, both CRA and NSF surveys report that the share of undergraduate CS degrees granted to women has been declining since the mid-1980’s (see also http://www.cra.org/info/education/us/women.html).

Table 3 looks at the percentage of faculty, both current and those newly hired, who are women.

[1] Women, Minorities and Persons with Disabilities in Science and Engineering, Tables C-5 and E-1; and National Science Foundation, Division of Science Resources Statistics. 2006. Science and Engineering Doctorate Awards: 2005. NSF 07-305. Susan T. Hill, project officer. Arlington, VA.

This article can also be found on the CRA website, at http://www.cra.org/info/taulbee/women.html