The Pipeline Crisis in Computing Taking the Initiative SIGCSE 2007 Symposium Covington, Kentucky March 9, 2007 Eric Roberts Professor of Computer Science, Stanford University Co-chair of the ACM Education Board Reframing the Issue • All too often, those of us who teach computing have looked at the declining interest in the discipline as an enrollment crisis. • This characterization is self-defeating and makes it harder to attract allies to our cause. • In a typical university, every department wants to increase its enrollment, and we become merely another player in a parochial game of resources. • The real concern is that we have a pipeline crisis in that we are producing far too few graduates to fill the growing number of positions that require computing skills. Judging by demand, we were producing too few graduates even at the top of the boom. • Failure to respond to the pipeline crisis will place significant constraints on the computing industry and compromise national competitiveness. The Looming Pipeline Crisis • The Bureau of Labor Statistics projects much faster growth in computing employment than in other science/engineering areas. A Graphic Indicator of the Shortage Annual Degrees and Job Openings in Broad S&E Fields 160,000 140,000 PhD Master's 120,000 Bachelor's Projected Job Openings 100,000 80,000 60,000 40,000 20,000 Engineering Physical Sciences Mathematical/ Computer Sciences SOURCES: Tabulated by National Science Foundation/Division of Science Resources Statistics; degree data from Department of Education/National Center for Education Statistics: Integrated Postsecondary Education Data System Completions Survey; and NSF/S Earned Doctorates; Projected Annual Average Job Openings derived from Department of Commerce (Office of Technology Policy) analysis of Bureau of Labor Statistics 2002-2012 projections Biological/ Agricultural Sciences RS: Survey of Graphic created by Greg Lavender at the University of Texas. Economic Utility of Disciplinary Degrees Working in the life sciences typically requires a degree in biology or some closely related field, but relatively few biology majors actually end up working in the field. • 80% of workers in the life sciences have degrees in the life sciences. • 14% of graduates with degrees in the life sciences work in those fields. SOURCE: National Science Foundation/Division of Science Resources Statistics, SESTAT (Scientists and Engineers Statistical Data System), 1999, as presented by Caroline Wardle at Snowbird 2002 Economic Utility of Disciplinary Degrees In computing, the pattern of degree production vs. employment is reversed. • 39% of workers in computing have degrees in computing. • 71% of students with degrees computing remain in the field. in These data suggest a significant underproduction of students with computing degrees at the university level. Why Other Sciences Should Be Concerned Though technologyWhile it is the itself ainformation discipline, computational powered revolution is all accelerating, this science serves to advance of science. The country has not yet awakened to the central most scientifically important and role played bypromising computational science and economically research frontiers high-end computing scientific, in the 21st century in willadvanced be conquered by social science, biomedical, and engineering those most skilled with advanced computing research; defense national security; and technologies andandcomputational science industrial innovation. Together with theory applications. But despite the fundamental and experimentation, computational science contributions of computational science to now constitutes pillar” of discovery, security,theand“third competitiveness, scientific enablingstructures researchers to inadequateinquiry, and outmoded within buildFederal and government test models complex the and of the academy phenomena—such as multi-century today do not effectively support this climate critical shifts, multidimensional multidisciplinary field. flight stresses on aircraft, and stellar explosions—that cannot be replicated in the laboratory, and to manage huge volumes of data rapidly and economically. . . . What We Need To Do • Develop greater understanding of the reasons behind the decline in student interest in computing disciplines. • Forge alliances with individuals and groups in other disciplines to bring new voices into the discussion. • Increase public awareness of the range of opportunities. • Press government and industry to support computing education. • Expand efforts to increase diversity. • Encourage experimentation in curricular strategies. • Develop tools and materials that can be used “off the shelf.” • Improve distribution channels for best practices. • Promote interdisciplinary curricular connections. • As Grady Booch encouraged us this morning, help students rediscover the “passion, beauty, joy, and awe” of software Reasons for the Decline 1. Students are insecure about the dot-com bust and offshoring. 2. CS curricula are seen as unexciting and lacking in flexibility. 3. Images of computing work—and workers—are often negative. 4. Students have changed in ways that decrease the appeal of CS. 5. Teaching computing in high school faces growing challenges. 6. Introductory courses have become more difficult to teach. Changes in Student Attitudes or Why Students No Longer Like Programming For much of our field’s history, programming was the most popular aspect of the major. That seems to have changed. • Students have adopted over time an increasingly instrumental attitude toward education. • For many students, opportunities for wealth are more attractive than simply having good prospects for a high-paying job. • A factor analysis by my colleague Mehran Sahami revealed an 88% correlation between the number of CS majors at Stanford and the average level of the NASDAQ the year before. • Students are primarily choosing careers that they perceive to fall on the capital side of the capital/labor divide. Despite the fact that software development is highly paid, it is generally viewed as labor. Some Encouraging Signs Matt Jacobsen, Senior, UC Berkeley And for those programming jobs,many the A common misconception is that reason it’s possible to sitting sit ininfront people think CS means front of of aa computer for day extended periods time be is computer all long. This mayofoften because in programming, CS we can learn newis things, the case for but CS a large achieve goals, and be creative. Every day! field. There are many applications It’s this last point that really drives that me, require CS skills thatask involve or no personally. If you any little passionate programming. . . . can "___ all day long", person how they it’s because that’s their outlet for being creative. From Dan Garcia’s “Faces of CS” web site. More Encouraging Signs • Many large universities have reported significant increases in enrollments this year. Some have recovered much of the loss from the past five years. Dot-Com Boom Echoed in Deal to Buy YouTube By ANDREW ROSS SORKIN Published: October 10, 2006 A profitless Web site started by three 20-somethings after a late-night dinner party is sold for more than a billion dollars, instantly turning dozens of its employees into paper millionaires. It sounds like a tale from the late 1990’s dot-com bubble, but it happened yesterday. Google, the online search behemoth, agreed yesterday to pay $1.65 billion in stock for the Web site that came out of that party—YouTube, the video-sharing phenomenon that is the darling of an Internet resurgence known as Web 2.0. . . . The purchase price has also invited comparisons to the mind-boggling valuations that were once given to dozens of Silicon Valley companies a decade ago. Like YouTube, those companies were once the Next Big Thing, but some soon folded. The Growing Challenge of High School CS • People who have software development skills command high salaries and tend not to teach in high schools for very long. • In many schools, computing courses are seen as vocational rather than academic. The NCAA, for example, no longer accepts computer science courses for academic eligibility. • Students who are heading toward top universities are often advised to take courses other than computer science to bolster their admissions chances. • Because schools are evaluated on how well their students perform in math and science, many schools are shifting teachers away from computer science toward these disciplines. • Teachers have very few resources to keep abreast of changes in the field. CS is Losing Ground in the AP Exam • The Computer Science exam is the only Advanced Placement exam that has shown declining student numbers in recent years. CS Is Tiny Compared with Other Sciences Computing Is Getting Harder Many faculty in our discipline believe that teaching computing has become more difficult. The contributing factors include: • Complexity. The number of programming details that students must master has grown much faster than the corresponding number of high-level concepts. • Instability. The rapid evolution of the field creates problems for computing education that are qualitatively different from those in most fields. Concern over these has sparked several initiatives including the ACM Java Task Force. If I had had to learn C++, I would have majored in music. —Don Knuth, October 11, 2006 Positive Initiatives • The National Science Foundation sponsored four regional conferences on Integrated Computing and Research (ICER) and launched the new Computing Pathways (C-PATH) initiative. • Several ACM Education Board projects are proving helpful: – – – – A brochure for high-school students The CC2001 series of curriculum reports The Computer Science Teachers Association A community effort to develop Java tools (the ACM Java Task Force) • There are many interesting ideas in the community that are showing promise: – – – – – – Mark Guzdial’s “media computation” strategy at Georgia Tech Stuart Reges’s “back to basics” strategy at the University of Washington Jeannette Wing’s “computational thinking” concepts Interdisciplinary curricula at a variety of schools The many efforts to enhance diversity from so many people All the good ideas that come out here at SIGCSE Dangers on the Horizon We have met the enemy and he is us. — Walt Kelly Unfortunately, the sense of crisis in recent years carries with it the risk that our community will adopt desperate measures that are selfdefeating in the long run: • Engaging in resource competition with fields that should be our allies in seeking to increase support of science and technology. • Changing our curricula in ways that might increase the number of students but will not meet the needs of their eventual employers. Every technical person in the industry with whom I’ve spoken is horrified by the prospect of reducing the emphasis on programming in the undergraduate curriculum. • Losing hope in the darkness before the dawn. Enrollments are already recovering in many institutions. This too shall pass, but only if we keep the faith and make it happen. The End The Pace of Change • The pace of change—particularly in terms of its effect on the languages, libraries, and tools on which introductory computer science education depends—has increased in recent years. • Individual universities and colleges can’t keep up. • In a survey by the Computer Science Teachers Association, high-school teachers cited the rapid pace of change as the most significant barrier.