Kjellquist Blog

The Computing Research Association just published some bleak data on student enrollments and computer science (CS) degrees being awarded. The report is the Taulbee Survey of Ph.D.-granting Computer Science (CS) and Computer Engineering departments in North America and has been conducted annually since 1974

I say bleak (unless you a CS degree) because much of the future strength of US technology will depend on CS graduates. My fear is that we are seeing a continuation of the loss of math and science education in the US

From the CRA bulletin

This article reports on CS bachelor’s degree enrollments and production among Ph.D.-granting departments in the United States since the late 1990s.

I interpret this to mean they are reporting degrees from the higher end schools i.e. “Ph.D.-granting”.

….the percentage of incoming undergraduates among all degree-granting institutions who indicated they would major in CS declined by 70 percent between fall 2000 and 2005.

After seven years of declines, the number of new CS majors in fall 2007 was half of what it was in fall 2000 (15,958 versus 7,915). Nevertheless, the number of new majors was flat in 2006 and slightly increased in 2007. This might indicate that interest is stabilizing.

Following several years of increases, the total number of bachelor’s degrees granted by PhD-granting CS departments fell 43 percent to 8,021 between 2003/2004 and 2006/2007

The sustained drop in total enrollments and student interest in CS as a major suggests that degree production numbers will continue to drop in the next few years.

The bulletin goes on to note

It is important to note that a steep drop in degree production among CS departments has happened before. According to NSF, between 1980 and 1986 undergraduate CS production nearly quadrupled to more than 42,000 degrees. This period was followed by a swift decline and leveling off during the 1990s, with several years in which the number of degrees granted hovered around 25,000. During the late 1990s, CS degree production again surged to more than 57,000 in 2004. In light of the economic downturn and slow job growth during the early 2000s, the current decline in CS degree production was foreseeable.

While I don’t disagree with the historical data, it doesn’t put the total picture in the proper context. With prior declines in CS enrollments and degree production there was not the simultaneous upsurge of China, India, Brazil etc ready to fill the void. Further, the previous reductions noted above occurred in the pre-Internet era when the world wasn’t nearly as flat (in Thomas Friedman sense). The drop in CS graduates will only continue to foster the ongoing out sourcing overseas. And, of course, the out sourcing makes CS less attractive; thus a downward spiral!