Practice
Computing Applications

The Decline and Fall of the American Programmer

In his 1991 book, The Decline and Fall of the American Programmer, Ed Yourdon wrote: "The American Programmer is about to share the fate of the dodo bird. By the end of this decade, I foresee massive unemployment among the ranks of American programmers, systems analysts, and software engineers. Not because fifth-generation computers will eliminate the need for programming, or because users will begin writing their own programs. No, the reason will be far simpler: international competition will put American programmers out of work, just as Japanese competition put American automobile workers out of work in the 1970s."1
Posted
  1. Article
  2. Author
  3. Footnotes

Did it happen? Well, no, not yet anyway. The late 1990s were characterized not by massive unemployment among programmers worldwide but by considerable shortages of such skilled personnel. This was caused in part by the demands of addressing myriad Y2K problems, particularly those relating to long-term legacy software. As these problems passed, the so-called "skills shortage" showed no sign of abating. Indeed, the demand for programmers, systems analysts, and software engineers is regularly reported by some companies to be in the thousands, and in large multinationals in the tens of thousands. So, for now it seems reports of the decline of the programmer are premature. But what of the future?

It is possible the term "skills shortage" is a misnomer, for there are probably many skills shortages. Discounting the so-called "IT skills" shortages in large sections of the work force in most developed countries, there are currently undoubted shortage areas such as networking technology (a technology that has increased in strategic importance throughout the 1990s), and platform-independent programming languages such as Java (which might be regarded as more transitory). It is difficult to believe the supply of graduates currently produced by our universities will keep pace with demand for graduates in these areas. It is also difficult to imagine that, in the long term, universities will be able to find and retain staff to teach these subjects as demand broadens. Some companies are trying to address these problems by entering into agreements with universities to take students on industrial internships (usually for one year), and to offer staff-limited consultancy opportunities. But given the scale of the problem to be addressed, companies and universities must enter into profoundly different arrangements to solve their respective problems to mutual advantage. For example, in the academic year 1988–1989, only nine computer science students (out of over 100) from the University of Kent accepted one-year industrial placements with U.K. companies. In 1999–2000, 22 students did so (six of them with U.S.-based Sun Microsystems). In 2000–2001, following the promotion of the university’s intern program and with the appointment of a full-time placement manager, over 60 students will take placements. Indeed, Sun Microsystems will take 25 of them. We are also in discussion with another company (seeking 50,000 engineers worldwide over the next five years) about a specialized degree program, tailored to their requirements, to educate large numbers of current and potential employees. Are these initiatives desirable, and will they be effective? In a U.K. context, they either represent a threatening 21st century brain drain or they represent an opportunity for inward investment by U.S. companies in U.K. software engineering skills. More likely, they will turn out to be both since we are now part of a global industry not limited by national boundaries.

These are potentially very significant developments by U.K. standards. In fact, it is possible the skills shortages will become critical before sufficient numbers of graduates can be produced through the traditional format of the U.K. educational system. Yourdon predicted companies would export their programming requirements abroad. India, for example, boasts a highly skilled work force pleased to be employed by major U.K. and U.S. companies (see Gupta’s article in this section). But perhaps unlike Yourdon, I am calmer about the future of the American (and British) programmer because I think U.K. universities, having survived since the 13th century (significantly longer than the industries with which we collaborate), will find a way to solve these problems in the long term. The solution will have three components, and these three components represent the challenges for U.K. computer science education in the next decade.

First, we shall educate more of our work force, increasing our participation rates from 35% of the 18-year-olds in higher education. Perhaps they will not be students in the traditional format (largely residential in a university setting for three years full-time), but more in the style of a modern apprenticeship with a less clear boundary between education and employment. Perhaps some students will enter in their 20s and 30s rather than at 18. To cope with the expansion, universities will have to find some way of diversifying their portfolio of activities, probably through building larger institutions in which staff have specialist roles (such as teaching or research, but probably not both).

Second, the academic practitioner—currently the preserve of the medical profession—will be born in the computing profession. Spanning the university-industry divide in this way will help address the current disparities between academic and industrial salaries, will increase the industrial awareness of staff, and, with so much changing so quickly, will provide a vital conduit for the flow of research ideas and practical applications.

Third, the interaction between universities and companies will be massive. They will be increasingly co-located (in the current style of teaching hospitals), but their co-location could well be virtual as well as physical, with network and communications technology providing the illusion of physical proximity (even from another continent).

The idea of a highly connected industrialized university (or is it universitized industry?) provides many opportunities, not the least of which are resource sharing, knowledge transfer, and the promise of students increasingly motivated by the direct application of their own learning. Such opportunities, however, are linked to the threats that always accompany technical shifts and a market-driven desire to remain at the leading edge of technology when sound educational practice suggests more conservative approaches. Plato deplored reading and writing because it devalued the memory. More recently, many in the U.K. have deplored the expansion of higher education because it opened up to many what had traditionally been the preserve of few. Such are the effects of change, and computer scientists, who by more nature are accustomed than most to embracing change, should welcome these changes. What Yourdon predicted as the decline and fall of the programmer was simply the descent of the rollercoaster we are all riding. We won’t fall off, but it may be a bit bumpy for awhile. Universities and companies just need to hang onto each other.

Back to Top

Back to Top

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More