Is diversity in computing a moral matter?

n Question: "Is UnderRepresentation Immoral?" Women and some minorities are under-represented in academic computer science and in professional computing more generally. Evidence for this assertion appears elsewhere (e.g., [2]), and we won ' t go into the details here. Instead, we focus on the underlying moral issue: Is there anything wrong (immoral, unfair) with so few women and minorities studying or working in computing? First we clear away two distractions. Most o f the data about this issue focus on the U.S., partly because of the special importance of equality and equal opportunity in American democracy. This paper also focuses on the U.S. Nevertheless, the arguments here can apply to many other countries (see [4, 6] in this special issue). Second, our analysis focuses on women in computing, not under-represented minorities. The circumstances of women and minorities are in general quite different. In computing, some minorities appear to be over-represented in computing and other minorities appear under-represented. We believe, however, that the core ethical issue fairness is important to everyone involved in computing. In this editorial, we explicitly address the issue of under-representat ion for women, knowing that work remains to be done on the details regarding specific minorities. We return to our original question: "Is the under-representation o f women in computer science immoral?" Some Say the Answer Is "No" A straightforward argument (that we disagree with) can be made for dismissing this issue on grounds that the prevailing under-representation is not the result of unfairness. The argument goes as follows: "Yes, there are relatively few women going into computer science. But this is not because women are discriminated against. Rather, they choose not to go into the field, despite the fact that universities and industry are actively recruiting them. I f women simply do not want to go into computing and freely choose not to, then there is nothing unethical about the current situation." This is not a convincing argument. For one thing, it claims that women are not discriminated against in computing on grounds that universities and industry appear to be actively recruiting them. The two, however, are not incompatible. It is possible that universities and industry are making conscious efforts to recruit women while at the same time placing those women in chilly environments in which they are overtly or covertly discriminated against. Official, overt discrimination has become less likely in the U.S. because of legislation and the potential to be sued. Still, there are other ways to make under-represented groups unwelcome and uncomfortable. Nevertheless, for sake of analysis, let's temporarily assume that there is no unfair discrimination in hiring and no conscious effort to discourage women from entering the field of computing. What comes into focus then is the interesting possibility that there is something about the field of computing that makes it less attractive to women. Suppose for the sake of argument that women on average find computing less enjoyable and more unpleasant than other activities. I f this were true and i f these features were essential to computing, then we might simply accept the current situation. We could claim that there is a mismatch between the interests and desires of women and the nature of computing, and leave it at that. No need to change anything. To use an analogy, suppose there is a particular job that required a worker to have a shaved head, and further suppose that women were less inclined than men to shave their heads to get the job. Here we would say that an inherent feature of the work influences choices, and the low number of women is not indicative of discrimination or unfairness. However, the ethical character o f the situation shifts dramatically i f we discover that the shaved head requirement is not truly essential to the job, or could be removed with a trivial design change in the workplace. Then we would wonder why this requirement was there at all. Furthermore, i f we found that the requirement was in fact detrimental to the effectiveness of the employee and didn' t improve the profits o f the employer, we would surely wonder about the fairness and the wisdom of the requirement. This analogy suggests that the situation with regard to women in computing may be more subtle than overt discrimination. Are the disincentives women experience inherent to computing? Are the aspects o f computing that women find distasteful essential? Or