Posner makes excellent points, so I will fill in at a few places. First, it is common for National Academy of Science sponsored Reports on economic and social issues, such as this one, to have few members of the NAS on the report committees. Unfortunately, it is also common for Reports on these issues to be poorly executed, and to be driven more by wishful thinking than by scientific findings. The low quality of many NAS Reports on economic and social issues goes back at least to the first such report I evaluated in the 1970's at the request of a Vice-President of the Academy, just after I was elected to the Academy. The Report was on the future of energy resources, contained virtually no economics, and was filled with common prejudices about how fast the world was running out of oil and other energy sources. I have felt since then that the NAS should not lend its prestigious name to reports on such issues.
Unfortunately, this NAS Report on women in science is no exception to the tendency of its Reports to be heavy on beliefs and weak on carefully documented analysis. To be sure, everyone who has seriously studied this question agrees that women in the past suffered greatly from discrimination in gaining entry into many professions, including but not limited to the sciences. In addition, however, there is also agreement that discrimination declined greatly over time, which is partly reflected in the data presented in this Report on the now substantial enrollments of women at technical schools such as MIT.
How much discrimination remains? The evidence is still unclear, so considerable disagreement remains over the respective roles of discrimination in access to education and jobs, women's responsibilities for childcare and other household activities, social conditioning, genetic differences, and possibly other factors. The report does not advance our ability to discriminate among these explanations. An Appendix to the Report discusses in an uninspired way various theories of discrimination, including mine, but no clear-cut conclusion is reached about which theory, if any, is highly applicable to the realities of women's position in science.
The summary of the Report says that it cannot be that women in academia, and sciences in particular, are now recipients of favoritism because affirmative action that selects candidates on the basis of race or sex is illegal. Well, legal or not, anyone who has sat in on academic departmental or divisional meetings, and my wife is also a professor, knows how often preference is given to candidates because they are women, even when male candidates have better records. Of course, not every professor or every department acts this way, but a strong and aggressive number of professors do, and deans and other university administrators frequently back their position.
The Report dismisses the importance of women‚Äôs interest in child rearing and other family activities in limiting their scientific accomplishments by stating that "many women scientists and engineers persist in their pursuit of academic careers despite conflicts between their roles as parents and as scientists and engineers". No one would deny that statement, but the relevant question is whether the considerable time spent by most women in child rearing is an important factor in their generally less outstanding achievements as scientists and engineers. Common sense and many studies suggest that the many hours spent on child rearing at least makes it much harder for women to produce distinguished research.
The Report recognizes that women take much more time off than men not only to take care of children after they are born, but also when children are sick, when a parent is needed to visit their children's school, and in other situations. The Report counters that over a lifetime men make up for this by taking more time off as sick leave. True, but women reduce their working time they could be spending on research at younger ages when scientific productivity peaks, while men generally become sick at older ages after productivity is on the decline.
Larry Summers was forced to resign as President of Harvard in major part because of his well publicized comments on why relatively few women are in scientific positions at the best universities. He attributed this partly to discrimination and the difficulty of combining family responsibilities with research. No trouble there. He got into trouble with many women's groups when he raised the issue of whether women on average had as much capacity as men to make outstanding scientific contributions. The Report denies that there are "any significant biological differences between men and women in performing science and mathematics that can account for the lower representation of women in academic faculty and scientific leadership positions in these fields". Account fully or only partially?
I am no expert on this evidence, and I do try in my own research to see how far I can go in understanding the different achievements of working men and women without assuming innate gender differences in capacities. Still, that is very different from claiming the evidence is fully persuasive on this point, or in more technical language in claiming that the variability in women's capacities is not less than the variability in men's, regardless of how their mean capacities compare.
Sweden probably has the strongest commitment to gender equality of any country. It implements this commitment with a liberal system of childcare allowances and facilities, a generous system of government paid leaves open to both sexes-with men required to take some of the leave- and a strong anti-gender discrimination attitude. I cannot speak with authority about Swedish scientists, but I can say with confidence that while there are excellent female Swedish economists, yet at younger, middle, and older ages, the best of Swedish economists are very predominantly men, perhaps even more so than in the United States.
To conclude, I have very strongly opposed discrimination against women in general and for academic positions in particular. While I am not sympathetic to strong government involvement in paid leaves for childbearing or for childcare facilities, I can see a possible case for some government actions along these lines. However, such attitudes on these issues do not justify a Report on women in science that does not really meet the fundamental criteria for a scientific Report. Instead, it provides further evidence on why the NAS should not be sponsoring Reports on economic and social issues.