T his contribution is based on a presentation made at the 2002 AAAS Annual Meeting, held in Boston, Massachusetts, entitled "University as Workplace: Complexity and Diversity."
Statistics on enrollments and degrees document the underparticipation of minorities in engineering, which is a representative segment of so-called SMET (or, if you prefer, STEM) disciplines. I would like to comment on why these conditions are amenable to change and indicate briefly how--through policy and local practice--change can occur.
The State of Student Human Resources on Campus
Science and engineering (S&E) crave the "best and brightest." The issue for these disciplines is not "market share" but "talent share," as The American Freshman data established long ago. As the college-age population approaches 80% women and minorities in the next 15 years, the very groups historically neither recruited nor supported in S&E will represent the pool of talent from which all professions (including medicine and law) must draw. As Teitelbaum (2001) has argued and Seymour and Hewitt (1997) have demonstrated, science fares poorly--even worse than engineering--in career prospects, lifetime earnings, and quality of student and professional life relative to its competitors.
Consider one other set of data: institutions that currently produce the greatest number of B.S. and Ph.D. graduates within each minority category of interest. We always hear about how concentrated federal R&D funding is, with 100 institutions accounting for 90% of federal R&D obligations and 30 institutions receiving the lion's share of that. The same applies to which universities produce the most Ph.D.s. But how do these lists change when the output is women and minority graduates--B.S. and Ph.D.--in engineering? Size matters here. We should look for "critical mass" or some indication that within an institution a cadre across cohorts has been created.
Of about 300 institutions that award engineering degrees (according to the Engineering Workforce Commission), 71 graduated fully half of the B.S. engineers in 2000. These same institutions in the aggregate produced engineers in the following categories: 35% of women, 47% of African Americans, 57% of Latinos, and 53% of American Indians. Yet only two universities produced a baccalaureate engineer from all four of those categories--MIT and the University of Florida. Fifty-two of the remaining 69 universities specialized in one group only. Except for B.S. degrees awarded to women in engineering, in institutions that awarded at least 500 such degrees in 2000, the list of top producers of minority B.S. engineers hardly overlaps with the list of major research universities.
Those awarding at least 25% of their B.S. engineering degrees to women were MIT, the University of Puerto Rico, the University of Michigan, Georgia Tech, and Cornell. For the other categories, by minority group, with a threshold of 200 graduates, different institutions appear, notably NC A&T, the University of Puerto Rico, the University of Colorado, Boulder, FAMU/Florida State, Florida International, and Oklahoma State.
At the doctoral level, the results are even more skewed, although the universities are more familiar as R&D performers. Of course, the threshold is far lower (at least 50 produced, with at least 20% female) because the numbers produced are far more modest. For women, Johns Hopkins leads the way, followed by Northwestern, Caltech, Princeton, and the University of Colorado, Boulder. Where African Americans represent at least 3% of the Ph.D.s awarded in engineering at an institution, the rank-order is Georgia Tech, NC State, Purdue, the University of Washington, and Stanford. For Latinos, the top five are Cornell, UT Austin, Caltech, the University of South Florida, and the University of Washington. Only one institution that awarded at least 50 Ph.D.s in engineering in 2000 had greater than 1% going to American Indians: the University of Colorado, Boulder (n = 4).
In all, only 45 institutions accounted for 60% of all the Ph.D.s in engineering awarded to women and underrepresented minorities, with no university producing at least one in each category.
Clearly, a few universities "get it." They succeed in recruiting and retaining to degree women or students of color. They are doing enough things right to be productive of those most disciplines and research universities have not historically developed.
At the undergraduate level, studies indicate that retention programs feature common elements (see, for example, Black Issues in Higher Education) that can be introduced and institutionalized to change the learning environment. Hersh and Benjamin (2001) go even further to suggest that a quality assessment of that environment should become a criterion of future investment in campuses (discussed further below).
Arguably, doctoral education is market insensitive, decentralized by department or program and driven by faculty interests and funding, not by demand for new professionals in particular disciplines, with certain skills, industry, or sector orientations (Chubin, 2002). If the human output reflects a broader range of intellectual interest and work orientation--i.e., a career path that does not necessarily emulate that of one's major professor--then heterogeneity by race, ethnicity, and gender adds value in several ways. The most immediate is that a diverse faculty is a sure sign of an increasingly diverse student body that a woman or person of color can indeed excel and achieve. Such role models are often pooh-poohed, but they can make a difference between persistence in and departure from a science or engineering course of study.
Community colleges have made strides in appointing women as CEOs (as reported in The Chronicle of Higher Education, February 2002). This sector, which employs the largest fraction of women and minority faculty members, has increased the percentage of women presidents from 18% in 1996 to 28% in 2001. But minorities remained at 14% of the 1200 presidencies over the same period. If progress is stalled in this 2-year sector of academe for minority administrators, then conditions at other institutions of higher education, we must conclude, are even worse. The dearth of Latinos in academic administration is especially noticeable. The overarching concern is leadership (Tapia et al., 2000): Only those accomplished persons in top positions control resources and can hasten change and both symbolically and materially increase opportunities for participation and participation of all in S&E.
How to Change and Other Questions
How the university diversifies itself as a campus community--from administrators and faculty members to new matriculants--will determine its long-term vitality as a research community. Particularly, how the university transforms a woman- and minority-dominant student population into full participants contributing professional achievements will determine its competitive advantage.
This must happen in a "post-affirmative action" era, where programs targeted to particular groups are legally challenged. This contributes to complexity, as most federal and state interventions must be institution-wide, serving all students.
As Hersh and Benjamin (2001) point out, universities could be required to show how they add value in converting raw SAT-certified talent, as well as those with 2-year college experience, into skilled science-based professionals. U.S. News and World Report rankings are reputational and rely on input variables such as endowment dollars and fraction of faculty members with Ph.D.s.
Assessing how the institution improves student access, knowledge, and skills would measure the difference between what students bring to the university and what they have attained as they exit degree in hand. Student performance would become a metric for faculty and department performance to be judged alongside research productivity and grantsmanship.
That the proportion of undergraduates opting for S&E majors has remained remarkably constant for 3 decades is hardly natural law. So, too, the retention rate: How can that be improved? After admitting these students, who cultivates them? Do universities receiving significant federal R&D funding have higher retention rates? Which departments excel and which lag? These statistics are knowable.
Universities must take a page from the corporate handbook, as it were. Industry diversity practices have been codified and replicated. Again, the crucial element is leadership: commitment from the top to institutionalizing and rewarding practices that change the environment and sustain dynamic, pluralistic, collaborative workplaces. Last October, the NAE Committee on Diversity hosted a "best practices" workshop, which recommends the following in a forthcoming report: link diversity to organizational goals and strategies, provide training to employees (but especially managers), recognize affinity groups, encourage outreach to all stakeholders, expand recruiting pools, benchmark and create metrics to monitor progress, and hold all accountable for results (i.e., through performance reviews and compensation).
Successful retention strategies have been identified. What will it take to try them at new sites? How will the campus signal that it values, along with research accomplishments, retention to graduation and close the gap nationally between the two-thirds of nonminority engineering students who today earn the baccalaureate compared with 40% of minority students?
People cannot be viewed as separate from the campus environment, its culture, practices, and future directions. In the absence of a commitment to "grow our own," we continue to import talent--a sacrosanct value in this democracy and especially in the history of U.S. science--and rely on ad hoc measures such as H1-B visas. This may be necessary, even as the threat of terrorism narrows our thinking, but it is no substitute for a national human resource development policy. Only a strategy of investment in native talent will prepare the population to ascend to positions of leadership in and outside the academic sector. Make departments accountable for succeeding in growing the S&E workforce instead of relying on the importation of talent.
Finally, if we are searching for more lofty solutions to complexity, we must also engage what I have studiously avoided mentioning because it is seen as adding complication and cost to, and overall diversion from, the university's core missions: How can we solidify the university role as facilitator of a more seamless K-16 education system by collaborating institutionally with schools, teachers, and students in their local communities?
Higher education, particularly the 2-year college sector, inherits the deficits of pre-K-12 education. If the achievement gaps between rich and poor school districts and between minority and nonminority students do not shrink, the pressure on universities--to remediate, to sustain quality, to educate all--will only grow. We need an environment where excellence and equity, like career and family, are intertwined and are not traded off. Diversifying the campus while connecting it to its various sponsors, suppliers, and consumers is a strategic necessity.
What the university must absolutely resist is declaring this mission too grandiose or unattainable. Such a concession will be the undoing of a sector that must lead, innovate, and take on new missions with the arrogance and the will to succeed.
D. E. Chubin, "The competition for talent" [letter], Science 295, 972-973 (2002). http://www.sciencemag.org/cgi/content/full/295/5557/972 
R. H. Hersh and R. Benjamin, "Assessing the quality of student learning in undergraduate education: An imperative for state policy and practice." CAE Policy Papers, Council for Aid to Education, New York, NY, 2001.
E. Seymour and N. M. Hewitt, Talking about Leaving: Why Undergraduates Leave the Sciences (Westview, Boulder, CO, 1997).
R. A. Tapia, D. Chubin, C. Lanius, Promoting National Minority Leadership in Science and Engineering, A Report on Proposed Actions (Rice University, Oct. 2000). http://ceee.rice.edu/Books/DV/leadership/index.html 
M. Teitelbaum, in Scientists and Engineers for the New Millennium: Renewing the Human Resource, D. E. Chubin and W. Pearson Jr., Eds. (Commission on Professionals in Science and Technology, Washington, DC, 2001), pp. 71-79.
Dr. Chubin is the Senior Vice President of Policy and Research with the National Action Council for Minorities in Engineering  (NACME). For further information, please e-mail firstname.lastname@example.org .