DIVERSIFYING THE SCIENCE & ENGINEERING WORKFORCE

On January 14 and 15, 2005, the Science and Engineering Workforce Project (SEWP) convened a conference at the Cambridge facility of the National Bureau of Economic Research (NBER) in order to discuss issues of diversity in the S&E workforce. Approximately fifty people attended, comprised of economists, physical scientists, mathematicians, psychologists, policy makers, program directors, sociologists, students, a lawyer, and a recruiting director from a Fortune 500 company. The presentations reflected the diverse backgrounds of the attendees, with talks ranging from highly abstract models of achievement, to tales of successful program efforts to address representation, to personal accounts of gender discrimination in the sciences. The lunchtime address by Harvard President Larry Summers captured virtually all of the mountainous press attention concerning the gathering, but in fact it represented only a tiny portion of the overall content presented at the conference.

PANEL I: INTRODUCTION – THE CURRENT SITUATION

Professor Richard B. Freeman of Harvard University began with welcome and greetings and then launched into his presentation with three main messages regarding doctoral level scientists and engineers: (a) something has “worked” in that there have been substantial increases in women and URM (under-represented minority) PhD proportions, (b) there is some evidence that policy efforts contributed to improved diversity, (c) but most of the gains can be explained by a normal supply response from the increase of female S&E undergraduate degrees. He showed data for the NSF’s Graduate Fellowship program documenting that the gender and ethnic gap in GRE scores of applicants has been shrinking since 1980. Freeman also presented a preliminary analysis showing that the distribution of minority graduate students in university S&E departments is very unlikely to have occurred by random processes alone. That is, URMs are somehow more favored to appear in some departments than others. Finally, he recommended that policy efforts be directed towards structuring S&E careers to be more compatible with the changing demographics of the workforce.

Next, Mary Frase, from the National Science Foundation -- Division of Science Resource Statistics, presented highlights of the NSF's Women & Minorities Report, confirming the gains in statistics showing the influx of women and underrepresented minorities (URMs) into the sciences in the past two decades. From 1990 to 2001, the total number of degrees awarded of S&E bachelors, masters, and PhDs all experienced gains (21 percent, 16.5 percent, and 2.0 percent respectively). The gains for URMs and women were much greater. The gain for each demographic group (American Indian, Black, Hispanic, or Women) was at least 20 percentage points higher than the overall percentage gain for the same degree level awarded. Despite substantial increases these groups still remain underrepresented in S&E when compared to the general labor market. Further, the picture varies depending on the scientific field. For instance, women are actually overrepresented in psychology occupations (65.6 percent), but only 10.9 percent of workers in engineering occupations were female in 2000. In 2003, women received 20.2 percent of doctoral degrees in engineering and 59.3 percent in the social/behavioral sciences. Catherine Didion of the International Network for Women in Engineering & Science (INWES) responded that when women do not enter the workforce in a scientific field, they are essentially lost to science, even if they had pursued degrees in an S&E discipline. Economist Charles Brown of the University of Michigan noted that URMs seem better represented in fields that have received more ample funding from the Defense Department.

Denice Denton, Chancellor Designate of UC Santa Cruz, delivered accounts from her career as an engineer when she faced discrimination as a woman scientist. One of her examples focused on the "lab bully" who is often characterized as an equal opportunity bully. But this type of behavior differentially affects those colleagues who are most marginalized to begin with. Male colleagues have a tendency to excuse abhorrent behavior by saying that a powerful professor acts like a jerk toward everyone, and therefore is not guilty of discrimination. Denton suggests that the climate for women and for all in academe could improve substantially if there were more support for “zero tolerance” towards the bullies and “jerks.” It is not enough to rely on ombuds personnel to root out creepy activity. Her conclusion, “We needed, and still need, cultural change in the sciences.” Finally, speaking as an educational leader, she discussed the importance of succession planning. Having served as dean at the University of Washington, she helped women achieve 16 percent of the faculty in Engineering, compared to 4-8 percent in peer institutions. Without planning for succession, a dean can watch gains come undone fairly quickly if the successor is particularly ignorant of representation issues.

PANEL II: RESEARCH UNIVERSITY EXPERIENCES

The second panel of the day began with Barbara Grosz of Harvard University who discussed the state of women in S&E at Harvard. She focused on women professors in the Faculty of Arts and Sciences where the proportion of female S&E faculty is about 10 percent, one-half of the female proportion in the entire FAS. When Grosz began this work on female under-representation in the early 1990s, there were only four tenured female professors in science at Harvard; it is now up to thirteen. To help make academic careers more attractive to women, Harvard, according to Dr. Grosz, has adopted some innovations designed to support child rearing including (a) delay of the tenure clock, and (b) special funds for daycare when traveling to research meetings. The federal government has tended to deny funding childcare costs to attend conferences, so the Harvard policy is helping to redress this implicit discrimination. For Grosz, junior faculty appointments will remain a significant area to get more women appointed to tenured chairs: currently 45 percent of male faculty and 69 percent of female faculty in science come from internal promotion. (Such a giant disparity does not seem to exist in the humanities and social sciences.) Then, she discussed the Radcliffe fellowship as a specific program which women found very helpful to their ability to thrive in academe. Twelve out of 45 recipients of this fellowship are in science. Professor Grosz advocated the development of role models, an inclusive team-oriented science education, and more research into understanding “pipeline” issues. She stated that issues of representation require constant vigilance against complacency.

Nancy Hopkins reviewed the gains in women faculty at MIT over the past ten years and the reforms that grew out of the Committee on Women Faculty in the School of Science at MIT, which she chaired in 1995. In summer 1994, there were 15 tenured women and 194 males with that status in the School of Science. By 2001, 22 women had tenured posts. The report of the committee found that tenured women faculty had been the object of subtle gender bias manifested as “the gradual marginalization of women faculty as they progressed through their careers . . . often accompanied by the women receiving less space and fewer resources for research, and lower salaries, and in their having little or no role in the important decision-making processes within their departments or the Institute.” Her analysis of the situation at MIT showed that even if MIT continued its affirmative hiring efforts, it would take 60 to 100 years for that institution to attain rough parity between males and females. She lauded former MIT president Charles Vest for recognizing the gender problem and supporting efforts to address quality of life issues in support of a diverse faculty.

Linda Hamilton Krieger spoke on implicit bias in faculty search committees, proferred Sexist Bullshit Bingo as a humorous way to highlight implicit discrimination in group situations, and spoke about the legal hurdles and legal protections in cases of alleged racial or gender discrimination. She discussed how Title IX, Title VI, and Title VII are pursued in the legal system. In her commentary on how the law defines discrimination, she explained the categories of individual disparate treatment, systematic disparate treatment, and disparate impact. Statutes have tended to prohibit discrimination but failed to define it, and thus the society has relied on case law to arrive at more precise definitional rigor. Professor Krieger wrapped up her talk by passing out the paper & pencil version of the Implicit Association Test that relates gender with humanities or the sciences. For instance, in general, people more quickly connect the words “chemist” with “son” than connect “chemist” with “daughter.” Krieger has worked closely with Harvard psychologist Mahzarin Banaji on IAT issues at the Radcliffe Institute during 2004-2005. The IAT has become an important tool in overcoming the reality that many people in ordinary polling situations try to give the pollster appropriate or so-called politically correct responses on the issues of racism or sexism, rather than their true feelings. IAT testing is conducted at a major website, https://implicit.harvard.edu/implicit/, which seeks to assess conscious and unconscious preferences.

LUNCHTIME SESSION

After participants were invited to help themselves to catered sandwiches, the session began with Larry Summers's lunchtime speech (Transcript).  Detailed summaries and analysis are widely available in press archives.

Following this, Shirley Malcom, who heads the Directorate for Education and Human Resources at the AAAS (American Association for the Advancement of Science), expounded on the prospects for affirmative action policies in the university world following recent rulings at the University of Michigan. She pointed out that many pro-diversity forces hailed it as a victory, but they failed to see how the opposition quickly mobilized to overcome its potential impact. Malcom discussed challenges ahead for science, as she entertained questions of shortages and oversupply of scientists. There are those eager to “push” women and underrepresented minorities into STEM fields that years down the road could lead to unemployment. National security issues are driving home the need for more scientists from the United States and less reliance on foreign/international students. But there remain problems. Postdocs are becoming a parking place for many young people whose scientific careers are essentially on hold. So much of the research money is increasingly going to older researchers. Moreover, many minority students shoulder levels of debt that are unacceptable. Malcom praised HBCUs for their traditions of internationalism and mentioned a science PhD from Egypt who had done her undergraduate work at Howard University. But she also indicated that there are problems with the domestic pipeline when some fields have such a high proportion of international students.

PANEL III: PSYCHOLOGICAL AND SOCIOLOGICAL ASPECTS OF OCCUPATIONAL CHOICE

Catherine Good of Columbia University presented data supporting the existence of "stereotype threat", under which performance differs depending on whether subjects are aware of existing negative stereotypes. For instance, Good reported that white males scored worse on standardized math tests when they were told that their scores would be compared to Asian males. The same type of effect was shown to exist with blacks taking diagnostic tests, women taking math tests, and Latinos taking verbal tests. Good suggested that the effects of stereotype threat may be ameliorated by fostering an incremental theory of intelligence, rather than a fixed view of intelligence. In terms of mathematical ability, “entity theorists” adopt the view that “intelligence is fixed.” The “incremental” view posits mathematical ability as “malleable,” a learned skilled that is honed through practice and interest. Professor Yu Xie responded that in many Asian societies parents constantly tell children that they can do better and that hard work will lead to mastery of mathematics. In the United States, he finds among parents and professors attitudes pervaded by the belief, “either you have it or you don’t.” The dominance of the innate view in the United States could be a significant obstacle to raising mathematical achievement.

Daniel B. Berch of the Child Development and Behavior Branch of the National Institute of Child Health and Human Development reported on the work of a variety of researchers funded under the Program in Mathematics & Science Cognition. He surveyed experiments in neuroscience and cognitive science concerning mathematics learning in general, and gender differences in the brain in particular. There is still much to be done in this preliminary research. Berch paid particular attention to the work of Diane F. Halpern of Claremont McKenna College, including her essay on “A Cognitive-Process Taxonomy for Sex Differences in Cognitive Abilities” from Current Directions in Psychological Science (August 2004). Reflecting on her own framework and a research project by Ann Gallagher, Jutta Levin, and Cara Cahalan, Halpern had observed that males did better when there was “a spatially based solution strategy,” but females were more successful when the “solution strategies were more verbal or similar to the ones presented in popular math textbooks.” For Gallagher and her team, wrote Halpern, “average performance of different groups on standardized tests could be minimized or maximized by altering the ways problems are presented and the type of cognitive processes that are optimal for generating solutions.” Halpern concluded that “everyone can improve in any cognitive area – that is the reason for education – and rapid changes in the proportion of men and women in some fields show that huge changes can occur across populations by changing educational opportunities and social expectations. People do not have to be the same to be equal.”

Yu Xie of the University of Michigan and co-author Kimberlee Shauman of University of California, Davis presented results from their book Women in Science, detailing their “synthetic cohort” analysis technique and “life course” approach. Shauman emphasized that the traditional “leaky pipeline” picture of gender attrition in S&E may be flawed since women also tend to enter or re-enter S&E at stages of life later than usually expected. Xie and Shauman presented a couple of alternatives explanations for the underrepresentation of women in S&E, and then presented their analysis of these alternatives. Their summary of the causes of the persistent gender inequities in S&E shows no support for hypotheses that involve a lack of mathematical preparation or training at the high school level. Instead, they found support for three issues, (a) gender gap in interest in S&E or in expectation of attaining an S&E undergraduate degree, (b) occupational segregation within S&E fields; for instance women are more likely to be employed in teaching universities, while men more likely to be employed in research universities, and (c) differences in gender roles in regard to raising children. (Note that marriage per se does not seem to matter much: “Married are women are disadvantaged only when they have children.”) While math and science participation is equal between males and females in high school, except for physics, there is a substantial drop-off in the transition to college: women are at that crucial moment half as likely as males to express an expectation of going into science and engineering.

PANEL IV: LEARNING FROM SUCCESSFUL PROGRAMS

Sue Rosser of Georgia Institute of Technology spoke on programs on her campus that promote the participation of women, POWRE and ADVANCE. Rosser stated that in her years of experience it has become clear that many women went to teaching rather than research institutions in part because the former are seen as family friendly. She noted that it is often very hard to make the transition from a teaching college to a research institution. Catherine Good wondered if some research could be conducted on postdocs, many of whom are bailing out of science when they recognize that its demands make it nearly impossible to have a family.

David Manderscheid, current chair of the University of Iowa Math department, reported on the success of its departmental initiatives to recruit underrepresented minority students into Iowa’s math program. From 1993 to 2004, the URM enrollment went from essentially 0 percent to approximately 22 percent out of about 120 graduate students. The initiatives were a faculty-led effort and included a three-week summer institute for incoming students, creation of a “mentoring community”, and regular help sessions throughout the year. Dr. Manderscheid noted that the department’s culture has shifted as a response to the graduate student compositional change, and that ethnic and gender inclusion “have become the norm”. Further, many of the non-minority students have “availed themselves of many of the initiatives initially created to help minority students,” and “the new environment has become a successful recruiting tool for the department.”

Sheila Tobias presented background on the expansion of PSM programs (Professional Science Master’s degrees) focusing on a range of scientific disciplines: biotechnology, bioinformatics, industrial mathematics, computational science, and forensic chemistry. She argued that the typical MBA lacks meaningful knowledge of science and engineering, and the PSM degree may fill the gap. Tobias also took up the early efforts to discredit affirmative action in the academy. She referred to Princeton University’s Richard A. Lester (1908-1997), whose study for the Carnegie Commission in 1974 claimed that affirmative action for women and African Americans threatened to lower academic standards. According to her interpretation of Lester, he believed that affirmative action could work for blue-collar jobs and industrial plants; but it would fail in the world of universities where the most exacting standards of excellence had to be maintained. Lester’s critique of affirmative action had many supporters at the time, but Tobias believes his thesis has failed the test of time.

DINNER SESSION

During the sit-down dinner at the Harvard Faculty Club, Nila Bhakuni, the new director of the Office of Technology Transfer at Rice University and a specialist on nanotechnology issues, gave some opening remarks and introduced the first of the dinnertime speakers, Jaline Gerardin (Harvard) and Saundra Quinlan (MIT & National Society of Black Engineers). Ms. Gerardin spoke about the difficulties of being a science major and also her perspectives on gender & science. She emphasized how much summer math and science camps had helped her progress in the field; but it was her sense that many of these programs are overwhelmingly male. High schools and many programs are helping to entrench gender under-representation at a very early stage. These intensive summer workshops help develop the advanced tools and networks for later success in science. Ms. Quinlan spoke about the decision making process that led her to choose MIT as a college and engineering as a major, and further on the benefits of participating in an organization such as the Society of Black Engineers. Bhakuni elaborated on the National GEM Consortium, an organization based at Notre Dame that has 50 corporations and 90 colleges among its participants. GEM, which stands for “Graduate Education for Minorities,” calls itself “the only privately funded nonprofit graduate education organization providing underrepresented groups the opportunity to complete graduate study at the nation’s best science and engineering programs while gaining practical work experience with leading employers.”

Following the student addresses, Anne C. Petersen, Senior Vice President for Programs at the W.K. Kellogg Foundation, spoke about the efforts of the Kellogg Foundation, (and other charitable foundations) in affecting the issue of representation in the sciences. The Kellogg Foundation has taken many initiatives, including the funding of a five-year, $2.75 million partnership with the University of South Carolina and six HBCUs to improve research on disparities in public health outcomes and ultimately increasing the numbers of minorities entering the public health field.

PANEL V: ACADEMIC EMPLOYMENT ISSUES

Catherine Weinberger from UC Santa Barbara presented an analysis of test scores to see if the S&E Workforce is comprised of only those people who are at the very top of the distribution in terms of “ability”. Her data was drawn from high school students’ math tests and the later outcomes of these students including college major, educational attainment and labor market outcomes at age 32. Prof Weinberger proposed three simple models for the distribution of “ability” in the workforce to test against the data. The conclusion: the distribution model where these workers are drawn from the upper 25 percent of test scorers is consistent with real data, while the model where they are drawn from only the top 10 percent of the distribution is not consistent.

Donna Nelson presented results from her survey of science and engineering faculty across the nation, demonstrating very poor representation of women and URMs in tenured faculty positions, much worse than the respective PhD proportions in the same fields. Looking at the top 50 science and mathematics programs in the nation, Nelson and her colleague found that 48.2 percent of B.S. recipients in mathematics were female in 2000; but only 8.3 percent of the faculty in this field were women as of 2002. For the same time period, 21.4 percent of physics undergraduate degree recipients were female, with women making up only 6.6 percent of the faculty in this discipline. In Computer Science, 20.5 percent of the PhDs were women, but only 10.8 percent of assistant professors were female between 1993-2002. In most fields, the disparities between PhDs and assistant professor representation were less dramatic, but were still significant. Astronomy, mechanical engineering, and civil engineering were the only three out of 14 fields that women could be said to be over-represented in the ratio between PhDs and assistant professorships. URM males earned 4.2 percent of the PhDs in chemistry, but only had 1.6 percent of the professorships in this time span. Mathematics was a different story, as URMs had 3.3 percent of the PhDs and secured 5 percent of the assistant professorships.

Cathy Trower, of the Harvard University Graduate School of Education, reviewed the salary and promotion gaps that still exist in S&E, despite concentrated efforts by many organizations and professional groups starting in the 1970s and through the following three decades. For instance, gender and URM disparities in faculty salaries compared to white counterparts have remained largely unchanged over the past 20 years. Looking at four-year colleges and universities as of 2001, she observes: “Studies have shown that gender disparities in faculty salaries have remained unchanged at approximately 20 percent over the past 30 years, with unexplained disparities of about 12 percent remaining after accounting for time in service and discipline for women over 40.”

Ronald Oaxaca of Arizona State University presented an analysis of faculty salary structure at a research university. As part of this work, he focused on gender pay gaps and other gender differentials in the characteristics of female vs male faculty. The “unadjusted salary for men exceeded that of women by nearly 16 percent.” Generally, male faculty tend to publish more articles and had more coauthors, while women won teaching awards at a much higher rate. He argues that teaching should be given improved compensation, and that this gender-neutral reform would benefit female faculty who are doing more of this educational heavy lifting. Oaxaca noted that their findings also “illustrate the erroneous conventional wisdom which holds that unexplained gender salary gaps will shrink if productivity/performance measures can be incorporated into salary regressions.” Further, he speculated that gender salary gaps at the full professor rank may be partly a result of a larger rate of men seeking outside offers as a strategy for salary augmentation, which women may not practice to the same extent.

PANEL VI: NON-ACADEMIC EMPLOYMENT ISSUES

Kirsten Roby from Microsoft Corporation described her company’s recruiting programs, and how she sees the benefits of having a mixed talent pool from which to select employees. Leadership development is a key challenge in cultivating diversity in industry, as it is in academe. Microsoft is facing the huge reduction in H-1B visas, and this is a problem because it is not so simple to do software development in India. Proximity is important for Microsoft, despite the outcry about global outsourcing conquering the planet. Microsoft expects a much more scientifically trained workforce, and it will no longer be hiring significant amounts of “liberal arts-types.” She also mentioned the Explorers Program, an intensive coding class and experimental program that brings 24 students around sophomore year to the company. She spoke of plans to double the size of the program and how it has brought many women and minorities to the company. Roby finished by admitting that Microsoft has had to change its culture. For too long the company behaved as though it were still a small start-up when it is now a corporate behemoth that hovers around Wall Street as the second largest company by market capitalization. The frenzied atmosphere and the stress on Social Darwinian categories such as who has the highest IQ must give way to new values. When so many people from the company’s earlier history had become wealthy, Microsoft recognized that it now had to do some new things for retention of good people. Microsoft wants to keep the informal atmosphere, creativity, and passion for getting things done from the earlier period; but there is much that can be done to bring a positive climate within a transformed corporate culture.

William Berry of the US Department of Defense highlighted the need for US citizen scientists and engineers to work in national security related activities, noting that the number of students eligible for security clearances pursuing such activities is small and declining. Turning to issues or representation, Dr. Berry’s data showed that female percentage of DoD S&E workers was smaller than the overall percentage of women DoD workers. The gap was only a few percentage points at most, much smaller than the comparable gap in the general (non-DoD) workforce. However, this is probably because the percentage of women working in the civilian DoD is smaller than the general population. As for non-white minorities, there was also a slightly larger gap between S&E workers and the overall DoD civilian workforce. The percentage of DoD employees who are minorities appears to be slightly larger than this same percentage in the general workforce. DoD lacks centralized recruiting, as is the case with Microsoft; labs and others are recruiting at the same time. David Manderscheid added that many science and math jobs at NSA and other agencies are 40-hour positions, unlike the 80-hour demands of contemporary academe. There is a growing cohort of PhDs who find that an appealing feature of work for government agencies.

Kjersten Bunker Whittington reported on her work analyzing the collaboration networks of scientist patent writers in academe and in industry. An interesting result: men tend to patent more frequently, while women have more cites per patent. Academic patenting (at least in biotechnology) is dominated by a small number of well-connected nodes. Since 1980, academic patenting has soared 700 percent. Whittington has collaborated on research with Boston University sociologist Laurel Smith-Doerr, who has found greater gender equality in smaller, dedicated biotechnology firms than in the large pharmaceutical companies. They are committed to investigating whether different types of work settings make a difference in achieving the goals of diversity.

Sharon Levin of University of Missouri and Paula Stephan from Georgia State University reported on their work examining the data for attrition rates from Information Technology (IT) careers depending on gender or ethnic status. They found that about 70% of those working in IT in 1993 were retained in 1999, the retention rate was higher for men than women (73% vs. 66%), higher for whites than African Americans (70% vs. 66%), and higher for Asians (70%) than whites (70%). Further, they found that African Americans leave IT occupations for other occupations; do not leave the labor force or become unemployed. Women leave IT occupations to leave the labor force or become unemployed, not to move into another occupation. The policy implication: programs directed towards retention will have differential outcomes depending upon group in question.

CONCLUSIONS AND OBSERVATIONS?

While women have been studying and entering into S&E fields more than ever before, there appears to be an impediment at the upper echelons of prestigious institutions. Specifically the proportion of women holding senior faculty positions at research universities is significantly below what would be expected given the growing pool of women with PhDs and records of academic distinction. For underrepresented minorities, there are gaps that can be even more acute than for females in general. For instance, out of the top fifty physics departments in the U.S., not a single one had appointed an African American female to a full professorship as of 2002.

David Goldston, Chief of Staff for the U.S. House Committee on Science, offered a few conclusions that he sought to be pursued in the future. First, he argued that fine distinctions matter, and there needs to be a recognition of differences between the challenges facing women and those confronting many underrepresented minorities. Second, he thought some of the perspective could be broadened to include a wider historical range of vision on underrepresentation. Earlier in the twentieth century, Jews and Catholics found themselves locked out of many positions at prestigious and mainstream universities, with both groups sometimes described as “races.” He noted that social class almost never came up in the conference. For Goldston, some types of attacks on meritocracy at the top can have harmful effects for those at the bottom. He cited the views of James Fallows that Japan has achieved its postwar economic miracle by having what might be called the best bottom 50 percent in the world. Japan did this in part by pushing its top to strive for the highest standards. Third, he called on academics to be careful about arid theorizing without paying attention to what goes on in daily practice. Finally, he thought some of the people in academe and at this conference had turned very ideological, and he argued that this should be an arena that strives for more non-ideological approaches to knowledge. He finds that Washington is a place permeated by ideological conflict, and he had hoped the university could be a place freer from this framing of information. For Goldston, he was troubled that the fury over President Summers’s views had led some to treat his observations as though they were a return to eugenics.

While willing to concede that Summers had not gone down the road to eugenics, John Trumpbour of the Labor & Worklife Program at Harvard Law School criticized the increasingly aggressive tendency in the human sciences to find biological explanations for inequalities, and he thought that President Summers had clearly been too dismissive of social explanations for underrepresentation. Sheila Tobias asserted that Summers had judged the work of so many of the scholars at this conference to be of no value in his ultimate conclusions, though Richard Freeman countered that the Harvard president did read some of the papers and would stay engaged with our debates. John Yochelson, the president of BEST (Building Engineering and Science Talent), thought there was some excellent social science analysis of the problem at the conference, but much more would need to be done in the areas of policy prescriptions and reform.

The ideological storms generated by this conference have led to much institutional activity and an array of task forces on improving diversity at Harvard and in the rest of academe. Despite the slow pace of change in achieving diversity, this conference showed that there are grounds and resources for hope. To cite just a few examples: three decades ago in the U.S., there were approximately 13 times as many boys in the seventh-grade than girls achieving a 700 on the Math SAT. Recently that gap has closed to 2.8 times. Though there may still be a long way to go, this progress indicates that there is an enormous amount of malleability in academic achievement. While some have tried to use the 2.8 times gap as a sign of a fixed biological gulf in math achievement, those who look at it historically see it as evidence that the best in academic achievement has yet to come. Again, when it comes to the production of PhDs among URMs, it might be observed that the highest five departments in Chemistry have 19 percent of the degrees going to URMs, while the lowest five Chemistry departments are awarding URMs the degree to a microscopic 2 percent. This variation indicates that some places are doing a considerably better job, and it might be helpful to find out what has been working in the few places with a history of higher numbers. This conference should be regarded as part of a continuing engagement with these issues. By promoting structures and practices that open doors to scientific opportunity, conference participants and members of the network will undoubtedly maintain a role in meeting this formidable societal challenge.