As US competitiveness is increasingly challenged on all sides, the forced attrition of women from the science, engineering, technology, and mathematics (STEM) workforce represents an annual cost of billions of dollars. This loss comes at a time when the United States is facing an absolute decline in entry-level engineers and growing rivalry from foreign innovators. Most discussions hold that gender equality is the primary benefit of, and reason for, getting more women into science. But this is not the primary benefit. Instead, the failure to expand women’s participation in science is not simply an issue of “feminism” or civil rights but increasingly a problem for US economic security.
Problem: Decline in US Technological Capabilities
For the last decade, a parade of reports has documented a slow erosion of the United States’ relative advantage in science and technology. The alarm sounded by the National Academy of Sciences analysis, “Rising Above the Gathering Storm,” is only the latest of these troubling surveys. After almost a century of near technological predominance, the United States has become a consistent net importer of high technology, shifting slowly from a US$22.4 billion high-tech trade surplus in 1990 to a US$134.6 billion trade deficit by 2005. The United States’ share of world science and engineering research publications has also fallen; while the United States now trails even in the use of high technologies, most of which were US inventions, including (per capita) the internet (9th), broadband (12th), and cellular phones (53rd).

The implications for US competitiveness should be clear to all. National strength in science and technology directly feeds US economic growth, industrial prowess, military might, and increasing living standards. Economists estimate that half of US economic growth since World War II has come from new technology, creating productivity improvements in every sector of its economy.

After peaking during the 1990s, the wellspring of US science and technology appears to have slipped into relative decline and is evident in the broader economy. Over the last decade, US patents as a percentage of world patents have fallen by one percent each year. And while per capita patenting rates are climbing within the United States (1.66 percent annually during 1996-2005), innovation rates are rising even faster outside the United States (2.31 percent annually from 1996-2005).

Perhaps more worrying is the fact that US high-technology small business formation has dropped in every sector. This is important because small business formation represents the traditional seedbed for new technologies and industries. Hewlett-Packard, Microsoft, Apple, and Google all began in garages and university dorms as small businesses, as did many of the telecom, internet, alternative energy, and even some of the biotechnology firms of the 1990s.

In contrast, foreign firms have vastly improved their scientific and technological capabilities. We have seen the rise of technological competitors in Ireland, Israel, Finland, Taiwan, South Korea, and a half-dozen other countries. Toyota and Honda now mass-produce the most advanced hybrid automobiles. Spain is home to Europe’s first commercial concentrating solar power plant and is a lead producer of wind power technologies, and Israel’s Checkpoint is the inventor and market leader of network security “firewalls.”

China, according to most analysts, now looms as the next major technological competitor to the United States. Although the data remain cloudy, China produces at least twice the number of engineers as the United States. In published research, China now ranks second in engineering and chemistry and third in physics and mathematics. High-tech production has been outsourced to take advantage of this labor supply, making China the world’s biggest exporter of telecom equipment, computers, electronic components, and now even the world’s largest producer of solar panels.
Diagnosis: Scientist-Engineer-Entrepreneur Gap
Why has this happened? Many observers blame globalization, but these criticisms are misplaced. Certainly globalization has allowed corporations to split up their research, development, and manufacturing processes, spreading them around the world. But while opening the door, globalization did not cause high-tech production, and the high-wage jobs that accompany it, to walk out that door. Indeed, globalization could have worked in the opposite direction by providing a springboard for US technology to dominate world markets. This leads to the question: Why did the flood of competitive prowess go out of the United States rather than in?

A lack of financial capital is not the answer. Controlling for inflation, total US research and development (R&D) spending is higher now than it has ever been, at or near record levels. However, critics are correct to point out that US R&D spending per GDP peaked in 2001 and has declined ever since. US R&D outlays now trail behind the per GDP expenditures of eight other high-tech competitors, including Japan, South Korea, Switzerland, Israel, and Taiwan.



The reason for the decline of the United States’ leadership role in technology appears to be that the country has drastically slowed its production of competitive STEM workers and entrepreneurs. As Intel spokesman Howard High put it in 2005, the top high-tech firms now “go where the smart people are.” Several recent studies have shown that high-tech multinational corporations heavily base their location and outsourcing decisions on the availability and quality of a country’s STEM workforce and of the research universities that produce them. Indeed, the data reveal the drop in relative US technological competitiveness to be highly correlated with a decline in the US STEM workforce. The United States had fewer college students pursuing engineering degrees in 2005 than in 1985, despite a rising undergraduate population. In 2000, more than 25 countries had higher percentages of 24-year-olds with degrees in science and engineering than did the United States. The country is losing its lead in science and technology because it is losing its edge in producing what high-tech companies call “smart people!”
Changing Trends: Men Leaving, Women Entering
The good news is that some components of the STEM workforce are improving. Reversing the traditional trend, more women than men are now entering college and have increased their pursuit of STEM careers. Women now make up just under 60 percent of both undergraduate and master’s degree students. In college, women now spend more time studying, earn higher grades, and achieve more awards and honors than men. As a result, US women have earned the majority of all bachelor’s degrees in science and engineering since 2000.

The bad news is that women are not distributed equally across the STEM workforce. Unfortunately, the greatest dearth is in those fields with the greatest workforce demand and biggest economic payoffs: engineering, computer sciences, and the physical sciences. At the highest level of training (PhD) women earned 67.3 percent of the PhDs in psychology, 55.1 percent in anthropology, and 58.7 percent in sociology, but only 46.3 percent in the biological sciences, 28 percent of the mathematics and statistics degrees, 26 percent in physical sciences, 21 percent in computer science, and 18 percent in engineering. Indeed, in engineering and computer sciences, the percentages of female students have reached a plateau or even dropped during the last decade.

This decrease in female participation rates accelerates after graduation in the science and engineering workforce. While more women are entering the STEM pipeline, they later leak out of this pipeline in record numbers. Despite grades and other academic attainments equal to or surpassing those of the men who remain in STEM, more women leave science and engineering compared to their male counterparts, resulting in very few women in senior and leadership positions in the STEM workforce.

The economic costs of this attrition are astronomical. When STEM workers leave the workforce, they take with them thousands of hours of expensive training and experience, often paid for by federal and state taxes. Conservative estimates (i.e. residency at a state school) put the economic cost of a newly minted STEM PhD at approximately US$500,000. Additional training and experience received on the job increase this figure exponentially. Multiplying this figure by the estimated 3,000 PhD-trained women who are forced out of the STEM workforce every year (i.e. are not fully participating in the STEM workforce) results in an economic loss due to attrition of approximately US$1.5 billion per year. This figure is one-quarter the NSF annual budget.
Finding the Leaks
Why do women exit the STEM workforce? The answer is not genetic disposition or lack of interest. If this were the case, then female STEM students would underperform their male counterparts in college and graduate school. The data show the contrary: women outperform men academically, receive more awards, have higher graduation rates, and better attitudes toward education. Interviews, case studies, and statistical research consistently suggest that, from the multiple forces acting on women to leave STEM, two primary factors stand out: balancing career and family and professional networks.

A primary source of leakage out of the STEM pipeline results from family obligations. For both male and female scientists, marriage and family create demands that can cut short a thriving STEM career. Women’s biological time clocks often mean that decisions regarding marriage and children cannot always be delayed until after their career has been well established. Therefore, they are often forced to choose, very early in their careers, between being a scientist or a mother, resulting in women being pushed out of science, engineering, and entrepreneurial careers soon after graduate school.



Dozens of studies document the struggle to balance career and family. For example, in a recent study of 450 female scientists and engineers employed at research universities, more than 70 percent cited balancing career with family as the most significant challenge facing their career advancement. Perhaps the most convincing evidence of the adverse effects impacting the STEM workforce: among PhDs, single men and single women participate about equally in the STEM workforce; but a married female PhD is 13 percent less likely to be employed than a married male PhD. If the woman is married with young children, then she is 30 percent less likely to be employed.

A second major source of leakage results from lack of networking and mentoring. Women STEM faculty report fewer opportunities and referrals from collegial networks to participate in the commercial marketplace by being asked to consult, serve on science advisory boards, and interact with industry. This handicaps female researchers, causing them to become less socialized to commercial science and have fewer chances relative to their male colleagues to resolve ambiguities that many scientists hold about commercial science. Various studies confirm that female scientists are less comfortable selling themselves and their science in the entrepreneurial manner needed for commercialization. These attitudes can be detrimental, considering the fact that patents are key to personal wealth and career promotion, and may impact university tenure.

The results show up dramatically in the patent data. The percentage of women granted patents ranks significantly lower than that of their male peers. Not only is the percentage of women obtaining patents lower than men, but it also ranks very low relative to the percentage of women in the STEM disciplines. A study of over 1,000 recipients of National Institutes of Health (NIH) training grants in cellular and molecular biology revealed that 30 percent of men compared to 14 percent of women recipients had patented. A similar study of 4,227 life science faculty found that 5.65 percent of the women, but 13.0 percent of the men, held at least one patent, despite no significant differences in publication patterns. The lower percentage of women obtaining patents appears to hold across sectors of industry and government, as well as academia. In the information technology sector, the situation is even worse. A 2007 study from the National Center for Women and Information Technology reported that from 1980 to 2005, 93.9 percent of US origin patents came from men who constituted around 70 percent of the US IT workforce.

If women scientists and engineers are not obtaining patents at rates comparable to their participation in the STEM workforce and at significantly lower rates than their male peers, then women are not participating in the new areas and directions for science and technology. This hurts women scientists and engineers who are left out of the leading edge work in innovation, and it hurts US competitiveness.
Solutions: Integrate Women into Science
If the US needs more STEM labor and entrepreneurs, then it has three policy options. One option is to increase the number of male scientists and engineers. But despite various programs and initiatives to attract students of both genders, the percentage of male freshmen intending to major in undergraduate STEM degree programs has persistently declined, by an average of -0.6 percent annually for 20 years.

Supporting increased immigration of foreign scientists and engineers is another option. However, this approach faces several practical obstacles. For example, since 9/11, visa restrictions and delays have shrunk the number of foreign STEM workers and students entering the United States. This occurred simultaneously with more high-tech entrepreneurial opportunities opening abroad to attract technology workers away from the United States. Then, of course, there is political viability. As the current political debates on immigration reveal, little enthusiasm or political support exists for increased immigration.

The third remaining option is to pursue policies that increase the US STEM labor force by attracting and retaining women in science and high-tech entrepreneurship. This will require changes in the culture of science. Like other career paths, scientific research must be made more pro-family, less belligerently confrontational, and given a greater focus on the social applications of science and technology. Many men in science, as well as women, would welcome this shift in emphasis.
Policy Recommendations
Studies repeatedly show that, for both male and female STEM workers and entrepreneurs, a main source of leakage out of the STEM pipeline occurs because of family obligations. A simple policy solution would be for grant-making organizations to allow all applicants (male and female) to allocate grant money towards child/family/elder care, as the Clare Booth Luce Professorships from the Luce Foundation currently do. This would not create “special rights” for a protected class, as the GI Bill did for veterans. Subsidizing STEM families in this minor way could provide payoffs far beyond the initial investment. Indeed, simply having parents as STEM workers has been shown to increase the chances a child will go into STEM.



Second, STEM departments at US universities should incorporate marketing, finance, management, and other business training into graduate STEM education. For STEM labor to become more entrepreneurial, scientists must understand the risks and processes of commercialization, lab management, and marketing ideas to venture capital. Interview research reveals that high-tech employers seek STEM workers who understand project management, leadership, and business skills such as the ability to read financial statements and write proposals. Women often do not receive this mentoring in graduate school, just as in the 1970s they often were not mentored to write grants. High-tech entrepreneurship has been a tremendous source of growth and competitive advantage for the United States. If the United States is to regain this advantage, then it must continue to provide incentives for productive entrepreneurship and to enact policies which discourage the diversion of entrepreneurial talent into less productive sources of wealth.

A third uncontroversial and low-cost solution would be for government to simply enforce the existing anti-discrimination laws (e.g. Title VI, VII, and IX of the Civil Rights Act). Men are no longer banning women from their academic laboratories (as Madam Curie was until her second Nobel Prize) or withholding research funding for their employment (as was done by the federal funders of the Stanford Linear Accelerator Center during the 1960s). But cultural and institutional biases do creep in to chill the climate for women scientists. For example, the 1999 MIT Report found an unequal distribution of resources between male and female faculty in measured variables of laboratory space, salary supplements, start-up packages, proportion of university funding, and even prize nominations.

Fourth, the National Science Foundation’s (NSF) existing ADVANCE program should be expanded and redirected. Since its inception in 2001, this tiny program has achieved fantastic results at 28 top US research universities. The existing ADVANCE model should be reevaluated and redesigned for scaling up to the next level of tasks. This successful model at NSF should be expanded to include the other federal agencies such as the NIH, Defense Advanced Research Projects Agency, and Department of Energy: NIH is considering one adaptation of the ADVANCE model to fit biomedical careers. Some thought should be given to designing ADVANCE-type programs that emphasize patenting and entrepreneurship, connecting industry with academia to disseminate best practices and to increase ties between the two research communities.

Fifth, more aggressive promotion of qualified women to science advisory boards, science journal editorial boards, and science policy positions would make them more visible to venture capitalists and industry. Mentors need to provide appropriate guidance for their graduate students to undertake riskier projects, assert themselves to sell their ideas, and make introductions into men-only networks.
Grassroots Solutions
Solving this problem is not a task reserved for policymakers. Readers interested in maintaining the United States’ scientific edge can take actions on their own that would have great effect if multiplied across the STEM community. While admittedly controversial, both men and women must jettison the attitude that women can approach a STEM career just like men. The balancing act between family and career must be recognized and strategically planned for while still in college. Scientists of both genders must also realize that networking and commercialization are not “selling-out” but can be part and parcel of a productive career in science and technology. Men must jettison the attitude that women face the same obstacles as they do. Women need no special rights but access to the same laboratory space, start-up awards, and male-dominated networks.

Finally, the United States must fight the portrayal of women scientists as a special interest group. This is hardly appropriate since women constitute half of the US population and now earn over half of the undergraduate degrees in science. Given that having STEM skills creates a key divide amongst income earners, winnowing STEM-trained women from the workforce because of poor mentoring or failure to institute family-friendly policies risks re-segregating women. Supporters can get involved by urging US Congressional officials and state legislators (since state governments fund state universities) to enact simple and fair policies such as those suggested here, and contacting alma mater universities, as well as their children’s universities and the NSF, to strengthen and enforce existing statutes. If we fail to rid ourselves of anachronistic cultural biases and outdated policies, then we will lose out to countries that are able to do so.