Though everyone recognizes there is a problem, during a generation when lots of efforts were made to increase diversity, the science, technology, engineering, and math (STEM) pipeline in academia remains primarily liberal white men.

That's not to say there haven't been efforts. Women and minorities are well-represented, though groups with less advocacy, like handicapped people and political conservatives, are routinely dismissed when they note that.

Is all that true? Andrew G. Campbell, associate professor of biology at Brown, argues that academia has become content with the thought that the STEM pipeline rests on flat terrain, passively and reliably conveying to the finish whatever quantity of students enter.  So if handicapped people or Republicans or ethnic minorities would apply for jobs at universities, they are just as likely to get them as anyone else.

Writing in BioScience, Campbell and a colleague argue that the pipeline has a steep rise against a gravity of endemic hindrances and cite data showing that for decades many students haven't made it to the top of the pipeline. What's needed to stem the leaks and backflow, is consistently applied energy all the way through the pipeline. Campbell and co-author Dr. Stacy-Ann Allen-Ramdial are only talking about women and select minorities, which ironically demonstrates the problem - selective blindness - that impedes true diversity in university's highest echelons.

It isn't numbers. Thanks to billions in taxpayer dollars giving out grants to boost STEM, there are 6X as many PhDs being pushed out each year as there are jobs in academia. The glut is so rampant some labs can advertise that post-docs work for free. Some labs even advertise a position, but you have to pay to be there. More STEM is not needed, but the authors don't accept that the best people are getting ahead. Science academia is not a meritocracy the way it is portrayed, they believe.

The data appear encouraging for ethnic minorities at the pipeline entrance - at least the only ones we are told to care about. Asians are 5 percent of the US population, so far smaller than black people, Latinos or native Americans but none of Asia counts as minorities: there are similar proportions of these underrepresented minority (URM) and non-URM incoming college freshmen (a little more than a third in each case) express an intent to study STEM subjects. Generally, however, black and Latino students are less likely to graduate than Asians, Indians or caucasian students.

While 24.1 percent of U.S. college freshmen came from these preferential minority groups in 2000, only 18.5 percent of bachelor's degree recipients did in 2004. For handicapped people, the numbers are also at least close to being representative, it's only after undergraduate that the real disparity kicks in. That is mimicked in Republicans, blacks and Latinos. Even women show a drop outside the life and social sciences. Grad school becomes a liberal white boys' club - as long as Asians don't count.

They cite National Science Foundation statistics which found that, after college in 2009, 36 percent of non-Asian minority students holding STEM bachelor's degrees left the field rather than starting a STEM job or graduate program, compared to 30 percent of comparable Asians and caucasians students. Non-Asian and non-caucasian minority bachelor's degree holders were more unlikely to earn doctorates. Non-Asian minority students earned 18.3 percent of the STEM bachelor's degrees in 2004 but only 12.1 percent of the STEM doctorates in 2010.

The diversity problem has gotten worse. It's obvious looking around a graduate school lab but it can be clouded by the presence of Asians and Indians. However, with handicapped people it is apparent because they stand out. For politically-oppressed minorities, it's harder to know, but you aren't going to find many conservatives under the age of 60. There are more Communist party members in academia than there are Republicans.

To serve the volume of students entering the pipeline, Campbell and Allen-Ramdial argue that educators and policymakers must enhance the conditions that will move them through the pipeline despite academic resistance. The authors propose four ideas based in educational research and practices that have emerged in recent years: alignment of culture and climate; partnerships between research and minority-serving universities; critical masses of minority students; and faculty engagement in diversity.

Diversity is listed as a core value in STEM fields and elsewhere, but that culture can become misaligned with the actual climate in which faculty members, administrators, and students work.

For one thing, it comes across as hypocritical. From the beginning of the process, schools use a 'secret sauce' for admissions, where Asians need higher scores than caucasians who need higher scores than blacks and Latinos. Then the contempt for outsiders is instilled early and augmented a pretense of latching onto every cultural fad. Diversity is a buzzword, in practice few departments follow it. Implicit stereotypes take their toll when people know the system is hypocritical.

Campbell and Allen-Ramdial recommend annual, confidential surveys of culture-climate alignment conducted by third parties, rather than the administration itself. They believe that if the administration knew of the problem, it would get solved. Larry Summers ran Harvard and said girls couldn't do math. Would a survey that showed women felt patronized at Harvard have gotten action by him? Unlikely, nor would the concerns of handicapped people claiming they wash out in interviews or political conservatives who feel they have been forced to hide their opinions or have their careers ruined.

Partnerships and focused recruiting - 'certain minorities are encouraged to apply' -  are les convincing. The authors argue that not only must non-Asian minorities be recruited, they must all come from similar 'backgrounds'. They say that universities should recognize that students need to be with others with whom they can identify and from whom they can find support. So not only are science labs supposed to recruit a non-Asian minority, they have to pay double so they can bring a friend.

"What is critical mass [in a program]?" Campbell said. "That number is when students feel the greatest sense of belonging. It doesn't have to be hundreds. It could be five."

Which means absolutely nothing.

Finally, the paper argues, diversity must be a goal embraced not only by senior administrators but also by faculty members. University presidents come and go, on average, every decade but provosts and medical school deans last forever, 20 years on average. Tenure-line professors may be there until they die. 

"Faculty members should be incentivized to engage more deeply in diversity by making it a meaningful scholarly activity, alongside research and teaching," Campbell and Allen-Ramdial wrote. "The opportunity to formally report on diversity-related activities as part of annual review and reward criteria for merit and promotion should be established."