Wednesday, October 16, 2013

Request for arguments against affirmative action

Allow me to start by posing a hypothetical situation:

A top-25 astronomy department has a major gender imbalance on their faculty. Let's say the fraction of women professors is below 10% of the overall faculty (This is a safe example since we don't actually know of such a department, do we? Right? Anyone?). 

Let's suppose that the upper administrators at said hypothetical university (e.g. the Dean of Sciences) would like to address this problem with a radical approach. If the astronomy department conducts a programmatic search for a woman junior professor and identifies a candidate that meets the high bar expected of the university and department, then a special faculty line will be made available that won't count against future departmental hires.

The reasoning for such a move is as follows:
  1. Studies show that departments with more gender balance can more effectively bring in excellent women students, postdocs and professors by offering a better overall environment for women to work. Increasing the number of women professors in the department counteracts unconscious bias, sexual harassment and provides the dept with advocates for progressive, family-friendly policies. 

  2. The increasing fraction of women among the department's student population would be better served by an equivalently diverse faculty. Many (but not all) women students would benefit from female role models, mentors, teachers and advisers in a manner that an all-male faculty cannot offer. 

  3. University and departmental policies and attitudes over the past century have greatly reduced the available talent pool by arbitrarily drawing from a population that was almost exclusively white male. This reduces the number of perspectives and backgrounds that contribute to creative solutions to long-standing problems. An expanded talent pool could bring break-through, game-changing solutions to many old problems while simultaneously developing entirely new fields of study. All of this will contribute to the department moving from top-25 to top-10 and beyond.
Okay, there are the arguments for a programmatic search for a female professor. I'd like to solicit
counterarguments from you, dear reader. 
  1. Why would this approach be a bad move for the department, university or field of astronomy?
  2. If this is a bad approach, what should the university/dept do instead?
Please refrain from one-liner comments sent from your iPhone. If you have a good argument, give it the time, thought and space it deserves. Similarly, please refrain from feeding the trolls by responding to one-liners. Don't give them the time and effort they don't deserve.

In addition to leaving thoughtful comments, I'd appreciate it if you send me a private email with your long-form counterargument, with the subject "Counterargument" so I can be sure to collect the full list. 

In future posts I'll present the counterarguments, along with my thoughts. This means you should bring your A-game, because I will not hesitate to wipe up any weak-sauce sent my way. 

Before commenting/emailing, please consult Graham's Hierarchy of Disagreement, below (credit Wikipedia). Specifically, please avoid emotional and/or ad hominem B.S. Keep it constructive and I feel certain that we'll all learn something from this thought exercise.


Anonymous said...

This is not a counter-argument, but rather a set of complicating factors to consider. The factors are: backlash, resentment, and/or lack of support from existing faculty. These types of hires do exist, so you should talk to people who have accepted them about their experiences. I have been told that people accepted into these sorts of positions can experience severe affirmative action backlash, including but not limited to: explicit comments that they were not qualified for the job and only got it because they fit x characteristic, reduced amounts of tangible and intangible support from the department, resentment or lack of respect from other faculty for having it "easier," etc. In the long run, it is *more* damaging for minorities if they are hired into positions where they are not supported and ultimately are driven to leave, by choice or by accumulated inequities. And the resentment can trickle down to junior members of the department: if postdocs, grad students, or undergrads get a whiff of the resentment, it affects department culture and the feelings of identity of students who fit the minority criterion. Furthermore, specific to the subject of women in astronomy, there is a current trend for departments to try to snatch "good" women young. I believe that this trend comes from a place of good intentions, but it effectively puts these women in a weaker bargaining position when it comes time to negotiate startup and salary, so it can ultimately lead to inequities in resources. If institutions do these hires well (with all the faculty on the same page about the hire and with a department chair willing and able to negotiate on behalf of the candidate), then I think it has the potential to be a very powerful approach. But my purely anecdotal evidence suggests that departments often do not do it well. (History: I was targeted for a hire very like what you describe, so I thought a lot about these issues and talked to several people who had been in the same situation. I didn't wind up taking the job for unrelated reasons -- I suspect I would have loved the job and department, but my passion and long-term goals were better served by taking a non-targeted position at a different university.)

Anonymous said...

Before attempting to post counter-arguments to affirmative action (I actually have some), I would like to stress a 'non sequitur' in the reasoning.

Assuming that points 1, 2, and 3 are true does not yields to the conclusion that affirmative action works, at least with the information given in the post.

In other words, if it is true that a dept is in better shape by having little or no gender inbalance, it is not necessarily true that affirmative action is the way to go. It might be that in those depts where there is no gender inbalance there was no affirmative action. Maybe they are like that by chance, maybe because there was a conscious attempt to counteract discrimination, which does not imply affirmative action.

Anonymous said...

I know someone who was hired under these circumstances. After teaching for a year at my university during a postdoc nearby, this person was offered and accepted a tenure-track position elsewhere. Within months, my university created a new faculty position to lure the person back here.

Today, this person is my department chair. He is universally respected on campus as an intellectual giant and a leader. He is the only full professor in my department and unanimously loved by his faculty. Resources fall down upon him like rain.

Of course, anecdote is not singular for data. We all should know that.

Furthermore, would hiring the Nobel laureate be considered discrimination against non-Nobel laureates? Would hiring Dr X's spouse be considered discrimination against all those scientists who didn't marry Dr. X? Would hiring the theorist be discrimination against the observers? In the end, we hire what we want the faculty to be, and there are powerful reasons to do so.

Anonymous said...

I think it would be quite naive to believe "a special faculty line will be made available that won't count against future departmental hires."
It may be true for a year or maybe several, but deans and other administrators change, sometimes quite frequently. It is inevitable that eventually during ~three decades decades lifetime of a successful faculty member, all existing hires will "count" when deciding whether to allocate additional hires to the department.

That isn't an argument against affirmative action, and I doubt it would change many people's opinions about whether they should go ahead or not.
However, I'd suggest that it is something that people should keep in mind, any time they are weighing promises from administrators.
That applies to more than just affirmative action hires, also spousal hires, target of opportunity hires, person in specific field we like hires, etc.

Note that there is also a counterarguement. A faculty hire in your department is worth two hypothetical hires in the future.

Even if you don't agree with the current priorities for your department's next hire (whether that be due to affirmative action or just sub-discipline), you should think twice before turning down an opportunity to hire any tenure-line faculty member who would is likely to strengthen your department.

Anonymous said...

This is a very bad idea. There are two ways this could go:

1) The department explicitly advertises something like "Seeking female astrophysicist". This is overt discrimination and is _illegal_ under Title VII of the Civil Rights Act, among other things. (The university lawyers would forbid something this blatant if they are worth their salaries; if they didn't, it could lead to an expensive lawsuit.)

2) the department has either an explicit or a "gentlemen's agreement" (HA!) with the administration to advertise a job to the unwashed masses but to only hire a female, eg. "Seeking tenure-track astrophysicist...women and minorities especially encouraged to apply". This is _still illegal_ if such an agreement exists. It would be harder to prove unless there was an email trail, etc, so probably the department could get away with it if they went at it strategically. If knowledge of such an agreement became public, however, male applicants turned down for the position would have standing to sue (for lost wages, etc), and would have a good chance of winning if there was a paper trail or if the administrators/faculty were forced to confirm the existence of such an agreement. Of course, if there were no paper trail, it would be easier to avoid conviction; one would only have to make vague points about the qualifications of the female faculty higher being higher than the plaintiff. (Note-I am not advocating breaking the law.)

The legal point here is that the job requirements of an astrophysicist, unlike the job requirements of a bouncer, can be equally well fulfilled by men and women (people disagreeing with this and arguing for employment discrimination in favor of women should perhaps recognize a contradiction in their reasoning). Therefore, employment discrimination on the basis of sex, as the post above seems to advocate, is illegal.

Finally, (to slightly alter Justice Roberts' words), one might question the wisdom of combating discrimination on the basis of gender by discriminating on the basis of gender.

-Malaka Akbar

Anonymous said...

I support affirmative action at least on the margins. For instance, I could not support affirmative action if every single department decided that they would only hire people of type x or women until the balance is reached. I doubt that few of us could.

That being said I would have two arguments against affirmative action in this case. First, affirmative action at least in the hiring of faculty puts the bulk of the cost on correcting historical imbalance on the people least responsible for it. Senior faculty who have arguably benefitted the most pay nearly no cost, while postdocs and other early career scientists pay a price. Given the rarity of faculty jobs, let us not pretend that there is no cost. I would argue that if we were to be more fair about this, we could do a two different strategies to spread out the pain.

1. Insist on mandatory retirement at 65. This opens up jobs that can be filled by women and minorities, while not impacting other young scientists too adversely.

2. Allocate 20% of the funding from national funding agencies to be directed to a special fund that is earmarked to support women and minority scientists. Similar to the NSF early career, but gear toward helping women and minorities really stand out in a competitive job market.

Second, while the emphasis on women in this article is commendable, I would argue that this is not the biggest diversity challenge anymore. Instead, it is the lack of underrepresented minorities at the mid and upper echelons of the academic ladder that is the real scandal. So I would argue that the dean should allow a more general hire: women or underrepresented minority.

On this final point, I would say that this can and does lead to situations that I would prefer not to be party to. For instance, I know someone who once asked me to suggest a few diversity candidates. I thought a bit and suggested person x, who I know and think highly of. A few months later, I ask about how the search went. "Well," this person said, "x is the wrong kind of ethnicity y." I was never sure what this means to be the "wrong kind" of a particular ethnicity. In this sense, hiring women is easier. Due to a quirk of biology, you are or you aren't, gender identity issues not-withstanding.

Anonymous said...

This is not a counterargument but rather a suggested ammendment to your process. It works better if the administration offers incentives rather than specifically dedicated FTE. If the search is open, but the admin is open to creating new (extra) hires for outstanding women candidates, there is an incentive for departments to hire women. This may seem like a cosmetic change, but it requires a department to actively seek the additional appointment and, in my observation, substantially improves the political environment for the incoming hire. They would not be there if a majority of the department did not want them to be.

Anonymous said...

It is not clear to me that the three reasons given for affirmative action lead to the conclusion that it is the right form of action.

In regards to reason 1:
Without disputing that gender balance is one way to achieve the goals stated in reason 1, it could be possible that there is a more effective course of action. For instance, it could be possible that diversifying or strengthening fields of research could be effective at attracting excellent women to the department. Perhaps building a specific program or activities to directly address bias, sexual harassment, and family-friendly policies could be more efficient than bringing in women as professors who are focused on research.

2) Considering most students do not end up becoming professors, it may be more effective to bring in role models and mentors that more accurately represent the future career paths of the department's students. Students are also better served by faculty who can attract top talent, grants, and are excellent mentors, advisers, and managers who care about their students.

3) I believe (although I could be mistaken) that studies have shown that the reduction in talent pool happens primarily before students enter university or grad school. Focusing attention and effort at younger students would be more effective at increasing the number of perspectives and backgrounds. Breaking down by economic background, race, or gender, are women the most disproportionately represented? Or is it people from lower economic backgrounds? Or specific 'races'? In terms of talent pools, wouldn't an international population be much larger and diverse and therefore important to attract?

All this being said, I do think it is important to address the gender imbalance, and sometimes in complicated situations embracing a effective method is acceptable when it's not obvious what the most effective method would be.

Dr. Krishna Kumari Challa said...

I feel both affirmative action and a sincere effort at individual levels should go hand in hand. Please read one of my articles based on this here:

Anonymous said...

"Increasing the number of women professors in the department... provides the dept with advocates for progressive, family-friendly policies."

We shouldn't assume women will take on the role of advocates for family-friendly policies. Neither should we dismiss the potential that men have as advocates for this cause. Any one woman may or may not choose to advocate about these sorts of policies. They may be more interested in other issues, like dyslexia or journal funding. They may not want to raise issues that may lead to them being perceived as weak or uncommitted, or challenging the department's norms. At the heart of the above statement is a stereotype about all women as mothers and carers. Although it is true in US society that the burden of care tasks primarily falls upon women, it is not true that all women are carers, and especially not true that any random woman must be one. There's also a stereotype about family-friendly policies about being about children, ignoring ageing parents, disabled siblings and spouses who just want to get through a whole weekend without you checking your email!

Anonymous said...

Counterarguments to gender-based affirmative action in faculty hires:

1. Equality of opportunity is a better and fairer goal than equality of outcome, especially at the end stages. There is not (and perhaps can not be) evidence that pools of talent and motivation are equal or higher for men or women. A goal of equal representation is a discriminatory bias.

2. Gender inequalities and discrimination are real. In our society, they pale in comparison to racial and socio-economic inequalities. It is damaging to pit one discriminated-against group against others. The strong focus on women in astronomy and the relatively weak focus on other disadvantaged groups effectively does this (pitting).

3. The argument that women need female role models reinforces negative stereotypes of women as weak and dependent. One rarely hears that men or Chinese or Europeans need a mentor who superficially shares their appearance and experience.

3b. It is commonly considered acceptable for women to prefer to mentor/hire, be mentored by, or collaborate with other women. Yet it is (and should be) unacceptable for other subgroups (excluding national and linguistic links) to prefer to working with people like themselves.

4. As resources become more limited, the backlash against affirmative action and related social engineering will certainly grow (see Europe). Making the current generation of males pay for the real and perceived sins of past generations is sure to create an increasingly divisive and toxic field.

5. Girls/women consistently outperform boys/men up to the collegiate level, at least. It is unclear that the current generation of men have a clear advantage that must be affirmatively acted against.

6. The oft-cited switched name/gender on applications bias appears real. It also appears statistically weak, with differences falling within one standard deviation. These studies also do not involve astronomers specifically. Astronomy hiring practices at the postdoc level and above are very different. In astro, if the applicant and their name (and their letter writers) are not already well known, then the odds of hiring are essentially nil. A study of this kind for grad student admissions in a mid-level astronomy dept. (i.e. where most successful applicants don't have papers they have presented at AAS) would be more instructive.

7. Once a perceived tilting of the scales exists, it is difficult to predict or stop the back reactions. The increase in (perceived) affirmative action hires can lead to an unconscious bias for white men in "open" hires. "It's this guy's only shot." This effect could explain the name-switching bias.