20 April 2017

Diversity is a bit of a politicized topic. For most of my blog posts so far, I've intentionally been avoiding topics like this, as discussions can often be unproductive. But that avoidance comes with the downside that I don't get the opportunity to talk about my views and allow them to evolve.

When people talk about diversity, there's a lot of components to it that are very different. In general, the goal of diversity initiatives is to create a society with minimal luck of birth. In the current world, there are a lot of ways in which your birth can affect your whole life. These include things like overtly racist or sexist policies, but also things like your family's economic situation and even the quality of your local school system.

Even within the goal of minimal luck of birth, there are different interpretations that often come into conflict. Should we work to minimize luck of birth for people right now or for people in 20 years, or even 100 years? Or is any sort of timeframe like this wrong and we should envision the ideal end state and work to minimize the time it takes to get there? It's not necessarily obvious that these ideas conflict, and it's difficult to give a straightforward example, but when discussing diversity it's important for every participant to understand where they stand on that tradeoff.

Affirmative Action

As a relatively young person, affirmative action is probably the diversity initiative that I'm most familiar with. It refers to the practice of preferring applicants from historically disadvantaged groups, usually in the context of college admissions.

First of all, let's be clear that any sort of preference does mean that the standards for qualification will be lower for the preferred group. As the applicant pool gets larger you'll have more flexibility in who you accept, so the difference between the standards can become smaller, but in practice the applicant pool for the disadvantaged group will be very small and as a result there are a lot of schools that have significantly different standards for applicants of different races.

I regard affirmative action as a bad idea. Especially for colleges, I believe that affirmative action is a negative for three groups of people, which add up to a huge section of students:

  • The people who get rejected in exchange for the preferred applicants. This group should be pretty obvious. They get less choice in schools, and they may end up going to a lower quality one.
  • The people of the disadvantaged group who would have been accepted regardless. In general, affirmative action creates the question "Did this person get accepted because of their credentials or because they are a ?" Even when the answer is the credentials, the fact that the question exists can be detrimental, especially since it takes a long time to learn the answer.
  • The accepted students that aren't part of the disadvantaged group. In my eyes, the biggest draw to a particular school is the other students that you interact with. Lowering the average quality of the students lowers the average quality of these interactions, and as a result lowers the quality of the education being offered.

In exchange for these negatives, affirmative action creates a world in which some members of a disadvantaged group will have access to a (presumably) better college education. The positive effect for a single student in this group is certainly much larger than the negative effects for a student in the other groups, but the negatives impact a much larger group of students. Is the tradeoff worth it? In my eyes, no.

In addition to my feeling that this tradeoff isn't worth it, I dislike affirmative action because it reinforces negative stereotypes. To illustrate, let's imagine an extreme example, where affirmative action creates a situation where the disadvantaged group (for brevity we'll call it group B, and the rest of the student body group A) is preferred so much that only 10% of the people from group B would get in otherwise. In the resulting world, 90% of interactions with group B students will be reinforcing the stereotype that group B students are worse. This feedback loop makes me believe that affirmative action programs extend the lifetime of these stereotypes.

Financial Aid

Colleges in the US are rather expensive, and the price is a big factor in many students' choice of school. Financial aid makes it so that students who would otherwise turn themselves away are able to attend. Unlike affirmative action, I think that financial aid is extremely beneficial and overall a great idea.

Unlike affirmative action, financial aid raises the standards for qualification by increasing the size of the pool of potential students. Whenever an extremely qualified student turns down a college because of price, every other student at that school suffers a little. Financial aid makes that situation less common.

Financial aid actually parallels the idea of a progressive tax system. The wealthy pay more taxes because money is disproportionately beneficial for the poor. As a result, a progressive tax is a fairer way to fund the government than a flat tax would be. Similarly, tuition serves as a way to fund a college. Students who come from wealthy families will be less pained by the burden of tuition, and as a result it makes sense for their tuition to be higher.

Financial aid tends to come in two flavors. One is financial aid for particular groups, such as minorities. The other is financial aid based on need, which is usually determined by the family's income. I strongly prefer need based financial aid, because the benefit of financial aid is preventing students from turning down the college due to cost, so the students receiving aid should be the ones who would otherwise be at risk of turning themselves away. Financial aid based on other metrics such as race can lead to confounding situations, where (for example), a financially secure black student might get aid while a struggling white student wouldn't.

That said, in the current world race and socioeconomic status are correlated reasonably strongly, so race based financial aid will still get a large amount of the distribution right, and I definitely prefer race based aid to no aid at all. I have also heard some horror stories of need based financial aid, where one parent has a good amount of wealth, but is unwilling to pay for college. In a perfect world, the cost of tuition shouldn't be a consideration for students at all, but it is a difficult road to get there.

Curriculum Changes

It's common to notice that in STEM (Science, Technology, Engineering, and Math) fields, you generally find significantly more males than females. One possible factor toward this phenomenon is that the education system pushes boys toward these fields and girls away from them. It's hard to say exactly to what extent this is true, but removing such an effect would be a good goal.

That said, it's important to avoid compromising the quality of education. For example, I read a newspaper article once that said programming classes appealed more to female students when there was more emphasis on fun and less emphasis on arcane symbols and text. Of course since the article was written for a general audience, it was probably simplified in a misleading way, but one has to wonder whether the resulting class is really worthy of being called a programming class. After all, a large part of a programmer's day is spent reading or writing these arcane texts. Furthermore, what the heck does it mean to emphasize fun and do boys not care about having fun?

Overall, I think that curriculum adjustments are a promising idea, but I am skeptical of the current initiatives on that front. I also believe that this should be a technique primarily aimed at lower education. Once you are studying to be an expert in a field, the things you are being taught should be the accepted best practice in that field.

Women-Only Events

Many competitions see a pattern of male-dominance. For example, only one woman has ever been in the top 10 chess players worldwide. In my own experience from math competitions, those are also dominated by males at the top level. As a result, some people have started women-only competitions.

It's hard to say whether I'm for or against these competitions. There's a very difficult balance to achieve here. Essentially, women-only competitions are trying to encourage women to enter into the competitive field by lowering the expectations for them. It's a similar idea as having local competitions, state competitions, and national competitions. They allow a wider audience to enjoy the sense of competition at a level right for them.

However, in practice women-only competitions have a very significant difference: they aren't considered stepping stones. When someone competes at a local competition, then if they win they will often set their sights on the state or national competition. This means that you get their attention with a lowered bar, and then turn that attention toward the top. When that happens, the field gains a valuable member.

[Math Prize for Girls] is a high school math competition that falls in this category. They offer larger prizes than almost any other high school math competition, and as such it makes it difficult to turn the resulting attention to the more prestigous competitions such as the USAMO. Essentially, the existence of such large prizes send the signal that that level is "good enough," and so it is reasonable for girls to be motivated to meet that level and then stop.

If there's one thing that humans are good at, it's meeting expectations with the bare minimum. I don't believe that Math Prize for Girls has had or will have any effect in making the US's team to the IMO include more girls. In fact, I am concerned that the lowered ceiling makes it even less likely for a girl to decide to pursue math to the extent of making the IMO team.