Tuesday, December 11, 2007

Biased Accommodations

As a society we tend to be much more accommodating of some commitments (e.g. religious or familial) than others (hobbies, etc.). Is this fair? Infinite Injury offers an example from UC Berkeley:
The clear sense these rules convey is that the instructor is expected to bend their rules if they might create difficulty or hardship for someone who wishes to respect a religious obligation but that a student who is going to be absent for some other extra-curricular activity undertakes a greater obligation if they want to miss class. Now one might justify such a policy on the grounds that some athletes or musicians are going to be out of town on a large number of dates or that religion is more important to people. However, it would be easy to give every student a certain number of absences they can exercise using the easier standard and there are many students who are more casual about the religious observances they ask to be excused for then athletes are about their games...

The things that we [non-believers] may really really care about get no accommodation while just someone has a ridiculous belief about some historical event we have to bend over to accommodate them. Now I fully understand that the potential for religious discrimination is great but if we weren’t implicitly endorsing religion as something more important than say a rocketry hobby we would use some fully neutral policy that gave everyone the chance to do what they really cared about.

That seems exactly right to me. For a more controversial example, then, consider the view that we ("society", i.e. employers and institutions) ought to make a special effort to accommodate those who choose to raise children. I'm sympathetic to this view. But is it biased? Why is the choice to raise children more worthy of accommodation than the choice to write a novel or compose a symphony in one's spare time? Perhaps we ought to be more accommodating in general, not singling out 'family support' as a uniquely worthy form of support. Fairer, perhaps, to enable individuals to pursue whatever projects are most important to them -- and for many this will happen to be childrearing, but for others it may not.

What is the best argument for singling out childcare? (I would look for consequentialist considerations, e.g. the impact that parenting has on the next generation. But let's bracket that for now.) Feminists sometimes claim that lack of support for women who want to both work and raise children is sexist. But it isn't entirely clear why this is so. (We can't always get everything we want. That's an annoying fact of life, not necessarily a sign of oppression.) Back to II:
The arguments given about the problems for women with babies in academia all focused on the extra time and energy women put into childcare. Now if women put more effort into children simply because they find raising children more rewarding (relative to men) the fact more women than men drop out to raise children is actually the desired outcome. It’s what would result from perfectly fair mutually beneficial trades. On the other hand if you think that the extra effort women put into childrearing isn’t the result of fair deals then the target should be on encouraging women to put less effort into childrearing, not making the unfair division of labor slightly less bad for women.

And a thought-provoking analogy:
Men are underrepresented in K-12 teaching. The reason most men abandon teaching is the difficulty of taking a high paying job in business and being a teacher. Therefore we should provide special benefits and accommodations to let men teach while still working as businessmen in the day. Obviously this argument is fallacious. If people are leaving some profession because they’ve found a better offer they don’t deserve special treatment as a result and it should only be fixed if luring them back provides a good value. Thus whether or not this is a leak we should be plugging is an empirical economic question and it’s only in the face of real data on marginal costs and productivities that we can answer whether or not we should address the ‘problem’.

Perhaps childrearing and religion are presupposed as normal components of the good life, and so it is thought that they should not be subject to trade-offs in the manner of our (other) chosen values. Writing a novel is a choice you make, and a somewhat peculiar one at that; having kids, on the other hand, is simply par for the course in a 'normal' life. This difference in normality may be thought to underwrite the special obligation to accommodate the one choice more than the other. But why should normality matter?


  1. Oh, Pierrot...

    The difference between childrearing and a hobby is that childrearing is a necessity for the perpetuation of the species. Someone needs to do it. if not women than men. But whoever is doing it is not simply performing an act for their own personal amusement that society could well do without.
    If all novel-writers chose not to write novels for the sake of their careers it would not be catastrophic for society. If all religious people decided not to engage in a religious practice because it came with heavy career costs, it would also not be a huge problem for society (prima facie). If all women and men decided not to have children (or worse, had them but did not rear them) the results would be disastrous. This does not show we should accommodate this way or that, but does show that your comparisons are quite out of place.

  2. If all people stopped farming food, that would be catastrophic for society. But that doesn't mean we must take extra-special care to minimize the opportunity costs of choosing to live on a farm, or to ensure that farmers (more than writers or musicians) can also pursue their other desires in life. So long as our actual arrangements produce enough food, they would seem just fine from this pragmatic perspective. And so it is for childrearing: "the perpetuation of the species" is not at all at risk. (It's not as though childrearing is this horrible burden that requires incentives in order to ensure that enough people do it.) So it's not at all clear why this is a relevant difference that you've pointed to, or how it renders my comparisons inapt.

  3. 1) A) when a person is taking care of their child the state places certain legal expectations on them - for example they not only choose to feed their child but they MUST feed their child or face the consequences. They become responsible to an extent for their bad actions (vandalism lets say) and sometimes even their failure to improve themselves (eg truancy). Parents who don't get custody get financial costs attributed to them in a special way - for example I might have to pay for a child and yet another person might not have to pay for the fact that they let their cat have hundreds of kittens. And if they did have to pay it would probably only cover a short time frame as opposed to the sort of time frame one might have to pay for a child.

    One could argue that if we expect to place these costs on parents we should compensate them - and that compensation would need to be pretty significant.

    B) From a Universal benefit point of view one could argue babies dont get benefits - but they should (so there needs to be a monetary compensation.

    C) and from a utilitarian perspective - if there is an optimal level of 'richness' (for a group) any specific long term cost might upset that equilibrium and force those individuals into poverty while under all the same pressues as others. So you might want to push them back to the optimal situation as best as you can with simple interventions - like a child subsidy. this could allows the parents to have the same income and ability to seek jobs as non parents - therefore competing on an even playing field and allowing optimal allocation (of course the world isn't fair and there isn't optimal allocation in reality - but still making it less fair will on average tend to hurt allocation efficiency).

    2) religions are usually accommodated because the organization making the rules knows it can defeat a hobby like "building paper planes" and doubts it can defeat a religion like "Islam". It is prudent for a university for example to not pick a fight with a religion. You can argue that we should pay no regard to that sort of pragmatism but if we did we could easily find ourselves fighting to the death over all sorts of unimportant issues - which would obviously be a much worse net outcome.

    In the end we create some rule and pretend that they have some sort of a right to make it les obvious that we are just capitulating to the force of their will and to contain that capitulation to a certain domain.


  4. I imagine that most religious people would be much more distressed by the impairment of their religious practice than their hobbies, because their immortal soul is at stake! This is a consequentialist reason not to interfere with religion - but what to do about someone with the (about equally justifiable) belief that their model rocketry is necessary for their salvation? How important is it to respect preferences based on wildly unlikely factual beliefs?

  5. Richard,

    I don´t think it´s the difference with regards to what´s normal that is underwriting the special obligation to accommodate a choice here, but rather that the fact of the ´normality´of a choice such as childbearing is trakcing something else. And what it´s tracking is that childbearing isn´t a choice in the same robust sense that, say, choosing to write an opera or to play softball is a choice. Of course, on a trivial level childbearing is a choice, as an individual man or woman can choose whether or not to bear children. But there are evolutionary and societal reasons why it is much more likely than not that any individual woman will have a strong desire. Now, these are not the whole story (in the case of evolutionary reasons) nor am I saying that these reasons are necessarily justified (in the case of societal reasons). But in general I take one argument to be that construing childbearing as a choice is misleading, or at the very least it is a choice in a much less robust sense than some of the other choices you identified. This, combined with the fact that childraising is in part a public as well as a private good, and that the costs of childbearing fall predominantly on women (again, I´m not jutifyin this fact, merely pointing it out), meaning that the effects of not accommodating this choice amount to effective discrimination, are the reasons why we should make special accommodations for those who raise children.

  6. If all people stopped farming food, that would be catastrophic for society. But that doesn't mean we must take extra-special care to minimize the opportunity costs of choosing to live on a farm, or to ensure that farmers (more than writers or musicians) can also pursue their other desires in life.

    But we often do think we must take extra-special care of this sort; if you live in a rural farming community you will often find precisely this sort of extra-special accommodation in matters of farming. It's just not particularly feasible to do much about farming if you live in an urban setting, or a rural setting where farming is not a major part of the local livelihood. (Even with rural farming communities different locations will have somewhat different systems of accommodation.) The difference with childbearing is that it is in no way localized, it's a concern whose importance everyone can recognize, and it is one that garners widespread sympathy because there are many parents in this world who fully understand the difficulties of it.

    Religious accommodation seems to me to be different entirely. Nobody accommodates religions consisting of one person. When accommodation is given to a particular person for religious reasons, what is really being accommodated is a relatively large, relatively unified group of people, with well-established institutions, engaged in, or at least convinced that they are engaged in (and in such a way that, as a matter of mere jurisdiction, so to speak, one doesn't have the right to say officially that they're wrong), a project of fundamental importance for self and society, which has a number of broad, vague similarities with other projects of the same sort (i.e., other religions). There is no parallel to "nonbelievers," who are not a unified group, and do not, as a unified group, have well-established institutions; and so with particular hobbies. There may be particular nonbelieving institutions, or particular hobby societies, that might in some places be candidates for some sort of accommodation along vaguely similar lines, but there is no overall parallel here. For the most part they would be different types of accommodation entirely.

    The disparity between sports, or music, and religion is a more interesting one, I think.

  7. Vanessa, I don't know what you mean by "choice in the robust sense." I can imagine a tortured artist whose drive to compose music is as strong as anyone's desire for children. You seem to want to move beyond the individual level, to a statistical claim about how common the desire is among women generally. But how is the group-level at all relevant? Surely what matters is each individual's ability to choose, and you grant that "an individual man or woman can choose whether or not to bear children." (I don't know why you call this the "trivial level", when it is the only level that matters.) Some people will have particularly strong desires, of course, but (i) the same is true of anything, as per my tortured artist example; (ii) that doesn't mean they aren't free to choose, it just means the choice of what projects to pursue will be really easy for them -- lucky! and (iii) I can't see why the demographic distribution matters, or why you would see any interest in the fact that "any individual woman" is especially likely to have this strong desire.

    Nobody thinks it's discriminatory that bar-cruising men are likely to spend more in their attempts to pick up partners. And of course you can equally tell a story about the evolutionary and societal reasons that men are especially likely to "have a strong desire" for sex. That hardly suffices to show that it deserves special accommodation.

    I'm yet to see any reason to think that special targeting is more appropriate than a more general empowerment (say through basic income or the likes) that individuals can then use in the service of whatever goals they please. People have all sorts of desires, and if you're going to privilege just one of them, that looks like a straight transfer which benefits the people with that desire, at the cost of those who lack it. My question then becomes: are there any impartial reasons why this should be done, or is mere factional / special-interest lobbying?

  8. This is simple: Religious groups have power and good child rearing practices have substantial externalities beyond the pleasure of practicing them, and also benefits with long time horizons, which people tend to under-invest in.

  9. OK, what i was meaning by choice in the 'robust' sense is eomthing like this: right now, I can choose to eat a sandwich, or not. There are no strong outside influencing factors. But if someone were to put a gun to my head, and say 'eat that sandwich or I'll shoot you,' it's still true that I have a choice, but it doesn't seem as 'robust' a choice as the first choice I pointed to.

    I understand your worry about what mattering is the individual's ability to choose, but I still think that what happens on the general level is important, and I think there has to be some way to factor in general facts about population groups into our reasonings. For example, its clearly the case that each individual can chose whether or nto to commit a crime. But the fact that, say, the NZ prison population is disproportionately Maori tells you something useful. So in this case, special targeting of programmes for offenders may be an especially effective way of reducing the prison population generally. So I din't think it's fair to dismiss the idea that macro-level generalisations are never useful or important because of generalizations are never true of an individual.

    It's funny you raised the male sex drive - one example I was thinking of is this: Imagine it were the case that for some reason this strong desire for sex meant that men's working lives were affected i much the same way women's wokring lives were affected by pregnancy. Would we still think of the desire for sex as a 'choice' and expect men to bear the consequences? After all, they could simply choose to never have sex, right?

    But I take your main point to be that even if my argument is that we should accommodate things that, for various reasons, people have less of a choice over, then there are all sorts of things which might fall into this category. I'm just going to fall back on practicality here, and say that it's very difficult to tailor policy to accommodate facts about individuals than facts about groups. The point is this: if we don't feel like we want to accommodate things that arew choices in the strong sense (I'm assuming this rather than arguing for it) then we need some way of establishing what kind of choices aren't. And the only practical way to do this is on the general rather than individual level.

    So, this is a bit confused, but maybe we can talk about it later.

  10. Vanessa - I agree that general facts can be important for some purposes. I also agree that not all decisions qualify as free choices in the sufficiently robust sense (external coercion, or even just a lack of reasonable alternatives, can undermine free choice). But I don't see any overlap between these two claims. It doesn't seem that - holding fixed the facts about the particular individual - mere statistical generalizations about others in their group could affect whether their own choice was properly robust in the relevant sense. At least, I'd need to hear a lot more about how this is supposed to work.

    "Would we still think of the desire for sex as a 'choice' and expect men to bear the consequences? After all, they could simply choose to never have sex, right?"

    Yes, exactly.

    "it's very difficult to tailor policy to accommodate facts about individuals than facts about groups."

    Yes, that's reason #4 to favour fully general (universal, non-targeted) benefits.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.