Saturday, February 7, 2009

Education and Number of Children

From a friend:

"How would you solve the demographic problem we have where educated people are not having that many children? One avenue we were thinking was tax policy, i.e. tax ppl with fewer years of education more for each child born, but that doesn't seem very nice."

So firstly, it strikes me that taxing people with fewer years of education more for each child born doesn't get more educated people to have children - it gets uneducated people to have fewer children.

It's unclear whether this would be desirable. On the one hand, it could have some powerful social effects. Children in the US cost a LOT, so reducing the number of kids people have saves a lot of money. That likely reduces greenhouse gas emissions, reduces individual indebtedness (increasing savings and investment, which stimulates long term growth and makes everyone better off), and, most importantly, probably results in substitution of some of that saved money towards education or healthcare, which has all sorts of good effects. Also, it's not politically correct, but that tax policy could lower crime pretty significantly as well (teenage children of low-income parents are, I believe, the most likely to commit crime)... in fact, there's a fascinating-if-ethically-sensitive study a while back that showed that legalizing abortion significantly reduced crime a decade and a half or two decades later (if anyone can link to that, please comment with it!) because young parents, who are more frequently low-income, could have less children. All of these can be promoted by things like your proposed policy of taxing uneducated people for children, as well as increasing access to and decreasing cost of abortions, condoms, birth control, etc.

On the other hand, we're running into a demographic crisis. Social Security, as constructed, is almost Ponzi-esque. The young generation buys in and has to save enough to subsidize the old, but to do that, there need to be more of us because economic growth isn't ridiculously fast and we have to spend some of our income on ourselves. Therefore, our generation needs to have even MORE kids to fund us getting old. Many other entitlement programs sort of work like that, too (Medicare, etc), though less explicitly than SS. Anyway... Uruguay and parts of Europe (i'm sure other places, too) are having trouble because they have lots of old people and not many young people to help pay for them, so they run into major economic trouble.

There's also all sorts of ethical arguments which can go both ways, so it's unclear whether you want uneducated people to have fewer kids.

In any case, the problem you suggest is for educated parents to have more children. There seem to me to be a few ways of doing that.

Firstly, you could try reversing your tax argument and subsidizing more educated people for having more kids. This would be controversial, because it functionally reduces the tax burden on wealther people (although having a child decreases their wealth).

It would also likely be inefficient, because educated people tend to be wealthier and you would probably need to subsidize them a lot to make them willing to change a major life decision like having children, especially given that educated people have children at an older age and are thus more mature about their decision-making. In other words, educated people are probably not that elastic on cost in their childbirth decisions.

You also are much more likely to see the psychological thought process of "I'd love to have a kid but I can't afford it" (preventing people from having children) than you are to see "I didn't think I wanted a kid but now that it's cheaper, I'll go for it!" (getting people to have children). Economically, they're usually identical... psychologically, perhaps not so much. A psych major knows more about this than I do.

So tax policy seems a little iffy.

A better idea, I think, may be to subsidize companies and organizations to help them provide strong daycare options. Many educated people make decisions on kids based on how it will affect their career. Increasing the availability of quality childcare options so that career impact is lessened could be good.

Subsidizing mothers directly for daycare is less optimal, though it would still work. One would expect educated parents to care about having daycare that is a) convenient, b) safe and c) developmentally positive. Subsidizing organizations instead of individuals means that there would probably be a greater number of smaller centers, while subsidizing parents doesn't provide that geographic and sizing effect. A greater number of centers means it's more likely daycare will be convenient, and smaller ones will make it easier for centers to be safe for each kid with personal attention to behavior and learning. I can't imagine daycare fixed costs are all that high, so any inefficiencies you have by having a greater number of centers (and therefore an increased number of employees over a direct-parent-subsidy case) would hopefully be offset by the educational, safety and convenience benefits to mothers and children.

Making babysitting tax deductible probably achieves a similar effect. You likely could only make it tax deductible up to a certain amount, or you'll face fraud. In any case, that could be good, too.


  1. Trevor, why is this a problem in the first case?


  2. Hm, not a member of the Pigou Club, I gather...

    When I get a chance, I'll do a fast lit review to make sure that (1) we've identified a real phenomenon and (2) that there's data which can explain why this is a problem. I think there's some strong intuition to why this could be at least correlated to some problems, but having some support is always nice.

  3. haha I love Pigouvian taxes on many things (Gasoline, cigarettes, trans fats, incandescent lightbulbs, water bottles, etc)... i just don't think they work with childbirth.

    BZ, children of educated parents tend to get educated themselves. More people who are educated results in progress - overwhelmingly, a small number of people tend to be responsible for a massive percentage of the progress in most fields, so increasing the number of people who could fall into that category helps science, politics, the economy, education, healthcare, etc. There are lots of specific arguments, but I think most of them break down into that general trend.

  4. Trevor, if you haven't already found the abortion study (by Donohue and Levitt), their original QJE paper is here: