[Crossposted at
http://www.bopnews.com]
Steven Levitt claims to be a rogue economist, this is incentivization in action: being an outsider economist is better marketting. But, in fact, the reverse is the case - Levitt is the hottest, not necessarily best or most brilliant, but most incandescent - orthodox economist out there. He does what most professional economists do: statistical econometrics. The difference is that he is able to rapidly pare down the problem to a manageable number of variables. He's not necessarily better at number crunching than other top economists - he's better at looking in the right place.
His recent book "Freakonomics" covers a variety of topics, but it is the theory on abortion that is attracting the most attention.
In itself, it isn't a new idea. In fact, Ronald Reagan was criticized for, on one hand, claiming to have cut welfare roles as Governor of California, and not mentioning that the single largest contribution to the fact that the "welfare chiselers" he got rid of the abortion bill he signed. So the question of the relationship between abortion and lightening the social burden is already in play.
What Levitt chooses not to comment on, and Andrew Leonard takes him to task for. is on what the correlation between lowered crime rates and abortion means. The implication that some would like to leave in the air is that the underclass is also the criminal class.
That's not what is going on here, Levitt's work says that all other things being equal, that in our society, abortion had the effect of reducing crime. Note the importance of that ceteras pares - it changes the entire frame of the question. The question asked and answered is not "is abortion good?" or "is there a permanent racial underclass that is the criminal class?" But "given two Americas, the one we are in, and one that is in every other respect the same, except that abortion is marginally harder or easier to obtain, what would the expected difference in crime rates be?" And the answer comes back in the form of a number: what is the chance you could differentiate that America from this one given crime rate figure only.
This may sound geeky, and it is, but let me compress it down: imagine you are in one room listening to some music, but it is something like classical music which is alternately very loud and very soft. Your mischeivious sibling or descendant of some kind, twirls the volume knob. How much does he have to twirl it before you notice, if he just moves it a little, how long will it take you to notice? That's correlation. And the knob is the question of the study.
So that Levitt's work shows that there is a correlation between abortions and crime rates doesn't even show what the usual phrasing "do abortions lower crime rates" says. What it does say is that we live in the wrong America - that we have made choices which leave us short of the optimal America. Why can I say this? Because there should be no question about the trade off here. Can anyone say "I want more unwanted pregnancies and therefore more crime?"
It also doesn't say that there is a permanent underclass, what it does say is that the sex tax doesn't work. What? Where did that come from?
Let me explain.
Abortion reduces one of the consequences of sex. A large fraction though by no means all, of the anti-Roe movement is also an anti-sex movement. Having pregnancy is supposed to increase the cost of having sex, and therefore create a marginal disincentive to having sex. The higher the cost, the less sex right?
And what Levitt's work shows is that this doesn't turn out to be the case. That instead, the cost of sex falls, at least in part, on the child that is born - because if the sex tax worked, and there was intergenerational altruism and rational actors, the mother would refrain from getting pregnant at a high enough rate to prevent there from being a correlation. In otherwords, even if restricting abortion does act as a tax on sex, it is clear, in the real world, that some of the people who it is targetted at, don't pay it in full.
In short, the sex tax does not work as a marginal disincentive, and one should be able to calculate the "crime tax" that results from this failure. That is, on average, what is the cost in crime?
<HR>
What is important to get at here is something that is missed by most reviews, but which Levitt knows, and many of the people talking to him do not. What Levitt has done is connect the micro-economic - having an abortion, growing up in a bad neighborhood, etc - to the macro-economic variable of the crime rate. he has connected the big and the little. The reason he can't make policy perscriptions is that the means of doing so - statistical survey of data - does not allow him to do so. He's not keeping his mouth closed because of a cop out, he's keeping his mouth closed because policy works on that black box in the middle.
And that middle ground is the realm of something which is neither micro-economics, nor macro-economics, hence what I term "meso-economics". What we have in the United States right now is a political language which does not correctly connect the foreground with the background. And not just on this issue, but on a host of issues.
And there are two primary reasons that we do not.
The first is that we are not taught how to connect the micro-economic picture to the macro-economic picture in 20th century modern terms. That is, we aren't educating people on how the century that has made much of the environment we lived in work. There is no theoretical break through in Levitt's hypothesis: it is straightforward application of intellectual elbow grease, along with an insight that one doesn't need to "polish the muffler" - that is, what he can eliminate from the picture and thus have a manageable problem. People don't understand this connection, and it isn't always easy to see how to make that connection.
The second is that the means we connect the two has, so far, not been measurable, but metaphysical. Even economics itself connects the big and the little together by means of a metaphysical creature, the "rational economic actor". But in politics we do so with "principles". But without being able to see the big and the little, these principles have less than nothing to do with what happens in that fog between the big and the little.
Since our whole system is on how to craft incentives that are enforceable by the big, on the little, that turn around and produce a new big that the little likes - this failure is one reason why we wallow around trying to solve the same problems year after year. We don't solve them, we slap some language down and leave people to muddle along the best they can. Often what we slap on is unusuable. For example, the "death penalty". As its name implies, it is meant to provide a marginal disincentive to commit crimes by making the possible outcome rather nasty. However, it doesn't do that. And what has happened is a rebellion against it, not one lead by liberals like me who see the death penalty as not serving any state interest and therefore unconstitutional on its face, but by the people who have to implement it. Many of these are conservatives.
Why is this? Because they find that the nature of the death penalty creates very strong incentives to warp the system. A death case is a big payoff for whoever brings it to trial and brings it all the way to the end of death row. Thus there is a perverse incentive to try, not the most dangerous criminals, but the ones that have committed the most telegenic crimes. There are others. And in each case, the changes made to avoid these problems lead to other problems. What is the truth beneath this is that the pathway of strategies between the beginning and the end lead inexorably into perverse incentives. The dominant strategies remove the clear linkage between action and cost, which was the heart of the death penalty advocates policy case. There is still the lust for revenge. But sensible people, even conservative sensible people, do not spend their lives meticulous documenting how they vindicated someone else's blood lust.
The middle realm then is one of behavior, of the real way that events happen. Economically, it is about finding ways to measure, and therefore model and understand, what goes on in this middle earth of economics. It is this that will determine whether policy ideas which, in theory, work in the micro-macro think world, actually work in ours. In economics we are learning what it means that "you can't get the-ah from he-ah."
<HR>
So what is necessary is two things. One one hand we need to teach people how the world of micro and macro think works. The standard issue textbook stuff of economics. On the other hand we also need to evolve our understanding of the mesoëconomic world (I just think that looks better than a hyphen, don't you?) and base our rhetoric more firmly in it. We have to have principles that match up to our understanding of how the world works, so that ordinary people may wrap their minds around what is being debated, and be able, in their own way come to terms with the choices that are being offered, and what those choices will mean in their daily lives.
It is this that will allow us to read a correlation like Levitt's and instead of arguing over where to put the knob on the stereo, be able to say "you know, we both don't like this music, let's put on something else." Because as Levitt notes: "decades of study show that a child born into adverse family environment is far more likely than other children to become a criminal." Meaning that we can patch this by reducing difficulty of obtaining an abortion, or we can realize that if there are persistent zones of adverse family circumstances, then someone, somewhere, wittingly or not, is profiting from the assymetry of opportunity. Afterall, almost no one likes to live in such circumstances, and if they are not, then there are fretters of the mind and society that are there.
And we are all better off if they are removed, rather than increased, by trying to implement a tax, whose "penalty" is the payoff that hundreds of millions of years of sexual reproductive biology aims at.
