Last week I wrote a post about how summer vacation exacerbated inequality and Conor Friedersdorf wrote a response. I procrastinated like crazy and didn't issue a rejoinder. But since Conor F has now revisited the subject over at his brand-spanking-new Atlantic Ideas Blog (and because I've never before had the opportunity to argue with someone else named Conor), I'll add a few more thoughts on the subject.

Conor F doesn't want to get rid of summer vacation for a bunch of reasons, some of which I won't quibble with (he says he wants his children to have summer as a "lovely reverie when I get to enjoy their company more than I do during the school year") and some of which I will. He writes, for instance, that he would prefer not to "rob the wealthy kids" of their edifying summer experiences. I'm happy to break bread with the libertarians, but I think that's putting it a bit strongly. Wealthy parents who are unhappy with public education are perfectly free to take their children out of public schools and put them in private ones, or keep them at home. This imposes a cost (citizenship often does, alas). But to the extent there's a concern here, I don't know why it shouldn't apply to the existing 180 school days just as strongly as to ones we might add in the future. And again: since the current school year doesn't stem from any modern policy considerations -- it's all historical accident -- what reason do we have to think it's the optimal one?

And that question -- what's the optimal policy? -- is really the heart of the matter, and the heart of Conor F's critique. He writes:

Give me Pareto optimality, please!

To which I can respond: No! The criteria of Pareto efficiency should not -- in many ways, cannot -- apply here.

Well, sort of. To some extent there is obvious room for Pareto improvement, and I would hope Conor F and I can agree on it. Let's assume we are allocating two goods -- test scores and vacation days -- among all the schoolchildren in the country. A Pareto improvement would be to increase the test scores without decreasing the vacation days, or vice versa. So, here's an easy way to do that: We spread the current number of vacation days out over the course of the year. Since the big problem with summer vacation from a educational outcome perspective is that it increases "summer learning loss" by putting the vacation days all in one big chunk, the low-hanging fruit would be to break the summer into small pieces to increase test scores. (That is, unless individual utility exhibits increasing returns to scale with longer vacations. But my intuition is that the opposite is true.)

The problem with taking the Pareto criteria any further is that we aren't talking only about increasing individual utility. We're talking about social utility -- about the benefits of lower inequality and the positive externalities of improved test scores. If our social welfare function were very simple -- say, if it just took the sum of all individual utilities -- then maybe we would have use for the Pareto criteria. But if you pick any other social welfare function, I'm not sure why we'd want to talk Pareto optimality at all.

Update 7pm: Conor F responds here (and, disappointingly, refers to me as "Mr. Clarke," thus making the debate less confusing). What he wants to see me address is "why we would assume two goods, test scores and vacation days, and weigh them against one another." I thought I had addressed that point. But allow me to belabor it!

 
I don't think we should assume two goods and weigh them. I introduced that example because that's what Pareto efficiency criteria are good for: weighing various allocations of goods among some group of individuals. But once you stop talking about individuals -- and start talking about different social welfare functions and justice and fairness and freedom and blah blah blah -- the usefulness of the Pareto criteria falls away. And if the debate is over the abstract value of greater equality of opportunity versus the abstract value of a poetic summer reverie, those criteria will be very useless indeed.

The only thing I will add is that I want greater equality of opportunity for reasons that I think Conor F should endorse. (And it is opportunity, and not outcome, that I am concerned with here. All the talk of educational 'outcomes' might be misleading in that regard, but make no mistake: Education is fundamentally about creating equal opportunities.) There are lots of interesting outcomes to pursue in the world, and it would be a shame for a large segment of society to be randomly prevented from achieiving them.    

Mikebaird's Flickr photostream seems to have several good pictures of children playing at the beach. Via Creative Commons.