Adam points to a new study estimating a 400K social value of a good teacher. This is in line with other recent estimates. I want to make a couple of notes on how to think about that.
1) Social value isn’t a feel good concept. Hanuschek limits himself to future earnings of the students. The other big drivers are always crime reduction and public assistance reduction. So we are saying better teachers lead to higher wages, lower crime and less welfare. This is a far cry from trying to put numbers on soft factors like civic engagement.
2) As big as these numbers are they actually accord with the way people behave. In areas where private school dominates, parents obsess over getting Johnny into the right private school. There are obviously some peer factors at work here but at its heart a school is primarily a collection of teachers and classrooms. Unless we think the classrooms are really, really good, then the parents are implicitly obsessing over getting the right teachers.
Some of the most gripping scenes in a recent controversial movie shows parents in tears and praying over whether or not their child will win the lottery for entry into a particular school. Rarely do you see people crying and praying that they will get into say, a T.G.I.Fridays. There are some fine dining establishments in New York and elsewhere that might be a different matter.
Yet, we notice that when people are on their knees hoping that God will intervene to get them into a place, that typically translates into a high market value of said place.
So we get big numbers, but people also act like their ought to be big numbers.
3) Not only are university professors paid more than K12 teachers but the pay per instructional hour differs wildly because professors spend so little of their time instructing. We can have a whole debate about why that it is but at the end of the day the market is telling us that the value per hour of professorial instruction is extremely high.
Yet, our science tells gives us strong reason to believe that K12 instruction should be more valuable. In general the older one gets the less impact good education seems to have. Giving someone a great college professor should be less impactful then giving them a great Kindergarten teacher.
Now, it could simply be that the technical nature of university instruction means that great professors are in short supply, while there are plenty of great Kindergarten teachers. Thus, the short supply drives the wages of professors higher. That implies, however, that there is enormous consumer surplus just raining out every single Kindergarten classroom.
That’s not how parent behave. They behave as if some Kindergartens are much better than others. With very low prices on Kindergarten instruction, that looks like market failure.
4) I want to leave as light a touch as I can here but let me simply say that when market forces control neither the quantity nor the price in a market, the observed prices can diverge wildly from the optimum even when those setting the price and quantity have the best of intentions.
That is, if finding the best teachers is worth upwards of half-a-million a year then something is seriously going wrong in the structure of education. However, with no market check its entirely possible for things to go really, really wrong.
Now since there is no restriction on the production of private K12 education we would expect it to eventually come to dominate. The key thing to remember, however, is that there are credit constraints. The children are the ones who are going to be making the extra money, not the parents. The children can’t borrow because kids can’t enter into binding contracts. The parents can’t borrow because they themselves won’t actually have any higher earning capacity with which to repay the loan.
You could try to arrange something where the parents borrowed and then sought to guilt the children into paying the loan back but there is serious moral hazard here since its fully and totally in the parents interest to simply default and let their kid walk away with a great education and no debt.
Thus you have a major problem where no one in the private market can make a creditable commitment to repay the schools for the future value they produce.
5) Where we do see something like strategic behavior emerging is in the housing market. By buying into an expensive school district parents are able to credibly amortize the value of education by linking it to the physical house. Though no one can repossess your kids education they can take back your educational golden ticket: a house in the right district.
Thus we would expect to see skyrocketing prices as people compete for this valuable resource through side channels. Indeed, it does look like that happens.

9 comments
Comments feed for this article
Tuesday ~ December 21st, 2010 at 1:32 pm
Ryan P
I’m not sure I get what you’re saying with (5). What is the mechanism through which parents can borrow against their kids’ future incomes in purchasing their house? That looks to me like they have a current revenue stream, they use it to buy a house, and the house includes the value of the schools (and specifically, the value to the parents, which presumably is primarily a sympathetic motivation wherein the parent values the child’s utility). I don’t see any commitment mechanism here over and above that.
Tuesday ~ December 21st, 2010 at 1:35 pm
sardonic_sob
I for one will never argue that the value of a good teacher to a child’s life is in many ways impossible to value fairly.
However, when I hear this kind of thing, I start thinking about the “Functionalist Revolution” in “The Roads Must Roll.” Or I just wonder, “Well, without plumbers we’d have cholera epidemics regularly. How much is it worth to society NOT to have cholera epidemics? Figure that out, divide by number of plumbers – is that relevant to the fair value of a plumber, yes or no? If the ambulance driver doesn’t get you to the trauma surgeon fast enough, the trauma surgeon’s services have zero value. Should the ambulance driver’s salary be in some way proportional to the trauma surgeon’s?” Etc, etc.
Tuesday ~ December 21st, 2010 at 1:39 pm
Ryan P
Sardonic,
Is that what Hanushek et al are measuring — the value of teachers as a whole, divided by the number of teachers? If so, then you’re right, that’s not a helpful number. But I thought they were trying to measure the marginal value of another good teacher, or of replacing one bad teacher with one average teacher. And yeah, I think marginal value is both a good target for wages and a reasonable measure of “fair value”.
Tuesday ~ December 21st, 2010 at 3:45 pm
Roland
I like the way you look for various measures of teacher value from other observed behavior. I have recently noticed a related fact. There is a view held by some that class size isn’t as important as other determinants of the effectiveness of teachers. After investigating several expensive private schools in the D.C. area let me just note that class size is very prominently identified as one of the benefits of attending these schools. Are we to suppose that these schools have bad information about what is valued by parents spending tens of thousands of dollars? Could these parents be mistaken in what they value? Or does class size really matter?
Tuesday ~ December 21st, 2010 at 4:18 pm
Karl Smith
The easiest would be that the parents are mistaken about what really does matter.
They may observe that smaller class sizes are correlated with better performance and think that this means smaller class sizes cause better performance.
However, it could be the case that better teachers command higher per pupil salaries and can be supported on smaller class sizes. Since, teaching a smaller class is preferable these teachers gravitate towards schools with smaller class sizes.
Thus the correlation builds but smaller class sizes are really just a way of saying higher paid teachers.
Tuesday ~ December 21st, 2010 at 4:38 pm
Outlying Observations
Better teachers prefer to be paid through smaller classes than through higher pay (public sector is usually paid more, so the classes have to be *really* small)? Maybe..
Tuesday ~ December 21st, 2010 at 5:03 pm
bullfighter
“There are obviously some peer factors at work here but at its heart a school is primarily a collection of teachers and classrooms.”
That is completely wrong. I guess you are not a parent, but have you also forgotten being a student? Student population (i.e., the socio-economic and educational status of the households from which students are drawn) accounts for most of the differences in quality between schools. And, to the extent that teachers prefer to teach good students, the placement of good teachers will also be driven by the student population. The total (direct and indirect) effect of student population may then account for almost all the difference in quality.
Tuesday ~ December 21st, 2010 at 5:37 pm
Matt
I disagree about the point on “professorial teaching”. As you say, professors typically spend a relatively small amount of time teaching, and are paid more than K-12 teachers. My impression is that that the teaching load at the university level varies inversely with the prestige of the school and hence with professor pay.
If professors are paid *inversely* with the amount of teaching they do, i.e. based on how much non-teaching work they do, this seems to imply the opposite of what you say on this point: that the market prices professorial teaching quite low, at least relative to professorial research, etc.
Wednesday ~ December 29th, 2010 at 6:53 pm
Wednesday December 29th, 2010 « Matt Alldian
[...] Behavior discussed a heavily linked study citing the value of good teachers. The number is 400k. This honestly [...]