Profile of John List

Although List is one of my favorite current economists, somehow I missed this one when it appeared a few days ago.  Here is an excerpt:

With $10 million from hedge-fund billionaire Kenneth Griffin, List will track the results of more than 600 students– including 150 at this school. His goal is to find out whether investing in teachers or, alternatively, in parents, leads to more gains in kids’ educational performance, Bloomberg Markets magazine reports in its April issue.

And:

He says he turned down an offer last year to become chief economist at Amazon.com Inc., the Seattle-based online retailer, partly because the company wouldn’t have let him publish the results of research.

There is much more at the link.  For the pointer I thank Michelle Dawson.

Comments

My money is on Parents. Or at least I should say that I bet the variance of parenting is much higher than schools and that marginal improvements in parenting skills will far outweigh the benefits to schools. However, where does early childhood intervention like daycare fit. Although, after a certain age I bet neither really matter all that much but that schools may overtake parenting in importance right before college.

Bill, parents don't have to tutor their children to be effective. I stopped asking my parents for homework help after the 2nd grade. It's a matter of "encouraging" your children to do well in school, in my opinion.

q, That's what I said, so thanks for the agreement.

The belief that teaching parents parenting techniques will improve student achievement is not altogether different from the belief that rigorous driver training will meaningfully reduce the rate of fatal auto accidents. The problem with both beliefs is that the undesirable outcome that we are seeking to avoid is only rarely caused by the sorts of things that education can address. People do not drive drunk, send text messages, and race on public roads, because they do not know that these things are dangerous. They do it in spite of knowing that these things are dangerous. Similarly, people are not "bad" parents because they failed to read the right parenting books. They are bad parents because they do things that they know are detrimental to their children's development.

I'd like to see a critique here, perhaps some good old libertarian perspective, that its wise to give loads of money to parents to educate their kids. How about we train each American to grow all of his own food? There isn't anyone questioning this cradle to grave paternalism at, of all places, Chicago? Its a pretty rough "Nudge." Is this a 'Lets create another Steve Levitt contest'? To top it off, it's being funded by this robber baron Griffin so he can use List for philanthropy porn.

Bottom line: kids at these relatively poorer schools aren't doing worse than their own socioeconomic groups were in the past (or even than some above them). It is simply that the higher groups keep pushing farther and farther out ahead. This blog should recognize this isn't a bad thing for anyone.

Experimenting on education: Good. This experiment: waste of $10 million. There is a complete disconnect here. List understands that experimenting is good, but THE QUALITY OF THE HYPOTHESIS MATTERS. This experiment is trying something random at the expense of $10 million. The idea is stupid. Why didn't List seek some informed opinions first?

Surely Ken Griffin wants to be a taker on the other side of this bet. No?

@dirk: Huh? The three groups (teacher support, parent support, and control) are randomly selected from those who elect to participate in the experiment.

There is something contradictory about applauding the scientific quality of this experiment while also suggesting that the scientific quality of the experiment will be apparent when the results start coming, because:

1) A well designed experiment is likely to result in a null hypothesis

2) A badly designed experiment is much less likely to result in a null hypothesis than a well designed one

List is correct in suggesting that there should be hundreds of experiments in education, but if the first one does not result in a null hypothesis, odds are it was a biased experiment.

I question anything John List says.

Thought it was another article from The Union.

"[...] and receive cash or scholarships valued at up to $7,000 annually as a reward. "

$7000/year seems like a substantial amount of money when given to a low-income family. Will this not essentially raise their socio-economic status? If I understand correctly, educational achievement correlates strongly with socio-economic status. Is the question here "If we give low-income families more money, will their kids do better in school?"

Comments for this post are closed