There is no evidence for the education unicorn

It is astonishing just how much agreement can be reached between an Age of Enlightenment Liberal and their nominally ideological opposite, Marxists, can agree upon when we also agree to deal with facts rather than emotional opinions.  From Pre-K Research is Mixed, Running to Discouraging, at Best by Freddie deBoer.

His subheading is "here's how bad analysis becomes conventional wisdom."  

At the NYT’s Upshot vertical, Claire Cain Miller says of universal pre-K

The bulk of the research shows that high-quality preschool tends to benefit children into adulthood, especially children from low-income families.

I find this a little brazen. Miller summarizes a corpus of research that even pre-K proponents admit is mixed but finds “the bulk” of the research encouraging. I know it can be annoying to report on controversies rather than to proffer conclusions, but this is a perfect example of when that’s warranted. I read an awful lot of ed research and am pretty plugged in to these arguments, and I’m accustomed to the inevitable “the research is mixed, but…” framing from pre-K advocates. Here, though, you just have a bald statement about the findings of the extant studies, one many in this fight would argue/concede is too rosy. Don’t take my word for it!

A little brazen?  I would have characterized it as an outright misrepresentation.  deBoer goes on to marshal the highlights of the consistent research over the past fifty years.  The bigger the study and the more rigorous its methodology, the less likely it is to find sustained beneficial impact from pre-K interventions. 

This is widely replicated and widely known.  Miller is simply being misleading at best.  

deBoer reaches the same conclusion I have after his extensive, but still tip-of-the-iceberg review.

Here is where he, and I, and the much replicated research all agree.  Genes (both in terms of IQ and behavioral attributes) dominate outcomes and become more influential as children age, hitting a sustained peak of approximately 0.80 at age 20 for IQ and 0.3-0.6 for specific behaviors.

Shared environments (home, school) account for about 10%.  Non-shared environments maybe 20-30%.  

We cannot expect schools, above a very low basis of safety, to make much difference in life and educational outcomes.  We also cannot expect family environment to make much difference, except in instances of extreme neglect, to make much difference.  

Genes drive the bulk of outcomes and we can have marginal influence at the margin through family and schools.  But it is pretty limited.  The best approach would be to acknowledge the value of accepted behaviors and more formally inculcate those through schools.  Behaviors are the attributes least determined by genes (though, depending on the particular behavior, they are still quite important.)

Provide a good safe school, focus on good behavior, and that's about it.  Teacher quality (above a bare minimum)?  Doesn't appear to matter.  Curriculum?  Doesn't appear to matter.  Computers?  Don't appear to matter.  Pre-K?  Doesn't appear to matter.

We appear to spend an inordinate amount of time on the least consequential interventions.  It is a terrible mistake.

deBoer's recommendation, which I endorse, is provide that what is truly beneficial - childcare for low income families.  Really, all families, but especially low income families.

The most noble purpose of funding pre-K is keeping kids alive and giving parents the opportunity to work during the day without spending all their earnings on childcare. As with afterschool programs, the trouble with predicating childcare on the idea that it raises test scores is that usually test scores don’t rise, and where are you then? Instead, we should advocate for childcare as such, establishing programs that put it in reach of more working parents, and defend it on those terms. 

A lot of ox would be gored in this approach but it is much, much cheaper and far more beneficial to low income families.   

The only thing stopping us doing so is a Mandarin Class, such as Claire Cain Miller, pretending that there is evidence to support a rosier ideal world that they prefer.  There is no such evidence but they will not accept no as the answer.  So they continue trying to improve that which will make no difference, wasting time, money, and making things worse for those most at risk.