A weird but interesting natural experiment

The European conquerors of what is today the United States have elected to make reparations to the original inhabitants in a very weird way: specifically, to allow them to open gambling casinos. Whatever you may think of this policy in general, for those tribes lucky enough to get a casino, it has undoubtedly brought economic benefits, or at least free money.

I think you'll only be able to read the abstract of this piece by Jones-Smith, et al but that's okay. They used data on Native American children in California from 2001 through 2012, and they compared kids based on the number of slot machines per capita in their school district. (Really!) Slots per capita correlates with the income of the Native American people.

It turns out that getting a new casino on your territory makes your kids less likely to be overweight or obese. This is pretty strong evidence that poverty causes childhood obesity. You can't do a randomized controlled trial of giving money to half the families and not the others, but this is almost as good -- it's what we call and "instrumental variable," a natural event of some kind that produces quasi-randomization of a population. Now, it's barely conceivable that some other factor is correlated with casino revenues that really explains the observation, but it's hard to think of what that might be.

Okay, why would having adequate income make kids more physically fit? It's easy to think of reasons: better nutritional quality, more recreational opportunities, less stressed out families. Maybe you can think of others.

Anyway, this is just one more brick in the ever growing wall of evidence that poverty can hurt kids for life. Once you are an obese child, you are very likely to end up as an obese adult, and that limits your economic potential as well as your life. I think we owe the Indians something better than the erratically awarded opportunity to run a casino, but at least we have learned something from this. Are you listening Paul Ryan? No, I didn't think so.