Tag Archives: ozone

Ozone Pollution More Dangerous than Previously Thought?

We’ve all heard about the ozone layer, but I reckon that most people know very little about ozone. I knew pretty much nothing about it until I read an article published by the National Bureau of Economic Research which tied ozone levels to reduced worker productivity.

It turns out that ozone is a molecule consisting of three oxygen atoms. It also turns out that ozone is known to cause respiratory problems. It is chock full of free radicals, and ozone doesn’t react well with cells in your lungs. Yuck. Health organizations (EPA, WHO, etc.) set exposure standards levels that should prevent long term effects. Of course, those standards are based on the science at the time of promulgation, and science can change, as is true for all health regulations.

However, as the NBER article shows, sometimes economics can reveal patterns that medicine doesn’t. Economists from UCSD and Columbia studied ozone levels in California’s central valley (a huge farming area) and compared those levels to farmworker productivity. It turns out that increased ozone levels are correlated to decreased productivity. Moreover, this productivity impact happens at levels well below the federal safety standards. So maybe the standards are wrong, and ozone is more toxic than people think.

I admit that this study is a little on the Freakonomics side of things: it runs a regression, sees a correlation and assumes causality. I have been critical of Freakonomics in the past (although not on this blog), because I don’t think you can just regress a boatload of data and then decide that you know why result A happened. There could be all kinds of other factors at play in this data; for example, maybe ozone levels are high when the weather is really hot, and farmworker productivity was down because of the heat, not because of the ozone at all. Hopefully the economists doing the study adjusted for that sort of thing, since zeroing out the noise of exogenous variable is a standard procedure in studies like this, but the article doesn’t say. Any way you look at it, this certainly is an interesting correlation that could bear further study.

As an aside, here is a critique of Freakonomics from the American Scientist; they are more qualified than I am.