Costs to the Economy, and the counterfactual

By Eric Crampton 30/10/2013


When it comes to mental illness, the incremental loss “to the economy” of diminished human productivity is only relevant to policy to the degree that it could be corrected at no compensating expense. Otherwise, cost claims amount to the difference in earnings between ourselves and hypothetical, perfectly rational gods. The irony of mental health advocates waving the $50 billion figure around is that they explicitly want to increase it by pouring more money into public mental-health programs. Give them $100 billion more and they would quickly be using that in their argument for the next $100 billion.

Colby Cosh on measures of the social cost of mental illness. He’s entirely right. While productivity costs seem plausible, they’re only really relevant where there’s something that can cost-effectively be done about it. The counterfactual isn’t a magic wand that makes everybody happy and productive; rather, it’s a costly ongoing process with uncertain likelihood of success. As Cosh then puts it, “We treat the mentally ill in the name of hope, not GDP.”

The problem of the counterfactual is bigger than that though. In most of these kinds of studies, we take as counterfactual somebody with average characteristics and outcomes. But if all that matters is GDP, that’s hardly the correct counterfactual. What’s the social cost of not getting exactly the right amount of education? The counterfactual there would be people who chose the income-maximising level of education given their underlying characteristics. What’s the social cost of taking more than the amount of annual leave necessary for GDP maximization? Define things with a workaholic as counterfactual. I’d bet you could get a horrible and horribly large social cost figure by measuring the “social cost of those elderly who fail to commit suicide immediately upon retiring”. The counterfactual savings in health care costs and pension benefits are enormous.

And Cosh gets it:

The real problem is what we mean by “lost productivity.” What is being postulated here is a sad, spiritually afflicted populace whose “economy” could be bigger if absolutely everyone were well enough to work to their full capacity. We all sacrifice some part of “the economy” to personal happiness or to the well-being of our families and loved ones. Probably we all sacrifice some to personal pathologies, whether or not they could be characterized as mental illnesses. Is being somewhat lazy a mental illness? How about having an IQ of 88? “Costs to the economy” are, in this sense, hard to take seriously: human imperfections defy accounting.

Exactly. If we go down this line, why aren’t we looking at the social costs of “being somewhat lazy”?

Previously (a few of the many many prior posts here…):