*Unsettled*, by Steven E. Koonin

A few of you have asked me to review this book, sometimes presented as a clinching case for climate contrarianism.  I thought it was fine, but not a great revelation, and ultimately disappointing on one very major point of contention.  On the latter angle, on p.2 Koonin writes:

The net economic impact of human-induced climate change will be minimal through at least the end of this century.

That is presented as a big deal, and yes it would be.  But “minimal”?  The economist wishes to ask “how much.”  The more concrete discussion comes on pp.178-179, which looks at twenty studies (all or most of them bad), and reports they estimate that by 2100 global gdp is three percent less due to climate change, or perhaps the damages are smaller yet.  Those estimates are then graphed, and there is a bit of numerical analysis of what that means for growth rates working backwards.  There is not much more than that on the question, and no attempt to provide an independent estimate of the economic costs of global warming, or to tell us which might be the best study or what it might be missing.  Koonin seems more interested in discrediting the hypocritical or innumerate climate change researchers than finding out the best answer to the question of cost.

To be sure, this is all a useful corrective to those who think global warming will destroy the earth or create major existential risk.  I am happy to praise the book for that and for all of its other corrections of hysteria.

But I just don’t find the Koonin discussion of economic costs to be useful.  The best estimate I know estimates global welfare costs of six percent, with some poorer countries suffering losses of up to fifteen percent, and some of the colder regions gaining.  There is high uncertainty about average effects, so you also can debate what kind of risk premium can be considered.  (I have myself written about how climate change may induce stupid policy responses, thus perhaps boosting the costs further yet.)  You may or may not agree with those numbers, but the above-linked paper provides plenty of structure for considering the problem further, such as modeling migration and adjustment effects across different parts of the world.  The Koonin brief meta-survey does not, it simply tells you that the junky papers don’t have the numbers to justify the panic.

So in what sense is the Koonin book useful for furthering my understanding of my number one question of concern?  Of course not every book has to be written for me, but at the end of the day it didn’t cause me to update my views much at all.