By Eric Crampton 13/10/2015


There are lots of potential catastrophes out there, each of which might be worth spending something to avoid. So how should we decide things in a world of multiple catastrophic scenarios?

Martin and Pindyck show that things get messy quickly. [ungated]

You can’t run a catastrophe-by-catastrophe cost-benefit assessment and just work your way down the list: as each has a big potential effect on current consumption (through taxes to provide mitigation) or on future consumption (if the catastrophe happens, or cumulative effects of the mitigation works), which catastrophes you want to avoid has to be decided simultaneously across all projects.

Here’s perhaps an easier way to think about it. Your expensive home renovation works might pass cost-benefit for you. Upgrading the car might also pass cost-benefit. But if you do the renovations, the car upgrade will no longer make sense because you won’t be able to afford to eat, and the same if you do the car upgrade before the renovations. So you need to look at them both at the same time. Worse, if part of the home renovations is to keep the garage roof from wrecking the car, then the car upgrade is pointless if you don’t do the renovations.

How things map out then depends on both your time preference rate and on your risk aversion. Here’s the results of some simulations over seven potential catastrophes, with their rough modelling of the effects.

1.bp_.blogspot.com-s6nXMiSaxncVhslW41J2MIAAAAAAABSA0qfQV3SmHP0cs400pindyck-02ab5e8ca3af657c48fbb7aa8e5e1c47b8be1b07-357c19bf8c9d5e70899eec7c0576ebb8ea2577d6

With low values for time preference and for the risk aversion parameter, it makes sense to pay to avoid a pile of potential catastrophes. As those change, so too does optimal mitigation.

I note that they left out the catastrophe that worries me: asteroid impact.