Disaster Bias

Today begins a conference on Global Catastrophic Risk I’m attending.  Coincidentally, Bryan Caplan recently pointed us to a new paper by Andrew Healy:

Using comprehensive data on natural disasters, government spending, and election returns, I show that voters reward disaster relief spending but not disaster prevention spending. This aspect of voter behavior creates a large distortion in the incentives that governments face, since the data show that prevention spending substantially reduces future damage. … Given mean annual prevention spending of $195 million and mean disaster damage of $16.5 billion, the regression estimates that a $1 increase in prevention spending resulted in a $8.30 decrease in disaster damage, and this estimate captures only benefits that occur in the five years from 2000-2004.

My and other conference presentation describe other disaster biases, such as not paying due attention to the very largest possible disasters, and not make extra preparations against human extinction. 

GD Star Rating
Tagged as:
Trackback URL:
  • Eric Falkenstein

    Having not read the piece, I wonder how they controlled for the fact that identifying disaster prevention expenditures by governments will have a hindsight bias, in that obvious pork barrel projects will be excluded in the regressions, even though the pretext of disaster prevention may have been crucial to their adoption. Consider looking back to a project passed in 1985, which involved considerable off-the-record wrangling and rhetoric. The fact that it is obviously a pork barrel project, after 22 years, implies very few would remember using ‘disaster prevention’ as a justification, because with hind sight pork barrel projects are solely driven by special interest groups.

    If we adopt a new strategy, I imagine everything is vaguely related to ‘readiness’, and so I have little confidence the marginal benefit/marginal cost ratio will 8.3. After all, where do you think all the increase in Homeland Security went? I could imagine it being very valuable if spent well, but I bet if one did the accounting, using the money as biomass for energy production might have been better. Like spending on health care prevention, it’s a nice thought, but I doubt it.

  • WTF

    “History books do not account for heroic preventive measures.” – Taleb

  • Grant

    I would imagine this bias is due to the complexity of prevention vs. relief. Governments are not very good at providing public goods that are at all complicated or difficult for the lay-person to correctly understand. Everyone can understand disaster relief, however.

    Take Katrina for example. Its a lot easier to get taxpayers to support relief of Katrina than it would have been to re-build the levies beforehand. Relief is something they can understand, but with levies they’d never know if they were getting hosed or not. In fact, tax money spent on levies might be portrayed as wasteful or corrupt by future candidates in New Orleans; after all, how is the layperson to know a disaster has been averted?

    I suppose its just another application of prediction markets.

  • Eric Falkenstein

    Katrina’s levee system missed a great opportunity when in 1977 local fishermen and an environmental group filed a lawsuit under the National Environmental Policy Act. A federal district court responded by enjoining the levee project pending the preparation of an adequate environmental impact statement, and so the United States Army Corps of Engineers abandoned its original design for the New Orleans levees and adopted instead a less ambitious design. This probably discouraged future efforts of this nature, though that is speculative. This highlights that there are many disasters competing for our attention, and they often involve direct conflict. What about saving the earth from global warming, but increasing the amount of phosphorus put into our water table via fertilizer for more ethanol?

    I sense catastrophe mongers think avoiding disaster is unambiguous, when indeed much of the spending will have to compare events that are so improbable, the relevant probabilities are within any reasonable standard error, making rational analysis of them impossible. Since catastrophes by their nature have probabilities from 1 in 100 for floods (specific to a town), to as low as 1 in 100MM if we are talking asteroids hitting the earth. There is no way to efficiently allocate resources between events because there is no way to prove your probability is correct to within a factor of 10 if not a hundred, and so we should expect institutional realities to become paramount. Competing disasters become political footballs, pretexts for things other people want (eg, jobs programs), and the probabilities will have very little to do with their prevention expenditures.

    of protecting the city from the storm surge created by Katrina. In other words, some commentators contend that, as a result of the lawsuit, the Corps redesigned the project in a way that failed to protect the city.

  • The problem would be solved if the Government wasn’t responsible for disaster prevention and instead *individual people* made choices about what disaster prevention they wanted to put their own money towards, or not.

    Of course this would lead to new problems — e.g., people would need to create knowledge about what disaster prevention they should spend their money on or not — and mistakes will be made. But under any system we’ll need knowledge. At least we’ll get away from more artificial problems like incentives of politicians.

  • I enjoyed Robin’s paper Catastrophe, Social Collapse, and Human Extinction linked from the earlier blog post. One of the points it makes is that in enormous disasters, it is likely that most deaths will be secondary to the main cause, occurring as the result of collapse of interdependent social mechanisms. This might suggest that society look at mechanisms to increase resiliency against large-scale disruption. We get many gains from economic specialization and exploitation of regional comparative advantage, but these measures put us at greater risk of extinction as well.

  • I see that Ron Bailey of the libertarian Reason magazine is blogging the conference. First two posts are TEOTWAKI!! and Will Humanity Survive the 21st Century?.

  • Perhaps the security theatre at airports *is not* in fact to the exceptions in the rest of their paper.