Free driver externalities

Martin Weitzman has a new paper out that introduces the concept of a ‘free driver’ externality in the context of climate change responses:

Climate change is a global “free rider” problem because significant abatement of greenhouse gases is an expensive public good requiring international cooperation to apportion compliance among states. But it is also a global “free driver” problem because geoengineering the stratosphere with reflective particles to block incoming solar radiation is so cheap that it could essentially be undertaken unilaterally by one state perceiving itself to be in peril.

It’s a really interesting idea but can it really be described as an externality? The distinction is important because the way you frame the problem defines the solution.

He is saying that the actions of one state to combat global warming could affect other countries, imposing an externality upon them. The paper goes on to say that a governance mechanism to resolve conflicts is required and propose a particular solution. That’s all fine, but why does Weitzman refer to it as an externality? Technically, perhaps it is since the actions of one party affect the welfare of another. But the common usage of the term isn’t that general: an externality is usually an unpriced effect upon another party.

Is this free-driver effect unpriced? No. The cost to other countries is reflected in their efforts to cajole the free-driver out of geoengineering the atmosphere. There are many ways that they could threaten the potential free-driver to impose large costs upon it. The fact that Weitzman spends much of the paper developing a governance mechanism demonstrates that. If the effect were unpriced why would the potential free-drivers ever participate?