The costs and benefits of the word 'externality': what I learned at PERC

by Andrew Balthrop, a PhD student in economics at Georgia State University and 2011 PERC Graduate Fellow.

Until this summer, the way I thought about externalities was pretty abstract: an externality was where one agent affected another agent’s utility without compensation. The solution was to get a social planner to tell each agent what to do, and society would be better off for it. I had never thought too deeply about why the two agents could not settle the problem themselves, or about the institutions that might help address the problem. It was merely an optimization problem with two possible solutions.

For most of the summer, I didn’t get why everyone at PERC hated the term. Often times a point was made that went something like, “If this is really a problem, why haven’t people dealt with it?”  I took that point as an assumption that the economy must always and everywhere be in equilibrium; I thought the senior researchers were ignoring the most amazing thing about the economy: its dynamism. People do get rich with new discoveries, the landscape is ever shifting, and if you don’t adapt, you get left behind.  But I will come back to this point.

What changed my view–what put me in my place as a graduate student–was listening to the senior fellows at PERC talk about Ronald Coase. I must sheepishly admit that what I had taken from Coase was only the first three pages of “The Problem of Social Cost.” I was exactly what the fellows railed against. My take away was that externalities were reciprocal, and that if “transactions costs” were low, and property rights clear, agents would achieve the efficient outcome. But I assumed the point of the paper was that  transactions costs were not low, and that property rights were not clear, and that is why you have to study Baumol and Oates. I didn’t understand that the court cases were a way to make property rights clear and transactions costs low–that the courts were a solution. I left that seminar with my mind blown, and a little embarrassed.

How I should have been interpreting the comment of “If this is a real problem, why haven’t people dealt with it?” is actually, “You are talking in abstractions here. The real world is not a blackboard problem, and people are not hapless. If this is a problem, there must be specific institutional friction(s) preventing agents from solving it. Tell me specifically what that is, or you have not understood the situation.” This is why PERC researchers don’t say externality. Or spillover.

My own research focuses on the extent to which neighboring oil producers interfere with each other's production. While at PERC I learned that merely quantifying this production interference was not sufficient to demonstrate a true externality. In order to do good research on the subject of oil and natural gas production, I would have to know the institutional environment inside and out. Inter-well communication isn’t enough to assure an externality; it is only a physics problem and people solve those all the time. For there to be an externality, something else has to be going wrong. For me to use the term, I need to identify specifically what that something else is, whether it is a peculiarity to oil production or oil law that makes contracting difficult, a regulation that incentivizes perverse behavior, or something else. And once I identify this problem specifically, there isn't much reason to use the term "externality."