Planners are quick to criticize roads and highway investments for the vast sums spent to build, operate and maintain them, often questioning the value of these subsidies. Recently, on a planning list-serve, these subsidies were labeled an "external cost" of automobiles, but they are not. I thought it might be a useful time to make this point more broadly because a crucial justification for public intervention in a market economy is to correct for market failures, including times when external costs exist.
Furthermore, fellow Interchange blogger Michael Lewyn provided his own contribution on the concept of externalities ("Externalities Meet Externalities"), and, with all due respect, I think he has confused a few important distinctions.
Defining what is meant by an "externality" or "external cost" is important because it creates a set of parameters for policy intervention. Following the conventional definitions used in economics, political science, public administration, and public policy analysis, an externality exists when a third party bears the unintended cost (or benefit) of a transaction or action even though they weren't part of the exchange. (My standard reference is the survey Dennis Mueller's Public Choice III, see page 25.) Because costs are born involuntarily, policy intervention is justified to mitigate or correct the harm. While external benefits are probably more common, they don't get much attention because they generally don't pose serious problems that must be addressed through public policy.
This brings me to the main focus of this post: Are the taxes levied to fund roads an external cost? No. For the most part, they are not costs imposed on unwilling third parties. Rather, roads are usually classified as public goods, which are products or services that are socially valuable but cannot (or will not) be provided privately. The subsidies used to provide roads are the explicit costs of providing these public goods for society's benefit in the same way that general taxes are used to support the court system. In fact, most communities vote on property and sales tax increases to fund roads, and they approve these quite often. Thus, subsidies to build, maintain, and operate roads are not properly called an external cost, or even an external effect, let alone an unintended effect, of automobile use. Since the vast majority of travel is on roads, any third party external costs as they relate to financing these transportation systems are likely very small.
A side bar is whether public funds on roads are costs "externalized" by users to the larger public through taxes and other subsidies. I'll address that in a Part 2 to this discussion that will appear next week.
Let's take the discussion of externalities a step further. In the same list-serve discussion, the BP Gulf Oil spill was called an external cost of our reliance on the automobile. Once again, the key question is: Why pays the cost? While the Oil Spill is a terrible environmental tragedy, that fact alone doesn't qualify it as an externality, or an external cost to using oil as an energy source. Nothing was inevitable about the oil spill, nor was it a necessary effect of the technology used to drill for the oil. Indeed, in principle, PB had "fail safe" technologies in place to prevent it. The spill may have been the result of negligence or rational risk analysis, but its potential was foreseeable, and to some extent anticipated, by the public and private sectors.
More to the point, to the extent BP pays for the full costs of the environmental and economic impact of the spill, the costs will have been internalized and by definition will not be an external cost. Only the portion of the cost born by third parties (presumably non-oil consumers) would be considered an external cost. The extent of this effect has yet to be determined, so it's not a foregone conclusion that the effects will be an external cost.
This is fundamentally different from air pollution from automobiles where the environmental costs were not captured or accounted for in automobile production or use. Public regulation was used to mitigate these effects, and the costs were only indirectly internalized by users and beneficiaries.
Similarly, policy choices are not properly classified as external costs either. Military expenditures to protect Middle East oil are policy choices, not externalities as defined by most social scientists.
Of course, other justifications for public regulation exist. Many of them are rooted in the traditional of planning. But planners should be aware that terms often carry very specific meanings and can be grounded in rich intellectual histories. When used outside those contexts, confusion can sully public discussion irrespective of the merits of the case being made.