Economist Writing Every Day

If you want an economist to support a government intervention, then there are two major sets of logic that they generally find attractive.

The first concerns rate of return and attracts narrower support. If the government can invest in a project in a way that the private sector couldn’t/wouldn’t and the payoff is bigger than the investment by enough, then the project should be built. 

The second set of logic is more accepted more broadly. If there is an externality, and the administration costs are small relative to the change in the externality, then the project should be pursued in order to increase total welfare.

I’m going to criticize and refine the second argument.  I was inspired by a student who wrote about education creating positive externalities for “all”. They kept using the word “all”. And I notated each time “not *all*”. While we might refer to something called ‘social’ cost and value, the existence of externalities does not imply that everyone is affected by the them identically. That’s a representative agent fallacy. The externalized costs and benefits are often irregularly distributed among 3^rd parties. This is important because government intervention can impose its own externalities depending on how the administrative costs funded.

I’ll elaborate with two examples that illustrate when an irregular distribution of externalities is a problem and when it isn’t a problem.

Electric Plant Pollution

The first example illustrates how resolving an irregular distribution of externalities can be resolved without issue. Consider a coal-powered electric plant that serves a metropolitan area and creates pollution. That pollution drifts east and passively harms residents in the form of asthma exacerbation and long-term ill health. The residents to the west are unaffected by the pollution, thanks to favorable weather patterns. Obviously, one would rather live on the west side, all else constant (importantly, all else it not always constant and there is a case to be made that there is no externality here).

To resolve the externality, the government imposes a tax per particle on the power plant at a low administrative cost. That’s nice and efficient – we won’t waste our time with means-oriented regulations. In turn, the cost of electricity increases for all metropolitan residents, both those in the east and in the west. Why is this appropriate? Prior to the intervention, the electricity users in the west were enjoying electricity at a low price, failing to pay for the harm done by their consumption. For that matter, the residents to the east are also paying the higher rates, but now they enjoy better health.

In the end, the externality is resolved by imposing a cost on all consumers of the good – which happens to be everyone. This circumstance is not pareto efficient, but it is Kaldor-Hicks efficient. Everyone now considers the costs that they were previously able to impose on others and ignore.

That’s the best case scenario.

Traffic Improvements

The second example illustrates how resolving an irregular distribution of externalities can be resolved in a way that imposes additional externalities. Consider a metropolitan area that has a highway. Some people use the highway, but not everyone. The highway is congested to a stand-still during peak hours on a daily basis. Every person who is trying to get to work or to an appointment is crowding the cars around them and slowing down everyone on the highway – that’s the externalized cost. One way to resolve the externality is to widen the highway, adding more lanes so that traffic flows more smoothly (this resolution encourages more highway use, but let’s put that to the side). In the short-run, let’s assume that the widening ameliorates traffic congestion.

How the widening is funded matters. If the money comes from a state general fund, then some people are paying for the widening who will never benefit from it. Even if the revenue comes from a fuel tax, there are still some people who don’t live near enough to the highway to use it and benefit. Indeed, because the tax is imposed imprecisely, a brand new fiscal externality is created to pay for the widening. The users of local and back roads were not imposing congestion on the highway users, yet they pay for the congestion relief.  In other words, the decision by highway users to impose congestion solicits a policy response that is borne by a separate set of people disproportionately. The intervention itself might improve total welfare, but some people bear additional costs that are not associated with their own behavior. That’s the definition of an externality!

Those local road users will now drive a less-than-optimal amount since they bear more cost for their actions – unrelated to the costs social cost that their behavior generates. That’s a transfer from one set of people to another in a manner that is not intrinsically necessary to improve welfare (there are other distortionary issues that I won’t delve into).

This second example illustrates that the resolution of an externality, while welfare-improving, can impose additional externalities in a way that is not necessary to internalize the externality. Imposing costs on people unaffiliated with the congestion is related to the nature of the funding source that policymakers choose. This is part of why economists prefer user-fees when possible. Those people who impose the externality of, say, congestion can pay for its remediation, such as with a toll. Then the user pays for the cost of their action and adjusts their consumption in correspondence with it.

Going back to the original example: no, education for some does not benefit “all”. Lower crime is not uniformly distributed across the population. It is enjoyed locally where the otherwise less-educated people are. The same is true for increasing shared civic values. The benefits are irregularly distributed. Those externalities might be worth mitigating in a subjective sense or even as judged by aggregate welfare. But let’s not pretend that those communities with higher incomes and educations or those who live in remote communities are benefiting in a way that is commensurate with their share of the tax burden.