The No harm principle, an ethical principle for economic policy advisors?

In a recent article in the New York Times, Harvard economics professor Gregory Mankiw points out that economic policy advice always relies on political-philosophical standpoints and, inspired by medical ethics, suggests that economists that give policy advice should apply the No harm principle rather than promote policy based on uncertain predictions and political-philosophical convictions. By applying his interpretation of this principle, he claims that economists should not endorse either the Affordable care act, or higher minimum wage because these are in fact policies that cause harm.

It is refreshing with an economist who recognises that there is no such thing as purely scientific, value-free economic policy advice, and it is interesting to consider whether ethical principles can be introduced to deal with biases inherent to policy advice and with uncertainties innate to economic predictions. However, Mankiw’s proposal is as biased as the policy advice he addresses, and his proposed version of the No harm principle is at best a poor re-articulation of his own ideological convictions.

According to Mankiw, the No harm principle implies the following when transferred to economic policy:

“[W]hen people have voluntarily agreed upon an economic arrangement to their mutual benefit, that arrangement should be respected. (The main exception is when there are adverse effects on third parties — what economists call “negative externalities.”). As a result, when a policy is complex, hard to evaluate and disruptive of private transactions, there is good reason to be skeptical of it.”

Mankiw’s argument takes some detours discussing utilitarianism and moral-philosophical thought-experiments and it is not crystal clear how he reaches his conclusion. The endorsement and interpretation of the No harm principle seems to primarily be based on three ideas: 1) Economists have very limited understanding of the consequences of different policies. 2) Because of these uncertainties and since interpersonal comparisons are difficult it is best to reject all interventions where someone is harmed. 3) Harm is, in the sense relevant to economists, equal to a disruption of private transactions.

I trust that Mankiw is right when he says that economists only have very basic understandings of how policies work, that unintended consequences is the norm, and that economics do resemble 19th century medicine where you would be wise to seek out a physician if you were ill, but where his remedies could well make things worse.

The other ideas are, however, highly problematic. It is true that there are significant, well-known complications surrounding both decision making in uncertain circumstances and interpersonal comparisons. Yet, to take the step from this to rejecting all policies that make things worse for some individuals is very extreme and not necessarily prudent. There are numerous situations with large uncertainties where the risks are higher if we don’t act than if we do act. Consider, for example, building codes in densely populated areas with high seismic activity, such as Istanbul, Los Angeles and Tokyo. We have, I assume, very limited understandings of the full effects of imposing regulations, and regulation is sure to be disruptive of private transactions, so some individuals are sure to be harmed in Mankiw’s sense. However, omitting to regulate this risks making things significantly worse for significantly more people. A prudential principle cannot, like Mankiw, rely on a strict distinction between acts and omissions.

To suggest that we reject all policies with partly uncertain outcomes that make things worse for some individual is everything but value-neutral and prudent, and rather an endorsement of very extreme political-philosophical standpoints: that the current status quo, whatever it is, is justified, and that regardless of the purpose, it is never justified to change any individual’s circumstances for the worse. These positions entail that new building regulations should be rejected, but also, for example, that all types of social security that rely on taxation and redistribution of resources are unjustified and wrong.

It is also quite creative to equate harm with disruptions of private transactions. Usually, a person is considered harmed if she fares ill, if she develops physical or psychological conditions, or if she is exposed to injustices, or abused. Or, indeed, if she dies in an earthquake. But do we harm potential buyers of heroin by disrupting heroin transactions? Do we harm citizens if we force them to live in safer buildings? Do we harm drivers when we enforce seatbelt legislation? We certainly limit their freedoms and sometimes increase the price of certain goods for them, but is it reasonable to speak of harm? Mankiw’s principle, it appears to me, is not so much a No harm principle as it is the laissez-faire principle with a different name. Intervention in human affairs is equated with harm, so that non-intervention can be promoted.

So how should one deal with the fact that economists fail to predict the consequences of different policies? Perhaps it would be desirable if economic policy advisors embraced ethical principles in order to deal with this, and perhaps they should even embrace the No harm principle, or a similar precaution principle. However, such a principle would need to be developed for the purpose. Mankiw’s sloppy application of the No harm principle is not convincing.

Economics need philosophy, and in particular, perhaps, ethics, but it doesn’t suffice to single out an ethical principle from a radically different field and apply this to ones own, and it does little to forward the debate if one in addition chooses to re-interpret the principle to such an extent so that it does nothing but express a well-known political ideology that currently divides our societies.

  • Facebook
  • Twitter
  • Reddit

2 Responses to The No harm principle, an ethical principle for economic policy advisors?

  • Dave Frame says:

    Agree that Mankiw’s attempt to apply a “no harm” principle is unconvincing.

    I’m sure it’s true that economics needs ethics, but I think ethics could do with being more comfortable with learning from economics, too – WAY too often ethicists ignore dynamic effects and incentives. Economics has uniquely valuable insights into both. For instance, ethicists often make demands to the effect that there is too much inequality in society (or across societies). And they conclude that we should soak the rich to address this (or encourage people to throw money at it), without considering how this sort of approach might play out, compared with alternatives. Turns out, soaking the rich is a bad idea in the long-run. Also turns out, simply throwing private money at development issues is insufficient in the absence of a reasonable policy environment. Economics helps explain why, and ethicists who ignore these explanations run the risk of arguing for bad policies, or at least policies that have bad consequences in the long-run (which is longer than the short-run).

  • Anders Herlitz says:

    Oh, I agree completely, and I think that there is a significant problem with the current state of affairs of ethics where (not even mentioning metaethics) there is on the one hand normative theory that seem predominantly occupied with right-making criteria (and thus can dismiss all empirics and work with immensely simplified thought experiments by stating that all of it of course should be taken account), and on the other hand practical/applied ethics that often rely on dominant systematic approaches in normative theory (that essentially are not construed with any practical purpose), and then work on particular cases. I am myself quite interested in whether ethics can say something systematic about how to deal with all the peculiarities so to speak of choice situations we (and policy makers) are doomed to meet: uncertainties, dynamic effects, interdependencies of distinct goods, biases, and so on. Economics (of various types) can be quite useful both for understanding these issues and for developing particular policies and making particular choices.

Authors

Subscribe Via Email

Affiliations