Efficiency: Positive or Normative?

Rad Geek asks an important and interesting question in a comment thread below: is efficiency a positive or normative term?

Let's first define our terms. Efficiency, according to economist Paul Heyne, is

a relationship between ends and means. When we call a situation inefficient, we are claiming that we could achieve the desired ends with less means, or that the means employed could produce more of the ends desired.

Therefore, when we call a situation efficient, we are claiming that it is not posssible to achieve the desired ends with less means, or that the means employed cannot produce more of the ends desired. In other words, an efficient outcome is the best way to reach our goal with the limited resources we have available.

For the definition of positive and normative, let's turn to a slightly less reliable source that nevertheless should give us an uncontroversial answer:

In philosophy, normative is usually contrasted with descriptive or explanatory when describing types of theories, beliefs, or statements. Descriptive (or constative) statements are falsifiable statements that attempt to describe reality. Normative statements, on the other hand, affirm how things should or ought to be, how to value them, which things are good or bad, which actions are right or wrong.

So is economic efficiency a positive or normative term? It certainly looks like it could be normative, as it is urging us in favor of doing something the best way, and best appears to be a value judgement. Doing X is inefficient - it is not the best way. Therefore, do Y instead - it is the best way.

But one thing is missing here. Efficiency is the best way of doing what? Efficiency tells us nothing about what goals we should have. All efficiency tells us is that if our goal happens to be X, then Y is the best way to go about achieving X, whatever X happens to be.

In other words, efficiency is a hypothetical imperative:

A hypothetical imperative, originally introduced in the philosophical writings of Immanuel Kant, is a command that applies only conditionally: if A, then B, where A is a condition or goal, and B is an action. For example, if you wish to remain healthy, then you should not eat spoiled food. Thus, a hypothetical imperative is not justified in itself, but as a means to an end; whether it is in force as a command depends on whether the end it helps attain is desired (or required). The opposite of a hypothetical imperative is a categorical imperative, which is unconditional and an end in itself.

So is an hypothetical imperative positive or normative? It looks like it's positive to me: it is a falsifiable description of the way the world works. It does not tell us how things ought to be (our goals), it doesn't tell us what is good or bad or what is right or wrong, except in regards to our previously chosen ends. Hypothetical imperatives - efficiency - are not value judgements, but a recipes for action, causal descriptions of the world in which we live.

Share this

Now if we return to the

Now if we return to the original post by Scheule that started the discussion, I think we find his use of the efficiency concept inexact, or at least underspecified. Here is the post in its entirety:

Perhaps Posner has spoken to it (query what Posner hasn’t spoken to) but it occurred to me that laws against suicide may be economically efficient, in that an individual’s suicide often imposes massive negative externalities on friends and family.

So, relying on the Heyne formulation, we must immediately ask whose means and whose ends?

Because laws are social constructs, it appears that Scheule is thinking about "society's" ends, but given the diversity of views about what society should be aiming for I think this is a mostly empty concept.

I think I agree with Ghertner's position, but I'm not sure I followed the philosophical points clearly. Efficiency, as a term of art in economics, is a positive term about the technical relationship between given means and ends. Any normative force that arises from an efficiency analysis comes from the desirability of the ends pursued.

Efficiency Micha Ghertner

Efficiency
Micha Ghertner discusses the nature of efficiency. I'd go a step further and claim that the contemporary use of "efficiency" in economics is almost exclusively applied to social welfare. (total social input ⇒ total social output) Social welfar...

Efficiency Micha Ghertner

Efficiency
Micha Ghertner discusses the nature of efficiency. I'd go a step further and claim that the contemporary use of "efficiency" in economics is almost exclusively applied to social welfare. (total social input ⇒ total social output) Social welfar...

1. Is the Chinese government

1. Is the Chinese government efficiently using market efficiency to efficiently maintain centralized control of the population?

2. I pay for efficiency in "factory" products, but only if the efficiency of the manufacturing process does not reduce the efficiency of the end use of the product.

3. I do not pay the gentleman who hand carved my cabinets for efficiency. I do however pay him for his inefficient proficiency.

This is one of the bones I

This is one of the bones I have to pick with the Coasean Social Cost Theory. Even if we put aside for the moment the argument that Austrians deny the very possibility of such quantitative, intersubjective comparisons regarding costs, the C.S.C.T. still takes for granted that society not only is allowed to, but ought to redistribute wealth in the name of efficiency.

How the argument from efficiency becomes a normative discipline is beyond me.

How could anyone reasonably

How could anyone reasonably argue that wealth redistribution is efficient!?
It easily encourages; idleness, laziness, growing inefficiency in the supported sectors, deincentivisation (did I get that one right?) or outright capital flight in the sectors from which it is taken. In my province of Canada, primary utilities, ferries, mass transit, medical facitlities, auto insurance, are all state supported monopolies, meaning that they are supported by redistributed wealth. In all cases endless cost overuns and budgetary problems are the case.

How could anyone reasonably

How could anyone reasonably argue that wealth redistribution is efficient!?

By just looking at a single moment in time and ignoring any effects redistribution may have on future behavior. It's based on the principle of diminishing marginal returns. If I have a billion dollars and you're flat broke, taking $100,000 from me and giving it to you will increase our aggregate utility, because $100,000 is insignificant to me but does a lot of good for you. So it's "efficient" in the sense that it increases aggregate utility for a single point in time. But, as you point out, the long-term results are anything but efficient.

Micha -- you're exactly

Micha -- you're exactly right, and that's just how I explain it to my students when I discuss the positive/normative distinction. Well, I don't give them the Kantian stuff, but I tell them that instrumental if-then statements like "if you want to live, then you should take this medicine" are positive statements. A lot of students have difficulty getting past the word "should," though, because some professor in their past told them that word automatically signifies a normative statement.

It is worth noting that, for many economists, efficiency becomes normative because they *do* embrace the value judgment in the "if" portion of the statement. That is, they take it as given that achieving some form of efficiency (Pareto or Kaldor-Hicks or whatever) is indeed desirable.

Neither Pareto nor

Neither Pareto nor Kaldor-Hicks strike me as very good measures of general well-being (yes, I know it's a vague concept), largely because they ignore diminishing marginal utility of wealth and thus tend to magnify the importance of desires of the wealthy. I can see a certain logic behind this---since the wealthy tend to be more productive, maybe we should prioritize their desires. But I'm still curious: Is there a measure of efficiency that takes this into account?