What is Efficiency?

If you’ve ever heard an economist expound on policy, you’ve surely heard the claim that this or that course of action is more ‘efficient’. We economists are prone to toss out this term as our trump card; who can argue they want *less* efficiency?

But just what does efficiency mean? First there is the conceptual division between allocative versus productive efficiency; then there is the separate issue of how to actually *measure* efficiency.

It is important to keep in mind that efficiency is a positive concept, not a normative one (see my post on economic methodology), and that is where much of the confusion lies.

Allocative Efficiency

This is usually the sense in which efficiency is used. Given a set of goods, what does it mean for the goods to be ‘efficiently’ distributed among some persons? A situation is allocatively efficient if all goods are allocated to their highest valued use.

Again, it is important to emphasize that this is not a statement of good or bad; it is merely a way of talking about the world. There are many other potential ways to allocate scarce goods: lottery, first come first serve, feats of strength, and so on. All of these can be debated normatively using different criteria, but this is beside the point of economic inquiry.

Markets are what allow allocative efficiency to be reached. When people are able to freely exchange goods, under the standard assumptions of microeconomic theory, then it follows that an allocative equilibria will be reached. (How this fares in the real world with real humans is another matter).

Productive Efficiency

Productive efficiency is not mentioned nearly as often as allocative efficiency, though it is potentially far more widely spread. When a market is productively efficient, goods are being produced at their lowest cost, and the price of a good is equal to its marginal cost of production.

In the ideal of perfect competition, all firms are using the lowest cost production technology; if they weren’t, they would have to charge more, and they wouldn’t be in business. So the idea of productive inefficiency is most relevant in non-competitive industries, such as government mandated monopolies, or industries with high barriers to entry, where something prevents those with the knowhow to produce goods at lower cost from doing so.

What standards do economists use to judge the efficiency of a situation? The two most common are Pareto and Kaldor-Hicks:

Pareto Efficiency

Simply put, a world is Pareto efficient if no person can be made better off without making someone else worse off. A Pareto improvement then is one that makes at least one person better off while making nobody worse.

What standard is used to judge better off or worse off? Usually with Pareto analysis that’s left up to the judgement of the people involved, and as such the concept is pretty useless in trying to judge actual policy making. Why? Well, *any* change that is made will upset somebody, somewhere. You know the type of person, the one who gets upset at others’ good fortune. As such, the world is already Pareto efficient; no Pareto improvements can be made.

Kaldor-Hicks Efficiency

Given the limitation of Pareto efficiency, the most commonly used efficiency concept is Kaldo-Hicks, which uses monetary equivalents to judge better/worse off. Benefit or harm from a policy change is quantified, and as long as society as a whole sees a net gain in wealth, that is considered a Kaldor-Hicks improvement. Hence, when K-H improvements can no longer be made, the world is considered to be K-H efficient. Another way of stating this is that the dollar value of social resources is maximized.

This is the standard governments use when in cost-benefit analyses, and this is the sense in which trade makes societies better off. Yes, there are some particular losers, but society as a whole is better off. In theory, the gains from trade can simply be redistributed to compensate those losers, but of course that’s easier said than done.


Narrow vs Broad Economics

In his economics text Price Theory and Applications Jack Hirshleifer makes a distinction between ‘narrow’ and ‘broad’ economics, and I think it’s a useful concept for elaborating on the ideas in the previous post on economic methodology.

I’ll quote again from Silberberg and Suen in The Structure of Economics:

What economists have in common with each other is a methodology, or paradigm, in which all problems are analyzed. In fact, what most economists would classify as noneconomic problems are precisely those problems that are incapable of being analyzed with what has come to be called the neoclassical or marginalist paradigm.

So we can broadly classify problems into economic and noneconomic problems, and further divide economic problems into narrow vs broad economics. The types of problems the marginalist paradigm can solve are of two sorts: equilibrium and optimization.

Narrow economics is what people traditionally think of as economics: acting within markets, supply and demand (equilibrium) and profit maximization (optimization), among other things. But starting in the 1960s and 70s, particularly with the work of Gary Becker, economic analysis was extended to nonmarket human activities: crime and punishment, the family, addiction, information, and politics. These fall under the purview of broad economics: any human behavior that can be analyzed with the marginalist paradigm.

One of the classic papers in this vein of analysis is George Stigler and Gary Becker’s ‘De Gustibus Non Est Disputandum’, or ‘In matters of taste, there can be no dispute’. The paper was a broadside against the limitations of ‘narrow’ economics, specifically the tendency to ascribe differences in human behavior to different preferences. Recall that the marginalist paradigm requires an assumption of stable tastes. As such, in instances where tastes appear to change, such as developing an addiction, standard economic analysis is powerless.

Stigler and Becker argue that simply throwing up your hands at the first sign of differences in tastes is too easy; indeed, with some hard work and clever models, the basic tools of prices and incomes can explain quite a lot with the seemingly radical assumption that ‘tastes neither change capriciously nor differ importantly between people’.

In the authors’ words:

On the traditional view, an explanation of economic phenomena that reaches a difference in tastes between people or times is the terminus of the argument: the problem is abandoned at this point to whoever studies and explains tastes (psychologists? anthropologists? phrenologists? sociobiologists?). On our preferred interpretation, one never reaches this impasse: the economist continues to search for differences in prices or incomes to explain any differences or changes in behavior.

And tying back into the discussion of economic methodology:

The choice between these two views of the role of tastes in economic theory must ultimately be made on the basis of their comparative analytical productivities. One the conventional view of inscrutable, often capricious tastes, one drops the discussion as soon as the behavior of tastes becomes important – and turns his energies to other problems. On our view, one searches, often long and frustratingly, for the subtle forms that prices and incomes take in explaining differences… If the latter approach yield more useful results, it is the proper choice.

What we assert is not that we are clever enough to make illuminating applications of utility-maximizing theory to all important phenomena – not even our entire generation of economists is clever enough to do that. Rather, we assert that this traditional approach of the economist offers guidance in tackling these problems – and that no other approach of remotely comparable generality and power is available. (emphasis added)

Again we see the influence of Friedman: the best theory is the theory that works best, regardless of the ‘realism’ of its assumptions. It seems obvious that human tastes differ, but if a theory that assumes they don’t can better explain human behavior, we should use that theory. It’s a deeper question as to whether we can take the usefulness of a theory as evidence of its assumptions; this is a central concern of the philosophy of economics (and science more generally). I won’t get into that now, except to say that one has to be extremely careful moving between models and the real world, and economists in the public sphere in particular often play too fast and loose.

The paper itself is quite interesting, is foundational to modern economics (over 5000 citations), and deserves a post of its own, so that’s what we’ll do next time.

Ants and the division of labor

Ants are one of the few of our fellow travelers* to engage in the division of labor, which is a key to economic development. Unlike humans, however, the ant’s division of labor is determined genetically: ants go through distinct life phases, certain ages during which they specialize in particular tasks. When you think about it, it is fairly remarkable that even with such basic elements of social structure so hard coded, ants are still remarkably adaptive and successful.

And maybe even more remarkably so, given that in many ant species, some workers seem to do nothing! I came across this article that finds that, at least for one species, some workers specialize in doing nothing. That is, when researchers observed inactive ants, they just sit around and do nothing all day; this isn’t an instance of workers resting between tasks, or that there is simply nothing to do.

Is this adaptive behavior that we just don’t understand? Is it evidence against a ruthlessly efficient interpretation of evolution? An illustration of the selfish gene hypothesis?

The answers to those questions could have bearing on how we decide to organize our own society.

* A term adapted from a quote that has stuck with me since I first encountered it, by Henry Beston: “We need another and a wiser and perhaps a more mystical concept of animals. Remote from universal nature and living by complicated artifice, man in civilization surveys the creature through the glass of his knowledge and sees thereby a feather magnified and the whole image in distortion. We patronize them for their incompleteness, for their tragic fate for having taken form so far below ourselves. And therein do we err. For the animal shall not be measured by man. In a world older and more complete than ours, they move finished and complete, gifted with the extension of the senses we have lost or never attained, living by voices we shall never hear. They are not brethren, they are not underlings: they are other nations, caught with ourselves in the net of life and time, fellow prisoners of the splendour and travail of the earth.”

Paid Family Leave as Intertemporal Transfer


Mike Konczal makes a point on Twitter that I hadn’t considered before, but seems to me to be correct:

This is in reference to a recent George Will column offering the standard conservative critique of paid family leave as merely a bill that will come due in the next generation. Mr. Konczal is right to point out that, even if there is no value add at all to paid family leave, it could be the case that future adults would prefer to be able to lend their parents some money so they could have an easier time in those first months of life. Given that time only flows one direction, such a transfer doesn’t seem to be possible without outside intervention.

An angle Mr. Will does not consider is that paid family leave improves outcomes for children, making them more productive future workers. To my knowledge, there is no slam dunk evidence as yet that this is the case, though the earliest months of life do seem to be very formative. Paid family leave could then in fact be a net benefit for society!

Of course, the opportunity cost to the employer has to be figured in as well. Funding paid family leave through an employer tax would be a huge mistake; you should always keep in mind the separation principle, that where a tax goes has nothing to do with where it comes from. There is no sense in adding an undue burden on to firms just for doing what they normally do: hire people. Instead, tax revenue should come from whatever sector a tax does the least harm.  Additionally, people should be paid directly by the government; again, there is no sense in involving firms here. It’s possible that firms should also receive some benefit as well, since they will be out an employee for several weeks.

In sum, paid family leave is something politicos of all stripes can agree on, provided it is structured correctly.

Supply Shock in Transportation Costs?

The Wall Street Journal noted in an article recently that food makers are facing higher input costs (leading to higher consumer prices), and this is driven in part by an increase in freight costs. This is the kind of thing that’s easy to overlook when thinking about the economy: I suspect most of us don’t consider the price of transporting various components of a good as part of its value add, but it is. Cheap transportation allows supply chains to have a huge geographic scope, and hence take advantage of specialization and comparative advantage.

But what about the increase in these transportation costs is newsworthy? FRED has data on the Producer Price Index for freight transportation, and there’s a clear spike in recent months:

fredgraph (2)

What could be driving this? A good first guess would be fuel prices, which are shown below (note the different axes!):

fredgraph (3)

While historically the two series are very closely linked, this recent runup seems to be unrelated to fuel prices. But obviously there has been some sort of supply shock. Later on the article gives a hint: new federal regulations aimed at increasing transportation safety have essentially reduced the number of hours drivers can work. It takes time to hire and train people, and as the economy heats up the marginal truck driver is probably finding better employ elsewhere (long haul trucking is a famously difficult job).

Looking at total truck transportation jobs, it’s interesting to note how employment has been flat since 2015 or so:

fredgraph (4)

An uptick has just begun; perhaps the beginning of hiring to offset the reduced hours per worker?

One question this story leaves us with is why these changes weren’t anticipated; if transportation firms knew that regulations were coming, they would presumably have begun hiring ahead of the implementation to avoid this spike in prices.

This article here is a year old but presumably covers the regulations in question.

Trouble in Congo?

This week’s Economist has an excellent article on the situation in the Democratic Republic of Congo, where tensions seem to be rising again and the threat of large scale violence is increasing.

I suspect most Americans are unfamiliar with the wars in Congo, or have just the vague notion that such wars existed. Probably most know of the Rwandan genocide, but don’t know that it was the kicking off point for a series of conflicts often dubbed ‘Africa’s World War’.

The best book I have read on the topic is Dancing in the Glory of Monsters, and what I really appreciated about it beyond lucid prose was that the author took pains to look at the structural factors that led to such horrors. It’s far too easy to put blame on the character of individuals (‘those people are just uncivilized/animals/etc’) and miss the much larger forces at play.

For one, the actual state capacity of the Congolese state is very limited and basically exists only in the west around the capital Kinshasa. In the east, the closest thing to a state are a series of warlords that operate more or less with the blessing of President Kabila; he couldn’t do much to stop them even if he wanted to.

In fact what makes the book great is that it is in many ways a case study of what types of social structures can lead to genocide and sustained warfare; the lessons he draws are not specific to Africa. Check it out!