The buzz about defining externalities

In intro to econ classes, most intructors spend at least a little time discussing externalities--y'know, what happens when markets fail to capture the full benefits or costs of a production or consumption decision.  I usually break the discussion into a two by two grid with consumption and production on one side and positive or negative externalities on the other.  Then I ask for examples from the class to fill in the grid--it seems students like to participate in class, or so I'm told. 

It's usually straightforward to fill in three of the four grid cells--

• Positive consumption externalities=vaccines, education
• Negative consumption externalities=second-hand smoke, public drunkenness
• Negative production externalities=industrial pollution.

The problems usually come in defining a positive production externality. A benefit to someone that is not fully captured by the producers--usually difficult because producers are usually pretty funny about finding ways to recover the full benefits of their production.  The example I usually end up using is that of orchards and honey bee producers...

The story goes like this:  Honey producers need flowers to produce honey and orcharders need bees to polinate their fruit trees.  So there are positive benefits to each to locate near each other. 

Even though I've used this example for years, I've always been skeptical because I was never quite sure that the honey producers could really get the bees back after they fly off into the orchard--and I've never really had evidence that honey producers and orcharders were in cahoots.  But, I'm skeptical no more:

A newly discovered virus might be behind a mysterious disorder that's killed billions of honeybees across the United States, scientists announced today.

Although they aren't certain that Israeli acute paralysis virus is the cause of this great disappearance called “colony collapse disorder” or how it got here, they've found the virus in more than 80 percent of the hives they tested.

“At best, we can say it's a candidate for a cause,” said W. Ian Lipkin, a Columbia University epidemiologist.

The virus is the best lead yet in a huge research effort to end the threat to honeybees, an essential link in our food supply.

In Ohio, farmers rely on bees to pollinate more than 70 crops, including apples, peaches, grapes, strawberries and pumpkins, valued at more than $14 billion a year.

Cool.

Comments

[Pedro Díaz]

Isn´t Einstein who said human being could hardly survive if bees disappeard?

Pedro

[mk]

Wait, well if you defined "production" as broadly as it probably should be defined, then one positive externality of, let's say, producing computer chips is that a bunch of local people are employed in the factory and can use that money to buy things in the local economy, which promotes growth. No?

[jim casey]

An externality has to be an unintended conseguence, as defined by Baumol and Oates. I do not think employment and multiplier effects fall under the category of unintended.

The General

[UMDAREC]

"Honey producers and orcharders in cahoots" has been pretty well recognized for 35 years. See Cheung (1973) and Muth et al. (2003). More work forthcoming.

[robertdfeinman]

I always harp on the consumption of non-renewables as something that the "externalities" model doesn't capture properly (I know it's not in the quadrant you are discussing).

It is one thing to, say, use river water to cool a power plant before putting it back into the river, but it is quite another to pump oil and burn it. One can never put it back. The traditional model that the value is based upon the cost of extraction ignores this externality, but so does the one that allegedly includes it. How does one price something that can't be replaced?

Substitution arguments only defer the issue by one step, you can't replace oil with coal if the coal itself is depleted.

Perhaps there is some branch of economics that I'm not aware of that deals with this, but so far all I see is the inability to face the issue. Resource limits are already taking their toll, the first signs are in arable land and fresh water. Where are the models and policies?

[Dan Cole]

Doesn't open-source computer coding create positive production externalities? True, those externalities are not unintentional, but it's not clear to me why that should render them non-externalities?

[ABC]

Chips, employment and growth: this is an example of what we call a "pecuniary externality" where the unintended consequences are mediated by the price system. This type of externalities does not cause any net social loss, because it simply causes resources to be redistributed to reflect society's new consumption and production decisions. No resources are left unused or inefficiently used. That's the reason why economists do not worry about this type of externalities. As a matter of fact, these take place all the time and they guarantee the virtues of the ideal market (when a long list of conditions exist): the fact that resources will be put to their best use.

We worry about the externalities that cause society's aggregate resources and well-being to fall below its potential maximum. Also bear in mind that the typical examples assume a given technology and a given amount of resources. That is, the "social pie" is constant. When the economy grows, the pie grows and externalities have to be reassessed for the new pie. These are the externalities that guarantee the environmental economist's existence and are called "technological externalities". The confusion between technological and pecuniary externalities was present in the profession when we first started to seriously look at them, but it dissipated over time, to the point that we simply say "externality" when we mean "technological externality".

[ABC]

Consumption of non-renewables: The fact that a good is non-renewable or that irreversibilities exist is different from the idea behind an exernality. In other words, a good might be non-renewable and yet no externality is present. I can't help thinking about when I was a teenager/young adult and gave up on frequent parties hoping that my hours of study would pay later (those days will never come back and incidentally I'm still waiting for the payoff...). The point is, if I fully account for the benefits and costs of my decisions, even if they involve an irreversible consequence, there will be no externalities. The problem becomes hairy when we fail to fully account for future costs and benefits (several generations ahead), and if someone can find a way to do this, I predict he/she will become very famous. We should also keep in mind that there is an intergenerational equity issue here: what is fair/unfair to future and current generations. But the concept of externality is not about equity/fairness. It is about efficiency (getting the most we can get with what we have available).

[Dan Cole]

More positive production externalities: when my neighbors plant and maintain beautiful unfenced gardens, which I get to enjoy at virtually zero cost; when a farmer imposes a conservation easement that prevents development for non-agricultural purposes (although would-be devleopers might consider the externality to be negative); owners of tree farms, which absorb carbon and produce oxygen (but when the trees are cut, they create negative externalities which offset the positive ones); the occasional inventor who does not obtain or enforce a patent or, in many cases, inventors whose inventions are so immensely valuable that the value obtained during the life of the patent comes nowhere close to capturing all of the social value created; creators of goods, such as dress designs, that are not (yet) subject to intellectual property rights, and so can be "knocked-off" by others and produced at lower cost; and the list goes on.

[joshua corning]

The example I usually end up using is that of orchards and honey bee producers...

You should change it to this one:

http://cafehayek.typepad.com/hayek/2007/09/good-luck-bls.html

[joshua corning]

The story goes like this: Honey producers need flowers to produce honey and orcharders need bees to polinate their fruit trees. So there are positive benefits to each to locate near each other.

I live in orchard country (apples, cherries, and pears) generally the orchardists either have there own hives or they hire people to put the hives on their land during pollination season. The later, I think, is more common.

and yes it is orchardist not orcharder.

If you don't believe me here is a picture of some rental hives on a couple of pallets:

http://www.mitzenmacher.net/blog/wp-content/uploads/2007/03/hives.jpg

[joshua corning]

(although would-be devleopers might consider the externality to be negative)

if that developer already has work in the pipeline it would be a positive one because the constrained market would raise prices.

If not then it is generally a wash...land prices are higher but those costs are just passed on to consumers.

[joshua corning]

An externality has to be an unintended conseguence, as defined by Baumol and Oates. I do not think employment and multiplier effects fall under the category of unintended.

Why?

Certainly it is not the intent of Acme corp to employ people. Its intent is to make money.

Your definition of unintended seems to require a degree of ignorance on the part of producer.

In fact if a producer could reduce costs (cutting labor) without loss of production it would have every incentive to do so.

[jim casey]

you are not a smart man - are you Josh?

[joshua corning]

you are not a smart man - are you Josh?

Which post are you referring to mister vague criticism?

[Eric H]

I have to agree with jim casey's first point: when a business buys labor, the effects are mostly internalized by both parties. When those workers then go buy products, again, the benefits are largely internalized in separate transactions.

To answer the original query, though, I think that on-the-job (OTJ) training can yield positive externalities when the training is not company-specific. For example, if you teach employees to use general-purpose technology (computers, networks, forklifts, etc.), and they in turn informally teach those skills to others (spouses, children, Boy Scouts), the result is not captured by either party to the original transaction, nor are they "captured" in subsequent transactions. I am specifically thinking of what happens in a developing country when a high-tech foreign company comes in and teaches skills that result in the development of a new industrial sector (in that country). Workers move to new employers and teach skills to other employees. It's called "learning by doing". I think this is reaching, though.

I am also thinking of the counter example in which companies train employees on proprietary processes that cannot be used outside the company. I recall reading about a comparison of Japanese and US workers in which the Japanese workers were more likely to identify as an employee of XYZ, while US workers are more likely to identify as an engineer or machinist. The Japanese corporations tend to train on their internal processes, while US employees frequently get training on general-purpose skills. The latter results in a more dynamic economy.

A better example might be the development of new ideas through R&D. Although mostly captured by the originator, the ideas eventually leak out and become copied or incorporated by others. An example might be the inkjet printer, first developed and successfully exploited by H-P, but which is now the basis of a large industry. Another would be the mouse, developed by Xerox, popularized by Apple, and used by MS.

I think the pecuniary externality, e.g. entering a market with a competitively priced product and dropping the price, is drastically underestimated by people who otherwise think that all externalities must be corrected by the state. I think this externality is what Tyler Cowen had in mind in a recent post on first v. second best economists.

[Oote]

What about regional externalities? For instance, large furniture stores are often located close to each other because costumers like to shop for new chairs 'n stuff where there are lots of different stores. So being a furniture store in one place poses a positive externality to other stores. Aren't such effects a major driver of agglomeration?

[reason]

What Oote says. In fact any city centre will do as an example.

[reason]

Read Jane Jacobs and think about network effects. Firm A has a factory that requires certain supplies. Some of these same supplies can be used by another factory, so the first factory can in fact enable other factories to open to due network effects.

But on a more trivial level, I remember as a child going every morning past a biscuit factory. Boy did it smell good.

[Christy]

I agree here with reason. The biscuit factory or any bakery and the aroma it produces has always been my favorite positive production externality. There is even a deli, where I live (its a chain but I can't remember the name of it at the moment) that has a sign in its window that advertises "free smells."

On the city center comment, I do not think that retailers locate next to each other so that consumers can shop around. Consumers no doubt like it but I would suspect that producers do not want consumers to shop around. I believe that the answer to this question lies in game theory. A first mover will undoubtably locate in the middle or city center in order to maximize the number of shoppers. The second mover's best alternative is to then locate right next door in order to split the market 50/50 if he moves farther away from the city center he will get less then 50% of the market. Assuming the stores are identical and that people only care about how far they have to drive. This is often why you see a Lowe's on one side of the street and a Home Depot on the other.

[Phaedrus]

Not sure if these are precisely externalities, but what about business that provide 'third spaces' for people to congregate and conduct business, such as Starbucks and other cafes, restarants, etc.

Also bookstores like Borders would seem to offer benefits to society that go beyond their business focus, whether it be as gathering places, reference 'libraries' or internet access points.