Prof. Bryan Caplan

Econ 370


Week 5: Markets with Externalities

I.                     Negative Externalities

A.                 Many choices have costly side effects that selfish agents do not factor into their decisions. Economists call these costly side effects "negative externalities."

1.                  Clearest-cut case: theft.

2.                  More relevant to IO: air pollution.

3.                  Contrast with: Worker safety trade-offs.

B.                 How do you measure negative externalities? The same way we always do: willingness to pay.

C.                How do you diagram negative externalities? In addition to the demand curve, draw a "social benefits curve." With negative externalities, the social benefits curve lies below the demand curve.

D.                Social optimum is at the intersection of the social benefits curve and the supply curve, but market equilibrium is at the intersection of the demand curve and the supply curve.

E.                 Thus, it is efficient to produce less than the competitive level.

1.                  Draw the deadweight loss.

2.                  Strange implication: Given negative externalities, monopoly can be a blessing in disguise!

F.                 Negative externalities are also often called "public bads," especially when the externalities are large relative to demand (so the socially optimal quantity is close to zero).

G.                The key: non-excludability.

1.                  There is no feasible way to exclude non-payers from the cleaner air.

2.                  Since you do not have to pay to use it, selfish people will not pay to use it.

3.                  And if no one will pay for it, why would selfish producers provide it?

H.                 Economists often call this a market failure, since self-interested behavior leads to inefficient results.

I.                     Insight from Friedman: "It is easy to misinterpret problems of market failure as unfairness rather than inefficiency... The problem with public goods is not that one person pays for what someone else gets but that nobody pays and nobody gets, even though the good is worth more that it would cost to produce." (p.278)

II.                   Positive Externalities

A.                 Positive externalities are the other side of the coin. Positive externalities are beneficial side effects that selfish agents don't factor into their decisions.

1.                  IO example: beauty products.

B.                 How to diagram? Draw a social benefits curve above the demand curve.

C.                Notice: It is efficient to produce more than the competitive level.

1.                  (Q: Why isn't this true in the absence of externalities?)

2.                  Draw the deadweight loss.

D.                Positive externalities are also often called "public goods," especially when the externalities are large relative to demand (so the equilibrium quantity is close to zero).

E.                 Non-excludability is once again the key attribute. If you can't exclude, there is no incentive to pay; if there is no incentive to pay, there is no incentive to produce.

III.                  Understanding Externalities

A.                 Many textbook treatments of externalities and public goods/bads also emphasize "non-rivalrousness" - the low or zero marginal cost of providing them. But I see this as a distraction, so we'll ignore it.

B.                 David Friedman's two caveats:

1.                  Must distinguish benefits from external benefits. (E.g. education).

2.                  Must include both positive and negative externalities in your calculations. (Important case: "pecuniary externalities").

C.                Important exception: Inframarginal externalities.

D.                Some popular and plausible examples: air pollution, national defense, highways and roads (especially local roads), law enforcement (especially victimless crimes)...

E.                 Some popular but dubious examples: education, health and safety, fire, R&D...

F.                 Some unpopular but plausible examples (depending on the society): censorship, persecution of religious minorities...

IV.               Correcting for Negative Externalities

A.                 A common initial reaction people have to negative externalities is: "Ban it!"

B.                 Obvious objection: The cure is worse than the disease. Many valuable activities (like driving) and even many activities essential to life (like breathing!) - have negative externalities.

C.                If they grasp this point, many people's next impulse is to set quantitative limits like emissions inspections, or technological mandates - like new emissions standards for cars.

1.                  A particularly crazy variant: "Best Available Technology."

2.                  Perverse effects of technological mandates: Since they raise the price of new cars, they encourage people to keep driving old cars that pollute a lot more.

D.                These approaches are highly inefficient. Quantitative limits and technological mandates ignore heterogeneity: Some firms can reduce pollution more cheaply than others; some people may value polluting more than others; some technologies may cost more than they are worth.

1.                  Application: Carpool lanes.

E.                 More efficient regulatory solutions that take heterogeneity into account exist:

1.                  Taxes

2.                  Tradable permits

F.                 Advantage: This gets you the same pollution level at a lower price. Firms that can easily switch to less polluting technologies sell their permits to firms where reducing pollution is expensive.

G.                Complication: Getting the margin right. A tax on cars reduces the number of cars produced, but does nothing to discourage people who own cars from polluting.

H.                 Further complication: If government has the power to tax negative externalities, political forces may lead it to tax all sorts of things with no negative externalities to speak of. More on this later.

V.                 Correcting for Positive Externalities

A.                 A common initial reaction people have to positive externalities is: "This is a job for government, not the market."

B.                 Obvious objection: Overkill. There is no need for government to take over the whole industry just because of some positive externalities.

C.                A much less intrusive option is for government to subsidize activities with positive externalities.

D.                Getting the margin right: Suppose there are positive externalities of voter education, but not math. If you subsidize ALL education, adjusting for the externalities of voter education leads to an inefficiently high level of mathematical education.

E.                 Further complication: If government has the power to subsidize positive externalities, political forces may lead it to subsidize all sorts of things with no positive externalities to speak of. More on this later.

VI.               Externalities, Property Rights, and Coasean Bargaining

A.                 The economist Ronald Coase pointed out that government action to correct externalities is often premature.

B.                 Another solution to the externality problem is to define property rights, then allowing parties to bargain. So long as "transactions costs" are low, externalities won't be a problem.

1.                  Caveat: Common sense ethics tells us to distinguish e.g. polluters from pollutees. But from an economic point of view it can be equally efficient to make polluters pay pollutees for the right to pollute, or have pollutees pay the pollutees to pollute less.

C.                Corollary: eliminating property rights can turn any situation into a public goods problem. (E.g. provision of food).

D.                Interesting IO applications:

1.                  Malls

2.                  Smoking sections


Prof. Bryan Caplan

Econ 370


Week 6: Markets for Research and Development

I.         Dynamic Efficiency: the Rate of Increase in Productive Efficiency

A.     Recall that by definition, a firm is productively efficient if at a given moment in time it produces at the minimum AC for a given level of output.

B.     Dynamic efficiency is, by definition, the rate of increase in productive efficiency. In other words, it is the rate of decline of the AC curve. If AMD improves its productive efficiency more rapidly than Intel, then AMD is more dynamically efficient than Intel.

C.    Dynamic efficiency is enormously important for the long-run standard of living of a society. Unless AC curves fall, you can only improve people's standard of living by building more capital goods and working more hours. ("extensive growth" or "working harder")

D.    If AC curves fall - if your rate of increase in productive efficiency is positive - then you can improve people's standard of living without building more capital goods and working more hours. ("intensive growth" or "working smarter")

E.     A little example: Right now, real per capita GDP in the U.S. is about $25,000. What will (real) per capita GDP be when your grandkids enter college in, say, 50 years?


Annual Growth Rate Per Cap. GDP in 50 years

-1% $15,125

0% $25,000

1% $41,118

2% $67,290

3% $109,598

4% $177,667

F.     Most remarkable feature of the modern world (last 2-3 centuries): Productive efficiency improves in a typical year.

III.                  How Productive Efficiency Improves

A.                 Return to the simple Bertrand model with two equally efficient competitors. Question: What happens if one firm's MC suddenly falls?

B.                 Answer: The lower-cost firm takes over the market and reaps monopoly profits.

C.                More important point: The firm only earns these monopoly profits as long as it remains the most efficient. If another firm comes along with even lower costs, it loses everything.

D.                Aspiring firms want to figure out lower-cost methods of production in order to become the next market leader; market leaders want to figure out lower-cost methods of production in order to remain the market leader. Competitive innovation is a "leapfrog" race.

E.                 This simple story highlights the two routes by which productive efficiency increases:

1.                  Individual firms reducing their costs

2.                  Selection lower-cost firms are more likely to survive and expand

F.                 Interesting empirical studies have compared these two channels. Big finding: Selection matters a lot.

1.                  In manufacturing, about 70% of efficiency improvement comes from selection

2.                  In retailing, practically 100% of efficiency improvement comes from selection!

G.                Notice that many people are likely to be angry when a market leader is replaced by an upstart. New ideas often hurt market leaders and the workers who depend upon them.

H.                 A vital point, well-emphasized in Cox and Alm: This "churn" is essential for growth. In the long-run, practically everyone is better off if we take the "hard-hearted" course and let failing firms and industries die.

I.                     This suggests that one of the main reasons socialism and government ownership fail is precisely because they leave almost no role for selection. Unsuccessful capitalist firms go bankrupt; unsuccessful government industries get more subsidies to keep them afloat.

IV.   R&D and the Free-Rider Problem

A.     But where do falling costs (as well as new products, which we will consider next week) come from? What makes them fall?

B.     For the most part, because someone figured out a better way to use existing resources. This sort of brain work is just another economic activity, often referred to as the market for research and development (R&D).

C.    Most people think of R&D as a specialized activity performed in high-tech labs. But R&D happens everywhere and anywhere someone uses resources to figure out how to improve productive efficiency (or improve the quality or variety of markets - but wait for next week for that).

D.    But what incentive is there to do R&D? If you invest to develop a new idea, mightn't everyone else just copy you? Example: Software piracy.

E.     Many people accordingly see positive externalities of R&D. The result can be an inefficiently low level of investment in R&D.

F.     On the other hand, a lot of R&D obviously exists. Only a tiny fraction stems from government subsidies. Why hasn't innovation ground to a halt?

V.     Solving the R&D Free-Rider Problem, I: Patents and Copyrights

A.     A patent or copyright legally protects the sole right of an innovator to use/rent/sell the idea he or she created.

B.     Legal differences: patents apply to inventions, expire after 17 years, and can be enforced against an independent discovery. Copyrights apply to writing, music, architectural blueprints, and so on, expire after the author's death plus 50 years, and cannot be enforced against independent discovery.

C.    How does a patent or copyright work? The owner of a patent or copyright operates like a monopolist: since he has the sole right to use/rent/sell the patented/copyrighted item, he can raise price above marginal cost and earn monopoly profits.

1.      How does the continued existence of obsolete technology limit the price a patent holder can charge?

D.    BUT: since there is free competition before the patent or copyright is awarded, these monopoly profits lead people to invest in R&D until P=AC.

E.     This works much like lobbying does (the monopoly profit leads people to spend resources to become the monopolist), but there is a crucial difference: Using resources to win the monopoly profit creates the idea!

F.     Do patents and copyrights reduce allocative efficiency? Yes, because the marginal cost of using a new idea is zero even though the average cost of creating it is positive.

G.    But by helping to solve the free-rider problem associated with innovation, intellectual property rights may greatly increase dynamic efficiency.

H.     Big advantage of intellectual property rights over government subsidies to R&D: Reward depends on consumers' ultimate assessment of the innovation's value. Government-subsidized R&D does not, leading to extensive funding of costly improvements that few people want.

1.      The absurdity of the "spin-offs" argument for the space program, etc.

VI.   Solving the R&D Free-Rider Problem, II: First-Mover Advantage, Secrecy, Imitation Difficulties...

A.     Many innovations do not (could not?) receive patent or copyright protection. Ex: The "invention" of the wholesale club, or jazz music, or the airport rent-a-car firm, or a new expression...

B.     Yet this innovation exists. How is it possible?

1.      First-mover advantage. The original innovator wins long-lasting name-recognition by being first. Ex: Hertz rent-a-car.

2.      Secrecy. The original innovator just keeps the innovation a secret. Ex: The formula for Coca-Cola (it isn't patented and therefore never expires!)

3.      Imitation difficulties. It often isn't easy to imitate an innovation in a cheap and timely manner. Ex: Aircraft. Or newspapers.

4.      Fame and other non-financial motivation. How is the World Wide Web possible? Almost all webpages can be viewed free of charge.

C.    Innovation can and does happen without copyrights and patents, although these are vital in some industries (drugs, books). How does government affect the incentives to engage in non-patentable R&D?

VII.  Population and Innovation

A.     Many people see population as a drain on living standards. One important factor this ignores is the positive effect of population on R&D and thereby economic growth.

B.     Consider a person considering whether to undertake a given piece of R&D. If the total monopoly profits an innovator wins exceed the total costs of doing the R&D, he makes the investment.

C.    Notice: If population doubles, then demand doubles, and so will the monopoly profits. The more population increases, the more likely it is that any given piece of R&D will be cost-justified.

D.    Ex: Suppose 1 year of research will be worth 1 penny to every person on Earth. What would the population of the Earth have to be to make it worth your while?

E.     This principle holds up empirically. The low population periods of human history were basically stagnant as far as R&D is concerned; the modern high-population world has an amazing level of dynamic efficiency. As world population has risen, total R&D has risen - and so has R&D as a percentage of output.

VIII.Coordination Games and Path-Dependence

A.     In a coordination game, everyone needs to do the same thing. It pays to just go along with what everyone else is doing. Examples:

1.      Which side of the road to drive on.

2.      Which language to speak

3.      Which computer operating system to use.

B.     It might be better if everyone switched at once. But: is there any way to get a lot of market participants to suddenly switch at the same time? It could be hard to arrange voluntarily.

C.    This insight has sparked the so-called "path-dependence" literature: "there is path-dependence"="this is a coordination game."

D.    An important claim in the path-dependence literature is that the market is "locked-in" to many inferior technologies for coordination reasons. Most famous example: QWERTY keyboard; also Beta VCR.

E.     But is this a serious problem for R&D? Probably not, because:

1.      Market leaders (e.g. Microsoft) can get a lot of people to switch at once.

2.      In relatively concentrated industries (like many high-tech fields), meeting to solve coordination problems is cheap and easy - or would be without antitrust. (e.g. Railroads and time zones)

3.      Patents can be sold, so there is no need for a good idea to die simply due to bad management.