This recent question about when monopoly is undesirable made me curious about when monopoly is actually desirable? I was reminded of an argument for the efficiency of monopolies when moral hazard effects create a backward-bending supply curve in competition, like in healthcare perhaps. But, for the life of me, I cannot remember exactly how this was justified or if I am just imagining the whole thing. Is anyone familiar with arguments like this?

I know you can create a backward bending marginal cost curve and then with the usual linear marginal revenue curve, you might have MR=MC at a greater quantity than where demand intersects marginal cost, but I don't understand how the backward bending supply is also a a marginal cost curve--then the curve isn't even a proper function.

Information is a funny thing, so I wouldn't be surprised if incomplete information made for strange results.

Edit: I am thinking of pictures like these. I am just having a hard time reconstructing the jump from a backwards bending supply or demand curve (I understand these--this is often taught in 101), I just don't understand the MB/MC interpretation below where it seems like you might get a higher equilibrium quantity with monopsony or monopoly.

enter image description here enter image description here

For these to make sense, MC and MB can't be functions of quantity unless I have a marginal cost in each direction? As in I could sell $q_1$ at a high price a low price, so am I approaching from above or below.


3 Answers 3


This definitely is not a complete answer. But I can imagine the case of increasing returns to scale. Or natural entry barrier. Increasing returns to scale is often discussed in international trade theory, but it can be also relevant here.

Example: think about commercial aircraft industry, dominated by Boeing and Airbus. Aircraft industry is seen as exhibiting increasing returns to scale, so monopolistic power those companies have (well, it is oligopoly, but that does not change the conclusion) may actually lead to cheaper quality aircraft.

Now there are people who really believe that all monopolistic powers are undesirable, but I believe those people are only minorities.

  • 1
    $\begingroup$ Airlines are also interesting because of the double marginalization problem. If flights are complements (connections), then monopoly is better than competition. $\endgroup$
    – Pburg
    Nov 23, 2014 at 10:33

I haven't seen the model you mention (so this answer should be taken with a grain of salt!) but, intuitively, I guess the story would go something like this:

Suppose a firm is selling insurance.

  • If price is very low then selling is not very profitable so supply is low.
  • Higher prices cause selling insurance to become more profitable so supply increases.
  • If insurance is very expensive then the only people who will be willing to buy it are the people who know they are almost certain to make a claim. Since selling insurance to such people is not at all attractive, the firm wishes to reduce its supply of insurance.

Thus, supply first increases and then decreases with price.

  • $\begingroup$ Doesn't this overlook the reality of the insurance market (at least in America) where most consumers are mandated by law (car) or practicality (i.e. terms of mortgage), or tax (health) to require a purchase? Not to mention many of the clauses that allow insurance companies not to pay (exclusions are a major part of any policy). I worked in the insurance industry briefly, and consider it one of the most distorted markets in existence, so I'm surprised by your choice of them as an exemplar industry. $\endgroup$ Nov 23, 2014 at 20:05
  • $\begingroup$ Well, of course the implicit model is highly stylised. But even in an insurance market consumers can substitute between alternative policies—either from different providers or from the same provider. The whole actuarial industry basically exists to ensure that insurance contracts are priced so that they attract a balanced profile of risk and do not suffer from significant adverse selection. $\endgroup$
    – Ubiquitous
    Nov 23, 2014 at 21:08

I don't recognize your moral hazard example, but there are a couple situations where a monopoly is desirable.

A monopoly is desirable when the cost of entry is close to zero or good substitutes are available, and the economies of scale are always increasing. One might argue that Amazon.com, Google, and other internet companies are cases of this. The advantage of being able to always go to the same source for everything makes the economies of scale high.

A monopoly is also desirable when the economies of scale are such that it is cheaper to pay the monopoly premium than to split between two producers. The classic case of this is with law enforcement. The cost of covering one more house on a street decreases as you cover more houses. The cheapest price is reached when one agency covers every house on the street. This would also apply to fire and ambulance services for the same reason.

Of course, if it is worth paying the monopoly premium, then it is generally worth paying the central planning penalty as well. Thus most of these services tend to be provided by governments rather than private enterprises. Even when they are provided by private enterprise, as gas, electric, cable, and phone connectivity are, they are often highly regulated as utilities.

Patents and copyrights award temporary monopolies to encourage production. Ideally the societal benefit of increased effort should outweigh the monopoly penalties. Of course, we can likely find individual instances where this is untrue.

Trademarks offer a relatively permanent monopoly to encourage brand building and responsibility. Without trademarks, the name wouldn't indicate who was providing the good or service, so determining quality would be more difficult. The trademark loses value if it becomes associated with lower quality, so the trademark owner is incented to maintain quality. Hopefully we wouldn't offer them unless the benefit outweighed the cost.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.