My textbook defines market failure as when "the production or consumption of a good or service causes additional positive or negative externalities on a third party not involved in the economic activity". That being said, I'd like to ask, Don't all activities produce externalities? For example, the production of oil will always have negative externalities, no matter how the government intervenes. Thus the market will always fail.

Now perhaps I misunderstood. Perhaps it means whenever there are any net externalities (positive - negative), the market fails. Please tell me if this is right.

Furthermore, from what I've read, I've gathered that the market fails whenever social cost $\neq$ social benefit. But this is a different definition from the one I previously mentioned (the one only dealing with externalities).

Please tell me what market failure actually represents. Does it take into account social costs and benefits? Or only external costs and benefits? And if it only takes into account externalities, will some markets fail continuously (like consumption of oil)?

  • $\begingroup$ Does the production of bread produce externalities? If so, please mention which. $\endgroup$
    – FooBar
    Dec 31, 2015 at 8:36
  • $\begingroup$ @FooBar its barely anything, but there are a number of gases that are produced during bread production (negative externality). Furthermore, (this is kind of a stretch, but still valid) the consumption of bread adds carbohydrates to the diet which ensures healthy functioning of workers, athletes, etc. $\endgroup$
    – Airdish
    Dec 31, 2015 at 8:50
  • $\begingroup$ to expand on my last comment, the healthy functioning of workers ensures that their productivities are high, which is better for factories and their profits. $\endgroup$
    – Airdish
    Dec 31, 2015 at 10:40
  • 1
    $\begingroup$ Small comment on your side question: if the degree of externality is trivial, so too is the degree of the failure. Many examples (i.e. bread) don't seem to be a -substantial- externality. $\endgroup$ Jan 2, 2016 at 4:47
  • $\begingroup$ @RegressForward so then in those trivial cases, action isn't even required, right? $\endgroup$
    – Airdish
    Jan 2, 2016 at 8:52

4 Answers 4


I have to intervene to say that market failure and externality are not the same thing. So I do not think it is at all correct to define market failure as

when "the production or consumption of a good or service causes additional positive or negative externalities on a third party not involved in the economic activity".

Externalities are but one example of market failure. Market failure is more properly defined as any situation in which a market, left to operate without any intervention, fails to produce the efficient (welfare-maximising) allocation.

Sources of market failure include

  • Externalities: if there is a negative externality then there will tend to be too much of an activity from a social perspective—resulting in inefficiency.
  • Market power: if the market is not perfectly competitive then firms will tend to increase price above marginal cost to increase their profit. This results in consumers not buying the good even though they are willing to pay more than its cost of production—which is inefficient.
  • Information asymmetries: If one party in a transaction has an informational advantage over the other then s/he will try to exploit it to the counterparty's detriment. This, in turn will lead to transactions taking place where it would be efficient for them not to (or to mistrust and the failure to realise efficient transactions).
  • Missing markets: sometimes efficient trades don't occur because the market simply doesn't exist. For example, there is no market to insure against the risk that an unborn child will be born disabled and requiring a lifetime of care even though many parents and their children would like such insurance (an argument often used to for the existence of state-provided social security schemes).

To address your actual questions:

"Don't all activities produce externalities"? Yes, but many of these externalities are priced. For example, if I buy an apple then you can no longer consume that apple, which is an externality. However, this does not result in a market failure because the price mechanism in a competitive market ensures that I get an apple and you don't only if I am willing to pay more for that apple than you are. So the apples go to the people who value them the most, which is the efficient thing to do. Since we are doing the efficient thing, there is no market failure.

So, when should we worry about externalities? We should check whether the net effects can cancel each other out. For example, suppose that the private benefit of some action was lower than the social benefit, but that the private cost was also lower than the social cost by exactly the same amount. Then the net effect would be that MPB=MPC at exactly the same quantity where MSB=MSC. The private individual would then take the socially optimal action and there would be no market failure. A market failure only occurs if the externality is such that MPB=MPC at a quantity different to that where MSB=MSC. Only then will the behaviour of the private individual (whose optimal action is to equalise private marginal benefit and private marginal cost) differ from that which is socially optimal.

A note on marginal benefit and cost:

When performing this kind of analysis, We typically assume that the objective is to maximise to total social welfare (green line), which is defined as the difference between the total accumulated benefit of the activity (blue line) and the total accumulated cost (red line):

total social benefit and cost

The marginal social benefit is the benefit society gains if we increase consumption by one unit. In other words, the MSB is given by the slope of the TSB curve. Similarly, the MSC (defined as the extra cost bourne by society if consumption increases by one unit) is equal to the slope of the TSC curve.

Now, we observe something interesting: the total welfare curve obtains its maximum at exactly the point where the slopes of the TSB and TSC curves are equal:

marginal social benefit and cost

In other words, welfare is maximised when MSB=MSC. This is not a coincidence for this particular graph, but rather is a far more general property.

This is actually quite intuitive. Suppose that MSB > MSC. If we increased consumption by one unit then society would get MSB units of extra benefit and MSC units of extra cost. Since MSB > MSC, this results in an increase in total social welfare. Similarly, if MSB < MSC then we could reduce consumption by one unit and society would save more in costs than it would loose in benefits. So neither $MSB>MSC$ nor $MSB<MSC$ can be consistent with maximised social welfare. Only when MSB=MSC do we find that there is no way to increase welfare by increasing or reducing consumption.

  • $\begingroup$ (+1) for stressing that it is net externalities that should enter the definition of market failure, if one would want to define it this way. $\endgroup$ Dec 31, 2015 at 13:58
  • $\begingroup$ @Ubiquitous Ok that cleared a lot of my questions. So then is the quantity where MSB=MSC always the place to go? Even if MPC > MPB? $\endgroup$
    – Airdish
    Dec 31, 2015 at 14:05
  • $\begingroup$ I just wanted to say that I agree with this answer OP. When I started writing mine this wasn't up yet and I guess it was written in the meantime. So do NOT see my answer as meaning I don't think this answer is correct or sufficient. $\endgroup$
    – BB King
    Dec 31, 2015 at 16:05
  • $\begingroup$ @S.Mo Yes, a social welfare maximiser always wants to reach the point where MSB=MSC. The private individual is a member of society and his private benefit and cost are included as part of MSB and MSC. So it's not like we are ignoring that person when we insist that MSB=MSC. $\endgroup$
    – Ubiquitous
    Dec 31, 2015 at 17:11
  • 1
    $\begingroup$ @S.Mo Yes, it's just like you said: if MSB<MSC at one point, and we decrease production, the TSC falls at a faster rate than TSB. I write T(otal)SB to be sure that we never confuse it with marginal social benefit. $\endgroup$
    – Ubiquitous
    Jan 2, 2016 at 9:32

To answer your other question of what properly defines market failure:

The market fails when the socially desirable outcome is not achieved through the market. Since the market decisions are made based on cost-benefit analysis, when social (net) cost/benefit = private (net) cost/benefit then private actors in a market will make the socially optimal decision. If this deviates, they won't and hence we have a market failure.

What does this market failure mean concretely? There is either not the right amount of production or not the right amount of consumption of a good, without the government intervening.

For example:

  1. A negative externality: Smoking. If everyone in the cafe would decide how much the smoker should smoke it would be very little as everyone has a cost (health har,), but no benefit. The smoker doesn't take into account other people's costs when he makes his decision. He smokes until: His own private marginal cost = his own private marginal benefit. However socially optimal would be marginal social cost = marginal social benefit. Since social cost is larger than private cost and social marginal costs are increasing in smoking, the socially optimal amount of smoking is less than the private optimal amount. So private decisions (as in a free market) do not lead to a socially optimal outcome.

  2. A positive externality: The producer of a good with positive externalities could be the builders of nice buildings which eveyone enjoys, because they're nice to look at. The builders gain a certain amount from the buildings. However everyone else also gains, but does not pay the builders for this benefit. Therefore that desire of the people is not priced in the builder's benefit and therefore the builder does not take it into account. So if society would decide they'd want more buildings than if the builder decides.

In general there are two broad categories of market failures:

  1. Allocation failures. This is the one we mostly talkk about in economics and refers to the correct allocation, production and consumption of goods.
  2. Distributive failures. This is when the market leads to an undesirable distribution of income, i.e. too much economic inequality. This is why most countries have progressive income taxes.

Allocative failures are not just externalities although they mostly are. Here are (most of) the kinds of allocative failures so that you can better understand how market failure is defined.

  1. Externalities (positive or negative)
  2. Public goods (will be underprovided privately, which is why governments provide them. They are basically postiive externalities)
  3. Market power such as monopoly or cartels. Most importantly natural monopolies as this problem can't be solved through more competition.
  4. Assymetric Information (Moral Hazard, Adverse Selection, etc)

Basically, look at all the assumptions required for the result that the market is efficient. See where those assumptions fail and voila you have found a form of market failure.

Note that not all goods produce externalities as you might think. Further note that yes, very often the market fails. Hence a great amount of markets are actually regulated. For whatever business you open you'll need to register it get a permit, etc. However very often we don't need a lot of regulation. Often the debates are not about whether to have any regulation or not, but rather the necessary extent of regulation. This is because regulation is costly and we don't want to use it if the benefits to it are too small compared to its cost.

  • $\begingroup$ Great answer! Just a few questions. What does it mean for a market to fail due to externalities? Because no matter how you regulate it, the process will still have externalities. Could it be that the marginal externality gets less and less as you increase/decrease production/consumption? And you mentioned "when social (net) cost/benefit = private (net) cost/benefit then private actors in a market will make the socially optimal decision". How can social cost/benefit = private cost/benefit? $\endgroup$
    – Airdish
    Dec 31, 2015 at 13:59
  • $\begingroup$ When there aren't externalities (or other problems I listed). Consumption side: society's benefit from me eating that bread is the same as my benefit from eating that bread (society does not lose or gain anymore than my utility there). Prodution: If there is a tax on carbon emissions of the production of bread, then the social cost of producing will equal the cost that the producer faces (his own cost + pollution cost to everyone).No pollution (externality) then his private cost is society's cost. Society doesn't have any other cost from production than flour etc, which he pays already. $\endgroup$
    – BB King
    Dec 31, 2015 at 15:02
  • $\begingroup$ Aren't distributive and allocative failure the same thing? You're treating money as somehow special, but it's not. It's fitting that distribution and allocation mean exactly the same thing. $\endgroup$
    – B T
    Mar 11, 2016 at 19:33
  • $\begingroup$ Allocation is about flows. Distribution is about stocks. $\endgroup$
    – BB King
    Mar 11, 2016 at 23:35
  1. There are positive externalities as well as negative ones. A good could have both and they could, in principle, sum to zero.
  2. In Foobar's bread example, "...the consumption of bread adds carbohydrates to the diet which ensures healthy functioning of workers, athletes, etc. " is not an externality, it is a benefit which is priced.
  3. Many goods are taxed or subsidized and if these are equal and opposite to their level of net externality then there is no remaining externality by construction
  4. Market failures can be de minimis in the dead weight loss they cause. When it comes to a consumer good, externalities are like taxes, their dead weight losses are (roughly, depends on the functional forms) increasing in the product size of the externality and the elasticity of demand. Therefore, if demand is highly inelastic (like for milk or bread) and the externalities are likely small (again for milk or bread) the total distortion of those externalities is likely quite small.

I therefore understand an externality market failure not merely as a circumstance where private marginal benefit $\neq$ social marginal benefit but where the inequality causes economically meaningful distortions. Remember that all models are wrong; the practical question is how wrong do they have to be to not be useful. The externality model is useful only when it describes a distortion that large enough to seriously distort demand.

Can you elaborate on what distinction you are making between social costs and benefits when you say "Does it take into account social costs and benefits? Or only external costs and benefits?"?

  • $\begingroup$ i think you misunderstood my "bread" comment. please read the extension i posted. and what do you mean by the distinction i made between social costs and benefits? $\endgroup$
    – Airdish
    Dec 31, 2015 at 10:41
  • 1
    $\begingroup$ Please see my update about social vs. external cost. I did not miss-understand your bread example. The benefits of bread on worker health is a private benefit and in therefore in the demand curve for bread. $\endgroup$
    – BKay
    Dec 31, 2015 at 10:43
  • $\begingroup$ its private for the worker, but its an externality for the factory owner who the worker works for $\endgroup$
    – Airdish
    Dec 31, 2015 at 10:50
  • $\begingroup$ That's not correct. It is priced in the worker's wage. $\endgroup$
    – BKay
    Dec 31, 2015 at 11:38
  • $\begingroup$ Ah yes, then what are some possible positive externalities for the production of bread? $\endgroup$
    – Airdish
    Dec 31, 2015 at 11:41

Your economic text book sounds poorly written. To be fair, it sounds like most econ textbooks are. The quote you wrote is a terrible circular definition of externality.

Certainly not all activities cause externalities, at least not significant ones. If I buy a rock from you, there's no externality. No one else is involved. Now you might have gotten that rock by stealing, which would be an externality, but presumably you got it off your own property or bought it from someone else who did that.

If social cost exceed social benefit, then you certainly have a failure. But in such a case its likely a government failure - markets can't really produce negative net benefit. That is, unless you consider violent groups (like gangs) to be market actors.

Ubiquitous's definition is correct enough, but I would define it more specifically:

Given a particular economic environment, market failure is a situation where the motivations of market-actors incentivizes behaviors that prevent the market from eventually reaching maximally efficient activity, or incentivizes behaviors that prevent the market from eventually reaching the maximum possible rate of efficiency increase.

Most writing on market failures have all sorts of myths and misconceptions baked into them. Ubiquitous listed the following things that are not actually market failures: information asymmetries, market power, and missing markets. None of those things are market failures because they don't prevent optimal allocation of products. For example, there's a reason some markets don't exist - we even have a phrase for it: "there's no market for it".

A lot of things that seem like market failures aren't when you consider the costs. Like information asymmetry. There is a cost to getting information and ensuring equal information. That cost may exceed the value. Without recognizing where there are costs, you can't understand what "optimal" is. Optimal is the theoretical "perfect market". Optimal is a real achievable alternative.

"Is market failure constant?"

I'm unclear about what you mean, but the answer is almost certainly no. When and how much market failure happens depends on the market environment, which is combination of the legal and social landscapes. If you change a law, you change where and how often market failures. Weird social pressures can do the same, as Freakonomics will attest to.

I wrote a full article on this topic here: https://governology.substack.com/p/the-role-of-government-part-1


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.