Below is an appendix to the first essay I’m working on mentioned in the previous post. It was a note to myself to work something out a few years ago. I was irritated with the automatic assumption that price discrimination (where a seller like Qantas or Telstra sells the same thing to different people for different prices) could be presumed to be efficiency enhancing as it often is.
Being associated with monopoly, price discrimination started off in economics with a bad name (see the Dupuit quote below). It is motivated by the monopolist trying to extend his profits further by selling to customers who simply won’t pay the monopoly price but will pay a price that generates marginal profit for the monopolist. At first blush this is efficiency enchanging because it extends the market. In economics 101 perfect price discrimination leads to the same efficiency outcome as perfect competition – because customers right down to the most marginal where marginal cost equals marginal revenue get served. The only difference between perect competition and perfect of ‘first degree’ price discrimination is an equity one – that the monopolist picks up all the loot. But there’s no efficiency cost to the economy.
The rub is that customers won’t tell you how much they’re prepared to pay for the simple reason that if you tell a monopolist that, he’ll charge you more. It’s in the makeshifts that occur to get disclosure of willingness to pay that create substantial efficiency losses (picture people staying on an extra Saturday night in one city to capture fare discounts, or waiting till midnight to make a call on a mobile phone. No doubt others have produced this analysis as it’s very straightforward – but I reproduce it for what it’s worth.
I should have incorporated this in what I posted last night, but didn’t. Here it is. For the record, I don’t think there are any very strong policy implications here because it would be exceptionally difficult to regulate ‘bad’ price discrimination without harming ‘good’ price discrimination. Looking around I expect there’s a fair bit more good done with price discrimination than harm.
The one policy conclusion I would float is that wholly owned government businesses should not engage in price discrimination where increased profits involve substantial efficiency losses.
Appendix One: price discrimination that reduces welfare
It is not because of the few thousand francs which would have to be spent to put a roof over the third-class carriages or to upholster the third-class seats that some company or other has open carriages with wooden benches…. What the company is trying to do is prevent the passengers who can pay the second-class fare from traveling third-class; it hits the poor, not because it wants to hurt them, but to frighten the rich. And it is again for the same reason that the companies, having proven almost cruel to the third-class passengers and mean to the second-class ones, become lavish in dealing with first-class passengers. Having refused the poor what is necessary, they give the rich what is superfluous.
Jules Dupuit, 1849.
Price discrimination will generally enhance efficiency to the extent that it helps focus the funding of fixed costs on those most willing to pay for them. Often it will also promote equity as those more willing to pay will often be wealthier. But though contemporary economic discussion frequently pays it little heed, price discrimination often involves substantial inefficiencies. The phenomena modelled in this box have become sufficiently common for the issue to be more than a theoretical curiosity.
Price discrimination requires knowledge of consumers’ willingness to pay. Because there is often no way for the firm to discover each individual customer’s willingness to pay, those charged lower prices are frequently provided with lower quality products. Where these products generate inconvenience or lower utility to the consumer, and yet save the producer nothing, a deadweight loss is generated.
Consider the requirement to obtain discounted airline tickets to stay in a destination over on a Saturday night. This requirement is not designed to reduce the cost of carrying those passengers, but to minimise the number of business travellers who access special fares for tourists a class of passengers who are typically more price sensitive.
Now of those people who take advantage of the package, some will want to stay overnight on a Saturday. Assuming they would not have travelled without the special discount fare, economic efficiency is improved. But there will be a group who do not want to stay over on Saturday night but for which the cost savings involved will justify some level of inconvenience. Moreover, the level of inconvenience to which those in the group of those inconvenienced would be prepared to go will vary from marginal inconvenience, right up to a level of inconvenience which all but entirely offsets the benefit of the fare reduction.
Further, some of those who take advantage of the special fare would not have travelled but for the fare. In the diagram below, they are from that group of passengers between Q3 and Q4. However there are also some passengers who would have taken the trip anyway. Of these some will be staying overnight on a Saturday in any event. They are represented as those passengers between Q1 and Q2. However passengers between Q2 and Q3 are those who would have made the trip anyway, who were not intending to stay Saturday night, but who change their plans sufficiently to take advantage of the special deal.
The triangles A and B represent the inconvenience to which passengers who take the special fare have gone. The deadweight loss that has been avoided by the price discrimination is B and C.
It will be seen on inspection that the sum of the former two areas may together be smaller than, equal to or larger than the sum of the latter two areas. Analogous triangles could be demonstrated where it is the firm which incurs costs in creating products for sale at a lower price. The same example is probably a good one. There is excess capacity for travel on a Saturday night, so the Saturday overnight specials could well drive up the opportunity cost of carrying the passengers!
Of course pointing to the possibility of market failure says nothing about its extent or of the costs and benefits of alternatives. Nevertheless it might sometimes be more circumspect for us not to conclude as a matter of principle that using price discrimination to cover fixed costs is necessarily welfare enhancing. My guess is that it would be unlikely in most cases for A + B to outweigh the efficiency gains from price discrimination in most circumstances. What seems much more likely however is that the size of A + B is substantial compared with the efficiency gains of price discrimination (B+C). The benefits of price discrimination might be seriously diluted where it can be anticipated that a substantial proportion of customers will ‘pay’ for those cheaper fares with their own inconvenience. Note these costs are in addition to the increased ‘search costs’ that price discrimination frequently foists upon consumers.