Price Discrimination

In its simplest form, price discrimination is selling the same product at different prices according to the buyers, volumes, locations, dates, payment methods, and so forth. This commercial practice has very bad press, as well as being prohibited under competition law.[1] And yet, it also has qualities of efficiency and fairness that merit consideration. To understand the difference of opinion between economists and lawyers on the subject, let’s take a closer look at the example of the British retail energy market.

Supplying natural gas and electricity to UK households

While from a continental perspective, the UK market supplying gas and electricity to the general public is a model to be followed, the UK authorities responsible for regulating this industry are not satisfied and repeatedly launch investigations to detect competition law infringements. One of the charges against the six major firms in this sector (the Big 6) concerns spatial price discrimination.[2]

To understand the situation, we must remember that the Big 6 are the heirs of regional companies after the privatization of the industry in 1990. Whether selling electricity or gas, these six companies and their smaller challengers are supposed to compete fairly across the entire British territory since 1999. Yet, this means forgetting consumer inertia. Each of the Big 6, in fact, has maintained a large portfolio of historical clients reluctant to change providers. This is also why, fifteen years after fully opening the industry to competition, the Big 6 are still big.

In fact, there are a significant number of residential customers who, although not entirely captives of their historical regional supplier, still remain deaf to competitors’ offers. For them to agree to leave the safety of their incumbent provider, they would have to be offered substantial discounts, which are hardly profitable in an industry where margins are already low. But there is a second kind of residential customer, younger, more active and experienced in price comparison shopping. These people have no difficulty switching suppliers to cut their bills, as they also do for their phone company, their ISP, and their banker. It is therefore not surprising that in such a market, every major vendor tends to offer prices that vary depending on postcode: in-area tariffs, for the region where the company has historically been located, target its least active customers, and ‘out-of-area’ tariffs for other regions where customers must be contacted door to door, knowing that there is little chance of winning over customers loyal to one’s competitors.

Equality, Equity, Uniformity

This practice has been deemed inappropriate by the UK’s energy regulator (Ofgem), which has intervened repeatedly in recent years to limit contract terms that are deemed anti-competitive.[3] In 2009, Ofgem thus imposed a non-discrimination licence condition (known as SLC25A) for a period of three years. However, while Ofgem expected an increase in competition, it observed that competition had in fact been reduced and so did not renew the policy.[4]

To take the image used by S. Littlechild,[5]  the UK energy regulator, as with most authorities in democratic regimes, tends to behave like Procrustes, that brigand of Attica, who laid travellers on a bed and then, wishing them all to be the same size, stretched the prisoners who were too short and cut off the feet of those who were too tall. Uniformisation has the advantage of simplicity, but also many disadvantages such as inefficiency and some forms of inequity.

Consider a market with two types of consumers. Some are willing to pay a higher price for the product, while others can only afford a low price. If sellers can differentiate between the two groups, they will ask the first group for a higher price than the one they give to the second group. If price discrimination is not possible, with a low single price sellers serve the entire market but lose the margins that they would have made on the top segment. With a high single price, the seller may increase profits but abandons the lower segment of the market.

Thus, we can see that:

i) A single price is ineffective when is set at high values, because it allows the seller to make profits by abandoning the least profitable customers.

ii) A single price is effective if it is low enough to serve everyone, but it may not be sufficient to finance the operator’s fixed costs. Moreover, it leaves a very high amount of money left over for those who were willing to pay a lot more. Billionaires pay the same price for gas and electricity as other customers.

iii) Price discrimination is effective because it encourages an increase in the collective surplus, but is unfair since the pockets of consumers are emptied to benefit the sellers’ shareholders. A flat tax on corporate profits for those who opt for this price system would enable redistribution of that money.

In the UK, the regulator hoped to shift from situation iii) to situation ii). But suppliers, who were required by SLC25A to offer the same price both in their traditional customer zone and outside of this area, did not lower the in-area tariff as Ofgem had hoped. Instead, they increased the out-of-area tariff, shifting from situation iii) to situation i). Observing the anti-competitive effect of this uniformisation, Ofgem abandoned the condition three years after creating it.

Fifty Shades of Price 

For a company, the ideal situation is one in which it has no competitor and knows the willingness to pay of each of its clients, who can not carry out transactions among themselves to exchange the product after having acquired it. Without any legal restrictions, a company serves all the clients whose willingness to pay exceeds the company’s costs and thus maximizes the consumers’ surplus . . . to its own benefit.[6] Outside of this situation, things go badly for the company. This is a boon for the economist, however, as knowing who wins and who loses with a price discrimination policy requires careful analysis of each individual case.[7]

It is easy to understand that competitors willing to offer discounts to high-paying consumers, the possibility of parallel trading (purchasing cars or medicine in market segments where prices are low and reselling them where the price is high), and sellers’ relative ignorance of buyers’ willingness to pay are all obstacles to price discrimination. To these, we must also add transaction costs, that is, the difficulty of managing a portfolio of clients with multiple pricing schemes.

The last two obstacles mentioned above can be reduced through information technology, particularly by using the enormous databases of consumer habits created by sellers (Big Data). In the energy sector, installing so-called smart meters and boxes, enabling users to control their use of appliances and heating, heralds a future of service offerings targeted to users’ specific consumption profiles coupled with personalized pricing. Big Data even fosters greater price discrimination and thus opens up a new field of investigation for economists, as well as for competition authorities who must remain vigilant about protecting personal data and the conditions for accessing these databases by small businesses.

*   *   * 

It is precisely because there is no general law on the effects of price discrimination that it is actually practiced widely in a variety of disguised forms without raising strong objections. In airplanes and trains, the extra inches granted to first class passengers to protect their knees are charged at a much higher price than their actual additional cost. Customer loyalty schemes, and the "we'll refund the difference if you find it cheaper elsewhere," are all various forms of price discrimination. The choice left to the customer to take a monthly phone plan or pay as you go is another. In the energy sector, the most widespread form of price discrimination involves pricing made of several parts, such as a completely fixed standard charge, a second fee depending on the amount of power signed up for, and a third based on consumption. All these pricing methods have great qualities in terms of efficiency, but we must keep a close eye on their potential redistributive biases. Here, as with many things, it is only abuse that is dangerous for the economy’s health.