Price discrimination is the practice of selling the same product to different buyers at different prices [BD]. This is only possible under certain conditions. First condition is that the company applying price discrimination must enjoy market power. The reason for this condition is that a company in a competitive market must accept the price P* set by the intersection of demand and supply in order to sell the quantity produced Q*. On the other hand a company with market power has a certain degree of flexibility on deciding the optimal price. Second condition is that the company is able to segment the market, in other words to identify different groups of consumers with different levels of price sensitiveness.
Companies set the level of price where marginal revenue is equal to marginal cost (MR=MC). If a company enjoys market power, it will set the price higher than marginal revenue. This implies that the price set by the company with market power will be higher than the marginal cost (P*>MC=MR). This fact causes a deadweight loss proportional to the difference between the price charged and the MC=MR. In other words the company will choose to sell the products at price P* and quantity Q* and lose the chance to sell an extra quantity Q’ at a lower price P’, still above MC, for fear of driving down the prices and the profit. If a company was able to divide the consumers into different segments with different price sensitiveness and charge each segment the maximum price each would pay, the company would have higher revenue and also the deadweight loss would be reduced, making the market more efficient. The extra revenue the company will make is P’(Q’-Q*).
If it is possible to directly identify the different groups, for example segmenting the market according to the geographic location or age, we speak of third-degree price discrimination. On the other hand if it was only possible to segment the market through purchase patterns and from that infer the segment, we speak of second-degree price discrimination. Common ways to perform indirect price discrimination are quantity discounts and Sunday night stays for air flights.
An example of direct price discrimination is the price Apple chooses to charge for downloading music from the Itunes Stores (ITS). The price differs substantially from US to Euro Area as we see in table 1.
| Country | Price for one download | Comparison (US = 100) |
| US | .99$ = .74€ | 100 |
| Euro Area | .99€ | 134 |
Table 1 - Prices and Exchange Rates taken on 10th April 2007 [ITUNES]
Apple enjoys market power because there are not direct (legal) competitors and it can successfully geographically segment the market by relating the credit card number to the country of residence of the customer. The high price differential (34%) is consistent with the theory.
It is more difficult to assess if applying a price differential between US and Europe is the correct strategy. Is Apple correct in guessing that US and Euro Area have a different elasticity of demand? In order to estimate the correctness of Apple hypothesis we could try to estimate the number of purchases per year per person across US end Europe. Apple does not publish these data but according to my calculations (see appendix for methodology) the number of estimated sales per person per year in the European countries served by ITS is 0.3 while in US is 0.5. We can estimate that Apple is cashing 0.4€ per person in US and 0.3€ per person in Euro Area for a total of 120m€ in US and 102m€ in Euro Area [see Appendix]. The gap in revenue per person and gap in total revenue suggests that the demand elasticity might not be so different to account for a 34% price differential. I would expect that Apple sought to get about the same revenue per person across the ocean. These calculations are just rough estimates since ITS cost structure and selling data are the best kept secrets at Apple.
An indication that the estimates made in this brief research are roughly correct and that Apple’s direct price discrimination strategy might not be optimal, is the fact that Apple has recently introduced the “My Album” service, where customers can purchase an entire album with a discount. This is an example of indirect price discrimination. In addition to that, Apple is studying the possibility to introduce a price discrimination [ECO07] based on perceived technical quality of the track. Apple will sell tracks with a higher sound quality at a higher price [FT030407]. Apple is also evaluating the possibility to sell the newest titles at a higher price [FT090407] which is one of the price discrimination techniques CD retailers normally use.
Bibliography
[BD] business dictionary on bestOfBizGold http://www.bestofbiz.com
[ITUNES] itunes online store as 9th April 2007
[FT030407] Financial Times (London); 03 April 2007; ALINE VAN DUYN; p. 28
[FT090407] Financial Times (London); 09 April 2007; EMIKO TERAZONO; p. 22
[ECO07] The Economist 4th April 2007 “A change of tune”
[FOR06] Forrester December 2006 “Few iPod Owners are Big Itunes Buyers” by Bernoff
[Apple06] Apple press statement 2 august 2006
[BUS07] “Fun With The Forrester Report and My Calculator” by Arik Hesseldhal
Appendix: Methodology used for estimating the number of purchases per year per person.
US: in [FOR06] it is stated that in 2005 0.18 tracks per households where bought in US. According to the US census there where 300 million persons and 110 million households in US in 2005. This means that about 0.5 track per person where bought in US in 2006. The revenue per person is calculated as (price per track) * (track per person sold) = 0.4€.
Europe: the total of the population of European countries served by Itunes is about 340million in 2005, according to Eurostat. In [Apple06] it is claimed that Apple sells about 100million tracks per year in Europe. This means about 0.3 tracks per person in Europe in 2005. The revenue per person is calculated as (price per track) * (track per person sold) = 0.3€.
[see also BUS07]
