An interesting article in the WSJ on e-book pricing and investigations of collusion. There are two models of e-book pricing. The first (the traditional model) where, given the publisher’s wholesale price, the retailer sets the price of a particular book. This is used by Amazon. The second (the agency model) where the publisher sets the retail price and the e-retailer takes a fixed percentage of that price when a book is sold. Apple uses this model with a retailer cut of 30%.
Now, I can understand why Apple would prefer an agency model. It is not a book retailer and really is just providing a ‘site’ to publishers. In contrast, Amazon has been retailing physical books on the internet for a long time. So there are issues of retail expertise and cost mixed up in the choice of pricing models.
But the concern is that the agency model, which appears to lead to higher prices, may be anti-competitive. This is interesting because, explicit collusion aside, basic economics tells us the opposite should be the case. With linear pricing, which I understand to be the case here, the agency model should be cheaper.
Why? Remember back to double marginalisation (probably in that third year economics undergrad course). With imperfect competition at the wholesale and retail level, the traditional model leads to two-stages of ‘mark up’. The wholesaler marks up their product above cost to the retailer, then the retailer again marks up the price when selling to the customer. This leads to higher prices, fewer sales and profits that are actually lower than the ‘joint profit maximising’ profit. In contrast, in the agency model, if marginal costs shares are not too different to the revenue shares, the wholesaler can set a price that maximises joint profit and then take 70% of that maximum profit. This should lead to higher profits for both the retailer and wholesaler but lower prices (because double marginalisation sets the price too high to maximise joint profit).
For e-books, marginal costs are probably close to zero at both retail and wholesale. So why do we see the reverse pricing?
- We actually don’t see the reverse. The cheaper prices under the traditional model are on a few books and is part of marketing. The average price under the traditional model is actually higher; or
- If I actually sat down and modelled this properly (using the mathematical tools that some readers appear not to like) I would find that there are other effects that my ‘third year undergrad’ intuition has missed.
I don’t know which of these is correct (or if it is something else entirely different) but I am happy to hear suggestions.