Price discrimiation, Usage-based pricing, and Pay-per-view

[J.F.Rowland@lboro.ac.uk]

Rick Anderson:
(snip).....

'the publisher can simply say "Five bucks an article.  Buy what you want." 
 This is, in fact, the exact opposite of price discrimination.' (snip)...

Something very odd has happened in this thread.  We were talking about
"price discrimination", which has been learnedly described by people like
Hal Varian and Andrew Odlyzko, and means charging different customers
different prices for the same product (like a seat on easyJet flight 12345
on 31st February 2004).  Then we got on to usage-based pricing, which is
the kind of thing PALS, COUNTER and other groups have been exploring very
carefully in recent times.  But suddenly we seemed to slip on to the idea
that usage-based pricing means (in Rick's words) "Five bucks an article".  
We have come round full circle to pay-per-view!

In the electronic era, most of the costs of a publishing activity are
fixed costs, which means that predictability of revenue is important for a
publisher's stability.  It always was, and it is so even more so now, when
the marginal cost of serving one more customer is almost nothing but the
fixed costs of production of the database must be covered.  So
pay-per-view is not a good idea for the *main* tariff, though it probably
is a good idea as an additional option, to enable the ocasual customer to
buy just the odd article.  "Usage-based pricing" can actually mean
something on a far less micro scale than "$5 per download".  It might
quite likely mean that what an institution pays for unlimited access to a
publisher's entire output for any one year depends on how much use that
institution's users in aggregate made of that publisher's entire output
throughout the previous year.  That is one possible basis for "usage based
price discrimination", though there could be many others.

Fytton Rowland, Loughborough University