Re: Usage-based pricing, a view

[Carl Anderson <ca25@drexel.edu>]

I think the observations are too mechanistic and assume that the burden of
paying is shifted to the end-user.  The print journal as we know it was
the perfect means of distributing a subject-oriented collection of papers
to a community interested in that subject.  It avoided the myriad
communications that would be necessary to inform all in the community of
every newly available article and to send to each just the articles of
particular interest.  Remember, journals predate postage stamps and were
the simultaneous means of distributing new research and informing
presumably interested individuals that the research had been done.

That mechanism is now well advanced toward anachronism.  We now have the
means of making virtually every article available to virtually every
reader, with means in place for readers to identify exactly the articles
they want.  Isn't it absurd that one reader should have to pay for a score
or a hundred or a thousand articles that he doesn't want just to get the
one he does.  Isn't it absurd that the largest universities can pay the
same amount for that collection of articles and get thousands or tens of
thousands of uses.  Isn't it absurd that publishers should use the same
formulae that were supposed to account for paper and ink and postage to
pay for servers and connectivity.

Use-based pricing doesn't necessarily mean that every reader pays directly
for every article obtained.  It can mean that a community pays
collectively - as universities and research institutions now do.  The
difference is in paying for what they want and not for what they don't.
Start with the present expenditure, and adjust for the future based on
experience.  Let's say a library pays $500 for a journal subscription this
year.  For next year it places $500 in a deposit account, from which $15
is deducted for each article downloaded by anyone in its community.  If
usage exceeds the $500 during the year, some previously agreed result
ensues (e.g., extra downloads are free, or the library is cut off for the
duration, or the library is billed for additional downloads, or whatever).  
If usage falls short of $500, then the library has a credit for the next
year and enjoys a pro-rata reduction in the amount to be deposited for
continued access.  If the bookkeeping seems excessive in this simple
example - and it would be - collapse all of a publisher's (or group of
publishers') journals into a pool, set the initial deposit at the combined
price of a library's subscriptions, and worry about over/underpayments for
the whole lot.

There's no reason for end-user behavior to change in such a model, so
there's no reason to expect any more abuses, misuses, or self censorship
than there is in the present model.  Professional and scientific
associations will have the same mission they do now to advance their
disciplines, so there will be no greater incentive than now to adjust
their publishing criteria (unless you want to accuse them of padding their
print publications now).

Deals can be superimposed to achieve whatever equity seems lacking, for
example reduced pricing for poorer countries.  The model itself eliminates
the huge existing inequities between large universities that can now
afford to subscribe to nearly everything and small colleges with limited
means.  Consortial arrangements are already ameliorating the disadvantages
that small colleges have faced, effectively bundling together the readers
at all consortial institutions.  Use-based pricing simply effects the same
kind of bundling at the article side of the equation, extending (with
greater precision) what the publishers are already trying to achieve
through FTE-based pricing.

Carl A. Anderson
Director of Electronic Resources
Drexel University Libraries
215-895-2771
Carl.Anderson@drexel.edu

____

On Wed, 10 Sep 2003, Ann Okerson wrote:

> There was a time when I put a good deal of thought into usage-based
> pricing. I couldn't make myself comfortable with it for a number of
> reasons, even though I certainly do recognize that it is often viewed as
> the most fundamentally 'fair' model.
>
> Here is a list of cons I came up with on the usage-based pricing question:
>
> 1) economic disincentive to usage, and so antithetical to both publishers'
> and librarians' missions;
>
> 2) incentive toward 'bad' behaviors, either to reduce usage or to 'hide'
> it in some way - just one example would be that this provides a strong
> incentive for one person to download a file and share it around, rather
> than having everyone hit the site for that same article. This both
> distorts usage stats and begins to touch on sensitive abuse questions;
>
> 3) likely to stir up debates and administrative hassles about what kind of
> usage should 'count' - Abstracts? Searches? The 500th time a particularly
> popular article is viewed? What about the second click to enlarge an
> illustration? Mistakes? Yikes! Seems like a disaster in the making...
>
> 4) harder to implement beneficial price discrimination and raises other
> fairness issues. OK, so a small research institute has relatively high
> usage, but does that make it fair to charge them the same as Yale? What
> about a big university in a developing country?
>
> 5) a good bet is that corporate customers of many key scientific titles do
> NOT use anywhere near as heavily as universities or research institutions.
> Usage-based pricing will lay bare the fundamental 'unfairness' of a dual
> pricing scheme designed to charge corporate customers more.
>
> The previous two objections can be overcome by simply deciding - more or
> less as is done today - that these are groups that will be favored or
> surcharged based on their category. I don't think the first 3 objections
> are as easily defeated.
>
> 6) what to do about first year pricing.
>
> 7) the unpredictability of year to year pricing. This latter is something
> librarians have worried about in the past, understandably because it could
> make for some very difficult budgeting.  But I would point out that it is
> also a downside for publishers. Not being able to ascertain the revenue
> from institution to institution, and year to year, presents a difficulty
> in managing costs. And don't let anyone tell you that things will smooth
> out over time. They might, but a given publisher would surely be subject
> to 'fat' years and 'lean' ones. A single 'blockbuster' article can skew
> the stats for a year for nearly everyone; and a single institution whose
> author/researchers have a particularly strong year of getting published
> will almost certainly see spikes in its own usage - and may especially
> resent paying a higher price, since the spike may be mainly related to
> works of their own faculty, the very ones they hate paying for in the
> first place!
>
> ***end**