The Social Value of Access to the OED

MARCH 17, 2006

I was pointed at this post about making a first edition of the OED freely available online. This is a wonderful idea but here I just wanted to talk about what the value of doing this would be.

In his email Kragen Sitaker guesstimates a social value of $293 billion for the project_[1]. While clearly done with his tongue firmly in his cheek I think such calculations are important enough to be worth debating and in this case think this amount is much too high for several reasons.

First and most important what is the value of gaining access to the OED. The figure of $295 per year for access he provides is for individual subscription. But the majority of people do not access the OED this way but rather through an institutional subscription. Thus this is not the average price of a user’s access but is the highest price point used by a price-discriminating monopolist. For an institution the price tag is probably in the thousands and this license covers all access by members (for example the recent edition of the Dictionary of National Biography costs around 3000 pounds per year at its most expensive). Given that most institutions have thousands or tens of thousands of members this puts the price per year at under half a pound (~$1) or even lower.

Secondly, if we wish to calculate the social value of access for the the whole world we cannot simply extrapolate the valuation of current users to every man, woman and child on the planet. Under basic demand theory, all things being equal, the current buyers are those with the highest willingness to pay for the good. The willingness to pay of the remaining consumers will be lower, and likely a lot, lot lower, than that of these original users. Picking a number out of thin air we might estimate that the total social value of all of the remaining potential users of the OED is approximately Y times the current level of sales where Y is less than 10.

Repeating the above calculations with these new figures yield:

30 million x $1 x (1/12) x Y x (1/0.1) =~ 3Y million

Setting Y to 1 yields around $30 million which is still a very large amount but a lot less than the $293 billion floating around in the original calculation.

[1] The original calculation

Here are some thoughts on monetary estimates of the value of this

Online access to the current edition costs US$295 per year and is
currently available to about 30 million people, for a total value of
$8.8 billion per year. [snip]

If we discount the $295 per year by a yearly factor of 1.1, which is
extremely generous, we get a total of $3059 for the next 30 years.
Adding it up to infinity, we get $3245.  If we use a more reasonable
(i.e. closer to unity) discount rate, we get a larger value.

Suppose we estimate the value of having access to the public-domain
part of the OED by reference to the version that Oxford has for sale,
discounted by:
- a factor of 6 to account for the fact that the people who have
  bothered to buy access at $295 per year are those who are unusually
  devoted to words;
- a factor of 3 to account for its incompleteness;
- a factor of 2 to account of it being out-of-date;
- a factor of 2 to account for getting page images instead of ASCII

This brings the total value of the public-domain portion down to $45
per person, or $4.09 per year per person.  Approximately 99.55% of the
world's population, or about 6.5 billion people, currently doesn't
have access to the OED.

This values the public-domain version at $26.6 billion per year, or
$293 billion overall.  (If you pick a lower discount rate, the $293
billion number becomes much larger.)  That means that every page I
scan, out of the fifteen thousand or so, produces about $19.5 million
of value for the world; that's about $9.8 billion an hour.  My hourly
wages have usually been less.