Re: The Harvards, the Have-Nots, and Open Access

From: Stevan Harnad <>
Date: Sat, 20 Mar 2004 15:10:43 +0000

>sh> [A]lthough the Harvards are somewhat better off than the Have-Nots
>sh> in their institutional access, *no* institution (or institutional
>sh> consortium) has remotely enough money to afford toll access to all
>sh> or even most of the planet's 24,000 serials: only to a small and
>sh> shrinking minority of them.
>sh> Nor is open access merely a matter of making tolls more affordable:
>sh> Even if all 24,000 journals were sold at-cost (zero profit) it would
>sh> still remain true of every article published that most of its would-be
>sh> users could not access it, because most institutions worldwide still
>sh> could not afford most journals, even at-cost...

Anonymous comment:

> I expect that at least one library, namely Harvard's, could afford all these
> journals, even with the publisher profits. But your statement would surely
> be correct if you changed "no institution" to "almost no institution."

I am happy to amend the statement to "almost no institution," But let
me also add these data:

According to the Ulrich/Bowkers Serials listing at

    "The 'Ulrich's Universe' consists of the approximately 175,000 active
    status titles that are included in the Ulrich's database. The Ulrich's
    Universe does not include ceased or suspended titles, or titles that
    are "forthcoming" titles (such as those announced for publication
    or announced but never published)."

Of those 175,000 serials, Ulrichs says it has about 24,000 "refereed serials":

    "As used by Bowker in the Ulrich's database, the term refereed is
    applied to a journal that has been peer-reviewed. Refereed serials
    include articles that have been reviewed by experts and respected
    researchers in specific fields of study including the sciences,
    technology, the social sciences, arts and humanities."

According to the excellent ARL statistics at
the *total* (refereed and unrefereed) serials holdings of
the top and bottom 15 of their 115 US/Canada research universities are:

Mean 31983.4652
Median 28454.50
Std Deviation 17103.5302
Std Error Mean 1594.91242
100% Max 106869.0
99% 90707.0
95% 69218.0
90% 53934.0
75% Q3 38121.0
50% Median 28454.5

Rank Institution Total Current Serials
1 HARVARD 106869.0000
2 ILLINOIS, URBANA 90707.0000
5 YALE 69664.0000
6 MICHIGAN 69218.0000
7 CORNELL 62077.0000
8 INDIANA 60019.0000
10 VIRGINIA 55843.0000
11 COLUMBIA 54958.0000
12 TORONTO 53934.0000
14 STANFORD 50056.0000
15 TEXAS 50014.0000
101 SOUTHERN ILLINOIS 17467.0000
102 VIRGINIA TECH 16679.0000
103 GEORGE WASHINGTON 16638.0000
104 QUEEN'S 16109.0000
105 MISSOURI 16073.0000
106 LOUISVILLE 16028.0000
107 MASSACHUSETTS 15260.0000
108 WATERLOO 15251.0000
109 TULANE 14998.0000
110 KENT STATE 14605.0000
111 DELAWARE 13541.0000
112 HOWARD 13102.0000
113 GUELPH 12637.0000
114 SASKATCHEWAN 11261.0000
115 MANITOBA 9865.0000

So Harvard, which is one standard deviation above the next richest
university (and over 4 standard deviations above the mean) still
purchases only 60% of the Ulrich's total. Now this could have various
explanations. (Maybe 40% are junk, with no academic interest.) And
the 24,000 refereed journals are only 13% of the Ulrichs total and 23%
of the Harvard total. And one has no idea what the priorities are.

But the stats do suggest that even Harvard's budget is not omnipotent --
and that it is a statistical order of magnitude higher than the next
richest (and 4 SDs above the mean).

I often speak about the "Harvards" vs. the "Have-Nots" in discussing
access and impact effects, pointing out that even if the Harvards are
somewhat better off in terms of access to the research output of the
Have-Nots, the impact of their own research output is still diminished
by the fact that the Have-Nots (which are in the vast majority) cannot
access it!

But I'm happy to add the qualifier "almost" if it is still warranted in
the light of these stats. (The zero-profit conditional is, after all,

What is important, I think, is to dissociate the access/impact problem
from the affordability problem, rather than to focus on how many
universities might be able to afford how much.

    "The Affordable-Access (AA) Problem and
     The Open-Access (OA) Problem Are Not the Same"

Stevan Harnad

NOTE: A complete archive of the ongoing discussion of providing open
access to the peer-reviewed research literature online (1998-2004)
is available at the American Scientist Open Access Forum:
        To join the Forum:
        Post discussion to:
        Hypermail Archive:

Unified Dual Open-Access-Provision Policy:
    BOAI-2 ("gold"): Publish your article in a suitable open-access
            journal whenever one exists.
    BOAI-1 ("green"): Otherwise, publish your article in a suitable
            toll-access journal and also self-archive it.
Received on Sat Mar 20 2004 - 15:10:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:23 GMT