Science 13Mar2020

SCIENCE sciencemag.org 13 MARCH 2020 • VOL 367 ISSUE 6483 1179

ILLUSTRATION: DAVIDE BONAZZI/SALZMAN ART

I

n the push for “open access” (OA)—

making scientific papers immediately

free to everyone—it’s easy to forget that

publishing costs haven’t vanished. They

have simply shifted from subscriptions

paid mostly by university librarians

to fees charged to authors. Those article-

processing fees (APCs), which can run several

thousand dollars per paper, raise concerns of

their own. Universities fear they could end up

paying more to help their scientists publish

their work than they do now for subscrip-

tions. Scientists who have small research

budgets fret that they won’t be able to afford

APCs. And some nonprofit scientific societies

that publish journals worry APCs won’t gen-

erate enough revenue to support other activi-

ties, such as meetings and training.

Now, two nonprofit publishers of promi-

nent journals have debuted new ways to sup-

port OA journals without shifting the burden

entirely to authors. “Everybody that we work

with is watching these two [new models]

closely,” says Michael Clarke of the pub-

lishing consulting firm Clarke & Esposito.

“There is not currently a good solution.”

One approach, called Subscribe to Open

and implemented this week by Annual Re-

views, would transform the nature of sub-

scriptions. To make a journal freely available,

institutions would be asked for a contribution

equivalent to their previous subscription—

minus a 5% discount that Annual Reviews

is offering to retain a critical mass of paying

institutions. To deter freeloading, Annual

Reviews says it will reimpose paywalls and

rescind the discount if not enough subscrib-

ers renew each year. It is planning to pilot

the approach in up to five of its 51 titles,

many of which are widely cited.

The Association for Computing Machin-

ery (ACM) launched a different approach

earlier this year. ACM is asking the insti-

tutions that publish the most papers in its

59 journals to pay more than they do now for

subscriptions—in some cases about 10 times

as much, or $100,000 per year. The higher

fees will allow all researchers at participating

universities to publish an unlimited number

of papers in ACM journals without paying

APCs. The average cost per paper will beat

the average market rate for APCs, the society

says. ACM is betting the approach will sus-

tain its journal revenue while it transitions

to making all the 21,000 peer-reviewed pa-

pers it publishes annually free to everyone.

So far, both approaches are getting a posi-

tive response. At Annual Reviews, some 90%

of subscribers have signed deals that, on

9 March, allowed the publisher to remove

the paywall from the Annual Review of Can-

cer Biology, says Richard Gallagher, presi-

dent and editor-in-chief. Annual Reviews

could roll out OA for up to four additional

journals this spring if other librarians accept

the model, Gallagher says. Curtis Brundy, a

library administrator at Iowa State Univer-

sity, which is participating in the pilot, be-

lieves that “hands down, Subscribe to Open

is our best option as an alternative to APCs.

It’s simpler to implement, and we don’t have

a lot of other models.”

Universities are also starting to embrace

the ACM model. In January, several that pro-

duce the most ACM papers, including the

Massachusetts Institute of Technology and

Carnegie Mellon University, signed 3-year

deals that lock in the higher payments. ACM

is optimistic more universities will follow.

Still, both publishers concede there are

risks. A big one is the free-rider problem.

“People will start to ask themselves, espe-

cially if we experience budget cuts in a given

library, ‘Why am I paying for this when other

people aren’t?’” says Lisa Janicke Hinchliffe

of the University Library at the University of

Illinois, Urbana-Champaign.

Persuading some institutions to pay more

than they do now is another challenge. ACM

calculated that it needed to do so because

two-thirds of its subscription revenue comes

from about 1700 institutions that publish

three papers per year or fewer in any ACM

journal. The rest comes from the roughly

1000 institutions that each publish more

than three papers annually. Once the jour-

nals become free, the 1700 less productive

institutions will have less incentive to sub-

scribe, and any APCs they pay won’t replace

the lost revenue.

Instead, under ACM’s plan for OA, the

11 schools that publish more than 75 papers

per year in ACM journals would pay the high-

est rate for unlimited publishing, $100,

annually, a 10-fold increase over current

subscriptions. In one of the lower tiers, the

roughly 90 universities that publish 20 to

29 papers annually would pay $35,000 annu-

ally. Overall, ACM estimates that if half of the

top 1000 institutions sign on, their journals

can transition to full OA within 5 to 7 years.

Whether the pricing structure works might

not be known for at least 3 years, when the

initial contracts expire.

Even if the models succeed, it’s not clear

they would spread. The Annual Reviews

journals, for instance, are unusual: Although

each appears just once a year and contains

only review articles, many are must-haves

for libraries.

The new experiments underscore that the

search for viable pathways to OA publica-

tion continues. The road to success likely

runs through university librarians willing

to dedicate some of their tight budgets to

testing new arrangements. “We want to sup-

port experimentation,” says Brundy, who re-

cently negotiated a deal under ACM’s new

model, which doubles what the institution

was paying annually, to $17,500. “It comes

down,” he says, “to how much of a commit-

ment a library has to open access.” j

Publishers try out alternative

pathways to open access

But can they overcome free riders and price concerns?

SCHOLARLY PUBLISHING

By Jeffrey Brainard

Published by AAAS