Transcript Comments on Designing the Microbial Research Commons

Comments on
Designing the Microbial Research
Commons: Digital Knowledge
Resources
Katherine J. Strandburg
New York University School of Law
PERSPECTIVE FOR COMMENTS:
IMPORTANCE OF SOCIAL NORMS AND
RESEARCHER PREFERENCES
HOMO SCIENTIFICUS PREFERENCES:
1) Performing research
2) Autonomy in research direction
3) Learning results of the collective research
enterprise
Scarce resources needed to satisfy preferences:
- Funding
- Attention of others
Access to these resources is mediated by
publication – if OA is to succeed it must align
with these preferences
I. OA JOURNALS
THREE PATHS TO OA
 Open Access Journals (perhaps based at
universities)
 Existing Journal Adoption of Open Access
approach
 Parallel OA manuscript repositories and
“proprietary” journals
THE IMPORTANCE OF IMPACT FACTOR
 Emphasis on high status publications
exacerbated by recent trend to quantify
publication records using impact factor
 Table 3 (p. 67):



IF of OA journals: 4.0 (with range up to 9)
IF of restrictive journals: 5.77 (with range up to 50!!)
IF of 50 trumps long-term belief in value of OA
 OA models cannot depend on scientists
foregoing publication in high impact journals


IF is path dependent and sticky – network effects,
preferential attachment, “Matthew effect”
Scientists unlikely to “vote with their feet” for the OA
mode
OTHER BARRIERS TO UNIVERSITY
PUBLISHED OA JOURNALS
 Problems with the law review model



Proliferation of journals b/c each university needs 1 (or
2 or 5)
Overly fine-tuned ranking of journals (rather than post
hoc ranking based on citation)  over-emphasis on
“placement”
Grad students are not law students



No need for publication venue
No time for journal editing functions
Is law review publication really faster?


Anecdotally, physics is 3 to 6 months
Microbial research?
 Not convinced of synergies with university
educational mission
JOURNAL ADOPTION OF OA?
 Unlikely b/c of bargaining power due to IF as
discussed above
 IP laws protect proprietary approaches and
reform is difficult
 Some movement is seen, but direct pressure on
high impact journals is difficult
 OA “tier” (e.g. Springer “Open Choice”)
problematic if payment competes with spending
on research
 Journal versions of OA not entirely satisfactory
MANUSCRIPT REPOSITORIES
 Circumvent the need to get journals to change
their practices

Need journal acquiescence only
 Separate things that universities can do easily
and well from things that are more difficult or
harder to dislodge


Good manuscript and good data mining, etc.
Hard copy printing, credentialing service
 Deposit can be mandated by funding agencies to
grant recipients


Solves collective action problem
Aligns incentives
MANUSCRIPT REPOSITORIES
 NIH Experience

Journals do not prohibit deposit in such repositories
 Federal Research Public Access Act of 2009



Recently introduced in the Senate
Mandates agencies to ensure open access deposit of peerreviewed manuscripts < 6 months after publication
Consistent with Obama administration Open Government push
 Mitigates concerns with database protection
statutes in Europe

No more sole source
 Could integrate with material/data repositories


Users of data must deposit manuscripts
data and materials associated with manuscripts must
be deposited
WHAT ABOUT PROPRIETARY
JOURNALS?
 May adapt to “service provider” role




page charges
Hard copies
Archival version
“better” or “premium” database services (competing
with the OA repository)
 May not be commercially viable



Scientific societies
Universities
“Knowledge hubs”
Could replace them, take them over, partner
Manuscript repositories = path to some OA outcome
II. DATA DEPOSITORIES
 Similar to issue of material and research tool
sharing (see earlier publications)
 Collective action problem – temptation to
withdraw w/o contributing
Scientist A
II. DATA DEPOSITORIES
Share
Other Scientists
Share
Don't Share
U(N)+M+R-C
U(1)+M+R-C
Don't Share U(N)+E–P
U(1) +E-P
U(.): value of the database, depends on N
M: first mover advantage regarding A’s data
E: incremental value of exclusive use of A’s data
R: reputational value of contributing, including attribution
P: penalty for not contributing
C: cost (including opportunity cost) of contributing
Contribute iff: R+P-C > E-M: U(N) doesn’t matter!
II. DATA DEPOSITORIES
• Roughly speaking, then, success of depository
depends on R+P-C > E-M
• Reduce costs!! (Cf. “Empty Archives,” Nature,
9/10/09)
– Easy formats
– No direct fees
• Provide rewards for contributing (e.g. attribution)
– Note these rewards must compete w/ rewards for
sharing informally with collaborators
• Provide penalties for non-contribution (funder
requirements to contribute)
• Depositories work best for interdependent data
II. DATA DEPOSITORIES
• Moral Hazard and Industry Scientists
– Withdraw w/o contributing problems may be much
greater for industry scientists
• Different motivations
• Less concern w/ reputation, funding, etc.
• Greater access to secrecy
– Should we be concerned?
• If so, may want to consider semi-commons approach
• Fee for service or data for data