Thursday, May 22, 2014

Top 31 IP scholars by h-index

Scholarometer is a citation-analysis tool developed by researchers at Indiana University Bloomington. It mines Google Scholar data to generate a scholar's h-index: the maximum number of articles h such that each has received at least h citations. (It also computes the hs-index, which normalizes an h-index by the discipline average, which is 14.8 for law.)

Dan Burk (UC Irvine) used Scholarometer to compute the h-index for full-time AALS faculty whose major emphasis is (at least arguably) IP, and he provided me with the results to post here. He noted that while Google Scholar is not a great resource for total citations, it is relatively easy to clean up the data for computing the h-index by removing misattributions and spurious citations. He also removed science publications for those who had them (including himself) because he was interested in law publications.

There are of course problems both with citation-based impact measures and with Google Scholar's data, which I will not rehash here. But with those caveats, based on Burk's calculations, here are the top 31 IP scholars by h-index (all of whom have an h-index well above the average of 14.8):

1Mark LemleyStanford73
2Pamela SamuelsonUC Berkeley43
3Lawrence Lessig*Harvard41
4Robert MergesUC Berkeley39
5Jerome ReichmanDuke37
6Dan BurkUC Irvine35
7Jane GinsburgColumbia34
8Rochelle DreyfussNYU31
8Yochai Benkler*Harvard31
8Rebecca EisenbergMichigan31
11Peter YuDrake29
12Gideon Parchomovsky*Penn26
12Raymond Nimmer*Houston26
14Peter MenellUC Berkeley25
14Thomas CotterMinnesota25
14Christopher Yoo*Penn25
14Michael Froomkin*Miami25
18Jessica LitmanMichigan24
18Dennis KarjalaArizona State24
20Wendy GordonBoston U23
21Julie CohenGeorgetown22
21Jay KesanIllinois22
21Michael Abramowicz*GWU22
24Michael MeurerBoston U21
24Jonathan ZittrainHarvard21
24Paul HealdIllinois21
24David McGowanSan Diego21
28Paul GoldsteinStanford20
28John DuffyVirginia20
30Arti RaiDuke19
30David PostTemple19

* Denotes significant legal publications outside of intellectual property.

Update: This table was corrected on 6/10/14 to add Thomas Cotter, who was missing from the original chart.


  1. Interesting - pretty steep dropoff - if the law average is 14.8, and the 30th highest here is 19, it implies a pretty vast middle between 10 and 18

  2. Personally, I'm trying to get my head around the steep dropoff from 1 to 2.

    But more to your point: remember that 14-ish is the mean for all law scholars -- all the Posners, Amars, Volokhs, Tushnets, Mnookins, etc., not just IP scholars.

    I've run about 150 IP names through the database -- by no means all the IP scholars in the AALS, but I long ago exhausted any I could find with scores similar to those Lisa posted. You would have to do a more comprehensive study to get the IP mean, but eyeballing it I wonder if the mean for IP scholars isn't closer to 9 or 10.

    (Btw, this would make sense if you remember that there is a seniority bias to the h score and most of the AALS hiring in IP has been done in the last 10 years or so.)

  3. I think that's right, Dan. Mine was 11, but given 7 years in the academy, I'll take it (though I think it included a couple people who are not me). Maybe the vast middle is more like 8-12. The one caveat is that other citation count studies show heavy weighing toward IP - but maybe that's about higher citation of a few articles, rather than the breadth that the h-score requires.

  4. Dan, how does this take account of self-citation? For people whose scholarship is more thematically unified, self-citations would be higher in general and would tend to accumulate more quickly into a higher h-score—e.g., paper 2 cites paper 1, paper 3 cites papers 2 and 1, etc.


  5. As far as I can tell, self-citations count, which is also interesting in light of recent evidence that men are 64% more likely to self-cite than woman. See (paywalled) or (non-paywalled).

  6. Lisa is correct that Google Scholar does not distinguish self-citation, or even more to Saurabh's point, good old fashioned CLS-style collusive citation.

    My suspicion is that a much stronger influence is non-peer citation. The results Lisa posted are limited to faculty from AALS member schools. I have tried running a few European academics through Scholarometer and they tend to fare comparatively poorly. My hypothesis is that this is because they mostly publish in professionally edited peer review journals where student notes are unknown, and American citation rates are heavily inflated by student citation.

    Some further evidence of this comes from this LSE study suggesting that senior legal scholars have an h-index of about 2.8 -- clearly not the case for American academics:

    Someone with a lot of time or a big research assistant fund might try culling citations to produce a peer citation study.