Thread by @drdrtsai

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier 1/ Problem is not specific to #EconTwitter, pretty much applies to many social/technical sciences (#SocTwitter, #PoliSciTwitter, #AnthroTwitter, #BioStatsTwitter) where theoretical and technical contributions are prized

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier 2/ In #EconTwitter, not uncommon for RAs to spend years collecting data and writing hundreds of lines of #RStats code. In return, they get named in Acknowledgments and (hopefully) a rec letter for grad school. But not authorship.

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier 3/ If you didn't prove the lemma, create the new conceptual framework, or devise the new estimator, your contribution is not a "real" contribution (or at least the kind that is valued in the field)

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier 4/ Related issue is that in some settings data collection has become commoditized eg lots ppl basically outsource data collection to @JPAL. Employees may hv totally different set of incentives, may not be motivated by authorship, etc.

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL 5/ Related issue #2, some contributions will get you tenure vs. some contributions will not. In #EconTwitter, the Tyranny of the Top 5 is such that a Top 5 is a must. Absolutely no way around it.
aeaweb.org/articles?id=10…

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL 6/ To some, #EconTwitter's emphasis on Top 5s borders on the extreme eg., would you give up an arm for an AER? How about a thumb? (@causalinf posted some insightful tweets awhile back re: shadow exchange rate, can't find it now)
onlinelibrary.wiley.com/doi/abs/10.111…

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf 7/ Some advise that untenured profs at top progs shld focus on Top 5s & not bother with other pubs. #EconTwitter web sites littered w/ unpublished working papers (some subsequently revealed to be influential) that were abandoned back in the day b/c couldn't get into Top 5 journal

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf 8/ Related issue #3, in some fields, having co-authors will "dilute" the value of a pub in relation to tenure. Some #AnthroTwitter depts require solo authored pubs for tenure, e.g. culanth.org/fieldsights/co…

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf 9/ Heather Sarsons (@saskatchewin) showed the dilution penalty in #EconTwitter is especially severe for women, and especially for women who co-pub with men:
scholar.harvard.edu/files/sarsons/…

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin 10/ BTW note that the finalized version of her paper that is forthcoming in @JPolEcon (which is...a Top 5) now has 3 co-authors tacked on. Posting the 2017 vs. 2020 acknowledgment paras here side by side, just FYI

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon 11/ Related issue #4, large primary data collection and study implementation operations are becoming increasingly important eg. field experiments/RCTs (cf. Duflo) in a field that heretofore relied on inventive secondary analyses.

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon 12/ All of these effects are mutually reinforcing. "Real" contributions = technical and theoretical advances that get you a Top 5. Co-authors dilute your contribution. Data collection is commoditized. Data collection is new, just apply existing standards of authorship.

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon 13/ Therefore unsurprising that there are few LMIC co-authors in #EconTwitter. Other fields eg #SocTwitter less hierarchical and a little less elitist but still same pressures apply, e.g. gotta get that @asr_journal or AJS (but you can get tenure at some institutions without it)

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon @asr_journal 14/ Maybe this issue less noticeable 20y ago w/ less pubs across disciplinary boundaries? #EconTwitter generally thinks #MedTwitter pubs are junk, so they don't pursue Lancet, NEJM, etc. cf. famous flowchart by @bill_easterly/@laurafreschi from 10y ago:
nyudri.org/aidwatcharchiv…

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon @asr_journal @bill_easterly @laurafreschi 15/ #EconTwitter now publishing in science/med journals more frequently. Dunno when this started, but cf. Beaman et al @ScienceMagazine 2012 (big survey; Indian data collectors...named in acknowledgments), Karlan/Zinman 2011. Science ed. (@tage_rai) actively seeking soc/behav sci

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon @asr_journal @bill_easterly @laurafreschi @ScienceMagazine @tage_rai 16/ Given these trends, perhaps these collisions across disciplines are just more noticeable now? Not sure the answer is to impose #MedTwitter conventions on other fields. For #MedTwitter PI to give authorship to RAs is very different compared w/ #EconTwitter PI to do same.

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon @asr_journal @bill_easterly @laurafreschi @ScienceMagazine @tage_rai 17/ Tons of variation in opinions about what specific forms of contribution warrant authorship -- across fields and between countries:
ncbi.nlm.nih.gov/pmc/articles/P…

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon @asr_journal @bill_easterly @laurafreschi @ScienceMagazine @tage_rai 18/ I'm mildly skeptical of "you need a representative of X" as a co-author type rules. Encourages tokenism, etc etc. Also, blunt hammer. Shouldn't there be reasonable exceptions? What abt secondary analyses? @LancetGH had a good discussion on this issue:
thelancet.com/journals/langl…

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon @asr_journal @bill_easterly @laurafreschi @ScienceMagazine @tage_rai @LancetGH 19/ This issue is a massive coordination failure that is too big for depts, authors, journals to tackle alone. Depts can't unilaterally declare new standards--need to coordinate w/ other depts and w/ Provost. (HMS Prof of Econ is supposed to be same as Prof of Medicine.)

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon @asr_journal @bill_easterly @laurafreschi @ScienceMagazine @tage_rai @LancetGH 20/ Journals can perhaps try to mandate by fiat (eg., AER data sharing, ICMJE pre-registration). Probably works better in highly concentrated fields where tyranny of Top 5 is a thing and where everyone is a price taker.

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon @asr_journal @bill_easterly @laurafreschi @ScienceMagazine @tage_rai @LancetGH 21/ Authors can unilaterally try to adopt better practices. I dunno. All I can say is this is very costly. Incidence of cost burden definitely not equally distributed in a "you be the change" type system. Good guys just watch the bad guys finish first.

@MiriamShuchman @nberpubs @WorldJSurg @BHedtGauthier @JPAL @causalinf @saskatchewin @JPolEcon @asr_journal @bill_easterly @laurafreschi @ScienceMagazine @tage_rai @LancetGH 22/22 PS. This discussion is entirely separate from the ethics scrum that is currently going on about this paper that @MiriamShuchman pointed out in a follow up tweet.