Sunday, July 03, 2016

Economist gekokujo


"Gekokujo" (下克上) is a Japanese word meaning "low overcoming high". It refers to when the people lower down in a hierarchy rise up and overthrow those above them. Rebellions, mutinies, and populist uprisings are all gekokujo. The word was often used in conjunction with coup attempts by lower-level military officers in the years before WW2. Japan, it turns out, is a lot less hierarchical - or at least, obedient - of a place than people think.

One culture I know that is very hierarchical is the econ profession. It has a centralized job market. Academic hiring is heavily weighted toward candidates from top departments. And it's sort of an open secret that old, famous professors can essentially publish whatever they want in top journals, which gives these old famous people enormous clout.

As in most academic fields, publishing in one of a few top journals is key to career advancement. But in recent years, more and more top-journal publishing has been dominated by one single journal: the American Economic Review:


In econ, the standards for what constitutes good research and what constitutes bad research are less clear than in the natural sciences (though far clearer than in the humanities). That adds to the suspicion among the lower ranks of the profession - grad students, younger profs, profs and students from lower-ranked departments, etc. - that successful economists got where they are by sucking up to old famous people. "It's a sleazy profession," a young and talented assistant prof once confided to me, over a beer. "I'd advise you to get out of it." (I ended up taking his advice.)

Anyway, it makes sense that econ would be ripe for some gekokujo. And it also makes sense that the rebellion would erupt from the Econ Job Market Rumors forum.

Almost everyone except EJMR regulars despises EJMR. The site has a fair amount of the sexism and other aggression so common on anonymous internet forums (it's been called "4chan for economists", which in my opinion is a bit of an insult to 4chan's creativity and sense of humor), but that can't account for the near-universal visceral hatred that most economists feel toward the place. Most people who tell me how much they hate EJMR talk about what one PhD student called its "insecurities, careerism, and insane status and hierarchy obsession." Essentially, EJMR vacillates between slavish worship and bitter disdain for the top people and organizations in the profession. These people deeply believe that they need to suck up to the big dogs to get the brass ring, and even though they go along with it - econ jobs, with their high salary, lucrative consulting gigs, and intellectual prestige, tend to be a pretty nice brass ring - they don't like having to do it. So if econ is going to have an equivalent of the gung-ho young Japanese officers who launched repeated coup attempts in the 1930s, it stands to reason they'd come from EJMR.

It also stands to reason that the target of the uprising would be the AER, one of the profession's most important gatekeepers. So I was hardly surprised to see a publication scandal involving EJMR and the AER. Essentially, EJMR people started accusing an AER editor of 1) favoritism and bias, and 2) overlooked citations. These are typical academic sore points, and of course this dispute obey's Sayre's Law. But I also see it as a proxy for a wider dissatisfaction with the centralized, hierarchical, and somewhat arbitrary nature of a lucrative and prestigious profession.

There's a possible political angle here too, though I don't want to play it up too much. The AER editor being accused of favoritism, and the young authors accused of being the beneficiaries, are all women. The paper is a reduced-form empirical paper of the type that is taking over more and more of econ publishing. And the paper's policy conclusions are that unemployment is bad and that poverty could be transmitted from generation to generation by mechanisms other than genetic ability.

Like Japanese army officers in the 1930s, EJMR people tend to be right-leaning types in addition to careerists. The few times I could bring myself to read EJMR, I always saw a combination of A) disdain for political liberalism, B) disdain toward "reg monkeys" (a pejorative term for economists who do reduced form empirics), and C) bitterness toward women in general and toward allegations of sexism in the econ profession. It seems possible to me that many smart young conservative men were drawn to study economics, attracted by its traditional support for free markets and its traditional skew toward the male demographic - but that when they arrived, they found a profession increasingly tilted toward the left, increasingly populated by women, and increasingly accepting of the use of reduced-form regressions that can easily be interpreted as making a case for government intervention. George Borjas, who came out in support of the rebels, is certainly known to lean pretty strongly to the political right. This episode of EJMR gekokujo might be part of a more general backlash against these trends by angry young conservative men who feel that the econ elite, by favoring women and/or liberal politics, is depriving them of their just desserts.

Or not. That was just a thought I had, and it seemed kind of interesting, so I wrote it down. Don't take these musings of mine as any kind of evidence that left-right politics or gender politics is actually involved. But whatever the larger political context (or lack thereof), it clearly seems to me like a case of econ's "bottom feeders" (as one top economist called EJMR's denizens) rising up against the masters of their lucrative, prestigious little universe. I wouldn't be surprised to see more such "incidents", as Japanese history politely refers to them.


Update

Regarding the sexism thing, someone sent me the following screenshot from an EJMR thread discussing this post:


I don't think this kind of thinking totally dominates EJMR, and I don't know how much it's motivating this particular uproar. But I do think there's a fair amount of it there, which is why I mentioned it. It's also why I didn't try to comment on the merits of the case...I don't like pseudo-corruption and favoritism, and it certainly seems like a pervasive problem in econ, but I'm also pretty wary of joining what might turn out to be a sexist witch hunt...

Also, check out some interesting comments from someone claiming to be a recent job candidate. He or she raises a great point, which is that in the age of empirical economics, data access is really important, and that many young and/or struggling economists are mad about top institutions getting preferential early access to data. Early access allows simple, easy analyses - which many non-top economists are perfectly capable of doing - to get published in top journals, because the data is novel. That, argues the commenter, is causing a lot of anger in the lower ranks of the profession, some of which might be getting displaced as anger against perceived nepotism (which is actually a different issue). That rings true to me, and I will definitely follow up on the data access issue...

52 comments:

  1. The last thing EJMR needs is more exposure, but at least you kept this to your blog and didn't advertise it on bloomberg.

    ReplyDelete
  2. Anonymous6:58 PM

    The revolt against the AEA mafia is nothing new. The Economic Logic blog has for several years lead a write-in campaign to get representation from other than the top 10 universities in the AEA leadership.

    ReplyDelete
  3. Anonymous7:48 PM

    Framing this as a "sexist bottom-feeder revolt" is far from fair. Just because they're sexist bottom-feeders -- which they definitely are -- doesn't make them wrong. There were major mistakes made in the handling of that paper. Of the four top-tier economists I've seen comment on this, none have disagreed with that. Specifically, Borjas and Hammermesh have directly agreed while Hoynes and Currie skirted the substantive question entirely and resorted to ad hominem.

    What remains is a confirmation of the bottom-feeders' obsession: being well connected is more important than doing quality research and the well-known connected take care of their own and suffer no consequences.

    ReplyDelete
    Replies
    1. This is an interesting question.

      What if A) the whole profession is riddled with mild pseudo-corruption, so that this scandal is actually par for the course, and B) EJMR people chose this target because they're angry at women and liberals?

      In that case, their complaint would be perfectly legit and correct, AND it would be unfair to go along with their witch-hunt.

      Interesting, right?

      Delete
    2. Anonymous6:06 PM

      So I am one of the unfortunate people who was asked to help moderate the site. I think most of the economists in this site are

      1. Engage in reduced form work. There are theorist and macro economist participants, but they don't make up a majority.
      2. Center left

      I think there is an very active alt right presence in the politics forum, much like 4-chan's random board attracts hacker and pedophiles. Neither group seems to be part of either sites original intended target audience.

      If my view of the site is correct : then the central question you are asking is what amount is discussion of this recent scandal is being fueled by sexism? I think that's a valid question.

      I don't think its a matter of really picking target because they are women/liberal. This is not the first time that ejmr has been used to discuss allegations of professional misconduct by influential and most of the targets of the past discussions were men. However, it could be that sexism is fuel for the witch hunt. Its probably 2nd order fuel. I think the main reason for anger is that this allegation was made visible and transparent. Others were not.

      Delete
    3. Thanks. Interesting stuff. (1) is something I didn't realize, I guess because back when I read the site a bit, I tended to read the macro stuff.

      And also: Oh God, they asked you to moderate EJMR?? That's like trying to turn fire ants into a string orchestra. "Unfortunate" is right!

      Delete
    4. Anonymous8:30 PM

      As someone who is somewhat familiar with EJMR due to being on the market this past year, I would agree that the sexism is second order.

      The three big recent "scandals" on that site are (a) the job market paper that controlled for log(NAICS) and still landed a good job; (b) 16 and pregnant; and (c) this paper in the AER. The author of (a) is male, there authors of (b) are one male and one female, and the authors of (c) are female.

      It's true that the only one of this to garner outside interest is (c), the all female paper. But it's also the one that's actually a scandal, where the others are very debatable. Some people are undoubtedly extra upset that they're female, but the universal and driving discontent is the high-ranked nepotism.

      And given the response of the profession (especially Janet Currie's) I really don't blame them.

      Delete
    5. Sounds about right. I think the root problem is that nobody really knows what's good research. Controlling for log(NAICS) is easy, but hey, maybe it yielded a really good insight (I don't know the paper in question). Or not. Who really knows? That means powerful people are even more free to publish whatever they feel like, which creates an incentive for nepotism.

      Would you say that sounds about right?

      Delete
    6. Anonymous12:17 AM

      Not exactly. I think the problem is that in empirical work people feel that there is bias coming from institutional affiliation. Researchers who are not part of the 'top' circles feel like they are held to a higher standard than those that aren't, its difficult to prove. That is what is difficult to prove.

      People at top places have more resources. This I think effects applied people more. Top researchers get preferred access to data. This allows them to commonly hide behind a defense that their work is cutting edge, because they were the first to use 'x' topic using a novel data set. Its often that 'x' topic isn't 'new', the empirical study of 'x' is. They were the first, they were allowed to take short cuts. The worker bees who work on these topics wouldn't be allowed to do so in subsequent studies. Since they weren't the first empirical study of the topic or didn't provide more interesting data, their publication won't go into a top journal.

      That is the heart of the matter with the current 'scanda'. You have a clear case where the traditional defense 'novel data' isn't true. Which has lead people to ask why is this in a top journal? The answer people come to is nepotism.

      The one thing I am interested in is to what degree do macroeconomist or theorist feel like that there exists nepotism bias in the profession. My feeling is those feelings are less common in those fields, because there is either less use of data or data used is from well known public data sets. There might be broader consensus on what 'good' research is.

      Delete
    7. Makes sense. The last paragraph of what you wrote is what I was trying to say. But the data access issue is really interesting. Do you think some of the anger over nepotism is displaced anger over preferential data access?

      Delete
    8. Anonymous7:31 AM

      How does one make valid inference or a "really good insight" using a non-ordered categorical variable (NAICS) as a continuous variable?

      Delete
    9. Anonymous10:04 AM

      For those not familiar, NAICS is a set of (arbitrary) industry codes, an update to SIC, where, e.g., 44 is retail sales establishments, 45 is wholesale, 10 is mining, etc.

      So adding a *slope* coefficient for NAICS, let alone for log(NAICS), because it makes your coefficient of interest look better, shows a complete lack of understanding of econometrics and such a person would unlikely be able to get a job at a top school.

      Unless the student comes from an influential school or influential advisers.

      Delete
    10. Anonymous10:46 AM

      There is no disagreement that including the natural logarithm of NAICS codes in a regression is not the correct way to control for industry. Doing so demonstrates a basic lack of understanding of the empirical research.

      The further problem is that the paper in question used proprietary data, was quickly "corrected" with *no* change in results (one would think that properly controlling for industry would change the results), and the researcher in question then got a top academic job.

      Thus the problem is *not* that quality was impossible to verify in this case. It is that in this, the rare case in which the lack of quality was *indisputably evident*, the person from the good program pushed by his advisers got the top job.

      Delete
    11. Anonymous1:14 PM

      Controlling for log(NAICS) yielding good insight? Oh dear, no, no, no. NAICS is the North American Industry Classification System. It creates a system of numbers whereby each industry gets a multi-digit code with the first couple of digits encoding a sector and following digits indicating industries and sub-industries. There is no ordering of sectors or industries. (How would one do it even?)

      The proper way to use NAICS in a regression is as a control to get fixed effects for each sector or industry... depending on the analyst's judgment. That means NAICS is treated as a nominal variable: you break the number into separate indicator variables, one for each sector. Instead, Emil just used the number and then (weirder) took the natural log of the number.

      To see how absurd this is, let's look at two NAICS numbers: 621492 (62 = healthcare, 1492 = kidney dialysis center), and 711211 (71 = arts and entertainment, 1211 = sports teams). There is zero meaning in the distance or level of these numbers: 62 vs 71 does not have 9 units of difference. These numbers could just as easily be letters.

      As for EJMR being sexist, I have also moderated there and there are some amazing discussions if you know where to look. Turn off politics, off-topic, and trash and browse all the other stuff. It isn't very sexists then. (Unless you are conflating discussions of sex with sexism... in which case we have bigger problems to overcome.)

      Delete
    12. Ah, I see what you mean. Yes, that's silly. Did that mistake critically affect the results of the paper, or was it just a random dumb mistake?

      Yeah, the two or three sex-related topics I saw on EJMR during the brief time I read it made me utterly despair for the sex lives of young economists... :-(

      Delete
    13. Anonymous3:30 PM

      "Did that mistake critically affect the results of the paper, or was it just a random dumb mistake?"

      That's where the similarities to the PP-MRS-HH scandal come in. Controlling for log(NAICS) is meaningless. However, in the original paper this "control" was significant. Then, the scandal hits and ES changed the regressions to properly control for NAICS as a nominal variable. And yet... the regressions were unchanged, the results still held, and all was good. That is incredibly unlikely.

      It gets worse when one realizes that his adviser was Rob Engle -- a Nobel Prize*-winning econometrician. Now, I know Rob and I've met Emil, so I don't think there was any malice. However, Rob is a very busy person and he might have (like many advisers) not given things a careful reading, holding off until before the defense. That is typical of faculty: that way we give students a chance to find things on their own and correct their own mistakes; it stops us from micromanaging students; and, we trust that they have absorbed our teachings.

      So it makes Rob and Viral and other people on his committee look bad. However, it makes ES look *very* bad. His only saving graces were a nice natural experiment (the Fukushima disaster allowing him to see effects on CDSs due to an exogenous shock on Japanese firms) and a nice new dataset from the OFR (which let him characterize how intermediation has changed in the CDS market). However, did he get preferential early access to the OFR dataset? I don't know, but I do know that I heard about his paper before I even heard they had made data available to researchers... so that might also be an issue here.

      Delete
    14. I still don't quite understand...the right approach is just to use industry fixed effects, is it not? If you accidentally log the industry codes, that's not going to change a fixed effects specification. And if instead someone's trying to use the code itself as a regressor, that seems wrong whether they log it or not.

      And yeah, the preferential early access thing seems like a big, pervasive issue.

      Delete
    15. Anonymous5:02 PM

      Using industry fixed effects is the right thing to do. However, let's say you had two NAICS codes in your data: 62 and 71. Most software, properly commanded, would make an indicator variable for sector 71 and use 62 as the baseline. Including log(NAICS) would, instead, add a column with two different values: log(62) or log(71). That's nonsense because the NAICS numbers do not have any inherent meaning.

      Delete
    16. Oh OK, I was assuming you'd just make indicators yourself.

      Delete
    17. Anonymous6:47 PM

      Hi Noah, see the table on page 67 of an old version of the jmp:

      http://faculty.chicagobooth.edu/workshops/financelunch/pdf/siriwardanejmp.pdf

      It is baffling...first to include the NAICS as a regressor and second to try a transformation of it as a regressor (what could be the purpose of this?)

      Delete
    18. Anonymous7:04 PM

      Well, you could make the indicator variables, but most statistical software can be easily commanded to do it for you. That's standard practice and the whole theory of this date back to Ronald Fisher and the field of experimental analysis. (Although, in truth, some results were known before Fisher.) That ES didn't even realize what fixed effects were is just the initial thing meant people were troubled by.

      Delete
    19. Haha, Fisher. Yeah. Well anyway, I'm just kind of amazed that people went through these papers and checked this carefully...how are there enough people, with enough free hours, to check the hundreds of papers that come out all the time??

      Delete
  4. It would be more interesting if the EJMR complaints have merit, AND EJMR offers a prejudicial forum. Why didn't a rationalist with the personality of a roomba catch the problems? Perhaps these insecurities provide the extra juice to breach competing abuses that were designed to stymie healthy minded people.

    ReplyDelete
  5. But Noah, the bottom-feeding gutter and smut is exactly what makes EJMR so enthralling and cathartic!

    ReplyDelete
  6. Anonymous4:18 AM

    Can we please stop this horrific sexism angle. Both in the teenage pregnancy case, and in this case the losers are females. Class is a female. The Houston PhD student girl writing about natural disasters during pregnancy (and most likely OP)? Female. the student who wrote the exact same article on teenage pregnancy with her paper ending up in the Southern Journal rather than the AER? Female. The authors of the Brazilian paper on teenage pregnancies who ended up in AEJ rather than AER? Female. The issue is not gender. HH is helping young women that she knows at the expense of other young women.

    ReplyDelete
  7. What value does "Gekokujo" bring to the discussion that is not brought by revolution, mutiny, or uprising? Is there a shade of nuance that the other words don't convey?

    ReplyDelete
    Replies
    1. I think so, yes. "Revolution" implies a transfer of authority from old to new, whereas "gekokujo" doesn't seek to invert the hierarchy, just punish those at the top of it. "Mutiny" is a bit closer, but is a little too specific to the military (in the coups of the 1930s, young Japanese army officers attacked government and business leaders with abandon). And "uprising" might work, but it is a little more general...you can have an uprising by people who never accepted a hierarchy in the first place, whereas gekokujo connotes being fed up with a hierarchy one had previously accepted...so gekokujo seems like a particular kind of uprising.

      In any case, it's fun to teach Japanese stuff via econ blog posts. :D

      Delete
  8. Anonymous1:38 PM

    "In econ, the standards for what constitutes good research and what constitutes bad research are less clear than in the natural sciences (though far clearer than in the humanities)."

    Or, alternatively, it means more pluralism, given subject matter that does note lend itself as easily as in the natural sciences to lab-like testing.

    I know this is the kind of thing about which you are entirely blind, but occasionally someone needs to point it out. (And please don't deny the physics envy--you have had it pretty bad.)

    ReplyDelete
  9. Anonymous5:42 PM

    Meanwhile, quantitative social scientists from other disciplines are laughing over the delicious irony of a bunch of economists complaining about economists who failed to appropriately cite past work and who claimed novelty where little existed.

    A good 50% of what passes for "new" discoveries in empirical economics these days is just rehashing -- without attribution -- ideas and empirical work done long ago in sociology, psychology, criminology, epidemiology, even anthropology.

    ReplyDelete
  10. I think we ( from the lowly ranked school) should go for 'we are the 99%.'

    ReplyDelete
  11. The EJMR people sorta remind me of GamerGate, well minus the death and rape threats, the doxxing, and swatting. However, since the worst abuses of GamerGate were perpetrated by a "few bad apples", and even without examples of similar antics by members of the EJMR crowd, I think it is safe to treat them the same way I treat those involved with GamerGate, with utter contempt.

    ReplyDelete
    Replies
    1. So basically: a pox on both houses? Remember, the anti-Gamergate crowd also issued multiple bomb threats on their conventions, doxxed them, harrassed at work, etc. And that one of their top "victims" turned out to be a serial liar: http://i.imgur.com/1ADQDgm.jpg

      Not saying there weren't a bunch of assholes on the pro-Gamergate side. But it seems to me to be scum & decent people vs. scum & decent people fighting each other, both with apparent moral claims.

      Delete
  12. Anonymous1:59 AM

    You mischaracterize the scandal (following the retractionwatch article which makes the same mistake). It was never about missed citations per se, which is indeed a minor issue. The issue is that every significant claim to novelty the paper had rested on those omitted citations, and subsequent revisions to the paper by the authors implicitly concede as much. What they're left with is essentially a replication study, with a minor, irrelevant, and methodologically unsound IV specification being reframed as the main contribution because all other claims of novelty were shut down. Given this post-acceptance reframing from answering a novel question to providing a deeply problematic alternative regression specification for a well studied question, the paper should have been re-assessed, but was not.

    It's certainly true that EJMR is filled with alt-right trolls and misogyny, but, as a participant in the threads throughout this scandal, what stood out to me was the relative lack of misogyny--most of the racism and misogyny on EJMR comes from non-economists, and they simply weren't interested in a technical discussion of IV strategies and publication standards (in the last week or so this has changed, since the economists have said their piece on the technical issues and the scandal blowing up with Borjas' comments has made it newsworthy enough to attract the trolls).

    I'd characterize the sentiment behind the scandal as anger about corruption and editorial misconduct from economists--which has been fed by the repeated failure of anyone in the AEA/AER hierarchy to address this issue--alloyed with misogyny that mostly comes from non-economists and has largely been absent from the formal critiques of the the paper and its handling.

    The question I've asked myself is, should we ignore a clear example of the sort of corruption and misconduct that almost certainly happens regularly in the profession and forego the opportunity the disincentivize bad behavior because the well has been poisoned by internet trolls? My answer is no.

    ReplyDelete
    Replies
    1. A result that was already well-known in other fields? A dodgy identification strategy? Published in a top econ journal?

      You don't say. :D

      What would be cool would be if we had some kind of open review system so that instead of posting in lame anonymous forums, smart people could point these things out for AER paper after AER paper.

      Delete
    2. Anonymous10:19 AM

      So "There are other people who do shady stuff, so let's do nothing in this case because the co-editor and authors are women." This is granite solid ethics. So I guess we just have to wait for this to happen when everyone involved is male?

      And for all the excoriation of Borjas that's happening now on Twitter, still no one has denied there was impropriety on the part of the AER editorial board.

      And I would be delighted to raise my concerns in a non-anonymous, sexism-free forum of some kind. But I'm young, not tenured, and in a profession where "who you know" is by far the most important thing, I can't afford getting thrown under the bus like Borjas can.

      Delete
    3. Anonymous10:56 AM

      Honestly, I'm surprised nobody has written up this story from the "economic imperialism" angle.

      It's pretty much being defended in terms of, "Sure, those health people have studied this issue, but they don't know how to identify causal effects properly! We had to wait for an economist to do it right!"

      Plus there is the general problem with cross-disciplinary research, which is that the authors and (probably) referees are all economists.

      Delete
    4. So "There are other people who do shady stuff, so let's do nothing in this case because the co-editor and authors are women." This is granite solid ethics. So I guess we just have to wait for this to happen when everyone involved is male?

      No, it's less about the sexism and more about singling someone out. If this kind of thing is near-universal (as some people tell me it is), the right approach is probably not to create lynch mobs that go after one or two culprits, but instead to work to change the system itself. Going after 1 or 2 people when 85% are guilty just automatically looks biased and personal, whether it's sexism or some other thing driving the personal enmity. Does that make sense?

      And I would be delighted to raise my concerns in a non-anonymous, sexism-free forum of some kind. But I'm young, not tenured, and in a profession where "who you know" is by far the most important thing, I can't afford getting thrown under the bus like Borjas can.

      Yeah, good point. There needs to be some better way for young people to complain about this kind of stuff (since all the old influential people are probably vested interests). One way is for people like me to start doing some investigative journalism, and for people to pass me anonymous tips...not sure if I'm ready to do that yet, but maybe. Another way is to get a pseudonymous blog (and be very careful not to get doxxed til you have tenure).

      Delete
    5. Anonymous4:49 PM

      Noah, your arguments normally make sense, but this is extremely sloppy. If you say this sort of stuff is widespread, please give examples of other papers with similar omissions. Otherwise you are just making stuff up without any evidence as far as I can see. I work in an empirical field and whilst there are certain authors that tend to bend the rules (Acemoglu, for one), most of the research I have seen in top journals is solid and at the very least does not suffer from the very obvious problems of the parental stress paper discussed here.

      Delete
  13. The AEA established a committee a number of years ago to look into the question whether journals were an institutional monopoly. As with their previous committee investigating the question of whether graduate education was training economists to be idiot savants, nothing changed.

    ReplyDelete
  14. Noah,

    As you know I am denounced from time to time on ejmr and as a result keep an eye cocked on it, although I only recently paid any attention to this scandal. I think you are right that quite a few complaints about the paper and the process surrounding it are overblown and involve matters that are not all that uncommon. It seems to me that even though some think the IV strategy was dodgy, there are really only two serious issues regarding the ethics or practices involved here, one explicitly mentioned and one not here.

    The explicitly mentioned one involves data access. According to Retraction Watch and some other sources, that should not have been a major issue for this paper. The data has been publicly available from the Swedes for some time. Indeed, this is part of the claim that the paper did not do much new because so many other researchers have already worked on this data source, finding most of the results reported in the paper, if not with its use of an IV strategy, which seems to be the current defense of the paper as contributing something new in the end.

    There certainly is a more general issue of data accessibility and people in privileged locations with high level supporters getting early access to such. But that does not seem to be the case here for this scandal.

    The other issue, obliquely hinted at but not explicitly stated has to do with the allegation that Hillary Hoynes is a coauthor of one of the coauthors of the paper in question. I do not know if this is true or not, but I have not seen anybody specifically deny or refute that claim. Supposedly the AER has rules against people with such close relations handling papers of each other, and I know some members of the AER board who are not happy about this particular aspect of the whole thing, although indeed a lot of these people do not want to comment too publicly. But this has been beaten on a lot in the ejmr commentary, and if it is true, then there is some substance to the complaints.

    This is a complicated matter and not all that simple. It is most certainly the case that journal editors and board members at many journals handle papers by their students and coauthors. It happens all the time, although this tends to be more common at lower ranked journals. As an editor, I have published and thus been ultimately responsible for "handling" papers by people I coauthored with or who were former students of mine, although I have always tried to "play it straight" with real refereeing processes and all that, and have in fact rejected plenty of papers by such people over the years. But it would be ridiculous to say that a journal could not publish any papers by anybody associated with anybody on the editorial board of the journal.

    So this is a tricky business. But I can see that especially at the very top journals, which are quite bureaucratic in contrast to many lower level journals where strong editors pretty much run the show, it is reasonable to make efforts to avoid even the appearance of favoritism. Given how important for careers publishing in those top journals is, it is understandable that many are charging corruption when they see this sort of thing going on in those journals. This would seem to be the most serious fire underlying all the smoke that has been generated in this scandal, for whatever it is worth.

    Barkley Rosser

    ReplyDelete
    Replies
    1. Anonymous3:21 PM

      They use confidential data that is not publically available (in that you can't just go to a site and download the data). Getting access to the data requires an application and approval. Also, a member of the research team has to be Swedish. This data is difficult, but not impossible, to obtain. So, not "publically available" as most people would use the term.

      Delete
    2. Pseudowoodo7:49 PM

      No, Barkley, the criticism is that they did the exact same thing as a prior paper. Not that the data set had been mined to death, that their method was literally the exact same. There's no excuse to misrepresent the facts when they're all over the internet.

      And if you think the Hoynes issue has been "obliquely hinted at" then you didn't even read the Borjas post.

      Delete
    3. P.,

      Well, then the criticism is wrong. The new paper has the IV strategy and the earlier one does not, even if you think the IV strategy is a pile of baloney.

      Regarding the "obliquely hinted at," I was referring to Noah's remarks and the comments here prior to mine. Heck it is not just Borjas who made the link, but the also cited ejmr thread(s).

      As they say on ejmr, you seem kind of stoopid, P.

      JBR

      Delete
  15. Anonymous4:26 PM

    'What if this behavior is universal--then this is a witchhunt'

    Really? Are we grading corruption on a curve?

    If you get caught stealing cookies from the jar, you are a thief. Full stop.

    ReplyDelete
  16. Anonymous4:48 PM

    why on earth do you economist think that you are special when it comes to academic sleaze ?

    geez....
    I am one of thousands of molecular biologists who has a strong, strong suspician that peer review was delayed so someone else could get out a paper ahead of me...

    My postdoc advisor told me this: as grad student, he made a very important discovery (it was) he sent in the MS, and after acceptance, asked his boss (nobel winner) can I go present at a meeting - the paper will be published in two weeks and no one can scoop us (this is back when it took two weeks to get revisions back from teh typist on carbon paper)
    He goes to the mtg, someone sees the data, and manages to do the experiment, write the MS and get it published first...

    I have been told, this is 2nd hand, that L Pauling, one of the most famous chemists of the 20th century, would add an author if the paper had only two names.
    The reason is, with two names, the citation would be Pauling and X; with three names it would be Pauling et al

    That you think economics is esp sleazy is yet another sign of economics hubris !!!

    ReplyDelete
  17. Anonymous4:56 PM

    Dietz Vollrath: "Not happy with a former co-author acting as editor - AER has plenty of other options"

    https://twitter.com/DietzVollrath/status/740952298122149890

    Vollrath has no disdain for female economists. He also lacks disdain for political liberalism (cf https://growthecon.com/blog/why-i-care-about-inequality/).

    There are people who are genuinely disappointed by this fiasco. No disdain for any subgroup. No ulterior motive. No tribalism.

    ReplyDelete
  18. Of course Noah, in doing his best Krugman impression, would immediately assume that the underlying issue is sexism, racism, or some sort of personal insecurity of the "right." Whenever antagonists are known to be conservative, the first thought always seems to be that they must be bigots of some sort. Only in certain cases, after intense scrutiny fails to finds some angle through which the entire issue can't be cast off as racism or sexism, do the issues at hand ever get examined. Noah writes that this notion was just a thought he had but it seemed "interesting," which is complete garbage. It's a thought he had because that's the first lens through which liberals have increasing been viewing every issue brought up by someone on the right. It's not interesting or novel. It's ad hominem.

    ReplyDelete
  19. Probably worth noting, as well, that most of the terrible work that doesn't get called out is written by men.

    ReplyDelete
    Replies
    1. Anonymous4:31 PM

      Really? Guys don't get called out? Wrong.
      Bruno Frey
      Emil Siriwardane
      Oberholzer-Gee
      Lemmon, Roberts, and Zender
      Fu
      Gompers, Ishii, and Metrick

      Then there are the male+female authored articles called into question:
      Reinhart and Rogoff
      Easley, Lopez de Prado, and O'Hara
      Ané and Geman

      If you are going to claim something, it should at least comport with the facts. EJMR is rude and the comments are sexist, but most of the papers criticized have been by men. Sure, lots of *threads* complained about Oster and Ostrom and other women, but substantive critiques of a particular paper? Definitely not focused exclusively nor even mostly on women.

      As for most of the terrible work that doesn't get called out... sheesh, there's too much terrible work to call it all out. And since the field is still majority male, even random sampling would leave more un-called-out work by males. You need a better benchmark to claim sexism.

      Delete
  20. I know how much you make and I know how much almost everyone else makes for a living if they work for a company. I have connections. And this information is NOT protected. I have every right to know it if I have the right connections.

    I know how much you make. I know how much most people make.

    Economists make descent money generally, but they're totally subservient to relatively dumb business people. Most successful business people are relatively dumb, but they have a combination of mediocre intelligence, suppressed consciences, oblivious nature and above all, arrogance.

    ReplyDelete
  21. Anonymous4:04 PM

    As a woman who is doing her bit to counter sexism in academia, I am saddened to see such an important agenda being leveraged to cover up a story of academic dishonesty and privilege. The crucial facts are these:

    - The paper that predates Persson and Slater's should have been properly and promptly acknowledged. It was not. Even now, the footnotes have a rather grudging quality to them.

    -The paper should NEVER have been assigned to Hoynes as per the rules of the journal.

    -When once the scandal broke, the journal should have acted swiftly to reassure the larger academic community that at the very least it takes its own guidelines to maintain academic integrity seriously.

    - Some economists have responded to the situation by saying that the allegations come from "nobody important." Arguments from authority have no place in academia.


    If Slater, Persson, and Hoynes authors are making the argument that men are getting away with the same sort of thing every day, they may be correct, but this is not the time or place to make that argument, and especially not in order to excuse oneself.


    Yes, people posting on the EJMR make horrendously sexist remarks and have a disgusting, puerile sense of humor. But that is something to address at a different time and in a different way. It is certainly not a point that should be made to avoid addressing the other issues at hand.

    Shame on everyone who is using sexism to get out of addressing the stickier issues of problems in the profession. This whole thing makes the profession look very bad. It is so short-sighted to not address this swiftly on the crucial facts and to instead bring sexism into the picture. It is doing a disservice to 2 causes at once: gender equality and academic honesty.

    I am sorry, Noah, and disappointed that you chose to write about it this way.

    ReplyDelete