Economic Logic

This is a bit late for the current job market for Economics PhDs, but say you have to choose among several job offers (lucky you). The departments are are all of equal prestige and working conditions, salary, and geographic environment are all comparable. The only difference is the size of the faculty. What offer should you take if you care about your future research output?

Clément Bosquet and Pierre-Philippe Combes say you should go for the uniformly good, more field diverse, and larger department. At least this applies to French academic economists. Even more interesting, they observe that department characteristics are as important as researcher characteristics. The initial placement of a researcher thus matters a lot for her future, and this may explain why there are so few success stories after an initially poor placement, at least in Economics. Of course, the academic market in France is very different from any other country, so I am not sure the results can generalize, but this is a start.

Think tanks have a horrible reputation everywhere by in the media. The reputation is because they are often very biased and sell out to their funders. The media is because think tank staff are willing to provide the expected sound bites to journalists, no matter what the topic. All this would be OK if think tanks were good at conducting independent research. It turns out the most prominent ones, do not do much of it on the hot topics they talk about, according to Dan Farber, who finds that they do not publish much of relevance (and this is not even considering peer reviewed research).

Interestingly, the picture is very different with respect to consulting. Looking at academics across all fields from five Spanish niversities, Pablo D'Este, Francesco Rentocchini, Liney Manjarrés-Henrìquez and Rosa Grimaldi find that getting grant money is positively associated with getting consulting contracts. In other words, good researchers also get consulting gigs. And in some fields, consulting is where the financial rewards of research really lie, especially in social sciences where grants are usually relatively small.

Whether you like it or not, Elsevier matters in the dissemination of research in Economics. By far the largest player in the field, it enjoys considerable market power (and a profit margin around 30% that comes with it). And even though journals are not at the frontier of research in Economics, it still matters what happens at Elsevier because it controls so many of the top field journals.

According to its web page on the global dissemination of research, Elsevier states:

We recognise that access to quality research is vital to the scientific community and beyond. For us this means providing support and the latest tools to maintain the quality and integrity of published scientific literature, achieving the widest dissemination of content, and embracing the opportunities of open access. We will continue to identify access gaps, and work towards ensuring that everyone has access to quality scientific content anytime, anywhere.

These are all nice words, but this is not all what Elsevier practices. First of all, all of the Economics content of Elsevier is gated, and academic libraries have to pay through the nose to let faculty access the content, including their own works. Even errata and retraction notices are gated. There is no open access journal in Economics, and even in other fields where it is available, the cost is prohibitive (usually US$3000, even more with color charges!), which cannot be justified in any reasonable way by hosting costs. Indeed, Elsevier spends considerable resources trying to keep potential readers away, by gating the material for the general public and making it difficult for individuals to buy subscriptions, especially for hard copies. All this management of subscriptions and filtering of web traffic would disappear with open access, making it much cheaper, not more expensive.

But this is not an issue only with Elsevier (Springer is much worse in this respect). Elsevier, with its market power is trying to kill any competition and any initiative that tries to open up the dissemination of research. For example, it was a huge backer of the Research Works Act in the US, which would have prohibited mandates that publicly funded research should be available in open-access repositories. Of course this generated a huge outcry from the scientific community (you know, the one that Elsevier claims to serve) and lead to a call for a boycott. This seems to have been successful, as Elsevier reversed its stance, thereby killing the bill.

Unfortunately, few economists seem to have participated in the boycott, which is probably why Elsevier continues to flaunt the research community with no remorse. For example, it has not updated the listings of its journals for over a year in RePEc, and still vigorously refuses to let RePEc perform citation analysis on its contents. Repeated attempts to get a reaction from Elsevier have unsuccessful from my part. My suspicion is that RePEc is threatening some of the products that Elsevier is pushing (Sciverse, Scopus), and the interest of the research community becomes second fiddle. From what hear, people are deserting the Economics desk at Elsevier, starting with its head, which makes you wonder who is in charge of "the widest dissemination of content."

To understand further what a fine business Elsevier is, here are some of my previous posts:
The evil empire strikes again
The evil empire strikes again (II)
Copyright and the lack of competition in academic publishing
Why I am boycotting Elsevier

Scientific publication is not the same as it was, now that we can easily access the literature over the Internet. No more trips to the library, much fewer waits for interlibrary loans, and no more chasing who took or misplaced the volume in the racks. But did all this change anything in the way we publish our results?

This is what Timo Boppart and Kevin E. Staub study by looking at the diversity of topics covered in journals and how the availability of on-line publication would have changed that. The idea is that on-line publication allows to discover and read more material, and one may in particular stray away from the usual topics. No doubt about that. But I wonder why Boppart and Straub have this focus on journals. After all, working papers is where its at in Economics, and journal readership has not really increased, I believe. The treatment variable is the share of cited articles available on-line the year before publication. This seems so wrong. There is no way it takes only one year from the literature search to the print issue. Not in Economics, where I would say it is a minimum three years, with really rare cases below that. In fact, a good share of mine took more than a year from final acceptance to actual publication. Then, what about working papers? This is what people read, not articles.

If you are an economist and maintain a blog, the best way to get an audience is to get into fights with pundits and journalists, and you then easily become a pundit yourself, losing the impartiality you are supposed to have as a scientist. The most popular economist bloggers are either openly libertarian or quite far right, or in reaction resolutely on the left. Almost all engage in politicking and wars of words and have basically abandoned the principles of impartiality their scientific upbringing taught them. The impartial scientists are drowned.

EconAcademics.org tries to rectify that, aggregating blog posts that discuss research, or at least refer to research. The list of monitored blogs is impressively long, yet it saddens me to see that the "popular" blogs are nowhere to be seen near the top of those who have discussed the most research, despite their considerable volume of posts. While I could be proud to be (far) ahead on this last list, it saddens me again how little consistent discussion of research there is. Out in the blogosphere, are we really so few economists doing this? Why can't the top economics blogs relate more to the results of their field?

A positive externality of EconAcademics.org is that is compiles a list of all papers I discussed, including where the ultimately got published. I have put the link in the sidebar and will come back to discuss it sometime soon.

A little less than a year ago, a controversy erupted about the publishing practices of Bruno Frey and his students. Indeed, they tend to repackage their research and submit it to multiple journals simultaneously (or sometimes successively), without cross-references and without alerting editors to this. This is in clear violation of the submission conditions of most academic journals and even goes against principles Bruno Frey has himself advocated in multiple (of course) publications: there is not enough space for everyone to publish on the one hand, and the pressure to publish leads people to (self-)plagiarize on the other hand. On his homepage, Bruno Frey crows about over 500 or 600 publications, depending on where you look, numbers that are completely surreal for any self-respecting academic economist.

The scheme blew in his face when some editors and some blogs started raising questions when very similar articles about the Titanic, with Benno Torgler and David Savage, appeared in four journals (some say there is even a fifth one in German, but I cannot verify). And the article was not even original, as a similar analysis was done and published 25 years prior and is now standard reading and exercise in statistics courses. Newspapers picked up the story, Frey went into denial but finally confessed to the editor of the Journal of Economic Perspectives, who published correspondence about the case and publicly admonished him for multiple submissions (he do not yet know about the prior literature). But that is only for this case, there are all the other ones. The University of Zurich, from where Bruno Frey recently retired, promised an investigation. That was sometime in the Summer. Since then, nothing.

One could suspect the University would do nothing, as Bruno Frey is the best ranked economist in German-speaking universities. And the prolonged silence clearly seemed to corroborate this. But rumors started circulating in the hallways, rumors that were not encouraging at all. But no evidence from Zurich.

Finally, I got good evidence from a reliable source. And it is indeed not encouraging. The University of Zurich mandated three prominent academics to look into the case. But the mandate was formulated in such a way that only the articles about the Titanic could be analyzed. The experts came to the obvious conclusion that unethical behavior was at play for this case. They could not mention the others and thus the University concluded that this was a one off miss-step. The University gave Frey a verbal admonishment, which does not go on his record, and did not release the report.

But this was not an one-off miss-step. Frey has been banned from the editorial board of Public Choice for a similar case of re-publication. He is by now banned from publishing in at least a half-dozen journals. To make matters worse, he has himself advocated to go against plagiarizers and others that unnecessarily take up valuable publication space. The investigation should have looked at his whole career, like when a scientist is suspected of fabricating data and all his publications are subject to scrutiny. And it is not like the information would be difficult to obtain, it is readily available and people have even compiled it, as documented in the FreyPlag Wiki.

For more about the case, you can read my past blog posts: 30 April 2011, 3 September 2011, 27 September 2011. Also, Olaf Storbeck's Economics Intelligence blog was the one that convinced the University of Zurich to finally (pretend to) act: 4 July 2011, 4 July 2011, 5 July 2011, 6 July 2011, 7 July 2011, 9 July 2011, 20 August 2011, 29 August 2011, 12 September 2011

A frighteningly low proportion of patents are granted to women, 7.5% in the United States, while only 5.5% of those that are commercialized are from women. Maybe they are of a more generous nature and realize how progress-crippling patenting can be these days. But I doubt this effect can be that strong. The elephant in the room is of course the low proportion of females in sciences and engineering professions. It cannot be the only explanation, because females have a higher share than 7.5%, but it is a start.

Jennifer Hunt, Jean-Philippe Garant, Hannah Herman and David Munroe claim the missing women has little to do with their proportion in the science and engineering fields. Rather, the culprit is the lack of women in science and engineering jobs that involve development and design, in particular electrical and mechanical engineering. What pushes women out? Lack of interest, abilities or discrimination? The paper is silent on this (except for hurdles in the promotion process) but ventures to say that correcting this would increase GDP by 2.7%. This number is based on the idea that if more women were working in those fields, there would be proportionally more patents, and GDP would be proportionally higher. I do not think this is that simple.

A large part of economic research is devoted to empirical studies, and the name of the game there is statistical significance. Once you get an interesting effect being significant, it becomes a study worth writing, and possibly publishing. If you cannot find an effect, then try another specification until it is statistically significant. Nobody will know how much you tried, the one that is significant is all that counts. If one cannot find a statistically significant result, a null result, publishing it becomes really difficult. This game of finding statistical significance is unfortunately misleading, as this hunt dominates theory or even common sense when choosing specifications, and often completely neglects economic significance. What if a statistically significant results is tiny even though it is precise? And what about a large effect that is statistically weak?

I am certainly guilty as well of confusing statistical and economic significance, including on this blog. Indeed, it is often difficult even understanding what the size of the effect is, because the specification does not allow one to relate the effect to something tangible.

The reason I am mentioning all this is that I came across a recent paper by Walter Krämer, who tries to revise the argument made by Stephen Ziliak and Deirdre McCloskey that any statistical significance is useless. While he seems to concur on the general abuse of statistical significance, he claims it can still be useful under some circumstances. And these are when you do exploratory testing, as it allows to discard unviable hypotheses or specifications. But one has to remember, even though this is elementary statistics yet so often ignored, one can only reject or not reject, but never accept a hypothesis. So even the exploratory testing has its limitations.