The value (or lack thereof) of low-ranked journals in economics

Last week I posted about the emphasis in economics on publishing in the 'Top Five' journals. Not everyone can publish there, obviously, due to space constraints. And even top economists can't publish every paper in the Top Five. So, what is the relative value of lower-ranked journals?

In a 2017 IZA Discussion Paper, Nattavudh Powdthavee (Warwick Business School), Yohanes Riyanto (Nanyang Technological University), and Jack Knetsch (Simon Fraser University) take that question a step further. They ask whether publishing in lower ranked journals might have negative value. That is, publishing in low-ranked journals might lower other economists' perceptions of the quality of an academic CV.

This idea harks back to a 1998 paper by Christopher Hsee (ungated version here), which showed that (quote is from Powdthavee et al.):

...people shown a set of dinnerware having 24 pieces in good condition, were willing to pay significantly more for these than another group of people were willing to pay for a set that contained 28 pieces in good condition but with another 11 that were broken.

Using survey data from 378 economists from economics departments in the top ten percent worldwide, Powdthavee et al. find something similar:

...it appears likely that the inclusion of lower ranked journals on an individual’s publication list will have a negative impact on the assessment of such lists by other economists. We found statistically significant differences between the higher average 1 to 10 rating that respondents gave to both lists having only eight higher ranked journals, and the lower average rating that other subsamples gave to lists containing all of the same eight higher ranked journals plus six more lower ranked ones.

Specifically, adding lower-ranked journals to a list that included two Top Five journals lowered the average rating from 8.1 out of 10 to 7.6 out of 10. Similarly, adding lower-ranked journals to a list that didn't include any Top Five journals lowered the average rating from 7.0 to 6.3 out of 10. However, it's not all bad news. When CVs including and excluding the lower-ranked journals were rated jointly (that is, the survey respondents saw both CVs), the effect went away. So, when CVs are being compared, the negative effect disappears. However, Powdthavee et al. note that:

There are, of course, occasions in which is it the results of joint valuations that will matter to final outcomes. Perhaps most easily imagined are comparisons between candidates for a position or honour – Candidate X vs. Candidate Y. But most others, such as those involving promotion, tenure, and selection of consultants and other experts, seem to be ones more likely to turn on results of single valuations. Further, even in cases of Candidates X and Y competition over a position, it is largely the results of single valuations that determine whether a person becomes a Candidate X or a Candidate Y.

It makes me wonder though, how publications outside of economics journals affect the perception of a CV. As someone with an appreciable number of publications outside of economics, of course I have a vested interest in knowing the answer to that question.

Read more:

• The emphasis on publishing in the top five economics journals

The gender gap and question time in academic seminars

The sources of the gender gap in STEM (and economics) are probably many and varied. However, one potential source that has been highlighted is the lack of role models (see my post on the topic here). If female senior students don't have female academic role models, then they be less likely to pursue an academic career, and if female junior faculty don't have female senior faculty role models, they may be less likely to succeed.

A recent paper by Alecia Carter (University of Cambridge), Alyssa Croft (University of Arizona), Dieter Lukas (University of Cambridge) and Gillian Sandstrom (University of Essex) looks at one example of role modelling - the asking of questions in academic seminars (with a nice non-technical summary on the LSE Impact Blog). Specifically, they collected survey data from 509 academics (mostly in biology or psychology), as well as observational data from "247 seminars, from 42 departments of 35 institutions in 10 countries".

They found that:

...a given question after a departmental seminar was more than 2.5 times more likely to be asked by a male than a female audience member, significantly misrepresenting the gender-ratio of the audience which was, on average, equal.

So, even after controlling for the gender make-up of the audience, and the hosting institution, and the gender of the speaker, men were more than twice as likely to ask questions as women. They conclude that:

[i]n the case of academic seminars, then, the fact that our data show women asking disproportionately fewer questions than men necessarily means that junior scholars are encountering fewer visible female role models in the field.

An interesting aspect of the paper is that they tried an intervention to increase the number of questions from female audience members. They first noted that there were more questions (and more questions from women) when the time allowed for questions was longer. They then manipulated some seminar lengths to allow for more time for questions. However, the intervention had no effect on the number of questions asked by female audience members.

Having read the paper though, there was an alternative intervention they could have tried. The number of questions from female audience members also appears to be higher when a woman asks the first question (even if you exclude the first question from the analysis). So, a simpler intervention would have been to ask the seminar moderator or chair to ensure that a female audience member asks the first question (if possible).

I've just gotten back from a conference in the U.S. I can't say I noticed a gender disparity in question asking there, but then again I wasn't looking. I'll definitely be paying more attention during departmental seminars and conference sessions in future.

[HT: eSocSci]

Book review: On the Third Hand

In the 19th Century, Thomas Carlyle dubbed economics "the dismal science". Unfortunately, the reputation of economics for being humourless has little improved since then. But that reputation is undeserved. Caroline Postelle Clotfelder's 1996 book, On the Third Hand: Humor in the Dismal Science, an Anthology, collects the best examples of economics humour across the time since Adam Smith (and up to the mid-1990s, obviously).

Most of the book is comprised of excerpts, and a lot of it is satire. Some of it has not aged well, and without better understanding the context of the time, I'm afraid the humour was lost on me. Other parts remain hilarious, such as Alan Blinder on the economics of brushing teeth (which can also be found here) and Arthur Levine on the benefits of government-sponsored sex. George Stigler's arguments against textbook publishers constantly bringing out new editions and the associated costs to society remain relevant today.

Some particularly famous pieces make it into the book. On such is the "Petition of the manufacturers of candles, wax lights, lamps, candlesticks, street lamps, snuffers, extinguishers, and of the producers of oil, tallow, rosin, alcohol, and, generally, of everything connected with lighting" (by Federic Bastiat, available online here), where the petitioners argue that they should be protected from "the intolerable competition" from the sun. Another, though some argument could be made as to whether it counts as economics, is Jonathan Swift's "A modest proposal" (available online here).

A particular highlight for me was Joan Beck's suggestions on how to reduce healthcare costs, which included do-it-yourself options for patients, end-of-year closeout sales on discontinued therapies, frequent-flyer-like programs where patients accumulate points, seasonal specials, his-and-her operations, family rates, and so on.

The anthology is a great collection of the humour and wit of economists past. Before life gets too dismal, it might be worth tracking down a copy of the book.

Intergenerational transmission of gender bias and the gender gap in STEM

More male students than female students choose to study STEM subjects (and economics) at university. In a recent post, I discussed two papers that could be interpreted as suggesting comparative advantage as an explanation. That is:

...even though female students may be better than male students in STEM subjects at school, we may see fewer of them studying those subjects at university (let alone taking them as a career), because female students are also better in non-STEM subjects at school, and they are better by more in non-STEM than in STEM, compared with male students. Economists refer to this as the students following their comparative advantage. Female students have a comparative advantage in non-STEM, and male students have a comparative advantage in STEM subjects.

However, there are other competing explanations. Students' choices are affected by their parents and by their peers. So, if their parents hold biased views about whether studying a STEM subject is suitable for female students (or if their peers do), then female students may be less likely to study STEM subjects. A recent working paper by Alex Eble (Columbia University) and Feng Hu (University of Science and Technology Beijing) looks specifically at the intergenerational transmission of bias (from parents to children).

Eble and Hu use data from 8,912 Chinese students in 215 seventh and ninth grade classrooms, who were routinely randomly assigned to classrooms. That means that each student's peer group (within their classroom) was randomly assigned. The key variables that Eble and Hu are interested in are the bias in the views of each student's own parents, and similar bias in the views of each student's peers' parents. They then look at how those two biases are related to students' own bias, aspirations, and academic performance. They find that:

...a one standard deviation increase in peer parents’ bias causes a 4.2 percentage point (8%) increase in the likelihood that a child will hold the bias... This transmission of bias appears to occur for both girls and boys in roughly the same manner... we calculate that going from roughly 25 percent of peers’ parents being biased to 75 percent of peers’ parents being biased generates an 18.9 percentage point (34%) change in the likelihood that a child will also hold that bias...

Children whose parents hold anti-girl bias are 29 percentage points (52%) more likely to also hold that bias, and again the transmission appears to hold equally for boys and for girls...

Ok, so there is intergenerational transmission of bias. How does that affect the gender gap? Eble and Hu find that:

...an increase in peer parent bias increases the gender gap... in girls’ perceived difficulty of math relative to boys’ by two percentage points, or 28 percent of the 7.2 percentage point gap between boys and girls in this variable. This pattern also holds for own parents’ bias, and the estimates are again more stark: our estimated coefficient of own parent’s bias on the gender gap is a 15.0 percentage point increase in the likelihood that a girl perceives math to be difficult, relative to the likelihood for boys. For boys, own parent’s bias is associated with a 6.1 percentage point decrease in the likelihood the child will perceive math to be difficult; the “total effect” for girls is an 8.9 percentage point increase in this likelihood...

As with perceived difficulty, our estimates of the effect of peer parents’ bias on boys’ and girls’ test scores, respectively, differ in sign. Boys appear to gain slightly (a statistically insignificant 0.05 SD increase) from a one standard deviation increase in the proportion of peers whose parents believe that boys are superior to girls in learning math. For girls, on the other hand, we estimate that a one SD increase in peer parent bias increases the gender gap - that is, reduces girls’ performance relative to boys’ - by a statistically significant 0.063 SD...

Here again the correlation between own parent’s bias and performance is much larger in magnitude - the scores of boys whose parents believe that boys are better than girls at learning math are 0.16 SD higher than for boys whose parents do not believe this, and for girls, having a parent who holds this bias pushes the child’s test score down, relative to boys’ scores, by 0.28 SD...

In other words, they find evidence that the gender gap in perceived difficulty of maths and in maths test scores can be partially explained by own parents' bias and peer parents' bias. Unsurprisingly, the effect of own parents' bias is the larger of the two. Interestingly, they also find that girl peers' parents have a bigger effect on girls than boy peers' parents do. However, in amongst the results they also find a potential partial solution, at least in terms of peer parents' bias:

These negative effects, however, decrease with the number of friends the child has in her class; a child with five close friends in her class appears to be entirely immune to the negative effects of peer parent bias that we have shown throughout this study...

So, it's not all bad news. If you want to give your daughter the best chance for success in STEM, start by minimising the bias in your own views. If you're worried about the biased views of other parents, to the extent you can you should make sure your daughter has many close friends in her class.

[HT: Development Impact]