Cai, Lontzek and I submitted our paper on the social cost of carbon in November, 2012, and received referee reports and a revision request from Monika Piazessi, the JPE editor handling the submission, in November, 2013. A couple of months later, we met with her to discuss the revision. In particular, I wanted to explain to her how we planned to include validation and uncertainty quantification methods into our paper, reflecting the focus of the RDCEP (http://www.rdcep.org) project which had supported this research. Shortly after I began my points, she cut me off and said:
“Ken, it’s nice to write papers for smart people, but that is not what the readers and referees of JPE want”.
It was a surprisingly clear statement by a journal editor but not unusual in my experience. I wrote it down as soon as I had pen and paper. If that was the only comment she made on this matter, we would be left with a “he said, she said” situation. However, in her editor’s letter to us in July in response to our revision, she wrote
“You really need to work to make this paper readable for the general audience of the JPE. That audience includes IO economists, theorists, labor economists, etc. I warmly recommend asking a second year PhD student (or, in fact, several students) for comments”
I must first say that I do not equate second-year PhD students to people outside the set of “smart people”, nor do I interpret Piazessi’s comments in that way. I interpret her use of the phrase “smart people” to refer to people who have the technical training to understand the methods in our paper. Her reference to second year PhD students indicated that they should be the target audience, and that I need to write the paper in a manner that recognizes their limited expertise in computational and applied math.
I have two points to make.
First, the idea that a paper in a leading journal should be written so that second year graduate students can understand it is a truly alien concept to me. I was first a graduate student in mathematics. We had to complete two years of courses before we could take the qualifying exams. Limiting a paper’s readership to second-year graduate students is unthinkable in mathematics, as it is in many sciences.
Second, the ability of second-year graduate students in economics is endogenous, depending on the training that they are offered. In fact, the same can be said of all readers and referees of JPE. Piazessi’s comment is really a comment on most current PhD programs, very few of which offer serious training in applying numerical analysis methods to economics. There is some effort to teach economists the methods used in their specific field, but Piazessi’s mandate was that we write a paper that could be read by a general audience that lacked training in the methods we used. Very few departments have ever attempted to give students training that would, for example, allow empirical IO people to understand the methods used in macroeconomics. This makes it difficult to write an economics paper using serious computational methods for a general audience. Let’s consider the history of recent decisions made by some of these universities.
Over the past decade, Karl Schmedders and I have taught computational courses at the University of Zurich. In 2020, the University of Zurich asked me to teach a course via Zoom. I have gathered the teaching materials, including the lectures, at https://kenjudd.org/teaching/university-of-zurich-2020/. We, along with Felix Kubler, have advised many graduate students and postdocs at UZH who used this training in their work. In fact, Thomas Lontzek was a postdoc at UZH during this time.
Penn State University is another example of a department that works to give its students computational skills. For at least ten years, it has a year-long computational course and I visit for two weeks each year.
UZH and Penn State are unusual. I have taught computational methods to economists in a variety of ways, including PhD courses I developed. However, some universities have become hostile to such training. In 2017, the Booth School asked me to teach my course there. All the details were set and a student told me that the course was even listed on the Fall, 2017, schedule. Provost Daniel Diermeier vetoed that. A few years ago, I proposed to the Stanford Economics department that I teach my course together with Lilia Maliar, who had taught it on her own. Under my proposal, I would provide any TA support and all the salary money would go to Lilia. When one cuts through the bureaucratic nonsense, I was told that Stanford (a decabillionaire) would agree to my proposal only if I gave it $7,000. I rejected this “offer”.
My paper with Cai and Lontzek was important for their careers. Therefore, I felt obligated to do what was necessary for that paper to appear in a top five journal. When I wrote the subsequent revisions and the published version, I did not compromise the technical aspects of our paper. I predict that many of my future papers will also be greeted in the same hostile manner. That is fine with me but young people will want to avoid such combat. I enjoy working with young people, but I will have to limit those collaborations.
Education is the answer to the shortage of “smart people” in economics. I have been saying this for over thirty years, and I will continue to say this. Perhaps I will some day find allies in this effort.