Development Impact: JDE now has you as the editor plus eight co-editors. How do you assign papers and coordinate with so many co-editors? Also, how are the co-editors are appointed?
Maitreesh Ghatak: There has been a steady increase in submissions since the current team started its term on October 2009. Last year the JDE received 1091 papers. Figures for the previous four years are : 918 (2010), 844(2009), 804(2008), 800(2007). Implicit in your question is a premise that having more co-editors is costly. However, the greater the number of co-editors who are differentiated by fields, the better we can match papers with the expertise of the co-editor. Moreover, given the level of submissions mentioned above this keeps the number of papers handled per co-editor to manageable proportions. For co-editors this is a major concern and this played a major role in the decision to expand the set of co-editors, especially since we have had a significant increase in submissions in recent years. These are the benefits. As your question mentions, this is not without cost. However, with the internet based editorial management system, my work as editor-in-chief is not significantly affected in terms of paper assignment as to how many co-editors there are, other than making sure everyone is getting a fair share of the submissions, and there is reasonable matching between the topic and the expertise of the co-editor. Yes, coordinating with several co-editors is costly, on top of coordinating with the publisher (Elsevier) and this is where my job involves a fair bit of work. However, we have an excellent group of co-editors in terms of collegiality and commitment, and I rely on my colleagues a lot for advice and suggestions. Also, we have a fairly decentralized system and once a paper is assigned to a co-editor, he/she is completely in charge of that paper and this too keeps coordination costs low. The choice of a co-editor is very similar to other journals I have edited (e.g., the Review of Economic Studies): names are solicited from a broad range of experts in a field and then discussed among the editors.
DI: It seems that the JDE is slow in getting papers out compared to many journals. In the latest issue (March 2012), the submission date for the first four papers range from November 2006 to December 2008, while they got accepted and became available online between March and May 2011. Some of this is authors not revising papers quickly, but even after acceptance, there is almost a year gap before publication. What, if anything, is the JDE doing to speed up the processing times as well as dealing with the apparent backlog?
MG: I startedas editor-in-chief in October 2009 and the current co-editors started at the same time or later, and so I cannot really comment on the processing time of papers that were handled by previous co-editors – which apply to the cases that you mention in the question. In general, there is a trade-off between quality of decision making (which includes inputs by the referees) and processing time. It is like the standard cost-quality trade off that you see in many other contexts. Delays happen most of the time (but not always) because of difficulty of getting referee reports in time. Good referees are in high demand and so they are often over-committed. This means they ask for more time, or they turn down our request and we have to find another referee. Also, it takes editors time to take decision on papers were the referee reports are conflicting. There are mechanical ways of improving turnaround times – for example, increasing the rate of desk rejections, but because turnaround is more visible and quality of editorial and refereeing inputs are less visible, focusing only on turnaround time is like focusing only on cost and not on quality. Having said that, all of us in the JDE editorial board are also authors (other than being editors) and we all have suffered from unreasonable delays and therefore one of our key objectives is to improve turnaround time without compromising quality. I am happy to report there has been significant improvements since we started. In 2011, for example, 12 weeks was the average processing time from submission to first decision (including desk rejects). The average revision time of papers that received a revise and resubmit was 41 weeks. As you can see, it is quite high and so we now request authors to resubmit within six months.
There is the other issue you have raised about the lag between an article being accepted and it being published. Recognizing this problem we have recently switched to article based publishing. It is publishing an article into an “issue in progress” as soon as it is finalised and not having to wait for the entire issue to appear. Until now, articles had to wait until a journal issue was fully complete to be assigned page numbers. Now, every time an individual article is finished, it receives a page range and is published online inside an Issue in Progress. Each finished article follows the previous one, until the Issue in Progress is filled with fully citable articles. Volume and issue numbering system will remain - as this is the industry standard, and it also provides context to when the article was published – acting as an indicator of the year.
DI: Continuing with refereeing, what do you think of the following innovations to speed up the review process in economics?
- Posting the time taken to review by referee on the website like some other journals (like the Journal of Financial Economics) have started doing.
- Giving referees a "guaranteed decision in 90 days" coupon if they return their reports by the editor’s deadline - with associate editors committing to provide quick reports in the event a reviewer is late.
a. Yes, this is a good idea and it is something we are discussing.
b. I am not sure if this is a great idea. First, it is unclear that you want an associate editor writing lots of reviews of papers because a review is late. Presumably the idea of having multiple reviews is to get independent assessments of the work. Second, I am not sure the author is better off if the associate editor doesn't know much about the literature or understand the contribution.
DI: Many of our readers are graduate students, recent PhDs, or junior faculty, submitting to journals for the first time. There are now a number of good development journals that compete with the JDE, which presents a dilemma for the researcher trying to choose between the JDE, AEJ:AE, EDCC, WBER, WD, and others. What’s the niche the JDE would like to carve in this field? What kinds of papers would you like to attract more to the JDE? Do you have any advice for authors that would help them in deciding where to submit?
You are right that there are many excellent journals that are potential outlets for development papers and we welcome this as it provides more choice to authors (including us!) and healthy competition. However, we would like to think that the JDE is the first choice among field journals in development. In any case, our goal is to provide the quality of reviewing, editorial decisions and timeliness that makes people think about submitting to the JDE ahead of other journals. Having said that, there are important differences between the JDE and some of the journals you have mentioned. First of all, JDE publishes in all areas of development whereas some of the journals you have mentioned are empirical applied micro journals (in some cases not restricted to development) and therefore, the pool of papers they draw from are not overlapping. For example, JDE publishes theoretical papers, papers on growth, trade and development, political economy other than empirical applied micro papers and this big-tent approach is one of the things that distinguishes it from its peers and this is also reflected in the current group of co-editors. Second, we are methodologically eclectic. We welcome papers that use latest methods like randomized control trials (indeed, one of our recently appointed co-editors Dean Karlan is known primarily for his work on experiments) but we also welcome papers that use other approaches such as structural estimation, IV, natural experiments etc. JDE is question-driven and not methodology or field (theory/empirics/macro) driven.
DI: What are your thoughts on trying to encourage more shorter papers on development - equivalent to the 3-5 page papers in medical and science journals for example. We could imagine these being particularly useful to report on methodological issues and survey design; to report the results from experiments that have very simple succinct findings; etc.
MG: I think this is an interesting idea and it is under discussion among co-editors. With clear, specific criteria for what constitutes a contribution, one can argue that there should not be a lower bound on the length of a paper that warrants publication. However, it is not clear if it is a good idea to simply replicate the medical journal approach -- that approach has a tendency to encourage papers to be narrow in focus and not explore alternative explanations in ways that are instructive and informative.
DI: As one of the leaders of applied theory in development, it would be great for our readers to get your take on what it takes to make an interesting applied theory development paper? Any common mistakes or pet peeves that you run across often in submissions?
MG: In the last decade and a half, developments in new empirical methods (such as randomized control trials) has led to a growth spurt in research in empirical micro development but I am pleased to report that I see a steady flow of theoretical papers and of late, papers that combine theory and empirics. I think there is general consensus that theory and empirics go together and the field will not thrive if one does only empirics or only theory. What makes a paper (either theory or empirical) interesting is the question it asks, and of course, the answer it provides. Just because it has a nice and elegant model or provides a very well-identified estimate does not make it interesting. So in my view necessary conditions for an applied theory paper being good are: it should be relevant to some important question or issue in development (so that it answers the general reader’s question ”Why do I care?”), and second, be interesting (in the sense of being non-obvious) and surprising. If something can be figured out without writing down a formal model, then it is not worth writing down as a purely applied theory piece. At the same time, an applied theory paper can be very subtle and non-obvious, but unless it has results that relate to interesting questions and topics in development, it is best to submit it to a theory outlet. Also, sometimes you see a theory paper that is designed to explain some particular fact. While this may pass both the above criteria, for it to be a good piece of applied theory, it should not be an ex post rationalization of observed facts. Theory not only explains why something happened but also, why something did not happen - for example, every theory of non-profits (or, sharecropping) is also a theory of for-profits (or fixed rent contracts). My sense is, in the coming years there will be a lot of papers that combine both theory and empirics, for example, using theory to design experiments, or using existing empirical evidence to calibrate theory models.
As the Co-Editor of the JDE the biggest mistake I see in submissions is some authors assume that just because a paper uses data from a developing country, or is a case study based on a developing country, that makes it suitable for the JDE. Some papers are too narrow in scope and unlikely to be of interest to the broad development audience of the JDE who are not necessarily interested in an in-depth study of a specific country or region or case. Some papers are not a good fit for the JDE either in terms of the topic or in terms of the methodology. We are aware that development economics is a broad field and many topics can potentially be considered relevant, and similarly, there can be many methodologies applied to answer a question of interest to development economists but still, the methodology has to be rigorous by current standards in the field, and the topic should be relevant to core issues in development.
Thanks for the opportunity to express our views in this forum.
DI: Thanks to Maitreesh. This is the third in our series of Q&As with journal editors, following on interviews with the editors of the QJE and WBER. If you have any burning questions for a future Q&A let us know in the comments.