Keeping contact with authors

August 26, 2008

One crucial aspect of RePEc are the regular mailing that participants get. Indeed, it is easy to forget that one participates in some initiative if one is not reminded about it from time to time. Thus, authors registered in the RePEc Author Service get an email every month with statistics, a list of new citations that were discovered, and some news about RePEc. Since we have started these emails, we noticed that authors have become much more diligent in making sure their profiles were up-to-date and that they have responded to suggestions made in the emails.

One issue that arises over time is that emails become obsolete, as authors move or institutions normalize email addresses. When their email address changes, authors are asked to log in under their old address (which is the username), then change their contact details to the new address (which becomes the new username). Creating a new profile with the new address leads to duplicates.

But some authors forget, as they do not check their old mailboxes. Eventually, the monthly emails bounce, and these authors are placed on the list of lost authors, along with a mention in their online profiles on EconPapers and IDEAS. This list has been very useful in keeping the number of bad emails down, currently 243 out of 17,300. as they and others come forward with updates. Sometimes, we also learn that some authors have unfortunately passed away. Their profile is kept online, hopefully in perpetuity.

Apart from the 17,300 authors with works listed in RePEc, the RePEc Author Service has also about 5500 registrants with no listed works. Either they chose not to claim the listed works, oversaw this feature while registering, or simply registered erroneously. These people get a reminder every six months. Emails that bounce are promptly removed from the service, as these profiles serve no purpose. Indeed, the RePEc Author Service is about assembling publication, affiliation and contact information of authors.

15,000 authors on the RePEc Author Service

December 15, 2007

The 15,000th author registered recently on the RePEc Author Service (which also has another 5,000 registered, but without any works in their profile). See a list of all those registered at EconPapers or IDEAS. This give us the opportunity to reflect on the coverage of this service: what proportion of academic economists is covered? Let me offer a few suggestions.

Assume that the works listed in RePEc provide a representative sample of all the works written by economists. Then determine how many of these works are listed in the profile of a registered author. By that account, about 40.1% have been claimed, and thus about 40% of the profession would be registered with RePEc. This latter number is in reality higher, due to several biases: a) some authors are not alive and cannot register; b) some registered authors have the unfortunate habit to remove from their profile working papers once they are published; c) some works listed are not written by economists, and these authors are less likely to register with RePEc.

Alternatively, estimate the number of authors in the world from the membership in academic societies. I guess the three largest societies are the American Economic Association (18,000 members), the European Economic Association (2,300 members) and the Econometric Society (5,500 members). Obviously, their membership overlaps, and not every of their members is an author. But not every economist is member either. Assume that adding their membership numbers corrects for all mismeasurements, then the RePEc Author Service covers 58% of the profession.

One can also observe a specific subsample of economists, those listed among the top 1000 by Tom Coupé. There, the RePEc Author Service covers 75% of the top 1000 by publications and 65% of the top 1000 by citations (which includes quite a few non-economists). But we have good reasons to believe these proportions are higher than for the whole population. Indeed the proportion is significantly higher for the better ranked within this sample, and we can extrapolate that those outside the top 1000 are less represented in the RePEc Author Service.

In summary, the RePEc Author Service covers between 40% and 75% of the profession. Possibly less, possibly more, likely in between.

Thanksgiving to Volunteers: Ivan Kurmanov

November 21, 2007

Ivan KurmanovAs the United States are celebrating Thanksgiving, it is time to celebrate our volunteers. With this post, we hope to start a regular feature that highlights the work that our volunteers do, sometimes unseen from the general public. RePEc is all built on volunteer effort, and we hope this feature will help these crucial people to get the recognition they deserve.

Today, we want to recognize Ivan Kurmanov, who has just left the RePEc team after being on board for over 10 years. As a undergraduate Economics major at the Belarussian State University in 1996, he noticed the work being led by Thomas Krichel at the now defunct NetEc, the precursor of RePEc. Thinking it was a great initiative, he volunteered to help out. Thomas quickly found something to get him busy: Writing ReDIF-perl, a perl module that validates the data contributed to RePEc by the participating archives and then massages the data for uses by RePEc services. ReDIF-perl has proven to be tremendously useful. Then, Ivan tackled the RePEc Author Service (then called HoPEc) that needed a lot of work, especially to iron out various bugs and performance issues. This was no easy task, as HoPEc was programmed in C++, while all other components of RePEc run with perl. Eventually, it became clear that a complete code rewrite became necessary.

Thomas managed to find a grant from the Open Society Institute to provide an open source author registration system, and Ivan started working full time on it. This is how the current RePEc Author Service was created, based on ACIS, which is now open source under a GPL license. ACIS performs quite complicated tasks, like pattern matching of names, which may include accents and other marks, or citation analysis with surprising efficiency. Another remarkable aspect of this project is that it is extremely well documented, unlike many other RePEc projects, unfortunately.

While technically Ivan was paid for part of his time with RePEc, we should still consider him a volunteer given all the tremendous work he has performed that went well beyond what would have been expected from the little money the grant provided. Also, he had to cope with often shaky Internet connections in Belarus. Ivan now works full time as a programmer, and we hope he will still listen in on RePEc and give his advice, and occasional fixes. Of course he leaves a void, and while Thomas Krichel is currently providing interim coverage, we are looking for a new volunteer to maintain and expand the code behind ACIS and the RePEc Author Service.