Following this introduction this article contains four main sections and a conclusion. The first section describes the current state of NetEc (current as of late February 1997). It draws heavily from a text I submitted to the MIDAS newsletter. The second section is concerned with the early history of NetEc. The third section looks in more detail at a specific NetEc project called WoPEc. It is based on a speech given at the Caisse des Depots et Consignations, Paris on 6th January 1997. The fourth section deals more generally with electronic research publications in Economics, and how WoPEc may impact on academic publishing in the future. All sections may be read independently but nobody should leave out the conclusion.
This article was written and submitted to CHEER to celebrate the fourth anniversary of NetEc on 1st February 1997. It is not part of NetEc's policy and the opinions expressed here are mine alone. It is dedicated to the people who have built and supported NetEc over the years. Some of them are mentioned in the following.
BibEc is a collection of bibliographic information about printed working papers. These are early accounts of recent research. Fethy Mili of Université de Montréal has donated information about his holdings to NetEc. There are over 40000 papers from almost 500 series. Most of the data comes from the collection that Fethy is holding at his "Centre de Documentation", but the National Bureau of Economic Research (Cambridge, Mass), the Centre for Economic Policy Research (London) (these two institutions are the world's leading working paper publishers) the Federal Reserve Banks "Fed in Print" database, as well as some individual institutions have contributed to the collection. The data can be browsed (if you have a lot of time to spare) or searched using a WAIS full text index. Markus Hatterscheid of Universität Bonn is responsible for the technical implementation of the project.
WoPEc is a BibEc for electronic papers. In addition to bibliographic information (author, title, abstract etc) all records in WoPEc contain information about the electronic manifestations of the paper (PostScript, Portable Document Format (PDF) and other more obscure format). On the web server, these are represented by hyperlinks. The user can fetch the electronic manifestations from the server where they are stored. This may be a machine overseas, in which case the user may need some patience. When the file is on the users' local machine, she can treat it at her convenience, i.e. (try to) view it on the screen, (try to) print it, or ignore it. The fact that the PDF format is gaining acceptance should mean that soon users can printed and view documents much more easily than it used to be in the past. Generating these documents should also become much easier when the major wordprocessors will have a facility to output PDF. This will happen with the latest versions during 1997.
To date there are over 3700 paper catalogued in WoPEc. Data in WoPEc can be searched using a purpose-built web interface to a digger whois++ database, or searched as a WAIS fulltext database. WoPEc is led by myself at the University of Surrey and managed by José Manuel Barrueco Cruz (on secondment from Universitat de Valencia).
CodEc is a WoPEc for computer code. So rather than cataloguing papers, CodEc contains information about computer code and programs that are useful for economists and econometricians. It provides source code for several programming languages (currently in C, C++ and Fortran) as well as programs written in some of the econometric and statistical software packages like for example Gauss, Rats and Shazam. CodEc also has programs for computer algebra systems (Mathematica) and executable binaries for DOS or Windows. It is lead and administered by Dirk Eddelbüttel (Goldman & Sachs Toronto).
WebEc is not a CodEc for something else. Rather it is a more general purpose link collection to all kinds of Economics information on the World Wide Web. It is a part of the World Wide Web Virtual Library, coordinated by the W3 consortium. It is organised according to the Journal of Economics Literature (JEL) classification scheme, a scheme that is widely used to order the sub-fields of Economics. In addition WebEc features some special categories that are more specific to the web. WebEc started as a personal link collection of Lauri Saarinen of Helsingin Yliopiston. His collection became soon known in the world as the "Helsinki Economics Link Page". Later the project joined NetEc and became WebEc.
BizEc is a WebEc that is more oriented to business resources. It also differs from WebEc in its classification method. It first sets aside Finish resources from the ones that interest the rest of the world. Second, within each of those categories, there is a matrix with topics in the column and frequency of data on the line. Thus it is possible to discriminate between short-run and long-run information about a given topic. BizEc is moderated by William Cardwell and Lauri Saarinen.
Last there is HoPEc. This project has some resemblance to WoPEc, not only by its name. Note however that the "Wo" in WoPEc is more pronounced like the initial consonant and vowel of "war", whereas the "Ho" in HoPEc is pronounced like the "ho" in "hope". The idea for HoPEc grew out of the concern about homepage publications in Economics. At the moment, most electronic working papers are stored in archives. An academic department of Economics or another research organisation opens an archive that stores the papers produced by authors affiliated with the institution. This is the traditional model of the working paper publication adopted from the days of print. However the web has favored the spread of homepage publications. These publications tend to be more volatile and therefore it would be difficult to include them in WoPEc. Hence the idea of a collection of links to these paper. Rather than cataloguing each individual paper, we catalogue the researcher. The collection uses the JEL scheme. It is led by Barry Schachter (Chase Manhattan Bank, New York City). The technical implementation was written by Ivan Kurmanov, an Economics student at the Belarussian State University.
The NetEc sites also carry copies of two other important resources for economists. First there is "Resources for Economists on the Internet", a large document listing detailed descriptions of almost all Economics resources that an academic could be interested in. It is written by William L. Goffe (University of Southern Mississippi). Second there is EDIRC, a list of Economics Departments, Institutions and Research Centers on the internet. This authoritative collection now lists around 1500 institutions and is growing strongly. It is produced by Christian Zimmermann (Université du Québec à Montréal).
None of these projects requires registration or a subscription fee. All projects except WoPEc are produced by volunteers. WoPEc has attracted funding from the Electronic Libraries Programme for 18 months. The next section explains how such large and complex project could come about by looking at the origins of NetEc.
My personal fortunes changed at the same time. I was offered a lectureship on a permanent contract by the University of Surrey. Before that time I had been employed on short-term contracts. The employment security gave me the chance to think more long-term. At the time the Janet-Internet Protocol System (JIPS) (is there anyone who remembers that?) had become available. The first gopher servers allowed to store data in a way so that an uninitiated user could easily retrieve it. The possibility of database queries on the gopher system as well as the distribution of full text papers were already visible, although not immediately available at the time. Clearly there was a large potential that this new medium was offering for the communication of research results.
My more immediate need was that I had no data. I wrote to the buslib-l list, and Fethy Mili responded that he could give me data for over 250 series. We agreed to start on 1st February 1993, the date I would be taking the job as a lecturer, initially operating a mailing list. I contacted the Rutherford-Appleton Labs for a listserv software but all they were prepared to offer (I understood that they could have chosen to do otherwise) was a ListRAL address operating on the Janet end of their network. A list called WoPEc was opened but never really operated. While waiting for a transfer to listserv I contact Hans Amman. He had a list email@example.com (yes, with a one) and offered to distribute the data on his list giving immediate access to about 400 readers. Thus we started on that list. The RAL list was never used, and closed in 1996. It continued to appear on several "all internet services in one book" guides that have the annoying tendency to perpetuate incorrect information.
With the distribution service via email working but more things being planned there was time to think about a structure. Although initially the project was called WoPEc it became soon apparent that when new services were added new project names would be required. Thus the idea to create a superproject came about, with a generic name as something that is concerned with using the electronic networks for academic Economics: NetEc. The project dealing with printed papers was called BibEc.
The BibEc project is the origin of NetEc. We very soon added a gopher server with all the historical data of Fethy i.e. data going back until the late 80s, the time when Fethy took up the job as Economics Librarian in the Department de Sciences Economiques at Université de Montréal. I had originally approached Manchester Computing Centre (as it was called at the time) and they kindly agreed to provide me with disk space, CPU time and most excellent computing support. The gopher server was set up by Geoff Lane and ran on a UTS system, a supposedly UNIXish operating system running as a guest of the Amdahl VM/XA service. Despite the technical limitations that this imposed Goeff brilliantly solved all the problems and the service remained open until the machine was shut and replaced by the CS6400 machine.
The movement of machine was completed in early 1995, but the gopher service did not run properly for the ftp transfers. By the same time the CodEc project was added to the NetEc service portfolio. The changeover to WWW took some time and was not ready until the spring of 1995. Markus Hatterscheid had at that time become an important volunteer on the project. Through this work he established a reputation as an expert on all matters WWW and was able to win consultancy contracts from it, demonstrating that working for NetEc as a volunteer could pay off. Later in 1995 the WebEc project was added. Discussions with the Americans intensified, Bill Goffe attended a meeting between me and Markus in Summer 1995.
Our first mirror was with Washington University. It opened in 1995 after efforts by Markus Hatterscheid. This mirror was quite simple, it copied the files from Manchester to St. Louis. An index was generated locally at St. Louis. With the opening of the second mirror in Tokyo we reorganised the whole way of mirroring, the result was called the Tokyo protocol. This is a rather complex mechanism, but it has served rather well. Mirroring will be a crucial issue in the future, in particular for the papers in WoPEc.
My theory is that economists have a built-in distrust of monopolies. In their book of tales, there are numerous accounts of the welfare loss that occurs when a monopoly supplier serves a market. They may also be afraid of the power accumulated by a person who controls a hard disk where all the world's recent knowledge produced by the discipline is stored. That is why a centralised system has problems to be set up. That does not only hold for an archive that stores papers, it also affects the collection of metadata about holdings on local sites, i.e. the work WoPEc is primarily concerned with these days. We have however established contacts with about 10% of archives, some universities and Federal Reserve Banks in the United Sates but mainly continental institutions. The uploading of information is conducted via ftp using a procedure called the "navette". The provider puts a description of the papers in ftp://netec.mcc.ac.uk/pub/NetEc/incoming. We then pick up that file, examine its contents, if necessary correct its syntax, install the information and then put the corrected file into ftp://netec.mcc.ac.uk/pub/NetEc/outgoing. The provider then picks up the file and uses it as the basis for further uploads. There is no need for any further communication between the provider and us. Thus it is relatively easy to provide information, and it should be in the interest of archive maintainers to disseminate their work as widely as possible.
Since April 1995 an ever increasing share of the data was collected by José-Manuel Barrueco Cruz. He came across WoPEc when working as a librarian at the University of Valencia. He wrote a message to the Manchester server in April 1994 and soon became a co-worker on the project. It was during a meeting with him in Valencia in April 1995 that the first steps were taken to convert the format that I initially designed to the one proposed by the Internet Anonymous Ftp Archive (IAFA) working group of the Internet Engineering Task Force. These were meant to be compatible with the whois++ protocol that was under development at the same time. That is not quite the situation and the status of the current input format for WoPEc is somewhat indeterminate still today; it is likely to change but not fundamentally. Fundamentally our standards compliance has been with whois++, but this is under review since we have problems with the whois++ software and we are told that LDAP is gaining more prominence.
In 1995 I applied for funding for the project from the Follet Implementation Group on Information Technology (FIGIT). Among the area in the programme WoPEc could have fitted into "electronic document delivery" or "access to network resources" or "electronic journals", but we submitted in the first area. FIGIT were so puzzled that they did not send us any response. In a later circular, institutions were invited to bid for a "preprint server" area, and I am glad to say that we obtained some funding for WoPEc. Writing the application took three weeks of my time, and the initial response from the funders requested that I obtain support letters from all leading departments in the United Kingdom.
WoPEc has much improved since the funded work started in August 1996. José-Manuel Barrueco Cruz was appointed project manager. He increased the stock of papers, associated controlled vocabulary with all the file format, improved documentation, wrote a web client to whois++ that is made for our needs. We are in the process of applying for prolongation funds. Please if you are using WoPEc's services let us know and write us a little message saying that you like it. This will serve as vital evidence to back up our application. In addition, we are very keen to increase the stock of British papers in WoPEc. If you wish to publish a paper on our web site please do not hesitate to contact WoPEc at WoPEc@netec.mcc.ac.uk. We will help you to prepare an electronic version of your working paper, and will make it available on the Manchester archive or on your own working paper site. We can also help departments to build departmental archives.
The name WoPEc comes of course from "working paper", i.e. an account of recent research published by the author or her department prior to submission to an academic journal. Thus the papers in WoPEc are not strongly peer-reviewed. Of course there is a selection at two levels, first the paper has to be accepted in the departmental working paper series and second the series is selected for inclusion in WoPEc. However at none of these stages the paper is studied in detail as one would expect it to be before the publication in a journal. Thus a system like WoPEc, even if the acceptance problem could be sorted out would not be a replacement of the journal system.
However the above paragraph is not quite correct. First WoPEc does not only contain working papers as its name would suggest. There are papers that are published by commercial publishers and that are in WoPEc in the sense that WoPEc contains a file whose contents when printed would be informationally equivalent to the printed publication of a conventional print publisher. There is for example the case of a particular researcher who has made all his published work accessible in this way. This could be copyright kamikaze in the sense that if one of the publishers of a journal in which get to know about the electronic deposit the author could be in for a law suit (This is the reason why we are not mentioning his name here). On the other hand the author may still hold the electronic copyright over the publications. It all depends on the particular type of copyright transfer that the author has signed.
Clearly from the preceding discussions, there appear a number of cracks that could considerably weaken the publisher's bargaining position in the system. First the authors could refuse to sign the copyright declaration, or sign it but cross out some of its clauses. As long as the academic leadership of the journal still wishes to publish the author's work there is a conflict between academic and commercial interest in the editorial office.
Second, there could be a whole "awkward squad" forming, a number of authors that will simply refuse to publish in a journal that would prohibit the free retrieval of an electronic copy. The squad already has an early member in Prof. Rasmusen of Indiana University, a law and economics specialist of high renown. Clearly initially only well established academics will be able to afford such a position, those who are tenured and already well known. But as the electronic medium becomes more wide-spread the advantages of keeping an electronic copy will become more important. So on the supply side there may be problems with authors either refusing to hand over copyright or demanding payment in return.
One way out for the publishers is the electronic journal. All publishers are eager to establish themselves in the electronic work and to demonstrate that the electronic communication revolution will leave them with a job to do. Elsevier now claim to make 20% of their business in electronic publishing, a figure that should be interpreted in the sense that they do 20% of their business in areas where some form of electronic publishing exists. An insider from the industry recently told me that although the publishers are experimenting with electronic publications none of the established players is actually making money from it. This could change over time of course. In Economics we have just seen the first issue of Studies in Nonlinear Dynamics and Econometrics, an electronic-only journal operated by MIT Press. If the contents of that journal were to be free, we would include it in WoPEc. It would be counted like any other paper series. As long as the contents of the paper are sold we are looking at electronic journals as a cost-cutting exercise by the conventional publishers that does not alter the fundamental problem of copyright usurpation of scientific knowledge by publishers.
On the demand side, things do not look good either for the publishers. Library budgets have been cut in many universities in the world. Cancellation of journals have been frequent. Publishers have increased prices well beyond the rate of inflation during the 80s and many librarians are angry about it. A top figure in the UK library world recently confessed to me that he could imagine the whole system totally collapsing in a few years time. We note however that the librarian may turn out in the end to be one of the supporters of the commercial publishing system. With no more journals to handle and everything on the network, large numbers of librarians will be on the street, particularly if they do not posses computing skills.
To run a system like WoPEc in its final stage, i.e. with publishers supplying information, only one librarian is required as an intermediate between publishers and readers. That is a person that associates codes with each series. Within WoPEc there is a code for each working paper series. The publisher of the series then adds the code of the paper within the series she is publishing. The combination of the two codes would make for an identifier for each paper that is published. Clearly there are some issues of digital preservation that arise if we wish to consider the permanency of the publisher, but if these are addressed a system with an identifier will be much superior to the current system that does not have any.
This is particularly evident when it comes to the issue of citation analysis. A good way to measure the impact of an academic publication is the number of citations it has received. However in the current system without identifiers, we can not refer to the paper's identity but rather to a list of its qualifiers i.e. the paper by this author in that journal in this year etc.. Even a large citation index like the one produced by the Institute for Scientific Information only lists descriptions of articles that were cited. With any search of citations the user has to decide which descriptions correspond to the same article. That is a task that a computer could not take over.
Thus the current system without handles hampers the development of a peer review by citation system. It is not a difficult task to construct software that would look through files to look out for citations, it is getting agreement on the citation system that seems to be the most important stumbling block.
Would losing the journals be a big loss? Clearly the peer reviewing process would no longer be organised by the journal. There is a value to ex-ante peer-reviewing. It filters out mistakes and improves the readability of papers. But all to often papers have to be cut to fit into space restrictions imposed by the journal, thus a working paper version may actually more helpful for those who wish to get to the nitty-gritty of the paper. And all too often mistakes still persist in the printed version of a paper. And once it is printed there is no way to correct it, even if it is spotted by an alert reader.
There are some ways by which an electronic peer review system can be implemented without the intervention of a conventional publisher. Some are already appearing in an embryonic form. For example there are a number of single subject web sites that deal with resources in a particular area of expertise, there is no reason why these could not over time develop into reviews or archives of papers, and there are archives for special topics like the Post-Keynesian Archive. Second, the academic societies could get themselves more heavily involved into electronic publishing. This is less likely since many are relying on revenues from the publishing process to subsidize other activities. It would also require organisational change that is more difficult than individual leadership. Here we find a parallel to the commercial publishing market. It is more likely that there are new players coming on the scene rather than the established adapting.
Guy Judge informs me that it is now an HMSO publication with editorial control from the University of Luton
The author may be contacted at T.Krichel@surrey.ac.uk