🚀 go-pugleaf

Hal.Varian@umich.edu writes:

>I'm going to be a participant in an NSF panel discussing how the NSF can help
>to encourage the development of Internet resources for economists.  As I 
>understand it, this is something of a brainstorming meeting.  If you 
>post your brainstorms on this issue here, I will convey them to the
>meeting participants.

One suggestion that I have is to develop an internet archival system
for working papers in economics.  I envision something like the
following:

1) Each interested department would maintain a local archive
accessible by e-mail requests (in the manner of netlib) and/or by
anonymous ftp.  Papers would be stored at the site in various formats,
but a postscript version of each paper would be mandatory.  A local
index searchable by gopher (and also retrievable by anonymous ftp)
would also be provided.  Departments would provide staff to maintain
their own archive.

2) Each department may choose to implement an internet mailing list
for notification about new contributions to its working paper archive.

3) A central site would maintain an index that merges the indices of
participating departments.  The central site would also maintain an
archive of abstracts (in ASCII format) that could be browsed on line.

4) Participating departments would be required to submit information
about new working papers (together with optional abstracts) to the
central site.

5) The central site would hopefully have generous file server and
gopher/ftp bandwidth capabilities.  The central site would also be
responsible for enforcing uniform standards for including information
in the central index, and would also provide staff for developping
support software.

6) Although the gopher system may be adequate for searching local
archives, it probably is not flexible enough to do the kind of keyword
searches that would be desirable on the central index.  More flexible
remote database access software might be available, or might be
developed explicitly for this project.

7) NSF funding would be required for (a) staff support during the
startup period, (b) a smaller amount of staff support on an ongoing
basis, (c) equipment purchases for high volume sites.  (Presumably
some sort of institutional matching for the equipment purchases would
be required.)  Equipment needs could probably be met using existing
facilities at many departments.

8) Depending on demand, the central site could support a mailing list
announcing new working papers in certain subject areas, consolidating
the listings of different departments.

9) Individuals could print their own copies of working papers on an as
needed basis, with duplicating/printing costs handled locally only for
those papers that are really wanted.

10) Departmental librarians (in an effort to control costs) could take
responsibility for taking requests to retrieve and print individual
papers and make them available locally for photocopying.

11) Some kinds of standards for what gets included in the indices is
needed.  At a minimum I think that the index should include, paper
title, names of all authors, keyword list, information needed to
contact the author, e-mail address or ftp site where paper is
available, date of latest version, journal citation for older papers
that have been published, indicator for whether or not an abstract is
available, list of available file formats at the archive site.

Justification: The current system, where departments circulate printed
versions of working papers by mail to all departments that
reciprocate, is wasteful of paper, office space, postage, librarian
time, and researcher time.  It is also somewhat exclusive.  Many
departments are cutting back on their working paper series in an
effort to control costs.  There are often significant delays between
the completion of working papers and time that they become available
in the libraries of other departments.  This time could be
substantially reduced.  All interested parties with internet access
would have equal, timely access to the latest research in economics.
The "old boy network" effect, where people only circulate their papers
to individuals whom they think might be interested, would be reduced.
Decentralized implementation would help distribute costs broadly,
minimizing the need for large initial outlays.

I suspect that we would have a system like this already if only
someone were willing to devote the time and energy to get it started.
Obviously there are costs associated with the transition to a new
system, but a coordinated approach would go a long way towards
minimizing these.

Another suggestion: Implementation of an online directory of
economists.
--
T. Scott Thompson              email:  thompson@atlas.socsci.umn.edu
Department of Economics        phone:  (612) 625-0119
University of Minnesota        fax:    (612) 624-0209