File downloads in 2008, by volume

Number of access to the home page of INVESTIGACIONES ECONÓMICAS

Papers published (and forthcoming) in INVESTIGACIONES ECONÓMICAS from 1986 are available in the Fundacion SEPI's web server in Acrobat pdf format. This allows us to count the number of downloads for each article, providing useful information about its dissemination.

However, some explanation is required for a meaningful interpretation of download statistics:

1. Online databases of Economics journals do not usually place article files in their own server, but rather provide a link (URL) to the journal's web server (the Investigaciones Economicas server in this case). The most relevant case is RePEc, which lists all articles published in INVESTIGACIONES ECONÓMICAS from 1986 (http://ideas.repec.org/s/iec/inveco.html). While RePEc provides download statistics for access to articles through its web page, file downloads are included in the overall statistics given by INVESTIGACIONES ECONÓMICAS. In fact, the INVESTIGACIONES ECONÓMICAS server cannot distinguish between file downloads coming through a RePEc web page and those coming from the Investigaciones Economicas web page. The link to the ftp server is the same in both cases.

Currently, the only exception to this rule is given by downloads through the web page of the Red de Revistas Científicas de América Latina y El Caribe (Red AlyC) (http://redalyc.uaemex.mx/). Its server does contain the files of recent INVESTIGACIONES ECONÓMICAS articles, with the publisher's permission.

2. The main bias arising in download accounting is due to robots or spiders. Basically, this is software aimed at indexing web pages or at copying other data servers. A single download of INVESTIGACIONES ECONÓMICAS articles by a robot amounts to more than 500 file downloads. Of course, this biases download figures upwards, in a way that could be misleading.

This bias could be partially reduced by removing IP addresses of robots which adhere to official robot lists (http://www.robotstxt.org/wc/robots.html). However, these lists are incomplete, so they do not allow complete removal.

INVESTIGACIONES ECONÓMICAS tries to analyze massive downloads in detail. With this aim, some files have been included in the data server without a direct link in the web page. Their downloads alert about non-discriminating action by a robot. All file downloads from their IP address are then removed.

3. This process could downward bias final download statistics. Accesses by robots could happen when a system administrator satisfies the demands from users of the proxy server interested in INVESTIGACIONES ECONÓMICAS contents. In such case, even though users find an article in the INVESTIGACIONES ECONÓMICAS (or RePEc) web pages, they download it from a proxy server (for example, a university server) and not from the INVESTIGACIONES ECONÓMICAS server. In other words, subsequent file downloads from the proxy server of an institution by final users will be undetected by the INVESTIGACIONES ECONÓMICAS server.

4. In the same way, when a user downloads an INVESTIGACIONES ECONÓMICAS file, it can be stored in the proxy server of his/her institution. Until the proxy server refreshes its contents (usually after a few days), all downloads of the same article by users accessing internet through the same proxy server are also undetected by INVESTIGACIONES ECONÓMICAS. This may be relevant, for example, after receipt of the e-mail messages regularly sent to alert of new journal issues, when many downloads of the same article are to be expected in a short period of time.