The two economists Klaus Wohlrabe and Daniel Birkmaier (both from the Ifo Institute for Economic Research in Munich) have published a new working paper in which they analyse the impact of open access publishing in economics on citations.
Their sample consists of articles from 2005 from 13 economic journals (including the top five journals). In addition to standard mean comparisons Wohlrabe and Birkmaier also use a negative-binomial regression model with several covariates to control for potential selection effects and quality bias. For their analysis they used citation data from three different databases, namely the Web of Science, RePEc and Google Scholar.
The results they retrieved are very interisting and might light the debate on open access publishing in academia.
Wohlrabe and Birkmaier took almost all articles from the
- American Economic Review,
- Quarterly Journal of Economics,
- Journal of Political Economy,
- Review of Economic Studies,
- Journal of Financial Economics,
- Journal of Monetary Economics,
- Review of Financial Studies,
- Journal of Business and Economics Statistics,
- Agricultural Economics,
- China Economic Review,
- Finanzarchiv, and
- Geneva Risk and Insurance Review.
Finally, the ecomists had 639 articles in their sample. In a next step they checked whether the article are freely available online by using Google Scholar. If a pdf-file was available on the first two pages they labeled the article as OA. Normally these articles comprehend either working papers or the (self-) archived original article. So they analyse the impact on citations for the “green” road to open access, but not for the “golden” road.
79 % of the 639 articles investigted are available freely on the web – that’s a large number. The largest share is given for Econometrica with 96% and the lowest with 33% for Agricultural Economics. Almost 50% of the articles appeared in a working paper series.
Subsequently, Wohlrabe and Birkmaier collected citations from the Web of Science, RePEc and Google Scholar in October 2012 for articles available open access and for those who are behind toll bars. For the full sample they found 508 citations for OA articles and 131 for articles with a gated access.
In a second step, the two researchers estimate the OA effect on citations by modeling citation count of a paper by using a negative-binomial regression model. The results they obtained point to the direction that articles published as OA receive on average more citations than those that are not. Wohlrabe and Birkmaier found a significant OA effect on citations counts for economic articles. This effect is robust across three different bibliometric databases.
In the articles conclusion, the two economists state:
One may conclude that researchers should make their articles freely available on the web to increase their citation count.
The full paper is available at MPRA.
Graphic: “Open Access logo designed by PLoS”. License: CC0 1.0