Saturday, April 11, 2009
This is a really interesting "working paper" called Demographics, Career Concerns or Social Comparison: Who Games SSRN Download Counts? in which two Harvard B-school profs, Benjamin Edelman and Ira Larkin, seek to determine whether academics "game" their download counts on SSRN and the motives for doing so.
As the authors note, there has been much interdisciplinary interest in understanding why people engage in deceptive acts. Until recently, most of that work focused on the economic incentives for engaging in fraud. However, researchers working in the social sciences have begun to explore the extent to which social standing, envy, cultural norms and similar factors motivate people to engage in dishonest conduct. In sum, these social factors may play a larger role than money in causing academics to "cook the books."
With respect to professors downloading their own papers on SSRN in order to inflate the numbers, professors Edelman and Larkin found the following:
First, we find strong evidence that gaming increases when it will increase a paper’s visibility on SSRN by putting the paper (or keeping it) on a 'Top 10 list.' Second, we find some evidence of gaming due to career concerns and demographic factors, but the results are limited. In the two years prior to a change in employment, an author is somewhat more likely to engage in download gaming, but assistant professors who are approaching a tenure decision are less likely to engage in gaming. Females and non-U.S. citizens are somewhat less likely to engage in gaming, as are researchers at low ranked institutions.
In contrast to the relatively limited results on economic and demographic determinants, we find strong evidence that envy and social comparisons play a strong role in predicting deceptive downloads. Higher levels of reported downloads for three separate peer groups – an author’s institution, other authors within an SSRN e-journal, and authors within an e-journal publishing papers on SSRN at about the same time as the author in question – are associated with 12% to 30% more invalid downloads. Our results therefore suggest that psychological determinants of gaming are important and understudied.
Ironically, I haven't been able to get into SSRN for the past few days or I'd provide a link to let you jack-up the download count on this article. Instead, I'm providing you with a link to the pdf version of the working paper available on one author's faculty page.
It's disheartening to think that this kind of gaming really goes on in academia. On the other hand, I guess I shouldn't be too surprised since the study merely confirms what we should already know; that even self-described intellectuals are, underneath it all, human beings with all the shortcomings and personal foibles that the rest of us struggle with.
Hat tip to the TaxLaw blog.
I am the scholarship dude.