Informal Collaboration in Financial Economics

On the left is a co-authorship network of financial economic papers. On the right is an acknowledgment network. (Image is a slide from Georg’s presentation at George Mason University.)

A classic application of network analysis looks at authorship networks. Co-authorship is considered a reliable proxy of collaborations and has been researched since the 1960’s [1]. Co-Pierre Georg and Michael E. Rose are researching what this looks like for financial economics research.

Georg and Rose are not taking the traditional co-authorship network route, because, according to Georg, in economics there is less co-authorship. He says that this is because it is a very long process, and research has shown that it is getting slower over time.

“Right now, it takes two to four years to write a paper, then another two to publish,” Georg said, during a presentation on his research at George Mason on Jan. 29, 2016. “Unlike biology or science, in economics there is a very high rejection rate [when submitting for publication].”

So, Georg and Rose took a different approach by looking at informal acknowledgements that occur near the beginning of the paper. These include:

• Individuals’ names, such as those of their colleagues who assisted
• Institutional informal collaboration, such as seminars where they presented
• Conferences (in the last sixty to eighty years)

Georg and Rose’s intent is to document the extent of informal collaboration, show the hierarchy flow of information, rank financial economists according to their centrality, explore determinants of centrality, and demonstrate that collaboration with central colleagues is valuable in the realm of financial economics.

Georg explains findings of their research.

In the first part of their research, Georg and Rose looked at papers from 1998 through 2011. Among their findings, they found that informal collaboration increases over time, and there is more informal collaboration in general-interest publications. Furthermore it was found that more recent papers from 2009–2011 were more diverse in their network than earlier papers from 1998–2000. In essence, the publications and the process became more democratic by including more individuals.

On the left is the network from 1998 to 2000. On the right is the network from 2009 to 2011. On the right, the giant component of the network is much larger than the network on the right and has more links. Overall, the later network is considered more “democratic”.

There are a lot of interesting components to this study. If you are interested in learning more about the study, their approach, and to interactively explore the networks, check out their website. Their site is probably the best examples of pre-socialization / user testing of academic research that I have ever seen.

[1] Kumar, S. (2015). Co-authorship networks: A review of the literature. Aslib Journal of Information Management, 67(1), 55. Retrieved from http://search.proquest.com.mutex.gmu.edu/docview/1647081914?accountid=14541