Work Progress

Since I got the data to match, I have been running regressions. First, I examine whether JSTOR affected economists citing patterns. I have 44,000 articles from 83 journals since 1985 where the first author is from a decent research institution (i.e., the 400+ that published 25+ articles in these journals over this time period). For each published article, I know the number of journals the author had access to through JSTOR and the number of his references that are from one of these journals.

1. With fixed effects for institution, citing journal, and year, I find that the number of JSTOR journals one cites to increases when the author obtains access to more JSTOR journals. In fact, I find that there are diminishing returns to adding access to more journals. This occurs for 'top' institutions and more mediocre institutions.
2. JSTOR access to more journals 'crowds out' references that are not available through JSTOR at about one-tenth the rate. And this also displays a diminishing rate.
3. Working paper and 'in press' citations increase over time but not with JSTOR. These are now findable over the Internet but not through JSTOR.

The really cool stuff is when I look at 'research productivity.' I aggregated all these articles by institution and year into some 10,000 observations to see if JSTOR increased the rate of publication. For the raw count of publications (controling for institution, journal and year fixed effects) publications rise and at decreasing rate when an author has access to more JSTOR journals. Cool! I need to investigate whether there is a differential effect for different quality of institutions. (Presumably, researchers at better institutions had better resources with which to find these citable articles.) I also want to investigated quality as well as quantity of articles. The usual measure is the citation weighted number of publications.

OK, so this is just economic research - who cares. But, suppose the same holds true in biology, physics, medicine, engineering, etc. Then it might be possible to claim that the Internet has increased the pace of academic research output. This university research is the precursor to the industrial innovation and invention that have increased our standards of living (See Ward & Dranove (1995)). If these results for economics hold-up and generalize, we might be in the midst of the golden age of research.

All in all, a banner day.

4/5