On Tuesday May 17, 2016, SSRN announced that it was being acquired by Elsevier. SSRN, the Social Sciences Research Network, is a widely used repository of scholarly articles that can be uploaded and downloaded by anyone. It is “open access” (that’s in quotes because SSRN’s approach to OA has always been partial and peculiar, and Monday’s news confirms that perception). Elsevier is, well, if you don’t know who Elsevier is, none of this will make sense to you.
Much of the reaction has focused on what will happen to all the papers in SSRN. Twitter has been alight and aghast with reactions, mostly of the form “this is not going to end well.” Many people are asking whether Elsevier will continue to do all the evil things it normally does: charge huge subscription costs to libraries, high download fees for single articles to individual readers, continue its assault on open access by shutting down access, etc. Some people are suggesting that it is time for a alternative, such as an ArXiv for social science research. All these things are good food for thought.
I would suggest to SSRN users, though, that they have nothing to worry about. I don’t think Elsevier cares about the papers. God knows they barely care about the 2000+ journals they publish, other than Cell and a few others—a fact which is obvious to anyone who has published in an Elsevier journal, or edited one. The papers in SSRN have no economic value, Elsevier can’t really claim ownership over them unless they were subsequently published in an Elsevier journal (which no doubt a substantial portion of them were). Some large number are probably published elsewhere which creates an interesting conflict. But ultimately Elsevier has such deep pockets that such law suits would disappear like lint and crumbs into their depths.
So what is the real value in SSRN? Data.
The data produced by SSRN is not terribly sophisticated stuff: number of papers and authors, number of downloads, number of citations, per paper and per author. Lots of other companies and services attempt to collect the same kind of data. But what makes SSRN specific is that it is a well known node in the network—we might say, in the discourse or mind-space—of social science.
Because it is “open access” and because it collects an impressively diverse range of social science in one place, SSRN’s data actually represents the world of social science scholarship reasonably well. There are lots of social scientists who don’t use it—none of my friends in sociology or anthropology do, for instance. Remember though: All models are wrong, but some are useful. SSRN represents better data about the impact of social science research than any single journal, or any publisher’s data (even Elsevier, with its hundreds of social science journals), because it has been built on the good will, apparent neutrality, and level playing field of an open access repository. Until Tuesday May 17, it was not tied to any particular publishing conglomerate or metrics company, and so its data represented social science research with a fidelity to reality that is expensive and hard to otherwise achieve–the result of the hard work of Gregg Gordon, his staff, and advisers to be sure.
Why is this valuable to Elsevier? Because it’s valuable to academics. Like the Impact Factor, which is also owned by a for-profit high-margin company—Thomson Reuters—SSRN is a valuable input in the bureaucracy of academic personnel. Academic administrators long ago gave up evaluating scholars based on quality or innovative research, and turned to evaluating “impact” instead. And impact is a sort of metaphysical quality that is not in the research itself, but in the circulation and reception of research—it can only be captured by metrics, which requires collecting data. The reason is obvious to anyone who works in the university: impact = higher rankings, higher rankings = more and better students, more donors, more reputation for the institution… all of which translates into the ability to hire more high impact researchers. If that kind of data is valuable to academic administrators, Elsevier is right to focus on collecting more if it, monetizing it, and selling it back to Universities. After all, the model of taking all the labor of academics, packaging it up and then selling it back to Universities and academics has worked really really well for them in the past.
Don’t take my word for it though, here’s the press release:
“Elsevier is actively linking data and analytics to its vast content
base in ways no other potential SSRN partner can match. By
connecting Mendeley, Scopus, ScienceDirect and its editorial
systems, they’re helping researchers get a more complete picture of
their research landscape. Institutions will also benefit with a
better view of their researchers’ impact.”
Translation: “helping researchers get a more complete picture” means something like “reducing impact to metric that can be applied to an individual scholar”; whereas “institutions will benefit” actually means “universities will pay big bucks for this data.”
What’s wrong with a big for-profit company producing such metrics? What does it mean if Elsevier owns this data? Here’s where there is no difference between the old private SSRN and the new publicly-traded for-profit Elsevier-SSRN. In neither case is something like “data skepticism” possible. The data is not available to the people or institutions or disciplines it purports to measure. It cannot be contested, it cannot be re-analyzed, it cannot be investigated, it cannot be downloaded. It just has to be trusted.
The problem with this—from any self-respecting academic’s standpoint—is that this is anathema to our very practice as academics: data is never the truth, data is never “raw,” data is always messy, dirty, constructed, interpreted, analyzed, re-analyzed, and uncertain. It’s garbage in, garbage out: if the data is bad, the analysis will be bad, and the only way to know that is for many people to see the data and contest it. At the end of the day, we academics prefer expert human judgment about data to unverifiable claims based on secret data that we cannot see.
Needless to say, this is not how for profit businesses think about data in the era of Big Data, data analytics, predictive analytics, etc. For a business, every data point is a potential strategic advantage in a competitive marketplace and they keep that data as close to the chest as possible. Think about the difference between the data the US Government produces about jobs numbers vs. the data that Monster.com or LinkedIn collects about the US labor market. Getting the former data is virtually a human right, getting the latter, well, good luck and Godspeed. We have the former because we need a universal, shared metric for policy making and for dispute about properly political issues like whether to invest in infrastructure or lower tax rates, or create stimulus. It doesn’t mean that LinkedIn or Monster.com should only use that data, or not collect their own (indeed, their data may very well be much better), but there is no way we are going to set public policy based on LinkedIn’s secret data—at least not in a functioning democracy. But that’s another issue.
Ironically however, this is exactly the world we academics have built for ourselves: a world where a very large number of judgments about quality, employability, promotion, tenure, awards, etc. is decided by opaque metrics collected by for-profit firms. I have no idea what Elsevier’s grand plans are for SSRN—they are probably a modest intermediate level of evil, and not the Advanced Evil they practice normally. But their bigger plan is to get out of subscription-based publication models all together. It’s a losing model: the pirates are doing a much better job, the academics want open access, the libraries hate them–no one would blame them for jumping ship (as clumsily and slowly as one of the world’s largest corporations can “jump ship”) into a strategy based in monetizing data and metrics. It’s obvious that they seek to do the same thing with their purchase of Mendeley and their expansion of their portfolio in the direction of metrics. And they are not alone: Research Gate and Academia.edu are competing for the same institutional attention and dollars by monetizing data about academics and their work produced through platforms that extract that data in return for “hosting” the nominally valuable part: our work.
So while we should be anxious about Elsevier’s immediate plans for SSRN—will it remain “open access”, will they do something evil with the papers—the real elephant in the room is that we ourselves–we scholars—are producing the market for data every time we insist on evaluating a colleague according to some hack metric like Impact Factors or SSRN downloads. It’s not that such metrics are bad in themselves (but cf. Goodhart’s law), but when they are inaccessible to skepticism or scrutiny, when they cannot be analyzed differently by different actors—we set ourselves up for a world were we buy access to data about ourselves that we cannot be sure represents us accurately, in order to make decisions—sometimes trivial, sometimes existential—about our careers and ultimately the quality of our work and the problems social scientists deem worthy of attention. SSRN will continue to be a valuable place to post papers—lots of your colleagues will see it there. Elsevier may even be magnanimous and keep all those papers “open access”—but we should in all accounts avoid taking their data seriously. Data skepticism needs to be part of the brave new world we are building, and that is not something the Elseviers of the world are interested in.