One of the core aspirations of open access (OA) publishing is that it levels the playing field for content consumption, eliminating the advantages that accrue mainly to richer nations and richer institutions in those nations — that is, those who can afford subscription products. The theory goes that because rich countries can afford subscriptions, they get richer still. OA could be portrayed as an attempt to counter the Matthew Effect — the tendency for the rich to get richer — by making publications free to readers. Hints of this as a tacit goal are legion in OA advocates’ writings.
The Matthew Effect occurs in many places in scientific research. Journals with strong impact factors tend to attract better papers, further strengthening their impact factors. Authors generating interesting results tend to attract more funding, increasing their chances of generating more interesting results. Institutions that attract star scientists tend to attract more funding and more star scientists. In many areas, the Matthew Effect appears to exert itself naturally.
And success in one area can lead to success in another area, which subsequently drives recursive success in both areas through feedback linkages.
Our analysis finds a clear correlation between the universities with the most employees serving on the NSF advisory committees and the universities that receive the most federal money. . . . Even when controlling for other factors, we find that for each additional employee a university has serving on an NSF advisory committee that university can expect to see an additional $125,000 to $138,000 in NSF funding.
Perhaps the most compelling portrayal of the Matthew Effect in NSF funding and representation comes in this statement from the report:
Twenty percent of top research universities got 61.6% of the NSF funding going to top research universities between 2008 and 2011. These universities also had 47.9% of the representatives on NSF advisory committees who came from top research universities during the same period. The next 20% of universities got 21.9% of the funding, and had 25.7% of the representatives. The bottom 20% research universities had just 1.0% of the funding and have 2.4% of the representatives.
This evidence of a link between rich universities, NSF advisory committee appointment, and further funding suggests to me that predicating publication on funding will reflect funding availability to a high degree. Such a system seems unlikely to make publication more available for less affluent researchers compared to free publication in subscription journals. In addition, the study found that coming from a big university or university system — so you could have more representatives on more NSF committees — also correlated to more funding. So not only do the rich get richer, but the big get bigger — and funding begets funding, which in an OA world, begets publication.
These thoughts are occurring to others. Recently, the American Historical Association (AHA) released a statement expressing clear cautions around OA publishing, with this kind of “fairness” issue at its heart:
Would the unfairness of unequal access be replaced by a different unfairness, one of opportunity to publish based on the availability of funds? . . . This different unfairness would be at least as pernicious as the current one. It would particularly diminish publication opportunities in fields where grants tend to be small and not central to the way research is done.
The authors also express the reasonable concerns that library subscription budgets will be raided to pay for author processing charges (APCs), and that flagship journals, which set a quality standard and provide translational zones shared by generalists and specialists, might be imperiled because the Gold OA model depends on low APCs, which creates what the authors call “perverse incentives” to publish more articles and lower quality standards.
Perhaps some innovative approaches could crack the problem. On the surface, PeerJ’s approach appears to tackle the Matthew Effect, with its low $99 annual membership per author. However, the total cost for a paper published this way can be about the same as the cost of a standard Gold OA paper (about $1,200, if everyone on a 12-author paper pre-pays, and about $1,600 if everyone pays only upon acceptance), and the requirement to review a paper (or comment on a paper or a pre-print) every year for every author creates a significant asymmetry — if you assume a 5:1 author:paper ratio (PubMed data show a 5:1 author:paper ratio*), it’s unlikely there will ever be as many papers as there are member-authors required to review. As PeerJ accepts papers, the reviewer pool grows arithmetically. Therefore, the pool of lapsing, non-reviewing members grows quickly. If this pool becomes too large, discontent could fester as renewals become commonplace, and PeerJ’s bond with its community could fray, which is likely why PeerJ retains the right to reconsider these cancellations — it doesn’t know what will happen here. Nevertheless, PeerJ’s attraction as a publication outlet will depend on the apparent quality of its reviewer and commenter pool, which is a function of ability to pay, and which drives further author willingness to pay. By creating payment barriers before publication, incentivizing payment upon submission, and integrating a quality bargain based on perceived quality of both published works and reviews/comments, PeerJ will be, at best, perpetuating the Matthew Effect.
It’s clear that OA publishing is moving into a dominant Gold OA mode — there are simply too many sound economic and business reasons for this path if sustainable OA is the goal, and pragmatists in the OA ranks accept this reality. With this as a baseline, we quickly leave behind the notion that OA publishing diminishes the Matthew Effect. As OA efforts lose cross-subsidization from existing subscription business — these subsidies are either direct (intramural, within the same business and books), or indirect (extramural, occurring because the expensive publishing efforts are still supported by subscription dollars, shielding new efforts from the full burden) — more and more funding will have to come from authors. And only those authors who can pay can play.
This is an interesting counterpoint to traditional subscription journals, which generally don’t require author payments except for special services. Free submission and publication seems, on its surface, a better way to fend off the Matthew Effect by decoupling publication from funding. Having this linkage broken may be healthier for science overall. Gold OA seems to create a new “Matthew Effect” linkage, however — between ability to pay and ability to publish.
In these early stages, it seems to me likely that Gold OA will actually exacerbate the Matthew Effect, in at least three ways:
- It will take more research dollars from governmental and NGO funders, leaving less on the table, which will rationally go to well-funded and proven researchers. Prudence will dictate these choices, which will reinforce the Matthew Effect around research funding.
- If prices for OA publication necessarily rise as subscription publication subsidies fade, then only those researchers and institutions already at well-funded programs will be able to publish in the best venues.
- As the more expensive venues become more desirable because they publish the best research and make the greatest impact, the cycle will become self-reinforcing.
I read a Tweet from an OA advocate a few days ago, apparently quoted from a talk at a PeerJ meeting in San Francisco — the Tweet asserted that OA isn’t a business model, but a distribution model. This is a worrisome gloss on two counts. First, you could just as easily claim that subscription publishing isn’t a business model, but a distribution model — which just shows that the distinction the speaker attempted is meaningless. Second, if we hide from the fact that OA is a business model, we’ll never face the reality that as a business model, it has pros and cons, and generates its own set of virtues and flaws. It is not perfect. It creates interesting conflicts of interest. It can be abused, manipulated, and it can have unintended consequences.
At this point, given the strong tendency of scholarly achievement to drive the Matthew Effect in funding, in publication, and in impact, it seems that OA publishing’s future will be as likely as not to perpetuate the problem and, if not thought about carefully and implemented judiciously, may actually exacerbate the “rich get richer” predilection of academic institutions, individual researchers, and prominent hypotheses by coalescing around current sources of funding and prestige, and inflating them through its business model.
* I think this significantly underestimates the average number of authors for research papers, as the PubMed data include editorials, letters to the editor, and review articles, all of which have 1-2 authors in most cases. If anyone has better information, please share via a comment.