Help support open science today.
Donate Now

Conflict between Open Access and Open Science: APCs are a key part of the problem, preprints are a key part of the solution

January 21st, 2020,

Subscription-based publishing business models inhibit open scholarly communication by making consumers of scientific outputs pay to access findings. Open access business models promote open communication, but the popular Article Processing Charge (APC) model is not perfect. APC’s can be a burden for under-resourced authors and, most critically, the model creates incentives for journals to publish as many outputs as possible. This open access business model conflicts with another theme of the broader open science movement—to improve rigor and credibility of research. Preprints solve this dilemma. Preprinting makes almost all papers freely available at very low cost. 

Wait a second.

The problem with APCs is that it incentivizes journals to publish as much as possible, whatever the quality; so the solution is to publish everything, whatever the quality?  

Yes, exactly.  

With preprinting, publishing is a relatively trivial act. Authors need only meet modest moderation criteria for their preferred preprint service. When most anything can be published, publication recedes as the key incentive. What takes its place? Evaluation. Journals have historically confounded publication with evaluation. If the paper meets the evaluation criteria, then it is published. Therefore, publication is the act that signals credibility for authors’ work and evaluation—peer review—is an impediment to achieving that reward. 

Preprinting separates publication from evaluation. Publication itself no longer signals credibility. If publication doesn’t signal credibility, then peer review is no longer a barrier for authors to overcome to get the reward of publication. Peer review becomes a service authors need to achieve credibility. 

When adopted at scale, preprinting can produce a virtuous system in which all content is openly available, and evaluation services (journal editors and reviewers) focus on improving the quality of that content and the visibility of relevant content to readers. Evaluation services help readers decide what to read rather than controlling their ability to read it. This avoids the dysfunctional properties of the subscription and APC business models. All content is available, and authors will seek, and be willing to pay, for peer review as a service based on the credibility that it can confer on the scholarship. Journals that provide weak peer review services will lose reputation for doing so. Journals that evaluate everything favorably will become weak signals of credibility and become low status. Journals that are tough but fair will confer the highest credibility for authors and be sought after. Journals that are transparent with their peer review processes will be more highly regarded because their credibility stamps are verifiable. Moreover, the other components of journals—typesetting, copyediting, marketing, etc.—can be unbundled and offered as unique services for authors to improve the quality, readability, and visibility of their work.

This does not, on its own, solve the problem of “who pays” and the potential inequities between authors and institutions with more or less resources to pay for these services. It does, however, create a marketplace in which price competition is based on quality of services, and the barrier to entry for under-resourced authors and institutions—posting to a preprint service—is very low. With unbundled services, economy and luxury markets can emerge for each, and interventions to address inequities can be focused on precisely the services that have the biggest impact for mitigating those inequities.

Preprinting therefore creates a marketplace that aligns the key goals and cultural incentives for open access and open science—rewarding authors and journals for the quality of their research and evaluation processes.  

Preprinting will accelerate the death of predatory journals

Preprinting at scale has a nice side effect on the Achilles heel of the APC model—predatory journals. Predatory journals are journals that charge a fee to publish without meeting basic presumptions of quality control via peer review. In a system built on publication as the key reward for authors, predatory journals can thrive. Unless the journal is widely understood as illegitimate, the authors get the reward that they need—publication—regardless of how the journal makes its editorial decisions. White-listing or black-listing journals may have some affect at discouraging predatory practices, but those solutions are fighting against the strong incentives on authors to publish.  

In an open system built on preprinting, the key lever available for predatory practice is removed. Authors don’t need journals to publish, they need journals to earn credibility based on their reputation for providing good evaluations. Focusing journals’ reputations on the quality of their peer review creates incentives for journals to increase transparency of their peer review procedures to demonstrate their quality. This will make it even more difficult for predatory journals because they require obfuscation. Sustaining predatory practices becomes much more difficult when the value of the journal is based on the service that it is supposed to be providing.  

To be sure, the preprinting-modified incentives system will not drive out predatory practices instantly. But, the selection pressures will become focused on the delivery of peer review as a service—the focal problem in predatory practices. If journals remain the gatekeepers on publication as the primary reward, the lure for authors to publish in predatory journals will persist.

An open system built on preprinting is much more cost efficient than the existing alternatives

Compared to traditional publishing, preprinting is a bargain. Reviews estimate costs for published articles at journals are typically in the thousands of dollars per paper depending on a variety of factors such as the selectivity of the journal (see, for example, here and here). Preprint service costs are multiple orders of magnitude less. arXiv, the longest running preprint service launched in 1991, reported an annual operating cost of $1,789,411 in 2018 and published 140,616 papers, or $12.73 per paper. arXiv’s 2018 costs included some new development, not just maintenance. In prior years, costs between $8 and $10 per paper were typical. Our organization, the Center for Open Science, launched a preprint hosting service in late 2016 and now hosts 26 preprint services such as INA-Rxiv for Indonesian research, PsyArXiv for psychology, EdArXiv for education research, and engrxiv for engineering. With shared infrastructure and accelerating growth, we are observing substantial economy of scale. In 2019, we forecast maintenance costs to be $208,076 with 26,143 papers published ($7.95 per paper). Our forecast for 2020 are costs of $229,225 with 33,650 papers published ($6.81 per paper). With continued scaling, we believe that we can reduce the cost per paper even further. 

It would be worthwhile to compare operating costs of other major preprint servers. We suspect that those data would reinforce our conclusion—a comparatively trivial investment in maintaining preprint services can make all papers available, and foster a shift in the scholarly communication business model that aligns goals for accessibility and goals for quality and rigor.

An open system built on preprinting facilitates a diverse services-based business model for scholarly communication

But wait! The costs between journals and preprint services is an apples-and-oranges comparison. Preprint services don’t typically offer peer review, copyediting, typesetting, marketing, or other value-added services that journals do. At present, all of these services are bundled as the cost of publishing and are built into the subscription or APC costs. Preprinting opens the door to decoupling these as discrete services that can emerge as competitive markets of their own. Journals can offer some or all of these services as a package or as individual services. The market will shift toward a community of services competing on providing their selected service with the highest quality as the most competitive price point possible. Authors will be empowered to choose their priorities in that marketplace. How much value does typesetting provide? Journals and other services will present options at their selected price points, authors will make value-based decisions, and the market will adapt. This system will also better enable evaluation of scientific contributions based on the transparency of reported outputs. Empirical evidence that is backed by open and FAIR data, that is well curated and connected to clear analytical code will be more credible, and journals that provide services to their authors that support these practices will see both of their reputations improved. These changes will be very healthy for creating competitive marketplaces for discrete services that could benefit from innovation produced by market competition. We suspect that a mature marketplace of discrete services will ultimately produce much better tools for scholarly communication at a much lower price than the present bundled solutions.


Following the early example set by economics, physics and allied disciplines, scholarly communities are recognizing the value of preprints for improving access and accelerating communication of their research. The comparatively modest investment in sustaining preprint services will have profound salutary benefits on enabling a comprehensive transformation of scholarly communication to one that promotes both rigor and transparency, and fosters healthy, competitive marketplaces in which providers compete on the quality of their services.

If you agree, you can help sustain the preprint ecosystem by supporting one or many of the services that you care about. Most of the preprint services hosted on OSF Preprints will gladly accept your donations to help cover the maintenance costs. See a list of those services here. As users of shared, open infrastructure, your support of one helps to advance the sustainability of all!


Recent Posts