The journal's business model

The CJPS is a hybrid journal in that its business model relies on subscriptions, but it also allows authors to publish immediate OA articles in exchange for a fee that covers the costs of editing, publishing and promoting the work. These fees, referred to as article processing charges (APCs) are currently set by the publisher (CUP), in agreement with the CJPS, at US$2,980. As of October 2019, only three authors have taken advantage of this OA option. The CJPS also has a “green” OA policy, under which authors can upload the accepted version (after reviews but before the final formatting) of their article to an institutional repository. Commercial sites such as Academia.edu and SSRN are not compliant with CUP's Green OA policy.

As is true of most academic journals, traditional individual and institutional subscriptions to the CJPS are increasingly being replaced with subscriptions through consortia (the so-called big deals) that CUP negotiates with libraries around the world. In 2018, CJPS circulation included traditional institutional and consortia subscriptions, as well as departmental and individual member subscriptions directly through CPSA/SQSP. A significant number of consortia subscriptions added since 2017 are low-cost or free subscriptions to institutions in Africa, the Middle East, South America and Asia through CUP's aid/donation program.

The journal's main sources of revenues are royalties from subscriptions, as well as from reproduction rights and permissions. CUP absorbs a significant proportion of subscription revenues in exchange for a range of professional services, including copyediting, production, printing, marketing and distribution. Additionally, CUP provides access to Editorial Manager, an online editorial software package that assists in the administration of the editorial process, and to Cambridge Core, its online platform for academic content. Other sources of income include reproduction rights that the CJPS generates from older issues that are not licensed with CUP (through JSTOR and ProQuest), as well as the ASJ grant from SSHRC, which constitutes approximately 20 per cent of annual revenues.

On the expenditure side, in addition to the production costs assumed by CUP, the CPSA provides an annual direct transfer equivalent to the SSHRC grant to the English-language editorial team. The SQSP provides direct support for the French-language editorial team. In addition to CPSA funding, the host university of the English-language editorial team supplies an equivalent or greater subvention to cover course release and other expenses. Without the dedicated work of the French and English teams, it would be impossible to produce a high-quality, reliable journal dedicated to supporting the Canadian political science community. The CPSA also directly incurs additional expenses related to the journal, including translation of all journal-related documents and web material and its clerical and financial administration.

While producing a quality peer-reviewed journal such as the CJPS is expensive, a significant portion of journal revenues is nonetheless transferred to the CPSA and SQSP, according to a mutually agreed upon formula. The net return to the CPSA from the journal accounts for roughly a third of the association's annual total revenue in recent years. The journal is therefore an important revenue source that supports, in part, a range of activities of benefit to CPSA members, including the organization of the annual conference, student travel grants, the association newsletter that has more than 4,000 subscribers, and the Praxis blog. The relatively stable income stream from the journal has allowed the association to offer low membership fees and to subsidize student memberships and student conference fees. This situation is not unique to the CPSA (Bull, Reference Bull2016). Many large academic associations in the social sciences rely heavily on income from journals to support their activities.Footnote ³

The impact of the changing landscape of scholarly publishing and, more specifically, the growing push for OA, will therefore be significant for the Canadian political science community. The CJPS and its proprietary scholarly associations rely heavily on revenues derived from journal subscriptions and reproduction rights. While it is extremely difficult to predict trends in subscriptions in the face of a changing OA market, CUP forecasts that over the next five years, overall revenue from the journal is likely to continue to decline due to the changing nature of the publishing environment, including growing pressure for alternatives to the subscription-based model. Maintaining the status quo may therefore be difficult in the medium to long term and may not be in the interest of the CPSA/SQSP and their members. On the other hand, a transition to OA will also have significant implications for the journal, its authors and its parent associations.

What is OA?

Simply put, Open Access means providing free and unrestricted online access to academic publications (BOAI, 2002). To qualify as OA, a publication must be freely available in a digital format, absent any subscription barriers or paywalls. The research should also be free of most copyright and licensing restrictions, meaning that it can be used, reproduced and disseminated at will, provided authors maintain control over the integrity of their work and are properly acknowledged and cited. Many OA advocates and funding agencies add a third criterion: that the research should be made immediately available, without an embargo period.Footnote ⁴

The purpose of OA is to make research, especially if it is publicly funded, as accessible as possible to the broadest possible audience. Several models have emerged to support these goals in the last two decades, each with its strengths and weaknesses (see Figure 1). While some models align more closely than others with the underlying principles of OA, none has so far established itself as the clear alternative to the subscription-based business model of most academic journals.

Figure 1. The Open Access Lexicon.

Gold OA

Gold OA refers to content that is freely and immediately available directly on the publisher's website, generally under a Creative Commons license that allows for free access and redistribution and, in many cases, for re-use in new or derivative works. Since publishers forgo any revenue under this scenario, they must find another way to finance production costs. In some cases, production costs can be minimized using in-house publishing and an OA diffusion platform such as Érudit. In most cases, and especially for general journals with a broad readership, Gold OA is sustained through an article processing charge (APC), which is a payment from the author, the author's institution or a granting agency to the publisher to cover all or part of the production costs.

The use of APCs means that the costs of publishing are essentially transferred from those who read the articles to those who produce them. The underlying assumption is that funding agencies and research institutions will absorb these costs through research grants to authors. While this does not necessarily pose a significant burden for researchers in science, technology, engineering and mathematics (STEM sciences), the model is more controversial in the humanities and social sciences, where close to 50 per cent of publications are not funded through research grants (Carling et al., Reference Carling, Erdal, Harstad, Knutsen, Lappegård, Skardhamar and Østby2018; Royal Historical Society [UK], 2019). Junior or unaffiliated researchers, as well as graduate students, are at a particular disadvantage under this model, as many cannot afford the APCs charged by some of the top journals (Carling et al., Reference Carling, Erdal, Harstad, Knutsen, Lappegård, Skardhamar and Østby2018; “Open Letter from History Journal Editors,” 2019). APCs for major HSS journals offering Gold OA vary considerably, but our findings suggest that it averages approximately US$2,880 per article.

Other concerns with this model include its potential effects on the quality of peer reviewing. Gold OA publishers have a greater incentive to accept as many APC-based OA articles as possible to support publication costs (Suber, Reference Suber2013). The emergence of predatory journals and “mega-journals” that publish several hundred articles annually on a broad array of topics are by-products of this “author pay” model of OA, under which the incentive structure for the publisher partly shifts from quality to quantity.

Green OA

Green OA (also known as self-archiving) is the practice of making a version of an article or book freely accessible online through an institutional or personal repository. The archiving of documents on commercial sites such as Academia.edu, SSRN or ResearchGate is not normally permitted under most Green OA policies. Publishers have varying copyright agreements that stipulate the particular version of the article that can be posted, as well as different timelines for when it can be posted. In some cases, only the author's version prior to submission for publication can be posted (preprint), either immediately or after an embargo period (6 to 24 months, in general). Several publishers authorize the posting of the accepted manuscript (AM) that includes revisions after having undergone peer review. Finally, some accept the posting of the version of record (VoR)— the published version—complete with volume/issue/pagination and the imprimatur of the journal and its publisher. Under CJPS’s publishing agreement with CUP, authors publishing in the journal can post the accepted version of their manuscript on an institutional repository without an embargo period.

Studies suggest that adopting a Green OA policy has only a limited impact on a journal's subscriptions (Houghton and Swan, Reference Houghton and Swan2013), although this is likely to change if self-archiving becomes more systematic (Anderson, Reference Anderson2017). For OA advocates, the problem with this model is that it does not fundamentally alter the cost structure of publishing. As such, Green OA is generally considered a transitional step, until a sustainable model more consistent with the principles and goals of OA can be adopted.

Interestingly, while an estimated 81 per cent of all research is currently published in journals that permit Green OA after a 12-month embargo, there is a large gap between the actual self-archiving done by authors and the self-archiving that is allowed in publisher policies. According to a review of existing studies completed by Björk et al. (Reference Björk, Laakso, Welling and Paetau2014), only 12 per cent of scientific journal articles were self-archived despite the fairly liberal policies of many publishers, although there is significant variation across disciplines. In short, the vast majority of academic authors do not take advantage of this opportunity. The effectiveness of Green OA policies is therefore likely to be limited in the absence of more coercive rules.

Hybrid models

Hybrid variants of Open Access have emerged in recent years in response to the requirements of a growing number of granting agencies related to OA. Under the most common hybrid model, journals maintain their subscription-based business model but allow both Green and Gold OA options, with the latter requiring authors to pay article processing charges.

The hybrid model offers a good compromise to a pure Gold OA model, but like its Green counterpart, it has been criticized for failing to directly address the rising cost of subscriptions (Matthias, Reference Matthias2018). Publishers may, however, charge twice for the same content (that is, both APCs and subscription fees, referred to as “double-dipping”).Footnote ⁵

Data suggest that the uptake of the Gold OA option in hybrid journals varies across subject areas but that it is generally much lower in the humanities and social sciences where, as previously mentioned, APCs face more resistance. As previously noted, very few CJPS authors have opted to pay the fee required to make their article freely available.

Delayed OA

Delayed OA allows free access to articles after an embargo period, either directly on the publisher's website or on an institutional repository. By definition, delayed OA is incompatible with a definition of OA that mandates immediate access (Suber, Reference Suber2009; cOAlition S, Reference cOAlition2018), but it is seen as a compromise for journals relying on subscription revenues. The Tri-Council statement on OA in Canada supports delayed Open Access, as does SSHRC's revised Aid to Scholarly Journals program. Embargoes are not accepted under the current version of the European Plan S (reviewed below).

The financial viability of this model is dependent on the desirability of paying for immediate access to journal material. This is likely to shift over time as content becomes increasingly available under delayed OA and libraries potentially forgo their subscriptions. Few studies have examined the long-term impact of delayed OA on the subscription base of journals, especially in the humanities and social sciences. King et al. (Reference King, Tenopir, Choemprayongchoe and Wu2009) suggest that about 50 per cent of all accessed and read STEM articles are at least a year old, although there is significant variation across disciplines.

Read and publish

Read and publish (R&P) is not, by definition, an OA model. Instead, it is a pooling mechanism that shifts the burden of APCs from individual researchers to their institutions, mostly through the leveraging of library budgets currently dedicated to journal subscriptions. In essence, an R&P is a negotiated deal between a publisher and a research institution (or a group of institutions) that allows researchers and students in that institution to freely access and publish in that publisher's journals. In exchange, the institution pays an annual fee to the publisher.

Several national research councils and institutes are currently considering pooling resources at the national level to negotiate R&P deals with the major publishers, including in Austria, France and Germany (Green, Reference Green2019). CUP has negotiated a number of R&P deals at the national or regional levels, including with the Max Planck Institutes, the Association of Dutch Universities and Academy Institutes, the Bibsam Consortium in Sweden, the Bavarian State Library and, most recently, the University of California. According to the University of Cambridge:

The R&P model is sometimes criticized for its lack of transparency (Poynder, Reference Poynder2018; Esposito, Reference Esposito2018) because the details of these deals are rarely made public. A recent study revealed that institutions in Germany paid Wiley approximately 26 million euros for an R&P deal to publish an average 9,500 articles per year at €2750 per article. Dutch institutions, on the other hand, are paying €1600 for a very similar deal with the publisher (Kupferschmidt, Reference Kupferschmidt2019).

A related concern about the read and publish model is that it does not necessarily fix the affordability problem libraries are facing. While it does pool costs, it shifts the financial burden elsewhere without actually reducing it. According to some, the R&P system maintains (and encourages) the big publishers’ current business models and serves to “lock their high prices into the new OA environment” (Poynder, Reference Poynder2018).

The asymmetry in the research outputs of different institutions is a further concern (Esposito, Reference Esposito2018). Well-funded universities such as Oxford and Harvard and national consortiums such as the Dutch Academy have much greater bargaining power than smaller institutions. The discrepancy is especially evident for teaching-oriented institutions that are consuming but not necessarily producing articles to the same degree. Unless there is some pooling of resources at the national level, this model may lead to greater polarization between institutions in terms of access to research and capacity to publish.

Trends in OA publishing

In March 2019, close to 13,000 academic journals from 129 countries were listed in the Directory of Open Access Journals (DOAJ, 2019). The number of Gold OA journals (with or without APCs) in the Directory has grown exponentially over the past 16 years, from about 300 in 2003 to over 13,000 in 2019. There are currently 4,707 Green OA repositories listed in the Registry of Open Access Repositories (ROAR), including 47 in Canada, 793 in the United States and 1,582 in Europe.

Open Access is a growing phenomenon that requires context. A 2017 report of the Universities UK Open Access Coordination Group using the Scopus database suggests the growth in the number of OA publications has stalled somewhat in recent years. In 2016, 38 per cent of academic journals indexed in Scopus were still under a classic subscription model. Another 45 per cent were under a hybrid model combining subscriptions and OA with APCs, while 15 per cent were entirely Gold OA (the majority of which charged APCs). These trends are confirmed by several studies on OA using different methodologies (Piwowar et al., Reference Piwowar, Priem, Larivière, Alperin, Matthias, Norlander, Farley, West and Haustein2018).

Our own analysis of some of the major Canadian-based HSS journals and international political science journals confirms these trends. There is a clear predominance of the hybrid model (see Table 1). Out of 30 journals surveyed, 21 offer both subscription-based access and Gold OA, with APCs varying from US$750 to US$3,300, for an overall average of US$2,950. Only one major HSS journal in Canada, the Canadian Journal of Sociology, operates under a fully OA model without APCs. Three journals, including the SQSP's Politique et sociétés, have adopted an OA model that provides free access to articles after an initial 12-month embargo.

Table 1. Status of Selected Canadian HSS and International Political Science Journals, 2019

Note: APC, article processing charge; CUP, Cambridge University Press; HSS, humanities and social sciences; MS, manuscript; PSP, Presses de Sciences Po; T&F, Taylor and Francis; UTP, University of Toronto Press; VoR, version of record.

Almost all journals also have a Green OA policy, although there is variation in the version of the article that can be uploaded to a public repository and in the length of the embargo. Under its publishing agreement with CUP, the CJPS has a comparatively liberal Green OA policy. Accepted manuscripts (unformatted) can be deposited to an institutional repository without an embargo period. It is also worth noting that there is no clear correlation between OA permissiveness and the impact factor of journals.

If we focus on individual articles rather than journals, we observe a similar but slightly slower uptake of OA. Depending on the methodology used, studies situate the total proportion of OA articles in scientific journals to lie somewhere between 30 per cent and 40 per cent. The European Commission's most recent report on OA trends (2019) found that 14 per cent of articles published between 2009 and 2017 in Scopus were available via Gold OA (through pure or hybrid journals) and another 24 per cent of articles through Green OA.

The uptake of OA is also not uniformly distributed across disciplines and countries. Open Access adoption is particularly low in the humanities and social sciences, where no obvious business model has emerged to support the cost of publishing quality peer-reviewed OA journals. The discipline of political science has been especially slow to adopt OA, ranking 34th out of 39 disciplines surveyed (European Commission, 2019). For example, between 2009 and 2017, 3.4 per cent of political science articles in the Scopus database were available via Gold OA and another 10.3 per cent via Green OA, compared with 18 per cent and close to 40 per cent respectively for biological sciences. Looking at country-specific data for the same time period, 62 per cent of Canadian-based articles in the database were behind a paywall, placing Canada 32nd out of 36 countries surveyed.

Maximizing the impact of research through its free and immediate diffusion is one of the core arguments for supporting Open Access (Rentier, Reference Rentier2018). Whether authors benefit from publishing in OA or hybrid journals, however, is contested. A recent study by Piwowar et al. (Reference Piwowar, Priem, Larivière, Alperin, Matthias, Norlander, Farley, West and Haustein2018) suggests that OA articles are cited 18 per cent more often than non-OA articles but that the impact is not consistent across all OA models. Articles in Green and hybrid OA journals are cited more often on average, while articles in Gold OA journals are, in fact, cited less often on average and even less than articles found behind paywalls. Gold OA arguably suffers from its association with predatory journals (Rentier, Reference Rentier2018). Contested is whether the citation advantage is due to OA or other factors, such as the nature and quality of the different publications (Li et al., Reference Li, Wu, Yan and Li2018).

There are few studies focusing specifically on political science journals, but Atchison and Bull (Reference Atchison and Bull2015) find a clear citation advantage for articles published in top-ranked political science journals with a liberal Green OA policy. For Canadian HSS journals, Érudit data between 2011 and 2015 suggest that articles with immediate OA are downloaded 2.3 times more often in the first year versus those under an embargoed paywall; the gap only closes after 5 years (Cameron-Peasant, Reference Cameron-Peasant2017).

OA mandates

Open Access mandates, under which granting agencies make OA a condition for receiving funding, are one of the key drivers of the adoption of OA policies. According to Larivière and Sugimoto (Reference Larivière and Sugimoto2018), at least 700 research institutions worldwide have mandated some form of OA for the work they support.

In Canada, as of 2015, the Tri-Agency Open Access Policy on Publications requires that peer-reviewed journal publications resulting from Tri-Agency (NSERC, SSHRC or CIHR) grants be freely accessible online (through public repository or directly on the publisher's website) within 12 months of publication (Government of Canada, 2016). APCs are an eligible research expense for grants from all three agencies. The Fonds de recherche du Québec–Société et culture adopted a similar policy for its own grant recipients in 2019. The CJPS’s hybrid model, which allows Gold OA with APC and the archiving of the accepted version of a manuscript on a public repository (Green OA), is currently compliant with this policy.

The new OA requirements of SSHRC's Aid to Scholarly Journals (ASJ) grant target journals rather than authors. To be eligible for public funding, journals must make their scholarly content freely accessible on the publisher's website, either immediately upon publication or with an embargo of no more than 12 months. The value of the grant is calculated based on a fixed rate for each approved peer-reviewed scholarly article to be published during each year of the grant period. SSHRC explicitly promotes the migration of journals to Canadian-based not-for-profit digital content and diffusion platforms, such as Érudit, with additional funding valued at CDN$5,000 per year. The agency does not consider policies allowing the archiving of manuscripts on public self-archiving repositories (Green OA) to be compliant with its new requirements. Journals that received a grant in 2019 have two years to comply with this mandate or they forfeit the final year of funding.

Reaction to SSHRC's new AJS requirements has been varied. Many OA advocates argue that it is too permissive in that it tolerates embargoes. Publishers and scholarly associations have responded critically, but on different grounds. The Canadian Association of Learned Journals (CALJ, 2019) and a number of journal editors and publishers believe that SSHRC may have underestimated the impact of the policy change on the viability of Canadian HSS journals, ones with a broad international subscription base but a limited capacity to flip to a pure Gold OA with APCs model. The perceived desire to support smaller journals with low production and diffusion costs—ones more easily supported on Érudit-type platforms—comes at the expense of ones with a more global reach and a more substantive subscription base. The inflexibility and one-size-fits-all nature of the policy has also been criticized by a number of Canadian associations and journals. As we discuss below, SSHRC may have also underestimated the destabilizing impact of this shift on scholarly associations such as the CPSA that rely on steady revenues from their journals to support other activities and their members.

Plan S

Plan S is an initiative of cOAlition S, a consortium created by the European Research Council that has won support of some of the largest European national funders and a number of charitable funders (Science Europe, Reference Europe2018). While not yet finalized, Plan S in its current version requires research funded by signatory agencies to be published in journals that make scholarly content immediately accessible, free of charge, to readers. Two OA models are explicitly accepted, Gold (with or without APCs) and Green, so long as a copy of the final manuscript or the version of record of the published article is immediately deposited in a publicly available institutional repository (cOAlition S, Reference cOAlition2018). In its current version, Plan S prohibits publication in hybrid journals that maintain a subscription base unless they have a plan to fully transition to OA by 2025.

The impact of Plan S is hard to measure at this point. Its strength lies in its broad endorsement by a number of funders in Europe, but many have faced strong negative reaction from their respective research communities and have backtracked. The short timeline and the lack of a clear transition plan are core concerns, but also noted are the substance of the plan itself and its possible repercussions (Kamerlin et al., Reference Kamerlin, Wittung-Stafshede, Dey, Wells, Gruden, van der Kamp, de Bruin, Holbrook and Hay2018; Else, Reference Else2018). Publishers of some of the most important scholarly journals, including Nature and Science, suggest that their upfront costs are too high to abandon subscriptions altogether and that they will not comply with Plan S unless the position on hybrid journals and APCs is softened (Else, Reference Else2019).

Others question the impact of Plan S on academic freedom, since researchers receiving funds from agencies that have endorsed the plan will have fewer outlets for publishing their research (British Academy, 2018; CALJ, 2019; AHA, 2019; University of Cambridge, 2019). In 2017, only 15 per cent of scholarly journals indexed in Scopus were Plan S compliant (UUKOAC, 2017). Further, the requirement for an open licence to share and adapt the work for any purpose (including commercial) could take away author rights to intellectual property (LSA, 2018).

Some of the strongest critiques of Plan S come from scholarly associations and journal publishers in the humanities and social sciences. Similar to the new SSHRC ASJ policy, critiques focus on the one-size-fits-all approach in Plan S, a model largely based on the experiences of STEM disciplines (Royal Historical Society [UK], 2019; University of Cambridge, 2019). Because APC-based OA has seen limited uptake in the humanities and social sciences, there is worry that a number of highly valuable and intellectually important journals will come under financial stress. In an open letter on Plan S, a group of HSS journal editors otherwise supportive of OA argue that Plan S “offers no suitable mechanism for the HSS community to transition their publications on a large scale to sustainable forms of open access” (Plan S Open Letter, 2018). Since a majority of researchers in the humanities and social sciences are unable to pay APCs, they argue that “it is impossible for existing subscription journals to flip to an author-pays model.” They further point to the fact that HSS journals normally serve relatively small research communities, often national in focus. They therefore cannot rely on high volume based on APCs to compensate for the loss of subscription revenues. A number of academic societies, editors and publishers in the humanities and social sciences have highlighted similar concerns, among them the Canadian Association of Learned Journals (CALJ, 2019), the University of Cambridge (2019) and the American Historical Association (AHA, 2019)—to name just a few.

As in Canada, concerns have been raised about the unintended consequences of Plan S on the ecosystem of national HSS societies. As the CALJ (2019) argues, “many scholarly journals in the social sciences and humanities (SSH) are published by not-for-profit scholarly societies, small and medium sized not-for-profit publishers, and university presses” that rely on subscription income to offer a broad range of services.

1) Adopt a Gold OA model

A shift to a Gold OA model is an “all-in” scenario, under which journals fully embrace Open Access. This is the riskiest scenario in both the short and long terms but the one most fully consistent with OA principles. It may well become increasingly adopted by political science or other HSS journals in the future, but to date very few have adopted an entirely Gold OA model, either with or without article processing charges. There is little evidence that the “pure gold” APC model is viable in the humanities and social sciences, except possibly for new journals that have yet to develop an established readership and without a long-held reputation such as that enjoyed by the CJPS. With some exceptions, these tend to be either niche journals that serve a specific purpose and audience or mega-journals that publish large numbers of articles (sometimes a few hundred a year) on a broad array of topics. There are also, of course, a growing number of APC-based journals in the HSS that are predatory journals, with low or no peer-review standards. Ultimately the key concern under this model is the capacity and willingness of our authors, especially graduate students, junior faculty and unaffiliated researchers, to pay APCs.

A clear, sustainable path to implementing Gold OA may well emerge in the coming years. Read and publish deals may, for example, evolve into a sustainable, more systematic model. For now, we believe that there are greater risks than benefits for journals with profiles similar to the CJPS in adopting full Open Access, with or without APCs.

2) Move to an OA-friendly Canadian-based platform

While a move to Gold OA is not viable for journals like the CJPS in the short term, they could consider embracing SSHRC's preferred course of action and move operations to a Canadian-based digital platform for editorial and diffusion purposes. Érudit is the most obvious option among several alternatives, and it returns approximately 75 per cent of subscriptions and licensing rights income to its journals. As mentioned, SQSP's Politique et sociétés uses the Érudit platform, a move that allowed the journal to access professional editing tools and software and facilitated production of a high-quality journal at minimal cost. Politique et sociétés is a hybrid journal employing subscription with a 12-month OA embargo. As such, the journal is compliant with SSHRC's ASJ requirements, although not with the proposed Plan S in its current form.

This option nevertheless has several disadvantages. The CJPS, for one, is a fundamentally different journal than Politique et sociétés. Although bilingual, the CJPS is a member of the English-speaking scientific community and therefore its market and reach are vastly different. This is reflected in the greater size of its readership and its subscription base. A key consequence of a move to Érudit or an equivalent Canadian-based platform would be to jeopardize the established international reach of the journal. Canadian not-for-profit platforms cannot compete at this point in time with publishers such as CUP in terms of promotional services, data analytics and access to international markets.

3) Maintain the status quo

Another option is simply to adopt a “wait and see” approach and adjust as needed. There is no clear risk-free option available for journals such as the CJPS under OA, and given that the hybrid model that is currently adopted by most journals works, there seems little incentive to change it. The risk is to lose access to potential authors in the medium term, including European researchers needing to abide by the requirements of Plan S by 2025. More significantly, it also means forfeiting the SSHRC ASJ grant as of 2021. Not all journals and associations are in a position to absorb such a significant and immediate financial loss.

There is nonetheless a certain degree of comfort with the familiarity of the status quo. The financial implications are not insignificant but at least they are predictable. This is the approach chosen by some Canadian journals, including the Canadian Journal of Economics, that have decided not to renew their SSHRC ASJ grant. In addition to immediate financial stress, this option also sends a strong message of resistance to the OA movement. Open Access has its disadvantages, but it nevertheless embodies a set of values that scholarly associations such as the CPSA and SQSP share in terms of diffusion and access to publicly funded research. Staying the course also potentially puts the journal on a path that may be hard to reverse over the long term, as the OA landscape continues to evolve.

4) Adopt an incremental approach

An additional option is to approach OA requirements with an open but cautious mind. As shown, a number of HSS journals have already moved to a hybrid format that includes a fairly liberal Green OA option for authors. There is no evidence that this has negatively affected their subscriptions (unlike the impact of reductions in library budgets), although it may affect royalties from rights and permissions in the long term.

Journals such as the CJPS may therefore opt to adapt to the new SSHRC AJS requirements, without entirely forgoing their subscription model. This would mean keeping their current hybrid model, with Green and APC options, while giving free access to articles after a 12-month embargo, beginning in 2021. Under this model, previously published articles would remain gated and continue to generate revenues from subscriptions as well as reproduction rights. The journal would remain eligible for the SSHRC grant, pending any additional policy changes.

There is a risk that adopting a 12-month embargo will negatively affect institutional subscriptions over time, but this impact is unlikely to be felt immediately, given that institutions will likely be slow to adjust to the change. The impact on subscription revenues will be somewhat mitigated for journals like the CJPS that are part of a publisher's subscription bundle, under which institutions subscribe to a large number of journals in a package deal rather than individually. The greatest impact of this option will likely be felt after a few years, when the number of freely available articles reaches a critical mass. Revenues from licensing and reproduction rights will then start to decline. Based on CUP's projections, the CJPS revenue losses are estimated at 15 per cent over the next five years under this scenario (CUP, 2019).

Adopting an incremental approach involves some uncertainty, but it provides journals with an ability to adjust as necessary along the way. Should the financial impact be greater than anticipated, the 12-month embargo could be reconsidered. This flexibility is absent under the full OA option, which is much harder to reverse once subscriptions have been eliminated. This option also allows for the development of alternative approaches to classic subscriptions, such as read and publish deals.