The role of open access in a changing academy: reflections on a new publishing paradigm

Now that Regional Studies, Regional Science is into its fourth year of publication, we feel it is a good time to reflect on the progress of the journal to date in relation to the wider publishing context and changing expectations about the role of open-access (OA) publishing. This editorial also marks the departure of Alex Singleton, one of our founding editors-in-chief, and the arrival of Stephen Hincks as a new editor-in-chief. Stephen joins Alasdair Rae and Heather Stephens (North American editor) on the senior editorial team and we welcome him to the RSRS family. At the same time, we extend warm wishes to Alex and formally thank him for his role in taking the journal from initial concept to the stage where we now have a growing corpus of interdisciplinary papers and very strong download figures. In fact, as we explain below, some of our papers are among the most viewed across the social sciences, thanks to the OA model and continued support from our publisher (Routledge/Taylor & Francis Group) and, of course, the Regional Studies Association. We are pleased to report that in March 2016 RSRS was accepted for inclusion in Scopus, the world’s largest abstract and citation database. This is an important milestone as we are one of the few OA journals in the social sciences to achieve this. Below, we also say more about who we are ‘open’ to, since this is often a subject of some debate and more than a little mystery. In this editorial we also make a renewed call for papers from people working in the broad areas of regional studies and regional science, at all career stages. We remain committed to our mentored route for publication, led by the early career papers editors, but we encourage all scholars – from doctoral students to professors – to submit their work. We offer field-leading review times (currently averaging 35 days) and your work will sit alongside papers by such luminaries as Rob Kitchin (Kitchin, Lauriault, & McArdle, 2015), Alan Wilson (2016) , Ana Rita Cruz (2014), Kevin Ward (Ward et al., 2015) and Martin Jones (Beel, Jones, & Jones, 2016). One particularly appealing feature of RSRS is that we publish a variety of paper types, from our ‘regional graphics’ of two or three pages (including maps or graphics) to full-length more traditional manuscripts of around 8000 words. We are also very happy to accept suitable papers of intermediate length, which are typically contributions of 3000–5000 words. We welcome more technical empirical papers, review papers, policy-focused papers and contributions that speak to a wide audience on


INTRODUCTION
Now that Regional Studies, Regional Science is into its fourth year of publication, we feel it is a good time to reflect on the progress of the journal to date in relation to the wider publishing context and changing expectations about the role of open-access (OA) publishing. This editorial also marks the departure of Alex Singleton, one of our founding editors-in-chief, and the arrival of Stephen Hincks as a new editor-in-chief. Stephen joins Alasdair Rae and Heather Stephens (North American editor) on the senior editorial team and we welcome him to the RSRS family. At the same time, we extend warm wishes to Alex and formally thank him for his role in taking the journal from initial concept to the stage where we now have a growing corpus of interdisciplinary papers and very strong download figures. In fact, as we explain below, some of our papers are among the most viewed across the social sciences, thanks to the OA model and continued support from our publisher (Routledge/Taylor & Francis Group) and, of course, the Regional Studies Association. We are pleased to report that in March 2016 RSRS was accepted for inclusion in Scopus, the world's largest abstract and citation database. This is an important milestone as we are one of the few OA journals in the social sciences to achieve this. Below, we also say more about who we are 'open' to, since this is often a subject of some debate and more than a little mystery.
In this editorial we also make a renewed call for papers from people working in the broad areas of regional studies and regional science, at all career stages. We remain committed to our mentored route for publication, led by the early career papers editors, but we encourage all scholars -from doctoral students to professors -to submit their work. We offer field-leading review times (currently averaging 35 days) and your work will sit alongside papers by such luminaries as Rob Kitchin (Kitchin, Lauriault, & McArdle, 2015), Alan Wilson (2016) , Ana Rita Cruz (2014), Kevin Ward (Ward et al., 2015) and Martin Jones (Beel, Jones, & Jones, 2016). One particularly appealing feature of RSRS is that we publish a variety of paper types, from our 'regional graphics' of two or three pages (including maps or graphics) to full-length more traditional manuscripts of around 8000 words. We are also very happy to accept suitable papers of intermediate length, which are typically contributions of 3000-5000 words. We welcome more technical empirical papers, review papers, policy-focused papers and contributions that speak to a wide audience on important contemporary topics. Our geographical focus does of course remain at the 'regional' level, but as stated in our 'aims and scope' this takes a broad view, from Chinese provinces such as Guangdong, with populations of over 100 million, to US counties or European NUTS-3 regions with only a few thousand people.
In the next section we reflect upon the issue of 'open', to whom we are open and for what purpose. These are important questions -too little discussed -but we think it is important to be clear here so that readers and contributors understand that OA publishing is not simply a box-ticking exercise to please funders or politicians but instead part of a fundamental realignment of how academic publishing operates and how knowledge is shared. Following this, we consider the wider OA context by looking at some data from the Directory of Open Access Journals (DOAJ), in which RSRS is included. This highlights the fact that, contrary to what we might expect, there are in fact thousands of OA journals, many of which have existed for more than a decade. Following this, we then look at RSRS in relation to our reach to date. It is very pleasing to be able to report that our most viewed papers are within the top 5% of all scholarly papers terms of the attention they receive. We end this editorial by highlighting the benefits of OA and making a renewed call for papers.

OPEN TO WHOM, OPEN FOR WHAT?
The first thing we want to emphasize here is that OA publishing is far from being a new phenomenon, but take-up varies hugely by discipline. In physics, mathematics and computer science, for example, scholars were 'doing' OA before that term was even in common usage and it is a firmly embedded practice within those disciplines. The arXiv project (pronounced as 'archive') was founded more than 25 years ago as an OA repository by the physicist Paul Ginsparg of Cornell University. Yet, it has taken several decades for OA to be more widely adopted. And, in our disciplines, the OA model often remains at the margins of mainstream publication practices.
When discussing OA and to whom we are making our research 'open', it is helpful to consider the incredible diversity of research conducted by scholars at institutions across the world. Although it might well be the case that a specific research agenda engages with a relatively small pool of scholars, our view is that the value of knowledge should not be measured according to the size of the audience but to whom it reaches and impacts. Disseminating through open channels is the only way to promote access for all. In the fields of regional studies and regional science, much recent work has focused on ideas related to 'smart cities', 'big data' or 'smart specialization'; concepts which have direct relevance to the real world and are widely adopted by city and regional authorities across the world. This raises the important question of audiences, and for RSRS we see our audience as varied and multiple, including: • other academics without access to traditional, subscription-based journals, and particularly academics working in developing nations whom this disproportionately affects; • professionals working in fields such as urban policy, regional policy, city planning and economic development; • students at all stages of study, from all parts of the world; and • members of the public with an interest in the subject areas we publish.
There are various levels of OA journals. 'Gold' OA journals like RSRS publish papers soon after acceptance and make them freely available in perpetuity to anyone with an internet connection; thus, our potential reach is much wider than traditional journals.
However, just being open and available is not enough. We need to think also about the value of OA. This raises the question of 'open for what?' We like to think of this in the context of 'knowledge as a public good', as described by Suber (2009). Like all scholarly publications, we seek to publish papers that address important questions or report the results of recent research, but we also seek to have influence beyond the bounds of academia. This is made possible by our open model but also through faster publishing and more flexible paper formats. In this sense, we are closely aligned to the Public Library of Science (PLOS) OA approach, which highlights 'accelerated discovery', 'public enrichment' and 'improved education' as the three mainstays of OA publishing. Therefore, as editors we do not make publishing decisions solely based upon the perceived impact of a paper (e.g., 'will this paper be widely cited?' or 'is this ground-breaking?'), but rather we focus on the quality of the underlying scholarship, the potential end use, and whether a paper speaks to a certain group of people or comments on a currently topical issue. This is partly because we do not need to worry about page budgets and space within a hard copy journal (we are entirely online), but also because we want to reach places and people that traditional journals cannot. The triple helix of discovery-enrichment-education is then, in a sense, the DNA of OA and we think it serves as a useful frame of reference for understanding why researchers might want to publish using this model. But what reasons do people have for publishing in OA journals such as ours? It seems that some of the most important to date are as follows: • a requirement of a funder (e.g., Research Councils UK in the UK or the National Science Foundation in the United States) • an institutional requirement (e.g., in 2008, Harvard University's Faculty of Arts and Sciences voted unanimously in favour of a blanket OA policy) • reach: scholars want to ensure their work has as large an audience as possible and Gold OA does this (see below for some of our download statistics) • speed: OA journals such as ours often offer much faster turnaround times than traditional outlets • to speak to audiences beyond your immediate, or known, network: we think of this as speaking beyond the 'echo chamber' • retention of copyright: as with most OA journals, authors of papers in RSRS retain the copyright under a Creative Commons Attribution licence.
This requirement-reach-speed-audience-copyright rationale has a variety of permutations, but scholars considering an OA approach should also be aware that there are a very large number of OA outlets, and not all offer the same quality or benefits. The next section takes a closer look at the 'state of open' by data published by the DOAJ.

'THE STATE OF OPEN': OPEN ACCESS BY THE NUMBERS
As of February 2017, there were 9453 OA journals listed on the DOAJ, a community-curated online directory that indexes and provides access to high-quality, OA, peer-reviewed journals. The current list does not include all OA journals in existence, but the list is very comprehensive for journals that meet their acceptance criteria (DOAJ, 2017). This database provides an excellent overview of the current state of OA publishing and is an invaluable resource for anyone seeking to learn more about it. We have produced a summary of 10 key statistics below in order that RSRS readers might better understand the global OA landscape: • 9543: total number of OA journals listed by the DOAJ.
• 120: number of countries that are home to an OA publication listed by the DOAJ.
• 937: number of OA titles from Brazil (the most), followed by the UK (809), the United States (650), Egypt (598) and Indonesia (538). • 42%: percentage of DOAJ titles using double-blind peer review (2973 out of the 7043 where the peer review process is known): 31% have blind peer review, 24% have peer review of some form and 1% have open peer review or editorial review only. • 52%: percentage of DOAJ titles with no article processing charge (APC) compared with 23% with an APC and 25% where it is unknown.
Regional StudieS, Regional Science • 4488: total number of publishers listed on the DOAJ.
• 3716: number of publishers who publish only one OA title (83% of the total). • 7: number of publishers who publish more than 100 OA titles.
• 4827: number of titles listed by the DOAJ where English is listed as the full-text language (51% of the total). • 3742: titles backed by a society or institution (40%); of these, the single largest institutional backer is the European Geosciences Union, which accounts for 31 OA journals listed in the DOAJ.
There are some clear indications that the contemporary OA publishing landscape is different from what came before. For example, the fact that Brazil has the most titles of any nation and that Egypt and Indonesia are both in the top five is perhaps indicative of a global shift in publishing dynamics, at least in terms of numbers. Most OA titles do not charge an article-processing fee (though most major titles do). A significant majority (83%) of all titles are the sole OA journal at a particular publisher. At the other end of the scale there are seven publishers with more than 100 titles, and 92 publishers with 10 or more. Our publisher, Routledge/Taylor and Francis Group, publishes 24 of the titles listed on the DOAJ, mostly in the subject areas of technology, science and medicine. Within this wider context it is easy to see that the term 'open-access journal' is used to refer to publications of many different types and models. Although the total number of OA publications is very large, not all OA journals are the same. RSRS is one of the few broad-based social science journals backed by an international learned society (the Regional Studies Association), published by an established global publisher, and with blind-peer review and a dedicated team of editors and associate editors who ensure the quality of peer review is comparable with traditional journals. This is not the case with many existing OA journals, of which there are many. This, then, gives rise to the question of whether this enables us to reach a wide audience. A closer look below at the data for RSRS illustrates that our reach does, indeed, extend far and wide.

RSRS BY THE NUMBERS: ARTICLE VIEWS AND ALTMETRICS
Although RSRS has only been in existence for a little over three years, our download figures are strong and scholars at all career stages have benefitted from publishing on our open platform. We examine this in two ways. Firstly, we consider raw article-level download metrics published on the RSRS website, and provided by Taylor & Francis. Secondly, we refer to altmetrics data to gauge the level of attention our papers have received.
Our most read article is 'Knowing and governing cities through urban indicators, city benchmarking and real-time dashboards' by Kitchin et al. (2017). This paper now has nearly 17,000 views in just over two years. This paper also currently has 21 citations according to CrossRef and it is our most viewed paper. Several other papers also have view counts well into the thousands, with Oliveira's (2015) 'Constructing regional advantage in branding the cross-border Euroregion Galicia-northern Portugal' having nearly 8000 views in just over 18 months. This paper was published via our mentored publication route when the author was in the process of completing his PhD thesis, and demonstrates that there is an appetite and audience for quality research by scholars at all career stages. We are particularly proud of the fact that we are giving voice to both established academics and emerging scholars.
Owing to a variety of factors, there is a lower level of published regional research on topics and areas relating to the global south, so it is pleasing to see that Rasul and Sharma's (2014) 'Understanding the poor economic performance of Bihar and Uttar Pradesh, India: a macro-perspective' is currently our third most viewed paper, with just over 5000 views to date. This supports our goal to have a wider reach and more diverse audience than would have been possible in a traditional publishing outlet. Beyond Rasul and Sharma, our top 10 most downloaded papers all have more than 2200 views, and within this there is a considerable range of subject foci, including papers on regional innovation systems (Komninaki, 2015), benchmarking city-regions (Calzada, 2015) and auto-industry plant closures (Bailey, Bentley, de Ruyter, & Hall, 2014). Furthermore, one of our featured graphics, on the 'Archipelago of Disconnection' of global internet use by Straumann and Graham (2015), is our sixth most viewed paper, with 2500 views to date. In terms of geographical focus, there is also a good range of diversity, with the top 10 features papers focusing on Ireland, Portugal, India, Australia, Greece and the Philippines.
There are a variety of ways to measure impact and reach, but in today's multifaceted and fastpaced publishing world, altmetrics provide a very useful barometer of a paper's success. Altmetrics are, put simply, non-traditional metrics that differ from commonly used measures like citations or impact factors in that they relate to individual papers and consider a much broader range of impacts. All RSRS papers have an individual altmetric score embedded within the home page for each article on our website. Altmetrics provide tracking and analysis of the reach of an individual paper and although the term 'altmetric' is generic, our altmetric tracking is provided by altmetric. com, the world's foremost altmetric tracking company. For each article tracked, an 'altmetric attention score' is computed, based upon a weighted combination of how many times a paper has been covered by a range of media sources, including the news, blogs, Twitter, Facebook, Sina Weibo, Wikipedia and LinkedIn. The largest weighting is assigned to news coverage and blogs, followed by Wikipedia entries and policy documents. Altmetric attention scores are reported as a whole number and -in general -if an article score 20 or more it is doing far better than most of its contemporaries.
Our most viewed paper (Kitchin et al., 2017) also scores highest on altmetrics, with an attention score of 233. This paper has 13 citations on Scopus, 152 readers on Mendeley (a reference manager), has been tweeted 727 times and is in the top 5% of all research outputs scored by altmetric, across all subject areas. The next highest paper scored by altmetric is Straumann and Graham's (2015) regional graphic, with an attention score of 101. This is particularly pleasing for two main reasons. First, it is a validation of our decision to include this paper type within RSRS from the outset. It shows that there is an audience for this kind of material and that, by offering a diversity of publishing experiences, we can attract scholars of the calibre of Straumann and Graham, leading digital researchers based at the University of Oxford's Internet Institute. Secondly, the data indicate that 81% of the attention came from members of the public, which is impossible to conceive of had this been published in a traditional, non-open outlet.
The altmetric attention score for Rasul and Sharma (2014) on regional economic development in India at 50 is also well within the top 5% of all research outputs scored by altmetrics. The data here indicate the 93% of the attention was generated by members of the public, the majority of whom were based in India. Although these data contain a measure of uncertainty, they provide further validation that the OA model of RSRS is one that can reach diverse global audiences and -crucially -people who we may not otherwise be able to influence. Given this situation, in the final section below we make a renewed call for papers from all parts of the world and all areas of regional studies and regional science.

A RENEWED CALL FOR PAPERS
It is clear from our review of OA publishing that there are many thousands of outlets jostling for position in an increasingly crowded knowledge space. Even within the social sciences, there are hundreds of potential OA outlets. But when we look more closely at the data, as we have done above, it becomes clear that despite the many and varied forms of OA publishing, there are actually very few dedicated social science OA journals relevant to regional studies and regional science scholars. Certainly, there is no other journal in our field backed by a major international learned society, published by a global publisher, and with peer review, with such strong article-level metrics. This then answers the important second question of OA publishing. The first, which we have answered above, is 'why should I publish in an OA journal?' The second question is 'where should I publish?' If, as stated above, discovery, enrichment and education are at the heart of the OA model, then it is essential that OA journals have scholarly standards, maintain the version of record of a paper, and curate content that is discoverable and digestible online. It is our hope that after reading this editorial more people will understand the benefits of OA publishing and be persuaded by its potential to enrich and educate new and diverse audiences. This being the case, we make a renewed call for papers here.
Like any scholarly journal, we are keen to ensure a good flow of papers, and this has been the case to date. In volume 1, we published more than 300 pages, and in volumes 2 and 3 this increased to more than 500 pages each. If we receive papers of sufficient quality, we will grow this further. We welcome submissions from authors working on regional issues in geography, economics, planning and political science -and beyond. If your research has a regional focus and speaks to the kinds of topic already covered in our pages, we encourage you to submit. As stated in our aims and scope, RSRS allows for a range of article types -including full-length research articles of up to 8000 words, shorter policy briefings of around 3000 words and 'regional graphics', which capitalize on the growing importance of data visualization in our field. We also continue to offer a mentored submission route for early career academics (doctoral students and recent doctoral awardees).
All papers must adhere to our threefold publication criteria -research must be within the scope of regional studies and regional science, have a commitment to methodological rigour, and be communicated clearly and succinctly. The latter criterion in particular is important in an open journal, since we are seeking to speak not only to academics but also to the wider world, which includes people who almost certainly do not value the often abstruse language found in many traditional journals. This being said, we end this editorial by making a renewed call for papers in the following areas: • We would like to see more submissions with an explicit policy focus. • We particularly encourage shorter submissions, of around 5000 words. We believe there is significant value in papers of this length as they often provide succinct summaries of important research. • We are also very interested in having more regional science scholars submit work to RSRS. We have a growing body of more technical regional science work in RSRS (e.g., Nishida & Yoshida, 2016) and we are eager to publish more. • We would also like to encourage scholars to submit what we might term 'risky' or 'different' papers that perhaps do not fit in other journals. An example here might be Makkonen and Williams (2016) paper on 'Border region studies: the structure of an "offbeat" field of regional studies' which we thought was a particularly useful contribution, but of a type not often published elsewhere. • Finally, we encourage contributions on current 'hot topics'. We do so because, unlike many traditional journals, we can get work online in a short period of time. Depending upon the length of the review and revision process, a paper could be online within four months of submission. This means that RSRS can help set the agenda on important contemporary topics. One such example might be the regional dimensions of 'Brexit', which are likely to reverberate far and wide.
This renewed call applies to scholars regardless of career stage or national origin. If you are considering publishing in RSRS but need further advice, the editorial team will be happy to discuss