The authors have declared that no competing interests exist.
This paper provides an institutional and empirical analysis of the highly concentrated market of academic publishing, characterized by over proportionally high profit margins for publishing companies. The availability of latest research findings is an important issue for researchers, universities and politicians alike. Open access (OA) publication provides a promising but also costly solution to overcome this problem. However, in this paper we argue that OA publication costs are an important, but by far not the only way for academic publishers to gain access to public funding. In contrast, our study provides a comprehensive overview of the channels through which public expenditure benefits big academic publishing companies. Furthermore, we offer the results of an explorative case study, where we estimate the annual financial flows of public expenditures in Austria for the field of social sciences. In all, these expenditures add up to about 66.55 to 103.2 million € a year, which amounts to a fourth of total public funding for this field. Against this background, we contribute to the debate whether and to what extent public subsidies are justified for economically successful companies.
In the course of the last 25 years the movement for open access (OA) spread the claim for free access to academic knowledge [
The overall success of the OA movement during the last two decades not least manifests in the vast amount of more than 14,000 and about 5 million articles listed in the DOAJ today (summer 2020). Additionally, big science funders started to mandate OA for its grantees (see 13 for the Austrian case) and new platforms for self-archiving (ResearchGate or Academic.edu) or illegal hosts (Sci-Hub and LibGen) challenge the traditional business model of academic publishers. Hence, the debate about OA is closely linked to the question of the public value and thus also the costs of scholarly knowledge as well as debates about the role of academic publishers in this field. While OA provides latest research findings for free for its readers, the costs are shifted to authors and/or their institutions. Particularly the latter–universities, libraries and consortia of both–are being confronted with increasing costs for traditional subscription fees as well as costs for OA publishing. Furthermore, the rising costs of so-called “Big-Deals” between academic publishers and distinct national consortia (e.g. the KEMÖ for the Austrian case, [
The market of academic publishing–comprising both OA and toll-access journals–overall is highly concentrated and potentially offers monopoly rents for the top publishing companies. Against this background several authors criticized the “black box” of costs for academic publishing for charging excessively, including double-dipping [
While most critical literature on academic publishing is focused on subscription fees, APCs and the debate on OA in general, i.e. the “revenue side” of academic publishing companies, there is hardly any literature on their respective “expenditure side”. Even critics hardly point to the fact, that academic publishers to a large extent benefit from the strong pressure to “publish or perish” [
Against this backdrop, our study is–at least to our knowledge–the first to provide a comprehensive overview of the direct and indirect channels through which public expenditure benefits big academic publishing companies. We complement this framework with the results of an explorative case study, where we estimate the annual financial flows of public expenditures in Austria in the field of social sciences. This way, we aim to provide an empirical basis for the question, whether and to what extent public subsidies are justified for economically successful publishing companies. Moreover, we also make suggestions for a more democratic and egalitarian form of knowledge dissemination and scientific progress, alike.
The remainder of the paper is structured as follows: Section two provides an overview of the field of academic publishing and introduces our model of four channels of access to public funding for academic publishers. In section three we present some characteristics of the Austrian academic publishing market and the institutional state of the social sciences in Austria. Section four offers the main results of our case study. In section five we discuss the empirical results and provide some science policy recommendations.
The role of OA in the debate about the market structure of academic publishing is ambiguous. On the one hand OA potentially challenges the very high subscription fees of conventional academic journals. On the other hand, OA publication comes with other kinds of costs, such as individual article processing charges (APC) or general agreements (“Big Deals”) including free publications for authors. “Big Deals” between Publishing Companies and consortia provide an illustrative example of power asymmetries between buyer and seller, resulting in partly very opaque agreements: “Unfortunately, neither the countries, contracts nor data can be released due to existing non-disclosure agreements–an ironic symptom of the challenges involved in creating a transparent scholarly publishing system” [
Despite the common argument that OA journals are forced to charge (high) APCs in order to maintain high quality standards, several studies reported only very weak or no correlation between quality of journals (measured in journal impact factors) and the level of APC. Contrarily, the level of APCs for publishing an article is more related to the market power of specific academic publishing companies and again differs strongly across disciplines [
In recent decades there is a trend to understand the academic field more and more as a competitive market, mediating the scarce resource of scientific prestige. The commodification of scientific knowledge not least manifests in the denomination of the field of academic publishing as “a market”. While we share the critical stance towards this process (e.g. [
Throughout, one can distinguish between four different actors within the debate evolving around OA, who have partly opposing goals, claims, possibilities and perspectives: (i) authors, (ii) publishing companies, (iii) (public) funding agencies (iv) universities and libraries and (v) the scientific community.
Against the backdrop of these mutual relations in the field of academic publishing, we first discuss the motivations of the main actors and then show what they offer and receive from each other.
The conflictual interests of different actors in the field of academic publishing strongly influence their stance towards OA. On the one hand a shift towards OA publishing provides a promising solution to the affordability problem of
As outlined above, the field of academic publishing reflects several trends of economization, numeric evaluations and impact rankings, which increased competitive struggles among researchers and institutions [
In all, the combination of the incentive structures of the current academic system and the intrinsic motivation of individual researchers and academic institutions, offers a very lucrative business model for a small number of top academic publishing companies. In order to provide a more systematic perspective on this process, we distinguish four main channels through which publishers can receive or tap into public funding:
A similar study was conducted by Lawson et al., who examine UK publishing markets including three broad types of financial flows (institutional income, subscription payments and APCs) between the different actors involved in scientific publishing markets [
Against the background of our analysis of power differentials and financial flows in the field of academic publishing, we employ our schematic “channels-model” for an explorative case study of public funding of research output in the field of social sciences in Austria. Hence, we contribute to the debate on the role of public funding in academic publishing and furthermore provide an original estimation of the actual annual financial flows in this particular market.
While much of the critical debate on academic publishing, i.e. the critique against the increase of APCs and subscription fees initially focused on the natural sciences, the social sciences were confronted with rising costs more recently as well. In a study conducted in 2013 the authors found that about 18% of social science journals enlisted in the DOAJ charge APCs, whereas this share is 80% in genetics [
The social sciences are composed of several disciplines dealing with human behavior in its social and cultural aspects. However, this broad definition aggravates the assignment of distinct sub-disciplines and researchers to the field. In order to arrive at a clear definition of social scientists in Austria, in this survey we use the classification of Social Science by the Austrian Science Fund FWF, which is based on international standards and includes the fields psychology, economics, pedagogy, sociology, legal studies, political science, human geography, media and communication studies and other social studies (see the
We decided to include only researchers with a PhD and thus ended up with a full sample about 1,500 social scientists in Austria. In sum, accounting for 427 out of 2617 professorships at 22 public universities in Austria, the social sciences constitute about one sixth of the Austrian research sector [
Big deals with publishing companies in Austria are negotiated by the KEMÖ (Austrian Academic Library Consortium). Currently 58 Austrian libraries are part of the consortium and their contracts include 61 publishers. Four of the big five publishers in general and social sciences (ACS, Wiley, Springer Nature, Taylor&Francis and SAGE) have contracts with the KEMÖ that include open access agreements. In general, the deals have a term of three years. The precise amount of money negotiated is confidential [
The aim of this study is to highlight structures of financial flows in scientific publishing for the field of social sciences in Austria. Based on the four-channels-model (see
This representation is based on [
The study rests on two methodological pillars: First, we examined several previous studies concerned with the topic of publishing costs in a systematic framework and categorized them in accordance with our four-channels-model. Secondly, we gathered primary data from a questionnaire study conducted among a full sample of Austria social scientists to supplement our analysis. We arrive at an estimation of public expenditures related to payments according to the four-channels-model by combining those two sources of information. The presented results heavily depend on the given institutional and cultural context of Austrian scientific publishing in social sciences. We therefore do not claim for representativeness or universality of our results. Still, the case study contributes to the overall discussion by illustrating an empirical example of theoretical considerations about power concentration and oligopolistic structures on scientific publishing markets [
We started by screening the existing literature on scientific publishing and public expenditures related to publishing. Several studies have dealt with the topic, although from slightly different angles and based on different data sets and time periods. In order to achieve better comparability, we limited our analysis to articles which calculate empirical estimates of expenditures of scientific publishing, excluding for example theoretical discussions of market structures. We ended up including 22 articles into the analysis and coded them with regard to statements made about the amount of subscription fees and libraries expenditures (channel 1), APCs and publication fees (channel 2), costs of reviewing (channel 3), and costs/article (channel 4). Most of the studies included in the analysis derive descriptive statistics and estimators of direct costs associated with scientific publishing, that is subscription fees (channel 1) and submission fees (channel 2). Whenever several countries or disciplines were discussed in the original paper, we focused on numbers derived for Austria and the field of social sciences. By combining several different studies and their results we hope to gain a broader and more realistic picture on the annual amounts spent on subscription and submission fees, as well as on the extent of divergence of these numbers between studies using different approaches and data sets. Additionally, we performed an online search for subscription and submission fees published on the journals websites to present the most recent official figures (Table 5).
Several constraints have to be taken into account when comparing the numbers from the literature depicted in
Subscription costs | Libraries expenditure | Publication fees | Average APC | costs of reviewing | cost/article | |
---|---|---|---|---|---|---|
[ |
€ 39,8447– € 225,8116 (*) | € 0.14–0.68 /page | ||||
[ |
€ 1,684.02–3,586 | € 1,242.45–1,391.54 | € 7,857.81–10,265.40 | |||
[ |
72% of publishers revenues | € 822,40–2,4672 | e. g. € 82.24 (paid to reviewers) | € 0.12/ page | ||
[ |
€ 252.47 | € 685.63 | ||||
[ |
€ 1,644.80–3,2896 | € 743.45 (€ 202.31–1,1061) | € 904.48 (*) | |||
[ |
average annual price increase: 13% | € 8.2 billion/ year (€ 575 Mio in SoSci) | e.g. € 2,467.20 (Springer) | € 500 (hypothetical cost estimate) | ||
[ |
€ 1,110.24 /paper (PlosOne); | € 485.65 (for OA publishers) | e.g. € 238.50 (Hindawi), € 24,6720–32,916 (Nature) | |||
[ |
€ 1 Mio/ university /year | |||||
[ |
€ 65–70 Mio/year | € 30 Mio (KEMÖ) | ||||
[ |
€ 30 Mio/year (universities); € 1.5 Mio/year (author’s pay) | € 3.9 Mio /year (*) | ||||
[ |
€ 218 Mio/year—260 Mio/year | € 92 Mio—144 Mio | € 1,100 | |||
[ |
€ 30 Mio/year | € 60–70 Mio. /year | € 0.9Mio—1.5 Mio. / year | € 3.5 Mio / year | ||
[ |
68–75% of journal publishing revenues from library deals | € 4,1145 (Cell Reports by Elsevier) | € 16.46–32.92 /page | |||
[ |
e.g. € 527.09 (economic journal); € 604.83 (AEA) | € 4.9 Mio / year / library on journal subscriptions; e.g. € 9.8 Mio for Elsevier’s collection | € 1,110.92 (PLoS ONE); € 1,657.50 (Economic Journal); | e.g. € 1,234.35 (Wiley) |
||
[ |
€ 1,6574– € 2,7624 | |||||
[ |
€ 418,408 (total costs of full OAfunding) | € 2,376,356 (funding by FWF) | € 1,000–1,250 | € 1,298 | € 4,816.18 (*) | |
[ |
€ 103.8 Mio/year (UK) | € 19 billion (*) | € 1,878.50 | |||
[ |
€ 2,037.63 (*) | €1,481.22–2,304.12 | ||||
[ |
€ 2,240.76–2,634.10 | € 660.55 | ||||
[ |
€ 597 Mio (EU), € 451 Mio (Big5) | € 1,526 | ||||
[ |
€ 2,147.01–2,653.11 | |||||
[ |
€ 41 Mio (€ 28.79/ review) |
In order to achieve better comparability, we converted all costs into Euro at the first published exchange rate in the year of the study. Converted numbers are marked with a (*). For example, if the number of articles and the total volume of revenues were given in the original paper, we calculated the corresponding amount of revenues/article or the mean value of different business model’s amount of fees for this table.
Since the vast majority of articles in the existing literature deals with channels 1 and 2, we decided to collect additional data on channels 3 and 4 in order to estimate implicit public expenditures related to reviewing and producing manuscripts, the main “goods” traded on academic publishing “markets”. Therefore, we collected primary data by designing a questionnaire and forwarding it to all social scientists (see the
In sum, we identified 1496 social scientists affiliated to Austrian universities and invited them to join our survey. The gross response rate was 15.7%, the net response rate 10.56%. Demographics of the participating sample are given in
GENDER | ||||||
90 | 51 | 1 | 16 | |||
POSITION | ||||||
21 | 25 | 44 | 49 | 1 | 5 | 13 |
DISCIPLINE | ||||||
21 | 4 | 35 | 39 | 13 | 4 | 46 |
The main focus of the questionnaire was the amount of time spent on writing a paper, i.e. time spent on producing the publisher’s final good, as well as the amount of time spent on reviewing, i.e. time spent on the quality control of the scientific output of others. This approach aims to capture the proportion of scientific activity in terms of working time that is actually connected with the production of a paper (or a review respectively).
An additional check on the individual’s motivation provides deeper qualitative insights on the incentive structure of individual researchers. When asking if and how important several motivational reasons are to actually perform a review, about 90% (89.87%) stated that financial incentives are unimportant or not important for their decision. Further, fear of potential harm on their personal careers if rejecting to write a review was also rather unimportant or unimportant to a majority of researchers (69.62%). On the contrary, contributing to the quality of science and personal interest in the topic were rated very important and rather important by about 85% of researchers in the sample. These qualitative insights to the intrinsic motivation of reviewers show that market incentives and mechanisms might be largely ineffective in this case, as the supply of reviews is practically independent from the price or compensation paid. Even if compensation is zero, scientists still review for the sake of the scientific quality process and out of personal interest. Other motives often stated in the survey and summarized in
The bars indicate the number of individuals rating the different motives according to their personal importance.
Several authors have tried to come up with an estimate of unit costs of an article for scientific publishing from the publisher’s perspective that is the cost side of publishing. However, such derivations evoke several problems, e.g. due to different structures and sizes of the individual publishers (for a list of arguments against unit cost calculation of articles see [
share of scientific work/day | articles published (2019) | hours/ review | numbers of reviews written (2019) for publishers: | requests for reviews / year | ||
---|---|---|---|---|---|---|
conventional | OA | |||||
46% | 3.37 | 7.35 | 6.5 | 1.05 | 12.43 | |
42% | 3 | 6 | 4.5 | 0 | 7 | |
7% | 0 | 1 | 0 | 0 | 0 | |
91% | 25 | 50 | 40 | 10 | 100 |
For comparison, in 2017 the Austrian Union of University Professors (UPV) published a study on the amount of time spent on actual research activity by university professors [
In order to come up with an estimate of indirect public expenditures on scientific output passed on to publishers, we used monthly gross wages for full time employment according to the collective agreement for public employees and adjusted them for the average numbers of hours employed in our sample and the average share of working time spent on researching and reviewing. The average employment indicated in the sample was 38.09 hours/week. Furthermore, the average wage per hour was calculated combining the 4 different wage models’ (Post-Doc, Associate Professor, Assistant Professor, Professor) minimum wage as derived from the collective agreement for public employees according to the distribution of the 4 positions among Austrian based social scientists (see
Position [Class] | number | Share of researchers | min. wage costs [€] | max. wage costs [€] |
---|---|---|---|---|
427 | 28.54% | 9,303.24 | 13,345.44 | |
112 | 7.49% | 6,787.88 | 8,029.62 | |
244 | 16.31% | 8,786.44 | 12,841.24 | |
713 | 47.66% | 4,842.62 | 7,905.96 | |
1496 | 100.00% | |||
The monthly wages were derived from the website of the union of public employees [
In this section we will proceed as follows: For the estimation of financial funds related to channel 1 and 2 we make use of the available information from our literature survey on publication and subscription costs. Channel 3 and 4 are estimated by combining these results with data from our case study. Hence, applying our four-channels model we come up with a rough estimate of the actual amount of money flowing from public institutions to private publishing companies.
As mentioned above, one considerable restriction in analysing subscription fees is that academic institutions, such as university libraries, often do not pay the list price due to internal deals that cover a great variety of journal subscriptions, often also including extra deals on free submission to some of their journals. Therefore, there is an almost universal demand for more transparency regarding true institutional costs of subscription fees in the literature dealing with open access and scientific publishing. We also tried to come up with actual numbers and contacted the KEMÖ several times, but did not succeed in getting additional information. Further, the range of numbers in
[
Overall institutional expenditure on subscription were calculated by [
The term submission fees (similarly publication costs) typically summarize various forms of payment for submitting to a journal, that is submission fees, Article Processing Charges (APCs), Open Access fees and others. A broad definition is given by any form of payment made by the author (the author’s institution respectively) to make the article available to the scientific community–this implies both, paying publication fees to a journal, or paying additional fees to make the article OA. APCs in the top journals in the field of Social Sciences range from € 2,495 to € 5,350 in 2020 according to the official list prices published on the publisher’s website, as presented in
Rank | Journal | Publisher | submission fees | APC |
---|---|---|---|---|
Academy of Management Annals | academy of management | 0 | No gold OA | |
Annual Review of Organizational Psychology and Organizational Behavior | Annual Reviews | n/a | n/a | |
Entrepren€ship Theory and Practice | SAGE | - | € 2,450 | |
Journal of International Business Studies | Palgrave | - | n/a | |
Journal of Management | SAGE | 0 | € 2,450 | |
Annual Review of Sociology | Annual Reviews | n/a | ||
American Sociological Review | SAGE | € 20.42 | € 2,450 | |
Annals of Tourism Research | Elsevier | - | € 2,352 | |
Information Communication & Society | Taylor and Francis | 0 | € 2,495 | |
Sociological Methods & Research | SAGE | - | € 2,450 | |
Political Communication | Taylor and Francis | 0 | € 2,495 | |
International Organization | Cambridge University Press | - | no OA | |
Environmental Politics | Taylor and Francis | 0 | € 2,840 | |
American Journals of Political Science | Wiley | - | € 2,750 | |
Political Analysis | Cambridge University Press | - | € 2,581.40 | |
Quarterly Journal of Economics | Oxford Academy Press | € 5,350 | ||
Journal of Economic Perspectives | American Economic Association | |||
Economic Geography | Taylor and Francis | € 2,495 | ||
Brookings Papers on Economic Activity | Brookings Institution Press | |||
Journal of Finance | Wiley | € 0–245 | € 2,962 |
Manually compiled list of top 5 of Web of Science ranked Journals in Social Science categories according to their Journal Impact Factor [
The findings of literature study on publications costs provides further insights into the structure and amount of submission fees. Some differences in the level of APCs can be attributed to different modes of access: For example [
Several studies have derived overall estimates for average per article APCs. The lowest estimate is given by [
Additionally, several authors came up with estimates of total costs of institutional expenditures on submission fees and publication costs. For example [
Downscaling the estimate of [
Although much research has been done in the area of subscription and submission fees, the role of the reviewing process has been little studied to date. [
In our sample, 94.94% of scientists indicated to have already produced a review in the course of their career. On average, approximately 16.7 inquiries for writing a review have been received in 2019 by a single scientist and 6.5 reviews for commercial publishers as well as 1.05 reviews for OA publishers have actually been provided in 2019, summing to approximately 7.55 reviews/scientist/year. The estimated time for one review was 7.35 hours on average, which is close to the estimates in the literature.
To calculate the value of reviews provided by Austrian social scientists in one year we multiplied the average hourly wage of min € 41.83 (and max € 62.23 respectively) with the average number of hours spent/review (7.35 h). This results in an approximate value of a single review of € 168 to € 239. Multiplying this with 7.55 reviews written (6.5 for commercial as well as 1.05 for an open access publisher on average in 2019) gives a public expenditure for reviewing of € 2,323 to € 3,456 / year / social scientist in Austria. This again multiplied with the total number of social scientists in Austria (1496) sums up to an annual total value of reviewing by Austrian social scientists of € 3,299,366 to € 4,908,779.
As mentioned above, only few estimates for unit costs of articles can be found in the literature, most of them focusing on the production costs for publishers to edit and publish an article. This way, most of these studies focus on average profit margins in scientific publishing markets and thus aim to identify costs and revenues of publishers. For example [
On average, our respondents indicated a share of research activity that aims at producing a scientific output (journal paper, chapter, book or other manuscripts) on a normal working day by 46%. Combined with the average weekly employment of 38.09 hours, the total number of hours spent on research activity by an average social scientist in Austria is approximately 911 hours/year, producing 3.37 contributions on average in 2019. One weakness of this approach is that non-published, or even unfinished papers are not considered, nor corrected for and that therefore the numbers presented might be overestimated for the numbers of hours spent on producing published articles. Nevertheless, we argue that research on unpublished work might indirectly benefit the published work.
The approximate value of a scientific contribution produced by Austrian social scientists is calculated with a minimum and maximum wage (see
The contribution of this paper is twofold: First our paper provides a comprehensive overview of the direct and indirect channels through which public expenditure benefits big academic publishing companies. In doing so, we build a four-channel-model of publisher’s access to public funding and thus contribute to the political economy of academic publishing. Second, we use this model for an empirical case study of the financial flows in field of social sciences in Austria. While the open access movement initiated an ongoing debate and several positive developments regarding channel 1 and 2 of our model–subscription fees, APCs and submission fees–channel 3 and 4, i.e. the free provision of peer reviews and research papers is largely understudied. Therefore, we supplement the analysis of the “revenue side” of publishing companies (channels 1 and 2) with their “cost side” (channels 3 + 4), which allows us to develop a broader understanding of the political economy of academic publishing.
More specifically, we focus on the role of academic publishing companies. We argue that a very small number of these companies benefit from the strong stratification logic, the “publish or perish” culture in academia and the academic practice of peer reviewing. During the last years, publishing companies have steadily increased the levels of APCs and subscription fees, which has led to severe challenges for universities and libraries. Moreover, recent studies have shown that these “big five” companies make use of the quasi-oligopolistic position to force libraries, universities and consortia to pay increasing fees in order to maintain their main role as provisionary of academic knowledge.
Against this background, our four-channels model identifies the implications of this huge power differentials between publishers and researchers as well as academic institutions on the one hand and several problematic incentive structures in academia on the other hand. Based on previous studies in the field of academic publishing and a questionnaire study of a full sample of Austrian social scientists, we estimate the annual public expenditures reaped by publishing companies. This way, we are able to estimate the amount of public funding which is explicitly (channel 1 and 2) or implicitly (channel 3 and 4) directed to a few dominant publishing companies.
Summing up the estimations for our 4-channels model, our estimation suggests that the Austrian state (indirectly) funds publishing companies in the field of Social Science with € 66.55 to € 103.2 million per year (
An estimation for the social sciences in Austria.
Although our results have to be interpreted with caution–not least due to the lack of transparency in the agreements between publishing companies, universities and consortia–our case study provides some novel insights into the political economy of academic publishing. We argue that in the highly concentrated market of academic publishing a small number of publishing companies (“big five”) benefit from the highly competitive academic culture and the intrinsic motivation of individual researchers.
Against the background of our empirical results we conclude that academic publishing is in urgent need of institutional reform. Apparently, despite the success of the open access movement in recent years, academic publishing is still a very lucrative business for a very small number of private academic publishers. Yet there are already some initiatives to challenge the problematic implications of the huge power differentials in the market of academic publishing and the problematic incentives for academic publishing, alike. While increasing obligations to publish open access could challenge the traditional business model of academic publishers, initiatives for a better and more comprehensive evaluation of research output [
(DOCX)
(DOCX)
(DOCX)
(XLSX)
We want to thank Bernhard Schuetz and Claudius Graebner for their comments and suggestions on an earlier version of this paper. Remaining errors are ours. Furthermore, we also thank Matthias Aistleitner, Carina Altreiter, Susanna Azevedo, Theresa Hager, Katrin Hirte, Anna Hornykewycz, Raphaela Kohout, Sarah Kumnig, Katharina Litschauer, Laura Porak, Ana Rogojanu and Georg Wolfmayr for valuable comments on earlier versions of this paper.