The Two-Way Street of Open Access Journal Publishing: Flip It and Reverse It

Round 1

Reviewer 1 Report

Summary

The article opens a new perspective on the current transition process taking place in scholarly journal publishing, bringing into focus and presenting evidence for the phenomenon of journals transitioning from Open Access back to subscription model. It provides detailed insights into the workings of scholarly publication with an extensive approach in data assembly and analysis. Discussion is focused on key findings of the results, mainly on the problem of sustainability of journals in an Open Access business model, the case of journals published by scholarly societies and the hybrid OA model in general. Overall a timely contribution to the OA transition discussion that could stimulate further research.

Broad Comments

Suggestion: make a reference to your dataset on github at a more prominent place in the article or even publish as supplement – it would be much easier to connect the underlying data to the figures provided in the article. It took me some time to find the relevant link in the references.

This brings me to my second concern: I am missing data labels in many of the figures. For most of the figures, these should be easy to add (figures 3, 5, 6, 7, 9, 10, 15). In Figure 4, which has already some labels, data labels for the set size related to the number of journals in each database could be added. Data labels could make Figure 8 confusing, though. In this case, a data table would be helpful.

Section 4.2.5. Analyzing the affiliation of journals to societies or research institutes, it would be interesting to know which or how many of these journals are published in-house by the societies themselves and how many are published by professional publishers in partnerships.

Section 5.3. You are reasoning that APC pricing is linked to several conditions, among them “to a large extent to the impact factor”, but you support this with only one single source, the Springer Nature IPO prospectus. I would like to suggest an additional source supporting this argument: Schönfelder, N. (2018): APCs—Mirroring the impact factor or legacy of the subscription-based model? DOI: 10.4119/unibi/2931061, as this would make a stronger case for your conclusion. But note that Schönfelder attributes the link more dominantly to pure OA than to hybrid OA, so some rephrasing of your argument could be necessary, unless you can find additional sources. Furthermore, the second part of the Springer Nature quote (lower APCs adversely affecting margins) does not point to the impact factor argument but to market mechanisms and monetizing.

Specific Comments

·         Line 173: Add “whereas” before “it is rare for journals to recover after a period of inactivity.”

·         Line 209: I suggest you begin a new section here. The following is not part of “Sustainability of OA Journals”, which you gave as background, but you present your research questions here and those should be given a new heading.

·         Line 257: How do you define OA in this context? Is the content in the journals free to read as a minimum condition or did you check for cc-by licenses?

·         Line 399: Reference [61] points to the Internet Archive Wayback Machine. Here I am missing a source for the academic disciplines you allocated to the journals. I assume from looking at the github dataset that you used the CTWS Journal Indicators file? What about the sources for the other information you added?

·         Line 407 / Line 559: Can you give a reason for 1985 as a turning point?

·         Line 547 / Figure 10: I agree that for health science the majority of flips occurred after an age of more than 15 years. But is this true for LS journals as well? It seems to me that here about half the journals concerned are above 15 years, the other half is below. Again, data labels could be helpful in supporting your arguments.

·         Line 781: Insert “however,” before “the uptake of hybrid OA is not high enough to ensure a sustainable flip to OA.”

·         Table 1: What is your source for the allocation of disciplines? (see also comment to Line 399)

·         Table 5: Add to the table description that it relates to Springer Compact only

·         Figure 13: I have to admit to difficulties in making sense of what you want to present. The criteria “before and after the flip” are not clearly represented. You might consider using boxplots instead, showing the distribution of APC prices in two datasets, one before the flip, and the other after the flip.

Comments for author File: Comments.docx

Author Response

Suggestion: make a reference to your dataset on github at a more prominent place in the article or even publish as supplement – it would be much easier to connect the underlying data to the figures provided in the article. It took me some time to find the relevant link in the references.

We thank the reviewer for their time and thoughtful comments to our manuscript. We have added the following references to the text, and hope that the data and code are now easier to find!

Line 247: For the purposes of this study, we created and curated a dataset—available on Zenodo [56]—that includes the following core data (...)

Line 417-18: The formal data analysis is tracked with R and R Markdown and is openly available on GitHub including version history [73].

We have also asked the Editor to add these references under supplementary files.

This brings me to my second concern: I am missing data labels in many of the figures. For most of the figures, these should be easy to add (figures 3, 5, 6, 7, 9, 10, 15). In Figure 4, which has already some labels, data labels for the set size related to the number of journals in each database could be added. Data labels could make Figure 8 confusing, though. In this case, a data table would be helpful.

In our view, adding data labels to the plots would negatively affect their legibility. The plots visualize the different distributions and shares clearly enough to see certain patterns, which is what we were interested in. The underlying  data and code is all openly shared for readers that are interested in absolute numbers or in how the plots were created.

Section 4.2.5. Analyzing the affiliation of journals to societies or research institutes, it would be interesting to know which or how many of these journals are published in-house by the societies themselves and how many are published by professional publishers in partnerships.

While we agree that it would be interesting to investigate the nature of the relationship between publishing house and scholarly society in greater detail, we think this lies outside the scope of this current paper. It can be quite challenging to determine whether the publisher distributes or publishes the societies’ journals as often this is not clearly stated on the journal websites. Moreover, we believe that the connection we established between publisher and scholarly society—even though on a basic level—already offers valuable insight in the context of reverse-flip journals.

Section 5.3. You are reasoning that APC pricing is linked to several conditions, among them “to a large extent to the impact factor”, but you support this with only one single source, the Springer Nature IPO prospectus. I would like to suggest an additional source supporting this argument: Schönfelder, N. (2018): APCs—Mirroring the impact factor or legacy of the subscription-based model? DOI: 10.4119/unibi/2931061, as this would make a stronger case for your conclusion. But note that Schönfelder attributes the link more dominantly to pure OA than to hybrid OA, so some rephrasing of your argument could be necessary, unless you can find additional sources. Furthermore, the second part of the Springer Nature quote (lower APCs adversely affecting margins) does not point to the impact factor argument but to market mechanisms and monetizing.

We have rewritten this part of the manuscript, and want to thank the reviewer for their helpful comment.

Line 173: Add “whereas” before “it is rare for journals to recover after a period of inactivity.”

We have added “whereas” to the text.

Line 209: I suggest you begin a new section here. The following is not part of “Sustainability of OA Journals”, which you gave as background, but you present your research questions here and those should be given a new heading.

We fully agree with the reviewer that it makes sense to have a separate section for the research questions, and have added the heading “2.3. Research Questions.”

Line 257: How do you define OA in this context? Is the content in the journals free to read as a minimum condition or did you check for cc-by licenses?

We agree that a definition of OA in the context of our paper would be useful, and have edited the text for clarity (see lines 265-267).

Line 399: Reference [61] points to the Internet Archive Wayback Machine. Here I am missing a source for the academic disciplines you allocated to the journals. I assume from looking at the github dataset that you used the CTWS Journal Indicators file? What about the sources for the other information you added?

We have added a wrong reference by mistake. We based the academic disciplines on the taxonomy provided by the National Academies of Sciences, Engineering, and Medicine, available here: http://sites.nationalacademies.org/pga/resdoc/pga_044522. This reference has been added to the manuscript.

Regarding the sources for “the other information” we added, as we state in lines 409-10 “the missing information was collected manually by browsing journal websites and the Internet Archive.”

Line 407 / Line 559: Can you give a reason for 1985 as a turning point?

We have added a short explanation and reference to line 415: “We chose the year 1985 as the tipping point because by the late 1980s and early 1990s the Internet had advanced to the point that scholars were able to publish free, digital journals online.”

Line 547 / Figure 10: I agree that for health science the majority of flips occurred after an age of more than 15 years. But is this true for LS journals as well? It seems to me that here about half the journals concerned are above 15 years, the other half is below. Again, data labels could be helpful in supporting your arguments.

As stated before, we prefer not to add data labels to the plots. However, we have edited the text (lines 557-63) following the figure for clarity.

We want to thank the reviewer for this comment as it made us look at the numbers again, and we found that we had made a mistake. We had put five journals that were 10 years old into the “11-15 year” period. We have corrected Figure 9 and Figure 10 accordingly. This correction is also tracked through GitHub, available here: https://github.com/njahn82/rev_flip_results/commit/5e95cb38ac42a165e035cba8a16b2dabf69e7584#diff-666c3911fa4bbbb416f8e46849713902

Line 781: Insert “however,” before “the uptake of hybrid OA is not high enough to ensure a sustainable flip to OA.”

We have added “however.”

Table 1: What is your source for the allocation of disciplines? (see also comment to Line 399)

We have also added a reference to the National Academies of Sciences, Engineering, and Medicine to the caption of Table 1.

Table 5: Add to the table description that it relates to Springer Compact only

We have edited the caption of Table 5, so now it reads “Number and relative share of hybrid OA articles in reverse-flip journals as part of Springer Compact, paid for by institutions reporting expense data to the Open APC initiative.”

Figure 13: I have to admit to difficulties in making sense of what you want to present. The criteria “before and after the flip” are not clearly represented. You might consider using boxplots instead, showing the distribution of APC prices in two datasets, one before the flip, and the other after the flip.

We have modified Figure 13. The axis labels now correspond with the phrases we used in the paragraph preceding the figure. We have also edited this paragraph and the figure caption for clarity.

Additionally, we have added a reference to GitHub, where the underlying data and code for specifically this figure is available: https://github.com/njahn82/rev_flip_results/blob/master/data/fig13_data.csv  

However, we do not agree with the suggested alternative of boxplots as we think there is much more value in seeing the individual journals as their own data points—especially when keeping in mind the highly varying circumstances of individual journals. In an earlier draft of the manuscript, we had actually used a boxplot instead (see https://github.com/njahn82/rev_flip_results/commit/9baf653c4810dd952e27a4e69f7fe0e83286ddc0?short_path=764fa7b#diff-764fa7b18e749836ef4e97f92541732b), but since we had to exclude APC-free journals (0 USD), we could only display very few journals, and decided in favor of the current version.

Reviewer 2 Report

This article offers a novel take on a subject of broad academic interest and offers conclusions that will contribute to the ongoing controversies that surround OA and commercial publishing. Only two suggestions for minor additions to the present discussion:

The constraints typically connected with 'external funding' of OA journals are briefly mentioned in ll. 192ff, but this should be expanded to include at least some mention of the policy/political/commercial constraints associated with governmental, non-governmental and commercial 'funders', e.g. which bureaucracies and agencies with which kinds of agendas are cor could be involved and for what kind of reasons? Otherwise the category 'external funding' begins to sound 'strings-free', which it isn't. 'Defense' funding might ring alarm bells, but 'medical' or 'charity' funding shouldn't go unexamined.

Rather similarly, and to be noted at the beginning and discussed at the end, academic publishing with commercial publishers is certainly profit-oriented (or non-profit oriented, in the case of Sage, which operates commercially) this 'industry' generally excludes the kinds of hugely lucrative commercial publishing 'opportunities' offered by internet services generally, such that Facebook etc. are really advertising companies, offering themselves as 'billboards' and charging what that market will bear. Academic journals from commercial publishers advertise only very limited services and products and don't - I imagine - generate much revenue. This needs noting, and could possibly change, of course. Any information on this important connection - remembering, as the article shows, that OA operates in a commercial world, however it works - would be welcome and is essential to the discussion.

Author Response

The constraints typically connected with 'external funding' of OA journals are briefly mentioned in ll. 192ff, but this should be expanded to include at least some mention of the policy/political/commercial constraints associated with governmental, non-governmental and commercial 'funders', e.g. which bureaucracies and agencies with which kinds of agendas are cor could be involved and for what kind of reasons? Otherwise the category 'external funding' begins to sound 'strings-free', which it isn't. 'Defense' funding might ring alarm bells, but 'medical' or 'charity' funding shouldn't go unexamined.

We thank the reviewer for taking the time to review our manuscript and provide thoughtful comments. We agree that external funding streams can come with certain “strings attached,” and that these are critical aspects for scholarly publishing. However, due to the complexity the issue, and with regards to what we cover in our paper, we think an in-depth examination would be difficult to do justice within the scope of our paper and would distract readers from what we are really concerned with here.

We have added the following sentence to briefly mention possible complications and refer interested readers to further readings: “Moreover, under certain circumstances, the dependence on external funding can make journals vulnerable to outside influences, such as commercial interests-through advertising (Tsai 2003)-or changes in the political climate in the case of public funding sources (Cattaneo & Corbellini 2010).”

Rather similarly, and to be noted at the beginning and discussed at the end, academic publishing with commercial publishers is certainly profit-oriented (or non-profit oriented, in the case of Sage, which operates commercially) this 'industry' generally excludes the kinds of hugely lucrative commercial publishing 'opportunities' offered by internet services generally, such that Facebook etc. are really advertising companies, offering themselves as 'billboards' and charging what that market will bear. Academic journals from commercial publishers advertise only very limited services and products and don't - I imagine - generate much revenue. This needs noting, and could possibly change, of course. Any information on this important connection - remembering, as the article shows, that OA operates in a commercial world, however it works - would be welcome and is essential to the discussion.