Home » Blog

What’s the return? Or…how is it possible that Open Library of Humanities works?

7 July 2017 3 Comments

Following on from my post yesterday a couple of questions popped up about collective and collectivist models in scholarly communications. Richard Poynder is skeptical, which left me nonplussed because from where I sit what I described is happening all over the place. Funders are looking at investment strategies, collectives are forming and some of them are growing very rapidly. Which brings us to the second and more concrete question. How is it that things like Open Library of Humanities, a collective funding model for journal publishing, and similar models like Knowledge Unlatched* for books, work?

The answer to this was, yes I do, but it keeps changing. But perhaps the time has come to lay that out and give a really concrete example. As I noted this is a weakness of our A Journal is a Club working paper, which has led some to think it only applies to a very narrow kind of journal.

The classical subscription journal club

In our “traditional” system we usually think of a journal as a club of subscribers. A set of actors band together to create a “club good”, one which is exclusive but broadly non-rivalrous. In the case of a subscription journal, this is access to content. “Access” here is meant broadly to include all the processes that lead to the creation of a journal, not just merely providing a key to existing content. However linking the funding of that content creation to access means that a certain class of free rider is avoided. The existence of that content is a collective (and in some ways at least partially public) good. The conventional analysis follows that it is only by making access to the collective benefits of that good exclusive that the classical collective action problems are avoided. That is, if the content were freely available then rational economic actors would not contribute, and the collective good would not be provided.

Buchanan, working in the 60s on club economics would have noted another important characteristic, particularly in the print world. That is that access is not purely non-rivalrous. With a fixed number of copies on the shelf, if there are too many members with access then there is a rising probability that when I go to the library I will not have access. Buchanan showed that for club goods, where access is usually non-rivalrous up to a point, but after that point there is increasing friction, which leads to a natural limit to the size of the club. Clearly as we move to the digital world, this friction drops radically (not to zero but certainly several orders of magnitude). This in principle allows that natural limit on club size to rise.

The observant reader will note that I’ve conflated two different clubs already. One is the club of subscribers to a journal. The second is the club of members of a particular library. I haven’t fully worked out the analysis for this, but it is possible to point to the way that big publishers have retained the link between pricing and historical print spend, and seeking to reduce the ability of libraries to combine holdings and expand their user base radically as an effort to prevent the natural club economics working its way through. Thus the serials crisis could be painted as a break down in the economic equilibrium of a series of close linked club-like groups that has been exacerbated by reducing market flexibility through contracts, pricing and big deals.

However, the main point I want to make here is that the link between the club that funds a journal (i.e. its subscribers, generally academic libraries) and those that actually use the content (and benefit from access as a club good) is not as obvious and direct as we often think.

Collective funding models through the lens of club economics

If we turn our attention to collective funding models that have recently developed, primarily Open Library of Humanities in the journal world and Knowledge Unlatched with books we can draw broadly two forms of analysis. The first is to reject the classical economic analysis that says these cannot work due to free-riders. In this form we note that libraries and universities are value-led organisations that are free to act in an economically irrational fashion in order to serve their mission. There’s an argument to be made for this analysis, and I think it might make a lot of sense in some cases at very large scale. For instance, it can make sense to see the community funding of Wikipedia as a case where at sufficiently large scale you can reach enough economic actors prepared to act in an irrational fashion to create a public good.

However, when it comes to budget negotiations and resource allocation in a cash strapped library or university setting an argument built around “we should be irrational” will only get so far. We can apply a club economics based analysis to projects like OLH and KU. That analysis simply leads to an obvious conclusion, there must be some club-like, and therefore exclusive, good benefit to the contributors. This benefit must have a value greater than the contribution. It is also likely that there will be friction in accessing that value that increases as the number of members increases. However, with more members the cost of contributions can go down, or alternatively more of the good can be provided. How friction, pricing of contributions, and goods provision scales with the number of contributors is crucial. So figuring out exactly what contributors are getting really matters.

Obviously exclusive access to content is not the club good. But neither is non-exclusive access. This is a collective, public-like good, not reserved to contributors. But what contributors get is a good feeling about making that public-like good available. This is a real benefit, particularly for brand driven organisations like universities, and for libraries within them that want to show internally that they support innovation in scholarly communications. This is an exclusive benefit, only available to those who contribute and have that contribution recognised. I doubt, for instance, that any universities are contributing anonymously to OLH. OLH assiduously recognizes each new member with dedicated postings and acknowledgement.

There is friction in access to this good. It’s great to be one of the first members of the cool kids club, it creates buzz, raises your profile, demonstrates your commitment to progressive change in scholarly communications. But once everyone is a member, the gloss wears off. When I first looked at the OLH model I thought this was what would cap growth. However it looks like I was wrong about that. Perhaps the momentum that has built up is its own driver, with institutions concerned about being left out? Classical economics is equilibrium economics and is generally poor at predicting the path towards equilibrium (such model systems are usually chaotic even in their approach to equilibrium). Overshoot is certainly possible, but equally possible is a shift in the perceived benefit.

A particular membership benefit for OLH is influence over its decision making. While an economic model would say that an investment in making future savings through subscription cuts is irrational (why not let others pay and reap the benefits anyway) investing in a level of control over the process by which that future saving is reached is certainly not. You might not be a first mover, but it is still worth engaging in having a say about the pace at which OLH grows, what gets included in the package, and why. Again, friction comes into play. The more members the less say any individual members have. In some cases membership differentiation can play a role here but having a system where certain libraries pay more to gain greater control seems antithetical to the values of the project and quite possibly detrimental to the good faith built up by contributors.

It would also be possible to explore other member benefits that relate to this control aspect. The ability to propose University Press journals for inclusion, first right of refusal to invest in particular projects that are of interest to specific libraries might be of interest. Other benefits might include access to expertise and support on publishing technology or financing. Access to these kinds of specialist resources was a big part of the success of the cases Mancur Olson describes in his book The Logic of Collective Action. In any case the major scaling challenge for OLH and similar efforts is exactly how to manage governance and strategy in ways that mean contributors have a stake that is meaningful to them and of value.

The contributions are small

One of the reasons this is working, and that the specific benefits are a little vague, is that the actual amounts contributed are relatively small. This is crucial, when the price of membership is low, the benefits can be a little diffuse. Feeling good about the contribution, a sense of being part of a movement for change, may well be enough. This works well because OLH has an efficient model for distributing costs, is growing membership, and thus keeping contribution pricing down even while it grows its journal portfolio. What happens though if OLH, or a similar system were to capture a larger portion of the market? Then contributions would rise as a proportion of library budgets. As the values rise the level of scrutiny rises and there will be a need for clear benefits.

It is the combination of the governance challenges with the rising proportional contributions that are likely to place upper limits on the scale of OLH and similar efforts. To be clear, flipping HSS journal publishing to a collective OA model does not need to be the end goal for OLH. It can declare victory as is in many ways, having demonstrated what is possible. But those of us who see such a system as a desirable model for funding the majority of scholarly communications will need to identify a route that navigates these twinned challenges of scale.

There are two questions I don’t have any answer for at all as yet. The first is what is the optimal ecosystem of such projects, one single system to rule them all with internal competition to constrain pricing? A small number of competing players or many smaller projects each building its own communities of interest? The second question, which is just as important, is what routes are feasible, and to what ecosystems? Just because a system is optimal, doesn’t mean we can get there. This in the end is to my mind the most important question. OLH and KU and similar efforts have already shown that non-access based subscription membership models can work. That means that membership benefits beyond access to content are feasible. But how we can institutionalise and grow them to the right scale to build a sustainable system of such efforts remains the big questions.

* Conflict of interest statement. I work on projects with Knowledge Unlatched Research as part of the work I do at Curtin University.


  • Brucecaron said:

    Certainly the OLH can point to both an optimal alignment between its goals and those of its members/supporters as well as an economy and efficiency in its practice. The question then becomes: how do for-profit publishers compete with the OLH? The key is to build a package of member benefits that fronts the goal alignment and provides no practical disadvantage (except perhaps impact factor). ESIP does this by creating alternatives for its members to propose online “workshops” that attract a broad variety of expertise, and that are available immediately and at no cost to the member. Instead of needing to propose to the NSF for a workshop in 9 months, ESIP members “workshop” online today and finish their work without any travel, and potentially within weeks. NOTE: the funders love this too!

  • crusbrid said:

    OK, I know nothing about this really, so this is just trying to understand a bit. Can we think of a library as a club for its readers, and a journal as the union of two clubs: one for its readers (the majority) and one for its writers. But what if you built the economics of the journal the other way round: the club is for researchers who want their work to be read and judged? Or, possibly, for research institutions (perhaps via their libraries) who want their researchers’ work to be read and judged?

    Those certainly seem like the criteria for OA journals, whether gold or green. Both seem to leave out some key potential club members: non-affiliated researchers (who are also left out of traditional journal clubs).

    I have a young relative, recently completed a clinical doctorate in psychology. A year or so after leaving the university where she completed her training, one of her pieces of work finally gets accepted for publication. Prompted by me, she asks for the Open Access option… but it’s £1,800! Too much, and no sponsor, so she has to opt for the traditional TA option… which I suppose makes her a free-rider on the traditional model, after being excluded from the new model!

  • Cameron Neylon said:

    So yes, there are lots of clubs interacting. One challenge is to work out when you have a single club with different kinds of members and when you have different clubs interacting in exchange. TBD on how one does that, tho at some level its a question of how you build your models.

    Currently we would tend to see a journal as a single club with different kinds of members. Buchannan’s work allows for this (and predicts it). The simplest case is when authors are readers (think traditional society journal) and we need to work out what happens as those two groups drift apart. My (not fully worked out) thinking at the moment is that what tends to happens is that coordination costs move from being off-balance sheet (community managed) to on-balance sheet.

    In this sense PLOS ONE and Nature both face the same problem. There is a large divergence between author and reader communities so internal (community based, tacit) knowledge that can efficiently do QA has to be professionalised, which happens in both cases.

    In my analysis an OA and a subscription journal don’t end up so different, but the models for payment mean that this scaling problem plays out differently for each.

    Now, all clubs (by definition) leave some groups out. I think the key club for your relative is not the journal but the institution/community one (which takes you back to the library). There are OA journals that function as a community club where your relative would have OA publishing privileges as a member, but in the subscription case, yes the external author is a free rider. Which is probably why we so fiercely police things around institutional affiliations I guess…

    Clear as mud?