Thursday, April 23, 2020

Funder Publishing Platforms

After posting Never Let A Crisis Go To Waste earlier this month, I can't resist a shout-out to Elizabeth Gadd's The purpose of publications in a pandemic and beyond:
The virus is reminding us that the purpose of scholarly communication is not to allocate credit for career advancement, and neither is it to keep publishers afloat. Scholarly communication is about, well, scholars communicating with each other, to share insights for the benefit of humanity. And whilst we’ve heard all this before, in a time of crisis we realise afresh that this isn’t just rhetoric, this is reality.
Below the fold, a few comments.

Gadd is right both when she describes the "two buts" that have stymied progress to open access ever since the Budapest declaration:
  1. But what about publishers and scholarly societies? How do we ensure they survive and that the economy isn’t damaged? (Subtext: publications are for profit)
  2. But what about academic careers? A good publication list is critical for promotion and tenure. (Subtext: publications are for credit).
And when she advocates for funder-based publishing platforms:
UKRI have been throwing money (about £150 million by my estimate) at funding eye-watering Article Processing Charges with haute couture journals for seven years. This is money that could have been used for actual life-saving research.
...
UKRI need to set up a funder-based publishing platform and say to recipients, if you want our money, publish your findings here. End of.

This is not a new idea, but a proven technology. Gates Open Research is a great example. Despite not being the only publication option for Gates Foundation recipients, the papers published in Gates Open Research are achieving a cites-per-paper rate on a par with the world average for medicine. Similarly, Wellcome Open Research is on a par with top quartile journals in both biochemistry and medicine in terms of it’s Scimago Journal Rank.
It has been obvious for many years that the funders are the only ones who can make open access a reality, by removing researchers' choice about how to publish their research. Preserving UK researchers' ability to spend other people's money on polishing their resume by publishing in "haute couture" journals isn't worth £21.5M/yr.

3 comments:

Michael Nielsen said...

I'm not sure either of these options is all that great. Funder-led publishing platforms are a good stopgap, but they become centralized bottlenecks for innovation. I elaborate on this point here: https://twitter.com/michael_nielsen/status/1101997243786117120

David. said...

Michael, your thread makes many good points, many of which I agree with. In an ideal world:

"we want a scientific publishing system where someone can start a new journal / repository / database (etc), & if it serves society as a whole better than existing solutions, it will rapidly grow to replace or augment them."

But, as you point out, we don't live in an ideal world. So funder publishing platforms may not be just a "good stopgap" but may well be the least worst alternative. Given my experience, I'm predisposed to favor innovation, but as Silicon Valley shows, "favoring innovation" is often a rhetorical shield for extractive behavior. So I'm skeptical when you write:

"what's the growth model for innovation? Is the market set up to enable the flourishing of many good new ideas that will benefit humanity? At the moment, it's not doing a great job, in my opinion."

I think Elizabeth Gadd was right to point out that the really important point is to enable scholars to communicate, and if they have to put up with slightly less functional tools to do it that's a price worth paying.

In my view the fundamental problem with your thread arises when you write:

"The underlying trouble is that researchers have a lot of power over buying decisions by university libraries, but do not bear the cost of those decisions (the libraries do). So they're price-insensitive. Such a separation of decision-making power from costs leads to bad outcomes."

and:

"payment for services needs to align incentives: the people benefiting from the services should be paying for them, to set up the virtuous feedback loop: genuinely better service => more revenue."

The fact is that the cost of an individual act of publishing and other at-scale digital services is so small that it has no effect on behavior. That is why at-scale services don't charge - they'd need micropayments (debunked by Clay Shirky in 2000 and Andrew Odlyzko in 2003). Instead they show ads (not a viable economic model for scholarly communication), aggregating the users to a scale worth paying for.

The costs of scholarly communication are in practice aggregated and imposed on the funders through grant overheads. Their goal is to generate research, so they see the benefits of that communication. So, by your argument, having the funders remove the author's choice so they pay the aggregated costs of communication rather than having the funds siphoned off to publishers' executives and stockholders is the right answer, or at least better than the alternatives.

Michael Nielsen said...

Thanks for the thoughtful comment, David.

Much of it I agree with. A couple of substantive points at which we differ:

(1) I disagree the price is "slightly less functional tools". In the short term that's true; over the long term it's vastly less functional tools.

(2) In some sense I agree with (and have made the same argument as) you before regarding the alignment of incentives in regard to the funders. Eg, the role of people like former head of the NIH Harold Varmus in pubmed, PLoS, etc. Those have been mostly positive developments, in my opinion, and certainly line up with the argument you make in the last paragraph of your comment. In practice, though, I'd prefer the alignment to be (a) much more direct, with decisions being made by individual scientists; and (b) as a result, allow much more decentralized innovation.

As a small example of the last point, physics has mostly moved to the arXiv as a de facto standard means of communication. I love and admire the arXiv. However, it remains very focused on pdf and similar formats. As far as I'm aware it doesn't allow the uploading of things like Jupyter notebooks, nor the kind of interactive visualizations published by distll.pub. In that sense it's something of a choke point for getting to better tools.

I'm reluctant to criticise them like this: they do an incredible amount with a tiny budget, and I know people there would like to do more, if only they had more money. But the structural point remains: as soon as you have a standard locked in, it can become a chokepoint for improvements in the tools.