Monday, 26 November 2018

Peer Reviewers Should Be Paid

Academia is an extended set of conversations all going on at once. We academics score status points for making a contribution that other people find interesting because it helps them with their often rather specific problems (say, about a new interpretation of Galileo's conception of physical laws or a new method for identifying pancreatic cancer cells). The more that other academics value your contribution (by citing it in their own contributions), the more status points you get for it. (Google's PageRank is based on the same system.)

I said academia was conversational, but these are not normal conversations. First because the standard intervention is a 10,000 word long monologue. Second because journals curate what is good enough to be allowed into the conversation using peer-review. These journals are another layer in the academic status economy. They try to publish those monologues of most interest to most people and thus most likely to increase the status of their journal within those academic conversations. The higher the status of a journal the more able it is to bestow prestige upon those who publish in it, and so the more academics will send it their most exciting ideas. A virtuous circle of prestige appears around the journal, which translates into outsized profits for its multinational corporate owner  - since every university needs to subscribe to it or risk missing out on the most exciting part of the conversation.

If we were building it from scratch, I don't think we would set up a system like this in which publicly funded research is privatised and commoditised, resulting in enormous profits for a handful of companies while systematically excluding ordinary citizens, journalists, and poorer country universities from access. There are many attempts to change things, from protests against price-gouging behaviours, to boycotts of the most loathed company, Elsevier, to outright defiance (pirate sites like scihub and libgen), to the creation of free Open Access alternatives. I wish them well.* Here I want to open up a new front that is not about access but how the product gets made: the treatment of peer reviewers


Peer review means asking 2-3 experts from the conversation a paper claims to contribute to whether it is any good. Peer review is crucial for the academic journal status system for two reasons. 

First, like a focus group, it provides an advance indication of how it will fair in the market for ideas and thus how profitable it will be for the journal's prestige. 

Second, quality control. This is crucial to academic publishers' justification of their enormous profits (35% of revenue is common). The market for blood products benefits many millions of people, but for it to work recipients need to trust that the blood they are receiving is the right kind and disease free. Likewise, the academic market for ideas we have developed requires that someone take on the tedious work of guaranteeing the basic safety of the products on offer. Even if you don't find this article especially helpful (and the overwhelming majority of articles never get cited), at least it won't make you ill (by corrupting your reasoning with false premises). Peer reviewers protect the reputation of the journal by carefully scrutinising articles for errors of methodology, reasoning, and coherence with better established facts.

Despite its importance, peer reviewing is nearly always unpaid. 

The brilliance of the academic publishing model is that private companies extract free labour from academics and translate that into profits by charging academics to access their own work. Yet the case of peer reviewers stands out as particularly unjust. 

The authors of articles are trying to get prestige points to increase their standing among other academics and advance their career (by swapping up to a more prestigious university or at least getting out of having to teach undergraduates). Moreover, this is exactly what their universities pay them to do, since the university shares in any prestige they gain. 

Journal editors sometimes receive a small stipend but most are just volunteers. Yet here too there is a payment in the currency of prestige. Being asked to edit a journal is a recognition of your importance that you can display like a badge, and it also grants you the power to confer prestige on others by choosing whether their work makes it into the oddly limited number of pages in your elite journal. 

Peer reviewers can't be paid in prestige points because the integrity of the publishing system requires them to remain anonymous. They are supposed to be motivated by an amorphous sense of 'collegiality' or 'service' to take on responsibility for keeping the whole academic project going. Yet given the contribution their work makes to the profits and careers of others, this lack of recognition amounts to a systematic exploitation of these honorable motivations. According to one report, journals currently extract £1.9bn in free work from reviewers every year. Peer reviewers are the student athletes of academic publishing.

Moreover, it seems that these honorable motivations are not flourishing. Journal editors are becoming dissatisfied with their volunteer workforce.
"Common bad behaviour from referees includes (but is not limited to!):
1) Failing to respond to invites in a timely fashion(where timeliness is calculated in days not weeks), even if it’s only to decline the invitation;
2) Agreeing to act as referee and to return the report within an agreed timeframe (in the BJPS’s case, four weeks), only to substantially exceed this timeframe (by weeks, sometimes months) and
i) asking for this substantial extra time for the weakest of reasons*;
ii) not communicating with the relevant editors whatsoever;
3) Returning a report long past the agreed timeframe, and that report being almost useless;
4) Not returning the report and not responding to emails enquiring about the report."
Those complaints come from an editor at the British Journal for the Philosophy of Science, who is particularly concerned that undermotivated reviewers are delaying publication decisions and hence career paths for junior academics. [Once again speaking to the importance of prestige in academia.] She suggests punishing bad reviewers by refusing to accept their submissions. I have a more practical suggestion with a better proven history of motivating people: Peer reviewers should be paid.

I envisage a system where journals pay a set rate per review that comes in on time and meets its requirements. $150 seems reasonable for a job that when well done can easily take most of a working day. (Perhaps with an option for a $50 bonus for a very fast or impressive review.) A carefully devised Uber style rating system could create a properly functioning piecework market.

Paying reviewers would be effective, affordable, and just.


Good quality work takes time, and time has to come from somewhere. It can easily take 6 hours to read a 10,000 word manuscript through carefully twice while making notes and then organising those notes into a forensically reasoned report submitted according to that journal's peculiar style. If the journal isn't paying for that time then it must come from somewhere else. From my working hours, when my university expects me to be teaching or writing new articles of my own. Or from my evenings and weekends, instead of spending time with my partner, friends, or hobbies. It is no surprise that people working for free don't prioritise the timely completion of review reports, or don't do a thorough job when they have to steal the time in fragments from their other obligations. If you want professionalism, you have to pay for it. 

Given the extraordinary profitability of academic publishers, there seems to be plenty of money available for paying reviewers for the work they do to make those profits possible. Moreover, this shouldn't affect the pricing of journals. These ruthless profit-maximising companies already pitch their prices at the highest amount they can get away with, which has the happy side effect that there is no room for them to pass on an increase in their costs to their customers. So the main effect of this would be redistributive: money would be transferred from profit-making publishers back to the academics doing the real work of maintaining the quality of the publishing system.

A formalised piecework system would certainly be quite disruptive of the status quo, and some valuable things might be lost. (Just as happened with the professionalisation of sports.) But not all of those disruptions would be negative. For example, it might provide an economic lifeline to the vast underclass of adjuncts currently working in precarious, under-paid, and often undignified circumstances to teach undergraduate classes so that star researchers don't have to. For another example, smaller journals run by volunteers (such as the journal I helped found*) would not be able to pay. But those journals are in the best position to make an honest appeal for unpaid volunteers since they are the ones genuinely organised around the pursuit of academic values rather than as multinational corporations exploiting academics' sense of moral obligation for profit. 


***

There is much to regret about academia's dependence on an economy of prestige whose workings are outsourced to profit-seeking companies. Even if such an economy of prestige were desirable, there are many better ways to organise one. For example, a Reddit style system where anyone can publish to a particular community and community members can vote submissions up or down and add their critical comments for the author to respond to. That wouldn't be perfect either, but it would be much faster, more egalitarian, extremely cheap, and much more in the spirit of academia as a conversation than what we have now. Paying peer reviewers is not such a radical proposal. It only aims to make a bad system somewhat better.


Notes
*I was a founding editor of an Open Access journal, the Erasmus Journal for Philosophy and Economics, proudly free to all readers and authors as ideas should be.