Publishers provide certain services (peer-review management, typesetting, brand badges, sometimes proof-reading or copy-editing, archiving, indexing) to the scholarly community.

Those services are of greater and lesser value, provided at higher and lower levels of quality, and cost greater and lesser amounts. Of course, we in the scholarly community want high-value, high-quality low-cost services. This is true whether the publisher in question is a multinational corporation with a multi-billion-dollar turnover, or a tiny boutique press run on a non-profit basis for the sheer love of the process.

Since the scholarly community (researchers, authors, peer-reviewers, academic editors, etc.) is spending money in exchange for publication services, and since publishers are providing publication services in exchange for money, it is clear that the goals of these two groups cannot be aligned. Any money that the scholarly community can save on publication costs is income lost to publishers; and any additional money that publishers can charge for their services is money lost to the scholarly community. I hope that so far, this is uncontroversial.

In the same way, if you sell me a second-hand car, then however well you and I might get on in civilian life — we might support the same football team, drink the same beer, discuss the same novelists, watch the same films — then for the purposes of that transaction, what is good for you (a high price) is bad for me; and vice versa. Note that in saying this I am not condemning or even criticising you. I am just stating a fact about transactions.

Now, suppose my wife and I sit down and decide that we need to buy a new car. We consider Hondas, Fords and Fiats. We weigh up various models on their merits, compare their prices with their features, and reach a decision on what we want to buy and how much we’re prepared to spend. We then approach the various Honda dealers (or, as we may have decided, Ford dealers or Fiat dealers). We negotiate with them to agree a price that we are happy with for a model that is in good enough condition. Different dealers compete with each other to win our custom by offering good cars at a low price. This is a functioning market.

What we don’t do is invite all the dealers to come and join us in our initial conversation. When my wife and I are discussing how important it is to us that our new car has variable-speed intermittent windscreen-wipers, we have that discussion in an environment quite free of car dealers telling us how great Fiat’s intermittent-wipe feature is. How could we possibly reach a coherent decision on what our own requirements are if we’re bombarded by the claims — some competing, some in collusion — of all the car dealers? And how can we think sensibly about what we’re prepared to spend if we’re surrounded by the dealers’ defences of the various financing arrangements they offer?

So in the same way, I feel that the scholarly community needs to figure out what publication services it needs, free of the influence of publishers who (and again this is not a criticism) have their own agenda. Then, when we know what we want, we can go to the publishers who offer the kinds of services we’re interested in, and invite them compete for our business on the basis of features and price.

But involving them in the initial what-we-want discussion can only lead to confusion, and a compromised outcome. Which is what we’ve seen for the last 50 years. This was the fatal flaw that led to the deeply flawed Finch Report and to the erosion of the RCUK’s initially very progressive OA policy.

As a side-note: my wife and I may end up deciding we don’t need a car at all: we might decide we can walk, or cycle, or take public transport. Car dealers would hate that: they would advocate against such an outcome with all their might if they were involved in that discussion. Which is why they can’t be.

Note. This post is adapted from a message to the Open Scholarship Initiative mailing list.