The marketisation of academic citations

I. The shared diagnosis

There is a cluster of projects I want to write about here—Bucket, PubChain, CiBit, CiteChain, the Impact Market, and others—and they all start from a diagnosis I share. A paper cited ten thousand times generates no revenue for its authors. The platforms that index and rank citations extract rent from a system that scholars produce for free. I do not think anyone who has looked closely would dispute the description. It is not a controversial claim, and it is not a new one.

What is new is the proposed remedy. They have converged, more or less independently, on the same answer, which is add a settlement layer. Make citation itself a thing that can be paid for, staked, or traded. Whilst I share the diagnosis, I do not share the remedy. And the more I think about why, the more I think the disagreement is not about tactics, or about implementation, but about what citation is.

II. Why citation is not a market failure

The current arrangement is often described as a market failure. It is certainly true that citation generates value: it signals quality, it distributes attention, and it builds reputation. It is also true that the value accrues to platforms, not to authors. But to call this a market failure is already to have chosen a remedy. It assumes the problem is the absence of a market, and that the solution is to build one.

I think that the economics say something different. Citation is a public good. It is non-rivalrous: my citing a work does not prevent you from citing it. It is non-excludable: once a work is cited, anyone can see the citation. More fundamentally, public goods are recognisably undersupplied by markets. However, this is not because markets are broken, but because markets are not the right mechanism for producing them. This is a standard observation and from it flows the principle that the remedy for a public good is not to make it rivalrous but to find a different form of provision.

So the proposed market in citations does, as far as I can see, exactly the opposite. It introduces rivalry where before none existed. A citation token is a scarce good and there are a finite number, and owning one excludes others. The market creates the scarcity it then claims to price. This is not a solution to the public good problem. It is a transformation of a public good into a private one.

Of course, one might object that the citation itself remains non-rivalrous, and it is only the token that is scarce. That is true as far as it goes. But the token is not a separate thing from the citation. It is the citation, re-described. To make the citation tradeable is to make the act of citing a transaction, and a transaction requires two parties with something at stake. This means that the relation is no longer simply recorded; it is priced. The distinction between the citation and the token turns out to be the distinction between a relation and a contract. I do not think you can have both at once, and I do not think the projects have quite reckoned with that.

III. What a market in citations would produce

Suppose citation becomes a tradeable asset. What follows?

First, the incentive to cite changes. A rational actor now cites not only because the work is relevant, but because the citation token has upside. The price signal enters the decision. This is not a hypothetical: it is the predictable outcome of introducing a financial incentive into a signalling system. The question is not whether scholars would behave this way; the question is what the system rewards.

Second, the signal value of citation collapses. If citation can be bought, then a citation no longer signals quality. It signals wealth, or strategy, or timing. Michael Spence showed that a signal is credible only if it is costly and the cost is correlated with the underlying quality. If citation can be bought, the cost is correlated with wealth, not quality. The signal ceases to signal.

Third, adverse selection follows. Akerlof’s “market for lemons” describes what happens when buyers cannot observe quality: bad goods drive out good. In citation markets, the same dynamic will apply. If citations can be bought, then low-quality work can purchase the same signal as high-quality work. The signal becomes noise. Ultimately, the result will be lower average quality than the non-market arrangement would produce, and not because scholars are more venal than anyone else, but because the structure makes the dishonest strategy the rational one.

Fourth, citation becomes a positional good. Fred Hirsch's analysis of positional goods shows that some goods are valuable only because others do not have them. Positional goods are zero-sum: one person's gain is another's loss. If citation becomes positional, the total value is fixed. The market does not create value; it redistributes it. And it redistributes it to those who can afford to buy in. Which is a rather familiar pattern.

IV. The performative problem

There is a further problem, which is not strictly economic. Goodhart's Law states that when a measure becomes a target, it ceases to be a good measure. Citation metrics were already imperfect. Making them financial targets accelerates their corruption. The measure does not just become inaccurate; it becomes actively misleading, because actors now have an incentive to game it. The projects acknowledge this. The Impact Market paper discusses "citation cartels" and treats them as a detection problem. But cartels are not a bug in the market; they are a feature of it. When citation is priced, cartel formation is rational.

V. What a citation is

There is a deeper question, which is not economic at all. What is a citation for?

A citation is a relation. It says: this work depends on that work. It is a trace of intellectual debt. It does not say what the work is worth; it says what the work owes. The distinction between value and price is not academic...and I mean that literally, it is not simply an academic distinction, though it is also that. Value can exist without a price. A relation can be real without being monetised. The market solution assumes that value must be priced to be realised. That is a theory of value, and it is the theory I am against.

Mauss, in The Gift, distinguished between the gift and the commodity. The gift carries the identity of the giver; the commodity is anonymous. Citation carries the identity of the cited work. To price it is to strip that identity. The citation becomes a commodity (a token with a price) and the relation it recorded becomes a transaction. I do not think this is an exaggeration. I think it is what the projects are proposing, and I think they would agree with the description. They would just disagree that it is a loss.

This is close to what Polanyi called a “fictitious commodity.” Land, labour, and money are not produced for sale, but are treated as if they were. Citation is like this: it is not produced for sale. It is produced as a byproduct of intellectual work. Treating it as a commodity does not simply price it; it changes what it is. I keep wanting to say simply and then catching myself, because there is nothing simple about it.

VI. The alternative

The alternative, if I am right about this, is not to build a better market. It is to refuse the market altogether.

Elinor Ostrom spent her career demonstrating that commons are not tragic. They are not failures of market provision; they are different forms of provision, governed by rules that communities develop. The register is a commons in this sense (or at least it tries to be). It records relations without pricing them. It treats citation as a relation to be held in common, not a good to be traded. It refuses the scarce good that the market would create. A work is in the register or it is not. There is nothing to win or trade, and nothing to lose.

This is a rejection of a particular theory of value: the theory that value must be priced to be real. A record can hold value without pricing it. It can show the relations through which works live without converting those relations into assets. Whether this is sustainable at scale is an open question. I do not know the answer. But it is a coherent alternative, and it rests on a different premise than the market projects: that the problem is not unpaid citation, but commodified citation.

VII. The difference

The market projects and the commons approach share a diagnosis. Both see that the current system extracts value from scholars and returns nothing. Both see that citation is a value transfer without a settlement layer. I think we agree about the description. The difference is in what we think the problemis.

If the problem is that citation is unpaid, the market is a solution.

If the problem is that citation has been made into a commodity, the market is the disease.

The market projects choose the first diagnosis. They build a settlement layer, create a scarce good and price the relation. They are, I think, sincere about it, and I do not want to write as if they were not. They are responding to an inequity within our current system.

The commons approach chooses the second. It records the relation without pricing it. It refuses the scarce good. It treats citation as a commons, not a market

The difference is not between people who care and people who don’t. It is between two theories of what has gone wrong. And it is a difference that matters, because the two remedies produce opposite outcomes. One makes citation legible as a price. The other makes it legible as a relation. You cannot do both. I do not think there is a viable synthesis here (and I do not think there should be). The question is which one you want.

I write this as someone whose own intellectual lineage is at stake in the argument. My citations are not an abstraction to me. They are the relations through which my work lives. To price them would be to change what they are: to convert the record of a debt into a claim. That is not a transaction I want to make, and it is not one I think scholarship can afford. Perhaps I am wrong about that. But if I am, I would want to know why..