The concept of performativity in STS and the social studies of finance is a powerful one but I’ve also found it problematic, in at least one big way, which is establishing its validity, which has always seemed difficult to me. Sometimes it just seems like a new term for good, old-fashioned theories of constructivism1. At other times, itseems like a powerful idea but finding data that demonstrates that something is “performed” seems really, really difficult.
For example, Donald Mackenzie’s case for the performativity of the BSM equation [pdf] consists of showing that (a) it was not a good predictor of option prices when it was invented, (b) that from 1970 to the middle of the 1980s, the BSM predictions started to converge with market prices as it started to be used more and more by actual market actors (with the help of calculating tables created by the BSM makers), and (c) the BSM again diverged from actual prices with the stock market crash of 1987, which Mackenzie calls “counter-performativity.” This account, by itself, should suggest that Mackenzie is not making an outright case for performativity.
Sociologist Ezra Zuckerman however suggests that the story of BSM is not about performativity at all (see footnote 5 in the paper):
Perhaps the best case for performativity theory is that developed by Mackenzie and Millo (2003; Mackenzie 2006) on how the Black-Scholes (BS) theorem “performed” derivatives markets in the sense that that market participants used the theory to enact a market that satisfied its predictions. But there are two interrelated problems with this argument: (a) BS was actually not developed as a theory of how pricing worked (“a camera”) but a piece of financial engineering that specified how pricing should work (“an engine”), so any enactment by the theory is equivalent to an engineer using a blueprint enacting a bridge; and (b) they cannot rule out the alternative that BS was simply a better approach to pricing options.
I don’t really want to engage with (b) because that gets us into the whole realism vs. constructivism debate, which is never a productive one. But (a) is interesting because it seems empirically verifiable. Was BSM intended to describe how prices worked, or as a prescription of how prices should work? The fact that reasonable people seem to disagree on this seems to me highly interesting, suggesting, perhaps, that how social actors articulate the description/prescription boundary itself might be something worth investigating empirically.
All of this is just a prelude to saying that the latest issue of Social Studies of Science has a interesting paper by Daniel Breslau (STS, Virginia Tech) where he examines the making of deregulated electricity markets. The abstract (emphasis added):
Recent work on the relationship of economics to economic institutions has argued that economics is constitutive of economic institutions, and of markets in particular. In opposition to economic sociology, which has treated economics as a competing disciplinary frame or an ideology, the ‘performativity’ literature takes economics seriously as a set of market-building practices. This article demonstrates the compatibility of these perspectives by analyzing the role of economics in the politics of market formation. It presents a case study of the formation of a new institution: capacity markets connected to wholesale electricity markets in the United States. The case demonstrates how economic framing shapes the politics of markets by imposing a specific set of terms for the legitimate conduct of the struggle over market rules.
Electricity markets are not my area of expertise (Canay — we’d love to have your thoughts!) but I really enjoyed reading his description of how actors themselves argued about the deregulation process, and how the market became a framing device in this series of arguments, and how, this in turn, pretty much guarantees that certain actors will come out winners. As he puts it:
this article will take advantage of the special features of its empirical case to ask how economics transforms market politics, shaping the terms of the struggle over market rules. [italics in the original]
What happened was that:
The conflict over the design had taken the form of a framing contest, with buyers and sellers supporting opposed frameworks within which to formulate and justify measures to assure reliability. The stakes of that contest were whether ongoing reliability concerns would be treated as a problem of market design at all, and whether the demand curve approach put forth by generators and PJM would be adopted. But once sufficient forces were lined up in support of the new capacity market design, with its downward-sloping demand curve, the struggle proceeded with the victorious market frame as its ground. This point was reached on 20 April 2006, when FERC formally accepted most of the elements of the RPM design as a ‘just and reasonable’ replacement for PJM’s existing capacity market.
The original debate was over whether the problem that the capacity markets were designed to fix was really a market problem at all. A group of actors argued that it was not. But ultimately, it was the market view that won out — and from then on, the struggle became one that took market design itself for granted. The “performativity” of markets here consists of how the rules of designing “good” markets themselves become the grounds of a political struggle, with all sides taking the usefulness of markets for granted.
The most fascinating part of the paper is about the demand curve, a piece of idealized mathematical machinery regularly used by economists. The demand curve (a plot of the price of a commodity against its consumption, or the amount of the commodity available) that fit the original market looked like this; it was, as some actors argued, “pathological.” (See the figure on the left-hand side.) In this demand curve, the price is either at a certain value when capacity falls below a certain level, or it is zero (when there is excess capacity). Certain actors (in favor of the deregulated markets) argued that this was clearly inefficient. As Breslau points out:
a vertical demand curve yields prices that swing wildly with very small departures from equilibrium. Indeed, this provides an explanation for the historic pattern of prices in PJM’s capacity market. Prices did seem to alternate from long stretches at very low levels to periods at which they hovered around the deficiency charge. During periods where prices seemed to be mired at a very low level, sellers of capacity complained that their revenues were inadequate. During the high-priced periods, LSEs took action, charging that owners of generation were wielding market power, withholding capacity to keep the prices high. The economists who described the shape of the demand curve pointed out that the shape of the curve provided a strong incentive for owners of capacity to wield market power, because a small decrease in supply of capacity could cause the price to leap from near zero to its maximum. (p838).
Their favored solution was the curve that looked like the figure on the right-hand side, something that came closer to the ideal market. Here the price would decrease in increments as the capacity/supply increased. The market would need to be redesigned in order for the demand curve to look more like this idealized version. But how to do this redesign?
Once the existing demand curve was described and related to the theoretical curve of a competitive market as a pathological case, it could become the object of a design intention. From deriving a demand curve from an analysis of the market rules, it was a small step to reverse the logic, working from a desired demand curve to reform market rules. [...] It was a small logical step to replace the implicit and pathological demand curve with an explicit one that mimicked the gradual downward slope found in a competitive market. A simpler version of such a curve had been developed and employed in the New York Independent System Operator since 2003, where it had been proposed by economists at the New York Public Service Commission (Paynter, 2004). And if there was to be a designed demand curve, there was no reason any longer to let buyers of capacity bid for themselves. The Independent System Operator itself, in this case PJM, would simply bid on the basis of the administratively determined demand curve, would purchase capacity at the market-clearing price, and then pass on the costs to LSEs. At each auction, only sellers would submit bids, consisting of offered quantities of capacity and the lowest price they would be willing to accept, from which a supply curve would be constructed. The intersection of that curve with the designed demand curve would determine the market-clearing price to be paid for all capacity. PJM management adopted this idea and promoted it in the discussions on redesign of the capacity market. Representatives of the generation side, that is, sellers of capacity, quickly registered their support for the design as well. (p 839).
In this view of the world, Breslau suggests, price was the problem. The price was sending the wrong signals — and by coming up with the right price, the designed market (with its new rules) would be an improvement over the previous one.
But consumers fought back. They suggested that the price was not the problem at all; rather the price merely reflected underlying structural problems of power generation and transmission. They pointed out that the proposed market redesign would increase electricity prices for consumers, and at the same time, increase profits for providers.
This resulted in what Breslau terms a “framing contest.” And in this framing contest, the same “facts” took on different meanings or moral overtones. The consumers framed the projected “windfall profits” of the providers to show that the market redesign was not the problem; that this solution would only benefit providers. The providers and the economists however saw the same projected higher profits as an example of a more efficient market, an example of “scarcity rents,” that would eventually lead to more capacity.
Ultimately, the economists won the framing contest. And here, Breslau uses an ANT framework, suggesting that as more and more actors lined up behind the “inefficient market” hypothesis, market redesign came to be seen as the favored solution.
The conflict over the design had taken the form of a framing contest, with buyers and sellers supporting opposed frameworks within which to formulate and justify measures to assure reliability. The stakes of that contest were whether ongoing reliability concerns would be treated as a problem of market design at all, and whether the demand curve approach put forth by generators and PJM would be adopted. But once sufficient forces were lined up in support of the new capacity market design, with its downward-sloping demand curve, the struggle proceeded with the victorious market frame as its ground. This point was reached on 20 April 2006, when FERC formally accepted most of the elements of the RPM design as a ‘just and reasonable’ replacement for PJM’s existing capacity market. (p843).
From this point onwards, the debate became a different one. It became one about setting the parameters of the demand curve. This, Breslau suggests, is how the performativity of economics works in practice. It provides a frame — the demand curve — about which the different participants can argue and reach some sort of agreement. An argument about the shape of the demand curve is an argument over the rules of the market. But the market itself is taken for granted.
As someone not very well acquainted with electricity markets, I really enjoyed reading the paper. Still, the article left me genuinely puzzled. Isn’t this a great empirical study of the hegemony of markets as a framing device — especially through the idealized demand-supply curves that economists produce so regularly? Do we need performativity at all in this account? And if we do, how exactly is performativity different from hegemony? Thoughts?