The edge of performativity…
August 26, 2011
I’d like to use this blog to clarify a conceptual area that is causing me considerable difficulty; I’m at the end of my theoretical tether and would appreciate any input that fellow readers can offer…
I’m trying to conceptualise the generative process of economic agency that gives rise to a) specific economic behaviour, and b) behavioural effects among a group of nonprofessional investors. Because these individuals are (or at least began as) investing novices, the achievement of any kind of competent investment appears to be the result of a configuration at the hands of investment service firms. On the other hand, there is a lot of incompetence too, playing into the hands of the behavioural theorists. My ethnographic surprise is not, however, that they make mistakes, but that they achieve any kind of economic agency at all.
Beunza and Stark’s Models, Reflexivity and Systemic Risk (http://ssrn.com/paper=1285054) has a good go at theorising a dynamic process as it occurs among skilled professionals, so I follow their lead. Behavioural Finance assumes that substandard investment decisions are a result of bounded rationality, low quality information, leading to the imitation of peers: “It is because agents are thus restricted that they lean on others” (B & S). The opposing view, stemming from Callon’s work, holds that market devices, tools, and models provide the basis for calculation; and the defects of social interaction are compensated by such devices. What does each position say about economic agency? The sociological claim is that economic agency is constituted through assemblages of devices and performative scripts. Behavioural finance, on the other hand, takes economic agency as a given, as does the efficient market theorising that it has displaced.
How then are we to theorise the proliferation of behavioural effects, empirically documented at the level of individual investors? There are two possibilities: that the assemblages are at fault, or that the users are at fault. B&S take these positions forward into a discussion of systemic risk, where faulty assemblages are represented by Black Swan type criticisms of inductive models, and faulty users are represented by traders’ decisions to ignore certain possibilities. They conclude, convincingly, that it’s a bit of both. The decision to ignore certain possibilities is a collective bit of imitation (qua behavioural finance) driven by the appropriate use of assemblages (qua Callon). It’s a clever argument, and I hope that I’ve read it correctly. It also gives the traders the benefit of the doubt in as much as it suggests that they only make mistakes when their clever use of clever machines tells them to. The traders’ existence as PhD holding self questioning rocket scientists at the cutting edge of finance gives a certain amount of credence to this claim.
In the more pedestrian example of non-professional investors, it may not be so clear cut. Empirical evidence tells us that these investors make lots of mistakes. It also implies that they are constituted as market agents by exposure to performative equipment in the form of e.g. trading platforms, charting engines and newspaper cuttings. In the case of one group of investors that I studied, the assemblages steer them towards a traditional kind of investing activity: long-term investing for dividends in small company stocks. These are individuals who don’t know anything about investing outside of what is thrown at them by investment service companies, and I would argue that they are therefore performed; however, because they are e.g. overconfident, or tend to follow their peers and not the dividend yield calculator, they manifest the effects identified by behavioural finance. The behavioural findings are unsurprising. But I’m interested in the achievement of any economic competence, and not the failure to achieve perfect results.
I would like to posit some kind of space for individual agency within the performative construction of economic agency. Putting it bluntly: Investor a presses the button as instructed and makes money, while Investor B chooses not to and loses. Perhaps this is unremarkable, as the argument might run that individuals are only being performed at the very moment of using a given device. (Here B&S’s argument is different in that it shows irrational outcomes as a result of systematic implementation of devices). But if we take a dynamic perspective, and we begin to question the existence of specialised economic agencies per se, then surely we must allow space for error, misjudgement, and imitation. Or maybe not? The argument is complicated further when other human agents (all fallible) are introduced into the assemblage. Those agents could produce performative scripts and devices that are sub optimal, in the sense that the devices reproduce the bounded rationality, poor information, or blatant error of their makers.
B&S introduce the entrenched position of this debate nicely: “This controversy extends a longstanding debate between economists, who underscore the challenges of individual decision-making, and sociologists, who emphasize processes of social interaction: economy versus society, calculation versus sociability”. But isn’t it the case that we are talking more about the assumption of competence? Behavioural finance bemoans the inability of individuals to reach a Platonic ideal of economic competence, the existence of which is never questioned. SSF, the Aristotelian, produces a relationship-based account of the existence of competence, yet seemingly doesn’t worry about the ability of agents to achieve it. (Okay, we don’t need the cheap philosophical references, but it is a blog after all…).
My question: SSF trades in quants, nerds, and pointy heads, people who can play Liar’s Poker while doing Black Scholes in their heads. People who don’t make mistakes. How should we cope with everyday mortals who do?