The model (also) in the world: extending the sociological theory of fields to economic models

ABSTRACT

The goal of this paper is to provide a methodological perspective on economic models that accounts for some sociological dimensions of economics, in two senses. Firstly, we are interested in how modeling is an activity that is constrained by the (implicit and explicit) rules underlying the accumulation of academic prestige within economics and at the same time can be a means to change these rules. Secondly, we are interested in how, for a given model, this dynamic can be influenced by the use(s) of that model outside of economics. We first provide a restatement of Brisset’s [(2018). Models as speech acts: The telling case of financial models. Journal of Economic Methodology, 25(1), 21–41] original contribution. We then put this clarified methodological perspective to work on a new case study, i.e. on the dual models at the frontier between behavioral and standard economics.

KEYWORDS:

• Models
• speech act
• Pierre Bourdieu
• finance
• dual models

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Nicolas Brisset is an associate professor at the Université Côte d’Azur (GREDEG-CNRS). His areas of specializationare the philosophy of social sciences, history of economic thought and economic sociology. More precisely, his research focus on three issues: the philosophy of economic models; the history of economic thought under the Vichy regime; and the limits of commodification.

Dorian Jullien is a research associate at the Université Côte d’Azur (GREDEG-CNRS). His work focuses on the history, philosophy and methodology of behavioral and experimental economics, with an emphasis on the role of language uses in economic rationality.

Notes

1 We only discuss Morgan, Boumans and Hédoin explicitly, but the issues we discuss also concern at least three literatures on models, i.e. in the philosophy of science (e.g. Suárez, 2004), in the philosophy and methodology of economics (see the references in Morgan and Knuuttila, 2012) and in the history of economics, especially in the ‘Historical epistemology’ approach (e.g. Lallement, 1984; Düppe and Maas, 2017; Stapleford, 2017).

2 The legacy of positivist approaches is still quite noticeable in contemporary analytic philosophy of language (see, e.g. Soames, 2010).

3 How to Do Things With Words was first published in 1962 (after Austin’s death in 1960) and is the reconstruction of a set of public lectures he delivered at Harvard in 1955 from his and some attendee’s notes.

4 Strictly speaking, in Austin’s terminology we should speak here of ‘total speech act’ because we include the effect of the speech act (the ‘perlocution’) in its definition, instead of ‘speech act’ tout court. In the example we have given this distinction is non-trivial especially for degrading and reassuring, which are both strictly speaking much more effects than speech acts proper. However, given our heuristic use of Austin’s work here that is not crucial for our arguments or for our applications in the case studies.

5 See Loewenstein and O’Donoghue (2007), Loewenstein et al. (2015), Fudenberg and Levine (2011; 2012), Fudenberg, Levine, and Maniadis (2014), Dreber, Fudenberg, Levine, and Rand (2016), Brocas and Carrillo (2008b, p. 2014), Alonso, Brocas, and Carrillo (2014).

6 For similar statements, see Camerer, Loewenstein, and Prelec (2005, p. 43, p. 56), Ashraf, Camerer, and Loewenstein (2005, p. 132), Camerer (2006, pp. 199–200, 2008, p. 55, pp. 62–63), Loewenstein, Rick, and Cohen (2008, p. 650 and p. 665), Angner and Loewenstein (2012, p. 676).

7 There are a few other studies of these models (e.g. Ross, 2010; Ross, Ainslie, & Hofmeyr, 2010; Ainslie, 2012; Alós-ferrer & Stark, 2014; Brocas & Carillo, 2014), but they all take a rather narrow angle of analysis that is not very useful for our purpose.

8 Both are originally published in the American Economic Review, with over 950 citations on Google Scholar for Fudenberg and Levine (2006) and nearly 700 for Bernheim and Rangel (2004). By comparison, Benhabib and Bisin (2005, Games and Economic Behavior) have over 250 citations, Brocas and Carillo (2008a, American Economic Review) have nearly 200 citations, and Loewenstein and O’Donoghue have nearly 400 citations for the original 2004 version of their dual model, which is the most cited version even though it is a working paper (later published in the interdisciplinary journal Decision with Sudeep Bathia as a new co-author). These data were collected in November 2018.

9 It can be argued that these are the main current uses of dual models in economics. We do not discuss them in details because we are more interested in the general change of conventions in microeconomics implied by dual models than the narrower corresponding changes that their uses imply in different sub-fields, i.e. the issues related to multi-utility representations in decision theory that underlies the contributions discussed by Lipman and Pesendorfer (2013) and the issues of so-called mixture models in statistical modeling that underlies the contribution of Andersen et al. (2008). The latter are nicely discussed from a methodological perspective by Don Ross (2016).