Markets, market algorithms, and algorithmic bias

ABSTRACT

Where economists previously viewed the market as arising from a ‘spontaneous order’, antithetical to design, they now design markets to achieve specific purposes. This paper reconstructs how this change in what markets are and can do came about and considers some consequences. Two decisive developments in economic theory are identified: first, Hurwicz’s view of institutions as mechanisms, which should be designed to align incentives with social goals; and second, the notion of marketplaces – consisting of infrastructure and algorithms – which should be designed to exhibit stable properties. These developments have empowered economists to create marketplaces for specific purposes, by designing appropriate algorithms. I argue that this power to create marketplaces requires a shift in ethical reasoning, from whether markets should reach into certain spheres of life, to how market algorithms should be designed. I exemplify this shift, focusing on bias, and arguing that transparency should become a goal of market design.

KEYWORDS:

• Markets
• algorithms
• market design
• bias
• transparency

Acknowledgements

The author thanks Luc Bovens, Elinor Clark, Enno Fischer, Jurgis Karpus, Donal Khosrowi, Sebastian Krug, Chrys Mantzavinos, Luis Enrique Segoviano Contreras, Lucie White, and Jannik Zeiser for helpful comments and discussions at various stages of this research. The author also thanks the audiences at the following events: the Soul of Economics conference, University of Zurich, 2019; the International Research Network Kick-Off Meeting ‘Justice and Interest’, Aix-Marseille School of Economics, 2019; the Philosophy of Economics Research Seminar, UNAM, Mexico City, 2020; the Philosophy of Social Science Roundtable, Emory University, Atlanta, 2020; and the 15th Biennial Meeting of the International Network of Economic Method (INEM 2021) in Tempe, Arizona. Financial support from the VolkswagenStiftung within the project ‘Bias and Discrimination in Big Data and Algorithmic Processing: Philosophical Assessments, Legal Dimensions, and Technical Solutions’ is gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 It is possible that for Hurwicz the market could sometimes also be a spontaneous order that developed in the absence of design, but the concept of spontaneous order does not appear to play an important role in his writings.

2 Roth (2018) adds that customs are also part of marketplaces, as local customs and norms may affect outcomes (see pp. 1610 and 1615). For our purposes, customs can be ignored.

3 See van Basshuysen (2021) for an attempt at providing a general methodology of market design.

4 It might be objected here that what is designed are not markets but mere marketplaces, and therefore market design is presumably consistent with the view of the market as a spontaneous order. However, since marketplaces are parts of, embedded within, larger markets, the latter are, at least partly, subject of design too. I thank an anonymous reviewer for pressing me on this point.

5 Brennan and Jaworski argue that especially anti-commodification theorists have tended to overlook that their arguments concern contingent features of particular markets, rather than the general features shared by all kinds of markets. This seems to motivate a similar claim: rather than debating the reach of ‘the market’, moral philosophers should focus on the ways in which particular markets are designed.

6 An example of an evaluation of a specific design is van Basshuysen (2020) who argues that removing barriers to kidney exchange programs is a moral imperative. Hitzig (2020) seeks to establish general claims, arguing that design economics should allow for democratic participation.

7 An exception is a paper by Friedman and Nissenbaum who discuss possible bias in the context of the medical match prior to its redesign (1996). We shall encounter their arguments presently.

8 This problem may be particularly severe in two-sided marketplaces, such as matching markets and some auctions, where, in Roth’s dictum, ‘you can’t just choose what you want, even if you can afford it: you also must be chosen’ (Roth, 2018, p. 1612). If an algorithm is biased in favor of one side in such a marketplace, members of this side will stick to the marketplace that tends to favor them, and will thereby eliminate outside options for members of the disadvantaged side.

9 These are matching markets, such as in Gale and Shapley’s marriage problem, in which there are no couples.

10 The algorithm also practically makes it a dominant strategy for applicants and programs to state their true preferences, making unlikely a different possible source of bias, namely bias due to strategic behavior (see below).

11 Technically speaking, the impossibility of finding unbiased stable matchings when the set of stable matchings is large is due to the fact that this set is a distributive lattice (Knuth, 1996).

12 Cf. Kominers and Teytelboym (2020) who discuss a subset of these factors as kinds of complexity that designers must overcome to achieve their goals. It is natural to think that the complexity of a design task can add to the epistemic opacity of the designed algorithm.

13 That opacity is relative to a cognitive agent is a general feature, see Humphreys (2009).

14 These data have in the meantime been released, but it appears that their implications have not yet been studied in detail.

15 Feltri (2020) suggested there were possible conflicts of interest arising from the academic and business relations between various market designers, including Paul Milgrom. Milgrom (2020) responded to both Weyl (2020) and Feltri (2020), claiming that the allegation that the auction favored private equity firms was unfounded, and rejecting the allegations that there were problematic relations among the designers.

Additional information

Funding

This work was supported by Volkswagen Foundation.

Notes on contributors

Philippe van Basshuysen

Philippe van Basshuysen is a postdoctoral researcher at the Leibniz University Hannover and research associate at the Centre for the Philosophy of Natural and Social Science at the London School of Economics. His research is in the philosophy of economics and public policy, with a focus on public health.