Dynamic exploits: calculative asymmetries in the on‐demand economy

• Date: 01 July 2020
• DOI: http://doi.org/10.1111/ntwe.12160
• Published date: 01 July 2020

162 New Technology, Work and Employment © 2020 Brian Towers (BRITOW) and

John Wiley & Sons Ltd.

New Technology, Work and Employment 35:2

ISSN 1468-005X

Dynamic exploits: calculative asymmetries in

the on-demand economy

Aaron Shapiro

On-demand service rms secure market power by cultivating

and operationalising calculative asymmetries between the

platform and labour. In this article, I analyse dynamic (or

‘surge’) pricing as an exemplary calculative technique. I show

how the asymmetrical application of price-setting allows

rms to leverage control at the aggregate level while maintain-

ing the façade of autonomy at the individual level, thereby le-

gitimising workers’ classication as independent contractors

but solving the coordination problems that the classication

introduces. The article complements and extends previous crit-

ical research into the platform or ‘on-demand’ service economy

by analysing how management scientists model and simulate

on-demand marketplaces. I consider management science to

be a calculative technique for optimising operational ef-

ciency. A critical review of management science provides nov-

el insights into platforms’ efforts to monopolise calculative

agency at the expense of other market participants. The arti-

cle concludes by considering implications for broader critiques

ofplatform labourmanagement.

Keywords: algorithms, calculation, control, management science, dynamic pricing, information

asymmetry, on-demand economy, platform labour.

Introduction

In the ‘on-demand’ platform economy, rms secure market power by cultivating and

operationalising calculative asymmetries between the platform and other market agents.

These asymmetries, understood as uneven distributions of calculative agency, derive

from market actors’ relative access to or paucity of calculative equipment. Calculative

equipment not only quanties the qualitative difference between market options (e.g.

the difference in price or quality between one commodity and another); it enables mar-

ket actors to anticipate the likeliest course of action associated with those differences

(Callon and Muniesa, 2005; Callon and Law, 2005). Calculative agency is not deter-

mined by market position; it establishes market position. And in the marketplace for

‘on-demand’ services, platforms hoard calculative agency.

Like most online exchanges, on-demand service platforms purport to generate value

as ‘middlemen’, connecting customers to ‘independent service providers’ while taking

a rent-like fee from each transaction (Bergvall-Kåreborn and Howcroft, 2014; Calo and

Rosenblat, 2017; Rossi, 2019; Jarrahi et al., 2019). However, two factors cast doubt on

the legitimacy of this arrangement, and especially the question of service providers’

Aaron Shapiro Research Fellow, Information Law Institute, New York University.

Dynamic exploits 163

© 2020 Brian Towers (BRITOW) and

John Wiley & Sons Ltd.

‘independence’ as workers. First, unlike other exchanges, on-demand platforms set the

price for the service, controlling what customers pay and what workers receive as pay-

ment for their labour. Second, research shows that service platforms exhibit substantial

information asymmetries, with market information withheld and disclosed strategi-

cally to serve platforms’ operational goals (Rosenblat and Stark, 2016; Shapiro, 2018;

Griesbach et al., 2019). Both of these factors undermine rms’ claim that workers are

not employees but independent contractors—a status that exempts the rm from sub-

stantial overhead costs while excluding workers from workplace protections (Cherry,

2015; De Stefano, 2016; Calo and Rosenblat, 2017; Kennedy, 2017; Collier et al., 2018).

Critical research tends to attribute these disparities to algorithmic control. But this

overlooks the more fundamental role that calculation plays in platform economics. In

general, calculation enjoys a special status in theories of marketisation (Callon, 2016).

Neoclassical economic theory holds that markets are themselves efcient calculators;

market players are inherently calculative and goods are inherently calculable; any

transaction in the market therefore proceeds only by the grace of calculation. By con-

trast, sociologists in the science and technology studies (STS) tradition understand cal-

culation to be the outcome of a complex coordination between market actors, materials

and institutional settings (Callon and Muniesa, 2005; cf. Çalışkan and Callon, 2010;

Callon, 2016; Callon et al., 2007; Mackenzie et al., 2007). Such coordination is never

neutral. Calculative practices ‘alter the power relations that they shape and are embed-

ded within,… enabl[ing] new ways of acting upon and inuencing the actions of indi-

viduals’ (Miller, 2001: 379). The question therefore is not whether a behaviour is

rational or irrational, nor whether actors’ reasoning is quantitative or qualitative, but

how calculative agency is distributed across and between market participants (Callon

and Law, 2005).

Building on this theoretical foundation, I argue that on-demand service platforms

exploit calculative asymmetries that they design into the market architecture.

Specically, I analyse dynamic (or ‘surge’) pricing as an exemplary calculative tech-

nique, showing how the asymmetrical application of price-setting allows rms to exert

control over labour at the aggregate level while maintaining the façade of autonomy

for the individual worker. This legitimises the rm’s classication of workers as inde-

pendent contractors while solving the coordination problems that the classication

introduces.

My argument is intended to complement critical research on platform economics,

which tends to focus on workers’ experiences with or resistance to algorithmic man-

agement (cf. Rosenblat and Stark, 2016; Briziarelli, 2018; Shapiro, 2018; Richardson,

2019; Rossi, 2019). Specically, I extend the analysis in two directions. Conceptually, I

account for the broader processes of calculation of which algorithmic control is but one

component; empirically, I study how on-demand markets are modelled in the eld of

management science, an intellectual arena that exposes platforms’ calculative logics.

Management scientists use mathematical models and computer simulations of on-de-

mand service markets to optimise strategies for operational efciency and revenue

maximisation. To the extent that these models and simulations enable the platform to

anticipate workers’ responses to inputs, their function becomes less like a ‘science’

than an ‘intellectual technology’ thatrms may wield to extract greater value from

customers or labour (cf.Dutton and Starbuck, 1971; e.g.,Kamat and Hogan, 2019).

In the next section, I discuss calculative asymmetries as they pertain to two compet-

ing views of market calculation, interface-markets and market-agencements (Callon,

2016). Where the agencement view helps us understand the contingent coordination

involved in the delivery of platform-goods, here I dwell on the interface-market to

unpack its role in constructing platforms as a market architecture. In the third section,

I discuss dynamic pricing as an emblematic calculative technique while situating the

practice as an exercise in managerial control. In the fourth part, I introduce the man-

agement science literature to demonstrate how dynamic pricing renders the workforce

calculable through labour-supply elasticities. In the fth section, I consider dynamic

pricing’s interaction with other management techniques. I conclude with a discussion