Do People Like Algorithms? A Research Strategy

19 Pages Posted: 19 Aug 2023

See all articles by Cass R. Sunstein

Cass R. Sunstein

Harvard Law School; Harvard University - Harvard Kennedy School (HKS)

Lucia Reisch

Copenhagen Business School; Zeppelin University Friedrichshafen

Date Written: August 18, 2023

Abstract

Do people like algorithms? In this study, intended as a promissory note and a description of a research strategy, we offer the following highly preliminary findings. (1) In a simple choice between a human being and an algorithm, across diverse settings and without information about the human being or the algorithm, people in our tested groups are about equally divided in their preference. (2) When people are given a very brief account of the data on which an algorithm relies, there is a large shift in favor of the algorithm over the human being. (3) When people are given a very brief account of the experience of the relevant human being, without an account of the data on which the relevant algorithm relies, there is a moderate shift in favor of the human being. (4) When people are given both (a) a very brief account of the experience of the relevant human being and (b) a very brief account of the data on which the relevant algorithm relies, there is a large shift in favor of the algorithm over the human being. One lesson is that in the tested groups, at least one-third of people seem to have a clear preference for either a human being or an algorithm – a preference that is unaffected by brief information that seems to favor one or the other. Another lesson is that a brief account of the data on which an algorithm relies does have a significant effect on a large percentage of the tested groups, whether or not people are also given positive information about the human alternative. Across the various surveys, we do not find persistent demographic differences, with one exception: men appear to like algorithms more than women do. These initial findings are meant as proof of concept, or more accurately as a suggestion of concept, intended to inform a series of larger and more systematic studies of whether and when people prefer to rely on algorithms or human beings, and also of international and demographic differences.

Keywords: Algorithms, algorithm aversion, algorithm appreciation

JEL Classification: K, K1, D11, D18

Suggested Citation

Sunstein, Cass R. and Reisch, Lucia, Do People Like Algorithms? A Research Strategy (August 18, 2023). Available at SSRN: https://ssrn.com/abstract=4544749 or http://dx.doi.org/10.2139/ssrn.4544749

Cass R. Sunstein (Contact Author)

Harvard Law School ( email )

1575 Massachusetts Ave
Areeda Hall 225
Cambridge, MA 02138
United States
617-496-2291 (Phone)

Harvard University - Harvard Kennedy School (HKS) ( email )

79 John F. Kennedy Street
Cambridge, MA 02138
United States

Lucia Reisch

Copenhagen Business School ( email )

Solbjerg Plads 3
Frederiksberg C, DK - 2000
Denmark

Zeppelin University Friedrichshafen ( email )

Am Seemooser Horn 20
Friedrichshafen, 88045
Germany

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
272
Abstract Views
1,449
Rank
214,246
PlumX Metrics