The Invisible Manager: Rating Systems and Labor

Collage Rating in Uber App
Collage: Rating in the Uber App

By David Carranza, Master student in Public and Global Health Program, Fulbright scholarship holder

Think of the last time you shopped online or looked for a new restaurant to try – did a user-generated review or star rating determine whether you clicked “buy” or decide to go to the restaurant? Reviews are not new, and feedback has been integral to all sectors of the economy for some time. However, the ever-growing digital economy combined with the ubiquitous use of apps has culminated in the proliferation of user-based rating systems.

As the transactions take place between strangers, user-generated ratings serve two essential purposes: it ensures that products and services are safe, legitimate, and reliable while also promoting enjoyable and customer-friendly sellers. Without direct supervision, rating systems have become the ‘invisible manager’ across the platform economy, essentially promoting those with good ratings, and pushing those with bad ratings out of the market. Customers are happier, and the best entrepreneurs crowd out the worst ones. This might seem fair and the logical conclusion of an online marketplace of strangers but relying on rating systems has its drawbacks.

First, reviews are prone to bias from customers.  Replacing a human resources department with something as simple as a five-star rating system, for example, allows customers to give a rating completely unrelated to the service or product. Poor ratings due to discriminatory perceptions of the platform worker have been increasingly documented (Lee, 2019; Ducato, Kullmann & Rocca, 2019). Poor ratings directly affect the ability to find or maintain work. In the case of Uber in the United States, an aggregate rating of 4.6 out of 5  or lower will shut out drivers from the app temporarily, denying drivers of work (Hill, 2019). The overall result is the rating system, which is ostensibly responsible for ensuring high levels of customer service and safety, is also responsible for frustration and anxiety among the drivers whose ability to find work is tied to their customers’ reviews (Rosenblat et al., 2017; Wu & Li, 2019).

Second, besides being subjective, reviews are in general predisposed to reputation inflation. This term explains the tendency of reviews to be skewed towards high ratings. (McDaid, Boedker & Free, 2019; Athey, Castillo & Chandar, 2019). In addition, there tends to be a positive drift in ratings over time, even if the same phrases are used such as ‘good job’ or ‘terrible’ (Filippas, Horton & Golden, 2019). A rating system that drifts towards the maximum ‘good’ rating will tend to diminish the accuracy of ratings both on and off the platform economy.

Overall, many rating systems are faulty and inaccurate due to reputation inflation and being a poor reflection of the service provided, while also being susceptible to discrimination and bias. Yet as ratings have been shown to help determine one’s ability to find work, the worker is governed by the invisible manager of a fundamentally flawed rating system.

References:

Athey, S., Castillo, J. C., & Chandar, B. (2019). Service quality in the gig economy: Empirical evidence about driving quality at uber. Rochester, NY: doi:10.2139/ssrn.3499781 Retrieved from https://papers.ssrn.com/abstract=3499781

Ducato, R., Kullmann, M., & Rocca, M. (2019). Five stars wars – european legal perspectives on customer ratings and discrimination. Rochester, NY: doi:10.2139/ssrn.3362470 Retrieved from https://papers.ssrn.com/abstract=3362470

Filippas, A., Horton, J. J., & Golden, J. (2019). Reputation inflation. SSRN Electronic Journal, doi:10.2139/ssrn.3136473

Hill, A. (2019). Ratings systems have returned to haunt the gig economy. FT.Com, https://www.ft.com/content/eb8b7c0e-ce71-11e9-99a4-b5ded7a7fe3f

Lee, D. (2019). ‘Thrown to the wolves’ – the women who drive for uber and lyft. BBC News Retrieved from https://www.bbc.com/news/technology-46990533

McDaid, E., Boedker, C., & Free, C. (2019). Close encounters and the illusion of accountability in the sharing economy. Accounting, Auditing, & Accountability, 32(5), 1437-1466. doi:10.1108/aaaj-09-2017-3156

Rosenblat, A., Levy, K. E. C., Barocas, S., & Hwang, T. (2017). Discriminating tastes: Uber’s customer ratings as vehicles for workplace discrimination. Policy and Internet, 9(3), 256-279. doi:10.1002/poi3.153

Wu, Q., Li, Z. (2019). Labor control and task autonomy under the sharing economy: a mixed-method study of drivers’ work. J. Chin. Sociol. 6, 14. https://doi.org/10.1186/s40711-019-0098-9