Monopsony in Online Labor Markets

By Arindrajit Dubé, Jeff Jacobs, Suresh Naidu, Siddharth Suri. American Economic Review: Insights, 2020

247 citations recorded by the source corpus.

Despite the seemingly low switching and search costs of on-demand labor markets like Amazon Mechanical Turk, we find substantial monopsony power, as measured by the elasticity of labor supply facing the requester (employer). We isolate plausibly exogenous variation in rewards using a double machine learning estimator applied to a large dataset of scraped MTurk tasks. We also reanalyze data from five MTurk experiments that randomized payments to obtain corresponding experimental estimates. Both approaches yield uniformly low labor supply elasticities, around 0.1, with little heterogeneity. Our results suggest monopsony might also be present even in putatively “thick” labor markets. (JEL C44, J22, J23, J42)

Subfields: Computer Science Applications, Electrical and Electronic Engineering, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/e53204edbb6e27a39830781c353e98e765836d46

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Science Applications

1. Saonee Sarker
2. Arun Rai

In Management Science and Operations Research

1. Erik Snowberg
2. Leeat Yariv
3. Panos Constantinides
4. Arun Rai

Citations in the corpus

• Testing the Waters: Behavior across Participant Pools (2018, background): “See also Dube et al. (2018), who estimate the low labor-supply elasticities on MTurk.”
• Next generation digital platforms : toward human-AI hybrids (2019, background): “…pertain to platform dominance and its consequences, including risk of monopsony where digital labor platforms use their market power to squeeze laborers into lower pay (Dube et al. 2018); and ambiguous net effects on income equality, productivity, and skill development (e.g., Agrawal et al. 2015).”