Hi, I am a Postdoctoral Research Fellow at Humboldt-Universität zu Berlin. My research interests are in applied microeconomic theory, specifically in the areas of organisational, personnel and behavioural economics. My work is funded by the German Research Foundation as part of the Collaborative Research Centre TRR 266 Accounting for Transparency.
harvey.upton(at)hu-berlin.de
When Should a Firm Employ its Workers? A Relational Contracting Approach (with Anja Schöttner)
Abstract: A principal contracts with an agent to repeatedly perform a task. The principal can choose between two organizational forms: employment or self-employment of the agent. The key distinction between the organizational forms is that, under employment, the principal is allowed to stipulate how the agent performs the task, whereas this is legally forbidden under self-employment. Due to incomplete contracts, the labor relationship must be governed by relational agreements, which are affected by the allocation of decision rights inherent to each organizational form. We show that employment typically leads to an overprovision of the agent’s action, compared to an underprovision under self-employment, and that employment relationships tend to be (i) more rigid and (ii) better suited to favorable production environments, relative to self-employment. We then use our model to analyze the impact of taxation, minimum wages and formal performance pay on the outcome of relational contracting and the ensuing implications for the principal’s choice of organizational form.
Wage Transparency in Labor Markets under Relative Income Concerns (with Dominique Demougin)
Abstract: We model an oligopsonistic labor market where firms with heterogeneous productivities hire workers whose preferences exhibit relative income concerns. We then analyze an increase in wage transparency that expands workers' reference groups. We find that larger reference groups induce less aggressive competition for workers, lowering average wages and increasing wage compression. This shifts employment away from high-productivity firms, reducing aggregate output and, when labor market frictions are low, total welfare. Our findings highlight a novel mechanism through which increased wage transparency may negatively impact labor market outcomes, by affecting wage competition between firms.
Implicit Incentives and Delegation in Teams
Management Science (2024), 70(7), pp. 4722-4741. [DOI, pdf and online appendix]
Abstract: We study an infinitely-repeated game of team production, where agents must supply costly effort under moral hazard. The principal also has the option to delegate an additional production-relevant decision to a team member. We provide conditions under which delegation changes the scope of peer sanction and thus influences the implicit incentives generated by the agents’ repeated interaction. Delegation can then become strictly optimal, despite misaligned preferences and symmetric information regarding the efficient decision. We show that implicit incentives under delegation are strongest in diverse teams and use our results to discuss various aspects of organisational design, including self-organised teamwork.
Individual versus Collective Bargaining under Relative Income Concerns (with Dominique Demougin)
Economics Letters (2023), 233, pp. 1-4. [DOI and pdf]
Abstract: We compare individual and collective bargaining when workers have relative income concerns and employment relationships are characterized by moral hazard. We show that collective bargaining internalizes externality effects that arise from other-regarding preferences. This improves firms’ abilities to create effort incentives and can therefore reduce inefficiencies associated with asymmetric information. We show that if relative income concerns are not too strong, both firms and workers strictly prefer collective bargaining.
Relative Income Concerns and the Easterlin Paradox: A Theoretical Framework (with Dominique Demougin)
Economic Modelling (2023), 127, pp. 1-10. [DOI and pdf]
Abstract: This paper analyzes the relationship between relative income concerns and individual well-being in an economy of heterogeneous firm-worker pairs characterized by moral hazard with respect to the worker’s effort. At the individual level, we use the standard principal-agent paradigm to study the design of incentive contracts when workers compare their earnings with the economy’s average wage. At the aggregate level, we define the economy’s equilibrium and prove that it always exists. We then analyze the impact of technological improvements and identify various channels through which economy-wide income growth can negatively affect an individual’s well-being. We show that in many cases, an increase in average earnings generates the Easterlin Paradox, whereby average expected utility for workers is stagnant or even reduced. We discuss empirical predictions resulting from our model and suggest how our analysis can help evaluate the effects of various policy interventions.
Relative Income Concerns, Dismissal, and the Use of Pay-for-Performance (with Dominique Demougin)
The BE Journal of Theoretical Economics (2023), 23(1), pp. 405-441. [DOI of final version at www.degruyter.com and pdf]
Abstract: This paper studies optimal incentive contracting under moral hazard when workers exhibit relative income concerns and compare their earnings with the economy’s average wage. We show that when firms have access to a rich performance measure, the optimal contract takes a binary form if effort is sufficiently low and a ternary form otherwise. We then use these results to investigate how contractual structure varies throughout the economy when firm-worker pairs are heterogeneous with respect to either their productivity, or the information system used to align incentives. We argue that our findings suggest that the incidence of pay-for-performance should be highest for jobs which significantly contribute to overall firm profitability and for which a worker’s performance is difficult to measure. These predictions appear largely consistent with recent empirical evidence.