The Allocation of Visible and Hidden Work: An Experimental Investigation
How to cite
Pan, Xiaofei. 2026. "The Allocation of Visible and Hidden Work: An Experimental Investigation." AEA RCT Registry. AEARCTR-0018818. https://doi.org/10.1257/rct.18818-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
We investigate how partial workload transparency influences distributive preferences using a controlled online Prolific experiment. Participants in the role of allocators decide how to distribute additional workload between a pair of peers (Participant 1 and Participant 2) who are randomly assigned differing initial task endowments. We systematically vary the information for each decision scenario: tasks include both visible (publicly observable to both peers) and hidden domains (private information to the assigned individual), while the allocator observes all domains. Utilizing a within-subject design across 12 distinct scenarios, allocators choose between three options that either equalize total cumulative workloads (by removing the initial endowment inequality), maintain initial workload inequality, or exaggerate it. Crucially, we vary whether the additional tasks are allocated into the visible or hidden domains across numerically identical decision matrices. This design allows us to cleanly isolate whether allocators prioritize true distributive fairness or the visible equality when peer workloads are only partially observable.
Experimental design
Allocators will decide how to allocate 6 additional tasks between two workers (Worker 1 and Worker 2).
The two workers are randomly endowed with a different number of initial tasks, such that Worker 1 is endowed with exactly 2 more tasks than Worker 2. The task allocations take place across two distinct information environments: Visible (where workers can observe both their own and their peer's assigned tasks) and Hidden (where tasks are not visible to the peers). While the allocator can always observe the full number of tasks assigned to both workers, the workers can only observe their peer's workload when those tasks land in the Visible domain.
To evaluate the effect of visibility, we compare how allocators choose to distribute these additional tasks when they assign only to the visible domain versus a numerically identical yet hidden task domain.
Allocators will make twelve distinct allocation decisions. In each decision, they must choose between three mutually exclusive options to distribute the 6 additional tasks:
• Favor 2 (Total Task Equality): Assign 2 tasks to Worker 1 and 4 tasks to Worker 2. This option completely removes the baseline endowment workload inequality, resulting in equal final cumulative workloads.
• Equal (Maintains Current Inequality): Equally split the 6 tasks (3 to Worker 1, 3 to Worker 2). This option preserves the initial baseline endowment inequality.
• Favor 1 (Exaggerated Inequality): Assign 4 tasks to Worker 1 and 2 tasks to Worker 2. This option actively exacerbates the initial baseline endowment inequality.
Design details
Cited by
No citation data available for this registration yet.