Evaluating Individual Contributions in Ugandan Businesses: Navigating the Performance Non-Separability Problem

Abstract

A central challenge in team-based organizations is the performance non-separability problem-the difficulty of evaluating individual contributions when only collective output is observable. This issue is particularly acute in resourceconstrained contexts like Uganda, where formal performance metrics are often unavailable. This paper explores how supervisors in Ugandan businesses navigate this dilemma, drawing on the theoretical lens of covariance-based inference (Uribe et al., 2022). We argue that in the absence of direct data, supervisors rely on counterfactual reasoning, inferring individual productivity by observing how team performance fluctuates with changes in membership. Through qualitative analysis of sectors including manufacturing, agriculture, and services, we identify two key factors that strengthen this inferential process: frequent employee rotation and clearly defined hierarchical authority. Rotation across teams creates natural experiments, providing supervisors with the variation needed to isolate individual impact. Meanwhile, employees in high-authority roles attract greater evaluative attention, making their contributions more perceptible. The findings suggest that instead of importing complex performance measurement systems, managers in similar contexts can enhance evaluation accuracy by strategically leveraging organic workplace dynamics. We propose practical implications, including the formalization of rotation practices, training for supervisors on implicit inference, and a heightened awareness of perceptual management by employees. This paper contributes to the understanding of subjective performance appraisal by theorizing it as a rational, inference-based process grounded in the observable realities of team production.