The credibility revolution established rigorous standards for causal inference by developing research designs that remove or isolate endogeneity. Yet in many consequential social settings — labor markets, public institutions, health systems, and political organizations — discretionary decisions do not merely confound outcomes; they constitute the very processes that generate them. Abstracting away from these decisions, as standard identification strategies require, changes what is actually being studied.
This project brings together researchers across Aarhus University from economics, political science, management, and the health sciences. The goal of this interdisciplinary group is to reframe endogeneity not as a methodological flaw to be eliminated, but as a theoretical feature to be incorporated into research design.
By combining theoretical frameworks for modeling agency with design strategies that keep discretionary decisions visible within the causal account, the project aims to close the growing gap between what current methods can identify and what social scientists most need to explain.
Banner image credit: Gabriel Rollenhagen, Nucleus Emblematum Selectissimorum (1611), Book 1, Emblem 13 (Remigio et Ventis Secundis, meaning “By rowing and with favorable winds”). Source: Internet Archive.
Tünde Cserpes,
Associate Professor, Department of Management,
School of Business and Social Sciences, Aarhus University
E-mail: tunde.cserpes@mgmt.au.dk
TBA
August 2026 - July 2028
