Five years of being wrong about specific things while right about the structural analysis. What the errors reveal about the limits of institutional analysis.
The Analytical Limits
This blog has been wrong in ways that are instructive about the limits of institutional analysis as a framework. It has been wrong about specific timelines — the institutional reforms that the analysis predicted would be difficult proved even more difficult than predicted, and the institutional failures that seemed imminent were delayed by the specific resilience mechanisms that the analysis underweighted. It has been wrong about specific actors — the individual decision-makers who made choices that the institutional analysis of their incentive structures suggested they would not make, and whose choices altered the institutional trajectory in ways that the structural analysis did not predict. And it has been wrong about the pace of cultural change — the shift in public norms around specific institutional practices has sometimes moved faster than institutional analysis expected and sometimes much slower.
The errors cluster around a consistent limitation: institutional analysis is better at describing the structural conditions that make specific outcomes likely than at predicting the specific timing, the specific actors, and the specific sequences through which those outcomes occur. The structural condition that makes reform difficult is analytically visible; the specific political moment when reform nevertheless happens is not predicted by the structural analysis alone. The individual who defies the incentive structure that the institutional analysis describes is always a possibility that the structural analysis cannot eliminate. These limitations are not failures of the framework — they are the inherent limits of structural analysis as a predictive tool.
What the blog got wrong was primarily in prediction rather than description. The structural conditions were described accurately; the specific events those conditions produced were sometimes wrong in their timing, their agents, and their sequence. That gap — between structural description and event prediction — is the permanent limitation of institutional analysis, and acknowledging it is the precondition for using the framework accurately.
Discussion