Political risk has long been theorized as a key factor that drives the allocation of foreign
direct investment (FDI). Frequently dened as the likelihood a host government will engage
in actions that threaten an investor's protability, political risk is thought to be remedied
by political institutions that allow countries to make more credible commitments to honor
their pre-investment promises. As a latent variable, scholars have struggled with ways to link
institutions to FDI ows in the absence of a measure of political risk. This paper presents a
dynamic hierarchical model of political risk that aims to address this problem by producing from
observable indicators a single time-series cross-sectional measure of political risk. Additionally,
this model produces estimates that allow scholars to measure within-country variation in political
risk, a previously unestimated quantity. The model itself is exible enough to be of use to
comparative scholars more generally to estimate latent variables at multiple levels of a hierarchy
from observable indicators produced at dierent levels of that hierarchy.