This study also uses the natural experiment method but examines the duration/benefit relationship using permanent partial impairment claims filed before and after four different legislative benefit changes in New York state.
So, when estimated using the natural experiment method without covariates, the most reliable estimates of the duration/benefit elasticities are 0.751 for severe impairments (1968; the 1965 estimate of 1.163 is significant at a lower level) and 0.238 for minor impairments (1978).
Unlike the natural experiment method, cross-section analysis does not rely on statutory changes in benefits.