This paper examines the effects of unemployment insurance on escape rates from unemployment using data from the 1998 Displaced Worker Survey. Transitions from unemployment to employment are modeled using a flexible representation of the baseline hazard function and allowing for discrete changes through time in the effects of unemployment insurance benefits, as well as those of the other covariates. The impact of unemployment insurance is also modeled using a time-varying benefits measure, namely, time to exhaustion of benefits. Potential biases stemming from reverse causation and unobserved individual heterogeneity are accommodated. Both approaches render transparent the major disincentive effects of access to benefits on re-employment rates while also providing evidence of time-varying effects of other regressors.