Determinants of Legislative Longevity and Survival in the United States Congress: Evidence from Cox Proportional Hazards and Weibull Survival Models

This study investigates the determinants of legislative longevity and survival in the United States Congress using survival econometric techniques. Drawing on congressional data from the 46th to the 117th Congress obtained from Voteview, the study examines how ideological positioning, party affiliation, chamber membership, demographic characteristics, and legislative participation influence the duration of congressional service. The analysis employs the Cox Proportional Hazards Model as the primary estimation technique, complemented by a Weibull survival model for robustness testing. The results indicate that ideological extremity significantly increases the probability of congressional exit. Specifically, NOMINATE Dimension 1 increased the hazard of exit by 54.1% (HR = 1.541, p < 0.01), while NOMINATE Dimension 2 increased the hazard by 12.4% (HR = 1.124, p < 0.05). Senate membership reduced exit risk by approximately 40.2% (HR = 0.598, p < 0.01), whereas greater legislative participation and older cohorts were associated with longer congressional survival. Party affiliation was not statistically significant. Marginal-effects estimates reveal that ideological extremity reduces expected congressional tenure by up to 1.84 years, while Senate membership increases expected survival duration by 4.36 years. Robustness checks using the Weibull model confirm the consistency of the findings, and proportional-hazard diagnostics indicate that the Cox model assumptions are satisfied. The study concludes that congressional longevity is primarily shaped by ideological positioning, institutional advantages, accumulated political experience, and legislative engagement rather than partisan affiliation. These findings provide important insights into the dynamics of congressional careers and the broader consequences of political polarization for democratic representation and legislative stability.