格林 面板数据讲义-20a-hazardmodels.pptVIP

Econometric Analysis of Panel Data Econometric Analysis of Panel Data 20A. Hazard and Duration Models Modeling Duration Time until business failure Time until exercise of a warranty Length of an unemployment spell Length of time between children Time between business cycles Time between wars or civil insurrections Time between policy changes Etc. Hazard Models for Duration Basic hazard rate model Parametric models Duration dependence Censoring Time varying covariates Sample selection The Hazard Function Hazard Function A Simple Hazard Function Duration Dependence Parametric Models of Duration Censoring Accelerated Failure Time Models Proportional Hazards Models Estimation Time Varying Covariates Unobserved Heterogeneity Interpretation What are the coefficients? Are there ‘marginal effects?’ What is of interest in the study? A Semiparametric Model Nonparametric Approach Based simply on counting observations K spells = ending times 1,…,K dj = # spells ending at time tj mj = # spells censored in interval [tj , tj+1) rj = # spells in the risk set at time tj = Σ (dj+mj) Estimated hazard, h(tj) = dj/rj Estimated survival = Π [1 – h(tj)] (Kaplan-Meier “product limit” estimator Kennan’s Strike Duration Data Kaplan Meier Survival Function Hazard Rates Hazard Function Weibull Model Weibull Model Survival Function Hazard Function Loglogistic Model Loglogistic Hazard Model Sample Selection Building a Likelihood for a Weibull Duration Model with Selection Building the Likelihood Conditional Likelihood Weibull Model with Selection Strategy: Hermite quadrature or maximum simulated likelihood. Not by throwing a ‘lambda’ into the likelihood Could this be done without joint normality? How robust is the model? Is there any other approach available? * * William Greene Department of Economics Stern School of Business +---------------------------------------------+ | Loglinear survival model: WEIBULL | | Log likelihood function -97.39018 | | Number of parameters