Semiparametric identification in duration models
This paper explores the identifiability of ratios of derivatives of the index function in a model of a duration process in which the impact of covariates on the hazard function passes through a single index. The model allows duration and the index to appear in a nonseparable form in the hazard function and includes a latent heterogeneity term which acts multiplicatively on the hazard function. The model allows covariates to be endogenous, that is to be correlated with the heterogeneity term. Quantile invariance, local order and local rank conditions are shown to be sufficient to permit identification of ratios of derivatives of the index function. The framework constructed in this paper is suitable for the analysis of identification in panel duration models with heterogeneity.
|Date of creation:||13 Nov 2002|
|Date of revision:|
|Contact details of provider:|| Postal: The Institute for Fiscal Studies 7 Ridgmount Street LONDON WC1E 7AE|
Phone: (+44) 020 7291 4800
Fax: (+44) 020 7323 4780
Web page: http://cemmap.ifs.org.uk
More information through EDIRC
|Order Information:|| Postal: The Institute for Fiscal Studies 7 Ridgmount Street LONDON WC1E 7AE|
When requesting a correction, please mention this item's handle: RePEc:ifs:cemmap:20/02. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Emma Hyman)
If references are entirely missing, you can add them using this form.