A class of binary response models for grouped duration data

Abstract

This paper explores the relationship between conventional models for binary response such as the probit and logit, and the proportional hazard (PH) and related specifications for grouped duration data. I outline a general class of hazard models for grouped duration data based upon the choice of period‐specific distribution functions, facilitating a thorough analysis of the implications of various specifications and consideration of various issues of model identification. This class of models nests, among others, the proportional hazard, probit, and logit specifications for interval survival. I consider the implications of various specifications for hazard behaviour, focusing on familiar specifications. While the specifications will generally yield results that are quite similar along a number of dimensions, there are significant differences. The probit model generates non‐proportional effects of variables on the discrete hazard, while the logit and PH tend to show only slight non‐proportionality. Furthermore, while the effects of variables on the derivatives are considerably larger for the probit specification, the time‐pattern of the probit effects is relatively insensitive to changes in explanatory variables. I illustrate these issues by providing an example taken from Katz's (1986) unemployment data from the Panel Study of Income Dynamics.

Exploring the duration of EU imports

“…While it can be shown that a stacked binary choice model employing a complementary log-log (cloglog) link function represents the exact groupedduration analogue of the Cox proportional hazards model, the more familiar logit and probit specifications do not imply the proportional hazards assumption (see e.g. Sueyoshi, 1995). Each of these models shares the virtues of the Cox model to allow for right censoring and a nonparametric estimation of the baseline hazard function.…”

The Personality Profiles of Early Adopters of Energy-Efficient Technology

“…If the discrete hazard rate P(t) is the conditional probability that a household i invests in energy efficiency in a particular year t, given that it has not adopted any energy efficiency measures before, straightforward logistic regression can be applied and the survival model can be specified as follows (Allison, 2014;Sueyoshi, 1995):…”

Public Health Insurance and Entry into Self-Employment

“…Therefore, we include twelve dummy variables indicating the duration of the current spell in paid employment in the vector xit. By conditioning on this very flexible specification of the baseline hazard, the probit model of the transition probability in equation (2), estimated on the data in person-period format, can equivalently be written as a general survival model, as described in Jenkins (1995), Sueyoshi (1995), and Caliendo et al (2010). We use annual data because the survey interviews occur once a year and the covariates are not available at higher frequencies.…”