Generalised residuals and heterogeneous duration models: With applications to the Weilbull model

doi:10.1016/0304-4076(85)90073-9

Journal of Econometrics

Volume 28, Issue 1, April 1985, Pages 155-169

https://doi.org/10.1016/0304-4076(85)90073-9 Get rights and content

Abstract

This paper deals with models for the duration of an event that are misspecified by the neglect of random multiplicative heterogeneity in the hazard function. This type of misspecification has been widely discussed in the literature [e.g., Heckman and Singer (1982), Lancaster and Nickell (1980)], but no study of its effect on maximum likelihood estimators has been given. This paper aims to provide such a study with particular reference to the Weibull regression model which is by far the most frequently used parametric model [e.g., Heckman and Borjas (1980), Lancaster (1979)]. In this paper we define generalised errors and residuals in the sense of Cox and Snell (1968, 1971) and show how their use materially simplifies the analysis of both true and misspecified duration models. We show that multiplicative heterogeneity in the hazard of the Weibull model has two errors in variables interpretations. We give the exact asymptotic inconsistency of M.L. estimation in the Weibull model and give a general expression for the inconsistency of M.L. estimators due to neglected heterogeneity for any duration model to O(σ²), where σ² is the variance of the error term. We also discuss the information matrix test for neglected heterogeneity in duration models and consider its behaviour when σ²>0.

References (11)

A.D. Chesher
Testing for neglected heterogeneity
Econometrica
(1984)
D.R. Cox
Some remarks on overdispersion
Biometrika
(1983)
D.R. Cox et al.
Theoretical statistics
(1974)
D.R. Cox et al.
A general definition of residuals
Journal of the Royal Statistical Society B
(1968)
J. Crowley et al.
Covariance analysis of heart transplant survival data
Journal of the American Statistical Society
(1977)

There are more references available in the full text version of this article.

Cited by (70)

What makes growth sustained?
2012, Journal of Development Economics
Citation Excerpt :
This problem, which is a general affliction of this and related literatures, is not easy to solve. Abrevaya and Hausman (1999) and Lancaster (1985) show that hazard estimates using the Weibull distribution are consistent even with measurement error in the dependent variable. However, our setup does not meet the requirements for this result (most notably, it applies only to non-censored data).
We identify structural breaks in economic growth in 140 countries and use these to define “growth spells:” periods of high growth preceded by an upbreak and ending either with a downbreak or with the end of the sample. Growth spells tend to be shorter in African and Latin American countries than elsewhere. We find evidence that growth duration is positively related to: the degree of equality of the income distribution; democratic institutions; export orientation (with higher propensities to export manufactures, greater openness to FDI, and avoidance of exchange rate overvaluation favorable for duration); and macroeconomic stability.
Testing for unobserved heterogeneity in exponential and Weibull duration models
2010, Journal of Econometrics
Citation Excerpt :
For this, Lancaster (1979) and Kalbfleisch and Prentice (1980) assume a conventional gamma distribution for the unobserved heterogeneity and test for its presence by measuring the variance of the gamma distribution. Chesher (1984) and Lancaster (1985) propose an information matrix (IM) test (White, 1982, 1994, ch. 11), as the information matrix equality holds in the absence of unobserved heterogeneity.
We examine the use of the likelihood ratio (LR) statistic to test for unobserved heterogeneity in duration models, based on mixtures of exponential or Weibull distributions. We consider both the uncensored and censored duration cases. The asymptotic null distribution of the LR test statistic is not the standard chi-square, as the standard regularity conditions do not hold. Instead, there is a nuisance parameter identified only under the alternative, and a null parameter value on the boundary of the parameter space, as in Cho and White (2007a). We accommodate these and provide methods delivering consistent asymptotic critical values. We conduct a number of Monte Carlo simulations, comparing the level and power of the LR test statistic to an information matrix (IM) test due to Chesher (1984) and Lagrange multiplier (LM) tests of Kiefer (1985) and Sharma (1987). Our simulations show that the LR test statistic generally outperforms the IM and LM tests. We also revisit the work of van den Berg and Ridder (1998) on unemployment durations and of Ghysels et al. (2004) on interarrival times between stock trades, and, as it turns out, affirm their original informal inferences.
Intershopping duration: An analysis using multiweek data
2004, Transportation Research Part B: Methodological
Citation Excerpt :
Such factors may include inherent individual preferences, household responsibilities/interactions, and other miscellaneous unobserved determinants. It is well established now that failure to control for such unobserved heterogeneity can produce severe bias in the nature of duration dependence and the estimates of covariate effects (Heckman and Singer, 1984; Lancaster, 1985). Table 4 provides the segmentation model results for probabilistically assigning individuals into the erratic and regular shopper segments.
This study examines the rhythms in the shopping activity participation of individuals over a multiweek period by modeling the duration between successive shopping participations. A hazard-based duration model is used to model intershopping duration, and a latent segmentation method is applied to distinguish between erratic shoppers and regular shoppers. The paper applies the methodology to examine the regularity and frequency of shopping behavior of individuals using a continuous six-week travel survey collected in the cities of Halle and Karlsruhe in Germany in the fall of 1999. The empirical results underscore the need to adopt a flexible hazard model form for analyzing intershopping durations. The results also provide important insights into the determinants of the regularity and frequency of individuals’ shopping activity participation behavior.
A continuous-time model of departure time choice for urban shopping trips
2002, Transportation Research Part B: Methodological
Citation Excerpt :
Such factors may include inherent individual preferences for certain departure times, household responsibilities/interactions not captured by the household variables included in the model, and other miscellaneous unobserved determinants. It is well established now that failure to control for unobserved heterogeneity can produce severe bias in the nature of duration dependence and the estimates of the covariate effects (Heckman and Singer, 1984; Lancaster, 1985). Table 3 shows the percentage variation in departure time hazard explained by unobserved heterogeneity for each of the six broad time periods in which the covariates and their effects are allowed to vary.
This paper proposes a continuous-time hazard duration model for urban shopping trip departure time choice. The time frame for the analysis of departure time is the entire day. The continuous-time model uses a non-parametric baseline hazard distribution, employs a non-parametric representation for the time-varying effect of covariates, and accommodates time-varying covariates. These econometric issues are important to recognize and consider in a departure time model spanning the entire day. The model also accommodates unobserved heterogeneity and recognizes the “rounding” of reported departure times by individuals in surveys to an integral multiple of 5 min. The continuous-time model is estimated using shopping trip data from the 1996 activity survey collected in the Dallas-Fort Worth metropolitan area. The effects of individual and household socio-demographics, employment attributes, and trip-related characteristics on departure time choice are presented and discussed. Application of the continuous-time model to forecast temporal shifts in shopping trip-making due to changes in socio-demographic characteristics and trip-chaining behavior is demonstrated. The use of the formulation as a powerful evaluation tool to manage and influence traffic patterns through dynamic congestion pricing control schemes is also highlighted.
NEYMAN'S C(α) TEST for UNOBSERVED HETEROGENEITY
2016, Econometric Theory
Optimal Investment, Monitoring, and the Staging of Venture Capital
2022, Venture Capital

View all citing articles on Scopus

^☆: This paper is an extensive revision of one first circulated in 1978 under the title ‘The Effect Omitted Regressors on the Analysis of Unemployment Durations’. The author wishes to acknowledge the kind hospitality of the Department of Statistics, Australian National University, while he was revising the paper.

View full text

Generalised residuals and heterogeneous duration models: With applications to the Weilbull model☆

Abstract

Testing for neglected heterogeneity

Econometrica

Some remarks on overdispersion

Biometrika

Theoretical statistics

A general definition of residuals

Journal of the Royal Statistical Society B

Covariance analysis of heart transplant survival data

Journal of the American Statistical Society