Heckman, JJ., ‘The Incidental Parameters Problem and the Problem of Initial conditions in Estimating a Discrete Time-Discrete Data Stochastic Process,’ in Structural Analysis of Discrete Data with Econometric Applications, ed C.F. Manski and D. McFadden (Cambridge, MA: MIT Press, 1981).

has been cited by the following article:

• TITLE: Gauging Risk Stability: A Simple Test Using Patterns of Workers’ Compensation Claims

  AUTHORS: Richard J. Butler, B. Delworth Gardner, Harold H. Gardner

  KEYWORDS: Risk Measure; Heterogeneity; Runs Test; Workers’ Compensation

  JOURNAL NAME: Journal of Financial Risk Management, Vol.1 No.3, September 28, 2012

  ABSTRACT: In financial risk management, whether for national and international entities, or within a firm, it is important to be able to decide what risks are altered by past experience (or intervention of a central banker at the national level, or a firm’s risk manager at the employer level) and what risks do not vary over time (risks from intrinsic heterogeneity). Policies aimed at changing intrinsic risk will obviously not be cost effective, though tools for identifying such unchanging risks may help to minimize losses associated with those risks. This paper outlines a simple test that allows the researcher to distinguish these alternative risk types. We present our test in the context of patterns of workers’ compensation lost day claims to test whether past experience explains repeated claims for some individuals, or whether some individuals exhibit an innate heterogeneity in claims-filing propensities. We find that a previous claim significantly increases claim probabilities in the future using easy to estimate and interpret “runs” tests on the claims. This suggests, for the risk management problem examined here, that early intervention to limit the effects of the first lost time claim may produce significant disability-cost savings with respect to future claims.