Keywords: Discrete dynamic models, panel survey data, hidden Markov models, estimation, forecasting, partial moments of an Edgeworth expansion, prevalence models, health incidence models, wealth change models, invariance tests, causality tests, income change models
References: Peter Adams, Michael Hurd, Daniel McFadden, Angela Merrill, and Tiago Ribeiro, Healthy, Wealthy, and Wise? Tests for Direct Causal Paths between Health and Socioeconomic Status, Journal of Econometrics (2003), 112, 3-56; Tiago Ribeiro, Discrete Dynamic Models: A HMM Approach to Estimation and Forecasting Using Panel Survey Data (PDF), Working Paper, 2002.
Description: This paper provides statistical methods that permit the association of socioeconomic status and health to be partially unraveled in panel data by excluding some postulated causal paths, or delimiting their range of action. These methods are applied to the Asset and Health Dynamics of the Oldest Old (AHEAD) Panel to test for the absence of causal links from socioeconomic status (SES) to health innovations and mortality, and from health conditions to innovations in wealth. We conclude that in this elderly American population, where Medicare covers most acute care and pension income is not affected by ability to work, the evidence supports the hypothesis of no direct causal link from SES to mortality and to incidence of most sudden onset health conditions (accidents and some acute conditions), once initial health conditions are controlled, but there is some association of SES with incidence of gradual onset health conditions (mental conditions, and some degenerative and chronic conditions), due either to causal links or to persistent unobserved behavioral or genetic factors that have a common influence on both SES and innovations in health. There is mixed evidence for an association of health conditions and innovations in wealth. The death of a spouse appears to have a negative effect on the wealth of the survivor; this is plausibly a direct causal effect. There is evidence for some association of health conditions with increased dissaving from liquid wealth for intact couples and singles. From these findings, we conclude that there is no evidence that SES-linked therapies for acute diseases induce mortality differentials. The question of whether SES-linked preventative care influences onset of chronic and mental diseases remains open. The Appendix to this paper contains the detailed model estimates.
Platforms: Multiple. Datasets were generated in SAS. Analysis was conducted using TSP, SST, and Matlab.
Support: Code was developed by Tiago Ribeiro and Peter Adams at the University of California, Berkeley. If you discover problems or have questions, please e-mail Tiago Ribeiro at email@example.com.
By clicking on the links below, you agree that you have read our disclaimer, understand it, and will abide by its terms and conditions.
Archive files: Datasets, data generation code, and analysis files are archived with the preprint version of the journal article. Please note: all source code development, testing, and analysis occurred on Sun Microsystems Ultra 5 workstations running Solaris 7. Researchers used SAS, TSP, and SST (Solaris). Ascii files are archived using standard Unix line-feed characters to end a line. End-of-line markers do not incorporate the standard DOS/PC carriage return (^M).
Archived by Grace Katagiri, 28 February 2003.
Last modified: 02/28/2003