Research & Analysis by Javier Meseguer
Correlation Patterns between Primary and Secondary Diagnosis Codes in the Social Security Disability Programs
This paper addresses impairment comorbidity among participants in programs that provide disability benefits. Comorbidity in this context is defined as the simultaneous presence of a primary and a secondary medical diagnosis. The author fits a high-dimensional Bayesian multivariate probit model with a 10% random sample of 2009 initial claimants (disabled workers, including individuals concurrently applying for Disability Insurance and Supplemental Security Income). The resulting correlation estimates provide evidence of strong impairment comorbidity patterns at the initial-claim level. Many of the findings mirror the epidemiological evidence, such as associations of diabetes with chronic renal failure, open wounds of a lower limb, peripheral neuropathies, and blindness/low vision. Other results are surprising. For instance, the correlation estimates defy the presumption of high positive association between mental and musculoskeletal-system diagnoses.
Outcome Variation in the Social Security Disability Insurance Program: The Role of Primary Diagnoses
This article investigates the role that primary impairments play in explaining heterogeneity in disability decisions. Using claimant-level data within a hierarchical framework, the author explores variation in outcomes along three dimensions: state of origin, adjudicative stage, and primary diagnosis. The findings indicate that the impairments account for a substantial portion of claimant-level variation in initial allowances. Furthermore, the author finds that the predictions of an initial and a final allowance are highly correlated when applicants are grouped by impairment. In other words, diagnoses that are more likely to result in an initial allowance also tend to be more likely to receive a final allowance.
This paper evaluates the out-of-sample performance of two stochastic models used to forecast age-specific mortality rates: (1) the model proposed by Lee and Carter (1992); and (2) a set of univariate autoregressions linked together by a common residual covariance matrix (Denton, Feavor, and Spencer 2005).