James E. Prieger, "A Flexible Parametric Selection Model for Non-Normal Data with Application to Health Care Usage", Journal of Applied Econometrics, Vol. 17, No. 4, 2002, pp. 367-392. This readme file contains descriptions of the data and programs used in the paper. All data and program files, which are in DOS format, are zipped in the file prieger.zip. Unix users should use "unzip -a". I. Files ========= name size ---------- ------- meps-1.tsp 10 KB meps-2.tsp 11 KB meps-3.tsp 7 KB meps-4.tsp 8 KB meps-5.tsp 11 KB meps-6.tsp 10 KB meps-7.tsp 9 KB meps-hospstay4.dat 121 KB meps-x5.dat 1799 KB meps-x6.dat 150 KB II. Data ========= Each data file is an ASCII free-format space-delimited file; missing observations (if any) marked with a period. Each file contains 14955 observations. All data are from the official Medical Expenditure Panel Survey (MEPS) sources, as made available on the MEPS web site . In the lists below, each paragraph contains the following information on a single variable: Variable name (as in the TSP programs); source MEPS data file; MEPS variable name (as in MEPS official documentation). Description of variable. Data file #1: meps-x5.dat -------------------------- RowName; --; --. Observation number. Intercept; --; --. 1.00 for all observations HospDur; MEPS HC-003; ipngts96+ipzero96/2. IPZERO96: A count of the number of inpatient events where the reported dates of admission and discharge were the same; these are "zero night" stays. IPNGTS96: the total number of nights spent in a hospital in 1996. Zero night stays converted into 1/2 night stays. HospNum; MEPS HC-003; IPTOT96. The total number of inpatient hospital stays that occurred at any time in 1996. privins; MEPS HC-001; PRIV1. 1="yes", 0= "no". Private insurance in general. Private insurance sources identify coverage in effect at any time during Round 1 of the survey. medicare; MEPS HC-001; MCARNW1X. 1="yes", 0= "no". Covered under Medicare. Edited for persons age 65 or over. Within this age group, individuals were assigned Medicare coverage if: 1. They answered yes to a follow-up question on whether or not they received Social Security benefits; or 2. They were covered by Medicaid, other public hospital/physician coverage or Medigap coverage; or 3. Their spouse was covered by Medicare. medicaid; MEPS HC-001; mcaid1x. 1="yes", 0= "no". Covered under Medicaid. Edited to include persons who did not pay anything for their Other public hospital/physician insurance when such coverage was through a Medicaid HMO or reported to include some other managed care characteristics. In addition, a small number of persons reporting AFDC or SSI coverage (questions included in the Round 1 interview for this purpose) were assigned Medicaid coverage. hmo; MEPS HC-002; see description. 1="yes", 0= "no". Yes if enrolled in an HMO through private insurance (PRVHMO1), Medicare (MCRHMO1), or other public insurance (PUBHMO1). condn; MEPS HC-006; CONDN (see description). Number of medical conditions recorded with CONDN per person. priolist; MEPS HC-006; PRIOLISTN (see description). Number of conditions on the "priority list" (PRIOLIST). These conditions were a priori designated as "priority conditions" due to their prevalence, expense, or relevance to policy. Includes cancer, diabetes, emphysema, high cholesterol, HIV/AIDS, hypertension, ischemic heart disease, stroke, arthritis, asthma, gall bladder disease, stomach ulcers, back problems of any kind, Alzheimer's disease and other dementias, and depression and anxiety disorders, and others. exclhlth; MEPS HC-001; RTEHLTH1 (see description). 1="yes", 0= "no". Perceived health status (RTEHLTH1) is "excellent". poorhlth; MEPS HC-001; RTEHLTH1 (see description). 1="yes", 0= "no". Perceived health status (RTEHLTH1) is "poor". adlhelp; MEPS HC-001; IADLHLP1. 1="yes", 0= "no". The Instrumental Activities of Daily Living (IADL) Help or Supervision variable. Is "yes" if person received help or supervision with IADLs such as using the telephone, paying bills, taking medications, preparing light meals, doing laundry, or going shopping. MidWest; MEPS HC-001; REGION (see description). 1="yes", 0= "no". Regional dummy variable. From REGION, Census region. South; MEPS HC-001; REGION (see description). 1="yes", 0= "no". Regional dummy variable. From REGION, Census region. West; MEPS HC-001; REGION (see description). 1="yes", 0= "no". Regional dummy variable. From REGION, Census region. female; MEPS HC-001; SEX. 1="yes", 0= "no". age; MEPS HC-001; AGE1X. Age in years. Black; MEPS HC-001; RACETHNX. 1="yes", 0= "no". From RACETHNX. Hispanic; MEPS HC-001; RACETHNX. 1="yes", 0= "no". From RACETHNX. educ; MEPS HC-001; EDUCYR1. Years of completed education. married; MEPS HC-001; MARRY1X. 1="yes", 0= "no". Marital status. employed; MEPS HC-001; EMPST1. 1="yes", 0= "no". Currently employed. privmcare; MEPS HC-001; --. 1="yes", 0= "no". Private insurance and Medicare. Computed as privins * medicare (see above). Data file #2: meps-x6.dat -------------------------- RowName2; --; --. Observation number. OfferInsCur; MEPS HC-001; OFERCM1X. Health insurance offered from the current main job. OfferInsPrev; MEPS HC-001; OFERNM1X. Health insurance offered through a job other than the current main job. Data file #2: meps-hospstay4.dat --------------------------------- RowName3; --; --. Observation number. HospStay; MEPS HC-003; see description. 1="yes", 0= "no". The total number of inpatient hospital stays that occurred at any time in 1996 (IPTOT96) is > 0. III. Computer Programs ====================== These TSP files were used for the estimations. See the files for additional documentation and comments. Meps-1.tsp Estimations for baseline model (first column of Tables 5 and 6), its nesting gamma-distributed individual durations model, endogeneity estimations and testing for the duration equation. Meps-2.tsp Bootstrapping of the endogeneity test statistics for testing the duration equation. Meps-3.tsp Endogeneity estimations and testing for the insurance selection equation (Table 9) Meps-4.tsp Estimations for FPS model and Lee's model (Tables 5 and 6), and Vuong's Test. Meps-5.tsp Conditional means, slopes, changes, percentage changes, and elasticities as in table 7 Meps-6.tsp Conditional moments tests for the exponential duration assumption (parametric bootstrap). See footnote 32. Meps-7.tsp Conditional moments tests for the exponential duration assumption (non-parametric residual-based bootstrap). See footnote 32.