HEALTH INSURANCE REVISION EXTRACT FILES 2000 through 2010 CPS March Supplement Notes In September 2011, the Census Bureau released revised figures on health insurance coverage from the 2000 to 2010 Current Population Survey Annual Social and Economic Supplements (CPS ASEC), reflecting enhancements to the editing process, including the assignment of a family health plan to all individuals in the household and the addition of a new variable to the allocation matrix (for more on the revisions see: http://www.census.gov/hhes/www/hlthins/data/revhlth/usernote.html). In September 2011, the Census Bureau released a new historical data series that incorporates the enhancements to the editing process. A working paper (found at: http://www.census.gov/hhes/www/hlthins/data/revhlth/SHADAC.pdf) released at the same time describes the revisions made and applies it to all years available: data files from the 2000 to 2010 ASECs. This guide is aimed at helping analysts who wish to use the Public-Use files from the CPS ASEC to produce the same revised estimates of health insurance coverage for years affected by the revisions. The following files are available as SAS datasets (.sas7bdat): 1. hlthins_sy2000_rev 2. hlthins_sy2001_rev 3. hlthins_sy2002_rev 4. hlthins_sy2003_rev 5. hlthins_sy2004_rev 6. hlthins_sy2005_rev 7. hlthins_sy2006_rev 8. hlthins_sy2007_rev 9. hlthins_sy2008_rev 10. hlthins_sy2009_rev 11. hlthins_sy2010_rev These datasets in conjunction with instructions contained in this README will allow for analysts to produce revised estimates from 2000 through 2010. It is important to note that the variables are not changing, but only the values on certain records are changing. Also, as the variables have changed throughout the years, all variables will appear on the datasets, but variables that are not available for a particular year will have missing values. Each extract file listed above is designated by the survey year identifying the corresponding CPS ASEC file to which the data should be merged (i.e. hlthins_sy2004_rev merges onto the 2004 data file for estimates of 2003). Each dataset contains health insurance variables used for the corresponding year. Below is a list of the variables that are found on the current extracts. COV_HI COV_GH HI CAID CARE MCAID MCARE CH_MC CH_HI CHAMP DEPRIV PRIV AHITYP1-6 OTHSTYP1-6 HI_YN HIOWN HIEMP HIPAID HITYP DEPHI HILIN1 HILIN2 PAID HIOUT PRITYP PILIN1 PILIN2 POUT OUT MON OTH OTYP1-5 OTHSTPER HEA AHIPER PCHIP HELSE1-6 HIELSE1-5 PRIELS1-16 I_HI I_DEPHI I_PAID I_HIOUT I_PRIV I_DEPRIV I_POUT I_OUT I_CARE I_CAID I_MON I_OTH I_OTYP I_OSTPER I_OSTYP I_HEA IAHIPER IAHITYP I_PCHIP I_CHAMP I_HELSE I_HIELSE I_HIEMP I_HIOWN I_HIPAID I_HIYN I_MCAID I_MCARE I_PRIELS I_OTHSTYP I_OTHSTPER Note: For the variable, CH_HI, when a respondent has coverage through someone outside of the household, an assumption is made that the source of coverage is employer-based coverage. MERGING THE DATA The extract files are person records, thus they are merged to the CPS data by the household sequence number PH_SEQ and the unique household ID PPPOS. Essentially, the SAS code should look like: PROC SORT DATA=cps02asec; BY PH_SEQ PPPOS; RUN; PROC SORT DATA=hlthins_sy2002_rev; BY PH_SEQ PPPOS; RUN; DATA rev_cpsfile; MERGE cps02asec (in=a) drop= (same variables as found on the extract you are using); hlthins_sy2002_rev (in=b); By PH_SEQ PPPOS; IF a and b; RUN; HOUSEHOLD-LEVEL VARIABLES There are four household variables, HMCARE, HMCAID, HCHAMP, and HHI_YN. If any person item indicated a “yes” then the household item is also a “yes.” Below is the code for these four variables. If anyone in the household has MCARE=1, then HMCARE=1; data hh_hic (rename=(PH_SEQ=H_SEQ); set pp_extract; by PH_SEQ; if first.PH_SEQ then do; HMCARE='2'; HMCAID='2'; HCHAMP='2'; HHI_YN='2'; end; IF MCARE=1 then HMCARE=1; IF MCAID=1 then HMCAID=1; IF CHAMP=1 then HCHAMP=1; IF HI_YN=1 then HHI_YN=1; if last.PH_SEQ; run; Note: The variable HHI_YN will NOT initialize to a “yes” unless individuals are over the age of 15 and COV_HI=1.