The 1968-1988 Compressed Mortality Data from the National Vital Statistics System of the National Center for Health Statistics consists of a mortality data file and a population data file. Each record in the mortality file has eight fields. Seven describe possible categories (age, race, gender, year, state, county of residence, and underlying cause of death) within which a decedent can fall and the remaining field is a count of the number of decedents matching on all seven categories. There are records for every death to a U.S. citizen for all years except 1972. The 1972 data is based on a 50% sample and is weighted by a factor of 2. The population file contains Census national, state and county population estimates. The age, race, and sex detail of the population file matches the mortality file.

Counts and rates of death by age, race, gender, year, state, and county of residence, and underlying cause of death (4-digit ICD code or group of codes) for more recent years can be produced by CDC Wonder. (Choose "Get Information" --> "Mortality")

Both ".Z" and ".zip" files can be uncompressed with winzip. In addition, ".Z" files can be uncompressed using the UNIX uncompress command and ".zip" files can be unzipped with pkunzip. The compressed data files are 4-6 Mb, and the compression ratio is over 90%. To check your ability to uncompress these files, download the small files compress.Z or compress.zip. These files give an example of how to read in .Z and .zip ASCII files into SAS for UNIX without decompressing the files. To download files in Internet Explorer, right click on them and select "Save Target As...". If the pdf documents appear to be all blank pages, get the latest Acrobat Reader at www.abobe.com.

NBER internal users can obtain the data from a UNIX shell at /home/data/cmf_new or on an NBER PC via Network Neighborhood --> NBER --> home --> data --> cmf_new

Updates and changes.

Data