Hi everyone, new to IPUMS. I wanted individual-level microdata at a yearly level that could be identifiable by city/MSA (i.e. CBSA FIPS). However, it looks like a lot of the earnweek and hourwage values are NA. I read here that this is because these values are restricted to respondents in their 4th or 8th month in the sample. However, even when restricting this, earnweek and hourwage still have many NA values. See here.
Using NBER MORG data from 2018-2023, I see around ~45% of earnweek values are NA, and more than 70% of hourwage values are NA. This seems like a substantial amount and I’m not sure why this exists. If someone could help me figure out why this is happening and/or point me in the right direction (w/r/t other data sets I can use) that would be great!
It sounds like you are trying to get annual estimates of weekly and/or hourly wages to aggregate at certain geographic levels, such as MSA or city. I will answer your specific questions about the CPS and then share ideas of how else you might capture annual income at the metro level. Note that the NBER files and the IPUMS files are derived from the same original source, but are different.
The variables EARNWEEK and HOURWAGE (and their replacements EARNWEEK2 and HOURWAGE2) are part of the CPS Basic Monthly Sample’s Outgoing Rotation Group/Earner Study (ORG) which are collected in the household’s fourth and eighth month in the survey (MISH). Any CPS household is in the survey up to 8 times over a 16 month period, and as such, the vast majority of respondents will report earnings at least once in a calendar year, but not in every month. Researchers tend to use the outgoing rotation groups variables because the pooled annual sample contains almost three times the number of observations as the ASEC. Note that the ORG variables only include only civilians age 15 and older who are currently employed as a wage or salaried worker (not self employed). The ASEC collects a wealth of data on respondent income as well. For details on geographic variables regarding your interest in MSAs, here is documentation on the metropolitan area codes (METFIPS) available in the CPS. Please note, the Census Bureau generally advises against sub-state analyses except for large metro areas. You can read more about the CPS sample design in Technical Paper 77 (specifically see page 52). Not all metro areas are available in the CPS.
Another option for you would be to use IPUMS USA or IPUMS NHGIS. IPUMS USA provides access to microdata from the American Community Survey (ACS) which has a large sample, includes variables on income, and has lower levels of geographic units. IPUMS NHGIS provides easy access to summary tables and time series of population, housing, agriculture, and economic data, along with GIS-compatible mapping files, for years from 1790 through the present and for all levels of U.S. census geography, including states, counties, tracts, and blocks. Household income is available annually using data collected from the ACS (also available through IPUMS USA).