Missing values in CPS harmonized data

CPS
1

Currently I work on a project that utilizes CPS-data from KCFed. I used Harmonized Variables (1976-present). However, there are a lot of empty entries in the table on several variables when I downloaded the data for different years. In particular, I’m interested in data on Outgoing Rotation Group (months 4 and 8), but for a lot of variables data is missed. Attached is a screenshot of the year 1980 as an example. Does anybody know, what is the reason for missing values and is it possible to access the full dataset?

I’ll be extremely grateful for any help and assistance!

Screen Shot 2020-06-15 at 22.15.51
Screen Shot 2020-06-15 at 22.15.512234×1010 262 KB

Here’s the list of variables I used as reference:

Harmonized Variable (1976-Present):

HRMONTH_abc - month of interview

HRYEAR4_abc - year of interview

PWSSWGT_abc - final weight

PEERNHRO_abc - usual hours

PUWK_abc - last week, did you do any work for (either) pay (or profit)?

PUERNH1C_abc - what is your hourly rate of pay on this job, excluding overtime pay, tips or commission?

V1AGEDG12_ab_n - age

V1RACE_ab_n - race

PRDTHSP_abc - detailed hispanic

PEEDUCA_abc - highest level of school completed or degree received

PRDTIND1_abc_ne - detailed industries

PRDTOCC1_abc_ne - detailed occupations

HRMIS_abc - month-in-sample

HRHHID_abc - household identifier

2

The non-ORG variables that you mentioned are available from IPUMS CPS for the time period you are interested in. The ORG data for 1979-1981 are currently not available via IPUMS, though we intend to make them available in the future. These data are available in the Merged Outgoing Rotation Group files, which can be downloaded from NBER. If you want to merge the variables from the NBER files into an IPUMS extract, you can do so by following the instructions on this blog post. You should be able to use the same linking ID’s to link the NBER MORG file to the Kansas City Fed’s CPS files.

3

Hey, Mattew! Thanks a lot for your previous response! I tried to merge data from IPUMS with the one from NBER according to the post in the link you’ve sent. I looked at MORG data here Index of /morg/annual for year 1979-1981. From these years I need data on the following variables: earnhr (hourly earnings at the main job) and uhours (usual hours of work at the main job). However, when I looked at morg79.dta, morg80.dta and morg81.dta I found that even in NBER data there are a lot of missing values for earnhr and uhours even though each observation supposed to correspond to individuals from Outgoing Rotation Group. Do you have any ideas how this can be fixed? I’ll appreciate your response a lot!

4

I don’t have much information about these samples. My guess is that there are additional universe restrictions for those variables in those years. I would advise looking at the original codebooks (available at the link you posted) and reaching out to NBER for information about data from their site.