Good evening,
I am working with CPS ASEC data and am interested in years 1978-1982 and 2015-2019. There exists a variable (asecoverh), that identifies persons that are part of the ASEC oversample. The way I understand it, is that there are two oversamples in the CPS ASEC: one is the hispanic oversample and the other one is the SCHIP oversample (is that correct?). Is there a variable that distinguishes which oversample an observation is part of? If not, is there another way to find out (possibly via another variable?)
Thank you very much!
You are correct that the ASEC oversample consists of the Hispanic oversample and SCHIP oversample. There is no variable that explicitly identifies the oversample group of these additional observations. The SCHIP oversample is first introduced in 2001, so there won’t be any cases in 1978-1982. In 2015-2019, households that do not include anyone who is Hispanic and do include someone under the age of 18 are likely the SCHIP oversample households.
Note, the SCHIP oversample does not result in additional variables , only additional observations for the variables already included as part of the ASEC. The SCHIP oversample was added to produce reliable state-level estimates on low-income children and measure the effects of SCHIP; while some research questions may justify handling these cases separately (e.g. linking CPS to leverage the panel component of the data), they are intended to be used alongside the standard sample.
You can find more information on the oversamples as they relate to linking CPS observations in section 2c of the working paper “Using the Annual Social and Economic Supplement with Current Population Survey Panels” from our colleagues Sarah Flood and Jose Pacas.
1 Like
Kari, I have a question about the 2001 sample expansion. Does CPS provide information about the actual survey month of the households that are not interviewed in March (e.g. preceding November, February, April)? I am asking this because some questions are asked as of the interview data and some as of the previous year. So I was wondering how the sample expansion affect the timing of these questions?
The CPS does not provide information about the month of the replacement ASEC interview for persons in the oversamples; however, you can identify persons who are part of the oversample with the variable ASECOVERP. To my knowledge the timing of the question remains unchanged. As you note, many ASEC questions refer to the past year; however, this typically refers to the last calendar year rather than the last 12 months. In these cases, there is no difference in the reference period regardless of when the person was interviewed. You can view the month when the ASEC was administered to each of the oversamples in this working paper (Table 2).