CINETHP in 2015 seems to be weird

CPS
1

Hello,

Looking at the case-count view of CINETHP, it seems that 2015 version of the variable is somewhat weird.

Excluding NIU, the percentage of having access to internet is 79% in 2013, but drops to 68 percent in 2015.

This is really weird because the rate remains about 80% in 2015 if I look at the ACS version of the variable.(IPUMS USA: descr: CINETHH)

Do you have any ideas why this is happening? Maybe this is because of the change in the interview questionnaire?

Thank you!
Peter

2

Unfortunately, the extreme variation in CINETHP is known and is due to the many changes in the the universe over time. The universe statements for 2013 and 2015 are quite different, so even though you have accounted for the NIUs, it is not surprising that your figures vary. The drop is likely due to the month-in-sample universe requirement.

The ACS version of this variable (CINETHH) has remained the same over time, moreover, the universe has remained the same. However, it should also be noted that in the ACS this is a household level question (in 2013 and 2015 CINETHP in the CPS was a person level question).