I am using ATUS data from 2018 to 2024 and found that, among respondents who are currently employed and whose ATUS interview day falls on a non-holiday weekday, about 14% report zero work hours. This proportion is higher than I expected, and I would like to gain a clearer understanding of the reasons behind it. One possibility is that some respondents do not work conventional schedules, for example shift workers who may be off on certain weekdays. I would appreciate any thoughts you may have on this. Thank you.
I looked at the ATUS samples from 2018-2024 using our online data analysis system and found that about 15 percent of ATUS respondents who were employed and whose time diary covered a non-holiday weekday, 15 percent spent zero time on work on their time diary day. Among full-time workers (FULLPART), 9.8 percent spent zero time on work. People who did not work on their time diary day on average work fewer hours per week (UHRSWORKT) than people who did work on their time diary day. Note that to obtain nationally representative estimates using survey data, including ATUS data, you need to apply sampling weights. Person-level analyses should use WT06, with the exception of analyses of the 2020 sample, for which you should use WT20.
We are not aware of any issues with ATUS data quality that make me call into question the accuracy of this estimate. If you come across something that appears to be a systematic issue with the data, please let us know by posting on the forum or emailing us at ipums@umn.edu.