Mean (UHRSWORK1) over sample of employed full-time

Hi ! I am trying to figure out mean of UHRSWORK1 for full-time salaried employees. I obtain the average is about 80 usual weekly work hours which is obviously a double of regular work hours in the US. What is the problem here?

I use the variable along with Outgoin Rotation Groups data, and the still problem persists when I restric observation to Month-ib-sample 4 (or 8) to exclude summing up of usual weekly hours by CPS respondent,

I resolved the problem, the excessive mean hours was just driven by outliers.