CPS Small Sample Size April and June 2001: using stata [pweight=wtfinl] - sudden spike in educational attainment

CPS
1

I think this has to do with July 2001 sample expansion. When I take the average of wtfinl (stata: sum wtfinl), the average weights are roughly:

March ~ 2300

April ~ 1200

May ~ 2300

June ~ 1200

July ~ 1900

This makes sense that July would be smaller than May or March: the sample population grows by about 20%, so it makes sense that the weights would shrink a bit. But why are the weights on average so small in April and June? When I get aggregate statistics (for example on education), the population is much small in April and June. This also causes spikes in educational attainment for just those two months, and then it goes exactly back to trend in July. What is going on here?

2

It was recently discovered that the household/person identifiers found on the 2000 population adjustments supplemental data file do not completely match the CPS household/person identifiers for April and June of 2001. We have contacted the Census Bureau for further insight into the non-matching identifiers. IPUMS-CPS is currently working on restoring the pre-2000 population adjustment WTFINL values, aiming to release these files by the end of the summer.

I apologize for any inconvenience this may have caused.