weights for partial sample in March CPS 2014

CPS
1

Comparing the 2010-2013 full samples and the 2014 5/8 sample (see below), I noticed that the unweighted person counts make sense (2014 is significantly smaller than the rest), but the weighted person counts for 2014 are basically the same as for previous years. Has WTSUPP been adjusted somehow for the 2014 partial sample?

below.

YEAR UNWEIGHTED N WEIGHTED N

2010 209802 304281698
2011 204983 306553211
2012 201398 308827259
2013 202634 311116170
2014 139415 313395422

2

Like previous March samples, the weighted 5/8 March sample in 2014 is nationally representative. Thus, as you suggest, applying the WTSUPP values to this sample will result in accurate national population estimates. Keep in mind that the decreased size of the sample will likely result in slightly larger margins of error.

Hope this helps.