oh. it may have to do with the size of the universe - S2702 is for the Total Civilian Non-institutionalized population. So the sample may be too small.
EDIT - I only looked at the first variable - I see there are more in that table than just the Total Civilian Non-institutionalized. But it does look like smaller population states are the ones missing. Perhaps someone with more expertise on that will reply. :)
you can try checking the methodology notes:
www.census.gov/.../