Thanks for posting, Bostonian.

It looks like about 1% of each cohort is double-skipped (or maybe single-skipped with a very late birthday) -- I would not have guessed it was that high. It looks like the frequency has grown over time.

The 12,000 15-yr-old graduate students who come from nowhere are also interesting. That can't all be people who don't understand the form, can it?

Data that was from the schools themselves would be better. The NCES seems a likely source, but I could not find it quickly.

Here is the same CPS data for Oct 2000:

http://www.census.gov/hhes/school/data/cps/2000/tab02.txt

Here it is for 1981:

http://www.census.gov/hhes/school/data/cps/1981/tab15.pdf

Double skipping seems to have become a lot more common.


Other years can be found here:

http://www.census.gov/hhes/school/data/cps/index.html

and here:

http://www.census.gov/hhes/school/data/cps/previous/index.html