As I understand it, overall the difference in unavailable population data rose from 1% in 2006 to 3% in 2011, a difference of only 2% of overall population from which data was unavailable. Is that correct?
On October 30th, 2014. See this statement in context.