We set out to replicate this analysis with the national household survey data. The results that we got were inconsistent with other research and were very odd. We went back and forth with Statistics Canada, as we sometimes do on these occasions, to see if we were misunderstanding something and if there was something that we were missing. Then we wound up writing a paper that basically said what these problems were, how it was inconsistent with other analysis, and that we couldn't do that analysis for the 2011 data in that way.
On June 8th, 2017. See this statement in context.