Firstly, I know that the associate minister has a thing or two to say about the definition of middle class, and I'm sure you'll have that discussion on Wednesday.
From purely a bean-counter and statistical perspective, though, I can say that there is no common definition of “middle class”. Any definition—attempted definition—would have to take in a range of considerations: family size; family composition, i.e., how many children you have; geography, where you live. Just with those three variables, there is a range of definitions around what is the middle class, which can range from some peripheral around the median income.... By extension, when defining “income”, is that before-tax income or after-tax income? You talk about a peripheral within the 50% or 150% of the median income.
Then, in adjusting for family size, there's this concept of adjusting for a per-adult equivalent. On a per-adult basis, is it $20,000 to $85,000? Is that after tax or before tax? That then changes those metrics.
I'm not trying to escape the question, by any stretch, but trying to emphasize that there are general variables in play when trying to attach a statistical definition to “middle class”.