I suspect the disconnect, then, is that the statement you quote appears to be referring to a tax on wealth, not income.

Moving the 2% to that puts it in the right order of magnitude; at that point, the discrepancy is likely to be in the guess of the retiring population size.