Mistigris yes - I made my post having read the entire report (that's why I mentioned I was paraphrasing): they explain about the median, but my point was that they say this gives a false data set, over-estimating non-working poverty, and under-estimating working poverty. They then say they correct for this.
However, it doesn't say how these corrections were obtained, or how they were made - this is why it's misleading. If such and such a study has identified that working poverty is under-identified by 5.2%, say, by the median method, then this should have been cited, and an explanation given that, as a result, the figures for working poverty had been increased by 5.2%. Without this information, the report is statistically meaningless, and the methodology is flawed as a result - that's why I said it would not be publishable in an academic journal, and why a student would fail if they submitted this. Data needs to be checkable/verifable, and the methodological process by which it is obtained needs to be transparent, correct/appropriate and also identifiable.