I don't think you're saying anything different, you're just missing how it proves the point. The key is that the number of people in the higher tiers increased

I don't think you're saying anything different, you're just missing how it proves the point. The key is that the number of people in the higher tiers increased, while the membership of the lower tiers went down. That's exactly what you said. But what it shows is people were shifting into those higher tiers. You're right that in all groups below $75K, the percentage of people in the group went down. While from $75K and up, the membership of each group increased (you've read that part of the data wrong, as the $75l-$99K group also increased.). That's precisely what my linked article says, and exactly what we'd like to see: people are shifting from lower groups into higher groups, across the board. EDIT: here's the spreadsheet from the census bureau I was working from: http://www.census.gov/compendia/statab/2012/tables/12s0696.xls http://www.census.gov/compendia/statab/2012/tables/12s0696.x... (Excel spreadsheet)