kurt_w wrote:I did say that the 50th percentile are "middle class", and I think any definition of "middle class" that excludes the -- you know -- middle of the income distribution is nuts.
Middle class is not defined by the median income. BTW, did you take a stats class on the road to your G.E.D.? (just kidding, professor.) Distributions are not all bell-shaped.
"Middle class" denotes a standard of living. Home ownership. Having some discretionary income to take a vacation. Most importantly, a measure of security in the dynamic job market.
In Brazil in the 1980s, perhaps 10% of the population was middle class. maybe that is where we are heading.
kurt_w wrote: But the fact that it responds more strongly to the 90th percentile doesn't mean that it stops there. We don't have a curve on that graph for the 95th or 97th percentile
No, the fact that the 90 percentile is 143K rather than 10M is critically important. It suggests that the politicians are responding to voters more than dollars. And btw, we have learned that 95 and 97 are not so special, you need to get about 99 to find wealthy people.
Households in the $100K to $1M range [...]
That range covers a full order of magnitude. It makes no more sense than lumping together households with $10,000 incomes and $100,000 incomes.
It ain't the numbers its the behavior. It's entirely possible, and in fact true, that people in 100K->1M band have substantially similar concerns.
A factor of ten at $10K of income is wildly significant.
A factor of ten at $10M of income probably makes zero difference in behavior.