Lots of studies have demonstrated a "keeping up with the Joneses" effect, in which the happiness derived from status, income, wealth, or possessions is relative to the circles one inhabits. If I'm making $1M a year in a company full of people making $100k, I'll feel like a baller. If I'm making that same $1M in a company full of people making $10M+, I'll feel like a loser. It's the proximate context, not the universal one, to which people draw their emotional comparisons and assess their financial worth.
Another well-studied effect is that lifestyle tends to rise to meet rising income levels. This is partially driven by the aforementioned effect -- rising income tends to put you in jobs, or in neighborhoods, or generally in physical proximity with other high earners -- and partially driven by the fact that you tend to develop more expensive tastes and habits as you climb the income ladder.
I'm writing this from my iPad on the train, and so I am too lazy to tab over and look up the particular studies or sources. But start by Googling C.N. Parkinson -- the same Parkinson who famously quipped that work expands to meet the time available for its completion -- who stated that expenditures rise to meet income. Plenty of people have taken this observation and run with it in actual studies.
The statistics I can find suggest that as a very broad population-level trend, having high income does correlate with having high net worth (though the causation could be complex). Here's some data from the federal reserve (skip to table 5 on p. 36): http://www.federalreserve.gov/pubs/feds/2009/200913/200913pa...
The relationship seems strongest at the top: the average net worth of someone with a top-1% income is 40x the U.S. mean net worth. People with 95th-99th percentile incomes are at about 7x the mean net worth; those with 90th-95th percentile incomes are at about 2x the mean net worth; those with 50th-90th percentile incomes are at 2/3rds the mean, and those in the bottom 50% of incomes, roughly 1/20th of the mean.
So the people with really high incomes (say, $300k+) seem to, as a general trend, also have substantial wealth, whether because they accumulated it, or already had it. It would be interesting to have more fine-grained data on net-worth distributions within each of those income categories.
I don't have a bookmark unfortunately, but I read first hand a couple of years ago a study mentioning that employees did spend more and more as their wages went up (ultimately leading to golden handcuffs).
One detail, though:
> somehow, many if not most people working in these high-wage jobs don't accumulate all that much money.
This sounds plausible, but I'd prefer to see it backed up by statistics.