Earlier quoted context omitted.
I've seen some research on "relative income happiness" that suggest having poor(er) people nearby should increase happiness. Perhaps it doesn't apply if the income cap is to large...
Or perhaps people aren't actually rational utility maximizing machines and might not properly anticipate the utility benefits from the change.
I'm guessing sorisos is spot on with their last sentence. There's a big difference between being a millionaire surrounded by middle classers and being a lower classer surrounded by extreme poverty. Often times with happiness research there are caveats and ranges to keep in mind but the nuances tend to get lost in the headlines.