What? That would be extremely unusual, if true. (The long tail on the upper end is unbounded, while on the lower end income is bounded at 0. So for income, usually mean >>> median.)
Did they get mean and median reversed?
= = = = =
Edit I drew a random sample of incomes for 1,000 people (between $0 and $1 million) and used Pytorch to update these incomes to achieve mean $196,000 and median $250,000. (I also heavily penalized income < $0.)
The shape of the distribution of incomes to achieve such a feat, as one would predict, is absurd. X axis is income in $, Y axis is the number of people with that income out of 1,000. (First image.)
If instead the economist got his understanding of mean and median reversed, you get the second image, which is zero-inflated but otherwise uniform in income.