n+1 > n. Let "n" equal "the number of dollars the person owns". If "A poor person is more likely to base his self-worth on how many dollars he owns than a rich person is likely to base his self-worth on how many dollars he owns" is true, then it stands that n+1 is a higher increase in self-worth for the 'arbitrary' poor person (compared to n) than is n+1 an increase compared to n for the 'arbitrary' rich person.

The poor person and the rich person have different values of n so their marginal rate is different. That doesn't comment on how much of their self-worth if a function of their total amount of money. This is sort of akin to how freshmen calculus students confuse a function being small with it having a small derivative.