This study tests the hypothesis that individuals’ cooperative behavior does not just depend on their available wealth but also on the shape of the income process which generates that wealth. Inequality is no longer observed in levels only. It also affects real incomes with those of the poorest not only stagnating but declining. I ran a pre-registered online experiment to look at the impact of income changes on cooperation. The experiment varies whether participants received increasing, decreasing or constant rewards for a task they had to complete repeatedly, while keeping their final budgets equal. Whereas I find no effects on charitable giving, I find evidence that contributions in a public goods game are affected by players comparing their reward profiles.