I think if you looked at a graph of U.S. welfare spending vs. the Gini coefficient over the last 40 years you'd see exactly the opposite of the correlation you suggest.