A growing literature on substitution between social programs provides consistent evidence that changes in the generosity of one program can lead to changes in enrollment on other programs. However, this evidence has been ignored in welfare analyses of social insurance programs. I demonstrate...