We consider the effects of changes in the distribution of a background risk on the optimal risk taking behaviour of a risk- averse decision maker. In particular, we suppose that the background risk deteriorates via a first- or second-degree stochastic dominance shift. Our contention is that such...