Why do people fight for their country? The risks are extreme, the payoff uncertain. In this paper, we argue that reciprocity is a key factor. Examining welfare spending in the US in the 1930s under the New Deal, we show that support for World War II became more common where welfare support had been more generous: war bonds were sold in greater volume, more men and women volunteered, and more soldiers performed heroic actions recognized by a medal. We use weather shocks in the form of droughts to instrument for agricultural emergency relief, and show that results hold. Because both war bond purchases and volunteering respond to welfare support, we argue that results cannot be driven by opportunity cost considerations. Data on World War I patriotic support shows that 1930s emergency spending is only predictive for
World War II support. Pre-New Deal droughts are also not correlated with patriotism after 1941.