*A swift rise in the demand for trade protection in the US and other advanced economies has paralleled a decline in world trade costs. Empirical research in the US has shown that labor markets that are more exposed to higher imports, have higher unemployment, lower wages, and vote in a protectionist direction on trade bills. We first empirically show if better market access to labor markets, moderates the positive effect of trade shock on protectionism in the US. Market access describes firms’ ability to sell goods to other regions and consumers’ ability to buy goods from other regions and depends upon regional geography. To quantify the mediating impact of market access, we derive a theoretical measure of market access in a multi-country, -region, and -sector, Ricardian trade model using the methodology developed by Donaldson and Hornbeck (2016). In the counterfactual exercise, we modify endogenous geography (i.e. domestic infrastructure) to see if the negative regional effect of trade shocks can be reduced, without reducing aggregate welfare in the US.

**This paper studies the consequences of charitable giving for both the optimal tax system and the optimal provision of public goods. Through warm glow, taxpayers derive utility from their personal donations. Aggregate contributions then benefit to all individuals through the public good effect of charitable giving. The government has two sets of instruments to maximize social welfare: nonlinear taxes or subsidies of both income and donations as well as direct contributions to the public good. Solving this joint problem, I provide optimality conditions for both tax rates and the government's contribution to the public good.
First I show that because of charitable giving, the welfare impact of behavioral responses to tax policy is magnified by an externality, due to the public good nature of donations, and a spillover effect, capturing the loop between changes in individual contributions and changes in aggregate contributions. Second, I prove that the externality channel can no longer affect optimal tax rates when the government's contribution to the public good is set optimally. Finally, I consider a realistic tax mix that combines a nonlinear income tax with either tax credits or tax deductions for giving, as it is the case in most OECD countries. Optimality conditions, expressed in terms of empirically meaningful parameters, are then delivered for these three tax instruments.