Abstract: My dissertation investigates two models of macroeconomics with heterogeneous agents. The first chapter analyzes a setup where agents are ex ante identical, yet receive idiosyncratic income shocks which make them heterogeneous ex post. A private information friction gives rise to incomplete risk-sharing as a constrained-efficient allocation. The second chapter again considers ex post heterogeneous agents: they have identical preferences but face idiosyncratic shocks to their earning capacity. There the focus is not on risk-sharing, but on the aggregate consequences for labor supply. Specifically, the first chapter examines efficient intertemporal consumption insurance with private information about income shocks. The novel feature is a monitoring technology available to the social planner: monitoring, though costly, can be used to mitigate the information friction. The chapter characterizes the optimal mix of monitoring and distortion of promised values used to provide incentives and overcome the private information friction. Efficient stationary consumption distributions are found to be nondegenerate in some cases. The second chapter is drawn from joint work with François Gourio. It attempts to reconcile the large volatility of aggregate employment with small micro estimates of the elasticity of labor supply. Our model of indivisible labor with complete markets shows that the aggregate Frisch elasticity of labor supply depends on the shape of the distribution of reservation wages. Even if most workers are wage-inelastic, the aggregate elasticity can be large if sufficiently many agents are close to their reservation wage. To evaluate this hypothesis, we estimate the model using monthly panel data drawn from the NLSY. This allows us to measure the aggregate elasticity implied by realistic heterogeneity. We estimate a Frisch elasticity around 1.5. Interestingly, this elasticity is countercyclical. Our model also has a natural cross-sectional implication: workers who are nearly indifferent between working or not are more sensitive to aggregate fluctuations. We find empirical support for this prediction: on average, the group of marginal workers, which makes up 22% of the population, accounts for 49% of aggregate fluctuations in employment.