What is the effect of asymmetric information about income on household decisions, resource sharing, and welfare? In this paper, I answer this question in four steps. In the first step, I develop a theoretical model that accounts for the possible existence of asymmetric information. The model predicts that households will partly mitigate the welfare cost of asymmetric information by incentivizing the wage earner to provide information about his or her true income. These incentives are provided by making the consumption share increase with reported income: the wage earner’s consumption share is high when reporting a high income and low when reporting a low income. Second, I derive a new identification result for this model. The result states that individual welfare functions can be recovered with at least one assignable good - a good that is observed at the individual level rather than the household level. The result also shows that the effect of asymmetric information on individual welfare can be identified separately from other variables impacting welfare. Third, I estimate the model using a survey of Bangladeshi day laborers. The estimation confirms the predictions of the model, providing evidence that the households in the data are affected by asymmetric information. Finally, I conduct three counterfactual analyses to document how asymmetric information interacts with policies. I compute the willingness-to-pay in each case.