Neoclassical microeconomic theory of the consumer postulates that a rational consumer maximizes utility subject to a budget constraint. The direct implication of this theory is that the source of the income does not affect consumption behavior. An emerging body of literature which builds on cognitive psychology however documents that individuals do not treat all income sources in the same way and may have different spending behaviors depending on the source of the income. Unlike previous studies which used observational data, we propose to use a randomized controlled experiment to investigate if the pattern of household consumption and saving differ depending on the source of income. We will randomly select treatment households and let some of the households earn income by doing a public work and let some others earn free money based on a lottery. The randomization will enable us to identify the impact of earned and unearned income on consumption and saving behavior of households robustly. This project will broadly contribute to a body of research in microeconomics investigating the relationship between income and consumption. It will also speak to the literature on precautionary savings, and the literature on aid in developing countries, with significant implications.