This study aims to examine the extent to which competing theories explain saving of low-income households in Individual Development Accounts (IDAs). Competing theories include individual-oriented perspective, social stratification perspective, and institutional saving theory. This study uses American Dream Demonstration (ADD) data collected at the Tulsa IDA program. Compared with the individual perspective and the social stratification perspective, institutional features explain a significant part of the variance in saving outcomes measured by average monthly net deposit (AMND) and deposit frequency ratio (DFR). Findings suggest that an inclusive asset-based policy should be designed with institutional structures encouraging low-income households to save.