Most studies of U.S. education borrowing constraints are based on an individual male household member and find that they have little effect on educational attainment. We argue that the correct unit of analysis is the attainment of all sibling intrahousehold resource rivals. We use the male...