In recent years, international organizations, think-tanks, and the social sciences have contributed to a dramatic expansion in the range of composite indices measuring concepts such as human development, governance, or social capital. This paper discusses challenges faced in the design of composite indices, and suggests the method known as matching percentiles as a solution to the problem of how to aggregate social institutional data. A set of indices are produced which measure dimensions of social development, understood as informal institutions such as the strength of civil society, intergroup relations, or gender equality, which contribute to better growth and governance. We then conduct a series of diagnostic tests which demonstrate the robustness of the new measures, and assess their degree of construct validity.