Selectorate theory provides an elegant and encompassing theoretical framework that makes predictions for several important puzzles in research on democracies and autocracies. Yet several critics lament shortcomings in the measurement of its key concepts: the size of the selectorate S and the winning coalition W. We suggest an alternative that exploits information on the power status and population shares of ethnic groups around the world. Specifically, we identify the size of the selectorate as the sum of groups' population shares that do not suffer from political discrimination, and the size of the winning coalition as the cumulative population share of those ethnic groups represented in a state's executive. Our proposal improves on existing work by providing a continuous operationalization of W and S, and thus to seamlessly bridge democratic and autocratic regimes, by not yielding observations in which the size of W exceeds the size of S, and by ensuring that S is strictly positive. We illustrate the usefulness by retesting and extending the claim that regimes with smaller winning coalitions receive higher levels of aid.