Measuring the district effect. Reflections on the literature