Norton Long's 1949 essay, “Power and Administration,” has a complicated legacy. First, analysis reveals both support for and important refinements of Long's arguments since the article's publication. Second, Long's claim has proven problematic that competition among agencies for power would bring more coordination and a cross‐agency sense of purpose to the federal government. Third, the bureaucratic pluralism that he explained and defended produced special interest biases that were off‐putting to large segments of citizens and thus helped create an unsupportive political environment for needed capacity building in the federal government. Fourth, by not considering how institutions “coevolve,” Long failed to warn that “horizontal power” building by individual agencies would provoke efforts by elected officials to enhance their control over bureaucracy in ways that, over time, diminished their collective sources of power. Finally, much remains to be done before what Long called a “realistic science of administration” incorporating the “budgeting of power” exists in public administration.