In a 1954 paper, A. C. Harberger claimed that the welfare loss from monopoly in United States manufacturing was less than one tenth of one percent of national income over 1924-1928. This led to additional claims of low monopoly welfare loss and eventually to a counter-argument in favor of adding a rent-seeking cost equal to part or all of the economic profit to this loss. These arguments assumed a passive role for government. In this paper, by contrast, governments are active and seek to maximize their political support. The political support maximum then depends on the nature of the political system and, in particular, on how inclusive this system is. The same is true of the monopoly welfare loss, which becomes largely the social cost of rent seeking plus the social cost of protectionism—or of protecting existing profits by restricting investment that would increase competition in markets where these profits are earned. Protectionist measures lower innovation and the growth of total factor productivity, but can still be a good source of political support in a political system where inclusiveness is moderate to low. This can explain not only the existence of inefficiency and of large monopoly welfare loss, but also their persistence and the persistence of large differences in total factor productivity between nations.