Matt Yglesias enters the fray:
The basic problem here is that contrary to the impression one gets from, say, Fareed Zakaria’s book, liberal autocracy, while certainly a conceptual possibility, doesn’t seem to be much of an empirical possibility. If you’re compiling a list of modern liberal autocracies, you’re going to start with Singapore and you’re going to end with . . . Singapore. Singapore’s nice and I wouldn’t be too eager to press for change there, but it’s also an unusual situation and it’s far from clear that there’s a generalizable model here. If you want to find examples of liberal autocracy as a stable governing model, you really need to look back to nineteenth-century Europe (or to some extent eighteenth-century Britain). Even there, you’ll find that the main theme was less autocracy per se than some combination of a restricted franchise and/or power sharing between elected parliaments and hereditary monarchs. The nineteenth-century United States in some ways fits this model as well, though nobody talks about it that way.
Food for thought.