In the lecture he defines inclusive institutions as having two components: (i) pluralism and (ii) political centralization.
Now, when most people talk about institutional development they like to focus on the first component. This is the crowd that will tell you that what Rwanda needed after 1994 was electoral democracy and free market institutions and all would be well (Did you know that Mali has elections this July?) But history teaches us that that is not the whole story. There is also the problem of Stateness (aka political centralization), and how it comes about.
Finally Acemoglu and Robinson are addressing what I think is a weakness in their inclusive-institutions-as-the-fundamental-cause-of-long-run-growth argument – the question of the need for political centralization through state building (and the messiness it entails). In the lecture, Acemoglu makes the argument that pluralism can actually lead to state building (or makes it easier) through the creation of “consensually strong states”; and that the process of state building can lead to pluralism due to the centralizing leader’s need to cut deals with local elites. In other words, there is reverse causation between pluralism and stateness.
The takeaway is that inclusive institutions result from the joint development of sustained economic growth and a wider distribution of political power.
This is all good, although I wish Acemoglu and Robinson explored the role of coercion in state-building. My reading of history may be completely out of whack, but when I look back in time I see a lot of conquests and forcible inclusion into states and a lot less “deal making.” Conquests not contrasts appear to be the modal MO of state formation.