This is a book with some very important, even profound, ideas about politics, institutions, the virtues of democracy and what it takes to realize them, but it is written so so very, very diffusely that it will will have next to no impact, which is a shame. Let me try to lay out the main path of argument, which is rather lost amid the authors' digressions and verbiage.
We live in big, complex societies, which means we are thoroughly interdependent on each other, and that we will naturally have different ideas about how our life in common should go, and will have divergent interests. This means that politics we shall always have with us. It also means that political problems are largely ones about designing and reforming the institutions which shape how we interact with each other. But because political problems are so hard, even if we could agree on what we wanted our institutions to achieve (which we don't), we can basically never know in advance what the best institution for a given problem is. (That markets should always and everywhere be the default institution is a claim Knight and Johnson carefully examine before rejected, whereas I would simply mock it.) We also can basically never be sure when changed conditions will make existing institutions unsatisfactory. Put this together and what we need is (to borrow a phrase) "bold, persistent experimentation", with meta-institutions for monitoring how the experiments are going, and deciding when they should be changed or stopped.
This is where democracy comes in. It has priority, not as a first-order institution for getting everything done, but as a second-order institution for checking on and revising other institutions. No other organizational form is as well-suited to checking whether an institution is actively working; some (e.g., markets and courts) are positively pessimized for monitoring their own performance. Democracy has, importantly, two crucial parts: voting, or some similar means of aggregating choices, and debate, the arguments which come before and continue after every vote. Democratic voting is a way of making choices. Democratic debate is a tool for cognition, for harnessing the dispersed knowledge of the citizens and their diversity of perspectives and insights. (There are appropriate citations to Ober and Page, among others, at this point.) The two together are a centralized competitive mechanism for discovering what to do, and for revisiting those decisions in the light of experience. Democracy does not produce agreement, and doesn't need to. Instead it gives us, for the most part, a shared understanding of what we disagree about.
This sort of competition is going to work better as we expand the pool of people who can contribute to it. This means not just having a vote, but being able to cast it as their reason, conscience and interests tell them, not according to the threats of their boss, preacher or family. But because debate is crucial to democracy, people also need to be able to participate in that debate effectively: they need to be able to read (or hear or see...) the debate, and write their own contributions. Others must not ignore or, worse, silence them just because of the kind of people they are (men, women, blacks, Jews, Muslims, Baptist, janitors, engineers...), though they might ignore them if they don't make sense. So Knight and Johnson recover the usual civil liberties — freedom of speech and of the press, universal suffrage through secret ballots, etc. — as conditions for making democracy work.
But remember that democracy is going to work better the more people can and do really ("effectively") contribute, especially to the debate. To use an analogy Knight and Johnson don't: in a sense I could run a marathon, since it is not legally or even physically impossible for me to go through those motions. But there are a lot of obstacles in my way: I have a time-consuming, sedentary job, I have 38-year-long history of not being very athletic and all the habits and physical drawbacks which go along with that. If I really, really wanted to run a marathon, I am young enough that years of dedication would probably get me there, but it would be very hard, and would get easier if I had a very different environment; it would be easier still if I could have had that different environment for a long time.
To participate in the democratic debate, people need a lot of skills and cognitive tools: literacy; numeracy; knowing what other people are going on about and why it matters to them; the cultural knowledge and rhetorical skill to argue effectively with fellow citizens*; knowledge of the world in general. Gaining all these skills and tools takes teachers and time. (Some people learn such things under extraordinarily bad conditions; expecting everyone to do so is like expecting every middle-aged office worker to become a marathon-runner.) Gaining these skills also takes a brain which is not too damaged by malnutrition, lead, stress, etc., or simply too inflexible with age. Even citizens who have these skills need free time, not taken up by getting a living, if they are to use them. Economic barriers matter too: if the cost of making oneself heard is owning a TV station, effectively we've limited debate to the friends and servants of TV-station-owners. But since democracy works better the more minds it can draw on, and the more diverse they are, that is not apt to be a good situation even for station-owners.
Making sure that everyone has an equal opportunity to participate in democracy would be very demanding, and we are very far from doing so. We are even far from making sure everyone has some non-farcical minimum of opportunity. We can and should move towards spreading those opportunities, and make democracy more of a reality and less of a mere promise.
This, then, is the main path of thought in The Priority of Democracy; I find it extremely attractive. I have left out a lot of detail, and a lot of side-roads, such as the discussion of institutional economics (drawing on Knight's earlier book), and everything about philosophical Pragmatism. It obviously matters a great deal to Knight and Johnson to see themselves as heirs to Dewey, and what they have to say about Pragmatism is interesting to me, but I don't see how it makes any difference. Someone who thought the whole Pragmatist tradition was rubbish from beginning to end could still accept all of Knight and Johnson's substantive arguments about politics, institutions, and experimental democracy. (They are after all very close to Karl Popper's ideas in The Open Society and Its Enemies; naturally, Popper is not mentioned.) On top of this, they spill a lot of ink in disputes with fellow political theorists and/or capital-P Pragmatists, and generally using five words instead of two.
None of this takes away from the value of what they've done. But all of this makes it harder to get to that value, and reduces the impact the book will have. (If I pressed it on friends or students, almost none of them would actually read it.) There is a good short, pointed book in here; someone should write it.
*: It's easy, on this basis, to argue for compulsory education strongly focused on assimilating all children to a single common culture. In other words, put this aspect of Knight and Johnson together with Ernest Gellner, and you easily get not multiculturalism but E. D. Hirsch. There's even a formally parallel argument for eugenics, but it'd be much weaker. (It'd be far more intrusive than melting-pot schooling, there's no diversity-preserving equivalent of bilingualism or code-switching, and, most of all, we have no idea of how to do it.) ^
Politics and Political Thought
With thanks to Henry Farrell for a copy of the book (Henry is not, of course, responsible for the use I've made of it)