Caplan's take on democracy can by summarized as follows: first, he accepts two arguments FOR democracy by democratic enthusiasts, 1. voters are largely unselfish; 2. politicians usually comply with public opinion. He then adds his point: 3. voters are irrational (they have "systematically biased beliefs", or in layman's terms, they have false beliefs). Caplan develops a theoretical framework to prove that it is in fact rational for voters to be irrational because the "price" of their irrationality is low in politics.
The book mainly consists of the following themes: 1. the history of people's economic misconceptions; 2. empirical evidence of systematically biased beliefs; 3. the "rational irrationality" framework and why systematically biased beliefs lead to democratic failure; 4. prescription for overcoming democracy's weakness.
I think Caplan succeeds pretty admirably in 1, 2 and 3, but he is relatively terse at 4. But this is understandable: if you take his arguments seriously, then unless every voter (or at least the "median voter") has a Ph.D. in economics (in fact, she needs to be a libertarian economist!), the outcome of democracy will not be efficient. Increasing the electorate's education, etc. level will somewhat mitigate the situation, but as Caplan himself proves, this is hardly enough (education is not sufficient to eradicate all systematically biased beliefs).
As to the book itself, it is quite readable. I knew about his work before reading the book, what surprised me was how he mixed it with the history of economics with his own research, with quotes and all.
It's also interesting to note that (at least according to my observations) mainstream public choice (the economic approach of studying politics) economists tend to downplay Caplan's work, maybe it is because Caplan's work cuts to the core of public choice (the "rational choice" approach)? Or maybe they really think his work is not much different than rational ignorance? Now that his book seems to have gathered a lot of publicity, maybe others will take a second look.
The only weakness of the book is the part that he repudiates the accusation that economists have "market fundamentalism". His point is basically 1. markets, when free of failures, will lead to efficient outcome (first, "positive", premise); 2. Caplan does not say this, but in most economists' thinking, there is also an implicit second, or "normative" premise, which is that efficient outcome is desirable. In fact, most economists tend to shy away from this conclusion and maintain that they only specialize in cost/benefit analysis and do not make such judgment, but from their passionate, enthusiastic and sometimes vehement arguments for free market, it is not too difficult to detect such deep-rooted belief -- that "free market is good". 3. economists do not always assume there are no market failures, therefore they are not "market fundamentalists". But this is typical economists' thinking: in order to argue with them, you must accept their first premise first, and implicitly also accept the second premise, then the debate about "market fundamentalism" naturally reduces to argument about whether there are market failures. But, they are people who do not accept even the first premise, and there are more -- on moral grounds, etc. -- having difficulty accepting the second. I am not saying I agree with these people (but I have not been blinded by the ivory tower yet, so at least I know the existence of such people and such views). It is very typical of economists to not even acknowledge such views, or pretend they do not exist. It is not an easy task to face these people face-to-face, listen to their arguments, then come up with your own arguments to correct their "biased beliefs", but a good economist should not be daunted. However, this is not a big blemish in an otherwise well researched and well written book, so I am still giving it 5 stars.