The Best Political Science Books

What is the mission of political science?

The mission is simply to understand the political world. I’m also a political theorist, so to change it in better ways is a subtext as well.

In your introduction to the Oxford Handbook of Political Science, you mention that political scientists should be concerned with norms as well as facts. Where is the distinction for you, then, between political science and political theory?

I do both: sometimes I do them together, sometimes I do them separately. Political theory used to be one of three things. One is the history of ideas of great, dead white men. Another is applied moral philosophy – what is equality? What is liberty? Why is it valuable? And the third, I will call pompous poetry.

It’s only in the last 15 or 20 years that people thinking about normative questions have started realising that in order to implement their recommendations, they need to actually work their will on the world. And how to do that is something that social scientists can help them with. An example: it looks like a good idea to have a separation of powers, but it turns out that sometimes, things which in principle look like great ideas blow up in your face. And that’s what the social scientists are good at explaining for you. So even if your primary concern is normative, you have a secondary concern with the empirics in order to achieve your normative goals.

That’s what I try to do in my work that’s a crossover between political theory and public policy work. A good little gang of people now are doing that sort of political theory as well.

That’s some recent history – could we sketch a more general history of political science?

Yes. The oldest of the books that I chose for today’s discussion is from the middle of the 20th century, and that’s because I’m supposed to be talking about political science, and modern political science emerged in the middle of the 20th century.

By modern political science, I mean a couple of things…One is big data, statistically analysed – rather than sitting around the common room and swapping stories about the last time you had lunch with the foreign minister.

The second is that it’s also theoretically informed. You have some formal model that tells you why this should go together with that. All the statistics tell you are correlations, and correlations aren’t causation, right? A and B are correlated, but that could be because C causes both of them. It may have nothing to do with any direct relation between A and B. You need a theoretical model that tells you why these things go together. So in political science, statistics join up with a sort of storytelling,  sometimes just discursively expressed but often formal modelling which is mathematically expressed.

Before we turn to your list, could you tell us a little about how you made your choices? It’s a big field to choose from.

Firstly, I confined myself to books from the mid-20th century onwards. Secondly, I assume that this is a discussion for people who aren’t specialists. Hence I tried to choose books that were accessible and didn’t presuppose any background knowledge or technical expertise. That meant I inevitably chose older and more classical contributions, rather than the very latest, up-to-date developments that you wouldn’t be able to make sense of without having read the five things that preceded it. And finally, I tried to choose one book to represent major sub-disciplinary fields: I’ve chosen books about electoral behaviour, political economy, comparative politics, international relations, and political theory of an empirically oriented sort.

Let’s talk about your first choice, which you describe in the Handbook as a ‘watershed’: Campbell et al’s The American Voter.

This was the beginning of modern political science – of empirically oriented, big data, statistically tested work. The book was published in 1960 but it drew on surveys taken during each of the three previous US presidential elections – a smaller number in 1948 and 1952, and a much larger number in 1956. So the book was largely based on the 1956 US election, and that will become important in a moment.

These were large samples of several thousand, randomly selected from American voters. They were all asked the same questions: it was a standardised survey, about 45 minutes long. It probed all sorts of attitudes people had toward politics, and various structural factors – race, geography, income – that might influence those attitudes. Finally, the data was coded and analysed at the individual level. That’s important because it allows you to assess the correlation between, say, age and voting, controlling for all sorts of other things at the same time – income, region, religion, race… You could factor out all these other confounding variables and see just what the relationship is between the two target variables.

The headline finding of the book is that the strongest correlate of a person’s vote is that person’s party identification, and the strongest correlate of a person’s party identification is their parents’ party identification. These authors were all psychologically oriented, and they see this as early childhood imprinting. They took identification really seriously: they saw it as an identity thing, where you say “I am a Republican” in the same way as “I am an American”, “I am gay”, “I am Catholic”.

So the combination of the identity emphasis and the early childhood socialization emphasis suggested to them that party identification and voting patterns are rusted on. You’re not going to change them. The issues, they found, don’t really matter – people don’t have much information about the policy issues, or any coherence ideologically in their attitudes. That’s the headline story, that people are psychologically driven rather than rational, issue-driven voters.

Has that story stood up?

No, the story – the specific findings – have not stood the test of time very well. Just four years after the book was published, the story took a serious hit when the Republicans nominated Barry Goldwater for president. Barry Goldwater was one of the few senators who voted against the 1964 Civil Rights Act. Black Americans had been rusted on Republican voters from the beginning – that was the party of Mr. Lincoln the Great Emancipator, after all – and overnight, they flipped and became a solidly Democratic constituency. So much for that theory of early childhood socialization and a fixed party identification driving everything.

How well-informed are voters? There is a big body of literature suggesting that voters are about as informed as they need to be to make rational choices. How many Americans could name all nine members of the Supreme Court? But do you need to be able to name all nine justices to know that you didn’t like them overturning Roe v Wade? You just need to know that the key justices were appointed by Trump, and if you want to get Roe v Wade back, you’d better vote against Trump, because he’s the guy who keeps putting these anti-abortion people on the Supreme Court. There are lots of models of low-information rationality, stories about how voters can use cues and shortcuts that help them arrive at the vote that they would have arrived at had they been perfectly informed.

Don’t read The American Voter for its substantive conclusions, necessarily, but appreciate it for its contribution in starting out modern political science in its big data, statistically analysed fashion.

Let’s go forward five years now, to your next classic choice. Could you introduce The Logic of Collective Action by Mancur Olson?

This one is an exercise in political economy. Indeed, Olson himself was an economist, but a heterodox one. As I said, the second strand of political science as a science is the provision of formal models that explain why the correlations we observe are there. That’s what political economy historically contributed to the development of political science. A lot of the formal models grew out of various strands of economics, whether it’s game theory or mathematical economics or microeconomics or public finance – or whatever.

Olson’s book is about collective action. He is interested in things like trade unions, where collective action is absolutely central. It starts from premises about rational self-interest and factual premises about the structure of payoffs and incentive structures, and from that, it derives an explanation of why it’s so hard to mount successful collective action, and why it’s harder for some groups than others.

What are the cornerstones of his explanation?

His key concept is public goods. Public goods are goods that if available to anyone are available to everyone, whether or not they paid for them. The classic example is the lighthouse: the same beacon shines out for all ships, whether or not the captain subscribed to its services. Lots of politics is about provision of public goods. Take national defence: the nuclear umbrella covered all Americans if it covered any of them. Or stopping climate change – you can’t stop it for one person without stopping it for everybody. And electoral politics provides public goods – if Labour wins, everybody gets Labour’s policies, for better or worse.

Now the problem with public goods will be obvious from the definition, really: free riding. Why should you pay for it if you can get it for free? How you induce people to contribute to the provision of public goods is a tricky question. When another economist, Paul Samuelson, first came up with the notion of public goods, he took it as an argument about market failure. If things are going to be given to you, whether or not you pay for them, then the market’s not going to work.

So how does Olson think we ever manage to provide public goods?

NATO is always offered as one of the main examples. In another paper, Olson developed a theory of alliances where NATO figures prominently. Why should people contribute their fair share, spending as much as they should on their own national defence, if the US and other major economies in NATO are going to be providing this protection against the Soviets regardless? Trump is going on about this again, and it’s a well-known problem. How did the lesser European powers get by paying too little? Well, the answer is easy: America used to care a hell of a lot about NATO, and was willing to pay more than its fair share rather than withdrawing its contributions and collapsing the whole Alliance. So that’s one way to get public goods: find somebody who’s super keen on the public good, and get him to pay whatever it takes and accept as many free riders as he has to.

There’s a story I quite like (not in Olson’s book) illustrating this way of providing public goods. Howard Hughes, the famous reclusive American millionaire, ended his days living in the penthouse atop his casino in Las Vegas. He was insomniac, and he wanted to watch western movies all night. Now, this was before videos or streaming or computers, and the only way to watch movies all night was for them to be broadcast on free-to-air TV. Hughes was a millionaire, so no problem: he bought himself a television station. He broadcast movies all night, and if anybody else wanted to watch them, it was no skin off his nose. Get one super keen guy to provide the public good for everybody: if other people want to pay, then great. If other people don’t want to, no matter.

Another way of getting people to contribute to the provision of public goods is what Olson calls selective benefits. Everybody gets the public good, and you can’t cut out people because they didn’t contribute. But what you can do is offer extras. For example, cheap health insurance was one way that unions used to get people to join. Unions push up wages for all the workers in the factory, whether or not they are members, but that cheap health insurance only goes to members. Political parties do the same sort of thing: patronage jobs and government contracts and public spending goes primarily, or maybe only, to people who supported the party in the election.

A third way is to have a small group. In small groups, the public good’s benefits are concentrated on a few people, and that makes it easy for people in those small groups to organise reciprocal I-will-if-you-will contribution strategies. If there are just a dozen of us, we can’t afford too many people not to contribute, or else the public good wouldn’t be provided at all, and the I-will-if-you-will strategy is a way of getting the public good provided.

That provides an explanation for what Olson calls the exploitation of large, diffuse interests by small and concentrated interests. Industries where there are very few companies – car manufacturers, for example – can get together and organize a lobby for new rules that make it cheaper for them to produce cars, but worse for consumers. Maybe the cars lack exhaust controls and so they are polluting, or they have crash problems. It’s easy for a small group of auto manufacturers to get together and lobby, and really hard for billions of consumers to organize any collective action to fight back.

And as with The American Voter, a legacy of this book lives on?

It’s spawned a huge follow-on literature. A lot of it is technical, some of it highly mathematical, but some of it is observational and sort of anthropological, looking at ways different groups in different societies manage to pool resources, like fisheries or aquifers or common grazing land. The small group story seems to be really important in most of those examples. Where it’s not a small group, you decompose it into smaller and smaller groups, and convert a large group problem into a small group solution. The most famous follow-on book from Olson’s is probably Elinor Ostrom’s Governing the Commons, which won her the Nobel Prize in Economics. Mancur was dead at this point, and a lot of us saw that as Elinor going and picking up Mancur’s Nobel for him.

Could you give readers a little context about the rational choice turn happening in political science around this time?

Yes. As I said in relation to The American Voter, those scholars were influenced by social psychology. The behavioural revolution was essentially a bunch of social psychologists invading political science. There was one famous dinner party at Charles Merriam’s house in Chicago in the 1930s: he happened to live next door to some psychologists, and he invited them around and introduced them to his political science colleagues, and they got all enthused that this was the scientific way forward for the social sciences. That psychology orientation carried through to the 1960s.

Then you have these marauding economists who came from next door and said, “Those are pretty sloppy, casual models of how people behave.” You know, it’s quite frustrating: when you talk to psychologists, every small tweak in their research design gets to be called a new theory. They don’t have a coherent body of premises that organizes things. So economists said, “You want theory? Here, we have theory: we have rational maximizing agents. Their maximand is a well-structured set of preferences. They do what is instrumentally best to achieve their most preferred outcome in a feasible set. And we can do a lot with just a very few simple assumptions like that for explaining behaviour. So why don’t you see if this helps you?” Olson’s book was one important contribution to that.

Your third choice is Graham Allison’s Essence of Decision: Explaining the Cuban Missile Crisis. Where does this one fit in?

This I chose as a very good example of interpretive international relations. It starts from the classic realist assumption in international relations: national interest is everything. Nations do what they do because it’s in their national interest to do so, rationally understood. And Allison offers alternative accounts to that, not to the exclusion of it, but to supplement it.

What makes it a really neat book is that it focuses on the Cuban Missile Crisis of 1962. US spy planes flying over Cuba discovered tell-tale signatures of Russian missile sites under construction, and just about finished at this point. Cuba is very close to the US. Any missile fire from Cuba would reach anywhere in the eastern half of the US within a matter of minutes, cutting down response time in Washington, which was strategically not good. The US couldn’t allow them to stay – but it was really tricky figuring out how to get rid of them, which involved trying to figure out what the Russians were up to. It was a dangerous time. President Kennedy thought at the time that there was a fifty-fifty chance of it all ending in nuclear war. So the case was of more than just academic interest.

So: what were the Soviets up to, and what should the Americans do in response? Well, the classic unitary rational actor model, the realist model, says that states have national interests and they act in an instrumentally rational fashion in pursuing them. Why did the Soviet Union install those medium and intermediate-range ballistic missiles in Cuba? Because Russia was falling behind the US in the nuclear arms race, and this was a way of catching up, cheap and easy. How should the US respond? On this account, there’s no point trying to talk Russia out of it. If it’s in Russia’s national interest, Russia is going to do it, and talking is not going to help. You have to either invade or bomb, and if you bomb, you have to bomb big, because you aren’t quite sure where all the various missiles and components and warheads might be. That’s the classic realist response, and of course, that’s what the Pentagon was arguing for.

But that’s not the story Allison tells?

There are other sides to the story. Allison offers two other conceptual models, as he calls them. One’s an organizational process model. On this account, states are complex organizations, consisting of many loosely linked sub-organizations. Each of those has a mission and competencies and standard operating procedures of its own. Thinking of it this way, we start asking different questions…

Why were the Soviets behind in the arms race, anyway? Well, it turns out that until just a couple of years earlier, missiles in the Soviet Union were under the command of the Soviet Ground Forces. They were interested in missiles that would complement their mission – the ground war in Europe – so they were interested more in short and medium-range missiles. They had no interest in intercontinental ballistic missiles capable of reaching the US, because that was not their job — and they were in control of missile procurement. That’s the organizational reason the Soviets had mostly short-range missiles that couldn’t reach Washington from Petrograd. Installing them in Cuba solved the problem of how to threaten the US with the missiles they actually had. But they had that problem because of an organisational issue.

If that’s what’s going on, what should the US do? I’ll tell you what they did do. It’s another organizational story. The US ended up secretly withdrawing its missiles in Turkey in exchange for the Soviet Union withdrawing its missiles from Cuba – and had to do it secretly, because they couldn’t be seen to be giving in to Russian blackmail. And here’s the organizational angle to it: President Kennedy was pretty pissed off when he realized that the missiles were still in Turkey, because he had twice before ordered them to be removed, and the army just didn’t do it. So forget the unitary rational actor model — the President was not really in command of this show.

Fascinating. And there was another model in Allison’s book?

The third model that Allison offers is the bureaucratic politics model. This model emphasises that the sub-organisations have goals and interests of their own, which often conflict with one another’s. What happens is just the outcome of ordinary bureaucratic political struggles. One agency is at war with another, and who wins determines what the policy turns out to be.

Looking through this lens, one question arises: why did it take so long for the US to realize that these missile sites were being constructed in Cuba? Well, there was a war going on between the Air Force and the CIA over who should have control over the spy planes. Until that struggle got settled, no flights took off – and this was the six-week period during which the missile sites were being constructed. Fortuitous, from Russia’s point of view.

Very fortuitous.

One last story: how did it all end? President Kennedy went public, demanded that the Soviets withdraw their missiles from Cuba, set up a naval blockade around Cuba and turned back any ship that had weapons missiles aboard. The Americans then got two official cables from Russia. One of them agreed to the President’s demands. The other one belligerently refused.

The first letter was all soppy and personal, talking about our children and grandchildren. The other was full of official prose. And Kennedy figured, “Look, the first letter came directly from Khrushchev. The second letter was negotiated in the Politburo. What I’m going to do is just accept the first letter as the official Soviet response, and ignore the second letter.” What Kennedy was effectively saying was, “I bet there’s a power struggle going on in the Soviet Union. I am putting my money on Khrushchev.” It worked. The Soviets withdrew their missiles.

There’s a sad epilogue for Khrushchev, who was overthrown two years later. Had he known that would be the upshot of his backing down in Cuba, Kennedy’s prediction of nuclear war may have been realized.

This insight into state bureaucracy brings us very neatly to your next choice. Could you tell us about Seeing Like a State, by James C Scott?

Jim Scott is a comparative politics scholar in the classic ‘poke and soak’ school. This involves deep knowledge of the history and culture and mores of a place: you typically would live there for a couple of years. In the old days, he would have been called a political anthropologist, I suppose, although that subfield has gone out of existence now, as best I can tell. He was in the Department of Agrarian Studies, as well as the Department of Political Science, at Yale.

To give you a flavour of what this sort of political science is all about, let me tell you about another of Scott’s books first. His first book was in response to a puzzle, something that really bugged agricultural economists. Agricultural researchers in the US had developed a strain of wonder rice, which had fantastically high average yield, but none of the peasants in Southeast Asia were planting it. The agricultural economists said, “What the hell? What’s wrong with them?” Scott went and lived among them for a while and came back with an easy answer. He said, “Look, they’re operating very close to the margin. They cannot afford a bad year – they’re like men standing in the river up to their neck, where one little ripple will drown them. Wonder rice is wonderful in the sense that its average yield is high, but the variance is also high: when it’s a bad year, it’s a really bad year. With traditional rice, you get less on average, but you can count on more or less the same every year. There’s your answer.” This is the sort of local knowledge puzzle solving that Scott enjoyed doing, and does so well in Seeing Like a State, which is his most general and most famous book.

What is its core story?

It’s a story about state simplification. States impose order on the places and people they’re trying to rule, with an aim of trying to control them more effectively. The extreme versions are the high modernism of the 1950s or the Soviet planning regime, but those are just the big obvious ones. More interesting, in a way, are earlier ways in which all modern states emerged and imposed themselves on messy traditional practices, mostly in order to be able to control and tax the people.

Scott kicks off with a story about scientific forestry. This involved clearing the land and planting straight rows of evenly spaced trees, making sure there was no underbrush growing up in between them, to make it easier to harvest. You would have all these trees of the same size, very convenient, they thought… But it was less good than it seemed, because the undergrowth actually contributed to the quality of the trees. Getting rid of the undergrowth meant your timber wasn’t nearly as good as it would have otherwise been. That’s his opening example of states trying to impose order and legibility for purposes of their own. But there are all sorts of things like that.

One example he gives is standardized weights and measures. He spent a lot of time in Malaysia, where if you ask somebody how far it is to the next village, they will tell you that it’s three rice cookings. Everybody knows how long it takes rice to cook, and how long it takes to walk there is more useful for you than knowing that it’s ten kilometers. Ten kilometers on the flat is very different from ten kilometers in the mountains. But the state’s not very satisfied with those sorts of weights and measures and distance indicators. They want some standard measures on the basis of which they can tax people.

Another example, again tax-related, is land tenure. A lot of places would have been communal lands over which people had particular use rights, with very special arrangements. So maybe one family could graze their cow on the pasture for this month, but only this month, and another could harvest blueberries this year but not next year. How do you tax that sort of use? Much better from the state’s point of view if you take this common land, divide it up between individuals and give them personal property rights – then you know who to tax.

A third example is permanent surnames. In a lot of places, until the middle of the last century, your surname was your father’s name, plus a suffix indicating son or daughter. And of course, that changed every generation – so you didn’t have any way of tracking genealogies across time. In the village, everybody knew, but if you’re trying to organize your population, you don’t really know who’s connected to whom. So at some point, the state put its foot down and said, “You pass on your surname to your heirs.” A continuous family name is the way that a state gets a better grip on the population that it’s trying to govern.

A final example is urban design. If you look at a map of an ancient city, it’s just a hodgepodge of roads going every which way, with no major road cutting through. It’s all well and good for the locals who really know the place intimately, but if you’re an outsider, it’s just impenetrable. If you had a rebellion on your hands, the locals knew all the hidey-holes, and the troops had no idea how to get through this place. So, you need urban design to control the population. Two things happened: one was that new cities or rebuilt cities were built on a grid structure. Easy to find your way around, easy to control. And in older cities, you superimpose big roads, like the grands boulevards of Paris. The point of the grands boulevards was marching troops down them, and they put some of these grands boulevards right through the old mazes that were the nests of revolutionaries.

So all these examples are ways that states simplify and impose order on their environment and the people, in order to better control them. It’s a story about mechanisms of domination.

You mentioned that Scott could be described as a political anthropologist, but that this isn’t a practising discipline anymore. Where should people look for more work like this?

They probably have all gone into what would be called comparative politics these days. Some of comparative politics is big-data driven, and comparison operates at a much higher level of abstraction, but there are still the know-one-place-intimately sort of comparative politics people that would be more like Jim Scott.

One of the reasons that political anthropology has fallen on hard times is that it was a branch of the foreign office during colonial times. Anthropologists would accompany and advise the colonial officers on how to deal with people, and as a byproduct of their day job advising the colonial officers, acquire information about the culture of the place which they would then write up in their academic books. We don’t have colonies much anymore, and the legacy of colonialism embodied in these political anthropology books is hard to handle. So people with a more anthropological interest in politics take refuge in comparative politics instead.

We’ve come to your last choice: Ana Tanasoca’s Deliberation Naturalized: Improving Real Existing Deliberative Democracy. This is your most recent choice, published in 2020. What’s it about?

This is a book of political theory of an empirically informed sort. As I said at the outset, political theory historically was either the history of ideas, applied moral philosophy, or pompous poetry. Empirically informed political theory is a new-ish genre, and there’s a growing body of work of this sort that I think is worth watching.

The normative starting point for Tanasoca’s book is the discourse ethics of Jürgen Habermas, who is a philosopher on the pompous-poetry end of the spectrum I was just talking about. When his discourse ethics got adopted by political theorists, they were trying to give it a more political angle: how would we actually use this in practice? How are we going to operationalize this – what sort of institutions should we set up? What do we need in deliberative systems to make them work the way that Habermas says they should?

So they started thinking about things like: we need to get everybody in the same room. We need to get them informed, but not brainwashed – give them neutral facts that pretty much everybody agrees upon. We need to have a moderator in the room to make sure that everybody is engaged in respectful mutual dialogue.

All this was in aid of trying to implement in the lab the discourse ethics that Habermas was talking about in his poetry. And activists, in turn, got really interested in this. They used these technologies that the political theorists were developing to advise on, and in some cases even make, public policy.

That’s exciting. What did it look like?

They did these experiments in various ways. The first was a set of Citizens’ Juries. Those were about twenty people, modelled on a jury in a law court, and they were given one particular problem to solve – “Should we build a hospital here or there?”, that sort of thing. One was on abortion in Ireland, and one was on constitutional reform in British Columbia. A single topic, maybe a big-ish topic, but just one; a small number of people; and discussions, typically over a weekend. Maybe that doesn’t sound like much, but frankly, a whole weekend spent talking about one single topic? That’s more time than you’ve devoted in a stretch to any decision in your life, right?

Jim Fishkin invented a model of 200 people in a Deliberative Poll. An organization called America Speaks scaled it up to 2000 people. But even in these bigger events, the deliberation actually happened at tables of 20 people or fewer. They all got together for lunch and for plenary speeches, but the real deliberative work was always at a 20-person table. These experiments had interesting results. They showed that people were better informed after the deliberation; that they stayed better informed; and that opinions changed as a result of deliberation, and stayed changed. People didn’t necessarily agree on what they should do, but they agreed on what they disagreed about, so it’s a better-structured conversation.

These looked like interesting exercises that you should try to adopt more widely – to nationalize, if you could. But that’s the problem. How do you scale up these 20-person events to 200 million people? That’s the challenge.

A huge challenge.

That’s where Tanasoca comes in. The scaling-up proposals that were on the table were not at all plausible. A lot of people had thought about trying online deliberation – at least we don’t have to pay the hotel bill of all the people that way – but everybody talking on one big chat? Who’s going to read all that? So another thought was, we can do it all online and use AI to moderate and summarize… It was starting to sound like fantasyland.

Tanasoca said, forget it, you don’t need any of that stuff. It’s easy because it happens naturally in conversation across the entire community. Her crucial insight is built on network theory. It’s all about friends of friends – I know one person, that person knows another person, that person knows another person. Assuming the information flows back and forth across that network in both directions, without too much loss, then in a sense the first person is in conversation with the last person, indirectly. They get information about what he knows, but also about what his concerns and interests and sensitivities are – not in detail, but in a general way. This happens naturally all the time, without any interference.

One sample calculation that I was struck by in Tanasoca’s book was that, given certain assumptions – and they’re not terribly heroic assumptions – it looks like everybody is networked sufficiently closely to 70% of other people in their society to get reliable information from them. (The key assumption is you’re all on Facebook, right? Everybody who’s on Facebook is just over three degrees of separation from each other, and information travels reliably, experiments show, across three degrees of separation. Once you get much beyond that, information loss kicks in.) If you’re in indirect contact with 70% of the people in the country, then it looks like the scaling-up problem has disappeared, or indeed was never there.

Of course, any model of deliberation requires not just that you be in contact with all these other people, but also that you’re prepared to listen and take seriously what they say, and be sensitive to their interests and all that. Tanasoca’s deliberation needs to assume that too, but so does any model of deliberative democracy. She was just solving the scaling-up problem, and I think that was a good solution to it, and an interesting contribution.

It’s so helpful to be able to talk more precisely about these networks – to know how connected we are, and at what degree of separation we start losing information…

The book was distinctive among political theories of deliberation for its heavy reliance on empirical information. You need a lot of facts about networks and how information travels through networks, and how many people with diverse points of view there are in your network. Are you in a bubble, or are you open to at least some dissenting voices in your network? All these things are empirical facts, and crucial to the account that Tanasoca was giving. To her credit, she saw the need for facts and dug them out. She didn’t just speculate and say, “Let’s assume.”

So this brings us back to marrying norms and empirics…

Yes, her book is an exercise in normative political philosophy as well as empirical political philosophy. She ends up with some recommendations for how to improve this real existing deliberation. There are suggestions for how to increase people’s exposure to other points of view, for getting people to discuss sensitive topics that ordinarily you might not talk about in public, and for avoiding mindless repetitions of things –  people just hitting ‘like’ without having read the entry, or Russian bots doing the same. So in that way, it’s a complete book: it starts with the normative problem, it goes through the empirics, and it ends up with normative suggestions.

As you say, it’s the newest of the books on my list. And insofar as people are going to be surprised by any of my recommendations, this is the one. About my other recommendations, people familiar with the field will say, “Oh yes, sure.” This is the one which will make them say, “Hm. I’ll have to go and read that one.”

An excellent orientation to the field. And of course, you also edited the Oxford Handbook of Political Science for anyone who wants a much more detailed birds-eye view. Before we go, I’m curious to see behind-the-scenes – how does a huge undertaking like that come about?

Okay, a shaggy dog story. I edited the Oxford Handbook of Political Science, which built on ten volumes of sub-disciplinary handbooks. I was the general editor of that series, and co-editor of a couple of those volumes myself. But before that series of books, there was a New Handbook of Political Science, which I co-edited. That grew out of the World Congress of the International Political Science Association. I was the Program Chair of that, and it was a lot of work. That was back before many people were on email, so there was a lot of faxing and snail mail… I decided I wasn’t going to go to all that trouble and have only the conference program to show for it. So I convened a series of “state of the discipline” sessions in that conference, and drew on those sessions for the New Handbook of Political Science. That sold very well – it sold 10,000 copies, and Russia produced another 50,000 for deposit in every library in the country.

So Oxford University Press saw this is a money spinner, and they wanted more. I was having drinks with the editor of OUP, congratulating me on the success of the new handbook – “Wouldn’t you like to do some more?”. And I said, “Yes, okay. What about some handbooks for each sub-discipline of the field? I’ll get a specialist in each to organize them.” These were all long books, and the market was American graduate schools. You have to sit field exams in three subfields of political science, and these handbooks would serve as a resource textbook for graduate students across the country.

It worked out pretty well. Then, at the end of the day, I consolidated those ten handbooks into my one Oxford Handbook of Political Science by choosing four chapters from each. That’s how I produced my second overview of the discipline as a whole.

Thank you for taking on the heroic task of slimming down the field yet again, to just five books. A phenomenal introduction to a really exciting field.