In my summer reading about the other social sciences (see here), the title that most intrigued me was Randall Collins’s Sociological Insight: An Introduction to Non-Obvious Sociology.1 The word non-obvious is what got me. As this blog demonstrates, one of my obsessions is figuring out whether political science has created any new knowledge. Another way to put it is whether we know anything that is non-obvious. Collins captures my obsession in his very first sentence: “The aim of any discipline ought to be two things: to be clear and to be non-obvious.”2
Collins goes on to lay out the main non-obvious insight in sociology. In his view, it is “that the human power of reasoning is based on non-rational foundations, and that human society is held together not by rational agreements but by deeper emotional processes that produce social bonds of trust among particular kinds of people.” Rationality does not work as a social glue because it is usually rational to cheat or free-ride on any agreement or social contract. More effective are the emotional bonds created by rituals. This idea can be traced to Durkheim, one of the founders of sociology. Collins goes on to show how these non-rational bonds operate in the prototypical area of religion and then in power, crime, and love (along with a chapter on creating AI written back in 1992).
Of course, the question that occurred to me is whether it is possible to write a similar introduction to non-obvious political science. Do we have a clear and non-obvious idea that constitutes the contribution of our discipline? My candidates for the main non-obvious insights in political science are (i) our conception of state, (ii) our distinctive ethics of political action, and (iii) our skepticism of either direct rule by experts or the people. These may not be quite as striking as sociology, but I will be happy to hear other suggestions.
The State and Violence
I would start the search for our non-obvious idea with one of political science’s ur-texts, “Politics as a Vocation”, written by the German scholar Max Weber in 1919. In it, he lays out two ideas that might form the essence of non-obvious political science.
The first is his definition of the state, the central institution of politics, as the “human community that successfully claims the monopoly on the legitimate use of violence within a given territory.” The key non-obvious element is the inclusion of violence. Presumably the point of politics is to avoid violence. Why then would we put violence at the center of politics, much less give one organization a monopoly on its legitimate use?
The reasons are essentially those that Collins points to. A community often needs to act collectively, most basically to defend itself against enemies. Even though it is rational to work together in this way, we can’t expect sacrifice to emerge simply through individual choice. Even if we agree that it is wise for all of us to contribute to our common defense, each one of us has an incentive to free-ride, to shirk their responsibility and let others sacrifice their energy, resources, and maybe even lives. Collective actions thus often fail or fall short of their goals.
Here is where violence, maybe better termed coercion, enters. A state that claims the right to use violence ensures that any potential shirkers are punished. Someone with a sword or gun shows up at your door if you fail to march off to defend the homeland or cough up your share of taxes. And the monopoly part of the definition makes sure that there are not multiple organizations shaking down the citizenry or fighting each other.
Of course, this all works best if the violence is rarely used and most citizens obey the laws. In functioning states, the violence standing behind it is invisible, which makes the insight non-obvious.3 But even in those better cases, violence lies beneath the surface and it doesn’t make sense to wish it away. Additional non-obvious insights come from understanding when and where the state emerges - ironically, fighting wars is one key stimulus for its emergence.
These arguments for a state challenge one obvious counter, frequently heard from libertarians - why do we need government at all? Doesn’t a government with the power of coercion ultimately just prey on citizens - shaking them down for exorbitant taxes and engaging in needless wars? In fact, one of the main scholars of the origins of the state, Charles Tilly, sees parallels between the state and a mafia protection racket and he is right to see them.
However, if the state is properly limited, as discussed below, the benefits should outweigh the costs. The danger of the strongest preying on the weakest or in Hobbes’s terms a “war of all against all” are bad enough that political scientists tend to believe that a state is justified. Our empirical evidence for this emphasizes life in places where the state is weak or non-existent, say, Somalia or Iraq (for a counter see Leeson). A monopoly on violence is better than a competitive market in violence. As one political scientist famously and controversially put it, “the most important political distinction among countries concerns not their form of government [ie, democracy or dictatorship] but their degree of government [ie, a state or lack of a state]”.
A Different Ethics
Weber concludes that same classic essay by connecting the existence of the selfsame state with a distinct form of morality and this is our next non-obvious insight. Our everyday lives are typically governed by what is called deontological morality. We follow a set of ethical rules or duties. This might be something like Judeo-Christian morality or more neutrally a version of the golden rule. Weber calls this an ethic of ultimate ends. We do good because God commands it or the natural law requires it. There is simply right and wrong as we teach our children.
But, according to Weber, a different form of morality prevails (or should prevail) in politics. This is the ethic of responsibility. Precisely because politics rests ultimately on violence, political actors need to consider the consequences of their actions. If they try to be good in the ordinary life sense, then others will take advantage of them. In politics it doesn’t make sense to turn the other cheek when wronged because then others will learn that you can be taken advantage of. Because of the high stakes of politics one needs to take a consequentialist approach in politics and calculate the likely results of any actions.
Failure to realize the demands of politics leads us to label politicians as bad people. They are liars and crooks. But good behavior does not equal good politics. Good politicians do things that we would never do in our private lives. I love the Danish series Borgen, which follows the trials of a fictional prime minister of Denmark. In one episode she must choose between forming a government that would have a chance to introduce policies that she supports and remaining loyal to her longtime friend and party colleague. She chooses the government (and throws her friend under the bus) because consequences matter more than being good.
The political theorist Michael Walzer calls this the dirty hands problem. Politicians will sometimes have to flout conventional morality - lie, cheat, and betray - because not doing so will lead to worse outcomes. To paraphrase Weber’s example, one’s belief in socialism may be morally correct, but it is possible that advocating socialism at a certain place and time will strengthen the forces of reaction and lead to more oppression. A convinced socialist might say that it is “not he but the world, or the stupidity of other men [that]… is responsible for the evil,” but they would be wrong. In politics, we must take responsibility for the effect of our actions.
For a contemporary application consider recent US elections. Many believed that Trump was an existential danger - to democracy and liberty if not world peace. If this was the case, then they should do anything they could to make sure that he doesn’t win election or reelection. This might entail stepping back from some positions that they consider right, say, support for BLM or higher gas taxes, which could alienate potential swing voters. As Weber put it, “a man who believes in an ethic of responsibility takes account of precisely the average deficiencies of people” who could be turned off by these stands.
Of course, some people do not have the stomach to act this way. They wish to be good Christians or follow their internal moral compass. That is fine, but it simply means that they are not cut out for politics, they don’t have the vocation. Only those willing to make the hard choices - sometimes doing bad to achieve good - have the vocation.
This may seem to lead us to a Machiavellian or might makes right view of politics. While politics does sometimes end up like this, Weber wants us to stop short of that view. The true politician does have goals and ideals which they pursue, but they pursue them with a realistic sense of consequences.
Here is how Henry Farrell eloquently puts it,
Weber is arguing against a specific kind of unworldliness, which assumes that from good actions only good things come, and from evil actions only evil. His claim is that the world of politics is at best a morally ambiguous one, in which wicked means can produce good outcomes. Those who fail to recognize this should withdraw entirely (as a truly religious vocation demands) from worldiness. Those who recognize this and are not pure creatures of politics who shift their positions according to interest and convenience, are, for Weber, genuinely heroic individuals, who have truly embraced politics as a vocation.
The Myth of the Good King and the Good People
Finally, it is tempting to think that most political problems are illusory or easily solved. This view comes in two versions. One is faith in a good king or leader who will put aside all the bickering and simply do what needs to be done. Politics is really administration, which is a matter of expertise and will. Once the experts or the honest are in charge, our problems are solved. This is the politics of the fairytale where we find a good king and then live happily ever after.
The other is faith in the people. Here the problem with politics is that it ignores the people. Elites are always corrupt and prey on the people at every turn. Power corrupts them even more. If we only put the people in charge, we would avoid the corruption of politics and provide real benefits. This might mean rule by referendum or through New England-style town meetings or perhaps even sortition, the random selection of officeholders by lottery. More feasibly term limits would put strict controls on politicians. In any case, this is the worldview of the populist, purportedly fighting for the people against the elites.
The non-obvious insight of political science is that both perspectives are self-defeating. The idea of expert rule fails in several ways. (See Dahl’s Democracy and Its Critics for these arguments in dialogic form.) Most centrally, major political decisions almost always depend on values. Experts can’t say which values are objectively better than others. Even a supporter of epistocracy - rule by the knowledgeable - like Robin Hanson argues that the first step is to tell the experts (or the prediction markets) what values they should be maximizing. Only then can we choose the right policies which fulfill these values.
Then consider what happens when we try to install expert rule. In the first place, we have to identify the experts. Perhaps a meritocratic exam would do this as in China, but the content of that exam would be controversial. And once we identify the experts, what motivates them to work hard and pick the right options? Knowing the good and doing the good are different things. Would they have the proper incentives to do what’s right? Why wouldn’t they use their power to enrich themselves and their friends?
Farrell, Mercier, and Schwartzberg go further. They note that human reason is much better at identifying the flaws in the arguments of others than the flaws in one’s own beliefs. The best ways of finding the truth therefore expose one’s ideas to criticism from others. Think, for example, of peer review in science or a jury trial. Democracy functions in the same way by letting people criticize and sanction their rulers.
But simply letting the people rule also fails. Perhaps most centrally, there is no “will of the people” to be represented. We can’t really say what the public wants - its views are inchoate, uninformed, contradictory, and changing. These are the reasons that political scientists tend to be skeptical that referendums can be used effectively. They are too easy to manipulate and rarely give a clear answer of what people want. Just notice what happened after Brexit. The British people, or 52% of them, wanted Brexit, but what exactly did that mean? It took several years and more than one prime minister to come to a provisional answer. Even elections don’t deliver clear verdicts. What exactly were voters telling us when they chose Trump? That they wanted a wall? That they disliked Hillary? That America should be great again?
Institutions also put limits on translating the public will directly into policy. Most famously, Arrow’s impossibility theorem shows how no voting system can satisfy a seemingly simple set of rules including that no one voter is a dictator. Other analyses lead to a set of “chaos” results where, depending on the voting system and the preferences of the public, the policy chosen can cycle - option A beats option B which beats option C which then beats A - and we can end up just about anywhere.
This is not to say that expert rule and citizen rule don’t have a place in a good political system. They do. Expert rule plays a role, for example, in technical matters like monetary policy, and there are arguments, like Garrett Jones’s, that perhaps we should increase their role a bit - “10% less democracy” in his terms. But moving fully to expert rule, sometimes called guardianship, has major hazards.
Similarly, citizen rule plays a role. Almost all political scientists believe that democracy is superior to the alternatives. But they define it in a particular way that mostly focuses on periodic elections rather than direct rule. The key is that citizens provide a check on representatives which incentivizes those representatives to do better. More generally, good systems of government tend to set up institutions which check each other. “Ambition shall be made to counteract ambition” in Madison’s terms. These checks can come from below - as citizens throw the bums out in elections - or they can come from the side - as executives, legislatures, and courts vie for power and check each other. However they are enshrined, they make it hard for tyrants to emerge and limit the downside of whoever is ruling.
If I had to sum up the non-obvious insight here I might call it a version of Ulysses and the sirens. Ulysses had himself bound to the mast so that he could listen to the music of the sirens without being tempted to jump into the sea. In the same way, putting limits on government makes it more effective. An unbound government whether in the hands of experts or the people ultimately fails.
Going further
What does this add up to? I wish I could reduce these points to one big idea, but maybe there isn’t one. Ansell’s recent book Why Politics Fails (which I summarize and review here) proposes five different ideas, so maybe I’m doing well to get it down to three.
Am I missing something? Or would it be better to say that there are political science insights, but not political science insight. Maybe we are more foxes who know many things as opposed to hedgehogs who know one big thing. And that might not be bad.
His blog, The Sociological Eye, is also excellent.
I won’t talk so much about what obvious means here. As Duncan Watts titled one of his books, “Everything is obvious”, and common sense leads us to believe that we understand more than we do. I suppose obvious is always obvious to whom and here I am imagining a perhaps naive John Q. Public. Earlier I wrote about a similar concept, the intuitive.
Collins takes this same insight further. He argues that the state is a product of deeper forces - the emotional bonds created by communal rituals - which induce allegiance. Political scientists typically put less emphasis on these rituals, but Collins makes a persuasive case for them.