Jason Brennan
Supposedly, on election day, We the People speak. We the People decide who leads and who will represent us. We reward the competent, punish villains, and choose replacements. We the People might not quite be the captains of the ship, but every few years we at least get to install a new captain and nudge the helm.
Supposedly, elections legitimise our government and our leaders. While queens, emperors, communist central committees, and dictators foist law upon their subjects, elections turn us into co-ruling citizens who exert equal authority. Elections ensure that those who lead answer to, act on behalf of, and carry out the Will of the People…or at least the will of the winning coalition.
In turn, elections supposedly create juicy normative problems for philosophers to solve. How should we balance voters’ conflicting preferences as expressed through their votes? Whom should representatives represent and how? When should representatives defer to what the voters want and when should they exercise their own judgment?
How we answer these questions – indeed, whether these questions even make sense – depends in turn upon how democracy works and what voters are doing when they cast votes. Our philosophical theories of democracy must be responses to the empirical work in political science, political psychology, and economics that tells us what voters are doing, what they know, and how they think. However, once we examine what this empirical work says, it turns out many of our pet theories of democracy make little sense and our many philosophical problems are illusory.
US students commonly take civics courses around sixth grade. Teachers push a basic theory of how democracy works. This theory – call it “sixth-grade civics” – means both to explain how democracy works and to provide a philosophical justification for democratic governance. More advanced philosophical theories tend to stick to variations of the sixth-grade model. But what if this model is simply wrong?
The model goes as follows: First, assume that most citizens have a set of interests or concerns. (These needn’t all be selfish interests.) Second, the model says that since citizens are motived to promote their concerns, they learn how the world works and what different policies might do. Third, on the basis of their interests and this information, they form ideologies, defined here as coherent and relatively stable sets of preferences about policies and about what goals politics should achieve. Fourth, they then examine the various political parties and candidates on offer, and on election day vote for those which best match their preferences, perhaps settling for weaker matches which have a better chance of winning. Fifth, since every voter acts this way, the winning parties and candidates will thus tend to match and will try to implement the shared or overlapping preferences of the winning coalition. Since politicians want to be re-elected, they will throw some crumbs to other voters as well. Sixth, if the winners failed to keep their promises, or if they did a bad job, come the next election, we’ll keep track and vote the bastards out.
If this model were true, it would seem to justify democratic rule. Democracy ensures that at least some of the people get their way. It incentivises leaders to listen to the people and to do their jobs well. It transforms popular opinion into law.
Even if this model were true, we would still reasonably worry about whether democracies enable the majority to ignore, oppress, or exploit the minority, or whether representative democracies allow special interest groups and government agents to exploit the masses. We can still ask whom leaders should represent and how, and when leaders should defer to the masses or substitute their own judgment. But the bigger problem is that sixth-grade civics is largely obsolete and mistaken. Philosophical theories which rest on it are too.
If you want to truly understand the majority of voters, think of sports team loyalties. It’s not as though most fans first learn about the sport and then select the team that best supports their values. It’s not like they switch teams when their teams suck. Rather, a person from Boston roots for the Red Sox and Pats because that’s what other Bostonians do. It’s sort of arbitrary and contingent that your local peer group picks the Mets over the Yankees, but once they do so, you enjoy all sorts of social benefits from sharing their fandom. Wearing the team colours and following the games helps to signal to others that you are loyal part of a particular network, and so can be trusted as a potential business partner, mate, or neighbour. Hating the rivals proves your loyalty all the more.
This, in short, is what most voters are doing when they form party loyalties. Let’s contrast sixth-grade civics with an empirical theory called “democratic realism.” As Kwame Anthony Appiah summarises it: “People don’t vote for what they want. They vote for who they are.” For most people, politics is not about policy, it’s about identity.
One widely known problem with the sixth-grade model is that most people know hardly anything about politics. They might know the price of a bus ticket or the local cost of day care, but they don’t know local or national statistics, whether crime is up or down (even in their neighbourhood), who is in charge, what the people in charge did or could have done, how the economy or different laws work or could have worked, or who their own representatives are. Most voters are ignorant about almost all basic political information. Some democrats pound the table here and insist voters can follow easily thought leaders or use heuristics. But that’s no help – you have to know which thought leader is reliable, which itself requires knowledge that most voters lack.
A less well-known but in a way more disturbing problem is that most voters are agnostic about politics. They don’t have political ideologies or real policy preferences. If we survey them about some issue, most will provide an “opinion” on the spot – they hate saying they don’t know. If we survey them a short time later, they’ll offer a different “opinion” but will deny they ever changed their mind. Based on decades of research like this, Donald Kinder and Nathan Kalmoe estimate that fewer than 1 in 5 Americans has a real ideology or real political preferences.
Instead, as Christopher Achen and Larry Bartels show, most voters vote for a party simply because that’s what people like them do. Voting Democrat proves to other college professors or Boston Irish Catholics that I’m one of them. Voting Republican shows other Southern White Evangelicals that you’re one of them. It’s just like rooting for a sports team and brings the same kinds of social benefits. Different identities become attached to parties for largely arbitrary historical reasons, and entire groups can switch parties overnight over essentially random events. Demographic groups do not usually vote for parties because that party is especially good for them or because the party’s ideology matches their own.
On the contrary, while most citizens are agnostic, even the more “opiniated” voters are usually following their party’s lead rather than picking a party because they independently agree with its ideas. Some voters learn what their party stands for today and then rationalise that they agree. For instance, surveys show Republicans voters switched from being apparent free traders to protectionist almost overnight once Trump became the presumptive nominee. These same voters sincerely but mistakenly report that they’ve always been protectionists. If this seems weird, think again about sports fandom. Last year, Patty at the Lansdowne Pub proclaimed that Tom Brady is the Greatest of All Time; now that Brady plays for Tampa, Patty declares that he always thought Brady was overrated.
This also explains political psychologists’ persistent findings that most people do not reason about politics in a scientific, truth-tracking way. Voters are almost impervious to evidence. Instead, they only look for or trust evidence that helps them believe what they want to believe; they dismiss anything that disconfirms their beliefs. Indeed, the smarter someone is, the better they become at rationalising that their side is always right. If voters were genuinely trying to promote their independent political interests, we’d have to say they are incompetent and irrational. But their bad reasoning skills pay social benefits. Remaining loyal to your team no matter what makes others on the team like you more and raises your status.
Some voters – including many readers of this magazine – do fit the sixth grade model. They have genuine ideologies and pick the matching parties. They do not simply parrot what their party says. But these voters are the minority, perhaps only a tenth of the electorate.
What justifies having this party rule rather than another? According to the prevailing view – and most democratic theories that philosophers have produced – it’s not merely that people happen to have voted for that party. Instead, it matters that we voted for you because we believed you’d promote our goals, interests, and policy preferences.
However, the realist view of political behaviour says few voters vote that way. Achen and Bartels conclude that elections, far from expressing the Will of the People or even the individual wills of individual people, are essentially random events. Voters don’t share their candidate’s politics or ideals and they don’t genuinely believe the candidate will promote their interests. For most Democrats, saying, “I’m pro-choice” carries no more ideological commitment than waving the Terrible Towel at a Steelers game. The statement might literally say “Women should have a right to an abortion,” but the typical Democrat means little more than, “Hooray Democrats!”
Consider what this means from the perspective of an elected political representative. A classic debate in political theory concerns whether the elected representative should act as a delegate, who does what her constituents want, or a trustee, who does what’s best for her constituents even if they don’t want it. The political philosopher Alex Guerrero suggests that the higher percentage of votes the winning representative receives, the more entitled she is to act as a trustee rather than a delegate. A bigger victory demonstrates trust in the candidate’s independent judgment.
But if the realist picture is true, it’s hard to make sense of this philosophical theory. Elections outcomes don’t quite represent the will of the people, not even the people who voted for the winning candidate. It’s not merely that the people don’t have a collective will, but even individual voters usually lack a political will other than that their political team – a team they support to impress their peers – wins. Winning by a large or small amount doesn’t usually show how much people believe in or trust you. It instead simply reflects how many voters of different demographics there are in your district.
Elections look something like this: Imagine that long before the election, a wizard cast a magic spell which arbitrarily made blue-eyed voters prefer Party A and brown-eyed voters prefer Party B. Imagine such “preferences” stuck around, even though A and B have no special tendency to promote the interests of blue or brown-eyed people respectively, over the years, people continue to have the “preferences” the wizard imposed upon them. Now, suppose it is election day. In Metropolis, thanks to random historical circumstance, 70% of the citizens are brown-eyed, so the candidate for Party B wins by a large margin. In Gotham City, only 51% of the citizens are brown-eyed, so the candidate for Party B wins by a small margin. It seems implausible on this picture to conclude that the Metropolis representative has greater legitimacy or a stronger mandate than her Gotham City colleague. For the same reason, then, we should not conclude a real-life representative’s margin of victory determines her mandate or determines whether she is a delegate or trustee. In short, the amount of support a candidate receives rarely reflects genuine agreement with her goals, endorsement of her platform, or trust in her abilities.
Let’s look at one final issue. In political philosophy today, one of the sexiest problems concerns “public justification.” Following John Rawls, Stanley Benn, and Gerald Gaus, theorists claim that coercive laws are legitimate only if they can in some way be justified in light of different citizens’ differing views of justice. We need to show that people with lots of diverse opinions have good reasons to endorse the laws.
However, it looks like this is something of a pseudo-problem for 90% of the electorate. Most citizens lack the kinds of opinions which public reason liberalism is supposed to adjudicate. They don’t have diverse political conceptions because they don’t have political conceptions, period. Worse, most citizens lack the kinds of normatively significant beliefs upon which public justification might rest. We cannot ground an argument for a policy on shared premises if, in fact, most citizens accept no relevant premises. We cannot offer them reasons that hook onto their beliefs if they have no beliefs.
This short essay probably won’t convince you to abandon the sixth-grade model. You’d want to read the relevant research first, which I’ve barely summarised here. But the upshot remains intact. How we philosophers think about democracy should depend upon how democracy actually functions. We can’t do democratic theory in isolation from the social sciences, and we’d best not insist on sticking to an outdated model of democracy simply because it lets us work on philosophical problems we enjoy. Questions about what justifies democracy, what leaders ought to do, and whether voters are equal co-authors of the law depend not merely on philosophical analysis, but on how voters in fact behave and why.
Jason Brennan is the Flanagan Family Professor at Georgetown University. He is the author of thirteen books, including most relevantly, Against Democracy (Princeton University Press, 2016), which has been translated into ten languages.