I just finished reading Bryan Caplan’s excellent book “The Myth of the Rational Voter”. It is an extremely valuable book for anybody who really wants to understand what drives democratic government.
Public choice theorists have long known about “rational ignorance” where people recognise their vote is almost worthless and so they don’t bother informing themselves about politics, political philosophy or public policy. The consequence is that most people can’t explain how the political system works, who their politicians are, the details of most policies, or the arguments underlying public policy debate.
Caplan’s addition to the literature is to say that people aren’t simply “rationally ignorant” but that they’re “rationally irrational”. Ignorance should lead people to making “random errors” where they are equally likely to make mistakes in either direction. But Caplan shows what political watchers already know… voters often make “systematic errors”, where many people are wrong in the same direction. The problem isn’t that people don’t know enough (ignorance), but that what they “know” simply isn’t right.
To explain this, Caplan suggests that people have “preferences over beliefs”, and once you think about this idea it becomes self-evident. There are certain ideas that people simply prefer to believe in. The most obvious one is religious, where people have a preference to believe the religion that they were brought up with: most Thais are Buddhist, most Indians are Hindus, most Serbians are Orthodox, most Colombians are Catholic, and most Iranians are Muslim. What an *amazing* coincidence!
Of course, if there is enough of a consequence, then many people can overcome their “preference over beliefs” and change their mind to achieve the better outcome. But if there is no consequence to your belief, then people will believe whatever makes them feel good.
Each vote has effectively no consequence so people stick with what makes them feel good. The problem with politics is that believing “whatever makes you feel good” can have a very negative consequence on society if the “feel good” policy is not the same as the “actually good” policy. Caplan outlines three areas where the average punter gets politics wrong: (1) an anti-market bias, where they don’t understand how markets work; (2) an anti-foreign bias, where they have overly-negative impressions about the impact of foreigners; (3) a make-work bias, where they have an overly-negative attitude towards new technology; and (4) a pessimistic bias, where they think things are worse than they really are.
Economics teachers encounter these biases every year and need to slowly (and painfully) try to explain to students that price controls don’t work, that comparative advantage means trade is good, that labour-saving technology makes us better off in the long run and that we are richer now than any time in history. The mistakes people make about these issues aren’t “random” in that half the class is “anti-foreign” and half is overly “pro-foreign”. The mistakes are systematically in one direction.
Understanding the bias
My complaint with Caplan’s book is that he doesn’t give sufficient attention to understanding why people have a certain “preference over beliefs”. Later in his book he seems to imply that we can fight to change these preferences, but I think that is naive. I suggest that evolutionary psychology explains what I see as the main political biases that people have:
1) conservative — people like to be accepted as “normal” and so will instinctively stick with the status quo and distrust big new schemes. The evolutionary reason is that the “trend-setter” often makes mistakes while the imitator can follow the good and avoid the bad.
2) like to be scared — everybody has a level of risk-aversion, but my theory is that we have evolved an instinct to be scared. The evolutionary reason is that the people who weren’t scared of tigers or starving ended up dead. The problem is that because “fear” is hardwired into our brains, as we solve previous problems we shift our fear onto new, less-serious, problems. Effectively, we are becoming more risk-averse.
3) need for control — because we have a sense of control over our own actions, we have developed an instinct to assume that if something is done, then somebody must be responsible for it. That is why it is comforting to belief in creation and it takes more of an imagination to consider an uncontrolled evolution of life. When something goes wrong, we want to find who is responsible. If nobody is, we often look for scapegoats. Consequently, people will assume that somebody should control the economy and it will take more of an imagination to consider the virtues of an uncontrolled market.
4) nice guy — while we have little incentive to research political philosophy, we do have a strong incentive to consider moral philosophy and the question of “how to live a good life”. Consequently, people have an instinct to judge politicians on what they know about, and support them on the basis of who gives voice to moral preferences. This gives rise to the common confusion between moral and political philosophy.
I think all of these instincts are understandable given the history of humanity, and are probably hard-wired into our minds. If somebody has enough of an interest in political philosophy and enough intellectual honesty they can overcome these biases… but due to “rational ignorance” most people will just stick with their biases.
The consequence is that we should expect politicians who play on your fears and then offer to fix them; to say that they won’t change too much or make risky decisions; and who confuse the moral with the political.
I think these four basic biases explain Caplan’s four specific biases (and more). I also think that the last bias captures another public choice theory — the Brennan & Lomasky theory of “expressive voting” — and fits it in nicely within the broader “rational irrationality” framework. Caplan played his theory as an alternative to Brennan & Lomasky, but I think they are part of the same explanation.
Where does this leave us? The logical consequence from the instinctive voter biases is that people will generally vote for a mix of the status quo and a little bit more statism. If voter preference is a major driver of politics, we should have a slow (due to conservative bias) ratcheting-up of the size of government (due to fear, control & nice guy bias), and the maintenance of bad (but popular) policies.
So is liberalism doomed?
Maybe, but not necessarily so. Democracy is one of the main institutions of our current system, but the other fundamental institution is jurisdictional competition, which forces politicians to also worry about the consequences of policy. Also, as Caplan says, politicians have “wiggle room” which they can use to follow an “intellectual zeitgeist” of sufficient merit. I think this explains the shift towards more free trade and market liberalisation in the 1980s (which were never popular).
Different sorts of democracy
Caplan also stresses that democracy can come in degrees. I also think it’s worth noting that democracy means different things to different people. The most crass definition is that 51% of people get to determine everything, but there are some other options, including (1) constitutional democracy; (2) limited franchise; and (3) decentralised democracy.
* When Alexis de Tocqueville famously praised American democracy, he was mostly praising the liberalism of the system where everybody had the freedom to pursue their own life story, and the civil society where people built strong voluntary communities. He was not praising “ballotocracy” where all decisions are made by majority rule, but a system of “constitutional democracy” where the government was strictly limited and so people were mostly free.
* A more controversial approach is to alter the franchise of who can vote. At one stage democracy only included land-owners, and later it only included men. Another option is for the franchise to only include people who are net taxpayers, or to exclude bureaucrats because of conflict of interest. And there is the option of having people pay for the privilege of voting through a poll tax. On the other side, the government could try to expand the franchise by forcing people to vote, even when they don’t want to… thought this seems likely to only make the problems of democracy worse.
* A final approach to adjusting democracy is to have the power of government limited to smaller jurisdictions. This achieves a number of improvements, such as (1) improving jurisdictional competition, which encourages more efficient governance; (2) increasing the value of people’s vote because there are fewer voters per jurisdiction; and (3) allowing people a more direct and personal relationship with their government.
Democracy is one of those buzz words that everybody uses. China and Iran call themselves democratic, and according to some definitions perhaps they are. But the distinctions between different sorts of democracy matters. Caplan has shown well that democracy can have a cost, and so it is wrong to automatically assume that we always need “more democracy”. That is a recipe for ever-larger and more populist government. What we need is a “smarter” and more effective democracy… similar to the one that was so admired by Tocqueville over 100 years ago.
(Cross-posted at my personal blog.)