Bryan Caplan  

When to Trust Your Superiors

Krugman on Scientific Method... TARP: A Gift for the Bondholde...
A while back, James Donald left a somewhat strident comment on EconLog.  The key passage:
When the superior rule the inferior, it is not only better for the superior, it is also better for the inferior.
Many readers will reject Donald's claim out of hand for its tone or undertones.  Yet there are definitely conditions under which Donald is correct.  Namely:

1. "Superior" means epistemic superiority - i.e., superior understanding of how the world works.  For example, superior understanding of the effects of government policy.

2. Superiors care at least moderately about the welfare of inferiors. 

The reasoning is simple: Suppose someone with power over you understands the consequences of various options better than you do.  If he cares about you at least moderately, he will typically make better choices for you than you will.  Picture the parents of a young child. 

If either the cognitive or motivational assumption is false, however, Donald is incorrect - and often tragically so.  Someone can be superior to you in health, strength, trust, or cooperation, but that doesn't enable him to make better decisions for you than you can.  And someone can be vastly superior to you in understanding, but if he doesn't care about you, he will typically make himself happy at your expense.  Picture a savvy slave-owner.

Public choice economists will be tempted to dismiss the optimistic case out of hand.  They're too hasty.  Within modern First World democracies, narrow self-interest tells us almost nothing about policy preferences or voting.  Instead, most people vote on the basis of perceived public interest.  Furthermore, there is strong evidence that people with more education and higher IQ have unusually sensible policy views.  The upshot, as I've often argued, is that even people with little education and low IQ are better off if people with little education and low IQ don't vote.

There's just one big catch: Some definitions of the "public interest" are more public than others.  Virtually no electorate puts more than trivial weight on the welfare of foreigners.  As a result, even well-informed voters often support immigration restrictions and aggressive foreign policies with hellish consequences for foreigners.  Many other electorates put trivial weight on the welfare of native out-groups.  Think about the way the U.S. historically treated Indians and blacks, the way that South Africa treated blacks under apartheid, or the way the Belgians treated the Congolese.  Some especially malevolent electorates actually put negative weight on the welfare of foreigners and out-groups.  Superior German technological understanding combined with the Nazi Weltanschauung was terrible news for the world's "Untermenschen" - and yes, the Nazis were (initially) democratically elected.

Blanket statements about the effects of rule by "superiors" give enlightened elitism a bad name.  When your epistemic superiors identify with you, deference is prudence.  When your epistemic superiors regard you as outsiders, chattel, or vermin, deference is folly.  Is superior rule better for all concerned?  It depends on the sympathies of the superior.

HT: Vipul Naik

COMMENTS (30 to date)
FredR writes:

I agree, and it's a really good point.

One question I've been toying with is: doesn't this line of thinking imply that we should do what we can to have equal proportional representation of all populations in elite positions in America, so that, for instance, our supreme court would be more protestant?

Mike writes:

Surely your conditions above are a bit too minimal to imply that rule by superiors is better? For example, suppose agent A cares equally about agent A and agent B and makes 90% of decisions correctly (epistemically). Agent B cares about agent B three times more than agent A and makes 91% of decisions correctly. This meets your two conditions (B is epistemically superior to A and B cares at least moderately about A) but rule by agent B is probably worse than a variety of other rules (rule by consensus, rule by A, etc.) according to most moral theories. In other words, rather than defending #1 and #2 above, you're really defending the view that the degree of epistemic variability is greater than the degree of variability in morals, in some sense. I think many egalitarians would argue that moral variability is high (i.e. people fail badly at caring enough about people not like themselves) and epistemic variability is low (a lot of people can make decent decisions, at least in an environment with strong institutions).

Bedarz Iliaci writes:

There is ontological superiority as well.
A husband is ontologically superior to his wife and he should rule her politically.
A military commander is ontologically superior to the soldiers under him.
A parent is also ontologically superior to his children and remains so even if and when he no longer possesses epistemic superiority.

The State is ontologically superior to an individual that belongs to it.

The ontologically superiority is not a matter of economic calculus and needs to be recognized as a first principle of society. That is, without children honoring their parents and wife obeying her husband, one does not get a lasting society in the first place and so on with an army where the General is not obeyed by the soldiery.

Tom West writes:

A husband is ontologically superior to his wife and he should rule her politically.

I think we need to invoke Poe's law here:

Poe's Law is an axiom suggesting that it's difficult, if not impossible, to distinguish between parodies of religious or other fundamentalism and its genuine proponents, since they both seem equally insane.

Bedarz Iliaci writes:

Epistemic superiority requires an inferior to know that he is an inferior. Thus there is a circularity that may turn vicious--a person may be so epistemic-ally challenged as to not know that he is an inferior and stubbornly refuse to accept his epistemic superiors.

The utility of the ontological superiority lies in offering the way out of this vicious circle and thereby defining unambiguously a superior that must be obeyed.

Paul Crowley writes:

A clear example of where both conditions and the rule holds is that we will take our cat to the vet against its will.

david writes:

@Tom West

I think he's serious - he has a whole blog to that effect. He didn't link to it but it's Googleable.

david writes:

He does identify a problem, actually. When you invoke drawing upon centuries of tradition to justify a boundary separating individuals who you care about from individuals you don't, there is always someone who wants to finesse a even more ancient, layered hierarchy - one that includes them and excludes you.

ladderff writes:

[Comment removed for rudeness. Email the to request restoring your comment privileges.--Econlib Ed.]

Sonic Charmer writes:

Golly, it sounds like would-be immigrants should think twice and stay away from all these dangerous 'epistemic superiors' in the U.S. Better safe than sorry.

Grant Gould writes:

I would have thought that the Hayekian notion of dispersed, local knowledge would put epistemic superiority away once and for all.

No matter how wise and caring the superior, that superior cannot be wise enough or know well enough how to care that we will all in general be better off bowing to him or her. It is the inferiors, their hands actually dirty with the details, who are epistemically equipped to make sensible decisions and to figure out what "caring about them" actually means in context.

There is no need to fill the superior's shoes with Nazis to get a bad outcome. If everyone bowed to the superiority of saints and geniuses they would still end up worse off than if they distributed that authority to the people with the local knowledge.

MikeP writes:

I would have thought that the Hayekian notion of dispersed, local knowledge would put epistemic superiority away once and for all.


A little epistemic superiority is a dangerous thing.
Drink deep, or taste not the epistemic superiority spring.

Of course, I implicitly read that in the original article: almost always the epistemologically superior action is to allow individuals to make their own choices.

One big exception would probably be writing Constitutions so that voters can't abuse individual rights.

Slocum writes:

@Grant Gould -- Exactly right. I was quite surprised that Bryan did not include that point in his post in the first place.

DougT writes:

Plato lives!

Wasn't it Whitehead that said most philosophy can be conceived as a series of footnotes to Plato? This is just Plato's notion of the Guardian class, who are ontologically superior to the warriors and artisans. The truly just society will have all three classes working in harmony, just as a healthy individual will have his intellect, spirit, and physical body in harmony.

You could look it up.

Phil writes:

I would ask James Donald to name a set of superiors he's willing to submit his life to.

A defensible version of the argument is more like, IF you MUST be ruled by SOMEONE, better that it be a "superior" in the sense that Bryan describes.

Bostonian writes:

Caplan wrote, "Within modern First World democracies, narrow self-interest tells us almost nothing about policy preferences or voting."

Blacks and Hispanics vote heavily Democratic because blacks and Hispanics have lower incomes and are more dependent on the welfare state. Whites vote Republican for the opposite reason. Government workers and union members are more likely to vote Democratic than private sector, non-unionized workers. People do tend to vote in the self-interest of their group.

Sonic Charmer writes:

To follow Bostonian's comment, we can posit two general self-interested forms of voting: (1) pure, rank, out-and-out tribalism; (2) because one sees oneself as an 'epistemic superior'.

I think Bryan's post (and the subsequent comments reminding us of Hayek) may have convinced me that this 'epistemic superior' thing is, like, almost as bad as tribalism.

egd writes:
Furthermore, there is strong evidence that people with more education and higher IQ have unusually sensible policy views...

There's just one big catch: ...even well-informed voters often support immigration restrictions and aggressive foreign policies

Well informed, highly educated voters tend to have good political views.

Well informed, highly educated voters support immigration restriction.

I oppose immigration restrictions.

Therefore, it follows that well informed, highly educated voters are wrong on the subject of immigration restrictions.

If you're correct concluding that well informed, highly educated voters are generally right, but wrong for supporting immigration restrictions, doesn't the same rationale support any other area of political disagreement with informed voters?

Ted Levy writes:

Those interested in Bryan's argument, including Bryan himself, may be interested in Jason Brennan's article on voting restrictions:

Joe Cushing writes:

I've come to believe that number 1 is true and number 2 is false about most politicians vs. the average voter or supporter. The average person believes in tariffs, soaking the rich, protectionism, hefty regulations, anti-gouging laws, and on and on. The average politician has to know that these things are not in the public's interest because if they didn't learn it in school, they have been told by the minority of us who have written them to explain it to them. Also, if they want to help a 3rd world country rise, they prescribe libertarian economic policies in their advice to them. So number one is true, yet our politicians regularly write and sign laws that go against the principals a number 1 politician should have. This means number two is false.

Saying number 1 is true and two is false, is another way of saying that most politicians are evil. I didn't always believe they were evil. I used to think they wanted to make changes for good but were naive or got sucked into the system somehow and failed. I don't believe this anymore. I know a disproportionate number of them are psychopaths and most of the rest are at least able to suppress feelings for the fellow person. Nearly all politicians are evil.

I used to question whether it was even possible for people to have evil inside them or if everything people did, they did out of a sense of good. Even Hitler believed he was doing good in a twisted mind sort of way. The same for the communist leaders. I don't believe that anymore. I believe they are evil.

Les Cargill writes:

South Africa is very nearly a failed state. It's beyond a poor example.

*The* important thing to consider about people being considered epistemic equals is that communication has a higher probability of occurring. That means even the "inferiors" will make errors that do less damage than had they been in some other arrangement. Even if it's not measurably true, it's probably better if the "superiors" can use persuasion rather than resort to "because I say so."

Mike Rulle writes:

There is some kind of inference here, beginning with Plato, that the better, smarter, wiser, etc., can somehow be those in political power. We know this not to be true and can virtually never be true. Therefore, we should lesson the desire for large government and let the better, smarter, wiser, etc., create private entities to employ and/or educate others. No force required and no need to obsess over some imagined known group of the better, smarter and wiser.

Methinks writes:

A husband is ontologically superior to his wife and he should rule her politically.

Oh, you poor thing! Women don't obey, they manipulate and have you think they're obeying.

Oh, men....

Vipul Naik writes:

I think this is a good first pass for sufficient conditions for Donald's statement to be true, but I imagine that Donald had in mind a somewhat different set of conditions. Namely:

1. The superior are epistemically superior.

2. The superior may or may not be too concerned about the inferior, but don't harbor either unconditional love or active malice toward them, at least not to a great extent (i.e., they aren't willing to undergo great sacrifices to either help or hurt the inferior).

3. The superior's actions are mostly to protect their own self-interest, not to help the inferior. But these actions have positive side-effects on the inferior.

For instance, the "superior" may provide better law and order, better roads, and more stable conditions for markets to operate, largely for their own self-interest. But these "public goods" are then also enjoyed by the inferior -- safer streets, more road connectivity, better conditions for markets to operate etc. In other words, I think that Donald's argument would rely on public goods, spillovers, and positive externalities, rather than a conscious attempt to help the inferior.

Just to be clear, I'm trying to formulate what I understand Donald's argument to be. I'm not endorsing it.

Matt C writes:

If we're going to have rulers at all, it's reasonable to say the best should rule. Alas, agreeing on who is considered best and therefore gets to rule, not so easy.

It is a little strange to me that a subset of libertarians, or at least former libertarians, are attracted to the idea of "strong" autocratic rule. There always seems to be an assumption that the new autocrats will enact "good" policies that will get the decadent lazy slobs in check, establish public order, and generally govern with an intelligent, long term, and general welfare maximizing perspective. But why on earth would you expect this? It is certainly not the typical result with autocrats now.

Trespassers W writes:

In reading this post, the following quote immediately came to mind, which I'm sure Bryan knows:

A leash is only a rope with a noose at both ends.

BZ writes:

OK, so a high IQ/public spirited selectorate is better than the tyranny of an ignorant majority. That's interesting, but I trust no system for the perpetual enfranchisement of the high IQ and public spirited to not soon devolve into something far worse than what was started with.

The public choice guys, Lord Action, and Madison were right: you can't trust men to govern men.

Bedarz Iliaci writes:

I quoted from Aristotle's The Politics:

A man should rule his wife politically, his children monarchic-ally and his slaves despotically.

Tracy W writes:

This misses the dynamic issue. Perhaps A is superior to B in the senses meant. But if B makes decisions for themself, then they will grow and perhaps become superior to A at least at making decisions for B.

Bedarz Iliaci's theory that a society in which women do not obey their husbands and children do not honour their parents is not a lasting society is an interesting one: certainly all the cultures around the world that I can think of are forever changing. Can Iliaci give an example of a culture where people do obey their husbands and honour their parents?

Aaronson writes:

Your actual intellectual superiors (mathematicians, physicists, and the like) seem to pick up their politics casually, directly or not from other campus folk who think they are your intellectual superiors--intellectuals, as described by Paul Johnson or Thomas Sowell.

Both groups think they are at least moderately benign towards everybody else; but aren't.

Trust these "superiors" and Bryan Caplans of tomorrow will have ... trouble.

Remember what Buckley said about the Harvard faculty and the Boston phone book. Alas, he was very likely right.

Comments for this entry have been closed
Return to top