Positive Liberty

One of those longish, shaggy-dog posts of the type that I’d write back when I got a dozen or so readers a day. At your own risk.

Dear ———,

First I must thank you. The article you recommended by Eliezer Yudkowsky was perhaps the most interesting and thought-provoking thing that I have been sent in a very long time. I read it during my commutes to and from work over the last two days, and I’m still trying to take it all in. Thank you so much!

I understand now why you would recommend it: The research into cognitive biases cited in this paper seems to be a direct challenge to the presentation I gave the interns last week — the presentation I made on logic, logical fallacies, and critical thinking. The core contention of the article seems to be as follows:

If I am selective about which arguments I inspect for errors, or even how hard I inspect for errors, then every new rule of rationality I learn, every new logical flaw I know how to detect, makes me that much stupider. Intelligence, to be useful, must be used for something other than defeating itself.

You cannot “rationalize” what is not rational to begin with - as if lying were called “truthization”. There is no way to obtain more truth for a proposition by bribery, flattery, or the most passionate argument - you can make more people believe the proposition, but you cannot make it more true. To improve the truth of our beliefs we must change our beliefs. Not every change is an improvement, but every improvement is necessarily a change.

The author is surely correct that the forms of logic are no substitute for fearless introspection, probity, and the courage to admit when one is wrong. I agree entirely. Yet I must insist that knowing the proper logical forms, and the typical deviations from them, is still important.

Imagine two alternatives:

a) Knowing something about logic, while having strongly held opinions.
b) Knowing nothing about logic, while having strongly held opinions.

It is hard to accept that b) is in fact the better alternative. For one thing, it leaves one defenseless against new and unrelated bad ideas, and these may appear at any time, about subjects for which you don’t yet have strongly held opinions. For another, even if one had the courage to admit when one was wrong, logical thinking and deduction from evidence is the only true way out of a bad idea: If I had the courage to admit when I was wrong, but not the logical acumen to think things through to the truth, then I would simply hop from one bad idea to the next, as often as my courage permitted it. This certainly isn’t a good way to think either.

Does the mere knowledge of proper forms of logic and argument cause one to hold one’s opinions more strongly, and therefore to wander into biased thinking? Perhaps. But I note that this was not a question that any of the research in Yudkowsky’s paper actually addressed.

Being politically knowledgeable — which the research did address — is not the same as being accustomed to the rigorous application of logic to evidence. The former appears to impart a bias, and this conclusion only makes sense to me: As one becomes more invested in “knowing a lot,” it becomes more and more important never to be proven wrong, and this will cause a well-informed commentator to be biased toward the opinions he already holds. But there is no research cited in the paper about logic per se. (Still more remote is the question of character, which I don’t believe is easily reduced to psychometrics. The surest path, and one that is well-evidenced besides, is to assume that all people tend strongly to be phonies. Ourselves included. We must struggle against our natures as vigorously as possible.)

The work on anchoring, scope insensitivity, and the influence of fictional narratives on decisionmaking, all suggest that great deal of public opinion is bunk, and that it should be no surprise that so much public policy is shortsighted, panicky, and disproportionate. This research fits well with the findings of the Public Choice School, which has exposed many such errors in public policy over the years.

One may be tempted to generalize these conclusions — people are both phonies and fools — and to argue therefrom that markets are irrational and perhaps even useless. Some reasoning along these lines will surely be correct, and much market behavior is truly irrational. But note that while actors in a market suffer from cognitive biases, state actors are by no means immune. They suffer too.

Here is the difference, as far as I can tell: Whenever a market actor behaves irrationally, his mistake creates an opportunity for arbitrage, entrepreneurship, and more efficient competition. Scope neglect in estimating the size of a potential market, for example, will allow rival firms to set up operations. A market actor’s error is also limited, in a sense, by the extent of his personal property — or at least it should be, in a well-functioning legal system. The same cannot be said of state actors. While they show all the same biases as market actors, the phonies and fools who call themselves the state also have guns with which to answer the competition, and they are not well limited by property rights.

All of this actually sits very well with Hayekian social thought, which takes for granted that everyone swims in a vast sea of ignorance, and that there are very few things about society or even our own lives that we truly comprehend. Leonard Read’s essay “I, Pencil” is the key text for grasping this concept, and if you have not read it, I recommend that you do.

When I am being honest with myself, I accept that I know vanishingly few things. I know the few books I have read, I know the rules of spelling and grammar to some extent, and I have a vague intuition about what might be interesting material for people to read or think about, at least if they happen to be libertarians. Oh, and I can cook a decent omelet. That’s pretty much it.

I could not, if left to my own devices, clothe myself, find good drinking water, grow enough food to eat, or build anything more than a crude hut to live in. (Given some of the real-life attempts I’ve seen at the latter, I’m guessing that “crude huts” are indeed rather sophisticated affairs.) I mailed a dozen Federal Express packages yesterday to destinations around the world, and I haven’t the faintest idea how they will travel there. I write these lines, and yet I can’t really explain how they will reach your computer, just five floors away.

Come to think of it, I can’t even cook a decent omelet: I haven’t the faintest idea how to raise chickens, nor how to fashion a skillet, nor how to make a gas range, and I can neither find nor extract natural gas from the ground.

Everything I need or am fond of — from antibiotics when I’m sick, to a fine gin martini on a Friday night — is in a sense perfectly mysterious to me. The same is true for everyone: The technicians who manufacture antibiotics, as a rule, will not be able to tell you whether social sciences research is new, interesting, or well considered. They leave that, wisely or not, to people like me. If only the social scientists were more modest about telling all the others how to run their affairs!

Such helpless creatures can only survive and flourish because of specialization, division of labor, and trade. When these things vanish, people die by the millions: Carpenters who can’t grow corn—farmers who can’t fix a broken transmission—mechanics who can’t bake good bread—bakers who can’t build houses. This is perhaps the key lesson of modern life: The two things that tie all of these people together, the two things that let all of them live in comfort beyond what their talents alone deserve, are markets and money. Disturb these, and all are left isolated, with only their own meager talents to go on.

Now I’m not sure that our author would agree, but I’m going to suggest the following to you: Our ignorances, and the cognitive biases that fill the voids, are profound reasons for adopting laissez faire as the default public policy. We are ignorant, but someone, somewhere may possibly know what to do. Let him.

Speaking of eggs and omelets, while reading that fascinating paper, I tried to perform a quick estimate of the annual egg production in U.S. — as you know, it’s one of the statistics the author says people are notoriously bad at guessing. I thought it would be fun to try to test my own cognitive biases. Here is my work:

My household consists of three adults. We buy about one dozen eggs per week. Owing to food allergies, I cook nearly all of our own food from scratch. We eat very little prepared food, so I disregard the number of eggs that go to prepared food products. I assume that our menus are typical of Americans everywhere, which may not be the case, but it’s what I have to work with. I also assume that no fresh eggs are ever exported, since shipping eggs overseas probably isn’t viable business model. (Or is it? How many processed eggs are exported? Probably quite a few. But I have no idea how many, or in what forms, or who might buy them, or in what amounts. Uh-oh.)

But anyway, assuming one dozen eggs per three people per week, with a population of about 300 million in the United States, we get 1,200 million eggs per week. Multiplying this by 52, we get 62.4 billion eggs per year. That’s my estimate.

Asked to assign a confidence interval, as the exercise also demands, I would do as follows: The paper warns us that we all set confidence intervals too narrowly, so I will aim for a wide figure. (It also warns us that, when people are warned of the universal tendency to assign too-narrow confidence intervals, they inevitably go right ahead and do it anyway.) How I will reach a wide figure, I’m not sure, but I’m going to try.

First, I will assume on the low end that many people don’t eat any eggs at all, and that my household’s appetites are just that rabelasian. I’ll guess only 5 billion at the lower bound. I will assume on the high end that many, many eggs are processed and exported — say, a third of the total, even though the whole business seems unlikely to me. This would get me around 95 billion.

Now let’s stop and think about this for a moment. The range between these two putative 95% confidence intervals is 90 billion eggs. Ninety billion. The average egg weighs 58 grams and has a volume of 53 mL. 90 billion eggs yields 335,456.5 liquid barrels of egg. Not really sure how much that is? Me neither, to be honest. Now let’s look up some statistics see how I’ve done…

As it turns out, in 2004, there were 89 billion eggs produced in the United States; this is the latest year for which I have data. Given the growth of the industry in previous years, I’ve probably missed my 95% confidence interval. But – I don’t even know that for sure!

All the best,

Jason

Trackback URI |