legionnaire

Before I get started, everyone go over to Aurelius's journal and say hello. He's a friend of mine from off the site who's actually been on SG longer than I have but has just recently decided to become active here. So go. Now.

I have something that's been bothering me for some time now that I think I'm finally able to articulate. It came up in the most bizarre (for me) of circumstances - thinking about car insurance. The reason why that's so strange is that I'm one of the lucky people who live in a place where I don't need (or want) to have a car, so I don't. Hence, no insurance. But for some reason I was thinking about it (my mind jumps around quite a bit sometimes) and I realized what it is that I hate - it's something that I consider to be a fundamental misapplication of logic, and as most of you are probably already aware, I'm a big fan of applying logic to basically everything. It makes me happy in a way that few other things can. This logical misapplication is the application of principles that were designed to describe group dynamics to something that is individual.

I'll elaborate a little because this is really the crux of the issue. Imagine looking at a very large herd of cattle, maybe ten thousand of them, from an airplane flying at ten thousand feet. It would appear to be a mass of indistinguishable objects basically heading in a single direction. However, if the airplanes altitude begins to lower, the individual cows begin to resolve and their movements can now be tracked singularly. At that level you realize that while much of the herd is moving together and that their group motion tends in a single direction, at any given instant in time the individual motion of a single cow cannot be predicted by the group as a whole. The cows are moving at right angles to each other, some are moving backwards, and some in the same direction as much of the group. But you would never be able to single out that one cow and say "I know with certainty that you will now move in this direction." It's not possible based on the knowledge that you have, because that knowledge is derived solely from your study of the group as a whole, not of any of the individual cows. Based on your observations of the group as a whole you might be able to determine a statistical probability of the likelihood that the cow would then move in a specific direction, but could not say for certain what it would do at any given time. In addition, unless you employed an additional logical fallacy, the Monte Carlo fallacy, you would not be able to accurately predict the behavior of the animal even given knowledge of its prior behavior. That is, unless you bothered to try and understand the animal's individual motivations and desires and tried to accurately develop a psychological model for the way that animal thinks. But if that were the case then you would be treating the animal as an individual, thus making the point that group behavioral dynamics cannot be successfully applied to individuals; a different set of rules and level of understanding are required for that.

This may seem rather esoteric, but it has real world consequences. As human beings most of us love to categorize things, subconsciously and consciously. Each categorization comes with its own series of excpectations that we believe will influence the way that person or object will behave. Take this for example. Imagine that I'm telling you a met a new friend today. Being inquisitive (I refuse to associate with people who aren't) you ask a series of questions about the person. Is it a man or a woman? How old is he or she? Tall or short? Black, white, latino, asian? High school diploma? College diploma? Advanced degree? Do they dress well? Are they attractive? These types of questions serve to build our own model of the way we believe that person to act based on generalizations that we associate with each of these categorizations. As a result, before you've even seen this person you now have what you consider to be reasonable expectations of what this person is like based on the set of criteria I've provided you with, combined with your own expectations that you associate with each of those criteria. But in reality those expectations could be (and often are) completely wrong. It should be a sign that this isn't always the best way to try and understand things that are new and different - but many people don't learn.

I believe this logical fallacy is one of the problems underlying many of society's ills including racism, classism, sexism and other faults of generalization. The problem is magnified exponentially when we develop culturally shared ideas about the sort of characteristics that certain groups share because it lends the notion an air of credibility. I may have a theory that everyone with curly hair likes the color yellow, but no one will give it any creedence because it's not part of our cultural lexicon. After putting forth this theory to a number of people and having it roundly rejected (and laughed it) it's likely that notion will die in my head as I start to realize the futility in trying to classify people based on their hair type. But take another statement that has equal (that is to say, no) merit, "Women are bad drivers." Some people will likely respond to this in anger and assert they know plenty of women who are perfectly good drivers. But others, even if it's a minority of the population, will agree. This is an idea that has managed to infiltrate our culture. That very fact lends it some measure of validity to people. "If so many people believe it then there must be some measure of truth to the idea." Sounds right, doesn't it? Maybe not - but we're all guilty of it to some degree. Women may be worse drivers than men as a whole (they're not, but for the sake of argument let's say they are) but even knowing that, say 51% of all accidents are found to be the fault of a woman is a statement that refers to women in general - a rather large group. If I meet an individual woman without knowing anything about her, does that mean that she's automatically a bad driver because she's a a member of that group?

This brings us back to auto insurance. Insurers use actuarial tables, which are basically massive compilations of statistical data, as a way to calculate premium rates. Using some of a (somewhat small) sampling of existing criteria including your age, your education, your location, the type of car you drive, your previous driving record and others they have a formula that calculates how much you should pay. In essence, they're saying that they are able to successfully predict your own behavior and likelihood of having a car accident based on the behavior of other, similar people. The real problem is this - this approach works for them because they don't deal with individuals. The insurer has contracts with many, many individuals which can then be treated as a group and statistically are likely to follow group dynamics,, so their model is accurate for the insurer's purposes. The problem is that people are not groups. They're individuals. Because of your age, race, sex, and location you are required to shoulder a larger or smaller burden for the compny because other people who share those few characteristics with you act, as a large group, more or less responsible than other groups by comparison. It has nothing to do with any of your individual characteristics. You could be the best driver in the world and never get into an accident, but because you have been categorized as a member of this group, misapplied logic dictates that your actions can be accurately predicted, which is entirely untrue.

Paying car insurance sucks but it's not ruining anyone's lives. But imagine how this scenario plays out in other arenas, like the judicial system, for example. It's no secret that in the US young black men are incarcerated at a significantly higher rate than their peers of other ethnicities, and receive steeper punishments for the same crime. By misusing group dynamics and applying them to individuals, judges are forcing people to suffer based on the erroneous expectation that their behavior can be predicted based on a few individual criteria and the person, along with all of his complexities, can be ignored because he has now been categorized.

Just like the cows in the example above, a person's behaviors can never be accurately predicted from instant to instant just because one is aware of how similar individuals have acted. Only a detailed understanding of the individual will be able to aid in this sort of prediction, and this takes time and effort. It's easier to just label people and forget about them. It's also illogical.