Monday, October 01, 2012

Traffic, Economics, and Constructal Law

There are a large number of different kinds of self-organizing processes, from chemical to biological to neurological to social. As we learn from the constructal law, these processes are fundamentally flows. And the flows create patterns. And the patterns affect the flow.

In each of these kinds of self-organizing processes, we see coordination of behavior. From this emerges patterns. In social systems, incentives matter. To understand the behaviors of a system, understand the incentives involved.

Traffic is a self-organizing process. A variety of orders emerge in traffic, ranging from steady flow to utter standstills. Traffic jams are a kind of order which emerges -- though most people would certainly classify it as an unwanted, even a "perverse" order. But how do traffic jams emerge?

Let us leave aside for a moment accidents, which will certainly slow traffic. They are hardly the only source. How many times have you been in heavy traffic and wondered what on earth was causing the slowdown, only to find traffic suddenly open up so that you are driving along as a more reasonable rate of speed -- only there appeared to be no reason for either the slowdown or the speedup?

The main cause of such slowdowns is the fact that when we are driving, benefits are concentrated while costs are dispersed. Worse, costs are dispersed behind us, where it won't affect us.

For example, if we find we are in a turnoff lane and we don't want to turn off, most of us will try to switch lanes, even if it means causing others to hit the brakes. In heavy-enough traffic, one person hitting the brakes will result in others hitting their brakes. This results in a chain-reaction slowdown that may only dissipate after a half mile or more. And this is assuming that nobody changes lanes, keeping the slowdown in one lane. Naturally, there will be those who want to continue going at the rate of speed they are going at and will therefore switch lanes. They gain, but at the expense of another lane slowing down. Thus there can be a chain reaction both down and across lanes. And all because people are responding to what will benefit us without considering the costs to others. And, like I said, those others are behind us, so we are typically left unaware of what we have done. And since we don't know any of those people, we probably don't care all that much, anyway. Further, any cost we may incur on ourselves is going to be much less than the benefit we gain by switching lanes.

Thus, the incentives overall are for us to switch lanes to our benefit. However, if none of us did so, traffic jams would be very rare. One does not see too many traffic jams on carpool lanes precisely because one cannot switch lanes on them -- they are one lane, and they are difficult to get on and off of. As a result, traffic moves much faster.

All of this is merely an analysis of what happens. Knowing this, I am not going to stop switching lanes, because since nobody else is going to stop doing so, I'm not going to make any difference on things other than to slow myself down getting to where I am going. And even if most people did stop switching lanes except when absolutely necessary to enter or exit, there would be a number of free riders taking advantage of everyone else following the rule, thus causing traffic to slow for the rule-followers. And since enforcement of any lane-switching rule would actually slow traffic more (not to mention be rather arbitrary, as it would only be enforced if an officer is around, and then the officer would have to read your mind to understand your intention in switching lanes, which is impossible, making the law impossible to enforce) from stops and from drivers changing their mind if they think they see a police officer, legislation would only make things worse. Other than trying to design roads that take human behavior and constructal law into conscious consideration, there is not much one can do.

And even with such conscious design, there will still be people making mistakes, slow drivers, fast drivers, etc. According to constructal law, differences in flow speed cause structures to emerge. People driving slower than the flow of traffic or faster than the flow of traffic, causing people to switch lanes and hit their brakes. But again, there is nothing that can be done about differences in speed. There simply cannot be a law stating that people have to drive exactly 60 miles per hour, no faster, no slower. People do have to enter and exit, etc. -- aside from the practical problems of enforcement. And if there is a traffic slowdown, are you going to ticket everyone?

Traffic acts as a good way of understanding incentive structures. There is no way of getting around the fact that benefits are concentrated on you while the costs are distributed onto a large number of drivers, most of whom are behind you. Certainly one could just deal with the fact that you made a mistake about what lane you should be in and exit and turn around or re-enter traffic further down. But how many people are realistically going to do that to the benefit of unknown strangers? If you haven't done so, you're one of those people who wouldn't. That's because the cost would be concentrated on you, while the benefits would be distributed to unknown strangers.

Politics works the same way as traffic. Benefits are concentrated, while costs are distributed. If the government gave me $300 million, that would benefit me greatly. But it would only cost $1 per person in the United States. It's not a big deal for you to give up a dollar, but it's a very big deal for me to receive $300 million. And this is the argument we repeatedly hear from politicians arguing for their pet projects, subsidies, etc. "Why, this program only costs ten million dollars." And when you break it down per taxpayer, it's not that much. But ten million here, ten million there, and soon you're talking about real money. This particular driver not switching lanes isn't going to make much of a difference on traffic, though his switching lanes may benefit him. One would have to get everyone to change their behaviors to get a change in traffic behavior. And that would mean a structural change in incentives.

While political sytems act like traffic, economic systems act the opposite of traffic. In a free market economy, costs are concentrated and benefits are dispersed. Any business owner who makes a mistake loses their business, but successful businesses distribute their benefits out to customers with the increase in goods or services, including increases in quality, etc. from competition. A free market would be like a traffic system in which somehow drivers who did things that would slow traffic were removed from traffic and made to start over again, thus making the flow of traffic improve for everyone else. How quickly would people learn what they needed to do to keep traffic flowing well? Yet, because of the incentive structure of free market economies, this is precisely what happens in the economy.

Of course, many prefer the traffic/political model because the concentrated benefits are easy to see, while the distributed costs are difficult to see in these models. On the contrary, in economic processes the distributed benefits are difficult to see, while the concentrated costs are easy to see. It is easy to see the benefits of the government stepping in to rescue General Motors; it is difficult to see the distributed benefits of letting it die off. It takes considerable amount of understanding of economics to understand the latter, so we should not be surprised people don't see it that easily. Like the lane-switching driver, the benefits are immediate and clear; the costs lay far behind among the unknown many. Worse, even if you do understand this, there's not a lot of incentive to do anything other than to continue to switch lanes to your own, concentrated, benefit.
Post a Comment