Fixing a hole
an illustration of the Collective Self Interest Fallacy:
the common but erroneous application of the concept of "rational self interest" to a multitude of individuals, without regard for the independence of those individuals.

Problem: the roads of Happyville are full of potholes

The town of Happyville, Montana just isn't living up to its name. The roads have deteriorated so much that residents driving on them are constantly damaging their cars from hitting cracks and potholes, which ends up costing them a lot of money. Although drivers in the town complain about the roads constantly, the roads just never seem to get fixed.

The mayor, Jim Miller, promises that he will find a solution to the roads. At a town meeting, he explains the situation. First, the good news: keeping the roads in excellent condition would cost $10,000 a year, which comes to only $10 per driver, far less than they tend to pay for constantly having to get their cars realigned and their bent rims and flat tires fixed. Even for $5 a year for each driver, the roads would be pretty good, with the worst of the potholes fixed.

The bad news, though, is that the town charter has some unusual rules, going back to the town's quirky founders. According to the charter, road repairs can't be paid for with tax money, since that would be unfair to those who don't own cars. Collecting money when people renew their vehicle registration would be acceptable, but only if the amount collected is decided by the people themselves, rather than forced upon them by the town government.

Attempted solution: let drivers chip in toward cost of maintaining roads

So Mayor Miller suggests that they do just that: let the drivers decide how much they want to chip in toward road fixing. When they register their cars, they will have the option of paying an amount of money toward fixing the roads, every dollar of which will go into road maintenance. Let them vote with their wallets as to how much money to put into fixing roads, he says. Susan Parker raises her hand. "Ok, but can you tell me why it's in my interest to put in money?", she asks. "Why wouldn't I just keep my money and let others foot the bill?"

The exasperated mayor explains, "Well, think about it....what if everyone decides to do that? The roads will stay in the sorry shape they are in. Is that what you want, Susan?" Others at the meeting look at Ms. Parker disapprovingly. "The point is, everyone benefits if everyone chips in. What could be simpler to understand than that?" Most of the attendees clap, and the measure passes easily.

The next year though, the roads are not much better, and the townspeople are still unhappy. The mayor throws it right back at them: "Well, you only have yourself to blame. We only collected, on average, 78 cents per driver. This proves that the town would really rather keep its money than spend it on fixing the roads." He fixes his eyes on Susan Parker. "If you really want the roads fixed, and you aren't chipping in your share, well, it seems to me you are being hypocritical".

But the same thing happens year after year, and the roads never get fixed. And most people just assume that Mayor Miller must have a point -- people must not want to fix the roads all that much, otherwise they would chip in more money. Susan Parker just doesn't buy into his logic though, and she continues to be a thorn in his side.

Better solution: require a road maintenance fee, but let the drivers set it by vote

How to vote for a number

Having a vote for a numerical value, such as the amount of a fee, seems like it might be complicated, but it isn't really. Each person simply writes down the number that they prefer, and then the median value is selected.

Why the median and not the average? Say I guess that others will vote for a $7 fee, but I would prefer it be closer to $10. If the average is used, I might be smart to vote for, say, $100, rather than my actual preferred amount of $10, in hopes of it swinging it further my way. That is, I can easily "game the system." Median, on the other hand, does not reward such insincere voting.
Eventually, Ms. Parker convinces enough of the townspeople that Mayor Miller is approaching the problem the wrong way, and she is elected mayor. At her first meeting as mayor, she suggests a new twist on Mr. Miller's plan. All drivers will pay the same amount toward road maintenance, as a required fee when registering their cars. But the amount of that fee will be voted upon by all drivers (see sidebar). Mayor Parker says she has run it by the lawyers, and they assure her that this plan still seems to be within the words and the spirit of the town charter.

The former mayor Miller scoffs at the new plan. "The people have already had a chance to say exactly how much they want to pay," he says, "and it's less than a dollar a year. As long as the people get to make the decision, the town won't collect enough money. Why should this new plan change that?" Needless to say, he is surprised when the vote results in collecting $8 a year per driver, enough to fix the roads quite well, if not to absolute perfection. "Well clearly the people in this town are not rational," he says.

Why one works so much better than the other

Why, then, the difference? What is wrong with the first mayor's logic? In both cases, it was the townspeople's choice as to how much to spend. Why is it that in the first case they "voted" for 78 cents per driver, while in the second case they voted for $8? As it turns out, in both cases the townspeople were behaving quite rationally.

Mayor Miller did not account for the tendency for people to be motivated to independently pursue their own self interest, a central premise of market economics and game theory. In fact, a game theorist would see this as a failure on Mayor Miller's part to understand and account for the Prisoners' Dilemma. In the first scenario, it would simply not be rational for a self-interested driver to voluntarily chip in money. The fact that everyone would be better off if everyone chipped in is irrelevant. What matters is whether an individual would be better off if that individual chipped in, regardless of what others do.

General form of the fallacy:

"Everyone benefits if everyone does X. Therefore, it is in everyone's self interest to do X."
For instance, if you chipped in $10, how much would your decision improve the roads, versus chipping in nothing? Would the increase in road quality from your $10 be worth the cost to you? Since there are a thousand or so drivers in town, the answer is almost certainly no. Your decision would cause you to pay an additional $10, which you could have spent on something tangible -- possibly a bottle of wine or a rib eye steak (or, of course, it could go toward repairing damage to your car!) But the amount your $10 affects the roads would be comparatively negligible to you. Most likely you wouldn't even notice a difference. The $10 cost from your decision is borne by you alone, while the benefit of 10 dollars' worth of better roads is divided equally among all thousand other car drivers, giving each a penny or two of benefit. This would be the case regardless how much others chipped in. In other words, if you decided to put in money, it would be an altruistic act, not a self interested one.

No longer a fallacy:

"Everyone benefits if everyone does X. Therefore, it is in everyone's self interest to form an enforced agreement with one another that each will do X if everyone else does X."
On the other hand, it is entirely rational (and non-altruistic) for you to vote to set the fee to $10. Now the cost of your decision is shared equally -- if it raises or lowers the fee, it does so for everyone, not just you. This means that if you choose to vote for $10 rather that $0, it might result in raising the fee about a penny, say, from $8.04 to $8.05. You pay a tiny amount more than you might have otherwise. The additional benefit you receive due to your decision is tiny as well, but it is in line with the additional one cent you paid as a result of your decision. The percentage you caused your own payment to increase (around a tenth of a percent) is met with an equal percentage increase in road spending.

To many, this might seem like painfully obvious common sense. It's certainly easy to see that something is wrong with the first mayor's logic when he said "if you want the roads fixed, make a voluntary contribution toward it." No one wants to be the sucker, and many are tempted to be freeloaders. This is doubtfully particularly controversial -- after all, Happyville, Montana is a made-up place, and no real world town would be foolish enough to expect a system based on selfless altruism to be effective for collecting money for municipal maintenance. But the point of this article is not to provide a solution to bad roads. Instead, it is to provide insight into other sorts of similar issues and situations, some of which may indeed seem controversial at first blush.

Other places where the fallacy shows up

In many other scenarios, people use identically flawed logic, often in situations that carry more emotional resonance than the mundane issue of road maintenance. How many times has someone appealed to your self interest to encourage you to do something, on the premise that if no one did it, a bad result would happen to you? For instance, "you should donate money to public television, because if no one made donations, the shows you enjoy watching would cease to exist." The flaw, of course, is that you can only make the decision for yourself, not for the group as a whole.

Below are a few more examples of things that well-meaning (but possibly naive) people are likely to say:

Now, cClick to magically correctrestore the above sentences to their erroneous versions.

Each of these expect people to make decisions that come at a cost borne by the decision maker alone, while the benefits of those decisions are divided among many. For instance, if you boycott a company's product that you otherwise would have liked to purchase, this has a tangible negative effect on you. But the likelihood of your participation in the boycott resulting in tanglible positive effects for you is negligible. Whether someone else's participation in the boycott provides benefits to you, or whether your participation provides benefits to others, is irrelevant to any self-interested decision to participate or not.

To put it in everyday terms, if we ask or expect people behave inconsistently with human nature, we tend to be frustrated and disappointed.

To put it in technical terms, a person should be expected to make a decision, if and only if the expected increase of reward provided to the decision maker, resulting from that individual's decision alone and regardless of what choices others make, outweighs the cost to that individual alone of making that decision.

Let me be clear, I am not in any way suggesting that people should not behave ethically, or that they shouldn't consider and be motivated by how their individual actions affect the community -- or the world -- at large. Nor am I suggesting that individuals are incapable of behaving altruistically. Don't look for an Ayn Randian "virtue of selfishness" argument from me. I firmly believe that selfishness is the exact opposite of virtue.

What I am suggesting is that altruism should be recognized as altruism, rather than suggesting that altruistic acts are actually self interested. Furthermore, I would argue that altruism alone is a relatively weak incentive, and therefore suggest that we stop wishing and waiting for mass altruism to actually solve important problems. Instead, we should seek out solutions that are more effective. In many of the articles to follow, I will propose specific approaches to addressing various problems, based strictly upon this reasoning. In most cases, these approaches will be applied to harder problems, and therefore will be less simple and obvious than the approach taken by Happyville.

© 2006-2007 Rob Brown