Greedy Algorithms and Politics

I’ve been a Libertarian pretty much since I’ve been able to vote. I’m for small government, free markets, and just about every personal/social liberty that doesn’t infringe on my own. However, I don’t kid myself that free market capitalism is without flaws.

At the core of free market capitalism is the idea that the market, with individuals and corporations all acting in their own best interests, will find far more optimal solutions to problems than any type of grand oversight. Furthermore, attempts to interfere with highly complex markets will usually have unintended or unforeseen side effects. While I believe this generalization to be largely true, I am reminded of a certain type of algorithm from Computer Science: The Greedy Algorithm.

Greedy algorithms are generally easy to understand, easy to parallelize, and (relatively) quick to execute. They might not always find the optimal solution to a given problem, but for many classes of problems they are the easiest way to find a close approximation. Tasks well-suited to greedy algorithms are those with an “optimal substructure.” By optimal substructure, I mean tasks that can be broken down into subcomponents where the best solutions for each subcomponent can be combined to produce the best solution for the whole. In many ways, capitalism has an optimal substructure: At each level, the incentive of wealth can drive people to create more of it. This, in turn, makes for a wealthier society, which provides more opportunity for others to create wealth. It’s a feedback loop which results in wealthier people, wealthier corporations, and wealthier countries.

However, there are problems for which a greedy algorithm is not a good fit. Consider a chess game: the capture of a queen might locally (for this move) seem like the best choice, but it could result in ultimately loosing the game. Or, consider the differences between the Elevator algorithm and the Shortest seek first algorithm. If the elevators in our buildings used shortest seek first (a greedy algorithm), then the average response time overall might very well go down, yet it would be at the expense of ignoring some floors for minutes or even hours during busy times.

If we can admit that there are problems with greedy algorithms in the computer science world, then it’s not a stretch to say there are problems with free market capitalism in the “real” world. The question is, what is the alternative?

For many, the obvious answer is government. However, government is a double-edged sword. For as much power as a government has to do good, it also has the same power to do bad. As Gerald Ford once said, “A government big enough to give you everything you want is a government big enough to take from you everything you have.” The founding fathers of the United States were well aware of this, which is why they tried to limit government. However, even if you could assume a benevolent government, you still have the problem of efficiency. As we’ve seen repeatedly, lack of competition within government makes for sloth and waste.

So, in closing, I’m a Libertarian because given the choice of government problems versus free market problems, I choose the free market. It’s not that I think free market capitalism is perfect, or even fair. It leaves a lot to be desired. However, I find that keeping the market in check with an even larger government is a much scarier proposition.

Perhaps, given this new era of technology and communication a new type of societal structure could emerge that addresses both the short-sightedness of markets as well as the sloth and size of government. Over the next few weeks I plan to speculate on the qualities such a society might have. Stay tuned!