Incentives: Defining and classifying incentives

My economics professors are typically surprised that I call myself a communist, especially since I tend to do very well in their classes learning capitalist economics. They really shouldn't be that surprised; Marx and Lenin make the case that the problems of capitalism are inherent to the theoretical structure of capitalism itself, and nothing in three years of undergraduate education — and by the end of this semester, I will have completed all the courses in basic capitalist theory — have led me even to suspect Marx and Lenin might be wrong, at least about the nature of capitalism.

Yesterday, my professor discovered, to his astonishment, that I really was a communist. He's cool with that (all of my professors, even the Libertarians and Republicans, have been cool with my communism), but he wanted me to think about one concept: incentives. He didn't make any argument explicitly, but I think the underlying argument is that capitalism has a complex and powerful set of institutional and systemic incentives, and these incentives have been the primary social cause of the massive industrialization and overall increase in the material standard of living that capitalism has achieved over the last 200-300 years. If we do away with those systemic incentives, we risk backsliding into the inertia, complacency, and inefficiency of the pre-capitalist era. Is that what I really want?

Marx and Lenin have, I think, adequately covered the problems with capitalism's systemic incentives. Briefly, by rewarding the accumulation of money and capital, capitalism requires ad hoc incentives (mostly government stimulus) to increase consumer spending when the economy has accumulated excess capital relative to aggregate demand. Ad hoc incentives can happen, but they are weaker, temporary, and don't have the self-reinforcing character of systemic incentives. The welfare state is an attempt to create systemic incentives (mostly fiat money, social transfers, and continuous government deficits) but the capitalist class will never give up its implacable opposition to the welfare state. They have been making gains against the welfare state for forty years, until it lies in tatters today. But that's not the issue: as bad as capitalism might be, communism could be worse.

But how do incentives actually work?

An incentive is a social institution that encourages some or all citizens to do what they would not do without the incentive (or discourages some or all to not do what they would otherwise do).

As I see it, there are five kinds of incentives. (Keep in mind that I am more-or-less arbitrarily defining labels, and I find discussions about whether or not I use the "correct" labels or "correct" definitions to be tedious. If you want to criticize arbitrary definitions, it's interesting only if you critique them at a more sophisticated level.)

First, there are psychological incentives, which individuals use to reconcile conflicts within their own minds. For example, some people generally have an internal psychological conflict between their desire to lose weight and their desire to eat. Some find it helpful to create psychological incentives, such as the social solidarity offered by organizations such as Weight Watchers, to reconcile these conflicting psychological desires. Although we can and do use many psychological incentives in a political sense, such as raising excise taxes on tobacco to reduce smoking, I don't see these kinds of incentives as being of primary interest in talking about large scale political-economic institutions such as capitalism and communism.

Second, there are natural incentives. We do not need social incentives to encourage people to do what they usually want to do. We do not, for example, need to incentivize people to eat food or come in out of the rain. People will do these things on their own. These incentives seem trivial precisely because we don't need them, but they have some meta-ethical significance, which I'll discuss later.

Third, there are coercive incentives. If, for some reason, a dominant minority wanted to get the majority to do something the majority had no natural desire to do. We would have to have some sort of social incentive if, for example, we wanted people to cut off their left pinky fingers. These incentives also seem trivial, because why would we want to encourage people to do what they just don't want to do? But these definitions also have meta-ethical significance.

Fourth, there are negotiated incentives. If I want something from you, I have to offer you something you want more than you want what I want. For example, if I want you to give me a bicycle, I have to give you a case of beer, which you want more than you want the bicycle. Since trade can make both parties better off, and since people by definition naturally want to be better off, people will naturally want to trade; there's a natural incentive at the abstract level to offer negotiated incentives at the concrete level. Since there's an abstract natural incentive to trade, we don't need to do much at the institutional level to support the idea of trading.

Fifth and finally, there are game-theoretic incentives. When there is a natural incentive to do something, but doing it works only if (most) everyone does it, then we need to create game-theoretic incentives so that people can trust others to do it. The concept of the promise is the most obvious game-theoretic incentive. When we can use promises as negotiated incentives, we can achieve a lot of long-term benefits that are extremely difficult, and perhaps impossible, without promises. However, promises work only if (most) everyone keeps her promises, and there are short-term natural incentives to break promises. If I promise to give you a case of beer tomorrow in exchange for a bicycle today, we get a benefit — i.e. it is more efficient for me to use the bicycle today than for you to use it — that we couldn't get without the use of promises. However, once I have the bicycle, I have a short-term natural incentive to renege on the promise of beer tomorrow. If even a few people don't keep their promises, the whole edifice of using promises collapses, and we lose the benefits of using promises.

So when we create game-theoretic incentives to keep our promises (such as the possibility that men with guns will come to my door, take possession of the promised case of beer, and give it to my neighbor who gave me the bicycle) we are not saying we need to be forced to keep our promises. We know rationally that keeping our promises has rational benefit. Instead, we are saying by adopting game-theoretic incentives that we are committed to keeping our promises, that we are eliminating the possibility of free-riding on the system; in other words, we use game-theoretic incentives to establish trust. And we require political-economic institutions to establish precisely these kinds of game-theoretic incentives.