Three cheers for the “warm glow”!

In the last post, I described how the existence of a welfare state in a democracy can be explained based on a simple economic model of human altruism, in which voters have a utility function that includes other people’s utility. By voting for a social safety net, we increase the utility of a lot of disadvantaged people at a relatively low cost to ourselves, because laws ensure that a bunch of other taxpayers also contribute their share. As long as the total increase in everyone else’s utility has enough value to us to offset the decrease in our consumption resulting from the taxes we pay, then we should be in favor of it.

Economists who write about this sort of thing call the social safety net a public good, like roads and a national defense. An ideal public good is something that anyone can enjoy without taking away from other people’s enjoyment of it. We are all more secure when our country is well-defended, and my enjoyment of that security takes nothing away from yours. If you value other people’s utility, then a social safety net gives you the enjoyment of knowing that the most disadvantaged people in your country are taken care of, and that increase in their utility increases your own utility. Your enjoyment of that fact takes nothing away from any other altruists’ enjoyment of that fact, which makes it a public good.

The same economists have also argued that government welfare programs are a more efficient way to help out the disadvantaged than private charitable giving is. Also, altruistically minded people should prefer to contribute to that effort through paying taxes, so they can be confident that all of the other affluent people in their county are also contributing their share. By contrast, a charitable gift is a one-off contribution by a single person, with no assurance that anyone else will follow suit.

Yet people still make private contributions to a great many different charities and cultural organizations. In the US, private giving amounts to about 5% of the country’s GNP—hundreds of billions of dollars every year. Economically it makes no sense, yet people still do it.

One explanation for this behavior is that people not only value the utility of others altruistically, but also derive some benefit from the act of giving itself—a warm glow from the thought of having voluntarily done their part to help others. James Andreoni (1989, 1990) constructed economic models based on this idea, which he called impure altruism. Pure altruism would be motivated solely by an interest in other people’s well-being. The altruism becomes impure to the extent that someone derives some pleasure from the act of giving that is over and above the rational value to them of other people’s utility.

Some people seem to get a kick out of pointing out how actions that appear to be altruistic are actually selfishly motivated. Taken to an extreme, this amounts to psychological egoism—the view that all human actions are ultimately self-motivated or egoistic, even when they appear to be altruistic. In this article in the Internet Encyclopedia of Philosophy, Joshua May notes that philosophers have widely rejected psychological egoism, and discusses various arguments against the idea. But that doesn’t keep amateur philosophers from smugly trumpeting the idea.

Has Andreoni put his finger on a bona fide example of an egoistic motivation for altruistic actions? Maybe, but not necessarily. What his models include in addition to the rational altruistic motivation is merely a preference for the act of individual giving. That could stem from the pleasure of giving itself, but there are other possible explanations. For example, people who give may not accurately evaluate the level of support that a worthy cause gets from government transfers, so they may be convinced that sufficient unmet needs exist to motivate individual altruism on their part.

But let’s say for the sake of argument that some normal human altruistic actions are motivated in part by a warm glow that comes from the act of helping others, over and above the actual benefit to the people being helped. Strictly speaking, that is indeed an egoistic motivation since it is driven by the pleasurable experience of doing good. But is it morally equivalent to typical egoistic motivations like hunger and thirst? The giver does seek a pleasurable experience, but the pleasure in the experience comes from helping another person in need. The motivation may be egoistic, but the behavior it produces is loving and caring.

If you were designing a social animal and you wanted to maximize its cooperative behavior, wouldn’t you program in precisely the sort of emotional response to the act of helping that Andreoni was getting at when he referred to a warm glow? To build an ideal human society, wouldn’t you want to stock it with people who experience as much warm glow pleasure as possible? And if we could discover social structures that bring out more of the warm glow in normal humans, wouldn’t we want to institute them in our own communities?

As long as the warm glow comes from sincere efforts to help other people, who would begrudge the helpers whatever pleasure they may get out of it? After all, that pleasure is encouraging them to behave exactly how we want people to behave. Do they have to do it dispassionately and in answer to high moral principles before we can admire them for their spirit of devotion? Not as far as I’m concerned—their warm glow gives me the warm fuzzies, and I want them to enjoy it to the fullest. Go glow!

© Joel Benington, 2011.

Altruism and the welfare state

This post is really The economics of altruism III, so you may want to check out the last two if you haven’t already.

One phenomenon pondered by economists who study the economics of altruism is the welfare state. In affluent democracies, voters routinely support social safety nets that they help pay for with their tax dollars and that mostly benefit people they have never met. What motivates people to give up their hard-earned money to support total strangers?

A common explanation is a sense of justice. This supposes that we think it’s wrong for some people to have much more than others, even if we are among the people who have more. So we’re in favor of government taxing affluent people more, and using the extra money to help out the disadvantaged.

Another possible explanation is a kind of insurance policy. While we may be doing fine right now, our lives could sometime take a turn for the worse. By voting in favor of a social safety net, our higher taxes now are buying insurance against some future catastrophe in our own lives.

Either or both of these explanations may indeed help explain this curious phenomenon. But can it also be explained simply in terms of other people’s utility factoring into our own utility functions?

Consider how the transactions of a welfare state work. The extra taxes you pay lower your total consumption (x_i from the equation in The economics of altruism). Your contribution is added to those of a great many other taxpayers, all of whom presumably are doing reasonably well or they would be net recipients of the welfare-state payout. The money collected from you and all the other taxpayers goes to people who presumably need it more than the people who are contributing the taxes do.

So your money isn’t a one-off payment that goes to a complete stranger in need. By voting in favor of a welfare state you ensure that everyone else pays too, in proportion to their ability to pay. Your share is just a tiny part of a much greater sum that can then be distributed to the needy, and increase their utility.

In the personal utility function that describes this situation, one factor is your own consumption, which is decreased by some amount depending on how much you are taxed to support the welfare state. The other factor is a set of the utility functions of all of the other people who pay taxes and/or receive benefits. There’s an equation for this utility function, but I won’t bore you with it.

Since the people who pay into the welfare state are all fairly affluent, the loss to their own consumption shouldn’t decrease their utility all that much. Since the people who benefit from the welfare state are more needy, the money that is transferred to them should produce a proportionately greater increase in their utility. The welfare state thus causes a lot of people to experience a small decrease in utility and a lot of people to experience a correspondingly greater increase in utility. Even though the vast majority of them are unknown to you, you know it’s happening because it’s been legislated and will be enforced.

What, then, is your trade-off in deciding whether to cast your vote in favor of a welfare state? One factor in your utility function is your own loss of consumption as a result of the taxes you pay. The other factor is the value to you of the increased utility of millions of complete strangers (the beneficiaries), against the much smaller decreased utility of another group of millions of complete strangers (the other taxpayers). Even if the well-being of each of these complete strangers factors only very weakly into your utility function, the sheer number of the people benefited by the welfare state may be enough to make you feel that your utility is increased by the existence of that welfare state, even though you have to pay some taxes to support it.

This line of reasoning can explain popular support for a welfare state without invoking a sense of justice, or a desire to insure against the future, or any other motivation. It can come down simply to a rational economic calculation by a person who derives some small value from the well-being of others. Given the number of people involved, that value can in fact be quite small and still outweigh the cost that they themselves have to pay in taxes.

I know, I know—there are arguments against a welfare state too. It may create inefficiencies, the higher marginal tax rates may discourage economic activity that would otherwise benefit everyone, it may create perverse incentives for the less well-off. But all that is a conversation for another day. My only point here is that we can explain the existence of a welfare state merely by supposing that voters have utility functions in which other people’s utility is a component.

© Joel Benington, 2011.

Using the economics of altruism

In the last post, I described how altruistic behaviors can be explained economically, in terms of utility functions where the two goods are our own consumption and another person’s utility. An indifference map based on such a function would predict altruistic actions any time we can give up some of our own consumption to increase another person’s utility enough that the new combination of our consumption and their utility that has a higher utility for us. If we can buy enough increase in their utility at a low enough cost to ourselves, we’d make that trade. How about some examples of this?

Say you run into a complete stranger who’s really thirsty and is one quarter short of the cost of a drink in a vending machine. No stores are open, so the vending machine is their only hope. Would you give them a quarter so they can slake their thirst? They’re a complete stranger, so you’re never going to get your money back or get any intangible rewards from them for your generous action, other than hopefully a hearty thanks. But the cost to you is small and the benefit to them may be disproportionately great. Does their utility factor strongly enough into your utility function that you would trade a small amount of your own consumption (however much you could buy with a lousy quarter) to make them better off?

I’m too polite to ask for your answer, but while you’re mulling it over I do want to acknowledge that the cost of the quarter may reward you in other ways, besides the increase in the other person’s utility. Humans are complex social animals, and all sorts of emotions about giving and receiving are programmed into us by some combination of genetic and social influences. Being generous in this situation may raise your self-esteem, and that may cause you pleasure. If you don’t give a quarter, you may violate some ingrained moral norm about helping out people who are in need, and that may cause you to feel guilty. Or you may be influenced by the slight chance of other people hearing about what you did in this situation.

Any of those motivations could factor into your decision. So if you do give that poor thirsty person a quarter, we can’t be positive that their utility does indeed factor into your utility function. But your actions can be explained in terms of the value to you of their utility—in terms of altruism. I emphasize this because some people delight in undercutting the idea of altruism by pointing out that these other motivations exist. But that fact by no means rules out that altruism is also a bona fide human motivation, and may explain a lot of what we do.

Another nice example is the ads you’ve probably seen showing deprived children in tropical countries that you can help feed for just pennies a day. You don’t know these children and probably never will, but the ads promise that a small decrease in your own consumption will dramatically increase their utility. Would you make that trade? Would that new combination of their utility and your consumption increase your own utility?

This example illustrates the well-established economic principle of diminishing marginal utility: if you have almost no money, then a dollar increases your utility more than if you already have a lot of money and get one dollar more. Pennies a day would increase the child’s utility far more than losing the same number of pennies would decrease yours, if you’re a fairly affluent member of a developed economy. So even though the child is a perfect stranger, you can buy a lot of their utility with a relatively small cost to your own consumption, which means the trade-off may increase your utility even if the well-being of a perfect stranger isn’t vitally important to you (a.k.a., only weakly factors into your utility function).

But if it’s really just the child’s utility that is motivating you, then you should be almost as likely to pony up to support a second child, and a third, and so on until you’ve given up so much money that a further loss of consumption would actually reduce your utility more than it would be increased by the knowledge that a child somewhere has been fed. That would happen eventually if you kept on giving. Each donation leaves you with less money than before, so each bit of money you give up should lower your utility more than the last one did. (That’s the principle of diminishing marginal utility in reverse, by the way.) But what are the chances that after just one donation you’d reach that critical point where the next donation wouldn’t be worth it to you? That would imply that you had been at a point where your loss of money and the child’s gain of food had almost exactly equal value in your utility function.

My hunch is that the vast majority of people contribute enough in these campaigns to support a single child and then stop, which implies that there are other motivations at work—not necessarily to the exclusion of a sincerely altruistic motivation, but at least in addition to one. People  may benefit from the boost to their self-esteem of knowing that they were generous enough to help out a needy child, and that benefit may more or less max out after one donation. A second donation would have to be motivated only by the value of a second child’s well-being, which may not be quite enough to balance the loss of consumption in their utility function. Or a picture of one starving child may only help the average person derive utility from the thought of one child being given food, and their imagination may not extend to a second. Or a single donation may be small enough that they don’t feel the need to carefully weigh the costs and benefits to themselves of making it, whereas giving in quantity may call for a deeper self-examination that they may not be up for at the moment. Truly, the human heart is mysterious and complex!

© Joel Benington, 2011.