David R. Henderson  

The Economics of Gifts

Top Economist's Headache... The Knowledge-Power Discrepanc...

The Institute for Humane Studies did an interview with George Mason University economist Chris Coyne just before Valentine's Day on the economics of gifts. Chris lays out one of the standard claims that economists have made about gifts: that to the extent the recipient values the gift less than the cost to the giver, there is deadweight loss. This is the conclusion I had reached in graduate school when studying economics got me rethinking so many things. But my friend Jeff Hummel pointed out that there really can't be a deadweight loss in the standard sense because the gift buyer is making a voluntary decision. Jeff writes:

Non-monetary gifts cannot EX ANTE cause deadweight loss any more than buying yourself such a "gift," unless the purchase is involuntary. The value the recipient places on the gift is virtually irrelevant, except insofar as it enters the utility function of the gift giver. Thus, if the recipient reports to the gift giver disappointment after receiving the
gift, this may cause some EX POST deadweight loss, just as any mistaken consumer purchase. Or even rarer, if the recipient actually places a negative value (rather than just a lower value than the price) on the gift, this MIGHT also generate some deadweight loss even if the giver is unaware of the fact. Notice, this last is most likely to apply to an
insulting monetary gift, which therefore cannot be most efficient, despite what Chris says. I know that the non-monetary gift as a supposed example of deadweight loss has even found its way into texts, but it is based on a misunderstanding of welfare economics.

What about Chris's second idea in the interview that the purpose of giving is to signal? I agree with that and I'm glad that years ago I came to that conclusion. I made piece with the idea of gift giving, a peace that has helped my 28-year-old marriage survive and, typically, thrive. But I don't agree that the key is for the gift-giver to signal by bearing a high cost. That might be true for some gift recipients; it's not true for the recipient I know best--my wife. What I've learned over the years is that the "signal" my wife values is that she is visible to me--that I've noticed what she likes and doesn't like. So, for example, one of my biggest victories, when I was on sabbatical in Australia, was buying for her Christmas present three tickets for the stage play, "Shirley Valentine," because that is one of her favorite movies. I rarely hit such home runs, but at least I understand the "game."

Jeff points out:

If the problem with monetary gifts were ONLY costly signaling, as Chris repeatedly implies, why wouldn't simply increasing the value of a cash gift solve the problem? He never adequately explains how asymmetric information has anything to do with gift giving, nor the exact signaling problem involved with a pure cash gift. It clearly involves much more than just the cost to the giver. [ME: I agree; thus my point above.] Here is Greg Mankiw's somewhat clearer, although still incomplete, explanation of what is going on. Notice, Mankiw does not claim that non-monetary gifts generate deadweight loss.

Comments and Sharing

COMMENTS (8 to date)
PrometheeFeu writes:

Here is an other theory of gift giving:

While we may have properly defined preferences over everything in this world, we don't know about all the things over which our preferences are defined. For instance, two weeks ago, I discovered a new kind if breadstick which I prefer to most other snacks but which I had not known about before. Gift giving helps with the discovery process. I might not be willing to just go and buy something I never tried before because I have very little information about my purchase ex ante. However, my wife knows me well and there are products which she knows well too. She might be able to make a reasonable guess as to what my preference for a certain item might be. So, she buys me an item which she thinks I might like as a gift allowing me to discover some new product.

PS: I am not presenting this as an alternative to the signaling model but as a supplement which explains why you might not go with the "safe" gift of something you already know the person likes.

Dominik writes:

Re: signaling that you know what someone else wants. Are you sure?

Dan Carroll writes:

I read an article a while back (can't remember the source) that identified gift giving as a means to cementing social ties. The term "signaling" wasn't used, and it seems an incomplete description. The best term I can come up with is ritual. It conveys a sense of belonging, social cohesion, and togetherness which are instinctive to humans (though the exact meaning varies by culture). The value of the gift is only relevant to the extent that it conveys that meaning. It provides utility to both the giver and the receiver.

I would surmise that individuals that are more socially oriented (extroverts) will value gifts more. Analytical introverts, like economists, will tend to view gift giving as an annoyance.

Shane writes:

Good point PrometheeFeu.

"What I've learned over the years is that the "signal" my wife values is that she is visible to me--that I've noticed what she likes and doesn't like."

Certainly seems true, David. Perhaps another signal a gift implies is that the giver has spent time and effort to please the recipient. In recent years when I lacked the money for decent presents I baked a cake for one friend, and made wooden products like combs and a clock for others. Recipients seem to appreciate the effort and time necessary to actually make something, instead of simply buying them. (Should anyone care, here's the clock!)

Tom West writes:

I have to absolutely agree with David's line

that I've noticed what she likes and doesn't like.

as a key to successful gifts,

I'll add that another aspect of gift giving is that is may allow certain recipients a self-indulgent pleasure that could not be achieved if the recipient purchased the item herself (as it would come with a strong degree of guilt). Receiving the item as a gift absolves the recipient of any responsibility for not "putting family first".

joshua writes:

The existence of anonymous gift-giving would suggest that it is not always about signaling.

Evan writes:

@Tom West

I'll add that another aspect of gift giving is that is may allow certain recipients a self-indulgent pleasure that could not be achieved if the recipient purchased the item herself (as it would come with a strong degree of guilt).
I think this is certainly true. A lot of members of my family have extremely bizarre, specific interests (Baroque era music, pulp magazines, etc.) that are very difficult to keep track of, so a lot of times we just point blank ask each other what we want. But we still love our gifts.

David Friedman writes:

Let me expand on the "signalling" theory.

In Becker's model of altruism, the altruist has the beneficiary's utility as an argument in his own utility function. That makes it in the interest of the altruist to have information about the beneficiary's utility function, in order that he can take actions that increase the beneficiary's utility and so, indirectly, the altruist's. Giving well chosen gifts is evidence that you are well informed about the recipient's utility function. Hence it is a signal that you are an altruist wrt that beneficiary which is less expensive to send if it is true than if it is false.

Why do you want to signal the beneficiary that you are an altruist with regard to him? Because of Becker's rotten kid theorem, which implies that it is in the rational self-interest of a beneficiary to take actions that benefit his altruist--even if the beneficiary is not himself an altruist.

So A gives B a (well chosen) gift to signal B that A is an altruist with regard to B, and by doing so gives B an incentive to take actions that benefit B when the cost to B of doing so is less than the value to A.

Comments for this entry have been closed
Return to top