Wednesday, January 11, 2006

Gift Certification

This year, as every year, my grandfather gave me a universal gift certificate for Christmas:

The cool thing about this gift certificate is that I can use it to buy anything I want, since they’ll redeem it just about anywhere. And I don’t have to worry about finding items that add up to just the right amount, because they’ll always give me change.

But seriously: Gift certificates have been around as long as I can remember, but only in the last few years have I noted the arrival of the broad-use gift certificates. You can, for instance, buy a Westfield gift certificate, which can be used at any store in any mall in the Westfield chain. Or you can buy an American Express gift certificate, which can used at any location that accepts American Express cards. The question is, what advantage do these gift certificates have over all of the following: regular gifts, single-retailer gift certificates, and cash?

The simplest – that is, most simplistic - economic analysis says that you should always give cash. Why? Because the recipient knows what she wants or needs better than you do. The worst case scenario is that she’ll use the money to buy whatever you would have bought her, in which case she’s at least no worse off. And otherwise, she buys something she wants more, in which case she’s better off. So if you’re hoping to maximize the recipient’s utility, cash is your best option.

When I teach this lesson to my microeconomics classes, I title it “Why my uncle the economist always gives me money for my birthday.” But as my nephews (and all my other friends and relatives) could tell you, I give them actual presents, not cash. What purpose is served by giving presents instead? Others have tried to answer this question, but here’s a quick summary of leading reasons:
1. To signal to the recipient that you care enough to spend time searching for a gift she’ll like.
2. To save the recipient the time it would take to purchase the gift. (But this argument depends on your having a sufficiently good idea of what the recipient would want. Also, if your opportunity cost of time is greater than the recipient’s, it would make sense to give cash and increase the amount to compensate for travel time.)
3. To make use of superior information – e.g., you might know, or strongly suspect, that the recipient would really enjoy Book X, which you’ve read and she hasn’t. If you gave cash, she might never try that specific book.
4. To accomplish some goal of your own – e.g., giving an electric razor to your husband who doesn’t shave often enough, or giving an educational toy to a child you think should study more often.
But if these reasons are good ones, then why give gift certificates? With respect to signaling, the gift certificate says, “I made a little more effort than just visiting the ATM, but not much more,” or “I know you well enough to find out where you shop, but not well enough to know what you’d buy there.” With respect to time saving, the gift certificate is useless because the recipient still has to go out and purchase what she wants. With respect to using superior information, the gift certificate only uses information about location (“I think Jenny might really like Pottery Barn if she would just try shopping there”) but not about specific items. And with respect to accomplishing a goal of your own, the gift certificate may utterly fail unless the chosen retailer has a very narrow selection.

I figure that gift certificates are intended primarily to send positive but weak signals to recipients. The implicit message is: “I thought of you, and I even spent a little time for your benefit, but I don’t pretend to know you very well.” But from the recipient’s perspective the gift certificate can be a hassle, even if its message is well intentioned. The loss in buying flexibility may swamp the gain in warm fuzziness. Hence the emergence of the broad-use gift certificate. This sort of gift certificate sends a slightly attenuated signal – something like, “I thought you merited a small bit of my valuable time” – in exchange for a relatively large, though not unlimited, increase in buying flexibility.


David Friedman said...

I have one other possible explanation, although I don't find it very convincing.

Suppose I'm a Becker Altruist, someone who gets utility from someone else's utility. It is in my interest to be well informed about that person's utility function, so that if opportunities to increase my utility by doing something for him appear, I will recognize them and act on them. Since I am well informed about his utility function, I can do a pretty good job of buying him a gift he will like without extensive research.

Giving him a gift is still, considered by itself, less efficient than giving him cash--but not by as much as if I were not an altruist wrt him. Hence a gift works as a signal of altruism--the essential requirement being that the signal is more expensive to send when it is false than when it is true.

Why do I want him to know I am an altruist wrt him? See the Rotten Kid Theorem.

Ben said...

Imagine a Christmas where everyone would give cash. What would be the point? Most of the gifts would simply cancel each other out. This works ,albeith with a delay, for birthdays too.

Blar said...

Why not give cash? Because a lot of people don't feel comfortable with money as a "gift". Relationships that involve the exchange of money are understood as a different kind of relationship than those that involve the exchange of gifts, with the latter kind of relationship valued more highly. Gift certificates muddy the distinction between the two kinds of relationships, but cash gifts go further towards effacing it.

For more details, see:

Fiske, A.P. & Tetlock P.E. (1997). Taboo trade-offs: Reactions to transactions that transgress the spheres of justice. Political Psychology, 18, 255-297.

Will Wilkinson said...


I think the lack of flexibility in gift certificates is a large part of the point.

If I get cash, I may spend it on something I "need," like my drycleaning, bils, or a tank of gas. A gift certificate goes into a different psychological account, funding the purchase of things you "want." That's why even a very general AMEX gift card can be nice in the way cash isn't.

And if the certificate is to a store you like and always wanted to buy something from, but considered frivolous or too expensive, the certificate gives you the pleasure of picking out something you wouldn't otherwise buy yourself. How is that not nice?

Last year, my parents gave me a gift card good at a nationwide chain of shopping malls. I ended up buying a very nice jacket at the Banana Republic at the local mall, which has given me a great deal of pleasure, but which I never would have bought had I been given the equivalent amount in cash.

Glen Whitman said...

DDF -- makes sense to me! That's a good reason for a kid not to shop at the Santa's Workshop craptorium at his school.

Ben -- good point. I would rephrase by saying that part of the value of a gift is *instant* enjoyment. This is a sort-of corollary of my reason #2.

Blar -- I think you're right, but the question is why? I think the reasons I've given, plus those of the other commenters, capture most of the logic behind your social psych explanation.

Will -- Yes, I intended to include that reason but forgot. It's an interesting point because it relies more on behavioral econ; people use guilt as a means of establishing self-control, so a gift is a way of giving someone a guilt-free indulgence. However, it's not obvious why a broad-use gift certificate is good for this purpose. The broader is the possible use, the more it's like plain old cash, in which case the recipient is more likely to regard it as augmenting his general purpose budget instead of his indulgence budget. I suppose that's an empirical question, though.

Blar said...

Glen, I don't think that the psychological approach based on Fiske & Tetlock is merely a restatement of the other arguments. Your reasons are all fairly specific mechanisms that apply to this particular situation, while they give a much more general theory about the way that people think. They say that people have four main models for thinking about social relationships that they apply to different situations in different ways, with the applications depending mostly on social norms. The relational models are ranked, with "communal sharing" (everyone pitch in for the good of the group) as the highest and "market pricing" (cost-benefit analysis) as the lowest, and whenever a "lower" model encroaches on the territory of a "higher" model this is generally considered morally offensive and threatening to the domain in which the higher model is thought apply. (The two intermediate models are "authority ranking" (hierarchies of authority) and "equality matching" (fairness and reciprocal exchange)).

This theory is supposed to help explain a broad set of findings, including why people are bothered by utilitarianism and novel applications of market principles (like markets in votes or bodily organs), in addition to people's uneasiness about giving a gift of money. It's not a perfect theory, and it's not the only explanation of why many people don't give cash, but I think it is an additional explanation (or, at the very least, it gives some content to the explanation "social norms").

Glen Whitman said...

Blar -- I guess I'm having a fairly typical economist's reaction to the work of sociologists (or in this case psychologists doing sociology): what they think does the explaining is what I think needs to be explained. Saying that people apply these social interaction categories seems a lot like begging the question to me. The real question is why they have those categories. Is there some rhyme or reason behind the social norms that enforce the category boundaries? I think the economic explanations go a long way (though perhaps not all the way) toward explaining the norms.

Blar said...

Glen, I would think that economic considerations say more about what we should do about our norms moving forward (should we maintain them or modify them?) than about how our norms originally arose. Social norms and cultural practices like gift-giving develop historically, changing over time, even while most economic factors remain pretty constant. They also might have deep evolutionary roots (like reciprocal altruism). Instead of just comparing cash and gift certificates, I think we'd get a moral accurate causal explanation if we first looked at why people were unwilling to give the gift of cash cash back before gift certificates were ubiquitous, and then looked at the history of gift certificates. I'm guessing that gift certificates caught on as replacements for gifts mostly because they had the advantages of convenience and efficiency. They were seen as acceptable gifts when cash was not because people were still able to think of them as analogous to gifts (that's one claim of Fiske & Tetlock - when an unfamiliar situation arises, people will categorize it based largely on analogies). I'm sure that the financial incentives of businesses that sell gift certificates accelerated the process. A good economic historian could say more (and probably already has somewhere).

Here's an addition reason for not giving cash: the fungibility of money. People who give money are unlikely to have their gift associated with a particular item or activity that the person uses / experiences / remembers / thinks about. Instead, there's a good chance that the money will be lumped into a more general mental account that is not associated with the giver (like "Christmas money" or "clothing money" or "disposible income"). The presence of a particular, identifiable gift makes recipients more likely to think about the giver and appreciate the gift. A concrete gift also helps the giver imagine the recipient using the gift and enjoying it. Givers are more likely to get utility from this kind of imagining than from a generalized altruism which brings them utility whenever the other person's utility increases, since 1) it's relevant to their relationship and 2) it's vivid and imaginable, not too abstract.

There are other forms of gift-giving for which most of the proposed mechanisms aren't relevant (Will's indulgence explanation is the only other one that consistently applies). A lot of times, people are given gifts that they specifically ask for. At other times, someone will give an IOU gift of, say, a camera, where they say "you pick out a camera and I'll buy it for you", or even "you pick out a camera and buy it, and I'll send you a check for what it cost." These arrangements are a lot like giving cash and letting the recipient choose what to spend it on, but they at least keep up the appearance of a gift. The main benefits seem to be creating an association between the giver and a specific gift, possibly allowing for an indulgence, and helping people believe that the value of the gift is not merely its market price.

As another example of how people like to imagine that they're giving something specific, rather than just cash, I know that charities often try to associate donations with some vivid purpose for donors to imagine. For instance, there are food pantries where, when you donate money, you pick out "foods" to "buy", and they'll tell you "you bought us X jars of peanut butter and Y boxes of macaroni and cheese", with X and Y based on how much money you gave and the cost to them of getting those products in bulk. This compromise between the efficiency of letting the food pantry buy its food in bulk and the warm fuzziness of picking out some food at the grocery store to donate lets donors think about all those kids eating peanut butter sandwiches (or whatever).

Glen Whitman said...

"I would think that economic considerations say more about what we should do about our norms moving forward (should we maintain them or modify them?) than about how our norms originally arose. Social norms and cultural practices like gift-giving develop historically, changing over time, even while most economic factors remain pretty constant."

Nope. I don't buy the idea that norms have evolved independently of economic considerations. I think many of our norms evolved in order to deal with economic circumstances, and norms have changed as circumstances have changed. This is the research program of Douglass North and other New Institutional economists.

And what's the viable alternative? That norms just "happen" to people? To me, that's not even an explanation. And I notice that when you're pressed, you too resort to essentially economic arguments that are similar to the ones I've already made. Your argument about fungibility, for instance, seems a nice corollary to the signaling argument. A specific durable item creates an ongoing and visible signal in a way that money does not. And you also say, "I'm guessing that gift certificates caught on as replacements for gifts mostly because they had the advantages of convenience and efficiency." Another economic argument, pretty much the same as I made in the initial post.

Vache Folle said...

I'd rather have a certificate than cash because the cash will go for necessities but the certificate will be spent on something I like.

I'd rather have cash or a certificate than anything from Bath and Bodyworks, Yankee Candle or any other thoughtless last minute gift store.

Blar said...

Glen, I'm not as hostile to economic arguments as you seem to think. Of course they help explain many things. I just don't think that they explain everything. In some cases, they don't even appear to be giving answers to the right questions. Most of the arguments here have been providing advantages of gift certificates that might compensate for the obvious efficiency advantages of giving money. What benefits do gift certificates have? Why might a gift-giver prefer to give a gift certificate rather than a check? My last comment offered another answer to these questions, which was related to the signaling argument and the altruism argument (although I'm not sure that I'd categorize my argument as entirely economic).

But if the purpose of this analysis is to explain social norms, then these arguments tell us very little. So people might sometimes decide that a gift certificate is a better gift than cash. But why would they feel uncomfortable about giving cash? Why would people often feel like giving money isn't even an option? Why should I have to worry about people being offended if I suggest that they just give cash? To answer these questions, I think you have to look at other sorts of analyses, including historical, psychological, and sociological. Some economic analysis might fit in here, but to even know how it does fit you have to venture outside of economics.

One good way to start is to look at the history of the practices and norms that you're trying to explain. Since I don't know the history of gift certificates, I made some educated guesses in my last comment and suggested that there were three main questions:

1. Why wasn't money a popular gift, back before gift certificates existed?
2. How did gift certificates gain popularity?
3. Why hasn't money become a more popular gift since gift certificates have become popular?

The answer to the first question is probably mostly sociological or psychological. The answer to the second question has two main parts: first, the advantages that gift certificates had over existing gifts at the time (which is primarily an economic factor, as you noticed), and second, the explanation for why gift certificates weren't held back by the same snares that had kept monetary gifts from being popular (which would be related to the explanation to question 1). The answer to the third question is more complicated, in part because monetary gifts aren't all that uncommon now, and in part because those norms probably are in flux, but I think that the answer is related to the answer to question 1.

The analysis could become much more complicated if we broaden the scope of our inquiry to the norms and practices of gift-giving of other cultures and eras, or even if we just look at different kinds of gifts within our society right now. Gift certificates are a common Christmas gift, and even monetary gifts aren't uncommon for events like Christmas, birthdays, and graduations. But if a married couple went over to another family's house for dinner, it would be very strange if they gave a check or a gift certificate rather than, say, a bottle of wine or something they bought on vacation. This is true even if they're giving something to the couple's teenage child, who got a birthday check from them earlier in the year. I don't know if I can explain the difference. This example suggests that simply comparing gift certificates and money won't explain our social norms. We need to be paying more attention to the context, to social factors, to relationships, even if these things are not free of economic influences.