Mar 212007

I just finished reading a very well-written, ingenious, and thought-provoking paper by Eric Goldman, “A Coasean Analysis of Marketing” (Santa Clara Univ. Legal Studies Research Paper No. 06-03)

I’m flattered he tried to make use of attention economics. However, I do not agree with its conclusions, nor indeed some of its premises. I also do not believe he has taken what I have argued fully into account. Here are some cavils.

Goldman takes it for granted that marketing in various forms can do good things, but argues that it imposes a cost on ordinary “consumers,” partly in attention paid. (I think that as far as attention goes, no one is a “consumer,” though possibly a fan, who must do active work to pay attention.) the economist Ronald Coase suggested that the way to deal with “externalities,” such as the cost (in lost “utility’) to “consumers” of marketing by firms was not necessarily regulation but some form of negotiation. Goldman believes this negotiation can be handled or dispensed with by a technological device he calls a “Coasean filter,” that would compare all incoming marketing messages with the particular consumer’s current utility as determinedl from an analysis of the consumer’s current location, her response to prior messages, her own communications, etc.

To start at the end, I think the COASEAN FILTER Goldman proposes is science fiction and will remain so. Basically, this device is to function like a highly attentive secretary (or mother?) virtually able to read your mind, and to know your preferences as well or better than you do. (Why couldn’t this device make purchasing decisions for you? Or find you a suitable spouse?) This requires a level of attention and personal alignment that few people ever get from another human, and is well beyond capacities of artificial intelligence for the foreseeable future. (I like to see announcements of art exhibits, for example, but one glance is usually enough to indicate to me whether further exploration would be warranted. For a device to understand my artistic taste at any given moment, it would have to be more discerning than the best art critic. )

I fear that Goldman’s notion, once put forward, will be taken seriously in policy and legal considerations of appropriate regulation. The result would be some very crude filter that would soon be overwhelmed by marketers unrestrained by current sensitivities. While this will eventually be righted as marketers with good sense restrain themselves for fear of losing potential customers, in the meantime everyone would be very disrupted.

I think there is much to learn from the fact that the no-call list has to be one of the most popular government programs ever. A ringing phone demands immediate attention. Then, normal standards of civility make it hard to hang up. Saying no to someone who acts familiar and friendly violates the norms again, so quite a substantial number of people find themselves actually ordering or buying what they did not and do not want. I think the objections Goldman offers to this method of regulation are wildly overstated. Most other forms of advertising are much easier to ignore, even spam.

Goldman also seem to underrate the value of tying marketing to active search rather than intrusion. It is known as shopping. Lots of people love the chase, more even than the purchases. They still buy a hell of a lot more than they need, and often even more than on reflection they want. Active shopping demonstrates that one is deserving of attention by those who note the suitability of one’s questions, taste and, sometimes, actual purchases. If it was a good salesperson who gave one attention, and to whom one paid attention, that may create a sense of obligation to pay for the joy of shopping by buying.

Goldman defines marketing in terms of selling, but then includes political calls (also annoying) the main intent of which is to influence voting, not buying. Why not then include any unsolicited attempt at attention getting? All efforts at attention getting are in competition to some degree. Most of us become reasonably good at filtering out less insistent forms unless they particularly connect. We can do this very much more easily than a device could. TiVo’s were designed to allow quickly passing over commercials, but according to a recent NY Times article most users apparently don’t bother to turn off the ads. I suspect this is just because they are already so good at ignoring them. Sometimes the ads are more entertaining than the programs, in which case they are probably available on YouTube, as well. But this just means the marketing message itself had to be encased in an interesting bit of what I call illusory attention; when we pay attention to it, that does not signal we have any interest in the product being marketed, any more than we indicate that by watching the program in which the ad was embedded.

I think Goldman greatly exaggerate the likelihood that one benefits from an ad in the sense of discovering something one ought to buy. Ads serve other functions. A correctly placed ad gets attention that any potential purchaser can hope will accrue to her if she buys and shows off the item.

Goldman is not alone in suggesting that “consumers” accrue definite amounts of “UTILITY” from purchases or encounters with marketers. Economists in my view are mistaken if they believe a single-valued, one-dimensional utility or utility function exists for any real person, as it might for an industrial firm. It is simply not the case that we go about maximizing utility as if such a function existed. We are neither that simple, nor that rational. For the most part, we muddle through, getting on jags or one sort or another that keep us from simply random drift. We make a huge number of attention decisions every day. We have no way of knowing whether or not each individually would enhance “utility,” because we must pay attention to know what we are paying attention to. We can only get some dim sense of whether our choices taken together make life good or not. At times, we realize that certain choices do not seem that good, and then with varying degrees of difficulty, we can try to change habits, which may or may not work out. But what we want from life varies all the time, in quite complex ways, which certainly cannot be captured by a single positive or negative number.

Coase seems to share the assumption that individual utility exists, and it is necessary for his argument. Well then, his argument breaks down.

  2 Responses to “Attention, Marketing, A "Coasean Filter, and Utility”

  1. Michael,
    This is a very interesting post. Is your concern more that you don’t think possible:

    “To start at the end, I think the COASEAN FILTER Goldman proposes is science fiction and will remain so. … This requires a level of attention and personal alignment that few people ever get from another human, and is well beyond capacities of artificial intelligence for the foreseeable future.”

    If your argument is simply, you don’t believe this is possible – would you be interested in the theory if you felt it was possible?

  2. Thanks for the great comments. Believe it or not, I agree with most of what you say.

    In particular, it’s completely fair to characterize my proposed filter as sci-fi. It may be that it never gets developed because of the limits of AI. But note my broader point; it may also not be developed because the regulatory environment may remove those choices from developers and consumers.

    As for the point about utility, I think we’re talking about different things. I don’t hypothesize a single one-size-fits-all utility curve for all consumers. Instead, my point is that each consumer has heterogeneous preferences that are private information, so any top-down approach of maximizing utility isn’t going to be able to consider all those idiosyncratic preferences. Having said that, I understand your point that consumers don’t manage their personal utility deliberately (and maybe not very well). Here’s where perhaps we disagree. I think consumers are intensely rational when acting; it’s just that we as outsiders can’t figure out what preferences they are maximizing. It’s also true that consumers don’t consider all relevant information in acting; but I work from the premise that consumers maximize the factors they consider.

    Regards, Eric.

 Leave a Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>