During the discussion following the death of JK Galbraith, the issue of advertising came up. In the Affluent Society Galbraith dismissed the idea that advertising is informative, and argued instead that it was used to manufacture demand for goods and services people would otherwise not want. The NYT obit suggested that Gary Becker and George Stigler had disproved this, a proposition that attracted some attention, mainly focusing on the work of Becker and Murphy.

Although Becker and Murphy don’t present it this way, their model actually supports Galbraith in most respects.

For a start, the idea that advertising could be informative is excluded by assumption. In the standard neoclassical model, adopted here consumers are supposed to know their own tastes (and the processes by which tastes may change over time) and to be fully informed.

Advertising is simply media content that is complementary with consumption of the goods advertised: this just says in economic terminology that people who consume the ads are more likely to consume the products or, even more simply, that ads sell stuff.

A possibly helpful analogy is that of salted nuts or pretzels on the counter in a bar. For obvious reasons, salted nuts are complementary with beer. And just as it makes sense for bar owners to make nuts available freely or cheaply, it makes sense for people selling a good to offer ads.

What about consumers? An obvious case of the Becker-Murphy story arises when the ads tell a story that enhances the subjective value of consuming the good in question. A pair of shoes that make you feel like a basketball star is better (for the target market) than a pair of shoes that just covers your feet. Becker and Murphy pay a fair bit of attention to this case, and so do people who comment on them.

But this isn’t the only way that ads can be complementary. Ads that make you discontented with your existing possessions, or reduce the subjective value of competing products work just as well.

The economic model presented by Becker and Murphy provides a simple and elegant way of distinguishing the two. If advertising is a good, which enhances the package of ad+product, consumers will be willing to pay for it. If advertising is a bad, consumers will have to be paid (or forced) to consume it.

An immediate consequence is that most of what we think of as advertising is a bad. We watch TV ads not because we like them, but because we are paid with the programs they accompany. This is the point made by the TV executive who said we were breaking the contract if we watched the programs but not the ads (he did say we could take a break to go to the toilet, IIRC).

Given the public good properties of financing broadcast TV, it’s possible to make a second-best argument that this social arrangement improves welfare on balance (I still need to work through this one, but for me at least, the price is too high, and I hardly ever watch ad-inclusive TV).

Billboards, though, are an unmitigated bad. If we wanted to look at them, we would pay to go and see them as with movies and concerts. And given that we are selling our attention to advertisers on TV and radio, those who force billboards into our field of view are taking that attention without payment, just like telemarketers making collect reverse charges calls.

An immediate policy conclusion, the exact obverse of the one about TV viewers watching ads, is that users of billboard advertising should be required to pay everyone who goes past. Given the transactions costs of implementing this, they should be taxed at rates comparable to the advertising charges of TV stations.