Useful introduction that outlines three ways advertising could be modelled: preference modification, signal of quality, advertising as a complement to consumption of the advertised good.

Ah, sorry, I focussed way too much on the first paragraph of your comment.

“Cancelling out” is what I’m suggesting by the signal jamming equilibrium….

I don’t think so, although I guess the ads could ‘cancel out’ their respective effects upon preference rankings. Maybe an ad for a Subway sald could move it above a McDonalds salad, though.

I’m pretty sure your model isn’t just going to need “Advertising can change preferences”. It’s going to need “Advertising ESPECIALLY changes our preferences towards things that are bad.” Cause otherwise, wouldn’t we just expect the idiot ad viewer to flip from “Oh, I love McDonalds” to “Oh, I love salad” if the next ad in the queue is for salad?

If most people (except the noble guys who set policy, presumably: Plato’s Golden Men who are our betters and whose benevolent rule we need) are blithering fools who’ll just do whatever the ads tell them, then we just expect more investment in advertising and we get a signal jamming equilibrium.

I certainly hope so – is this a question of how much conscious effort they put on when advertising is near?

Mmmm, I want food – if only McD’s had real vegetarian options :/

More importantly, I think, advertising is likely to change the ranking of bundles and the optimal consumption bundle. Do people rationally account for the change in preferences as a consequence of consuming the advertising and choose advertising accordingly?

Surely it depends on how the advertising has changed the individuals relative ranking of goods.

If the adverts do so by increasing the subjective value associated with the product this is a good thing – if the adverts do so by misinforming or reducing the subjective value associated with its competitors then this is a bad thing.