John Quiggin has a nice post on the differences or lack thereof between Galbraith and Becker and Murphy on advertising --- namely, both assume advertising is uniformative but B&M allow for the possibility that consumers might like it.
John Quiggin suggests that because we are 'paid' to watch ads on television -- essentially by the lure of television programs -- that it must be the case that ads are bad and do not provide satisfaction.
While I don't disagree with this conclusion (I think that is likely to be right because people spend money to avoid ads), the economic logic behind is argument is incomplete. To see this, suppose that everyone loved ads but some people liked them more than others. This means that a broadcaster of ads only (with no TV shows) will not attract all viewers. That may reduce their advertising revenues and so it is conceivable that even in this case the broadcaster might offer the inducement of television shows in order to attract more ad-viewers.
Thus, observations that people watch television shows does not prove that they like or hate ads. Now, if there was an explicit monetary payment that would be another matter.
What might be a better proof that people hate ads is where we don't see them: e.g., in books.
[Update: John said that he needed to ponder the welfare implications of a model of broadcast television with advertising. No need. Simon Anderson and Stephen Coate have done all the work in a paper published in the Review of Economic Studies in October 2005.]