Thursday, May 22, 2008

Media Economics as Stochastic Process, OR: Perhaps Richard Posner Was Right About Something After All...

One of everyone's least favorite Posner screeds is his critique of the media, published (where else) in the King of All Media, a.k.a., the New York Times. The article prompted so much annoyance that Bill Keller was moved to write, and then print, a letter to himself in response.

And yet what Keller calls Posner's "trademark theory of market determinism" is starting to look more and more appealing as an explanation of some of the sillier aspects of the media today. And status in general, particularly the sorts that require the uninformed public to buy in to things.

Consider the question on the minds of many today, including my two very favorite bloggers (Helen and our own Belle), to wit, Why The Hell did the NYT publish that horrible atavistic driveling snot-nosed vapid putrid and profoundly self-indulgent in a way that only those who are Young and Talented and in New York City and have been told that they are Young and Talented and in New York City for So Very Long by so many Sexy, Sexy People that they think they can turn contents of their garbage can into Writing That Reveals Essential Truths About The Essential Human Condition All The While Maintaining A Properly Ironic Attitude Because All The Other Cool Kids Are Doing That Too And Anyway Some Professor I Vaguely Remember From My Elite Undergraduate School Once Said Something Cool About Modernity piece from Emily Gould (no, I will not link it). For that matter, Why the Hell, one might ask, does the NYT publish that modern love column? And Why Gawker? For that matter, WHY?!?

And the answer, my darling friends, readers, google searchers, and criminal stalkers, is really quite simple and Posnerian: because people read that shit.

Why, perhaps, is the more interesting question, is it that people read that shit?

But the answer to that question requires a level of coldness and trademark Posnerian market determinism that I'm not sure I can handle.

Let's see.

om mani padme hum
rational choice rational choice rational choice
f(u)= β1x1+ β2x2...βnxn
homo economicus

There. I feel better. NOW:

Suppose that ordinary people, call them media consumers, have a very poor grip on quality in media production -- they know they have preferences over the sorts of things that famous people produce, but they don't know very well what will satisfy those preferences ex ante (that is, the public has no taste). Media institution endorsement serves as a signal for media consumers that they'll like a given product. Moreover, media consumers have a taste for status: some part of their utility function for the sorts of things that famous people produce is increasing in endorsement of the famous person in question by some media institution.

Now suppose that there is a strict ordering of people by status (Sp), where status is endorsement by media institutions, measured in probability of being endorsed by a given institution (Πi), weighted by the influence of that institution, measured in Readers (Ri). That is:

SpiRii+1Ri+1... + ΠnRn

where p indexes people and i indexes media institutions. (I imagine R itself will have to be weighted by status, i.e. of readers, too, but that'll be a different variable, and can probably be proxied by income... at any rate, we can drop that for -- HAHAHAHA -- simplicity.)

Now let Tp=person p's talent, and suppose that for all p, i, Πi, p=Tpi, where epsilon;i= media institution i's error. Further assume that theΠi, ps are correlated such that for all i, p, if Πi, p increases, so does Πi+1, p, etc.

Then all we need is for εi to be high in one, or a very few, influential media institution. Suppose that the errors work as follows: in each time period (sliced however one likes) media institutions do a search of the available writers (actors, models, etc. -- call them Potentially Famous People (PFPs)). Each institution selects the best available PFP, with the noise in its judgment modeled by εi. Say that εi is a draw from some suitable probability distribution whose bounds depend on the competence of i.

Then, in each time period, there's some positive probability that someone totally worthless gets published somewhere highly correlated with high influence institutions. That person's status goes through the roof, and, as a result, media consumers form a demand for his/her output. Media institutions being aware of this demand, and market driven, they have an incentive to publish that person. Said person's status increases accordingly. Keep iterating, until Fame.

There. Cory Kennedy, Emily Gould, and... dare I say ... Richard Posner. Explained. You may send my Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel in care of Satan. (I'm sure the economics committee has his fedex coordinates on speeddial.)