Spin in the Information Age

“When you trust your television what you get is what you got, ’cause when they own the information, oh they can bend it all they want” John Meyer, Waiting on the World to Change

I hate to rehash old posts (OK, “hate” is a strong word) but I always thought the song was cowardly. “We want to change the world, but it will be hard because vested interests have the upper hand, so we’ll just wait and hope something good happens.” Yeah, that’ll get it done.

I caught an article earlier today from Joe Klein over at Time titled “I Don’t Know.” It’s mostly Klein rambling on about how he doesn’t like Romney, but he makes a couple of good points. First, he says that anybody claiming to “know” who will win is full of it. On this we agree – but I prefer to put it in probabilistic terms. That is, nobody knows with 100% certainty. But that is not the same as saying “I think it’s a 50-50 proposition” – I don’t, and quite a few other people don’t. But to be sure, none of us really knows who will win.

The people who know the best are the very ones who won’t talk about it, at least not honestly. The campaign staff in each of the major campaigns probably have the best models, the best analysis, the best read of anybody on what will actually happen. They pay for better polling, more specific polling, more precise measures of performance. Further, they do these things so they can go out and impact the outcome, meaning they still have some control (though less as the days go on).

The realization that they have the best read on the race is interesting given Klein’s second good point: “I tend not to trust anything that comes out of presidential campaigns at this point in the process, especially attempts by said campaigns to ‘analyze’ how things are going.” I couldn’t agree more.

Campaigns spin. They spin like crazy. They always put a positive read on how things are going for their candidate. They even try to add an air of “surely you must know that we know far more about what is happening in the race [stipulated] and therefore you must trust us when we tell you that we’re going to win this thing rather easily.” But it is just spin.

They’re doing their job, mind you. They want to win. Even if it is a last ditch effort against daunting odds, their best chance is to smile and construct some story about why things are going to turn in their direction, to give some obscure statistic from the field that is “better than last time” (and hopefully by a lot) that makes it sound like they’ll win. The last thing you want is for supporters to be disheartened – because that will end the race.

Given that, the counter argument is also true – you’d like to dishearten your opponent’s supporters if at all possible.

We’ve talked in the past about polling and how one should not look at it with rose colored glasses. Pollsters don’t actually care about being right until the end. Only on election day can we call them on their polling prowess. Until then they can fill those poll reports with whatever fictional data they choose – to the end of effecting whatever outcome they want (if they actually want an outcome). That’s not to say the polls are junk, or that all pollsters are political shills. They’re not. It simply means that you can’t really trust it until the rubber meets the road, which isn’t until seven days from now. (Or, really trust that any polling errors are honest errors.)

Of late, both the Romney and Obama campaigns have been spinning the fact that Romney surrogates are buying airtime in Minnesota and Pennsylvania. Why would they do that? I’ve heard a number of explanations – all plausible. The Romney camp will say it’s because Minnesota and Pennsylvania are in play. The Obama camp will say it’s because Romney has figured out that Ohio is a lost cause and is desperately trying to put some other state into play. Outside watchers will say that it’s because Romney just has that much money flowing in and they really can’t do more advertising in Ohio – it’s saturated. That one was an interesting explanation to me.

How much is spin worth? How much is it worth for the base to get excited? Is it worth dropping $2 million in Pennsylvania advertising just to make supporters in Florida, Virginia, and Ohio feel like they’re going to be part of a “wave election” – giving them more intensity? Maybe. You can rest assured that Romney has gotten more free airtime out of that $2 million expenditure than all of the hundreds of millions he’s spent elsewhere.

But if $2 million is a drop in the bucket, then what else could you spin? You could certainly spin polls (commission your own, or just outright bribe a pollster – demanding that they shade results in your direction until the last day before the election), but that (i) is already going to happen in some corners anyway, for free, and (ii) it’s too open.

Every day during election season I visit Real Clear Politics just to see what the polls are doing and what the pundits (left and right) are saying. I like politics. When I go to the RCP homepage I’m given a quick trifecta of information in the right hand column:

  • RCP National Average (currently Romney 47.9 to Obama 47.1, Romney +0.8)
  • Favorability Ratings (currently Romney +6.5 to Obama +3.3, trending Romney)
  • Intrade Odds (currently Romney 36.8 to Obama 63.3)

Intrade is a cool little gambling site out of Ireland that allows you to make bets on political (and other) events. The bet is simple, if you think Obama will win the election you buy a contract for an Obama win that settles at $10 per share. If the odds are “Obama 63.3% chance of winning” then you’d be able to buy that share for $6.33 and get $10 if he does win, $0 if he loses. It’s a market, so you can only purchase what others are selling and only if you can make a price match with them.

There, right there on the RCP homepage, and mentioned in newscasts all over the place in the 24-hour news cycle, is an indication of “what the crowd thinks will happen” … but is it?

Free market capitalists love the price mechanism. We believe that it is the best way to settle the value of goods and services. We believe, I guess, in the “wisdom of crowds” as some have put it. (Note, I don’t think it’s necessary to believe in such a thing to support a free market – instead I simply hold that one must believe that an individual’s ability to determine his own preferences is greater, in general, than the government’s.) By that logic one might look at Intrade and say that the free market believes Obama will win, so he’s the favorite.

But I will point out that national polling does not confirm the Intrade numbers. A few months back the overwhelming majority thought Obama would win, but now it’s about even – actually a slight edge to Romney depending on the poll. No worries though, we know enough to point out that the population of those polled and the population of people willing to “put their money where their mouth is” are not the same, and can easily come to different conclusions.

But what would it cost, I wonder, to manipulate the Intrade market to make it look like your guy is winning? I mean, you get free mentions on talk news if you’re up in Intrade, so what’s it worth? And what would it cost? I can’t tell you the answer to the former, but the latter isn’t that tough to cipher (for a  rough estimate, anyway).

The Intrade market for “Barack Obama to be re-elected President in 2012” has an open interest of 1,347,763 shares as of this writing. At $10 a share that’s $13,477,630 “on the table” for an Obama victory. The “Mitt Romney to be elected President in 2012” contract has an open interest of 1,386,355 shares. There are other contracts with similar readings, such as “Democratic Party candidate to win 2012 Presidential Election” – 56,186 shares. These are small by comparison and simply serve to capture sentiment before the nominations and/or obscure events. If we just take the top two contracts we’re dealing with a shade over 2.7 million shares, or $27 million to “change hands” when the contracts are settled.

Is that a lot of money? Well, to me it’s a lot of money, but in an election where each side will spend in the neighborhood of $1 billion is it still a lot? Especially if one considers that it is completely unnecessary to pony up $27 million to manipulate the market. I’m looking at the site now and for a scant $67,000 I could move Obama’s “probability” from 63.3% to 65.1%. And it’s not all sunk either. Suppose that the actual believed probability from the heard is Obama at 45% – I’d only be out 20 cents on the dollar if I liquidate when we roll back to stasis.

Alexis Levinson has an excellent piece on this over at the Daily Caller, pointing out that it may have been tried in times past. Reading the article one quickly comes to the conclusion that it (a) could be done but (b) would have to be thought out ahead of time – which I can only imagine it would be. So would it be worth it? Well, if for a few million dollars (say $5 million to be safe) I can gradually manage the price of shares for my preferred candidate and keep them elevated up to the election then I might conclude it’s worth it. If, that is, I really do have billions to throw around and a $2 million advertising buy is “throw away” money, and heck – I might actually make money on the proposition.

Does that mean I think somebody has manipulated Intrade? Not at all. The whole point of “investing” or even “speculation” is to believe that you can figure out price movements before everybody else. From that standpoint alone one must conclude that the market doesn’t always “know best” or at least “know first” – otherwise one’s belief about underlying probabilities holds little value. This claim is only bolstered by the notion that, if somebody wanted to, they could manipulate the market for what amounts to a drop in the bucket from the standpoint of a campaign.

The last few weeks have seen significant and dramatic moves in the Intrade probabilities, with Obama dropping and then “correcting” upwards. On 14 October he dropped as low as 58.1% chance of winning, and then pushed back up to above 66% in three days. Then again he dropped back down, hitting 52% on 24 October, only to correct upwards back to above 64% on 26 October. These type of fluctuations aren’t unexpected in the “endgame” when new information is coming quickly and is always more relevant than anything we’ve heard before. But they are interesting.

Who knows, maybe the whole thing is a sideshow with Obama and Romney surrogates dumping money in to manage appearances. Or, maybe (and this is the most likely case) the candidates are out of Intrade and the market is just responding to what it sees. I suspect it will keep responding in fairly dramatic moves over the next week.

Until then don’t believe what the campaigns are saying. We’ll all know the final answer soon enough (unless, of course, this is another “2000 election” that takes months to resolve).

This entry was posted in Uncategorized. Bookmark the permalink.

One Response to Spin in the Information Age

  1. I thought this was interesting and relevant: Silver offers a $1000 bet to Joe Scarborough on an Obama victory. This *would* be a great way to distinguish a biased attempt at manipulating public opinion from an unbiased *projection* of public opinion… except that I think Silver’s very visibility limits the credibility of the move. In other words, he has a lot of eyes on him, and so just like your discussion of Intrade, a $1000 bet is relatively small change if its real motive is to keep Republicans out of the booth next week.

    Fortunately, however, no one pays any attention to me :), so there is no similar potential for ulterior motive: I am also willing to make it interesting. Any readers/writers care to take my $1000 even-odds wager that Obama will win?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s