My own rational market hypothesis

A couple things I read the past few days:

(1) PPP (a Democratic pollster) did the best of any pollster in terms of predicting the 2012 election. In fact, the polls they did themselves finished first out of 28 sets of polls, the ones they did for Daily Kos finished second. That seems pretty remarkable to me.

(2) Models that predict higher rates of global warming (up to 8 degrees Fahrenheit) have been performing better than other models for climate change thus far.

I’ve filed this under The Dirty Fucking Hippies Were Right, but that’s not really my point here. I believe in some kind of rational market for national dialogue, where pundits et al. generally behave in the way that is most beneficial for their careers. Right now, that generally means either taking a far-right Fox News position or a “both sides do it” position.

This incentivizes predictions that are more friendly to the right. A pollster who believes, correctly, that more Democrats than Republicans will vote in a presidential election will call an election more accurately but will also be pilloried for TEH LIBRUL BIAS. I suspect that even within the world of climate science, which is probably more insulated from this than establishment media, the same principle applies albeit to a lesser extent.

I tend to believe that nearly everything works this way. That doesn’t mean liberals will always be right, because predictions are hard, especially about the future, but it does mean that nearly everything we hear through any kind of official channel is slanted to the right relative to what a straight analysis would be.

Share On Facebook
Share On Twitter
Share On Google Plus
Share On Pinterest
Share On Reddit

42 replies
  1. 1
    Maude says:

    Did this post just show up or is my paranoia kicking in?

  2. 2
    Regnad Kcin says:

    gresham’s law at work, b*tchez

  3. 3
    300baud says:

    PPP (a Democratic pollster) did the best of any pollster in terms of predicting the 2012 election.

    Couldn’t this be a bit of stopped clocks being right twice a day? If they lean D but the Obama ground game brought out more D than outside observers could have predicted, that would be cancelling errors, not genius.

    Totally with you on climate change, though. If I were making a prediction on climate change, I’d be pretty conservative about what I said. I figure it’ll go like measuring the charge on the electron:

    http://en.wikipedia.org/wiki/O.....ethodology

  4. 4
    schrodinger's cat says:

    I think the media, always comes down on the side of the conventional wisdom, which tends to skew toward the institutions that hold power.

  5. 5
    Schlemizel says:

    For climate, and to some extent Nate’s work on polls, what is actually written is more along the lines of:
    “Here are what the numbers say”
    “Here are some assumption we can make based on past performance/trends/trig functions/what have you”
    “Here is what we consider the doubt factor or what is unknowable”
    “Here is what I think the outcome will be”

    Thats what professionals do. Reporters take that very last one and ignore everything above it to get a sensational result – particularly if the things above were not done or done badly for nefarious reasons.

    Professionals might be wrong & so they tend to be conservative with that last one. Whores want attention so they tend to be extreme with that last one. The Press loves to ignore everything else & claim there is no way we can know which is right

  6. 6
    Ash Can says:

    But how long will that behavior be “most beneficial” for their careers? With the electorate changing, I have to think these bobble heads and bloviators will be competing for a shrinking viewership/readership over the upcoming years if they keep toeing the RW line. I guess we’ll have some idea of whether these geniuses recognize this in a couple of years, when they once again have to choose between Karl Rove’s math and reality when reporting on the midterm elections.

  7. 7
    Schlemizel says:

    @300baud:

    Perhaps but they were more accurate in 08 and 10 also. Gallup and Rass both were heavily skewed towards the GOP in each of these 3 cases but got closer and closer to the actual results as election day neared.

    I think we can draw a tentative conclusion that PPP has a very good model and that Gallup and Rass do too but chose to put their thumbs on the scale

  8. 8
    schrodinger's cat says:

    BTW is Jonathan Karl of ABC a wingnut. He was on ABC’s this week and when the Democrat on the panel said that Romney was the leader of the party, he quickly jumped in to correct him. His take on the deficit and the so called fiscal cliff was also quite hacktacular and a repeat of GOP talking points.

  9. 9
    Marmot says:

    I believe in that rational market hypothesis too, DougJ. But why do its incentives push toward the right? I’ve never had a good grasp on that part.

  10. 10
    Schlemizel says:

    @schrodinger’s cat:

    Please to be pointing out to this poor stranger in a strange land – which person on ABC not a wingnut?

  11. 11
    blingee says:

    Yawn, way to buy the orange satan propaganda DougJ. Knew I could count on you to get it wrong.

    Nate Silver has PPP coming in at 15th out of 23. Near the bottom along with Gallup and Rassmussen. I can’t be bothered to see how Kos interpreted (manipulated) the data to make them #1 or 2 but I’ll trust Nates findings more than I will the website that uses PPP…mkay.
    http://fivethirtyeight.blogs.n.....tial-race/

  12. 12
    MikeJ says:

    With the polls, it’s interesting that Republicans grabbed a Democratic concern and distorted it and got much, much more mileage out of it than we ever did.

    Democrats complained that polls didn’t accurately reflect the voting population, skewing towards old white people that had landlines.

    Republicans looked at the results of polls and complained about the outcome(there are more Democrats than Republicans!) rather than about the process.

    Democrats wanted the demographics of the polls to match the electorate. Republicans wanted more republicans.

  13. 13
    roc says:

    What I’ve noticed is the more general rule: “polite” analysis is tilted toward the most impolite party.

    Because everyone else is (by definition) more polite, they will prefer not to inadvertently insult/call out/upset the impolite participant. So they will push for the result that’s close to reality, but least likely to cause a stink.

    When it comes to the press, you simply swap polite/impolite with the likelihood of a given set of potential viewers to rage-quit/boycott/etc. So, printed political analysis becomes tilted toward the group most likely to cancel their subscriptions if they read things they don’t agree with. Etc.

    But it’s not purely left/right in American political terms. Plenty of liberals have similar blind-spots around e.g. nuclear power, GM foods, etc. And they drag ‘polite’ analysis away from “no, really, our best science says this can be as, or more, safe than anything else” to “concerns remain” and conspiracy theories and everything else.

  14. 14
    schrodinger's cat says:

    @Schlemizel: Ok may be I am noticing it because I did not have any TV for the past couple of years. Even now I just get ABC, Fox and 4 PBS channels.

  15. 15
    MikeJ says:

    @300baud:

    If they lean D but the Obama ground game brought out more D than outside observers could have predicted, that would be cancelling errors, not genius.

    You shouldn’t base your samples on a predetermined percentage of Dems and Republicans. If you want to know how many Democrats are going to vote, go ask 1,000 people what party they identify with.

    Republicans were pissed because the *answer* to the poll didn’t match what they wanted.

  16. 16
    Marmot says:

    @roc: That’s all true. But any idea why conservatives are most likely to rage quit/create stink/freak out?

  17. 17

    @Marmot: Because it gets results. Because they have successfully defined the nation as “center right” which conveniently reflects a mirror image of how they view themselves. In short, narcissistic personality disorder writ large.

  18. 18
    cervantes says:

    Scientists (unfortunately) generally don’t think much about public opinion, and aren’t very engaged politically around the consequences of their work. Climate science isn’t comparatively conservative because of public pressure, but because that is the basic nature of science — to be very cautious about drawing firm conclusions. Particularly in this instance, in which we are dealing with extreme complexity, in both the vernacular sense and a technical sense. The technical sense trips up a lot of communication with the public – the weather is so-called chaotic, which means it’s unpredictable at time scales of more than a few days, and frequently moves outside of its usual range. The climate is a much longer term, larger scale (both spatially and temporally) construct. That’s why scientists don’t want to say that any particular even is a consequence of climate change.

    On the other hand the probabilities, or if you will frequencies, of particular kinds of weather events do change with the climate. We’re just beginning to understand this, and some of these changes may be locally against wider trend, e.g. excursions of arctic air to the south produce unusual local cold. All this is hard to explain to the innumerate general public. That’s just a fundamental challenge.

  19. 19
    Marmot says:

    @ranchandsyrup: I think you’re right that many of them do it on purpose, because it gets results. Doubt those are the majority, though. How about your average wingnut retiree, say? I doubt there’s much strategy in that mind.

    Those conservatives are my relatives, and I’m fascinated with why they behave that way. It’s as if they don’t understand the idea of objectivity as a goal.

  20. 20
    1badbaba3 says:

    Numbers make my brain*hurt*.

    But I agree with Der Blingle, the GOS at #2 is the giveaway. So, once more into the breech – Frak those firebagging motherfuckers! Always trying to take credit and hog teh gloree. He’s *our* Metrosexual Black Abe Lincoln! We saw him first!

    Mom, Markos is lookin’ at me!

  21. 21

    @Marmot: I think the average wingnut retiree knows well how much influence they wield. The older voting bloc is disproportionately powerful due to participation. They have the 2010 success story from the midterms and a belief that they must save ‘Murica from the godless libtards.

  22. 22
    roc says:

    @Marmot: It looks like they’re disproportionately likely to freak out because the ‘themes’ of the last two election cycles hit on or near the things they freak out about.

    And that was largely by design, as Republican strategists decided they long-ago lost demographics and now need to win on turn-out, and so they hit hard on the things that drive conservatives to the polls. And those things tend to be those they freak out about. (socialism, redistribution, liberal media bias, reverse-racism, etc)

    But my whole point is that there’s no reason to believe that will always be the case. If/when the Republicans realize that not even turning-out-the-crazies is effective, their personality quirks won’t be so disproportionately visible.

  23. 23

    … but it does mean that nearly everything we hear through any kind of official channel is slanted to the right relative to what a straight analysis would be.

    National dialogue has resembled this since I first started paying attention in 1984.

    @schrodinger’s cat: Maybe this is why.

  24. 24

    Agreed on the structure of the debate. (Republicans don’t care what centrists, or liberals, or scientists, or anyone who knows anything about anything think about them, so they’re always too far right; “centrist” commentators always blame both parties equally, so they are always wrong; Democrats care passionately what “centrist” writers think, so they are always too far right.)

    Unclear to me whether this is mostly cultural (my feeling), or they’re following orders from their paymasters, some combination, or whichever other factors.

    Colbert to Pres. Bush, 2006: “Now, I know there are some polls out there saying this man has a 32% approval rating. But guys like us, we don’t pay attention to the polls. We know that polls are just a collection of statistics that reflect what people are thinking in ‘reality.’ And reality has a well-known liberal bias.” Famous quote, but funny that it’s actually about polling data.

  25. 25
    Heliopause says:

    pundits et al. generally behave in the way that is most beneficial for their careers.

    While this pressure certainly exists the more important one, I think, is the selection bias at the front end of the process. That is, people are selected to be pundits in mainstream outlets because they have safe, mainstream views to begin with.

    I suspect that even within the world of climate science, which is probably more insulated from this than establishment media, the same principle applies albeit to a lesser extent.

    Do you have any evidence for your suspicion that climate scientists are dumbing down their models as an appeal to center-rightists? Do you also suspect that Nate Silver kept saying that Romney had a chance not because he really thought so but because he worked for a mainstream outlet?

  26. 26
    Schlemizel says:

    @schrodinger’s cat:

    You are lucky I guess. The networks all have problems in this area. Even the token lib, George Judus Maximus Stephenoplis, couldn’t wait & started stabbing Clinton in the back on his (george’s) way out of the White House.

    There is no “moderate middle” right wing meme too stupid for him to not carry

  27. 27
    CW in LA says:

    @MikeJ:

    With the polls, it’s interesting that Republicans grabbed a Democratic concern and distorted it and got much, much more mileage out of it than we ever did.

    That’s standard operating procedure. Remember after the national embarrassment of the 2000 election when everybody said, “Man, we need to fix our electoral procedures to make sure nothing like that ever happens again”? Rethugs immediately started shrieking about “Voter fraud” and illegal aliens voting and such rubbish. And that became the big issue that “electoral reform” came to revolve around.

  28. 28
    danimal says:

    I’ve noted for several years that climate change models are consistently wrong: the actual negative impacts of climate change are much worse than the projections imply. They are written down and archived, look up the 10 year projections made in 2001.

    It’s frightening, Eventually the business community will realize climate change denial costs them money, and will cost much more money in the future. Until then: stalemate. Hurricane Sandy did more for climate change acknowledgement than a hundred studies by climatologists. We need the studies, but they aren’t going to mean much politically until the business community and, frankly, the Republican Party interest groups start seeing the economic impacts.

  29. 29
    Metrosexual Manichean Monster DougJ says:

    @Heliopause:

    Do you also suspect that Nate Silver kept saying that Romney had a chance not because he really thought so but because he worked for a mainstream outlet?

    Yes, I do. Sam Wang wasn’t saying it.

  30. 30
    NotMax says:

    predictions are hard, especially about the future

    Say what?

    Predictions, by definition, are about the future, as the “pre” in predictions delineates. Accuracy and/or lessening margin of uncertainty over extended timeframes are hard.

    Oracles don’t oraculate about the past. Nor do diviners so divine, not forecasters so forecast.

  31. 31
    MikeJ says:

    @NotMax: You’re not a fan of Niels Bohr, are you?

  32. 32
    jayackroyd says:

    One of the things that has me in stark raving terror is that I know how conservative scientists are. Predicting catastrophe is a bad career move. You always want to hedge the prediction.

    IAC, the existential threat we face is climate. And our leadership is worried by Zion, and making some money outta the fracking business.

  33. 33
    Omnes Omnibus says:

    @Heliopause: I did not interpret DougJ’s climate scientist remark as meaning that results are dumbed down, but rather that there are incentives toward toward reporting the more “conservative” analysis of the data.

  34. 34
    NotMax says:

    @MikeJ

    To paraphrase (okay, bastardize) Heisenberg, it is uncertain.

    :)

  35. 35
    RSA says:

    @CW in LA:

    Rethugs immediately started shrieking about “Voter fraud” and illegal aliens voting and such rubbish.

    Speaking of which, Dean “unskewed” Chambers is back with his “analysis” of how Romney lost. Fraud, of course. (Via TPM.)

  36. 36

    @Metrosexual Manichean Monster DougJ: I dunno. Silver came out of baseball. Any given at bat, Roger Clemens wasn’t likely to get a base hit; but it wasn’t a zero percent chance.

  37. 37
    ThatLeftTurnInABQ says:

    @BruceFromOhio:

    National dialogue has resembled this since I first started paying attention in 1984 1972.

    My version goes back another 12 years and the story is the same.

    Honestly, I don’t think we’ve had a dominantly liberal media-pundit complex since probably 1968 at best.

    Before that, as best I can tell from reading Perlstein’s accounts of the period, the conservative myth regarding “our liberal media” originated as a valid complaint that the news media during the 1950s and 60s was biased in favor of liberal Republicans over conservative Republicans, which came out when they (the media) were covering GOP intra-party conflicts. Even at the best of times it was never that much of a partisan pro-Dem bias or a non-partisan pro-liberal bias.

  38. 38

    @danimal:

    We need the studies, but they aren’t going to mean much politically until the business community and, frankly, the Republican Party interest groups start seeing the economic impacts.

    The problem with placing hopes on “the business community” is that, to the extent such a community exists, it is not known for being particularly far-sighted. If you explain that their current behavior will makes a lot of money for two years, but after that disaster, they will take the money for two years nearly every single time.

    When the economic impacts arrive, they will turn to the federal government to bail them out. They do it every time. And, sadly, the federal government does it every time.

  39. 39
    kay says:

    I feel like it was more than bad polling. It was NOT SEEING. Nate Silver wasn’t looking for a particular electorate, so he was open to the idea that there was the big group of voters who might come out for Obama. Pundits weren’t like that. They had already announced that Obama’s coalition was fractured and disappointed, so they approached everything they saw with that pre-conceived notion.

    What remains weird to me (and I might love this, I haven’t decided) is that you can run this huge elaborate on the ground campaign in this country, like Obama did in swing states, and not be SEEN by national political press.

    THAT is wild. It was hardly underground or secret, but it was somehow invisible to them. They just kept talking about “enthusiasm” for Romney, and there was this whole other thing going on, below their radar or something. It’s not like the voters just appeared out of nowhere.

    They were there all along :)

  40. 40
    Heliopause says:

    @Omnes Omnibus:

    Omitting or downplaying a portion of your results because of social pressure is indeed dumbing it down.

    @Metrosexual Manichean Monster DougJ:

    Then you’re essentially saying Silver deliberately cooked his model for the sake of the NYT, since he simply reported the results of his model. This seems fairly absurd, since his model also said that McCain had a chance in 2008, before Silver was working for the NYT.

  41. 41
    El Cid says:

    Which more affects your comfort, earnings, and position as a figure in the ‘news’ media — your accuracy and truthfulness, or pissing off corporate elites, the wealthy in general, and right wing politicians and activists?

    In recent decades, the safer selfish choice (made consciously or otherwise) is to prioritize avoiding the latter at the cost of the former.

  42. 42
    Wolfdaughter says:

    @kay:

    My friends and I are all liberals. Some of my friends were very worried about the election. I was worried but less so. I kept reminding them of Nate’s poll aggregates. Some of my friends also quoted conventional wisdom which decrees that minorities and young people just don’t turn out in proportion to their numbers.

    I said to them that they heard it first from me: that young people and minorities were hugely pissed at the Republicans and yes, they would turn out in much higher numbers than before and that most of those votes would go to Obama. And I was right! And thank deity of choice that I was.

    I also knew back before the Iraq war that there were no significant WMDs, that the war would be one huge cock-up, the Arab street would be galvanized, etc.

    If you see the big picture you can see these things, and if you also read widely. I’m no rocket scientist, just a humble retired medical librarian. But I’m good at trends.

Comments are closed.