61 comments

[ 3.2 ms ] story [ 111 ms ] thread
Imagine you honestly believe in intelligent design. You probably know most people believe the opposite, so when you encounter poll designed by people you don't know you assume they believe evolution theory is true, so if you want to win money you say whatever it takes, no matter what you believe.

So IMHO this experiment measured how well people predict what experimenter thinks is correct, and not how people beliefs change when incentivized.

To control for this divide the group that gets money into 2, and say to one subgroup you're republican think-tank, and to the other you're democrat think-tank, and compare the results :)

In your metaphorical question, there is actually a correct, factual answer and it's not intelligent design.
I'm guessing ajuc agrees with you, but it's completely beside his point.
Yes.
I'd say that admitting something is generally accepted as true (which you're agreeing the respondents did) is the first step towards admitting that it might actually be true.

The truth isn't subjective, and it seems that the survey elicited that from people to an extent.

That's just, like, your opinion, man.

I happen to know for a fact that I did not fart at a party in 2001, but that's not generally accepted as true.

Well, by popular opinion and downvotes, apparently it's totally up for debate whether Saddam had WMDs in Iraq.
This is certainly a factor that needs to be controlled for.

To me, the most interesting part of the study is that, by rewarding correct answers and confessions of ignorance, they got the most accurate results.

I would like to see that effect studied for its statistical significance, and for the technique to be applied to polling if it proves to in fact increase the accuracy of the signal.

Something like 'This survey asks questions about facts and opinions. Your compensation depends on correct answers to factual questions. Answering "I don't know" to a factual question is worth N% of a correct answer.'
This is an additional complicating factor, but it's not some fundamental problem with these type of studies. There are facts that essentially all people agree to be true when provided with all the information ("objective facts"), and the question is how people's subjective beliefs about these facts are correlated with political party when these people do not have ready access to all the information (as we usually do not).
Imagine you honestly believe in intelligent design. You probably know most people believe the opposite, so when you encounter poll designed by people you don't know you assume they believe evolution theory is true...

That's a poor example that demonstrates your own ignorance about the USA as a whole.

In the USA more people believe in creationism (God created everything 6000 years ago) than intelligent design. And far more people believe in intelligent design than evolution.

See http://www.gallup.com/poll/155003/hold-creationist-view-huma... for the numbers. It is worth noting that belief in evolution is within a factor of two of belief that Obama is a Muslim (about 20%) and belief that Elvis is still alive (around 7%).

Yep, I assumed I encounter representative portion of USA on internet, apparently I was very wrong.

USA I am disappointed.

Maybe thinking global warming is a scam would be better example.

EDIT: on the other hand I never stated the poll takes place in USA :)

Maybe thinking global warming is a scam would be better example.

Better, yes. Good?

Depending on how you ask the question, between 2/3 and 3/4 of Americans believe in global warming. However that figure drops to under half of Republicans.

(Lest you think I'm merely insulting Republicans here, let me point out that most Democrats do not understand what economics 101 says about the effects of rent control. Both parties show a lot of ignorance about the world, merely different parts of it.)

You are missing the forest for his tree example.

He was using creationism as a means to illustrate a point: that the responders probably tilted their responses towards what they thought the researchers believed instead of what they themselves believed.

Obviously I understood his point.

My point remains that the USA is much less accepting of science than most educated people in a bubble of similarly educated people will realize.

This effect doesn't go away entirely though (as was admittedly stated in the article).

In political betting markets, you can almost always find a +EV bet by betting on the most likely candidate to win because those who are on the side of the underdog always overprice their candidate's chances, some of this can be attributed to buying into the neck-and-neck rhetoric.

http://en.wikipedia.org/wiki/Favourite-longshot_bias

But is there any way to exploit this effect? I am a big believer in federalism and even further devolution of power to smaller polities. It's a good tradeoff of liberty vs. society: If you don't like the way your city is run, it's a lot easier to move to the next town over than it is to leave the country. As a result, despite being a pretty laissez faire guy, I'm much more tolerant of, say, San Francisco banning plastic shopping bags than if the EPA were to do so.

It also has the pleasant result of giving voters much more direct control over the way the government that affects them most is run. Their vote really does matter; it is, in many ways, a "bet" on which policies will provide the best outcomes for them.

it is, in many ways, a "bet" on which policies will provide the best outcomes for them

Which is why many people shy away from it. Voters don't want to be responsible for their votes and their actions. More and more we see the populace wanting to hand over control of their lives to someone else. They live vicariously through celebrities and movies while eagerly voting to redistribute wealth via the Nanny State.

Occupy Wall Street was right in tune with this no-repercussions thinking. Capitalism was squarely in the OWS crosshairs since Capitalism is the best way we have of keeping track of good and bad decisions.

The forecast is for more of the same. The TEA Party was the last gasp of a desire of a minor segment of society that wanted to reduce the size of government and increase individual Liberty. They've pretty much been shouted down by the thought leaders with the biggest megaphones in our society.

The juxtaposition of this comment on an article about how bullshit levels go way down when bullshitters are financially responsible for their statements practically blows my irony meter up.

Capitalism was squarely in the OWS crosshairs since Capitalism is the best way we have of keeping track of good and bad decisions.

Sigh. OWS was protesting the fact that the government handed over billions of dollars to Wall Street banks while middle America got shafted. Either you don't know what capitalism is or you're just spouting talking points. Or possibly both.

OWS and the Tea Party were the opposite sides of the same ideological coin. They went about it different ways but they were both concerned with fairness and ending governmental oligarchy.

The OWS Manifesto completely contradicts your contention.

http://www.peoplesassemblies.org/2012/05/may-12th-globalmay-...

That manifesto completely contradicts itself mostly. I'm not sure which of my contentions (that OWS was protesting against the huge bailouts or that they are of the same ideological bent as the Tea Party) you're referring to but there are enough points in the linked manifesto to support either including :

We are living in a world controlled by forces incapable of giving freedom and dignity to the world´s population (if, indeed, they ever were). A world where we are told ‘there is no alternative’ to the loss of rights achieved through the long, hard struggles of our ancestors.

Yes, point 1 is clearly socialist in nature but point 2 is clearly democratic in nature. This is the inherent problem of trying to singularly define a movement that is made up grassroots organizations, each with their own agenda. But the theme behind all of them is the social and economic inequality that resulted from the bailouts. The societal themes that brought both OWS and the Tea Party to life were the same. How they went about addressing them may have been different but the underlying hopes were very similar.

If OWS were protesting the government handing over billions of dollars, it would have been called "Occupy the Capitol Building".
Depends on who you think is in charge, the government or Wall Street.
I disagree, I think that MY tribal group is totally right no matter what and YOUR tribal group is totally wrong no matter what. I'm gonna namedrop the concept of 'austrian economics' while I'm at it in order to prove that I'm a detached, rational analyst of the situation.
You completely missed the point of the article. All positions aren't equivalent and equally chosen when consequences are on the line.

I was drawing a connection between the economic system of Capitalism and consequences vs the Nanny State of big government which avoids consequences.

Yes, you imputed nothing but noble motives and plans to your tribal group, while for the other tribal group, you set them up with not only a bad plan but explicitly bad motives!

Then brought in some keywords about capitalism (they're trying to destroy it!), the Nanny State (that's what they want!). Frankly, I'm surprised you restrained yourself from accusing them of supporting Josef Stalin.

From here, the conversation goes into nutpicking, you'll find some dreadlocked retard who actually says he opposes all corporations while his parents' investment account pays his college tuition. I'll respond with the "keep government away from my medicare" lady. And so on.

Meanwhile, no common sense reform happens in Washington.

you set them up with not only a bad plan but explicitly bad motives

They aren't my plan or motives, so don't blame me for them. I simply labeled them. I think you're more offended that I'm daring to point them out than anything, which is a shame because you should be offended by the people pushing low accountability agendas.

Meanwhile, no common sense reform happens in Washington

You're hardly demonstrating how Washington can do things any better when you spew a bunch of vitriol against me that had zero justification to support the low accountability agendas that are actually in the context of the article this whole thread is based upon. Then you threw in some hyperbole that wasn't even close to being in my original post, referencing Stalin and Austrian Economics.

The problem of federalism is that a lot of polities will be shitty and the rest of the country won't stand for it. Segregation would still probably be legal in many parts of the country if the federal government hadn't put the states in their place in the 1960's and 1970's.

Also, how do you reconcile glorification of federalism with the fact that the level of corruption and general terribleness in the big state governments is far worse than that of the federal government, despite the federal government being much larger and more extensive? Would you really rather devolve power from the federal government to California, Illinois, New York, Texas, or Florida? Ohio, Georgia, Massachusetts? Sure, as you say you can move, and that's a fair point, but the only states that are arguably well-run are the ones that are also insignificant in terms of population and GDP (or have economies based on unsustainable things like federal spending (Virginia), oil resources (the Dakotas), etc).

(comment deleted)
Both really good points. First, I'd mention that lots of polities are pretty shitty right now in a variety of ways, and yet (mostly) continue to be left to their own devices. True injustice gets the citizenry's gander up to be sure (and rightfully so!), but there doesn't seem to be much agitation for the Federal government to dictate to failing cities like Detroit or to crummy school districts in Kansas where they don't teach science when they don't like what it says.

As to corruption: I agree that state governments are more corrupt, but I still like the tradeoff. In a hypothetical world where state and local governments had more power (read: more money directly under their control), I'd hope that the media and the voters would pay closer attention. Probably a dream, but even so, just because a system is less corrupt doesn't mean it's making better decisions. A bureaucrat in Washington can have great intentions, but if he screws up badly, it's the whole country that suffers the consequences. Most importantly, there's no way to explore the counterfactual of a nation without that bad decision having been made. Obviously states and localities never provide perfect "controls" in the laboratory of democracy, but watching the future of, say, Texas vs. California will provide us with some good notions about the practical effects of very different views of the role of government.

I'm not really arguing for a wholesale devolution of power. I have no problem with the Feds continuing to build highways and run the FDA. But the extent to which state governments (mostly voluntarily!) are becoming addicted to a range of "free" Federal aid and its associated strings is a bothersome development.

Segregation would still probably be legal in many parts of the country if the federal government hadn't put the states in their place

Segregation would still probably exist without the feds stepping in therefore let's spend trillions of dollars a year driving our economy into the dirt.

Then on the other side, you have the Dred Scott Decision, the Fugitive Slave Act and many other Federal moves that prolonged slavery in this country. Would it not have been for the northern states and churches, the Federal Government would have been able to keep slavery itself in place.

Bottom line is that we want to maximize freedom for individuals and small group decision making wherever possible. Larger entities should only step in when individuals or small groups can't work it out for themselves.

The federal government decisions that prolonged slavery were animated by the interests of states that would without a strong central government still have slavery, and impacted mostly those states, so I'm not sure that helps your argument much.
So now you're falling back on "interests" of the states when the actions of the Federal government are the ones we have on hard record.

I could just as easily counter Rayiner's contention that the Feds were responsible for ending Segregation by saying that the interests of the states, localities, and citizenry were the actual force behind the end of Segregation.

You actually think that these states would still have slavery? To this day, thirty years after the last country in the world (Mauritania) abolished it?
Segregation would still be de-jure lawful in Alabama, in the same way that sodomy laws were still technically enforceable in Texas until struck down by SCOTUS.
Another commmenter hit this, but it deserves underlining.

You can't ask the question of whether certain statements, to the people giving them, are matters of fact or political opinion, and then have "correct" answers which are provided by others. If you have correct answers, you already know that certain opinions are based on error. The only thing you can do with such an experiment is measure how well people can guess what the judges think to be true. As it turns out, they did fairly well.

I could care less about this example, but let's take the WMD question. Old WMD was found in Iraq. http://en.wikipedia.org/wiki/Iraq_and_weapons_of_mass_destru...

A partisan wouldn't argue the facts. He'd simply point out that, depending on how you phrase the question, the answer could be interpreted either way. The questioner has to rephrase "After the second Gulf War, were WMD found that were created in between the two wars?"

This easily goes to extremes, and questions end up being extremely detailed in order to resolve matters of "fact" This is why the role of fact-checker is a waste of time. Most all of the time, politicians aren't lying, they're simply phrasing things in enough ambiguity to create the desired impression. (This actually works for me as "bullshit", but it's not lying, and it's not giving out factual errors)

I find this line of argument very unconvincing. Would you apply it to the question of whether inflation and/or unemployment went down during Regan's term? Do you really think this has anything to do with the precise definition of those quantities?
I think you picked a poor example, because inflation is hard to measure and hard to define.

I know lots of people who argue with our current definition of inflation using reasonable-sounding arguments.

Inflation is only hard to define for those who want there to be something called "inflation" that predicts recessions and is caused by the government. Naturally it's hard to find anything to nail down within those parameters.

In the rest of the world inflation equals price increases and there are standard metrics that have been used for decades. Most economists would probably agree that those metrics have several well-known flaws. But we can agree that inflation according to the definitions and metrics that 99% of the world uses did or did not change by X or Y in a given time period.

The fact that the existing metrics are flawed mean that it's perfectly reasonable for a person to both:

(a) Believe one thing (b) Understand that the "accepted fact" contradicts their belief

If no money is on the line, this reasonable person will state their belief. If money is on the line, they are more likely to state whatever they think the "accepted fact" is, regardless of their true belief.

In politics, there are many things for which minor differences in interpretation can cause a "fact" to change between true, false, and unknown. Inflation is certainly one of those things.

The question "did X go up during Reagan's term?" is probably correctly answered "yes," at the same time "did X go down during Reagan's term" is correctly answered "yes." Because some things move both up and down, and the question isn't explicit that it is asking about the overall trend.

EDIT: said the opposite of what I meant to say

Thus far, I'm agreeing with you. Facts have nothing to do with the rhetoric being spewed. It's all just tribalism.

As a progressive, I had believed that we could all reach some kind of agreement, if only we could talk it out. Not understanding why that wasn't working, I started researching (Don't Think of an Elephant, What's Wrong with Kansas, etc) and experimenting (practicing talking points on right wing family and friends).

It's been illuminating.

One of my brothers was advocating higher effective tax rates, deficit reduction, and so forth. So I said "So you'd support returning to Clinton era policies?"

The answer surprised me.

He said no. I asked why not. After some musing, he told me Clinton era economic growth was a "false economy". Um, okay. If he had something about bubbles, I'd totally agree. But all he had was "false economy". Probably something he heard on the radio.

I've had similar discussions about healthcare. Every single value, position, complaint, etc would suggest the belligerent would support universal coverage with a single payer. But dare suggest it and there's a strong reaction. Hell no, that's socialism. Or whatever.

I've decided there's no profit in attempting persuasion.

The only successful strategy is to organize better than your opponents. The purpose of rhetoric (messaging, persuasion) is to motivate one's supporters or neutralize opposition.

To really get at this, you need to get people to make predictions about the future, not just about what some panel of judges thinks "correct" (even if I'd be more inclined to agree with the judges on any particular issue).
If I may attempt some supplemental notes to your reply:

What I took out of the puff piece (it's really a book advertisement, which if you need a quick heuristic, features a herd of sheep on its cover - something I instantly view as at best landing it in "pop science / economics / sociology done badly") is that they offered an incentive to game the system. The author is attempting to make a 'free market' tragedy of the commons argument to suggest that voters should be given incentives to vote.

The problem is two fold:

#1 Voters already have these incentives, but are informed by others (who certainly don't have altruistic goals) that these do not exist[1]

#2 Democracy, in America (arguably everywhere), is already driven by this mechanism, just not at the level that the author is addressing. It is called Lobbying. To use a quick metaphor, you pay to set the questions people get to vote on - this shouldn't be news to anyone with a moderate or above political understanding.

As for the WMD claims, beyond the simplistic level of common media, the question of Iraq's ownership of chemical weapons was never in doubt (the USA, France and so on sold them to Saddam / provided the engineering expertise to build them), it was whether or not Saddam was a) building more, b) reducing stockpiles and c) willing to use them ever again. Hans Blix was on record[2] as stating that Saddam was obviously not attempting a and c, and was attempting b[2].Thus, again, the people setting the question didn't like this conclusion, so re-framed the question as widely as possible to be proven correct.

A much more interesting question over all of this is how to separate beliefs from politics, or at the very least maximize the scientific basis for informed participation in society.

And that's not going to be solved by a book with a picture of sheep on the cover.

[1]It's called "living in a society", fire, police, health services etc which I realize will not be popular with a Libertarian readership. [2]http://www.telegraph.co.uk/news/worldnews/middleeast/iraq/70... - this is a right-wing source

Voters have no incentive to vote correctly. The probability of your vote affecting the outcome is so fantastically small (read: too small to be represented by a double precision floating point) as to be irrelevant.
From where I sit, voters have precisely one concrete incentive beyond vague feel-goodery: bragging rights. (Granted, people can and do bullshit about how/if they voted, but that's a minority case.)
In my jurisdiction, there's a mandatory manual recount triggered by a close result every single general election.

If you think your vote doesn't matter, you haven't been paying attention.

I should have clarified that what I said holds for national or state elections.
While I agree with the spirit of your post, doubles can represent numbers far, far smaller than that. They can represent the probability of two people randomly choosing four atoms out of the entire universe, and each picking the same four atoms.
Crunch the numbers with some reasonable estimates - you'll find that the probability of two people picking the same atoms is considerably more than the probability of one vote mattering.
This is probably true at the National level; however on individual State / local government levels this isn't true at all.

You can argue that the original tripartite division of US government doesn't scale to modern population levels, but your argument is otherwise extremely weak.

> And that's not going to be solved by a book with a picture of sheep on the cover.

Attacking a book by its cover?

Are you unaware of the "WAKE UP SHEEPLE" meme?

Suggesting that voters (or "the masses", or "the proletariat") are a flock of dumb animals that need herding / shepherding is as old as the hills.

Either you're unaware of this, or don't find it problematic. I find the latter to be less edifying than the former.

I don't think your argument hold. I think this more about whether person has his/her own skin in the game.

Lets take example, of a person working for company X and there is a claim Y which is related to company X. Lets imagine that the company is Google and the claim is "Google App Engine is one of the best cloud services".

So when that person is on Hacker News, then he/she will agree with that statement. However, if that person start preparing a new company he/she will be neutral about statement "Google App Engine is one of the best cloud services" and will evaluate other services. Why? Because he/she has her own skin the game.

Giving people a chance to save face, with "I don't know", deflates partisanship. USA cold learn from eastern cultures about this.
A major problem with this experimental setting, and the proposal that bets signal belief, is that they ignore the system the bets are placed in.

As students observe, "the right answer" depends on who's doing the judging. Most students have learned to answer not what they believe, but what they believe the examiner wants to hear. This is a fairly important skill, as many examiners are incapable of asking even unambiguous questions.

On the other side of the equation, information signalled in a bet is tainted by the signaller. The answer they give is not necessarily their "honest belief", but rather what they wish to signal as such. Much of high-stakes information, such as in politics, commerce, and warfare, is valued opaquely, for the purpose of enticing onlookers to act on the bets that they do see, which can be counter to the interests of the bettor.

Getting people to make decisions that directly impact them is a good way of figuring out which beliefs they actually use to predict the behavior of the world and which beliefs they only profess but in some way know are false.

Consider someone who claims "I have a dragon in my garage.". When someone asks to test that claim by looking, you say it's an invisible dragon. When they then say they'll listen to the dragon breathing, you say the dragon doesn't need to breathe. If they offer to throw a bag of flour over it and see the shape of the flour-covered dragon, you say the dragon is permeable to flour. Despite saying you have a dragon in your garage, all your predictions are based on not having a dragon in your garage. If you actually thought you had a dragon in your garage, you'd say "sure, let's see"; you should be perfectly willing to test that claim and accept the result. (Credit to http://lesswrong.com/lw/i4/belief_in_belief/ for the dragon example.)

The same lines of experimental questioning can generally distinguish between "I truly believe this and use it as a model to predict how the world works" and "I profess this but when it comes time to predict how the world works I do so using a model that actually works.".

On the other hand, if you build a system that rewards people for how strongly they profess their beliefs, and how much confirmation and support they get from others professing the same beliefs, you can give strong positive reinforcement to beliefs that do not model the world.

There is a problem with the invisible dragon.

If your believer is aware that the strongest alternative to their belief that there is a dragon in the garage is that there is no dragon, they will tailor their explanations to your perspective. That doesn't prove they share your perspective deep down if they could just admit it to themselves, it just means they understand you.

It doesn't matter whether they understand the other person's beliefs or not; the question is, what do they actually believe experiments will show about the world, before performing the experiments? If they truly expect that there's a dragon in their garage, they'll expect results consistent with having a dragon in their garage; if they don't actually believe that, they'll make predictions consistent with not having a dragon in their garage.
They believe that you will see X because you believe Y, while they will see P because they believe Q. They believe that Y-ers' worldviews cause them to interpret P as X, and blind them to seeing X as P. They don't believe that there is a neutral substrate of observation that everyone can agree on beforehand.

I don't completely disagree with what you're saying, I am just trying to point out that the game goes many levels deep. A strategy that leads to victory on level 1, "tell me more about this dragon", is not going to work on level 2.

Too bad there isn't anymore a great prediction market. This could be a nice opportunity for bitcoin.