Last week, Britain had a referendum to decide whether or not to replace its current “first past the post” electoral system with the alternative vote system (AV). During the campaign, the No to AV campaign claimed that changing to AV could cost £250 million, in part because voting machines would be introduced with it. Yes campaigner and member of parliament, Simon Hughes claimed that this was false and that the No campaigners knew it was. He asked the electoral commission to stop the No campaigners from lying.
Similar appeals are often made by other frustrated political disputants. But the idea that electioneering politicians should be allowed to say, and voters to hear, only what the electoral commission deems to be true and honestly believed is outrageous. It would make election outcomes depend on the judgement, not of the voters, but of the electoral commissioners.
The proposal is also unnecessary. As anyone who has argued with blowhards will know, there is an easy way of showing that someone does not really believe what he says. Challenge him to a wager. Demand that he put his money where his mouth is.
If the No campaigners really believe that changing to AV would cost £250 million, they will be willing to bet on it. By offering the wager, and having it declined, Mr Hughes would expose their insincerity. Equally, Mr Hughes’ failure to suggest the wager may tell us something about his own alleged certainty on the matter.
Politicians should generally be obliged to bet on the outcomes their various claims. This would discourage their lying which, incredible as it may sound, is even more widespread than people working with the standard definition of lying realise.
Lying is not a matter of saying something you do not really believe. This is because, on matters subject to doubt, we believe “both sides of the debate”. For example, I believe that Osama Bin Laden is dead. But I am not certain of it. Or, in other words, I believe, to a small degree, that Bin Laden is alive. So, if I said “Bin Laden Lives!”, though I would be lying, this would not be because I do not believe it; I do believe it a little. I would be lying because I believe it with less confidence than my assertion suggests.
Once you see lying as misrepresenting your degree of belief, it is clear that politicians lie incessantly. They pretend to a level of confidence that they cannot really feel. Forcing them to bet material sums of money on their claims would encourage them to reveal their true confidence.
For example, Tony Blair said that Saddam Hussein almost certainly possessed weapons of mass destruction (WMD) despite the available evidence making the proposition far from certain. If someone had bet him £100,000 at odds of 5:1 – surely attractive to someone of Mr Blair’s professed certainty – he would have lost £500,000 when the WMD failed to appear. Or, had he refused to wager at these odds, we would have known what to think of his certainty.
Or consider the politicians who bailed out Greek sovereign bond holders on the ground that this would prevent “contagion” and further European sovereign debt crises. This was an implausible idea. What odds do you think those who so enthusiastically peddled it would have taken on a €100,000 bet that there would be no more euro bailouts within two years? What odds would Mervin King, Governor of the Bank of England, take on a bet that inflation will be below 3% in a year? What odds would Donald Trump have taken a month ago on President Obama’s place of birth?
Of course, compulsory betting is an imperfect path to honesty. Much of the nonsense that politicians talk is not amenable to verification and so cannot be bet on. How will we know, for example, if Britain really has become “fair” under the influence of our government’s social mobility policies? And politicians may sometimes be willing to take the betting losses caused by their misrepresentations. Indeed, they could do deals that mean they do not have to. The holders of Greek government debt would surely be glad to cover the gambling losses of the European politicians who gave them taxpayers’ money.
Imperfect but not worthless. If a politician were a long-run loser in his compulsory betting, we would know that he was either a well funded liar or a fool. Knowing which may be required to make a moral judgement about him. But not to know whether we should take him seriously.
In theory, this is a great idea, and I have always tended to use the phrase “so, would you care to wager money on that?” as a means of foreclosing on annoying arguments.
In practice, I cannot wait to see the creation of a market of hedge vehicles for politicians wishing to lower their exposure to losses via exposed lies, and the consequent creation of derivative markets in political bloviation. What could possibly go wrong?
I have found there are a few common responses when I’ve challenged people with this logic:
“It’s too risky for me.”
If the wager is sufficiently small relative to their net worth, this shouldn’t matter even if they’re highly risk averse. It may be quite hard to prove this to them, though.
“I can get better returns elsewhere.”
Mostly applicable for when an interest-free escrow is mutually requested. This is often suspect, since their stated beliefs often imply that they could get an expected ROI in excess of 20% per annum. To counter this, try seeing if you could bet on the success of their portfolio or asking if you could invest with them.
“I don’t gamble/gambling is wrong.”
They might be lying, but unless you have evidence of them gambling elsewhere it’s hard to unveil their insincerity. The best counter-argument I can think of is equating pseudo-gambles like driving a car with formal wagers, but the differences between them make for an easy “they’re just not the same!” rebuttal, even if that’s somewhat weak.
In principle I think this an excellent idea. The “No” campaign was woeful in terms of lack of real argument and scaremongering. Anything that would help stop that would have to be a good thing. It is unfortunate that this sort of campaigning seems to be so effective.
In some cases it could backfire. In the above case, it is clear that the costs of AV need not be nearly so high as £250M, but of course they could be if the most expensive options were always chosen. How would you verify that the statement “The cost of AV will be £250M” is false? Well if the “Yes” campaign wins and AV gets approved, then the costs could possibly be measured. If AV is not implemented, then the costs are never measured. Someone has to arbitrate (possibly the electoral commissioners?), or the bet is off.
But if the argument is persuasive, then just making it, and backing it up with a massive bet, makes it much less likely that AV would actually get approved. The bet makes the lie seem even more persuasive. The bet could well be viewed as a price well worth paying to convince people of your lies. If the lie wins, and AV is rejected, you can argue your case with the arbitrators, or the bet is off anyway. The bigger the bet, the less likely you are to have to pay it.
When a politician makes a statement, they are gambling with their reputation. Blair gambled his and lost, at least with the UK public. It doesn’t seem to have damaged his earning potential. It is really hard to tie them down to a categorical statement, even without a bet. If caught out in an outright lie, they will probably have to resign. Getting one of these creatures to agree to a bet is similar to getting them to give a “yes” or “no” answer – just about impossible.
Two thoughts.
You suggest that recipients of the Greek bailout would be willing to cover the wagers. Therein lies the rub. If politicians can lay off the wagers then we either have a simple bribery or we have some form of plutocracy. It only works if the politician has his own well-being on the line.
Mutatis mutandis doesn’t this suggest that everyone should pay taxes and no-one should be exempt? Isn’t paying a tax a bit like wagering the policy will do at least that amount’s worth of good?
Just as a point of fairness, JW presents no reason to suppose that Blair would have spurned the proposed wager. The availability of evidence “making the proposition far from certain” is no proof that Blair ignored or lied about that evidence. You need look no further than some of the posters on this board to see that evidence and argument often has little effect on sincerely held belief.
reminds me of this
http://www.smbc-comics.com/index.php?db=comics&id=2221#comic
It is easy to be persuaded to believe what you want to believe. A wager may make you re-examine the evidence, to see how selective you have been in examining it.
A very simplistic and naive article. I’m willing to make 2 wagers.
I’m willing to bet anyone $ 100 that the unemployment rate will rise above 33% by Dec 2014 with the Ryan plan passed by the house.
The logical terms of the bet are:
If the house bill becomes a law as is and there are no legislative changes affecting it between now and 2014 and the unemployment rate is higher than 33% by Dec 2014 I win.
If the house bill becomes a law as is and there are no legislative changes affecting it between now and 2014 and the unemployment rate is never gets above 33% by Dec 2014 I loose.
If the bill is modified before becoming a law or if there are any legislative changes between now and 2014 that affect the law, then of course all bets are off.
As my second wager not only am I willing to bet $ 100 on the unemployment rate going to 33%. I’m willing to bet $ 1000 that I won’t have to pay out on my bet.
Will A’s response misses the point. Whyte proposes the wager as a proof of good faith. The purpose is not to enrich or impoverish the bettor but to let us see if he means what he says. Will A’s wager is so freighted with caveats that it clearly will not convince anyone of his sincerity. So it falls into the other fork of Whyte’s dilemma: “we would have known what to think of his certainty.” It has thus served a purpose, and in this case quite well.
@ Ken B:
I always miss this point. So please let me know the fair/appropriate terms of the bet on the Paul Ryan budget plan and what it would do to unemployment.
It seems to me there are 2 possibilities:
1) Such fair/appropriate terms exist.
In which case you can enlighten me.
2) Given the all of the variables of politics/economics it really is never possible to come up with fair/appropriate terms to such a bet that both sides of a wager would agree to.
In which case any wager proposed would be a sign of insincerity.
Another point would be that such bets are usually impossible to verify to the satisfaction to both parties of the wager. Let’s say in 1987 2 parties made a bet on whether repeal of Glass–Steagall would cause a huge housing bubble that would result in unemployment going above 10% sometime before 2011. Who has won this bet? Does the loser have any arguments that he can make that he didn’t loose the bet?
It seems to me that the majority of proposed bets on the political/economic effects of some proposal would never be made because terms for the bet could never be agreed to.
And even if such a bet were made, it would never be possible to verify who actually won the bet. “The repeal of Glass–Steagall didn’t cause the housing bubble unethical mortgage brokers caused it.”
So judging the sincerity of someone based on making wagers on these issues seems simplistic at best, flat out wrong at worst.
@Will A: Yes you do miss the point. (We agree this a common occurence.) The point is not if the wager is fair. The point is to filter out obvious bullshit. What the wager would or would not tell us about the world is irrelevant; what it tells us about the true beliefs of the proponents of a policy are the interesting bit. Hence attempts to avoid the wager with caveats etc are themselves quite revealing enough: the wager need not even take place.
This is just the difference between expressed and revealed preference gussied up in a flashy proposal. But the difference is real enough.
Ken B:
Would it be possible to state the type of wager Paul Ryan could make to show how sincere he is that his bill will reduce unemployment and lower healthcare costs?
I bring up Paul Ryan because he is serious and sincere.
There are still some diehards that maintain there is no global warming. Perhaps a wager on global temperatures over the next few years would result in a little more focused examination of the evidence?
By the way, what happened with the bets on the extra half boy problem? Did you get a stats prof to adjudicate, or did everybody drop out in the end?
@Harold: Good example. Of course the world has cooled slightly in the past decade; I will take your circa 1999 bet…. Plus the real issue isn’t whether the globe is warming but WHY and how big a problem it really is.
Ken B. I was careful to say only that there were some who believe there is no global warming.
We could construct a wager – how about the temperature will lie within the IPCC predictions for the next 4 years? You could argue that if it were at the lower end of those predictions it could also be consistent with a world that was not warming.
Had I inserted the word “anthropogenic” it would become more contraversial, and almost impossible to verify.
@Harold: There IS no global warming. There HAS been. There was quite a bit from 1970-2000. There MIGHT BE again, but in the past 8 or ten years and seemingly this year there is none right now. It depends on what the meaning of “is” is to coin a phrase.
If you look at the projections made in about 1999 or 2005 by IPCC models they have already been falsified. (Which does not prove there will be no warming in the next 10 years, only that the models have a problem).
“The point is not if the wager is fair. The point is to filter out obvious bullshit.”
If you just want to filter out obvious bullshit, you could reject everything. That’s guaranteed to reject 100% of the bullshit.
Of course it will also reject 100% of the good ideas. Something which filters out the bullshit but which filters out far too much other stuff as well is not useful. So it actually does matter if the wager is fair–an unfair wager will be far too overexclusive because it filters out the bullshit and just about everything else as well.
Another problem is that diversification in investments is a good idea. A politician is already betting part of his reputation on those statements; betting money as well is a bad idea because he’s putting all his eggs in one basket.
Ken B.
There is an interesting post about betting on climate change here:
http://www.realclimate.org/index.php/archives/2005/06/betting-on-climate-change/
It is from 2005, so a bit old now.
“It depends on what the meaning of “is” is to coin a phrase.”
Well, my bet for serious money would be that the mean global temperature was consistent with IPCC predictions.
I believe that there has been and continues to be a warming trend. The general variability means that short term cooling is entirely consistent with this. See
http://data.giss.nasa.gov/gistemp/graphs/Fig.A2.gif
for global temperatures. The 5 year running average temperature continued to rise during the last decade. If you start in 1998, then it appears flat, but that is cherry-picking and caused by a particularly hot year. There is no valid reason to look only at the data for that period.
There is a flat bit at the end there – this could be because global warming has stopped, or due to fluctuations. The IPCC predictions include a small chance of very low or no warming – particularly over short times, but I am happy to have a little flutter on the likelihood of the trend asserting itself in the next few years. I would make a simple bet that the this graph (as it is updated) starts to slope upwards and continues to rise over the next few years, in line with the trend. If you believe that warming has ceased, then I presume that you would bet that the line continues flat.
“If you look at the projections made in about 1999 or 2005 by IPCC models they have already been falsified.”
I haven’t looked back at this in detail, but I would be willing to bet that this is not the case. This one should be relatively easy to check.
Harold:
Ken B won’t take your bet because he isn’t sincere in what he is saying. If his concern is that that there is randomness involved in the weather, he could easily ask for odds from you.
Ken B:
I still think that there might be a point to be made that given the variables involved in these types of issues, the refusal of someone to take a bet doesn’t mean that he is insincere.
@Harold: “If you believe that warming has ceased, then I presume that you would bet that the line continues flat.” Which is an odd thing to presume considering that I specifically and *more than once* disavowed such a conclusion.
Of course picking 1998 specifically is cherry picking, as is picking 1972 or whatever unusually cool year alarmists pick to start their graphs. Your 5 year rolling average is better, but it is cherry picking to want only 5 year periods excluding 1998.
I will wager that temperature predictions from the IPCC made during the year Al Gore warned us we had only 5 years left — this was more than 5 years ago, we can look it up — are off by more than 5% from the IPCC’s mean value as of 2010. (Thus if they predict 3-5 degrees warming the mean is 4). Please note I am not wagering on the future; I am betting the models were wrong already. $2 via paypal.
Ken B. ” Which is an odd thing to presume considering that I specifically and *more than once* disavowed such a conclusion.” Well, not quite. True, you have avoided stating your beliefs about future climate, but you also say “there IS no global warming.” You have said “There MIGHT BE [warming] again.” I am asking you to to bet on the strength of this “might”. If you think it quite likely that there will be warming again, then you would not bet. If you think it more likely that your stated belief (no current warming) will continue, then you would take the bet. My position is that the warming has continued, so there IS warming, but it has undergone a short term fluctuation which masks the effect for a short period only.
I will take a bet that the temperatures for 2010 are outside the temperatures the IPCC projected. The scientists on the IPCC are aware that there are fluctuations in the temperature, and therefore include a margin of error consistent with the data. I do not know precisely what this margin is, but it is not reasonable to impose an arbitrary error (5%), different from the authors’ own. So pick a year and we can look back at the projections and see if current temperatures are outside. I have not recently checked these projections, so I could be wrong.
@Harold: Here’s a chart showing IPCC predictions with the IPCC’s own error range (in pink)
http://wattsupwiththat.files.wordpress.com/2009/03/akasofu_ipcc.jpg
The dot is 2009, not 2010, but the pink zone comes from the IPCC prediction. This looks like a pretty clear failure of the predictions. So I pick whatever year these predictions were made.
Other failed predictions include these:
“In 2005, the United Nations Environment Programme predicted that climate change would create 50 million climate refugees by 2010. These people, it was said, would flee a range of disasters including sea level rise, increases in the numbers and severity of hurricanes, and disruption to food production.”
You seem to confuse a skeptical demand for accuracy in predictions with a rejection of AGW. That just ain’t so. Al Gore and his epigones demanded radical action on the basis of very shaky — and now disproven — predictions. Thst suggests to me that the skeptics are the more empirically minded ones. Gorism is a parlay: it’s caused by man-made C02, it’s rapid, it’s irreversible, and it’s catastrophic. Each bet is open to some doubt. This debate is just about ‘rapid’.
If you want to start discussing people who have said wrong things about climate change we could be here forever, and I bet(!) I could find quite a few examples of “skeptics” who got it wrong. I don’t want to get into a wide ranging discussion of who said what, and I don’t care much about Al Gore. The only issue for now is whether the IPCC predictions have been falsified.
To illustrate the point, I offered a bet that the 2010 temperature was outside the predictions, although this does not “falsify” the IPCC predictions, even if it were so.
Your graphic falls short of proof, but if you can find a properly referenced example so we can see the original prediction that shows the same thing then you will have won this bet. It is important to use the expected range for individual years temneperature, not for the average. I think you may struggle, I have seen other graphs (on what could be called “skeptics” sites), eg:
http://www.climatechangefacts.info/
showing prediction from Hadley Center. The error bars clearly include a level period until about 2010.
Another piece that pre-dates 2010, but is as good a piece as you are likely to get, published in Science. It compares actual data with IPCC predictions in 2001. It concluded that IPCC underestimated the actual temperature rise, and temperatures up to 2008 were well within the expected range.
From a BBC article, where the reporter claimed that global warming has flatlined: http://www.bbc.co.uk/blogs/paulhudson/2009/11/global-temperatures-and-the-fu.shtml
“The authors say that they have been able to simulate near zero or even negative temperature trends for intervals of a decade or less, due to the model’s internal climate variability. But importantly, the simulation’s rule out zero trends for intervals of 15 years or more.”
So if the trend stays flat until 2015, the models are wrong. Currently we are still within the limits of the model. The next few years will have to be scorchers (globally) if the models are right.
Clouds
With the new results just published in the recognised journal Geophysical Research Letters, scientists have succeeded for the first time in directly observing that the electrically charged particles coming from space and hitting the atmosphere at high speed contribute to creating the aerosols that are the prerequisites for cloud formation
This is important because sunspots are main contender for “the really big thing the IPCC like totally left out dude” http://science.au.dk/en/news-and-events/news-article/artikel/forskere-fra-au-og-dtu-viser-at-partikler-fra-rummet-skaber-skydaekke/ It’s fun to hear skeptics making jokes that amount basically to ” the sun affects weather, who knew?” But the point really is a good one; it has generally been assumed that the sun is a constant as far as climate goes. But it isn’t, sunpots vary quite a lot. If that variation matters ….
@Harold: Yes indeed many skeptics were wrong. Does that mean the IPCC was not wrong?
You cite a prediction from Hadley. I thought we were debating the IPCC predictions. And those from the early 90s too.
I am quite happy btw with agreeing that many “no warming” predictions and models were wrong. My position is we don’t know enough to make confident enough long term predictions to justify the sort of action Al Gore wants (and that we have time to invetigate all kinds of options anyway); the more models that fail the better that stance looks, no?
Robin Hanson eat your heart out! :-)