Food isn’t about Nutrition Clothes aren’t about Comfort Bedrooms aren’t about Sleep Marriage isn’t about Romance Talk isn’t about Info Laughter isn’t about Jokes Charity isn’t about Helping Church isn’t about God
Regulating is not about getting it right “The problem with risk weighted bank capital is that assets assigned the lowest risk, for which capital requirements are therefore low or nonexistent, are those that have the most political support, sovereigns, residential mortgages.” Paul Volcker, 2018 https://subprimeregulations...
So what you're left with, in the words of Frederick Bastiat, is everyone attempting to live off everyone else.A great recipe for moral and economic destruction.
relationship isn't about relationshiplove isn't about feelings for anotherHelping isn't about caringCaring isn't about othersHouses aren't about shelterQuestions aren't about learnngBeing isn't about what isOpinion isn't about epistomologyHappiness isn't about well-beingGood isn't about desireJudging isn't about discrimination
Thanks! I'd expect plenty of political identities (neighboring cities & countries, ethnic & religous groups) in unproportional systems, and at some point there is diminishing return for a larger and larger menu, right?
Proportionality seems to just be a stable way to organize politics. Every major group gets a vote so nobody can complain too much. Besides, also in decisions between humans (e.g. in distributive situations) proportionality seems to be popular.
Can somebody explain why politics being about status leads to proportional representation?
The only thing I can come up with is that in proportional representation every bigger group gets a piece of the pie, so nobody has to feel like a complete loser.
This comment came through on my phone about 9:20 Eastern Time. At that point Trump was showing some strength, but even North Carolina was doubtful at that point. Your comment was the first inkling I had that something was up. I'd told my students clearly that the betting markets were showing that Clinton would do better than Obama had in 2012.
There was a certain amount of variance in the betting markets, and some of them had Clinton at 75-85%. 90% was basically the high end for prediction markets -- it was individuals, like Sam Wang, who were assigning higher probabilities.
For my personal estimate, I basically averaged between Silver's estimate, which he supported with actual data, with the prediction markets, so I estimated about a 75% chance Clinton would win. While one might be able to come up with causes in hindsight for what happened, I still think this was a reasonable estimate the day before the election.
However, note that if something has a 10% chance of happening, it happens about a third as often as something that happens 30% of the time. Given the weakness of humans at prediction, these are fairly close together. So there really was not much difference between Silver's estimate and prediction markets. This is why I didn't offer to bet again when you decreased the odds to 90% -- because it was getting uncomfortably close to what I thought were the real odds, especially since I am risk averse, and especially if you would want a profit margin.
Wang's estimate, and a few others like his, are the real outliers here. If something has a 1% chance of happening, it is 10 times less likely than something that has a 10% chance, and 30 times less likely than something that has a 30% chance. So the 99% figure was much farther from 90%, in the way that matters, than the 90% from Silver's estimate.
A 99% estimated chance of a Clinton victory a few days ago was outrageous, and the figure itself, given the actually uncertain situation of the country, is strong evidence that Wang predicted according to his political preference. And he has already said that he is very upset with the result, which is additional, although weak, evidence of the same thing.
A final point that weighs in the same direction: on Twitter last night he says that Nate Silver was right about there being a high degree of uncertainty. That seems like a good thing, admitting where you were mistaken. But what does it imply about his model? Silver gave explicit reasons, and explicit historical examples, of the kind of variance that would be needed to allow for a Trump victory. Wang basically said, no way the polls can be that far off. That was just wishful thinking.
But the prediction markets said 90%. Would you say the bettors were relying on their political preferences? (The prediction market people must know something went terribly wrong. David Rothschild stopped posting when he had Clinton at .90, confidently pontificating about her "firewall." Then the bottom fell out, and David fell silent.)
Silver got it reasonably right, but he was an outlier (with his own seriously flawed track record).
Probably what needs questioning is the whole idea that you can predict politics by "aggregating" data. Maybe serious political prediction requires substantive (sociological, historical, psychological) insight. That's what Sam apparently lacks, as does the betting population.
That Wang predicted according to his personal preferences is a serious charge. Do you have evidence?
Big reputational hit against trying to judge reality using political preferences.
At the same time Sam Wang said there was a 99% chance of a Hillary win, Nate Silver said the odds were 65-35, which was an actually reasonable estimate of the facts, as opposed to a partisan preference.
It's election day 2016. How about "Voting Isn't About Facts".
I paraphrased this from an article recommended on Marginal Revolution entitled "A New Theory for Why Trump Voters Are So Angry — That Actually Make Sense" (here's the URL https://www.washingtonpost....
This blog isn't about writing
Regulating is not about getting it right “The problem with risk weighted bank capital is that assets assigned the lowest risk, for which capital requirements are therefore low or nonexistent, are those that have the most political support, sovereigns, residential mortgages.” Paul Volcker, 2018 https://subprimeregulations...
So what you're left with, in the words of Frederick Bastiat, is everyone attempting to live off everyone else.A great recipe for moral and economic destruction.
Agreed, first past the post is more about having enough support for collective action.
relationship isn't about relationshiplove isn't about feelings for anotherHelping isn't about caringCaring isn't about othersHouses aren't about shelterQuestions aren't about learnngBeing isn't about what isOpinion isn't about epistomologyHappiness isn't about well-beingGood isn't about desireJudging isn't about discrimination
The issue isn't what identities you can have but what you can affirm via voting.
Thanks! I'd expect plenty of political identities (neighboring cities & countries, ethnic & religous groups) in unproportional systems, and at some point there is diminishing return for a larger and larger menu, right?
Proportionality seems to just be a stable way to organize politics. Every major group gets a vote so nobody can complain too much. Besides, also in decisions between humans (e.g. in distributive situations) proportionality seems to be popular.
Politics lets people project an identity to people around them. Proportional representation offers a larger menu of identities for this purpose.
Can somebody explain why politics being about status leads to proportional representation?
The only thing I can come up with is that in proportional representation every bigger group gets a piece of the pie, so nobody has to feel like a complete loser.
This comment came through on my phone about 9:20 Eastern Time. At that point Trump was showing some strength, but even North Carolina was doubtful at that point. Your comment was the first inkling I had that something was up. I'd told my students clearly that the betting markets were showing that Clinton would do better than Obama had in 2012.
There was a certain amount of variance in the betting markets, and some of them had Clinton at 75-85%. 90% was basically the high end for prediction markets -- it was individuals, like Sam Wang, who were assigning higher probabilities.
For my personal estimate, I basically averaged between Silver's estimate, which he supported with actual data, with the prediction markets, so I estimated about a 75% chance Clinton would win. While one might be able to come up with causes in hindsight for what happened, I still think this was a reasonable estimate the day before the election.
However, note that if something has a 10% chance of happening, it happens about a third as often as something that happens 30% of the time. Given the weakness of humans at prediction, these are fairly close together. So there really was not much difference between Silver's estimate and prediction markets. This is why I didn't offer to bet again when you decreased the odds to 90% -- because it was getting uncomfortably close to what I thought were the real odds, especially since I am risk averse, and especially if you would want a profit margin.
Wang's estimate, and a few others like his, are the real outliers here. If something has a 1% chance of happening, it is 10 times less likely than something that has a 10% chance, and 30 times less likely than something that has a 30% chance. So the 99% figure was much farther from 90%, in the way that matters, than the 90% from Silver's estimate.
A 99% estimated chance of a Clinton victory a few days ago was outrageous, and the figure itself, given the actually uncertain situation of the country, is strong evidence that Wang predicted according to his political preference. And he has already said that he is very upset with the result, which is additional, although weak, evidence of the same thing.
A final point that weighs in the same direction: on Twitter last night he says that Nate Silver was right about there being a high degree of uncertainty. That seems like a good thing, admitting where you were mistaken. But what does it imply about his model? Silver gave explicit reasons, and explicit historical examples, of the kind of variance that would be needed to allow for a Trump victory. Wang basically said, no way the polls can be that far off. That was just wishful thinking.
But the prediction markets said 90%. Would you say the bettors were relying on their political preferences? (The prediction market people must know something went terribly wrong. David Rothschild stopped posting when he had Clinton at .90, confidently pontificating about her "firewall." Then the bottom fell out, and David fell silent.)
Silver got it reasonably right, but he was an outlier (with his own seriously flawed track record).
Probably what needs questioning is the whole idea that you can predict politics by "aggregating" data. Maybe serious political prediction requires substantive (sociological, historical, psychological) insight. That's what Sam apparently lacks, as does the betting population.
That Wang predicted according to his personal preferences is a serious charge. Do you have evidence?
Big reputational hit against trying to judge reality using political preferences.
At the same time Sam Wang said there was a 99% chance of a Hillary win, Nate Silver said the odds were 65-35, which was an actually reasonable estimate of the facts, as opposed to a partisan preference.
Although I recently said odds overwhelmingly for Clinton, I'd bet on Trump now. Probably I'll be wrong again.
[Added.]Big reputational hit for the prediction markets and polls.
It's election day 2016. How about "Voting Isn't About Facts".
I paraphrased this from an article recommended on Marginal Revolution entitled "A New Theory for Why Trump Voters Are So Angry — That Actually Make Sense" (here's the URL https://www.washingtonpost....
The quote in my post was from a footnote in a page of Chapman's web book.