People who buy cars can often get independent evaluations on them from auto mechanics. Same with house inspectors. More generally, consumers can get help in evaluating products from Consumer Reports, while bond investors can get help evaluating bonds via bond rating agencies like Moody’s. Charity donors can use GiveWell.
There are already lots of indices like this. For example, Heritage has the Index of Economic Freedom. Transparency International has the Corruption Perceptions Index. I'll bet there are more indices for things like free speech, government stability, gender equality, you name it. You could aggregate these indices into an overall quality measure, or let people choose their own weights for various aspects of the index in order to figure out the optimal country for them or something like that.
BTW, one thing that can happen once you set up an index is that national governments will try to game the index, e.g. figure out how the Index of Economic Freedom is calculated and figure out how to increase their score in order to be more attractive to investors. Similar to college rankings I guess.
As I said in the post, the most valuable stats would be those under substantial control of a recent administration. Basic human factors are often not good for that.
Basic human factors could be used such as poverty, access to water, sanitation, healthcare, good food etc
The last thing we should do do right now is further entrench a misleading metric like GDP and rate our governments on that.
Humans, should of course be good at rating these human factors in their own lives but perhaps we are only aware of them relatively, i.e. how much it has increased or decreased recently and this has a disproportionate affect.
You gave examples, some of which are already done and really quite well known and I'm surprised you didn't even try to justify why your corruption index would work better than all the existing corruption indexes or provide information which is not already in the S-factor. This post seems especially wheel-reinventing. I am also not sure there are *any* pieces of information which are relevant to migrants or emigrants and which are not already well known and predictable from the S-factor. The Value of Information is the decisions it changes; what decision would a subsaharan African migrant make differently about Germany given some new corruption index?
Can we agree that announcing relatively random pieces of info about states is unlikely to seem very useful to voters, but in the vast space of such things there are probably some that would be influential and useful? I listed some specific things I thought were promising but you didn't respond to those, you just pointed out other things you didn't think promising. We can agree that there are many things that don't seem promising.
Very rarely. As consistent with the Outside View. And you are ignoring all of my other points as well in favor of making the extremely weak 'well it's *possible* isn't it? you can't prove it isn't possible!' claim. (This is an argument you are usually extremely strongly opposed to, and I recall you mocking it with reference to time travel etc just a few days ago on Twitter...) No, I can't, but you have provided zero evidence that this is particularly plausible or workable, and I've provided lots of real reasons to think it's especially unworkable.
I like the idea in principle, but I do not see it surviving a pubic choice critique. We have attempted to rate financial instruments and institutions and see heavy political pressure (e.g., when a government's bonds are downgraded). This will only be worse, even if we somehow get passed the what-ought-we-measure question.
We *might* have some hope to make an existing private rating of governments based on freedom, transparency, equality, etc. focal -- something analogous to US News rankings of colleges and universities. Or maybe one could convince an eccentric billionaire to fund a blue-ribbon panel to create a measure with sufficient fanfare and pomp to gain notice.
Overall, I see your previous suggestions to focus on prediction markets for outcomes we might wish to see as much more promising.
The Outside View certainly does imply that. In fact, I can't imagine any stronger evidence short of actually providing the supposed new kinds of info and them not using it...
Lots of ratings are based on objective data that voters are totally ignorant of. For example, perceptions of economic inequality within countries are deeply inaccurate, and yet statistics on them are easily accessed. Statistics on salaries also. Or what about health statistics, or local doctors/lawyers/hospitals' track records? When they are available, they aren't used. The Outside View suggests that since voters and immigrants don't use them now, they won't use them in the future either. They don't care. And for good reason, given what goes into making decisions, the S-factor, and public knowledge of most of these topics.
Ratings by Moodys arguably are useless. What new information does Moodys access? If nothing, then by EMH, they are useless because they don't contain any information that the market & everyone don't already know. Consumer Reports, on the other hand, does actual testing of products for which there is rapid turnover and illiquid markets (and even there, despite reading decades of CR issues - I think my insurance company sends me them, because certainly I never subscribed to CR myself - I struggle to think of any decisions I actually made in part or in whole because of CR testing).
Yes of course there are many rankings. But that's not the same as a rating based on objective data not known to local voters. Yes of course voters already know their own perceptions, but do they know how much money local politicians made last year over their salaries? Would you similarly argue that ratings by Moody's or Consumer Reports are useless, because everyone already knows that stuff too?
There are a lot of international rankings. Freedom House, the UN, World Bank, the Economist's, etc. For the most part, they go largely ignored except as clickbait factoids useful in partisan arguments ('did you know the USA is ranked only 20th in the world for maternal mortality?') and once in a great while, something like the Corruption Index or 'ease of doing business' rating will shame a Third World government into improving things (or perhaps more accurately, trying to game the rankings). Things get measured, but doesn't seem much like they get done. However, most people are profoundly uninterested in the rankings (except inasmuch as it's useful for partisan arguments online, of course), and I've never heard of someone exclaiming 'did you know Singapore is #1 on the just-out 2017 ease of doing business index? I'm going to move there right away!', and I've certainly never heard of, say, subsaharan African migrants poring over the countless existing rankings to decide whether to accept Merkel's invitation or head for Sweden instead.And they aren't necessarily new: an individual hardly needs a corruption perceptions index (most of the corruption indexes are just based on survey opinions) to know if people perceive a country as corrupt, since they already perceive a country as corrupt or not. So no one really uses them.
There is also the problem that most rankings are completely redundant. As Thorndike observed, "in human nature good traits go together" ( https://www.gwern.net/Every... ); he in fact examined exactly this problem by compiling a large number of metrics about American cities' amenities and lifestyles and economic variables and health and crime etc (_Your City_/_144 Smaller Cities_) and finds that, unsurprisingly, they all correlate strongly and there is a general factor of city quality, which we'd just call socioeconomic status (or intelligence). This is also true of international rankings. You can always predict what they will be, no matter what the variable is: roughly, the Scandinavian countries will be at the top, the USA/UK and richer parts of EU and East Asia will be underneath them, the oil petrostates may or may not overperform (depending on whether it's a metric they can buy their way upwards on), the middle income countries will be below them, and then there will be a long underbelly of South American/African/South Asian countries, and the basketcases like North Korea will be at the bottom. Maternal mortality, per capita income, freedom of the press, kilograms of chocolate consumption per capita, electricity consumption, number of tourists, corruption perceptions, feminism... whatever. (I recall an article, whose link I can't find at this moment, that activists attempting to construct 'feminist' indexes of global rankings for shaming & activism purposes, are annoyed that their rankings are so redundant with existing ones of non-feminist metrics.) So if you know one variable, you know a lot about all of them. And to the extent that any country deviates from its overall SES factor, the residual may not be important to any decision making (not, of course, that anyone is using them for any real decision making but let's pretend this might happen at some point) and there is little VoI to knowing the discrepant index - if you've decided to relocate to Switzerland, what do you care if an index suggests that its carbon consumption per capita is actually 5 ranks higher than you thought? So, indexes are not very informative.
If you'd read the post, you'd see why those sort of indices aren't what I'm talking about.
There are already lots of indices like this. For example, Heritage has the Index of Economic Freedom. Transparency International has the Corruption Perceptions Index. I'll bet there are more indices for things like free speech, government stability, gender equality, you name it. You could aggregate these indices into an overall quality measure, or let people choose their own weights for various aspects of the index in order to figure out the optimal country for them or something like that.
BTW, one thing that can happen once you set up an index is that national governments will try to game the index, e.g. figure out how the Index of Economic Freedom is calculated and figure out how to increase their score in order to be more attractive to investors. Similar to college rankings I guess.
As I said in the post, the most valuable stats would be those under substantial control of a recent administration. Basic human factors are often not good for that.
Basic human factors could be used such as poverty, access to water, sanitation, healthcare, good food etc
The last thing we should do do right now is further entrench a misleading metric like GDP and rate our governments on that.
Humans, should of course be good at rating these human factors in their own lives but perhaps we are only aware of them relatively, i.e. how much it has increased or decreased recently and this has a disproportionate affect.
The point is to help voters evaluate their recent local government performance, not to help people rank nations overall.
You gave examples, some of which are already done and really quite well known and I'm surprised you didn't even try to justify why your corruption index would work better than all the existing corruption indexes or provide information which is not already in the S-factor. This post seems especially wheel-reinventing. I am also not sure there are *any* pieces of information which are relevant to migrants or emigrants and which are not already well known and predictable from the S-factor. The Value of Information is the decisions it changes; what decision would a subsaharan African migrant make differently about Germany given some new corruption index?
Can we agree that announcing relatively random pieces of info about states is unlikely to seem very useful to voters, but in the vast space of such things there are probably some that would be influential and useful? I listed some specific things I thought were promising but you didn't respond to those, you just pointed out other things you didn't think promising. We can agree that there are many things that don't seem promising.
Very rarely. As consistent with the Outside View. And you are ignoring all of my other points as well in favor of making the extremely weak 'well it's *possible* isn't it? you can't prove it isn't possible!' claim. (This is an argument you are usually extremely strongly opposed to, and I recall you mocking it with reference to time travel etc just a few days ago on Twitter...) No, I can't, but you have provided zero evidence that this is particularly plausible or workable, and I've provided lots of real reasons to think it's especially unworkable.
I like the idea in principle, but I do not see it surviving a pubic choice critique. We have attempted to rate financial instruments and institutions and see heavy political pressure (e.g., when a government's bonds are downgraded). This will only be worse, even if we somehow get passed the what-ought-we-measure question.
We *might* have some hope to make an existing private rating of governments based on freedom, transparency, equality, etc. focal -- something analogous to US News rankings of colleges and universities. Or maybe one could convince an eccentric billionaire to fund a blue-ribbon panel to create a measure with sufficient fanfare and pomp to gain notice.
Overall, I see your previous suggestions to focus on prediction markets for outcomes we might wish to see as much more promising.
All through history most everyone has neglected info that they've had. Yet we have in fact found new useful info sources.
The Outside View certainly does imply that. In fact, I can't imagine any stronger evidence short of actually providing the supposed new kinds of info and them not using it...
Just because there is info that many people don't use doesn't imply that it might not help to provide new kinds of info.
Lots of ratings are based on objective data that voters are totally ignorant of. For example, perceptions of economic inequality within countries are deeply inaccurate, and yet statistics on them are easily accessed. Statistics on salaries also. Or what about health statistics, or local doctors/lawyers/hospitals' track records? When they are available, they aren't used. The Outside View suggests that since voters and immigrants don't use them now, they won't use them in the future either. They don't care. And for good reason, given what goes into making decisions, the S-factor, and public knowledge of most of these topics.
Ratings by Moodys arguably are useless. What new information does Moodys access? If nothing, then by EMH, they are useless because they don't contain any information that the market & everyone don't already know. Consumer Reports, on the other hand, does actual testing of products for which there is rapid turnover and illiquid markets (and even there, despite reading decades of CR issues - I think my insurance company sends me them, because certainly I never subscribed to CR myself - I struggle to think of any decisions I actually made in part or in whole because of CR testing).
Yes of course there are many rankings. But that's not the same as a rating based on objective data not known to local voters. Yes of course voters already know their own perceptions, but do they know how much money local politicians made last year over their salaries? Would you similarly argue that ratings by Moody's or Consumer Reports are useless, because everyone already knows that stuff too?
There are a lot of international rankings. Freedom House, the UN, World Bank, the Economist's, etc. For the most part, they go largely ignored except as clickbait factoids useful in partisan arguments ('did you know the USA is ranked only 20th in the world for maternal mortality?') and once in a great while, something like the Corruption Index or 'ease of doing business' rating will shame a Third World government into improving things (or perhaps more accurately, trying to game the rankings). Things get measured, but doesn't seem much like they get done. However, most people are profoundly uninterested in the rankings (except inasmuch as it's useful for partisan arguments online, of course), and I've never heard of someone exclaiming 'did you know Singapore is #1 on the just-out 2017 ease of doing business index? I'm going to move there right away!', and I've certainly never heard of, say, subsaharan African migrants poring over the countless existing rankings to decide whether to accept Merkel's invitation or head for Sweden instead.And they aren't necessarily new: an individual hardly needs a corruption perceptions index (most of the corruption indexes are just based on survey opinions) to know if people perceive a country as corrupt, since they already perceive a country as corrupt or not. So no one really uses them.
There is also the problem that most rankings are completely redundant. As Thorndike observed, "in human nature good traits go together" ( https://www.gwern.net/Every... ); he in fact examined exactly this problem by compiling a large number of metrics about American cities' amenities and lifestyles and economic variables and health and crime etc (_Your City_/_144 Smaller Cities_) and finds that, unsurprisingly, they all correlate strongly and there is a general factor of city quality, which we'd just call socioeconomic status (or intelligence). This is also true of international rankings. You can always predict what they will be, no matter what the variable is: roughly, the Scandinavian countries will be at the top, the USA/UK and richer parts of EU and East Asia will be underneath them, the oil petrostates may or may not overperform (depending on whether it's a metric they can buy their way upwards on), the middle income countries will be below them, and then there will be a long underbelly of South American/African/South Asian countries, and the basketcases like North Korea will be at the bottom. Maternal mortality, per capita income, freedom of the press, kilograms of chocolate consumption per capita, electricity consumption, number of tourists, corruption perceptions, feminism... whatever. (I recall an article, whose link I can't find at this moment, that activists attempting to construct 'feminist' indexes of global rankings for shaming & activism purposes, are annoyed that their rankings are so redundant with existing ones of non-feminist metrics.) So if you know one variable, you know a lot about all of them. And to the extent that any country deviates from its overall SES factor, the residual may not be important to any decision making (not, of course, that anyone is using them for any real decision making but let's pretend this might happen at some point) and there is little VoI to knowing the discrepant index - if you've decided to relocate to Switzerland, what do you care if an index suggests that its carbon consumption per capita is actually 5 ranks higher than you thought? So, indexes are not very informative.
Ideally it would be voters, but I'm open to many other possibilities. Media buyers would be nearly as good.