“This is the best of all possible worlds” can be said by an optimist or a pessimist, for the former with emphasis on “best,” for the latter on “possible.”
As I'm sure you know but didn't mention, institutions and even more, systems of institutions, can and often are trapped at local optimums. A coordinated shift to a better system can be very hard to achieve.
Economics seems like it ties into evolution. Someone tries a new business model, and if it works, it sticks around, maybe obsoleting others. Evolutionary theory can’t yet come up with a way of computing how long a process should take. The same thing seems to apply in economics: isn’t it possible that there really could be better institutions, but they haven’t been made yet? The old saw about an economist rejecting the idea that there’s money on the sidewalk (because someone would have picked it up by now) seems to apply here as well.
I think you can rescue all of these by seeing them as like, what the system tends to in the limit over time.
Biological evolution selects for reproductive success, firm evolution for profitability. It isn't clear what characteristic of a society is being selected for.
Over very long time frames, I think it’s something like “fostering prosperity and not chaos or stagnation.” This is what I think makes the “social pendulum” swing.
If you’re conquered, you can’t be prosperous, your stuff will get taken from you. So prosperity requires that you don’t lose. If you could theoretically conquer others but don’t maintain prosperity, I think you’ll fall apart internally.
The only viable path to keep existing in the future is, I think, to keep growing in prosperity. In economic situations that aren’t positive sum, people get real nasty. National defense is an instrumental subgoal for sure. But I’d say that also depends on prosperity.
* More prosperity means better and more weaponry and mercenaries for defensive purposes. Offensive costs more, and empires have a cost too, especially when they are more primitive
* More prosperity increases the chances of paying the Danegeld
Just playing devil's advocate and it feels like a shoddy job of it.
>>> If finding out which of two brands of car you should choose costs more time and money, perhaps renting and driving both, than the information is worth, it is rational to choose at random.
It seems to me most often the case that what people end up doing isn't choosing at random, but choosing based on tribal signalling. In fact, even when people do go through the effort of researching a future purchase decision, tribal signalling is there in the subconscious with a heavy thumb on the scales.
People end up buying the car that they think will most help them look like they fit into a certain category, or avoid some other one, rather than the car that would actually serve them best as a reliable/efficient/capable machine to get from place to place in.
I believe this is the whole reason why Stellantis exists! How many other companies, how many whole sectors of the economy, can be attributed to this phenomenon?
I can see how from a satellite view it might look to an economist like people are rationally pursuing their own ends, even if rational irrationalism is one path getting there, but knowing something about these products and how they're made and marketed, and knowing something about how so many people manage (actually, mismanage) their finances, and paying attention to the social posture so many people wear on their shirtsleeves, I think my tribal signalling explanation is more powerful.
"People end up buying the car that they think will most help them look like they fit into a certain category, or avoid some other one, rather than the car that would actually serve them best as a reliable/efficient/capable machine to get from place to place in."
If fitting into that category is important to them, isn't that rational — especially if alternative cars vary a lot in that dimension, only a little in how well they drive?
Rationality is an assumption about individuals, not groups. Every member of the tribe might be better off if they identified with a better car but each member is better off buying the car the others identify with.
English drivers might be better off if they all drove on the right but any driver who did would be worse off.
There are in fact individual drivers who import cars designed to be driven on the other side. Within the country to which the cars are imported, those cars often appeal to an enthusiast niche who are willing to deal with the headaches of having to jump through extra regulatory hoops, of not being able to find certain spare parts as easily, and maybe of not being able to use drive-through windows or something. If such importations were not done by a few enthusiasts but by a substantial chunk of the population, many of those inconveniences might be reduced as well.
Now, I assume there would be more car accidents if 50% of drivers were driving right-hand drive cars and 50% had their steering wheels on the left, but if everyone in that situation at least had their steering wheels where they thought it important to have them, would that be a situation in which everyone is being rational?
If rationality is simply "pursuing whatever is important to you" without regard to whether that "whatever" is actually *good* in some objective sense, then it seems rationality is being treated (by economists) more just as a measure of "how readily can people fulfill their every whim" than what I'd think of as its more common definition, something like "considered behavior resulting from an accounting of pros and cons based on available information".
Maybe economists *do* share my definition but conclude that if a person has only one thought in his head, and that is "does this product help me fit into the group I fancy myself part of", then that counts as "available information"? As George Carlin said, someone with a 100 IQ is pretty dumb, and half of all people are dumber than that! But that doesn't seem aligned with the rest of what you said. Besides, the "accounting of pros and cons" part seems important: it implies a comparison of A to B--a *ratio*, which is where the word "rational" comes from.
In a modern society not ruled by warring tribal factions but by equality of individuals before the law, how rational is it to base so many of one's decisions about what to buy, what to think, what ends to pursue, with whom to associate, etc. on signalling their group identity--especially when this is happening subconsciously, without an explicit rational process?
I was thinking the same thing, that it would seem very odd to claim that South Korea and North Korea both have optimal institutions, considering how different their institutions are.
But given that its “unique situation” is serving the interests of Fat Face, its dictator, the single biggest of which is his ability to remain in power?
The problem is always defining what 'better' means. Better for whom? In a conflict, there are at least two different interpretations of what 'better' means. The (utilitarian) economist now attempts to reconcile these conflicting perspectives objectively in order to arrive at the 'best' solution (i.e. the one that provides the greatest benefit). This is impossible to do neutrally.
Obtaining information is part of the subjective utility calculus. At some point, the actor stops gathering more information. Why should the economist continue to gather information for them? Like the politician, the economist is an actor with opportunity costs who wants to be paid and remunerated, and pursues his own interests. Information is not neutral, but is placed in a chosen context. Context is always just a part of the world, but never the whole thing. Without context, information is impossible. Information remains speculative. Information can be false.
Take tariffs or money supply expansion, for example.
There will always be winners and losers as a result of such political interventions. How could one objectively determine which result would be better for the collective as a whole? There will always be division.
Ultimately, an economist (with his inevitable personal bias) can only influence how people see things , just like any other writer, filmmaker, artist, philosopher or politician. Here in Germany, leading economists tell biased tales that you can choose to believe or not. Whose advice is correct?
In short, 'better' is not an objective metric. It depends on perspective -- just like rationality. There are billions of rationalities out there.
You can decide which economist to believe by examining the internal logic of the argument. Once you understand comparative advantage, some common arguments about tariffs no longer make sense. That does not, by itself, tell you whether tariffs are desirable but it could easily change your opinion on the subject.
It isn't as simple as "support the policies that benefit you" because what policies benefit whom is one of the things that an economic analysis might show you. In the case of tariffs, some people are made worse off who do not realize it and some people believe they are made better off but are mistaken
We do not live in the best of possible worlds because a finite population can make a finite number of efficiency improvements to a finite count of areas.
My preferred way to understand institutions is similar to the way I understand biological features produced by evolution - ways of dealing with something or other that worked well enough to persist. The situation has probably changed somewhat since they evolved, and they may never have been optimum. All they ever needed to be was "good enough".
When it comes to institutions, they need not be good enough <em>for everyone</em>. As a trite example, slavery may persist even when everyone agrees that it's not good for the slaves.
One implication: the institutions we have today are unlikely to be optimum.
On the other hand, tossing them out and replacing them with something random, or the latest fad, probably won't help - particularly if we don't understand all the constraints and benefits.
I'm not sure whether economists can be particularly helpful in suggesting improvements that would actually improve things. They tend to be far too prone to discounting anything they can't measure in monetary terms. And they often persist in seeing people as disconnected individuals, in ways that don't appear to me to mirror the real world.
But extra ideas are always useful, even if they don't work.
"They tend to be far too prone to discounting anything they can't measure in monetary terms."
What sorts of things are you thinking of? Anything people value can, conceptually, be measured by what someone would give up to get it. There may sometimes be no way to learn what that value is, but that's true of lots of material things as well.
Individuals obviously do not "tend to take the actions that best serve their objectives", even with complete information, because processing the information can be often beyond their capabilities. In any given state of a chess board, both players have access to all information just by looking at the board. But no player will *always* make the best possible move. Because actually processing that information is too difficult.
The same goes for many other decisions in life. Even if a person were to have access to complete and perfect information about every car they could buy, or every career they could take, they would likely be unable to process it properly and would still make mistakes.
One of the advantages of capitalism is that it is easy to imitate people who are succesful, and the people who are most succesful gain more resources, and can then be succesful on a large scale. I think that makes capitalist societies tend towards being more rational. But individuals can still be very irrational, especially in new situations.
The first is "tend to." The assumption is not that people always make the right choice but that assuming they do works better than any alternative, at least for random strangers, that we have no theory of irrationality with as much predictive power as the assumption of rationality.
The second is that the same approach I applied to information cost can, in principle, be applied to processing costs.
Kahneman's _Thinking Fast and Slow_ can be seen either as a theory of irrationality or as a theory of rationality expanded to include processing costs.
Are you claiming that people do not tend to take the actions that best serve their objectives, or are you going further and saying they can't take the actions that best serve their objectives? I can't tell how strong your claim is, although it seems pretty strong.
Seems like a faulty assumption to me. The very fact that people survive and thrive seems a good refutation of them not being smart enough to make good choices. "Best choice" is a bad term since it assumes clarity in what is best versus merely good enough. "Very far from the rational ideal" also assumes it is easy to objectively establish "the" rational ideal.
Yes, I did get that. What I am getting at is what that means operationally. In your model, do people manage to act that are only marginally related to achieving their goals, or tend to act in ways entirely contrary to their goals, or in ways so contradictory to their goals that knowing their goals is no guide to their future behavior?
Relatedly, the question is also what the nature of people's goals is. How specific and defined. Your car example comes to mind, as even with all that information if the goal is merely "have a reliable car that looks pretty nice and is comfortable and gets me where I am going for not too much money" there are likely wide ranges of indifference between cars, all of which are good enough. A much more highly specific goal requires more optimization and mental load to achieve.
I have thought about a variety of alternative mechanisms but I don’t think any of them solve the problem. An individual selected by sortition has no special competence and, before he is selected, no more reason than a voter in our system to acquire it. If the vote is shortly after sortition he has little time to change that. If there is a long gap, that’s an opportunity for interested parties to get at delegates and influence them. Sortition offers one more opportunity for the system to be corrupted, by control over who is chosen. And if the number chosen is small it introduces a source of random error.
I have a preferred system but it is preferred mostly on the grounds of elegance, not because I think it would produce substantially better results.
David, You are right: "An individual selected by sortition has no special competence".
But your observation masks the heart of the matter: **THE PROBLEM OF THE INITIAL POOL** in random choice. True randomness cannot preclude the selection of a legislature composed entirely of incompetents.
*** As you might expect, I HAVE SOLVED THIS PROBLEM.***
Venice used sortition for 1100 years (697-1797). Their astounding success was from their solution to the PROBLEM OF THE INITIAL POOL. How? The initial pool was all aristocrats, all carefully vetted and watched, with propertied skin in the game. (Daniel J. Smith has documented the system.)
My solution: Have each of the 3,144 counties in the USA impose its rules for its citizens seeking to enter the initial pool. Their rules can be TOTALLY ARBITRARY. For example, counties in Utah might require each entry to be Mormon. Other counties may even (outrageously) exclude women or blacks. In that case, the opprobrium is so great that none will do so, but we allow it. In practice, these rules will guarantee the standard of excellence determined by the county, as each avoids offensive rules and strives to pick the best, as their lights see it. Randomness will weed out the absurdities and will skew the final selections to favor reasoned excellence.
Why is it in the interest of the political institutions of the county to select someone who will vote for the candidate best for the nation?
Other alternatives:
The voter will be a target for bribes, so choose someone you want to benefit or someone who will give a cut to the county legislators who supported rules that favor him.
The voter can trade his votes for favors for his county so choose someone who will be good at doing so.
There will still be political parties. Each county arranges to appoint people of its preferred party, so sortition ends up equivalent to the electoral vote system by counties.
BONUS! from non-stratified sortition: NO GERRYMANDERING.
A county whose population gives them several House Representatives could not skew its House district map: All of their allotted Reps would be in the one random pool, and that number would be randomly drawn for House membership.
Only non-stratified sortition ANNIHILATES elective politicking and all its ills. E.g.: No more billions of dollars spent to promote a particular candidate (the pool is too large to buy off all of them); no more bogus campaign laws with their associated lawfare; no more back-room cabals at the State and national level; no more demagogic media poseurs; no more years-long campaigning circuses; no more beggar-thy-neighbor political handouts; no more of the lying bill of goods as represented by a so-called political platform; no more manipulation of votes; no more disputes over voting rights or voting access; no more voter fraud; no more of the shibboleth of "representation," whereby skin color and gender outweigh the content of one's character.
You say: "Why is it in the interest of the political institutions of the county to select someone who will vote for the candidate best for the nation?"
>> One word why: Subsidiarity. The "national interest" is a fiction when divorced from the interest of each county. Virtually no voter is competent to judge nation accounting (e.g., to know the folly of tariffs and fiat money printing), but counties, once cognizant that their duty is suddenly meaningful, will not establish standards of plunder; they will establish standards of excellence that pre-emptively exclude the media poseurs and demagogues of the current day.
You say: "The voter can trade his votes for favors".
>> Categorically false. The size of the electorate, as well as of the initial pool, makes it impossible to bribe so large a number. Sortition – note well, WITHOUT demographic stratification – ANNIHILATES electoral politics and all its ills.
You say: "There will still be political parties."
>> Categorically false. There are essentially TWO party platforms (by Duverger's Law) in the current Uniparty electoral scheme. Under my proposal, there will be essentially 3,144 platforms (one for each county) – composed not of a beggar-thy-neighbor bill of goods, but of a standard of excellence.
You might get a giggle of my Chartertopia system. Each district elects three representatives, who cast as many votes in the legislature as they won in the election; I call it proxying. This makes them much more representative, for what it's worth, and also makes stuffing the ballot box much less useful, since a handful of votes is just a handful of votes and cannot change the 2nd place lose-everything candidate to the 1st place winner-takes-all candidate. It also makes it a lot harder for politicians and pundits to predict votes and how vote-swapping affects bill passage. It also encourages people to vote even if the polls put their guy way back in the pack, since if he wins, say, third, he'll get all their votes; and if he loses, the amateur will get more votes.
But the sortition angle is what you reminded me of. Every voter can throw a name in the ballot box, and one is drawn at random and proxies all the remaining election votes. I call him the amateur. He serves several purposes. First, being a voter, he probably voted for one of the three winners, but there's a decent chance not, and that gives him a better chance of representing everyone who did not vote for a winner. Second, all bills have to pass by 2/3 majority in both proxies and head count, so the professionals have a hard time ignoring the amateur.
I don't know how many amateurs would actually want to be a representative and have to uproot themselves for a full term. Can they afford to take that time off from their job, will their family put up with it? But if there are 10,000 voters in an election, there's bound to be someone who accepts the job.
Interesting proposal. However, all such schemes fail for 2 reasons: They preserve ELECTIVE POLITICS, and they preserve the MYTH OF REPRESENTATION.
In his 2008 book, Gaming the Vote, William Poundstone has cataloged many of these schemes. He doesn't cover recent malignancies like AADV (Approve, Approve, Disapprove Voting), BAWV (Best, Alternate, Worst Voting), the D21 Janeček method, blah, blah, blah, yadda, yadda – the schemes are legion and all doomed to fail for the 2 reasons just given.
**ONLY SORTITION ANNIHILATES ELECTIVE POLITICS.**
By sortition I do NOT mean the modern variants such as James Fishkin’s Deliberative Polling, the 2015 Irish Constitutional Convention (on marriage for sodomites), the 2018 Irish Citizens’ Assembly (on abortion), the French citizens’ climate assembly of 2019-2020, or Brett Hennig’s UK Climate Assembly of 2019-2020. All of these FAIL because they address the PROBLEM OF THE INITIAL POOL by "tweaking" it to assure "representation" (for gender, race, wealth, for left-handed red-heads[?]).
**BUT REPRESENTATION IS A MYTH.** Does the black Hank Johnson, who fears that a ship load of Marines will capsize the island of Guam, best "represent" his majority black Georgia district? Does the black Sheila Jackson Lee, who thinks that the moon is made up "mostly of gases" (and who astoundingly sits on the House Science Committee’s space subcommittee), best "represent" her majority black Texas district?
**SPECIFICALLY, REPRESENTATION BY DEMOGRAPHIC STRATIFICATION IS A MYTH.**
My system is that each voter chooses a candidate to represent him, the candidate casts a number of votes in Congress equal to the number of voters he represents, and a voter can change his candidate at any time. If Congress is still conducted in realspace, any representative with more than X votes gets a seat and representatives with too few votes can combine to get a shared seat, with any rule they agree on to determine who occupies it when.
I've thought of similar schemes, but couldn't think of any way to have anonymous voters. Being able to change their vote means needing credentials. Maybe modern crypto methods can get around that. The system would be a high priority hacker target.
One of the other tricks in my legislature is that every district selects their own election day, so legislators still have terms but the concept of sessions vanishes.
An entirely different scheme is to treat votes like money -- they can only be cast once in the legislature. If a legislator begins the term with 100,000 votes, he could use them all at once on a bill he really likes, or spread them out over the term. It would sure be more interesting, but I think a terrible idea when a small bloc husbands their votes until the end of session when almost everybody else has very few votes left, then introduces and passes some atrocious legislation.
Everyone shows the evidence that he is a voter, gets to roll up a private key/public key pair. The public key is registered with the electoral authorities as the public key of a voter, but not which voter. Instructions are signed using the private key, which only he has.
David, this can't possibly work. You have absolutely no standard of excellence informing the voter in his selection of his representative. This scheme would turn into a contest of the meretricious, with media phenoms ruling the electorate – a total carnival of freaks riding their 15-minutes-of-fame. Welcome JoJo Siwa as Speaker of the House, and RuPaul as Majority Leader of the Senate!
I didn't say it would work, I said, in another comment, that I didn't think any version of democracy solved its problems. Your comment is true of the present system as well.
As I'm sure you know but didn't mention, institutions and even more, systems of institutions, can and often are trapped at local optimums. A coordinated shift to a better system can be very hard to achieve.
Economics seems like it ties into evolution. Someone tries a new business model, and if it works, it sticks around, maybe obsoleting others. Evolutionary theory can’t yet come up with a way of computing how long a process should take. The same thing seems to apply in economics: isn’t it possible that there really could be better institutions, but they haven’t been made yet? The old saw about an economist rejecting the idea that there’s money on the sidewalk (because someone would have picked it up by now) seems to apply here as well.
I think you can rescue all of these by seeing them as like, what the system tends to in the limit over time.
Biological evolution selects for reproductive success, firm evolution for profitability. It isn't clear what characteristic of a society is being selected for.
Over very long time frames, I think it’s something like “fostering prosperity and not chaos or stagnation.” This is what I think makes the “social pendulum” swing.
Why does prosperity beat ability to conquer other polities and not be conquered yourself?
If you’re conquered, you can’t be prosperous, your stuff will get taken from you. So prosperity requires that you don’t lose. If you could theoretically conquer others but don’t maintain prosperity, I think you’ll fall apart internally.
The only viable path to keep existing in the future is, I think, to keep growing in prosperity. In economic situations that aren’t positive sum, people get real nasty. National defense is an instrumental subgoal for sure. But I’d say that also depends on prosperity.
* More prosperity means better and more weaponry and mercenaries for defensive purposes. Offensive costs more, and empires have a cost too, especially when they are more primitive
* More prosperity increases the chances of paying the Danegeld
Just playing devil's advocate and it feels like a shoddy job of it.
>>> If finding out which of two brands of car you should choose costs more time and money, perhaps renting and driving both, than the information is worth, it is rational to choose at random.
It seems to me most often the case that what people end up doing isn't choosing at random, but choosing based on tribal signalling. In fact, even when people do go through the effort of researching a future purchase decision, tribal signalling is there in the subconscious with a heavy thumb on the scales.
People end up buying the car that they think will most help them look like they fit into a certain category, or avoid some other one, rather than the car that would actually serve them best as a reliable/efficient/capable machine to get from place to place in.
I believe this is the whole reason why Stellantis exists! How many other companies, how many whole sectors of the economy, can be attributed to this phenomenon?
I can see how from a satellite view it might look to an economist like people are rationally pursuing their own ends, even if rational irrationalism is one path getting there, but knowing something about these products and how they're made and marketed, and knowing something about how so many people manage (actually, mismanage) their finances, and paying attention to the social posture so many people wear on their shirtsleeves, I think my tribal signalling explanation is more powerful.
"People end up buying the car that they think will most help them look like they fit into a certain category, or avoid some other one, rather than the car that would actually serve them best as a reliable/efficient/capable machine to get from place to place in."
If fitting into that category is important to them, isn't that rational — especially if alternative cars vary a lot in that dimension, only a little in how well they drive?
Rationality is an assumption about individuals, not groups. Every member of the tribe might be better off if they identified with a better car but each member is better off buying the car the others identify with.
English drivers might be better off if they all drove on the right but any driver who did would be worse off.
There are in fact individual drivers who import cars designed to be driven on the other side. Within the country to which the cars are imported, those cars often appeal to an enthusiast niche who are willing to deal with the headaches of having to jump through extra regulatory hoops, of not being able to find certain spare parts as easily, and maybe of not being able to use drive-through windows or something. If such importations were not done by a few enthusiasts but by a substantial chunk of the population, many of those inconveniences might be reduced as well.
Now, I assume there would be more car accidents if 50% of drivers were driving right-hand drive cars and 50% had their steering wheels on the left, but if everyone in that situation at least had their steering wheels where they thought it important to have them, would that be a situation in which everyone is being rational?
If rationality is simply "pursuing whatever is important to you" without regard to whether that "whatever" is actually *good* in some objective sense, then it seems rationality is being treated (by economists) more just as a measure of "how readily can people fulfill their every whim" than what I'd think of as its more common definition, something like "considered behavior resulting from an accounting of pros and cons based on available information".
Maybe economists *do* share my definition but conclude that if a person has only one thought in his head, and that is "does this product help me fit into the group I fancy myself part of", then that counts as "available information"? As George Carlin said, someone with a 100 IQ is pretty dumb, and half of all people are dumber than that! But that doesn't seem aligned with the rest of what you said. Besides, the "accounting of pros and cons" part seems important: it implies a comparison of A to B--a *ratio*, which is where the word "rational" comes from.
In a modern society not ruled by warring tribal factions but by equality of individuals before the law, how rational is it to base so many of one's decisions about what to buy, what to think, what ends to pursue, with whom to associate, etc. on signalling their group identity--especially when this is happening subconsciously, without an explicit rational process?
So North Korea has the best system for its unique situation?
I was thinking the same thing, that it would seem very odd to claim that South Korea and North Korea both have optimal institutions, considering how different their institutions are.
Literally best? Perhaps not.
But given that its “unique situation” is serving the interests of Fat Face, its dictator, the single biggest of which is his ability to remain in power?
It’s likely up there…
I've heard your subhead phrased as "An optimist believes this is the best of all possible worlds; a pessimist fears that this is true."
The problem is always defining what 'better' means. Better for whom? In a conflict, there are at least two different interpretations of what 'better' means. The (utilitarian) economist now attempts to reconcile these conflicting perspectives objectively in order to arrive at the 'best' solution (i.e. the one that provides the greatest benefit). This is impossible to do neutrally.
Obtaining information is part of the subjective utility calculus. At some point, the actor stops gathering more information. Why should the economist continue to gather information for them? Like the politician, the economist is an actor with opportunity costs who wants to be paid and remunerated, and pursues his own interests. Information is not neutral, but is placed in a chosen context. Context is always just a part of the world, but never the whole thing. Without context, information is impossible. Information remains speculative. Information can be false.
Take tariffs or money supply expansion, for example.
There will always be winners and losers as a result of such political interventions. How could one objectively determine which result would be better for the collective as a whole? There will always be division.
Ultimately, an economist (with his inevitable personal bias) can only influence how people see things , just like any other writer, filmmaker, artist, philosopher or politician. Here in Germany, leading economists tell biased tales that you can choose to believe or not. Whose advice is correct?
In short, 'better' is not an objective metric. It depends on perspective -- just like rationality. There are billions of rationalities out there.
You can decide which economist to believe by examining the internal logic of the argument. Once you understand comparative advantage, some common arguments about tariffs no longer make sense. That does not, by itself, tell you whether tariffs are desirable but it could easily change your opinion on the subject.
It isn't as simple as "support the policies that benefit you" because what policies benefit whom is one of the things that an economic analysis might show you. In the case of tariffs, some people are made worse off who do not realize it and some people believe they are made better off but are mistaken
We do not live in the best of possible worlds because a finite population can make a finite number of efficiency improvements to a finite count of areas.
My preferred way to understand institutions is similar to the way I understand biological features produced by evolution - ways of dealing with something or other that worked well enough to persist. The situation has probably changed somewhat since they evolved, and they may never have been optimum. All they ever needed to be was "good enough".
When it comes to institutions, they need not be good enough <em>for everyone</em>. As a trite example, slavery may persist even when everyone agrees that it's not good for the slaves.
One implication: the institutions we have today are unlikely to be optimum.
On the other hand, tossing them out and replacing them with something random, or the latest fad, probably won't help - particularly if we don't understand all the constraints and benefits.
I'm not sure whether economists can be particularly helpful in suggesting improvements that would actually improve things. They tend to be far too prone to discounting anything they can't measure in monetary terms. And they often persist in seeing people as disconnected individuals, in ways that don't appear to me to mirror the real world.
But extra ideas are always useful, even if they don't work.
"They tend to be far too prone to discounting anything they can't measure in monetary terms."
What sorts of things are you thinking of? Anything people value can, conceptually, be measured by what someone would give up to get it. There may sometimes be no way to learn what that value is, but that's true of lots of material things as well.
Individuals obviously do not "tend to take the actions that best serve their objectives", even with complete information, because processing the information can be often beyond their capabilities. In any given state of a chess board, both players have access to all information just by looking at the board. But no player will *always* make the best possible move. Because actually processing that information is too difficult.
The same goes for many other decisions in life. Even if a person were to have access to complete and perfect information about every car they could buy, or every career they could take, they would likely be unable to process it properly and would still make mistakes.
One of the advantages of capitalism is that it is easy to imitate people who are succesful, and the people who are most succesful gain more resources, and can then be succesful on a large scale. I think that makes capitalist societies tend towards being more rational. But individuals can still be very irrational, especially in new situations.
There are two answers to your point.
The first is "tend to." The assumption is not that people always make the right choice but that assuming they do works better than any alternative, at least for random strangers, that we have no theory of irrationality with as much predictive power as the assumption of rationality.
The second is that the same approach I applied to information cost can, in principle, be applied to processing costs.
Kahneman's _Thinking Fast and Slow_ can be seen either as a theory of irrationality or as a theory of rationality expanded to include processing costs.
Are you claiming that people do not tend to take the actions that best serve their objectives, or are you going further and saying they can't take the actions that best serve their objectives? I can't tell how strong your claim is, although it seems pretty strong.
That in practice individuals perform very far from the rational ideal, because they are not smart enough to identify which choice is the best choice
Seems like a faulty assumption to me. The very fact that people survive and thrive seems a good refutation of them not being smart enough to make good choices. "Best choice" is a bad term since it assumes clarity in what is best versus merely good enough. "Very far from the rational ideal" also assumes it is easy to objectively establish "the" rational ideal.
Yes, I did get that. What I am getting at is what that means operationally. In your model, do people manage to act that are only marginally related to achieving their goals, or tend to act in ways entirely contrary to their goals, or in ways so contradictory to their goals that knowing their goals is no guide to their future behavior?
Relatedly, the question is also what the nature of people's goals is. How specific and defined. Your car example comes to mind, as even with all that information if the goal is merely "have a reliable car that looks pretty nice and is comfortable and gets me where I am going for not too much money" there are likely wide ranges of indifference between cars, all of which are good enough. A much more highly specific goal requires more optimization and mental load to achieve.
So how are you defining goals here?
David,
Regarding your 'coarse control of the democratic system,' i.e., the failed tool of majoritarian voting: Have you ever explored sortition?
I have thought about a variety of alternative mechanisms but I don’t think any of them solve the problem. An individual selected by sortition has no special competence and, before he is selected, no more reason than a voter in our system to acquire it. If the vote is shortly after sortition he has little time to change that. If there is a long gap, that’s an opportunity for interested parties to get at delegates and influence them. Sortition offers one more opportunity for the system to be corrupted, by control over who is chosen. And if the number chosen is small it introduces a source of random error.
I have a preferred system but it is preferred mostly on the grounds of elegance, not because I think it would produce substantially better results.
David, You are right: "An individual selected by sortition has no special competence".
But your observation masks the heart of the matter: **THE PROBLEM OF THE INITIAL POOL** in random choice. True randomness cannot preclude the selection of a legislature composed entirely of incompetents.
*** As you might expect, I HAVE SOLVED THIS PROBLEM.***
Venice used sortition for 1100 years (697-1797). Their astounding success was from their solution to the PROBLEM OF THE INITIAL POOL. How? The initial pool was all aristocrats, all carefully vetted and watched, with propertied skin in the game. (Daniel J. Smith has documented the system.)
My solution: Have each of the 3,144 counties in the USA impose its rules for its citizens seeking to enter the initial pool. Their rules can be TOTALLY ARBITRARY. For example, counties in Utah might require each entry to be Mormon. Other counties may even (outrageously) exclude women or blacks. In that case, the opprobrium is so great that none will do so, but we allow it. In practice, these rules will guarantee the standard of excellence determined by the county, as each avoids offensive rules and strives to pick the best, as their lights see it. Randomness will weed out the absurdities and will skew the final selections to favor reasoned excellence.
Why is it in the interest of the political institutions of the county to select someone who will vote for the candidate best for the nation?
Other alternatives:
The voter will be a target for bribes, so choose someone you want to benefit or someone who will give a cut to the county legislators who supported rules that favor him.
The voter can trade his votes for favors for his county so choose someone who will be good at doing so.
There will still be political parties. Each county arranges to appoint people of its preferred party, so sortition ends up equivalent to the electoral vote system by counties.
BONUS! from non-stratified sortition: NO GERRYMANDERING.
A county whose population gives them several House Representatives could not skew its House district map: All of their allotted Reps would be in the one random pool, and that number would be randomly drawn for House membership.
Only non-stratified sortition ANNIHILATES elective politicking and all its ills. E.g.: No more billions of dollars spent to promote a particular candidate (the pool is too large to buy off all of them); no more bogus campaign laws with their associated lawfare; no more back-room cabals at the State and national level; no more demagogic media poseurs; no more years-long campaigning circuses; no more beggar-thy-neighbor political handouts; no more of the lying bill of goods as represented by a so-called political platform; no more manipulation of votes; no more disputes over voting rights or voting access; no more voter fraud; no more of the shibboleth of "representation," whereby skin color and gender outweigh the content of one's character.
You say: "Why is it in the interest of the political institutions of the county to select someone who will vote for the candidate best for the nation?"
>> One word why: Subsidiarity. The "national interest" is a fiction when divorced from the interest of each county. Virtually no voter is competent to judge nation accounting (e.g., to know the folly of tariffs and fiat money printing), but counties, once cognizant that their duty is suddenly meaningful, will not establish standards of plunder; they will establish standards of excellence that pre-emptively exclude the media poseurs and demagogues of the current day.
You say: "The voter can trade his votes for favors".
>> Categorically false. The size of the electorate, as well as of the initial pool, makes it impossible to bribe so large a number. Sortition – note well, WITHOUT demographic stratification – ANNIHILATES electoral politics and all its ills.
You say: "There will still be political parties."
>> Categorically false. There are essentially TWO party platforms (by Duverger's Law) in the current Uniparty electoral scheme. Under my proposal, there will be essentially 3,144 platforms (one for each county) – composed not of a beggar-thy-neighbor bill of goods, but of a standard of excellence.
You might get a giggle of my Chartertopia system. Each district elects three representatives, who cast as many votes in the legislature as they won in the election; I call it proxying. This makes them much more representative, for what it's worth, and also makes stuffing the ballot box much less useful, since a handful of votes is just a handful of votes and cannot change the 2nd place lose-everything candidate to the 1st place winner-takes-all candidate. It also makes it a lot harder for politicians and pundits to predict votes and how vote-swapping affects bill passage. It also encourages people to vote even if the polls put their guy way back in the pack, since if he wins, say, third, he'll get all their votes; and if he loses, the amateur will get more votes.
But the sortition angle is what you reminded me of. Every voter can throw a name in the ballot box, and one is drawn at random and proxies all the remaining election votes. I call him the amateur. He serves several purposes. First, being a voter, he probably voted for one of the three winners, but there's a decent chance not, and that gives him a better chance of representing everyone who did not vote for a winner. Second, all bills have to pass by 2/3 majority in both proxies and head count, so the professionals have a hard time ignoring the amateur.
I don't know how many amateurs would actually want to be a representative and have to uproot themselves for a full term. Can they afford to take that time off from their job, will their family put up with it? But if there are 10,000 voters in an election, there's bound to be someone who accepts the job.
Interesting proposal. However, all such schemes fail for 2 reasons: They preserve ELECTIVE POLITICS, and they preserve the MYTH OF REPRESENTATION.
In his 2008 book, Gaming the Vote, William Poundstone has cataloged many of these schemes. He doesn't cover recent malignancies like AADV (Approve, Approve, Disapprove Voting), BAWV (Best, Alternate, Worst Voting), the D21 Janeček method, blah, blah, blah, yadda, yadda – the schemes are legion and all doomed to fail for the 2 reasons just given.
**ONLY SORTITION ANNIHILATES ELECTIVE POLITICS.**
By sortition I do NOT mean the modern variants such as James Fishkin’s Deliberative Polling, the 2015 Irish Constitutional Convention (on marriage for sodomites), the 2018 Irish Citizens’ Assembly (on abortion), the French citizens’ climate assembly of 2019-2020, or Brett Hennig’s UK Climate Assembly of 2019-2020. All of these FAIL because they address the PROBLEM OF THE INITIAL POOL by "tweaking" it to assure "representation" (for gender, race, wealth, for left-handed red-heads[?]).
**BUT REPRESENTATION IS A MYTH.** Does the black Hank Johnson, who fears that a ship load of Marines will capsize the island of Guam, best "represent" his majority black Georgia district? Does the black Sheila Jackson Lee, who thinks that the moon is made up "mostly of gases" (and who astoundingly sits on the House Science Committee’s space subcommittee), best "represent" her majority black Texas district?
**SPECIFICALLY, REPRESENTATION BY DEMOGRAPHIC STRATIFICATION IS A MYTH.**
My system is that each voter chooses a candidate to represent him, the candidate casts a number of votes in Congress equal to the number of voters he represents, and a voter can change his candidate at any time. If Congress is still conducted in realspace, any representative with more than X votes gets a seat and representatives with too few votes can combine to get a shared seat, with any rule they agree on to determine who occupies it when.
I've thought of similar schemes, but couldn't think of any way to have anonymous voters. Being able to change their vote means needing credentials. Maybe modern crypto methods can get around that. The system would be a high priority hacker target.
One of the other tricks in my legislature is that every district selects their own election day, so legislators still have terms but the concept of sessions vanishes.
An entirely different scheme is to treat votes like money -- they can only be cast once in the legislature. If a legislator begins the term with 100,000 votes, he could use them all at once on a bill he really likes, or spread them out over the term. It would sure be more interesting, but I think a terrible idea when a small bloc husbands their votes until the end of session when almost everybody else has very few votes left, then introduces and passes some atrocious legislation.
Everyone shows the evidence that he is a voter, gets to roll up a private key/public key pair. The public key is registered with the electoral authorities as the public key of a voter, but not which voter. Instructions are signed using the private key, which only he has.
What prevents stuffing the virtual ballot box with public keys, since they are not verifiable once issued?
Sorry for being so dense. Maybe I'll figure it out by the morning....
David, this can't possibly work. You have absolutely no standard of excellence informing the voter in his selection of his representative. This scheme would turn into a contest of the meretricious, with media phenoms ruling the electorate – a total carnival of freaks riding their 15-minutes-of-fame. Welcome JoJo Siwa as Speaker of the House, and RuPaul as Majority Leader of the Senate!
I didn't say it would work, I said, in another comment, that I didn't think any version of democracy solved its problems. Your comment is true of the present system as well.