In a democracy, voters pick which candidates become government officials. In this context, we usually admire those whose efforts help voters to become better informed, and to vote more altruistically. After all, we generally praise altruism. And we see how much harm could from ignorant voters who have no idea what candidates might do, which actions would have what effects, and who would be helped or hurt by them.
A "simple selfish ignorant retrospective" voter as you describe them would be far BETTER informed than the average voter. The average voter is hypnotized by tribalist rhetoric that has no connection to empirical results at all.
You argue that the result would be skewed, but why would that skewedness necessarily be for the worse? "Maybe" it's better for policies to swing towards some kinds of voters on reasons other than numeric?
How would this apply to sortition-based democratic systems? My understanding is that the appeal of them is that they neither encourage nor benefit from voter awareness campaigns, with the idea being that the selected electors are expected to become educated on the candidates only after being chosen. It would seem that this circumvents the issue of voters getting sucked into coalitions but still involves them becoming much more educated on the election.
Could you point me towards any reading that elaborates on this? I'm quite interested in the concept and presently am considering integrating it into my work (governance for decentralized autonomous organizations). But I have not found any robust arguments against it so far, the people at https://democracywithoutelections.org/ seem pretty convincing.
The objection here is specifically to *partial* altruism (i.e. towards a mere subset of other voters), right? So one solution would be to stress that "altruistic" voters are only good if they are *impartial* in their altruism.
1) Maximal Short Term Selfishness Voting (Third World, Black Machine Politics Urban Government)
2) Maximum Altruism (Woke Philosopher Kings trying to built perfect society)
Each fails, both because they make coordination impossible. The first rather directly. The second because accurate information on peoples preferences is obscured by altruism and signaling.
The ideal would be someone that tried to show some altruism towards things they know and some degree of long term planning, without descending into ideological fancy.
The main virtue of democracy is that it provides feedback when political leaders fail badly and provides for a peaceful transfer of power. Narrow political power is subject to insider abuse that can get out of hand because the costs can be externalized beyond the narrow political power players with impunity.
Yes, you can construct some models where creating more informed voters are bad. But it's easy to see that purely retrospective voting gets you into lots of trouble.
For instance, the purely retrospective voter will always favor the canidate who raids future benefits to offer benefits now (eg enter into long term deals with costs in the future once out of office).
Clearly to avoid this kind of impact you want at least some intelligence in a voter. At the very least enough intelligence to understand that if a policy makes stocks plunge, even if they aren't invested in them, it might not be ideal. So sometimes more intelligence is good sometimes bad. The hard question is what situation we are in and I'm not convinced you give us much reason to suppose we are in the case where it's harmful not helpful.
Ok, but that only works in a country where people own enough stocks. It's going to fail in many developing countries.
Also the government can simply inflate the currency to increase the nominal value of the stocks or literally pass a law saying stock prices can't decrease (outlawing trades below the level when they took office).
Point is you are no longer asking them to judge based on their current welfare/consumption but to predict their future welfare. Sure, in the present US it's a simple rule but only because people are smart enough to realize that currently stock prices for track future expectations but that under other conditions (price floor) they wouldn't.
Am I right in taking from this that these problems only arise from there being asymmetric distributions of information and altruism across (groups of) voters?
If so, should these campaigns not be seen as beneficial provided that they reduce those asymmetries relative to the status quo?
Not sure what status quo you have in mind. Relative to the extreme retrospective voting scenario, this definitely increases variation. People vary a lot in their abilities to form group coalitions and to get better informed about policy.
By status quo I mean whatever the situation is before you do anything. It's not like in the real world there has ever been an election with literally all voters in the extreme scenario. Is it not better (for this problem) to have less variation in average information and altruism (across groups) than to have less total information? (Except for energy/resources expended during campaigns, but I'm assuming those are dwarfed by the budgets elected officials will distribute)
There's too much noise in your ideal system. If most of the variation in the electorate's status over one election cycle is due to factors outside the control of the officeholder (probably true), and if (hypothetically, in your ideal system) the electorate is voting for the incumbent based on whether they saw an improvement during his term, then what we'd see is that roughly 50% of the time the incumbent would be kicked out of office for factors independent of his performance. If there are two parties that means they would change place randomly every term. The party with better performing policies would hold office only *slightly* more often than the other party. That's not a good system. The pressure on a politician to adopt the best policies for their electorate would be very small.
The % of incumbents rejected is a free parameter; voters could set it to whatever they want. And I just don't think it is true that most of the variation in aggregate life satisfaction is out of politician control.
I am thinking, if the voter's life satisfaction rose compared to his prior expectations, he votes for the incumbent, and if it fell, he votes against the incumbent. Assuming the voter's prior expectations were unbiased, that's a 50% chance of voting against the incumbent.
Setting the % of incumbents rejected to a different number - by, say, taking the initial 50% chance of rejection, and flipping a coin, and if the coin lands heads then do not reject the incumbent - doesn't really change the fundamental randomness of the voting. In effect it's roughly the same as increasing the term length.
Variation in aggregate life satisfaction might be within the control of politicians in general, but it's mostly out of the control of any *individual* politician. Suppose that a senator does some deal to bring a factory to their state. Let's be generous and say that's 500 in-state factory jobs. Compared to the baseline variation in job numbers at the state level, that's barely anything. What if a war or pandemic happens nationally? What if some new invention like electricity, automobiles, computers, or AI starts taking off? What if the stock market crashes? Those things are mostly out of the control of the individual senator, and could have a much bigger impact than 500 jobs.
Just from a numbers game, suppose there are a dozen politicians at local, state, and federal levels who are "responsible" for a voter. Even if the politicians all together have total control over the voter's life (not the case), on average any one politician only has 1/12 share of that control. 1/12 is a lot closer to 0 than it is to 1.
Well, in my voting experience, those whose soi disant efforts are to "help voters become better informed" are generally rent-seeking hucksters, well-meaning fools, or politicians interested in gaslighting away their disastrous (or corrupt) choices -- so yeah no, I have not admired such efforts at all.
The best kind of voters, among who I'd live if I could, would be a surly skeptical lot, who found voting a tiresome form of necessary self-defence, like participating in a neighborhood watch, who more or less instinctively voted against any new-fangled scheme unless it were obviously and transparently in their self-interest. The only things government could then get done would be stuff about which hardly any rational person could cavil -- build a bridge across this river between two polities, defeat this Gorn invasion, hang that murderer. Anything else that seemed good to some majority could only get done by a social mechanism that didn't require the application of violence (or threat of same) to coerce the minority. It baffles me that it isn't generally understood that government is violence: if you wouldn't personally put a gun to someone's head to force him to do X, then you shouldn't be asking goverment to do X, because there is no significant moral difference.
Typo: "yet more wasteful rest-seeking efforts". - I agree with Dmitrii: it is sweet and dandy (not 'less laudable') if better informed voters are more relevant to politicians than ignorant ones + why should voters be "altruistic" ever - not even the rich are really accused (by most) for voting in their best interest (buying extra-influence is another story) - And happy Xmas/solstice everyone!
Interesting example of how "logic is a hindsight" can give rise to bad worldbuilding. Possibly, all is not lost. The frame is important.
"Logic is a hindsight" is a type of bias. Developing or suggesting worldbuilding based on gamey theory like this is all very well, and I am not arguing against the truth of the matter here (hindsight is so 2020) but there can be no hindsight without the exact issues you are using hindsight to criticise. If you look only backwards perhaps no one would have children, then there would be no one to make these errors. Survival is a trade-off of bias against bias against hindsight against the just-so story of #ourownlives. Overcoming bias is to be dead.
"Democracy tempts each voter " no that's life not democracy, or least where democracy is an outcome of being alive among others, which gives rise to "democracy" not so much in the first place, but is part of its founding.
Now we need to know about bias, knowing about biases… —that's a gateway drug to enlightenment….
But, what your are describing is more like how accents develop, than how we talk in the first place. And that there is some ideal of elocutions to be found through proper practice and the right traditional techniques, to some perfect neutral schwa we can all say, but then there will be no vowels at all and we would be back grunting, spitting and gesticulating .... pointing back over our shoulders... Generally some capital with an army decides what is the neutral schwa is (despite IPA) and everyone else has a dialect. That morality by other means.
But using what you have said more positively, I would argue that it would be good for politicians (by which I mean anyone engaged in political processes) to use the hindsight gained here in discussion, to look forward (guess/trialanderror/iterate) to actively police against narcisissts and psychopaths, which is what our moral processes are geared to police in the first place. However they seem particularly bad at keeping those OUT who are attract to those exact same positions. And who exploit exactly what you are talking about, the fact we cannot even say their name in this discussions except briefly and in passing with a rather economic "free-rider" term, when it should be a headline is how they have us completely conned. "they"of course being us.
A "simple selfish ignorant retrospective" voter as you describe them would be far BETTER informed than the average voter. The average voter is hypnotized by tribalist rhetoric that has no connection to empirical results at all.
People mostly vote party line regardless of whether they are better off or worse off. Carter had the best jobs creation record in his first four years of any president since, and he was not re-elected. Biden has the best jobs creation record by this point in his term of any president after Carter, and his polls don't look good. https://en.m.wikipedia.org/wiki/Jobs_created_during_U.S._presidential_terms#/media/File%3AJob_Growth_by_U.S._President_-_v1.png
You argue that the result would be skewed, but why would that skewedness necessarily be for the worse? "Maybe" it's better for policies to swing towards some kinds of voters on reasons other than numeric?
I didn't say skewness made it worse.
"Less laudable" certainly seems close enough? Could you please elaborate on the difference?
How would this apply to sortition-based democratic systems? My understanding is that the appeal of them is that they neither encourage nor benefit from voter awareness campaigns, with the idea being that the selected electors are expected to become educated on the candidates only after being chosen. It would seem that this circumvents the issue of voters getting sucked into coalitions but still involves them becoming much more educated on the election.
Yes, random selection avoids these problems, though it induces others.
Could you point me towards any reading that elaborates on this? I'm quite interested in the concept and presently am considering integrating it into my work (governance for decentralized autonomous organizations). But I have not found any robust arguments against it so far, the people at https://democracywithoutelections.org/ seem pretty convincing.
The objection here is specifically to *partial* altruism (i.e. towards a mere subset of other voters), right? So one solution would be to stress that "altruistic" voters are only good if they are *impartial* in their altruism.
Most accept that altruism needn't be uniform across humanity. So not clear how you'd tell voting-coalition motivated altruism from others.
Let us consider the two failure modes of voting:
1) Maximal Short Term Selfishness Voting (Third World, Black Machine Politics Urban Government)
2) Maximum Altruism (Woke Philosopher Kings trying to built perfect society)
Each fails, both because they make coordination impossible. The first rather directly. The second because accurate information on peoples preferences is obscured by altruism and signaling.
The ideal would be someone that tried to show some altruism towards things they know and some degree of long term planning, without descending into ideological fancy.
The main virtue of democracy is that it provides feedback when political leaders fail badly and provides for a peaceful transfer of power. Narrow political power is subject to insider abuse that can get out of hand because the costs can be externalized beyond the narrow political power players with impunity.
Yes, you can construct some models where creating more informed voters are bad. But it's easy to see that purely retrospective voting gets you into lots of trouble.
For instance, the purely retrospective voter will always favor the canidate who raids future benefits to offer benefits now (eg enter into long term deals with costs in the future once out of office).
Clearly to avoid this kind of impact you want at least some intelligence in a voter. At the very least enough intelligence to understand that if a policy makes stocks plunge, even if they aren't invested in them, it might not be ideal. So sometimes more intelligence is good sometimes bad. The hard question is what situation we are in and I'm not convinced you give us much reason to suppose we are in the case where it's harmful not helpful.
If investment is private, voters will own stocks, and see lowered stock prices as hurting them.
Ok, but that only works in a country where people own enough stocks. It's going to fail in many developing countries.
Also the government can simply inflate the currency to increase the nominal value of the stocks or literally pass a law saying stock prices can't decrease (outlawing trades below the level when they took office).
Point is you are no longer asking them to judge based on their current welfare/consumption but to predict their future welfare. Sure, in the present US it's a simple rule but only because people are smart enough to realize that currently stock prices for track future expectations but that under other conditions (price floor) they wouldn't.
Am I right in taking from this that these problems only arise from there being asymmetric distributions of information and altruism across (groups of) voters?
If so, should these campaigns not be seen as beneficial provided that they reduce those asymmetries relative to the status quo?
Not sure what status quo you have in mind. Relative to the extreme retrospective voting scenario, this definitely increases variation. People vary a lot in their abilities to form group coalitions and to get better informed about policy.
By status quo I mean whatever the situation is before you do anything. It's not like in the real world there has ever been an election with literally all voters in the extreme scenario. Is it not better (for this problem) to have less variation in average information and altruism (across groups) than to have less total information? (Except for energy/resources expended during campaigns, but I'm assuming those are dwarfed by the budgets elected officials will distribute)
There's too much noise in your ideal system. If most of the variation in the electorate's status over one election cycle is due to factors outside the control of the officeholder (probably true), and if (hypothetically, in your ideal system) the electorate is voting for the incumbent based on whether they saw an improvement during his term, then what we'd see is that roughly 50% of the time the incumbent would be kicked out of office for factors independent of his performance. If there are two parties that means they would change place randomly every term. The party with better performing policies would hold office only *slightly* more often than the other party. That's not a good system. The pressure on a politician to adopt the best policies for their electorate would be very small.
The % of incumbents rejected is a free parameter; voters could set it to whatever they want. And I just don't think it is true that most of the variation in aggregate life satisfaction is out of politician control.
I am thinking, if the voter's life satisfaction rose compared to his prior expectations, he votes for the incumbent, and if it fell, he votes against the incumbent. Assuming the voter's prior expectations were unbiased, that's a 50% chance of voting against the incumbent.
Setting the % of incumbents rejected to a different number - by, say, taking the initial 50% chance of rejection, and flipping a coin, and if the coin lands heads then do not reject the incumbent - doesn't really change the fundamental randomness of the voting. In effect it's roughly the same as increasing the term length.
Variation in aggregate life satisfaction might be within the control of politicians in general, but it's mostly out of the control of any *individual* politician. Suppose that a senator does some deal to bring a factory to their state. Let's be generous and say that's 500 in-state factory jobs. Compared to the baseline variation in job numbers at the state level, that's barely anything. What if a war or pandemic happens nationally? What if some new invention like electricity, automobiles, computers, or AI starts taking off? What if the stock market crashes? Those things are mostly out of the control of the individual senator, and could have a much bigger impact than 500 jobs.
Just from a numbers game, suppose there are a dozen politicians at local, state, and federal levels who are "responsible" for a voter. Even if the politicians all together have total control over the voter's life (not the case), on average any one politician only has 1/12 share of that control. 1/12 is a lot closer to 0 than it is to 1.
Well, in my voting experience, those whose soi disant efforts are to "help voters become better informed" are generally rent-seeking hucksters, well-meaning fools, or politicians interested in gaslighting away their disastrous (or corrupt) choices -- so yeah no, I have not admired such efforts at all.
The best kind of voters, among who I'd live if I could, would be a surly skeptical lot, who found voting a tiresome form of necessary self-defence, like participating in a neighborhood watch, who more or less instinctively voted against any new-fangled scheme unless it were obviously and transparently in their self-interest. The only things government could then get done would be stuff about which hardly any rational person could cavil -- build a bridge across this river between two polities, defeat this Gorn invasion, hang that murderer. Anything else that seemed good to some majority could only get done by a social mechanism that didn't require the application of violence (or threat of same) to coerce the minority. It baffles me that it isn't generally understood that government is violence: if you wouldn't personally put a gun to someone's head to force him to do X, then you shouldn't be asking goverment to do X, because there is no significant moral difference.
Typo: "yet more wasteful rest-seeking efforts". - I agree with Dmitrii: it is sweet and dandy (not 'less laudable') if better informed voters are more relevant to politicians than ignorant ones + why should voters be "altruistic" ever - not even the rich are really accused (by most) for voting in their best interest (buying extra-influence is another story) - And happy Xmas/solstice everyone!
Typo fixed; thanks.
Interesting example of how "logic is a hindsight" can give rise to bad worldbuilding. Possibly, all is not lost. The frame is important.
"Logic is a hindsight" is a type of bias. Developing or suggesting worldbuilding based on gamey theory like this is all very well, and I am not arguing against the truth of the matter here (hindsight is so 2020) but there can be no hindsight without the exact issues you are using hindsight to criticise. If you look only backwards perhaps no one would have children, then there would be no one to make these errors. Survival is a trade-off of bias against bias against hindsight against the just-so story of #ourownlives. Overcoming bias is to be dead.
"Democracy tempts each voter " no that's life not democracy, or least where democracy is an outcome of being alive among others, which gives rise to "democracy" not so much in the first place, but is part of its founding.
Now we need to know about bias, knowing about biases… —that's a gateway drug to enlightenment….
But, what your are describing is more like how accents develop, than how we talk in the first place. And that there is some ideal of elocutions to be found through proper practice and the right traditional techniques, to some perfect neutral schwa we can all say, but then there will be no vowels at all and we would be back grunting, spitting and gesticulating .... pointing back over our shoulders... Generally some capital with an army decides what is the neutral schwa is (despite IPA) and everyone else has a dialect. That morality by other means.
But using what you have said more positively, I would argue that it would be good for politicians (by which I mean anyone engaged in political processes) to use the hindsight gained here in discussion, to look forward (guess/trialanderror/iterate) to actively police against narcisissts and psychopaths, which is what our moral processes are geared to police in the first place. However they seem particularly bad at keeping those OUT who are attract to those exact same positions. And who exploit exactly what you are talking about, the fact we cannot even say their name in this discussions except briefly and in passing with a rather economic "free-rider" term, when it should be a headline is how they have us completely conned. "they"of course being us.