I might at some point write up a more in depth discussion, including the (mostly helpful) feedback I received. Having thought some more about it, an especially interesting aspect is that part of the government is responsible for designing the examinations, which means we have a recursively self-improving system which needs to maintain its original purpose. Sounds familiar?
All bureaucracies eventually wind up as psychopathic UFAIs; that’s not news to me. But it’s an interesting idea at any rate. A similar idea (but one that’s not as radical) is to have voters take aptitude tests. That is, in order to be a full citizen capable of voting, you must take an examination that measures the very basics of knowledge about politics, business/industry, history, and psychology. The great thing about this is that the idea isn’t to measure a very high level of intelligence, so the questions would be easier to design and could be made more unbiased. I imagine that even with very simple questions (“How many senators are there?”, “What is the availability heuristic?”), at least 90% of people would fail the test. You must take the test every year and you are obviously allowed to train for the test. The goal of the test isn’t to make sure the voter base is highly intelligent and rational. The goal of the test is to eliminate the people who are definitely lacking in intelligence.
I think this simple idea would improve the state of things a lot.
How about a counterproposal: If you want to subject some other citizen to laws made by people who are voted for, you need to grant them the right to vote (unless they are considered incompetent on a general level).
Failing an aptitude test will not get you out of having to pay taxes or going to jail, nor will it get your gay marriage recognized.
All bureaucracies eventually wind up as psychopathic UFAIs; that’s not news to me.
If you are saying that my proposal will inevitably eventually go wrong, I think that’s a little pessimistic. Caution is certainly required, but there are extra precautions that can be taken, such as the part of the government that decides the tests being non-self-modifying, or a constitution, or a supermajority of citizens being able to impeach the government. I’m not saying that these do not have their own downsides, but I don’t think its time to give up, and I also don’t think its more dangerous than a democratic bureaucracy.
A similar idea (but one that’s not as radical) is to have voters take aptitude tests.
I would describe this as “representative technocracy” as opposed to “direct technocracy”. I would however raise some concerns about taking the test every year—revision is a lot of work, and you select for people who really care strongly about politics, rather than just intelligence (I, for instance, would not be able to answer “How many senators are there?”) at first glance this might be a good thing, but it could also give a disproportionately large voice to people with extreme views, as perhaps centralists would not care as much.
Part of my proposal was that the first stage of exams would be administered as part of the exams people take when leaving school, like the SATs, and that universities and employers would also care about the results. Later stages would be given to people who had a real chance of pursuing a career in politics, so obviously they have a motive to work hard.
Unfortunately, tests for voting have been tried before, and were used to stop blacks voting in the US. Now, these tests were designed to be unfair, most of them were oral (and therefore not anonymous) and initially whites did not have to take the test. None of these things need apply to a future voting test, but nevertheless requiring voters to take tests seems to have gained a very bad press because of this.
(I, for instance, would not be able to answer “How many senators are there?”)
I tend to think that knowing the answer to a question like this is important when voting for your senator (and there are 100 senators − 2 for each state).
and you select for people who really care strongly about politics
If caring is paired with knowledge, this isn’t necessarily bad.
tests for voting have been tried before, and were used to stop blacks voting in the US.
Yes, this the primary criticism I have received from people about this idea, and I agree with you that none of these things need apply to a future voting test.
If caring is paired with knowledge, this isn’t necessarily bad.
I’m a little worried that the people who shout the loudest are not the wisest, but the most ideolgically driven, therefore the most blinded by ideology. Maybe it’s the extremists who really care, the centralists are often more ambivalent.
I tend to think that knowing the answer to a question like this is important when voting for your senator (and there are 100 senators − 2 for each state).
I should point out that I’m not actually american, and the equivalent situation is slightly more complex in the UK, because the number of MPs keeps changing.
That’s an interesting idea. Care to elaborate?
There was a substantial discussion of it I started here:
http://lesswrong.com/lw/l5r/nonstandard_politics/bi1g
I might at some point write up a more in depth discussion, including the (mostly helpful) feedback I received. Having thought some more about it, an especially interesting aspect is that part of the government is responsible for designing the examinations, which means we have a recursively self-improving system which needs to maintain its original purpose. Sounds familiar?
All bureaucracies eventually wind up as psychopathic UFAIs; that’s not news to me. But it’s an interesting idea at any rate. A similar idea (but one that’s not as radical) is to have voters take aptitude tests. That is, in order to be a full citizen capable of voting, you must take an examination that measures the very basics of knowledge about politics, business/industry, history, and psychology. The great thing about this is that the idea isn’t to measure a very high level of intelligence, so the questions would be easier to design and could be made more unbiased. I imagine that even with very simple questions (“How many senators are there?”, “What is the availability heuristic?”), at least 90% of people would fail the test. You must take the test every year and you are obviously allowed to train for the test. The goal of the test isn’t to make sure the voter base is highly intelligent and rational. The goal of the test is to eliminate the people who are definitely lacking in intelligence.
I think this simple idea would improve the state of things a lot.
How about a counterproposal: If you want to subject some other citizen to laws made by people who are voted for, you need to grant them the right to vote (unless they are considered incompetent on a general level).
Failing an aptitude test will not get you out of having to pay taxes or going to jail, nor will it get your gay marriage recognized.
If you are saying that my proposal will inevitably eventually go wrong, I think that’s a little pessimistic. Caution is certainly required, but there are extra precautions that can be taken, such as the part of the government that decides the tests being non-self-modifying, or a constitution, or a supermajority of citizens being able to impeach the government. I’m not saying that these do not have their own downsides, but I don’t think its time to give up, and I also don’t think its more dangerous than a democratic bureaucracy.
I would describe this as “representative technocracy” as opposed to “direct technocracy”. I would however raise some concerns about taking the test every year—revision is a lot of work, and you select for people who really care strongly about politics, rather than just intelligence (I, for instance, would not be able to answer “How many senators are there?”) at first glance this might be a good thing, but it could also give a disproportionately large voice to people with extreme views, as perhaps centralists would not care as much.
Part of my proposal was that the first stage of exams would be administered as part of the exams people take when leaving school, like the SATs, and that universities and employers would also care about the results. Later stages would be given to people who had a real chance of pursuing a career in politics, so obviously they have a motive to work hard.
Unfortunately, tests for voting have been tried before, and were used to stop blacks voting in the US. Now, these tests were designed to be unfair, most of them were oral (and therefore not anonymous) and initially whites did not have to take the test. None of these things need apply to a future voting test, but nevertheless requiring voters to take tests seems to have gained a very bad press because of this.
But yes, broadly I approve of your proposal.
I tend to think that knowing the answer to a question like this is important when voting for your senator (and there are 100 senators − 2 for each state).
If caring is paired with knowledge, this isn’t necessarily bad.
Yes, this the primary criticism I have received from people about this idea, and I agree with you that none of these things need apply to a future voting test.
I’m a little worried that the people who shout the loudest are not the wisest, but the most ideolgically driven, therefore the most blinded by ideology. Maybe it’s the extremists who really care, the centralists are often more ambivalent.
I should point out that I’m not actually american, and the equivalent situation is slightly more complex in the UK, because the number of MPs keeps changing.