An ideal government

Democracy will not work. It does not work and will never work. Specifically any voting system is subject to Arrow's Impossibility, and democracy cannot violate any precept thereof without destroying itself.

To be a dictatorship is to not be a democracy, so I can safely ignore the idea of a democracy that is a dictatorship. That said, states such a Stalin's Russia or Saddam's Iraq tended to be highly corrupt and inefficient, and if you constantly keep people sated with fake security they will not secure real security for themselves. Of course if the individuals under the government do not protect the government, the government, being the group of individuals under itself, will also be unable to protect itself.

Next off, a democracy without Unrestricted Domain. This again is split into 3 distinct clauses I can violate. IE-
A democracy that does not take into account all individual preferences. In other words, a democracy with a limited vote, similar to America's original vision. The main flaw of this system is that people with power will use their power. IE, it is implausible to expect that you will be able to indefinitely keep the vote from being used by all groups. People will naturally vote, that additional voters who agree with them should also be able to vote. By compromising and giving various special interests support and various causes in return for extra voters, they will be able to expand to fill all limits. Those left out of this expansion will be completely unrepresented by their government and will thus be disloyal, hateful and in pain. As the unrepresented populace grows it will use progressively larger quantities of force in order to seize power. A strange alien government could probably repeatedly genocide such groups, but, as humans, we are naturally empathetic and kind, and cannot bring ourselves to build such a government. We could design a government that secretly carries out its genocides, (like Russia) but once again, a government without oversight will immediately fall to corruption.

A democracy without a complete ranking of all issues, IE a democracy wherein the government does not compare all options. The problem here is this- that on all issues a choice must be made. IE somehow, someone or some group must make the choice. In essence, what is proposed is merely a tiered democracy in which votes are located on different planes. X can be compared to Y, but not Z because Z is located in another castle. We can see some of this in the idea of division of powers/check and balances. The supreme court handles some issues while the president handles others and the congress handles others. Unfortunately this achieves very little if each branch is generated from the same group, and faces the same problems above if they are generated by separate groups (think house of commoners and lords from the past.) Of course you can generate certain policies via local populations and others via global population- thus ignoring the preferences of people unrelated to any given issue. Unfortunately people will tend to push an issue upwards until it receives support- IE if the city doesn't want X then you can still ask the state to mandate that every city build X. Fundamentally, if votes determine the government, then everyone will have an effect on every issue, regardless of cup games and vote shuffling.

The third axiom I can violate to avoid universality is to have a system wherein a person's votes do not necessitate any given result, IE the senate of the Roman Empire. Unfortunately there's a reason we call the Roman Empire an Empire- it was not a democracy.


Secondly a system that does not maintain independence of irrelevant alternatives. At first this seems very different from the original axiom, but fundamentally it's not. Instead of allowing a few people to vote on all issues you're allowing everyone to vote on a few issues. IE, your restricting the vote to the few people who care about the issue. Here we see many of the antics of modern American democracy- For instance, imagine a candidate as a conglomeration of issues. One candidate supports abortion while the other opposes it. Thus any person CAN vote for or against abortion. But, imagine that the same candidate who supports abortion also supports higher taxes and more spending. Note that a voter restricted to the two candidates will soon realize that due to caring more about economics, he IS NOT ALLOWED to vote on abortion. This will result in subsets that care heavily about an issue always winning any debate on the issue. (IE a teacher cares a lot more about schools than anyone else and will only support candidates that propose more spending on teachers. The teachers then push through smaller class sizes, higher wages, lower standards that are more specific to the teachers that exist to keep out competition by stopping the removal of the old and restricting access to the new.) Such a system can only result in larger government each voting cycle. As such, if you lack independence of irrelevant alternates then you are part of the communist system.

Finally we come to Pareto efficiency. IE a society in which preference of A over B does not necessarily lead to A being chosen over B. This is interesting, and assumes some sort of automata that can intelligently come to a conclusion given data. What we have here is not decisions being made by voters- it's decisions being made by the system, a amorphous object granted sentience by the votes fueling it. Note that such an automation need not be technically advanced- a machine in which black and white stones are placed into a box, and then a number are removed at random, would display a certain level of intelligence. While ignoring Pareto efficiency we can easily drift away from democracy to systems such as- Everyone votes, then a super intelligent robot decides on a fitting society. It's not a dictatorship, because the robot has no opinion, it merely takes votes and gives results. Everyone votes on everything (the robot could even simulate votes using its super-intelligence by predicting voter behavior without actually having to ask the voter personally) so all individual preferences are used. Since the robot is just a highly sophisticated program it will give the same results from the same vote every time. Assuming the robot chose each issue independently it could avoid being affected by irrelevant issues. This however is nowhere near a democracy, and any similar system inferior in amplitude is inferior to the max amplitude super robot. So, that leads to -Democracy as super robot, IE a system that is designed to be the closest thing we can build with modern technology to our super robot.
What we must understand about such a system is that in it we are already entrusting ourselves to a foreign intelligence- IE, if we are willing that an automata make our decisions for us then that automata has no advantage over any other decision making body besides that of the quality of the decisions made. Imagine then the alternate super robot- A super intelligent machine that contains its own motive power, and designs society to suit its predefined precepts. Any nobility, should be, in essence, our highest tech attempt at designing said super robot. As such, for our democratic super robot to succeed it must be superior to our noble super robot.

So we come to the question of- Just what is our democratic super robot? Answer- The bill of rights and representation. We compile a list of preferences, then throw out ones we don't like. Then we categorically link answers we do like to answers we like on issues wherein we received answers we didn't like. The problem here, is that it gives too much power to the representatives. With this power the representatives can influence the very votes they are supposed to be processing. Soon enough, we receive a system that is non-democratic, wherein the government feeds the populace lies, and the populace then turn around and say whatever is asked of them. The decisions are then made from the top down- Even though they are not enacted until the decision moves from the bottom up. IE just because a decision is approved by the bottom does not make a system democratic. The people in the Soviet Union were very supportive of the Soviet Union- Even as they did everything within their power to escape the lives produced by the Soviet Union. We cannot afford to trust our politicians because once trusted, they will abuse said trust. (Even if the politicians originally given said trust were trustworthy.)

So here I present my noble super robot- Wherein a group of people are selected via intelligence and personality testing. By scientifically designing the test and putting it through millions of test cases and simulations ahead of time, we can give it a high degree of accuracy. It will select people who are honest, intelligent, ruthless and kind. They will not hesitate to protect us in war or against criminals, but will not seek out enemies in peace. They will defend the environment without destroying the economy. They will enrich the poor but will not harm the rich. They will do everything every lying politician has declared he'd do during every election- Because he would be mentally suited to the task. However, the most important thing that the test could measure is this- That the selected nobility would not destroy the system. The chosen participants would not take their children who are mentally distinct from them and put them into power even though the test said they were ineligible. They would not extend federal powers into local disputes, nor would local nobility seek to elevate themselves to federal positions. Such people do exist, but we will never find them through democracy- a democracy will only elevate people who claim these properties- Only science can sort people based on who actually has such properties.