[kijin]

> I'm not saying technocracy is a wonderful system of government either, but it won't be hard to do a better job than democracy.

My objection to that statement is that designing political systems is not about doing a better job (on average). It's about not doing a worse job (ever). Rawls calls this "maximin": you choose the option that maximizes the minimum possible value. The maximum possible value is irrelevant, and the average doesn't matter, either. When lives are at stake, you always choose the safest option, the one with the lowest chance of descending into tyranny.

I'm sure technocracy has the potential to soar higher than populist democracy under many circumstances, but unfortunately it also has the potential to crash just as hard as, if not harder than, most democracies can ever dream of. Without giving others the power to overrule technocrats, it's only a matter of time before technocrats become autocrats. Anyone in a position of power who can't be fired is a Stalin in the making. I don't see how that's "not worse" than any democracy. A populist democracy has a reliable way to throw a political leader out of his job or at least get him to change his mind: throw up a shit storm like we did with SOPA. It's dirty, it ain't easy, but once you manage to do it, it works like a charm. And in the end, keeping little Stalins at bay is all that matters.

Power tends to corrupt. Any political system that depends on trusting a group of people to do things right will invariably fail because of corruption, and technocracy is no exception. The only solution is to design a political system that harnesses the power of distrust. When all alternatives are necessary evils, you choose the most controllable option because a controllable evil is almost always less bad than an uncontrollable evil. Of course, evils aren't easy to control even under the best circumstances, but do your best to find one that you can at least put on a leash.

So the best government isn't one that is not subject to control by any political interest, because objectivity and neutrality in politics are never going to be anything more than daydream. Rather, I think the best government is one that is open to control by so many conflicting interests in so many different ways that the vectors eventually balance one another out. Politics is meant to be dull, boring, and utterly predictable. Because when politics gets exciting, people die! It's OK if you want to give up your own life for a cause, but don't put others' lives and human rights in peril.

tl;dr for technical readers: the performance benefits of putting a bunch of spinning platters in RAID 0 ain't worth the inevitable data loss. Especially if your data == human lives.

[4891]

Yours is the standard post-cynical view of democracy. It's common among very smart people. Think Churchill's "two cheers for democracy". It's what I used to believe. I don't now.

"Power tends to corrupt." If this is true, then democracy may well be the maximin solution. But is it true? The word "autocrat" makes people think of Hitler and Stalin. They don't think of Frederick the Great or Deng Xiaoping.

Power often corrupts, but not always.

1. When power is insecure, it is dangerous, like a cornered wild animal: totalitarian spying against potential conspirators, suppression of dissent, brutal retaliation against political opponents. Secure power has less need for such behaviour. As Bismark said of the press "they can say what they want, and I will do what I want".

2. When power is a "family business", the leader's time preferences change and they want to ensure they pass a prosperous country onto their children. That's why old-school monarchs, for all their faults, were less rapacious than third-world dictators - the dictators knew their time in power was limited, so they made out like bandits while they could.

3. Many people are actually quite nice and will use their power to help their fellow men (or at the very list, fellow countrymen).

Consider that Communist China has transitioned from Maoist schizo-state to pragmatic autocratism and now to some kind of quasi-democracy, all without major revolutions. I see the real danger of technocracy being Brezhnevism, not Stalinism, but the same thing has happened with modern day democracy. (The current "democratic" system is designed to prevent charismatic populist politicians (ie, potential Hitlers and Stalins) from making much significant difference - how much personal impact over government policy did Bush have, for example? Or, for that matter, Obama? - and instead most power rests with hundreds of unelected but government-funded agencies.)

"So the best government isn't one that is not subject to control by any political interest, because objectivity and neutrality in politics are never going to be anything more than daydream. Rather, I think the best government is one that is open to control by so many conflicting interests in so many different ways that the vectors eventually balance one another out. Politics is meant to be dull, boring, and utterly predictable. Because when politics gets exciting, people die! It's OK when it's your own life, but don't put others' lives and human rights in peril."

I definitely agree with your last 3 sentences. The thing is any kind of true democracy is going to be full of exciting politics. Your ideal of a system where the vectors balance out is savvy (were you a British colonial administrator in a previous life?) and is also quite close to the current system. Designing a government that is ineffective by design is a good way to prevent tyranny. Unfortunately, it also makes it unable to make difficult-but-necessary decisions. Example: who in the USG actually has the power to reduce the deficit? Will this situation change in the next 20 years?

[kijin]

While I agree that some people who enjoyed absolute power were actually rather decent, I doubt that the modern world is conducive to producing more of them. The degree to which power corrupts a person depends a lot on external variables, and the variables that have been at play in the last 60 years or so have produced a lot more malevolent dictators than benevolent ones. Guido van Rossum is the only person that I can think of who actually lives up to the BDFL title in today's world, and that's probably because his realm does not physically exist ;)

I also wonder if the possibility of getting another King Sejong is important enough to risk getting another Vlad the Impaler, even if the chances were as good as 50/50. Even if power only corrupts 50% of the time, or even 10% of the time, maximin still says that you should choose the safer option.

I'm not saying that the current form of American democracy is the safest option. Every day it seems to move further away from the maximin solution. But if so, that's only a reason to find and implement the maximin solution, not a reason to move even further away from it. If all the power actually rests in the hands of 10000 unaccountable bureaucrats, the solution is not to give more powers to Obama and thereby end up with 10001 unaccountable people. The solution is to make those 10000 bureaucrats fully accountable, in addition to Obama, so that we have 10001 fully accountable people.

> The thing is any kind of true democracy is going to be full of exciting politics.

As long as my life, limbs, and human rights are not at the mercy of some megalomaniac (or any particular interest group), I'm sure I'll get used to whatever excitement "true democracy" brings. I'm assuming, of course, that this "true democracy" includes strong constitutional protections of my rights so that nothing that happens in ordinary politics can violate them. Democracy doesn't mean everything gets decided democratically, there are certain things that even a majority shouldn't be able to do.

> Your ideal of a system where the vectors balance out is ... quite close to the current system.

There are too few vectors in the current system. We need way more vectors to cancel out supersized incumbents like the military-industrial complex, the oil industry, Monsanto, the MAFIAA, etc. Putting together a short list of interested parties and deciding all matters among them is just corporatism. We need something more open and flexible, where EFF (for example) has a realistic chance of outcompeting Sony in the marketplace of ideas.

> it also makes it unable to make difficult-but-necessary decisions.

Big decisions usually shouldn't be made in a hurry. So as long as it's not 100% impossible for the government to make decisions, I think it's OK for there to be lots of checks and balances in the way. (This would be a different situation from what we currently have, where unaccountable bureaucracies are responsible for a lot of inertia.)

I also don't think it's right to move in the direction of facilitating the production of massive decisions on a national scale. In my picture of "true democracy", decision-making responsibilities would be decentralized unless the federal government absolutely needs to get involved, and most local decisions would follow an agile and rapid-iteration model. Take health care reform for example. Did we really need such a protracted, ideologically charged, misinformation-filled, national debate about it? Vermont actually wanted to implement their own single-payer system, and a few other states had similar but slightly different ideas. Why couldn't we let those states implement their own health care systems first, compare results after a couple of years, make changes along the way, and get other states on board over the next decade or two with plenty of opportunities for A/B testing between them? Answer: Ego. Obama and Hillary wanted to finish the job themselves, get it exactly right the first time, and take credit for it.

A very large proportion of the stalemates that make Washington ineffective could be eliminated if people refrained from saying "But there's this teeny tiny provision here that I don't completely agree with..." and just agreed to reiterate every year. Can't agree on the debt ceiling right now? Release the current version now and release the .1 version by the end of the year. You ask how the government can make massive decisions. I ask why such massive decisions are needed in the first place. Of course this won't work all the time (wars, for example, can't be easily canceled once begun), but the thing about necessary evils is that you don't want to make them appear any more necessary than they strictly are.