I've just realised that the same applies to government but, um, in reverse.
What you want from a government is legislators and administrators who are competent and not evil. Democracy is quite bad at locating such people, since it reduces everything, ultimately, to a popularity contest; and there is nothing in particular that assures a strong correlation between popularity and either of the desired qualities. Far better to select leaders on meritocratic principles, since this should at least select for competence. It's my tentative conclusion that China has lately been working on this approach, and has been showing some pretty favourable results, overall. Of course, one can easily criticise China for its lack of transparency and its high rate of corruption.
At the same time, it seems that democracy has been outperforming most of the alternatives, on average, even though it has a marked tendency to elect, from time to time, malicious idiots, or worse.
I have to ask myself, then, if the power of democracy comes not from its dubious method of choosing leaders, but the facility with which it eliminates them when they prove to be unacceptable.
So perhaps, as with education, we have it backwards. Maybe the form of democracy we want is that candidates for office should be drawn from a well educated, well prepared, pre-groomed elite; and the function of elections should be to determine, every few years, which officials should continue in their positions, and which should be removed - or indeed, put on trial.
I think I like that idea quite well.
There doesn't have to be a single source pool for officials; universities could produce them independently, so long as they could be ranked according to some more or less objective criteria (one could imagine actually checking wither the candidates knew things about economics, sociology and engineering, for example) originating from a politically independent body. Candidates could be chosen by examinations written by an independent body. There would probably still be political parties, but the officials wouldn't need to belong to them, a priori; rather they would organise campaigns for or against the officials being evaluated according to their pattern of past behaviour on the issues the parties care about.
Just, you know, a thought.