One important step is to identify the relevant factors that need to be considered to determine what kind of governments are stable. Educational level, technology, wealth, cultural diversity and antagonism, etc. A recent paper in Science is an interesting example: “Ethnicity and Conflict: Theory and Facts”, Joan Esteban, Laura Mayoral, and Debraj Ray, Science 18 May 2012: 858-865. [DOI:10.1126/science.1222240]. It took on the question of how ethnic diversity affects stability. They found 1 ethnic group is very stable, 2 less stable, 3 the least-stable, and then increasing stability from there on. But this depended on whether the rewards of ruling were “public” or “private”. “Public” rewards are where the central government gets all the tax money and decides how to spend it. The more money you spend at the federal level vs. the state level, the more likely you are to have ethnic conflict.
Re. democracy, that does not exist. There is no government in the world where the citizens choose legislation, render legal verdicts, or do anything by popular vote other than elect people to do those things for them.
One aspect of government is information theoretic. (A government is very like an AI architecture. The earliest and simplest ones used pure top-down planning and information flow; later we learned how to design more distributed, bottom-up AI architectures and governments.) How much of the information present in the population is available for governing? Policy should maximize information flow towards decisions. This would involve identifying, for any issue, which people provide positive information, ensuring that they’re motivated to provide truthful information, and summing up their information in an optimal way. “One person, one vote” is far from optimal, and possibly worse than random.
One important step is to identify the relevant factors that need to be considered to determine what kind of governments are stable. Educational level, technology, wealth, cultural diversity and antagonism, etc. A recent paper in Science is an interesting example: “Ethnicity and Conflict: Theory and Facts”, Joan Esteban, Laura Mayoral, and Debraj Ray, Science 18 May 2012: 858-865. [DOI:10.1126/science.1222240]. It took on the question of how ethnic diversity affects stability. They found 1 ethnic group is very stable, 2 less stable, 3 the least-stable, and then increasing stability from there on. But this depended on whether the rewards of ruling were “public” or “private”. “Public” rewards are where the central government gets all the tax money and decides how to spend it. The more money you spend at the federal level vs. the state level, the more likely you are to have ethnic conflict.
Re. democracy, that does not exist. There is no government in the world where the citizens choose legislation, render legal verdicts, or do anything by popular vote other than elect people to do those things for them.
One aspect of government is information theoretic. (A government is very like an AI architecture. The earliest and simplest ones used pure top-down planning and information flow; later we learned how to design more distributed, bottom-up AI architectures and governments.) How much of the information present in the population is available for governing? Policy should maximize information flow towards decisions. This would involve identifying, for any issue, which people provide positive information, ensuring that they’re motivated to provide truthful information, and summing up their information in an optimal way. “One person, one vote” is far from optimal, and possibly worse than random.