If it's possible to mathematically characterize the relationship, and the relationship is non-linear, then there exists an optimal hierarchical grouping algorit

If it's possible to mathematically characterize the relationship, and the relationship is non-linear, then there exists an optimal hierarchical grouping algorithm for maximizing empathy in nested groups (that balances the loss due to group numbers vs the "generation loss" from traversing hierarchical levels. I think the difference in sizes between the organizational groups of different countries goes a long way towards explaining the vast differences in political flavor. How many people are in the smallest political constituency you identify with? Dozens? Hundreds? Millions? This is the range of scales we're dealing with. What would it look like if we layered government so that at no level was any group larger than tribal size (~150)? In the US you'd only need 4 levels to cover the entire population, roughly corresponding to: neighborhood (150 people), town (~20,000 people), city/state (3-4 million people), country (327 million people). At that scale, each representative can be on friendly personal terms with a) everyone in the constituency they represent and b) every other representative at their level. So the access to government ought to be much better with this system. In the US, local and federal governments are divided - there's not a lot of continuity between the town government and your friendly local congressperson.