> but you can always model the system as zero-sum if you normalize the total amount of wealth in the system to 1 after every transaction
Yes, if you "normalize" the world to fit your conclusion, you'll generally find that your conclusion is true. If we just normalize everybody's wealth to 1, then all wealth inequality disappears. Problem solved.
The fact that you need to normalize the total amount of wealth suggests that you recognize that the total amount of wealth changes, which means the system isn't zero sum.
> The power law is created by the preferential attachment of capital to existing capital. Even a weak preference creates an extreme inequality.
> If we just normalize everybody's wealth to 1, then all wealth inequality disappears. Problem solved.
You misunderstood the use of the word "normalize". The point of normalization is to change the absolute magnitude but keep the relative magnitudes the same. Perhaps I should be more specific: if you have two values, 6 and 4, which sum to 10, and you'd like to normalize the total to 1, then you multiply each by 1/10 so that you end up with 0.6 and 0.4.
> The evidence doesn't really bear this one out.
Write some code to simulate this. I think you'll find that preferential attachment does in fact create a rich-get-richer phenomenon. For example, take a look at this simulation coded in NetLogo (http://ccl.northwestern.edu/netlogo/models/PreferentialAttac...).
Yes, if you "normalize" the world to fit your conclusion, you'll generally find that your conclusion is true. If we just normalize everybody's wealth to 1, then all wealth inequality disappears. Problem solved.
The fact that you need to normalize the total amount of wealth suggests that you recognize that the total amount of wealth changes, which means the system isn't zero sum.
> The power law is created by the preferential attachment of capital to existing capital. Even a weak preference creates an extreme inequality.
The evidence doesn't really bear this one out. https://www.gsb.stanford.edu/faculty-research/publications/f...