3 ms·

> The aides work directly for the senator and thus the state, not for a private organization. Isn't it common for regulatory capture to happen with a "revolvin

> The aides work directly for the senator and thus the state, not for a private organization. Isn't it common for regulatory capture to happen with a "revolving door" where people move between industry and government, setting up future benefits for themselves on every cycle? e.g. Say Microsoft is really good, make a law that de facto requires their software, later happen to get hired there. At least it worked once. I just don't think there is any special structure that can prevent corruption without human effort.

BobaFloutist 3y ago
It's more common for people to move from government to industry, and then utilize their insider knowledge and contacts to benefit their company.
godelski 3y ago
A revolving door is circular. Like a aide becoming a lobbyist and then becoming an aide again. But typically it is more like what we see with supreme court clerks, flow in one direction. Using their experience, insider knowledge, and the networks they have created to become more influential. But I'm not sure how we can avoid this in any structure without significant consequences. Like if you say a government employee must spend X time before entering the private industry, what are they supposed to do during that time? I guess we could give them a pension? But this is why I am connecting the concept of democracy to that of federated systems. Think of the fractal nature of this. We can draw groups around things and talk about consolidation. We need to think about graphs and networks. It isn't all there is to the problem, but it is part of it. Let's pretend we have two groups that have equal knowledge and equal amounts of power. So in a perfect world, if these two were benevolent entities they would be equally effective. Group 1 is a singular politician. Group 2 is a politician with their aides. For all intents and purposes these two groups are equal other than the number of people. Is one group easier to corrupt than the other? Most certainly! Group 1 only requires that you corrupt a single individual. Group 2 the power is distributed (we don't need homogeneity here, but if it helps, that's the clearer case), and so you have to expend more effort to corrupt that system. This scales and takes on a fractal like nature! It then becomes even more beneficial when you consider that there are many adversaries. There is a weird effect where you can actually have all keys to power (politicians, aides, whatever) be corrupted BUT where no single adversary is able to take control over the state! Because they themselves are in competition! If you have only a few nodes that need to be corrupted, it is much easier to perform a total graph capture because the power you have is also proportional to the number of nodes you control. But if you divide the power across many nodes, and you cannot capture nodes instantaneously and/or at no cost, then (assuming there are competitors) the steady state solution is having at least 2 entities capturing the graph, not 1. Essentially the steady state corruption solution is an oligopoly rather than an autocracy. We might call what we have a Plutocracy, but even though I dislike Plutocracies, I think they are better than autocracies. I don't think there is a special form of government that can prevent corruption. Certainly not one that doesn't require action to maintain its state (which is difficult to rely on). So instead I think the main focus should be on failure design. Failure design is one of the most important concepts in engineering because it is taking control in how a system fails rather than letting it happen randomly. In something like a government you have to balance the distributed nature as well as the centralization. Centralization (with autocracy being the extreme) has the benefit that one can respond fast because they have absolute control. Direct democracies on the other hand have the opposite problem where the power is spread too thin and thus it is difficult to do anything because you have to do a full graph transversal anytime you want to make any decision. You need the balance for efficiency, not to mention that people aren't absolutely equal in skill and knowledge. But placing high importance on failure design in embracing the chaos and unknown of the system. Knowing that you can't control all parts and that once you've set the system in motion, you no longer have control. It evolves on its own.

BobaFloutist 3y ago

It's more common for people to move from government to industry, and then utilize their insider knowledge and contacts to benefit their company.

godelski 3y ago

A revolving door is circular. Like a aide becoming a lobbyist and then becoming an aide again. But typically it is more like what we see with supreme court clerks, flow in one direction. Using their experience, insider knowledge, and the networks they have created to become more influential. But I'm not sure how we can avoid this in any structure without significant consequences. Like if you say a government employee must spend X time before entering the private industry, what are they supposed to do during that time? I guess we could give them a pension? But this is why I am connecting the concept of democracy to that of federated systems. Think of the fractal nature of this. We can draw groups around things and talk about consolidation. We need to think about graphs and networks. It isn't all there is to the problem, but it is part of it. Let's pretend we have two groups that have equal knowledge and equal amounts of power. So in a perfect world, if these two were benevolent entities they would be equally effective. Group 1 is a singular politician. Group 2 is a politician with their aides. For all intents and purposes these two groups are equal other than the number of people. Is one group easier to corrupt than the other? Most certainly! Group 1 only requires that you corrupt a single individual. Group 2 the power is distributed (we don't need homogeneity here, but if it helps, that's the clearer case), and so you have to expend more effort to corrupt that system. This scales and takes on a fractal like nature! It then becomes even more beneficial when you consider that there are many adversaries. There is a weird effect where you can actually have all keys to power (politicians, aides, whatever) be corrupted BUT where no single adversary is able to take control over the state! Because they themselves are in competition! If you have only a few nodes that need to be corrupted, it is much easier to perform a total graph capture because the power you have is also proportional to the number of nodes you control. But if you divide the power across many nodes, and you cannot capture nodes instantaneously and/or at no cost, then (assuming there are competitors) the steady state solution is having at least 2 entities capturing the graph, not 1. Essentially the steady state corruption solution is an oligopoly rather than an autocracy. We might call what we have a Plutocracy, but even though I dislike Plutocracies, I think they are better than autocracies. I don't think there is a special form of government that can prevent corruption. Certainly not one that doesn't require action to maintain its state (which is difficult to rely on). So instead I think the main focus should be on failure design. Failure design is one of the most important concepts in engineering because it is taking control in how a system fails rather than letting it happen randomly. In something like a government you have to balance the distributed nature as well as the centralization. Centralization (with autocracy being the extreme) has the benefit that one can respond fast because they have absolute control. Direct democracies on the other hand have the opposite problem where the power is spread too thin and thus it is difficult to do anything because you have to do a full graph transversal anytime you want to make any decision. You need the balance for efficiency, not to mention that people aren't absolutely equal in skill and knowledge. But placing high importance on failure design in embracing the chaos and unknown of the system. Knowing that you can't control all parts and that once you've set the system in motion, you no longer have control. It evolves on its own.