As the organization increases, you are more likely to find people who can band together to collude to corrupt some small corner of it. Number of corrupt candidates increases, number required does not.
Amazon only needs to lobby one federal legislature’s members and only need to battle one federal regulatory body. If all of the power were at the state levels, it would have to scale that up 50x.
Example: greed.
There is such a thing as unorganized corruption. Example, in lots of third world countries where their infrastructure and systems are so bad it breeds corruption. No need for a top down version of it.
Yes, I agree that you can see examples in some parts of the world and in some time periods (including today) where even uncoordinate corruption is very damaging to a society. But - and this is important given the context in which I started talking about this - it does not appear to be a pathway to regulatory capture nor the formation of giant corporations in the sense that was proposed up-thread, in which a powerful "federal" organization is used to enable such organizations to exist.
Greed won't direct an organization in a unified direction, but that just exposes a vulnerability where an outside party can exploit the greed at an individual level to direct the organization in a unified direction of its choosing.
But yes in the context in which you have laid down you are correct I believe.
No such thing is required beyond the common stance of 'venal self-interest' regardless of organizational mandate. Without checks and balances (and penalties) you can get a loosely coupled "corruption" at all levels with little effort.
This sort of claim seems to me fundamentally rooted in some specific ideas about large human organizations, specifically of the governmental kind. One of the key ideas it requires to be true is that the corruption endemic to small-scale institutions (think your local mayor, perhaps (depending on your state) even your governor) will automatically apply to much larger scale institutions (such as those administered by the federal government).
I'm not going to deny that this could be true - it clearly could. But the change in scale, the change in mechanisms of influence that are required, the change in the possible side-effects of such corruption being uncovered ... these all point in the opposite direction, in my opinion.
Amazon and Walmart are not the victors in a game of regulatory capture. They are the beneficiaries of the election of senators, representatives, presidents and the appointments of judges who explicitly do not believe in much regulation. Even so, they are subject to more oversight than one of my local ranching families in NM, who can get a (former) governor to re-route millions of dollars in state & federal funding to a road they would prefer.