When I say regulations do not work, I do not mean that they do not succeed in controlling or regulating the economy. What I mean is that regulations do not work in the way that naive supporters of state regulations believe that they do. I mean that regulations do not overcome market failures (as that term is commonly used) to produce more efficient outcomes. Regulations do not succeed in protecting either workers or customers except incidentally and not better than could be done under the free market.
The argument is rather simple. The people most motivated to become regulator are those who personally benefit from regulating. There is also an idea we learn from game theory, reciprocity. People who are in the industry would be more qualified to regulate it on the basis of their knowledge. But since they come from that industry, they are going to be more willing to reciprocate to those who helped them get where they are. The flip side is also that regulators will need to find a job after they leave the government sector. Building a network by appeasing firms in an industry is the best way for a regulator to do that.
Even if a regulator has the best of intentions, she can only act on the information she has. But she only has information about the past and only incomplete information since much of it is personal. She can never have full information about the present.
In order to regulate an industry, a regulator is going to need to have regular contact with members of that industry. Stockholm syndrome and its converse Lima syndrome illustrate that people in an antagonistic relationship can over time develop sympathy for one another. Since regulators are often creating win-lose situations, since they have sympathy for the industry they regulate and the firms they work with, will create situations that benefit their friends in the industry.
This is why absurd regulations exist and why they will always exist. And regulations will never work.