A recent development is that societies should be run by an expert government. In ancient times, Plato dreamed of such a government in The Republic, and the Chinese thought they had implemented such Dynasties through civil service tests. In the last hundred years this idea has been adopted by Socialists and referred to as "scientific socialism" and called "Progressive".
This idea is based on the argument that everything is best controlled by rational study and logical implementation by central experts. The rise of the college educated, elite class has supported and made possible the implementation of this expert, central state mechanism. The term "scientific" has been used to cloak the enterprise in the success of science in the material world.
The problem is that intellectual activities based on analysis and logic are flawed in many ways that make their results unreliable for central control. The "hard" sciences compensate for these flaws by reductionism, severe attention to detail, and insistence on testable correspondence with reality. Other intellectual use of analysis and logic have no such checks.
The first problem is that all such effort is based on abstractions, generalities, and statistical categories. These are dangerous from the start because their correspondence to real things is pretty much unknown and untested. The most rigorous and profound logic on these is only speculation unless by accident the abstractions correspond to reality in a specific case. They cannot be relied upon in general.
The second problem is that societies are far more complex than any system ever successfully analyzed by a hard science. The social data measured are at best superficial and must leave out many relevant factors. Usually the differences within an analyzed category are as great or greater than between the categories. As a consequence, any programs or policies implemented based on such analyses are pretty much doomed to failure and unintended consequences.
The third problem is similar to Chaos Theory in the hard sciences. There are many problems where perhaps the abstractions and inferences fit adequately, but tiny differences in the data or assumptions lead to wildly different results. This of course makes social solutions based such results unstable and dangerous.
The fourth problem of course is the human element. Power seeking politicians and the self-interest of the intellectual class promote expert centralization. Once in place, it is no surprise that all "rational" solutions work to their advantage, and few contrary solutions see the light of day.
The solution to such problems throughout history has been the rise and fall of societies. The safety factor was that everyone was not under the control of one authority, and failure was localized. The rise of central authorities and their unification through worldwide organizations such as the UN is worrisome because it eliminates the safety factor.
The system of self-government that evolved in the United States was bottom up, so that failure was localized at the individual, local, or regional level. This made the US the most robust and successful country in history. Groups, businesses, and local governments in the US were run by practical people of experience, not by intellectuals situated at a central authority. Such locals can respond with on-the-ground knowledge of timely detail that is impossible with central control. If they misjudge or fail, the damage is localized and others can continue.
As a consequence, we must reverse the centralization trend in the US and worldwide by returning to self-control and self-government at the local level and strictly limited power at any higher level.