For me, the readings make a lot of very similar points, but all of them have their own nuances which make many arguments actually very different. One thing many of them agree on, specifically Fukuyama and Carl & Victor, is that the power and importance that civil service has is dramatically overlooked by many. Fukuyama takes this point and then describes the ways to measure the effectiveness and actually just how difficult that is. Though he goes through problems with all the different approaches, the main theme seemed to be the outcomes and products of a bureaucracy are all so endogenous that it is difficult to measure precisely what works and what doesn’t. He strongly asserts, however, that a largely autonomous bureaucracy is the strongest bureaucracy. While I understood his reasoning for this, I feel like this argument is very situation based and that in many places a largely autonomous bureaucracy would have too many risks to be the best option.
I agree that the effectiveness of autonomous bureaucracy depends various from case to case. It relies on the magnitude and nature of issuing such directives by the principal. There could be multiple legitimate principals issuing overlapping and conflicting directives. As for example, in China, duplicate functional agencies are found where one reports following a chain of command to the concerned ministry while the other reports to municipal or provincial governments.