As I explained in a post a few years ago called "The bureaucracy paradox", the American sociologist Charles Perrow famously wrote "...the more bureaucratized an organization, the more possibilities there are for decentralized decision-making...You decentralize, we might say, by centralizing." When I first came across this statement in Perrow's 1977 article "The bureaucratic paradox: The efficient organization centralizes in order to decentralize", I have to admit that I was thoroughly confused. In many ways, the traditional bureaucratic organization is hierarchical in nature, and decisions are made by a relatively small number of individuals who work at the top of the hierarchy. Bureaucracy, as a word, is almost synonymous with centralization. However, as I've experienced working in different organizations with different organizational structures and cultures, I have come to appreciate how much truth resides in Perrow's statement.
According to most authorities on the subject, when it comes to organizational structure, centralization versus decentralization refers to the locus of decision-making authority. In a centralized organizational structure, individuals who lie at the periphery (or, as some describe it, the bottom of the hierarchy) of the organization send information to those individuals in the center (the top of the hierarchy), where a decision is made and sent back to the individuals at the periphery to execute. Decentralization flips this around, giving individuals who lie at the periphery (the bottom of the hierarchy) the authority (and accountability) to make decisions. Retired U.S. Navy Captain David Marquet refers to centralization as "pushing information to authority" and decentralization as "pushing authority to information".
As with just about everything, there are certain advantages and disadvantages to centralization versus decentralization. Proponents of centralization, as an example, will often argue that it makes the process of gathering and processing information much more efficient. Centralization also circumvents coordination problems and organizational politics ("Not everyone is going to be happy with the decision, and that's perfectly okay"). Conversely, proponents of decentralization argue that by giving individuals at the periphery (to use the description above) the opportunity to provide input and take part in decisions increases engagement, builds trust and cooperation, and ultimately increases the chance of a successful implementation, particularly one that may be controversial.
Hala Altamimi, Qiaozhen Liu, and Benedict Jimenez studied (see "Not too much, not too little: Centralization, decentralization, and organizational change") whether the degree of centralization influences the implementation of four different types of organizational change in their national study of U.S. city governments who were confronted with a severe budgetary crisis in the aftermath of the Great Recession of 2007-2009. These local governments responded to the crisis by significantly changing the organizational structure (reorganization), contracting out services previously provided by the local government (service contracting, or outsourcing), adopting new technology (technology adoption), or instituting performance management with defined metrics (performance information).
These investigators found that neither full centralization nor full decentralization works well, particularly during a period of crisis or time of constraint. Instead, those organizational structures that have elements of both may perform the best, particularly when organizational change is required. Just as important was the finding that context matters. When organizational change is likely to be disruptive (reorganization or technology adoption), an organizational structure that leans toward moderate centralization is likely to be preferred. Conversely, if organizational change is less disruptive in nature (service contracting or performance information), an organizational structure that leans toward moderate decentralization is preferred.
Apparently, this contingent approach to the centralization/decentralization debate has been known for quite some time. Henri Fayol argued in 1949 that, "...the question of centralization or decentralization is simply a matter of proportion, it is a matter of finding the optimum degree for the particular concern." Others have labeled this contingent approach, "decentralized centralization" or even a networked approach (see also, Stanley McChrystal's "Team of Teams" approach).
The best analogy may be historical. From 1781 to 1789, the United States were governed by the Articles of Confederation, a truly decentralized model in which the 13 states (at the time) existed as more or less distinct entities with very little in the way of central oversight. The U.S. Constitution divided power between a federal (or central) government and the state governments. While the state governments still had significant autonomy, they were now also subject to a more powerful central government. The Constitution gave certain "powers" (authority) to the federal government, such as authority over national defense, currency, interstate commerce, etc. In this way, the U.S. Constitution decentralizes governance across today's 50 states, allowing decisions to be made at the local level, while at the same time, centralizing authority in the federal government to maintain unity, enforce rights, and manage national priorities. In other words, "decentralized centralization".
I think we've covered enough for today. I am still trying to wrap my brain around this topic, so I will likely keep coming back to it in future posts. For now, I think the take-home message is paraphrasing what Perrow said, "You centralize so that you can decentralize..."
No comments:
Post a Comment