The central problem in any bureaucratic organization, whether in the public or private sector, is how to delegate sufficient authority to agents who have expertise and are close to sources of local knowledge, while at the same time maintaining overall control over their behavior. In an ideal organization, the principals set the organization’s overall goals, but the agents are given substantial autonomy in implementation. The need for delegation increases as the organization’s size and complexity increase. Further, as the complexity and skill requirements of the external environment increase, the organization’s need to delegate will also grow.
The need for bureaucratic autonomy can be illustrated by the US Army’s incorporation of the concepts of “mission orders” and “commander’s intent” into its combined arms doctrine in the years following the withdrawal from Vietnam. (Those interested in a more detailed account should look at the Rand study that I wrote with Abe Shulsky, The Virtual Corporation and Army Organization, MR-863-1, 1997.) The experience of the First World War had led France and Germany to opposite conclusions with regard to operational doctrine. The French placed great emphasis on coordinated fires and minute synchronization of the movement of large forces, which required a high degree of centralized control over lower-level units. The German General Staff, by contrast, began to experiment toward the end of the war with storm trooper battalions, highly trained small infantry units that were given a high degree of autonomy to probe for and exploit weaknesses in the enemy’s positions. While Germany lost the war, this idea evolved into the Blitzkrieg tactics of Heinz Guderian and the General Staff of the 1930s. The resulting doctrine of Auftragstaktik empowered panzer units to operate under their own initiative subject only the most general of orders from higher echelons. Needless to say, the German model of highly delegated authority, when combined with appropriate technology (the tank and mobile radio), showed its worth against the more centralized French model in the summer of 1940.
During World War II, the US Army tended to be more centralized than the Wehrmacht, though individual commanders like George Patton had internalized German maneuver warfare doctrine. (In general, fire support, intelligence, and logistics–all things that the US is relatively good at–tend to drive centralization.) Overcentralization of the command structure was a big problem in Vietnam, where President Johnson was reported to have picked individual bombing targets. As the Army sought to reform itself after 1975, it refocused on conventional land warfare, and began to take seriously the incorporation of the principles of Auftragstaktik into its own doctrine. This resulted in a revision of its combined arms field manual, FM-100-5, in 1986, which placed emphasis on the concept of “mission orders” and decentralized execution.
The idea behind mission orders is that high-level commanders should focus on strategy and operations, while delegating as much authority as possible for actual implementation to the lowest possible levels of the command structure. The “commander’s intent” (i.e., the mandate issued from the principal to the agent) should be expressed as simply and clearly as possible, with every effort to avoid micromanagement. In the words of FM-100-5, “it is essential to decentralize decision authority to the lowest practical level because over-centralization slows action and leads to inertia.” Or, as another document from the Army’s Training and Doctrine Command (TRADOC) put it, “Battalion commanders must believe that they have the trust and confidence of the brigade and division commanders. They cannot waste time second-guessing their decisions and worrying about whether they are meeting their commander’s intent… Commanders must have the ability to listen as well as the ability to dictate because subordinates have things to say which are pertinent.”
The generation of officers trained under this doctrine achieved some remarkable military successes. These included the operations in the deep desert by Gary Luck’s 18th Airborne Corps and the 24th Mechanized Division during the 1991 Gulf War, and the Thunder Run by the 3rd Infantry Division that took Baghdad during the 2003 Iraq War. In all of these cases, commanders on the ground seized the initiative in ways unanticipated by their higher-ups.
Delegated autonomy necessarily involves risk, both for the agent and for the principal. The best military organizations understand that junior officers need to be given the “freedom to fail”: they will never innovate if they fear being punished later for taking risky decisions. (Ariel Sharon took a big risk and lost in the Mitla Pass during the 1956 Suez War; this didn’t stop his subsequent rise through the ranks of the Israel Defense Forces. By contrast, commanders in many Arab armies like those of Saddam Hussein were terrified of being second-guessed by their political superiors, and therefore failed to take initiative or risks.)
So what does this have to do with contemporary governance issues? In my view, mission orders represents a good example of how to define an appropriate degree of bureaucratic autonomy in a government organization. There is a “sweet spot” between too much and too little autonomy which organizations have to hit. No lower-level commander should set strategic or operational goals–that is the job of the senior commanders, operating at the behest of their political principals. On the other hand, the lower levels of the organization need to be given as large a degree of autonomy as possible, since they are close to sources of local knowledge, and have the expertise to actually implement their orders.
Many of failures of modern government are due to a failure to hit this sweet spot. The German and Japanese militaries during the 19th and early 20th centuries, as well as the civilian bureaucracies standing behind them, were famous for their elite training and autonomy. But they were too autonomous: the younger Moltke in 1914 and the Kwantung Army in 1933 essentially began to make German and Japanese foreign policy on their own, with disastrous results.
In democratic political systems, there has been a longstanding tension between the principals (the people and their elected representatives) and their bureaucratic agents. The system’s legitimacy depends on the fact that citizens and voters are the ultimate principals who get to mandate the behavior of their agents. But democracies are constantly tempted to over-mandate their agents, or else pass down overtly contradictory mandates. Outside the military, the US government has experienced numerous examples of excessive bureaucratic autonomy (of which PRISM would seem to be a leading contemporary example) and of the failure to permit sufficient autonomy. More examples to come.