War is too important to be left to the generals.
While Austin is not the candidate I was hoping Biden would select—I have been an advocate for Michèle Flournoy—I admire his military career. My objection to Austin’s expected nomination is based on an important principle: The U.S. military should always remain under civilian control. The National Security Act of 1947, which created the Department of Defense, prohibited military officers from becoming secretary of defense within ten years of retirement. A later amendment reduced that to seven years. Austin retired in 2016.
There is a good reason for the requirement that recent officers not run the Department of Defense. Former Secretary of Defense Robert Gates wrote in his memoir that “I would often tell [General David Petraeus] that Iraq was his battlespace and Washington was mine.” This is because military officers are taught to be apolitical, while the secretary of defense is a politician who has to weigh the interests of the Department of Defense as an institution, geopolitical concerns, and domestic political realities.
War is an instrument of policy, fought by troops but won by politicians. In addition to battlefield violence, the successful prosecution of a war also requires the correct allocation of resources, the proper mobilization of industry and the economy, and the political support of the people. It requires boosting public morale through ideological mobilization. Military officers are not trained for this. And they should not be. As Peter Feaver, a leading expert on civil-society-military relations writes:
[Political concerns do not fall] within the scope of unique military expertise; on the contrary, [they hinge] on political or strategic considerations for which the professional competence resides at the presidential level. It is the president’s job to generate public support, and it is the president’s job to place the policy bet on how much the American people will tolerate.
Feaver mentions the president, but his advice extends to the secretary of defense as the president’s chief assistant on military matters.
As Samuel Huntington put it, “the peculiar skill of the officer is the management of violence.” In a democratic system like ours, officers can certainly debate strategy with their civilian superiors, but they must not engage in the formulation of policy—ever! The secretary of defense, by contrast, ought to have experience with policy development.
There is another place for the president’s senior military advisers: the Joint Chiefs of Staff. They are there to represent the institutional interests and views of the military. Naturally, they come with their biases, biases that are important and that deserve to be represented. The secretary, though, is not supposed to be an institutional figure within the department. By design, the secretary of defense is supposed to be an outsider, free of institutional bias.
A recently retired senior officer may be too close with the military—and is almost certain to be viewed that way by the department’s civilian staff. That was a problem with James Mattis, President Trump’s first secretary of defense. Mattis, a Marine four-star, had only retired from active service in 2013. As a result, his 2017 appointment required Congress to pass a bill to waive the requirement just for him. In Mattis’s case, House and Senate Republicans voted for the waiver unanimously (with the exception of Justin Amash in the House), while Democrats were split.
Waiving the law for Mattis made sense at the time: Donald Trump was a toddler and needed adult supervision. As Eliot Cohen said in his testimony in support of the waiver, “the president has to have somebody they listen to. I guess I do tend to believe that President-elect Trump will be inclined to listen to General Mattis.” In the extraordinary circumstance of a president so clearly unsuited for the office, it seemed that having a secretary of defense with the experience and steadiness of Mattis warranted the waiver. Even the Democrats who ultimately voted against the waiver made a point of expressing their respect for Mattis and their understanding of Trump’s deficiencies.
Unfortunately, the expectations for Mattis’s tenure were not met. First, as with the other so-called “adults in the room,” his moderating influence on Trump was hard to detect. True, there were some instances, not widely known, in which military operations with potential for disaster were avoided because Mattis dissuaded Trump. But anyone who hoped that Mattis could help make Trump more presidential, or at least less unpresidential, was disappointed.
Second, Mattis never quite fit into his suit as secretary; it always seemed as though, in his mind, he was still wearing a uniform. The leadership experience that served Mattis well at the highest ranks of the Marine Corps was a poor fit for leading a cabinet department. The modest National Defense Strategy he published—which noticeably focused on the “lethality” of the force, language you would expect from a Marine—was heavy on the operational capabilities and light on the strategy side. The problem of Mattis’s closeness to his service was compounded by the fact that the chairman of the joint chiefs, General Joseph Dunford, was also a Marine. The two were old friends. So the chairman of the JCS often reinforced, rather than constructively complemented, Mattis’s views. It did not help that Mattis never made any serious push to have the president fill the vacant under secretary and assistant secretary positions to add high-level civilian insight and oversight to the department.
As Jim Golby wrote soon after Mattis left the Pentagon, his tenure as secretary
(1) blurred the lines of authority between civilian and military, as well as between active-duty and retired military; (2) enabled the rapid erosion of civil-military norms; and (3) widened gaps between the military and American society as well as between the military brass and elected political leaders.
Having said this, Golby concedes that Mattis’s tenure might have been worth it in the context of putting a brake on Donald Trump’s worst impulses. But the Mattis record shows why it is important for the secretary of defense not to come from the Pentagon’s own bureaucracy.
And yet: Here we go again.
The divide between the civilians and the military in the Pentagon has sometimes been characterized this way: Every civilian is a “Cohenian”—that is, they agree with Eliot Cohen’s thesis that strategy belongs in civilian hands after consulting the military. And every officer is Huntingtonian—they agree with Samuel Huntington that the military alone should formulate strategy. Cohen’s thesis fits a liberal democracy better, but Mattis governed as a Huntingtonian. We should be concerned that Austin—or any retired senior officer—will do the same.
And Austin has his own analogue of Mattis’s problem with Dunford: The current chairman of the joint chiefs of staff, General Mark Milley, is an Army officer who is close with Austin. As a West Point graduate whose career was defined by the Army, Austin will inevitably face the reasonable suspicion of the other services that he and Milley will pick favorites against their interests.
Moreover, the context that seemed to justify a waiver in Mattis’s case—the unsuitedness of the president for the office—does not apply in Austin’s case. In Joe Biden we have the prospect of an experienced and level-headed commander in chief.
Mattis and Austin are not the only men to have required a congressional waiver. In 1950, President Truman nominated George C. Marshall, who had been Army chief of staff until 1945, to serve as secretary of defense. Marshall needed the waiver even though he had already served as secretary of state under Truman.
Marshall was secretary of defense for only one year, but it was a fateful one. On his watch, Douglas MacArthur began drifting away from the administration’s strategy in Korea, publicly criticizing the commander in chief. Marshall understood the problem, but his military background and his old friendship with MacArthur clouded his judgment. He tried to navigate the problem passively and to protect MacArthur—but that only encouraged MacArthur further, which led to one of the biggest controversies of Truman’s tenure.
When Congress debated the waiver for Mattis, the ranking member on the Senate Armed Services Committee, Jack Reed (D-R.I.), said:
Waiving the law should happen no more than once in a generation. Therefore, I will not support a waiver for future nominees, nor will I support any effort to water down or repeal the statute in the future. This requirement has served the nation well for the past seven decades. It is up to this committee to ensure that the principle of civilian control of the armed forces, which is the bedrock of civilian-military relations, remains the defining tenet of our democracy.
That is the level of commitment to civilian supremacy that one should expect from our leaders. Members of Congress should insist on holding to that standard again.