An earlier version of this op-ed said that the Goldwater-Nichols Act of 1986 names the chairman of the Joint Chiefs of Staff as the National Security Council's exclusive military adviser. That act names the chairman as the president's principal military adviser.
Generals Shouldn't Disagree Publicly With the Commander in Chief
The president, the Constitution tells us, is the commander in chief. But is it true?
In a speech in London on Thursday, Gen. Stanley McChrystal publicly intervened in the debate over Afghanistan. Vice President Biden has suggested that we focus on fighting al-Qaeda and refrain from using our troops to prop up the government of President Hamid Karzai. But when this strategic option was raised at his presentation, McChrystal said it was a formula for "Chaos-istan." When asked whether he would support it, he said, "The short answer is: No."
As commanding general in Afghanistan, McChrystal has no business making such public pronouncements. Under law, he doesn't have the right to attend the National Security Council as it decides our strategy. To the contrary, the Goldwater-Nichols Act of 1986 explicitly names the chairman of the Joint Chiefs of Staff as the National Security Council's principal military adviser. If the president wanted McChrystal's advice, he was perfectly free to ask him to accompany Adm. Mike Mullen, chairman of the Joint Chiefs, when the council held its first meeting on Afghanistan this week.
But Obama did not extend the invitation, even though McChrystal was leaving Kabul and could have gone to Washington easily. Instead, Obama asked the general to report to the council via a brief teleconference.
News of McChrystal's position had been leaked to Bob Woodward and was published in The Post early last week. But it is one thing for some nameless Washington insider to engage in a characteristic power play; quite another for McChrystal to pressure the president in public to adopt his strategy. This is a plain violation of the principle of civilian control.
McChrystal seemed curiously blind to this point. He emphasized that the president had "encouraged" him to be blunt when making his grim report on Afghanistan. But future presidents won't be so encouraging if they know that their commanders might create political problems if they think that their recommendations will be overruled. Instead, they will insist that their commanders tell them only what they want to hear. Confidentiality is a condition for candid communications between commanders and the commander in chief.
McChrystal was almost cavalier in dismissing this point. After praising his superiors for encouraging straight talk, he laughingly suggested that "they may change their minds and crush me some day." This is precisely backward: Generals shouldn't need to be told that it is wrong to lecture their presidents in public. Perhaps McChrystal was misled by the precedent set by Gen. David Petraeus, who strongly supported President Bush's military surge in Iraq in 2007. Though Petraeus publicly endorsed the surge, this happened only after Bush made his decision. Petraeus was backing up his commander in chief, not trying to preempt him.
Nevertheless, precedents have the habit of adding up. Unless McChrystal publicly recognizes that he has crossed the line, future generals will become even more aggressive in their efforts to browbeat presidents.
We have no need for a repeat of the showdown between President Harry Truman and Gen. Douglas MacArthur over Korea. Truman faced down his general the last time around, but it was a bruising experience.
Though McChrystal may feel "crushed," he should show more self-restraint. Indeed, his breach should provoke a broader discussion of the meaning of civilian control in the 21st century. It may well make sense for the Pentagon, or a special commission, to frame more concrete guidelines so that we may avoid future breaches.
The writer is a professor at Yale Law School.