Can the “Secret Government” Save Us?

A national security expert says professional bureaucrats held back Obama. Can they do the same to Donald Trump?

President-elect Donald Trump speaks during a meeting at the U.S. Capitol on Thursday in Washington, D.C.

Zach Gibson/Getty Images

In October 2014, the Ideas section of the Boston Globe published a piece under the headline, “Vote All You Want. The Secret Government Won’t Change.” The story, based on an interview with Tufts University professor Michael J. Glennon, quickly became a runaway hit on social media. In it, Glennon explained that President Barack Obama had failed to follow through on some of his key campaign promises—for instance, closing down Guantánamo Bay and ensuring Americans’ privacy was protected during the war on terror—because a legion of professional government bureaucrats prevented him from doing so. These individuals, Glennon said, steered policy to an extent that most Americans, with their idealistic assumptions about the power vested in elected officials, could not imagine.

Glennon has credibility. He’s the former counsel to the Senate Foreign Relations Committee as well as a consultant to a number of congressional committees and the State Department. As the Globe pointed out, his book, National Security and Double Government, was blurbed by people who had worked in the Department of Defense, the White House, and the CIA. That’s part of why the interview—in which Glennon stated that the “American people are deluded” in their belief that “when they vote for a president … policy is going to change”—was such a hit. It seemed like proof that there really was a secret government calling the shots—at least in the realm of national security, which Glennon focused on—while the politicians whose faces we know paraded in front of us on television like so much window dressing.

I thought about Glennon’s thesis last week, after it sank in that Donald Trump had won the presidency and would soon be installing his pack of pet goons at the top of the country’s federal agencies. It struck me as a reason for optimism: If in 2014 the notion of a “secret government” felt like dismaying evidence that the system was rigged against Obama’s ambitious vision for change, in 2016 it feels like a promise that Trump won’t manage to single-handedly destroy America’s institutions.

I called Glennon at his office at the Tufts Fletcher School of Law and Diplomacy to find out if I was right to feel reassured by his thesis. At first, I got exactly what I had come for. Then, to my chagrin, Glennon stopped saying things I wanted to hear. Our conversation, which has been condensed and edited for clarity, is below.

Leon Neyfakh: In your book, you argue that there are essentially two governments in Washington. There’s the one made up of elected officials, who are put there by the people, and then there’s this other group that is composed of, basically, professional government workers who are actually in charge of setting policy. Do I have that right?

Michael J. Glennon: My book is an application of a theory devised in the 1860s by Walter Bagehot in his book The English Constitution. According to the theory, there were two sets of institutions in Britain. The so-called dignified institutions—the House of Lords and the monarchy—presented a façade of authority and decisional power that enjoyed legitimacy in the eyes of the British people, and a set of concealed institutions—“efficient” institutions—that operated behind the scenes and actually formulated and implemented governmental policy.

I take this theory in the book and ask whether the progeny of England might also be analyzed through this dualist framework. And the suggestion in the book is yes. The “dignified” institutions are the ones that are constitutionally established: the presidency, Congress, and the courts. And in the realm of national security, the “efficient” institutions actually engage in the process of defining national security and formulating measures to safeguard it.

And what are those institutions?

They are not, in fact, institutions so much as a network of managers made up of several hundred individuals who direct the intelligence, national security, law enforcement, and military departments and agencies of the U.S. government. [Their existence] explains the strange continuity in national security policy from the Bush administration to the Obama administration. The first three pages of the book are a list of policies—from drone warfare to prosecutions of whistleblowers to the invocation of state secrets privilege, mass surveillance, you name it— which have been, for all intents and purposes, virtually identical through the two administrations.

That in a nutshell is the nub of the thesis of the book. The obvious question it poses, which I suppose may be impelling your call to me, is what are the implications of this for the Trump presidency? And to what extent can we look to this managerial network to serve as a check on potentially repressive policies in a Trump administration?

Before we get to that, I’m wondering: Who are the people in this managerial network? What are their jobs and where do they come from?

The short answer is they come from several different sources. Some are careerists. Some are in and out-ers, meaning they’re people from law firms or academia who get appointments in the national security bureaucracy at managerial levels for a specified period of time. There’s no one career path. Some are presidential appointees, but that number is fairly small in the national security realm. It’s much lower than most people realize. And among those appointees, many, many are carry-overs from the prior administration. The list of people in the Obama administration who also worked in these jobs in the Bush administration is long. And those who did not were heavily influenced by the carry-overs.

One thing that’s surprising about your thesis is that you’re suggesting these people are more powerful than the elected officials who are putatively in charge of directing them. Why is that? Most people aren’t more powerful than their bosses.

The short answer is that policy in these sorts of organizations is not top-down. That’s the myth under which the public operates. Policy proposals percolate up from within the bureaucracy, and inevitably policy proposals that are generated at the top are met with resistance for a wide variety of reasons. The authors of the extant policies typically are still present in the organization, or their protégés are. They have a vested interest in the continuation of policies that they helped create. So the policies develop a life of their own.

I would have assumed that when a new administration takes power, a lot of the people who staff federal agencies are replaced with people whose thinking is more in line with that of the new president. Is that not true?

It’s largely untrue. It is true, of course, that there is a turnover among the most visible appointees, be it a new national security adviser and secretary of defense and secretary of state and director of the CIA. But the second- and third-level jobs are not going to have any turnover for a long period of time, and may not have any turnover at all. Charlie Savage’s book describes this brilliantly, in case after case after case. Read the chapter on Greg Craig and Eric Holder sitting down for the first time with the leadership of the NSA and CIA to learn about the mass surveillance programs. They have virtually none of their own people under them who can brief them on what they’re about to hear or what questions to ask. So you know, [James] Clapper and his crowd comes in and unloads with lots of jargon and references to programs Craig and Holder have never heard of before. They don’t have time to engage in tracking down the citations and references to figure out whether any of these programs are legal, what the history of them is, who’s thought them up, where they originate. They’re still trying to figure out where the bathroom is. I’m sure that’s exactly how it’s going to be at the outset of the Trump administration. They’ve got to rely very heavily on the bureaucracy for these answers, and people who have been there and had experience and have participated in the formulation of these policies.

So that gets us to Trump, who, as someone who has never worked in government before, certainly doesn’t know where the bathrooms are. What should we expect in terms of continuity on national security when he takes office?

Well, the first question is, what does he want to do? And what does he tell the bureaucracy to do? He’s told us at various points during the campaign some contradictory things about what his policy directives are going to be. So take a worst-case scenario. Suppose he really tries to deliver on the most draconian policies he has described, like reinstituting waterboarding or “taking out the families” of people he suspects of being terrorists, both of which he has said he’d do. What is the effect going to be within the agency? What happens if he tells the intelligence services to institute these sorts of policies and they say no? I mean, the short answer is he fires them. He fires whoever says no and causes those people to be replaced with people who say, “Yes, Mr. President.” And there’s no difficulty in finding people who will say that.

There’s a long line of individuals willing to replace somebody who is insubordinate. And there is a strong incentive to be subordinate and compliant, because you could lose your job, lose your income, be ostracized by your friends and professional associates. How many people in Washington are really going to stand up and say no? That just really doesn’t happen. If it did happen somehow that the directors of these programs ended up successfully defying a presidential order, what would happen to the legitimacy of democratic government? Suddenly you’ll have reversed the power flow that is set out in the Constitution. Instead of having power flowing from the constitutional institutions to the departments and agencies created by the president and Congress, you have power flowing from them to the presidency. There are other twists and possibilities we can talk about, but I just don’t see it working. I don’t see bureaucratic checking as a realistic way of stopping a populist authoritarian president.

OK. That is not what I expected you to say. Everything you’ve told me up to this point made it sound like there’s this bureaucratic inertia that comes from having this network of managers that we can count on to create continuity.

Well, let me continue. There’s another possibility, and that is that the managers would not engage in direct defiance of a presidential order, but rather that, backstage, behind the scenes, they would slow-roll what the president wants to do. They would effectively present the president with an intensified form of the bureaucratic inertia and torpor that has, in the past, caused the bureaucracy to be such an impediment to changing national security policy. So the question is, would that work as a check? Yes. That phenomenon is present. Whether it’s intended or not, that’s the way bureaucracy works—it’s part of the plumbing, and you can’t turn it off by twisting the faucet. The sluggishness and slowness is part of the system and it could be more protracted if there’s a measure of intentionality behind it.

However, while that could work in the short term, it’s doubtful it would work in the long term. Ultimately, the leadership of the bureaucracy—the Trumanite network, as I call these several hundred people—does have essentially the same overarching objective as the leadership of the Madisonian institutions, and that objective is to appear in the eyes of the public to be on the same page. Because when there is a public dispute between the two, they lose legitimacy. So there’s a strong incentive to remain in sync, and over the medium and long term, you would most likely find a kind of power-sharing arrangement in which the leadership of the national security institutions comes to adapt and accommodate themselves to president Trump’s policies.

I don’t really understand what you’re saying. Wasn’t the point of your book that Obama was overpowered by this network of managers, in terms of following through on his objectives going into his administration? It sounds like you’re saying now that Trump will be able to impose his will on them.

Well, it’s purely hypothetical at this point, but the situation you’re imagining is that the managers of the national security institutions in a Trump administration would seek to play the role of a check and to do so publicly, by defying a presidential order. They haven’t done that in the Obama administration. The Obama administration and the Trumanite network have been on the same page. But we’re hypothesizing a situation in which these institutions play a very different role, and that they are not publicly in sync. When that condition ceases to obtain, in Bagehot’s theory, the consequences are that the whole structure falls to earth. That’s really the situation that we’re speculating about, in which the national security bureaucracy stands up and says no. And that did not happen during the Obama administration.

So then why didn’t he do all the things he said he wanted to do? Weren’t you just telling me this bureaucracy, this network of managers, was the reason Obama didn’t close Guantánamo and all the other things on his list?

As I understood your question, you’re raising the specter of a public confrontation in the Trump administration versus the normal, quiet bureaucratic inertia.

No, I’m asking if that quiet bureaucratic inertia will be enough to rein him in. If we were having this conversation eight years ago and I was someone who was terrified of Obama’s proposals on Guantánamo and Afghanistan and the NSA, you would have told me, “Well, he probably won’t be able to do those things,” right?

What’s different is you’re speculating about the occurrence of a public split between the network of managers and the president. That never happened in the Obama administration. We’re talking about the network of managers publicly, overtly, possibly with great fanfare saying “no.” And that goes to the core of the condition required to make double government work, which is that the dignified and efficient institutions operate on the same page, that they operate in sync. If there’s an open split between the two, the legitimacy of each is undermined. That’s the difference between the situation we’re speculating about and the one that prevailed during the Obama administration. There was no public split.

Why not?

Why did the Obama administration not stand up to the NSA or the CIA? If you buy Bagehot’s theory, one of the principal reasons is that the dignified institutions—the Obama administration—would have lost legitimacy by publicly splitting with the efficient institutions, and it’s essential to the maintenance of legitimacy that they appear to be operating on the same page.

OK. I guess I don’t understand why the Obama administration made that calculation but you don’t think the Trump administration will. Is it specific to Trump? Are you saying he, specifically, is the kind of leader that would overpower this network of managers?

Well, according to your hypothetical, the one we’ve been talking about, it’s not the Trump administration that jumps off the page—it’s the CIA or the military that says no, right?

Sure, but didn’t they say no with Obama? Isn’t that the whole point of your book? That their inertia was so powerful that they could prevail over the president’s will?

I wish I could send you a copy of the book before you write this.

I don’t know if that would help. I mean, correct me if I’m wrong, but the thesis of your book is that there are these people who work for the government, and they’re gonna do what they’re gonna do, and they’re gonna do it regardless of what the elected officials want, because they have these quiet ways of asserting themselves. But you’re saying that won’t hold under Trump. Right?

If Trump directs the Central Intelligence Agency to resume waterboarding and people say no, or President Trump directs the military to take out the families of suspected terrorists …

You’re saying there’s no quiet way to oppose that?

It’s very difficult to speculate about what hasn’t happened yet.

OK. Well, I am not reassured.

You’re not reassured in which direction?

I thought you were going to tell me that there’s enough power in this network of managers that we don’t have to worry about the prospect of the Trump administration realizing our worst fears.

I’m sorry to disappoint you.

I’m sorry too. I think the reason that interview in the Globe was so resonant with people is that it suggested our government is much less democratic than it seems—that we elect these people but actually they’re relatively powerless against the bureaucracy they oversee. With Trump in power, I would actually like that to be the case.

A lot of people would like that to be the case. I think it’s a false hope.