by John Bolton;
The next day, Monday, August 24, I concluded, amazingly, there had been no meeting. There was certainly no media coverage of a meeting, although French Foreign Minister Jean-Yves Le Drian had met with Zarif for over three hours, joined at one point by Macron.73 The French told Kelly Ann Shaw that Macron gave Trump a one-on-one readout of those conversations in Trump’s hotel room but apparently informed no one else. When I talked to Mulvaney just before Trump’s first bilateral that morning, he said he didn’t think there had been a meeting with Zarif. I e-mailed this news to Pompeo at about ten thirty a.m., saying I couldn’t rule out a phone call, and I also wasn’t sure whether Kushner or Mnuchin might have met or spoken with Zarif, to create a future channel of communication. (This latter hypothesis was something that I believed agitated and worried senior Israeli officials, and which of course made Pompeo livid.)
I don’t know if I alone talked Trump out of meeting Zarif, but the decision was enough for me to travel on to Kiev rather than return home. Still, how much longer could it be before Trump made a truly bad, irreversible mistake? Yet again, we had only postponed, and maybe not for long, the day of reckoning.
CHAPTER 13 FROM THE AFGHANISTAN COUNTERTERRORISM MISSION TO THE CAMP DAVID NEAR MISS
I knew what I wanted to achieve in Afghanistan, and Trump’s other senior advisors shared my two objectives. Tied for first, they were: (1) preventing the potential resurgence of ISIS and al-Qaeda, and their attendant threats of terrorist attacks against America; and (2) remaining vigilant against the nuclear-weapons programs in Iran on the west and Pakistan on the east. This was the counterterrorism platform1 we wanted to pursue in early 2019.2 The hard part was getting Trump to agree and then stick with his decision. If these objectives were presented poorly, or timed badly, we risked another outburst in which Trump would demand we withdraw everyone immediately; not presenting them meant withdrawal by default.
Zalmay Khalilzad’s ongoing negotiations with the Taliban constituted another layer of complexity. Pompeo believed he was carrying out Trump’s mandate to negotiate a deal lowering the US troop presence to zero. I thought this was clearly bad policy. In theory, the US government opposed any such arrangement unless it was “conditions based,” meaning we would go to zero only if: (1) there were no terrorist activities in the country; (2) ISIS and al-Qaeda were barred from establishing operating bases; and (3) we had adequate means of verification. I thought this was touchingly naïve, much like the Pentagon’s views on arms control: we make a deal with some gang of miscreants, and they adhere to it. How nice.
From the outset, Pompeo insisted it was the Pentagon that wanted a deal with the Taliban, to diminish threats to US personnel as we reduced our presence; without such an agreement, the risks to the shrinking US forces were too great. But again, I thought this was an almost childlike view. I never understood why such a deal gave us any real protection from a group of terrorists we had never trusted. If the Taliban, ISIS, and al-Qaeda concluded from the plain evidence of palpable US troop drawdowns that we were withdrawing, and that raised risks to our diminishing forces, what would those terrorists conclude from a piece of paper that expressly said we were going to zero by October 2020?
“Conditions based,” in the Afghan context, was like an opiate. It made some of us (not including me) feel good, but it was merely a temporary, ultimately hollow, experience at best. I doubted there was any deal with the Taliban we should find acceptable, given their track record. If “total withdrawal” was the target, violations of the “conditions” would not change that outcome, given Trump’s view. Once we were on the nosedive to zero, that’s where we would finish. But if continuing negotiations bought us time to prepare and then maintain a sustainable counterterrorism presence, then play on.
Shanahan, Dunford, Pompeo, and I all believed that the sooner we could brief Trump on how such operations would work in practice, the better. A briefing was scheduled for Friday, March 15, and preparations for it began in earnest. Knowing how much was at stake, we had a prep session in the Tank on the Friday before. Curiously, Pompeo had John Sullivan attend in his stead; perhaps he didn’t want to reveal what the actual state of play in diplomacy with the Taliban was before the Trump briefing, which would be consistent with his practice of sharing as little about the negotiations as he could. I wasn’t bothered by his absence, because I had concluded Afghan diplomacy wouldn’t matter much in the long run anyway. My horizon was more limited: how to make the Defense Department’s presentation to Trump the following Friday as effective as possible, thereby persuading him we needed to keep substantial counterterrorism resources in country. Typical of its briefings, the Pentagon had prepared charts and slides making everything more complicated than necessary, even when they were conveying “good” news to Trump, such as the substantial reductions in personnel and costs the military would achieve. I urged Shanahan and Dunford to pay close attention to Trump’s reactions as they briefed and not simply bull ahead reading the slides and charts. This was the Mattis style, and I had heard the stories, true or not, of Trump tuning out long, exhaustive briefings by McMaster. No need to go through that again. I thought everyone agreed, but the test would come the following week.
On the morning of March 15, Trump called Ethiopian President Abiy Ahmed Ali (who became the 2019 Nobel Peace Prize winner in October) to express condolences for the recent Ethiopian Air crash (leading to the worldwide grounding of all Boeing 737 Max airplanes). While Trump and I were talking, waiting for the call to go through, he raised Afghanistan and said, “We’ve got to get out of there.” It was not an auspicious way to start the day. I called him back after the Abiy call to explain that the briefing would largely consist of the Pentagon’s showing how it had redefined the Afghan mission based on his instructions to reduce the US presence. “It won’t hurt the negotiations, will it?” he asked, worried that pulling down our numbers would indicate weakness, and I said it would not. He asked that I ride with him in the Beast to the Pentagon, along with Pence, which provided another chance to take his temperature and address any concerns, but the conversation during the ride turned out to be largely about North Korea.
In the Tank, Shanahan said the briefing would explain how to maintain our counterterrorism and other wherewithal now that the Taliban was negotiating. Trump immediately interrupted to ask, “Do we weaken our hand in the negotiations by saying we’re dropping our forces?” I had called both Shanahan and Pompeo right after my earlier exchange with Trump, and their answers were well prepared. They said the timing was actually perfect. As is often the case, decimating the opposition often makes the survivors more eager to negotiate. Trump then discoursed on how bad previous commanders had been (unfair, but a frequent complaint), not to mention Mattis’s performance, despite Trump’s having approved the rules of engagement Mattis requested. “How are the negotiations going?” Trump asked, but cut almost immediately into Pompeo’s response, raising the endemic corruption among Afghan officials, especially Afghan President Ghani and his purported riches, although unfortunately confusing Ghani with former President Hamid Karzai, as he did constantly. Hopefully unnoticed, I signaled Dunford to jump in to say that the reduction in violence accomplished by our current strategy meant we could fulfill our ongoing counterterrorism and other missions in practice with reduced resources, even without agreement from the Taliban. “We listened to you,” Dunford said to Trump, which was a good line in the post-Mattis era. Dunford also said we could handle any weakening of the Afghan government, if that occurred, and focus on al-Qaeda and ISIS, the real terrorist threats to America. I pointed out that weak central government was Afghanistan’s historical default position and would be nothing new for the inhabitants.
Dunford further explained the need to maintain a counterterrorism presence for the broader region. As he launched into his charts and slides to show how our ongoing Afghan operations would be staffed and costed, Trump said, “There are still a lot of people there,” but fortunately he went on to say, “Having no one i
s dangerous, because they [the terrorists] tend to form there and knock down buildings,” which was exactly the point. Trump repeated one of his hobbyhorses, namely that it was cheaper to rebuild the World Trade Center than to fight in Afghanistan, inconveniently ignoring the loss of life in the 9/11 attacks, not just the cost of rebuilding. It also ignored the reality that a Trump withdrawal, followed by a terrorist attack, would be devastating politically. Dunford pressed on, saying our military pressure kept the terrorists from reconstituting and was like an insurance policy. He had no precise timetable in mind, but the Pentagon would allow the diplomatic reconciliation process to set the timing. I thought we were approaching dangerous ground here, again opening the question of whether we should be in Afghanistan at all. The discussion meandered around for a while, with Trump asking me why we were fighting in Iraq and Afghanistan but not Venezuela, which at least showed everyone else in the room what he really wanted to do.
Following more chatter, Shanahan turned to the cost reductions that maintaining the counterterrorism capability would entail, but before he got too far, Trump broke in to complain about Congress’s refusal to fund the Mexico border wall. Then he was off: “Why can’t we just get out of Syria and Afghanistan? I never should have agreed to the other two hundred [in Syria], and it’s really four hundred anyway.” Dunford explained that other NATO countries would hopefully contribute to the multilateral observation force in Syria, and Trump responded, “We pay for NATO anyway,” which in turn produced another riff about Erdogan and what he was doing in Turkey. Then after literally forty-five seconds back on Afghanistan, Trump asked, “Why are we in Africa?” He soon made it clear he wanted out of Africa altogether, expounding for some time on our $22 trillion national debt, followed by the problems of our balance-of-trade deficits, followed by complaining, again, about how Nigeria received $1.5 billion annually in foreign aid, as he said the President of Nigeria had confirmed to him in an earlier visit, even though they wouldn’t buy US farm products. After more Africa discussion, Trump returned to Afghanistan, saying, “Get it [the annual cost] down to ten billion dollars [a figure derived from the resources required by our thinking for the continuing US presence], and get it down fast.” That led to the costs of US military bases in South Korea and how much Seoul should contribute to defraying the expenses. Trump said happily we had extracted $500 million more from the South in negotiations at the end of 2018 (it was actually around $75 million, as pretty much everyone else in the Tank knew). He still wanted the South Korea payment to equal US costs plus 50 percent. And whatever our political differences with Iraq, Trump reminded us that, having paid so much to build our bases there, we were not leaving.
After a bit less than an hour, as we were wrapping up, Trump asked Dunford in front of everyone, “How’s our Acting Secretary doing?” Dunford, obviously stunned, came back quickly to offer, “Here’s where I say what a great job he’s doing,” and everyone laughed. I just wanted to get out the Tank’s door, so Pompeo and I simply stood up and started pulling papers together. Others then rose, and we all walked to the Pentagon’s River Entrance, where the motorcade was waiting. Pence and I rode with Trump in the Beast, where the conversation was mostly on the Max 737 grounding. Back in the West Wing, I called Shanahan to congratulate him on the outcome and say he should take the rest of the day off. It wasn’t pretty, as I also told Dunford when I called him a few minutes later, but it was definitely a win. Dunford said, “[Trump] now believes we’re listening to him, which he didn’t believe before,” which was emphatically correct. Pompeo, with whom I spoke the next day, also believed the briefing was a win.
The problem still not resolved, however, was the Taliban negotiations. At a breakfast on March 21, Shanahan and Dunford brought a chart showing several ways the State Department had departed from what the Pentagon believed were the agreed negotiating guidelines. Most troubling to me was that the State Department’s negotiating objectives were completely detached from what I considered to be our real objectives: being fully capable of preventing a resurgence of terrorism and remaining vigilant against the nuclear dangers of Iran and Pakistan. The resource levels Trump had implicitly approved at the Tank briefing were not near what we would need in a major crisis, but, in my view, at least we would still have several bases in Afghanistan, and the possibility of rapidly scaling up our capabilities. Pompeo and Khalilzad, however, were still negotiating as if we were withdrawing entirely. We may have started there in November, but we had clawed ourselves a long way back, and Shanahan, Dunford, and I feared losing that progress. Within thirty minutes of the breakfast, I called Trump and said it was his decision whether to let Khalilzad and the State Department act with complete independence in the negotiations, but I thought it was dangerous for what Trump said he wanted. “I don’t even know who he is,” Trump replied of Khalilzad. “Do what you think is best.”
That same morning, I met with Khalilzad, someone I had known, as I have said, for nearly thirty years. He said Pompeo had ordered him not to communicate with me because I was undermining Pompeo with Trump. That happened to be untrue, and I wondered if Pompeo’s real motivation was who would get credit for Afghanistan, a common Washington phenomenon. If true, it was misplaced. I didn’t think there was any material chance the negotiations would produce an acceptable result, so I was hardly eager to get “credit” for the outcome. Khalilzad agreed to an informal meeting with all the officials involved to work out the misunderstandings before he left Washington to resume talks with the Taliban. I thought this was a good decision. Several days later, however, Pompeo called to complain that senior Pentagon officials and Lisa Curtis, the National Security Council’s Senior Director for South Asia, were interfering with Khalilzad and should leave him alone. Normally, such meetings constituted “interagency coordination,” but Pompeo saw it as meddling. No wonder we had internal Administration problems on Afghanistan.
In the midst of these difficulties, we received one piece of extraordinarily good news, on April 12, when the International Criminal Court’s “pretrial chamber” handed down a thirty-two-page opinion rejecting its prosecutor’s request to open an investigation into the conduct of US military and intelligence personnel in Afghanistan. I spoke several times with former Congressman Pete Hoekstra, our Ambassador to the Netherlands, where the court was located, in the Hague, and found him nearly as surprised as I that we had succeeded in stopping this miscarriage of justice. I had long opposed the court,3 and I had given a speech early in my tenure to the Federalist Society in Washington about why the Administration rejected it as a matter of principle, and the steps we were prepared to take against the court if it presumed to target American citizens.4 I called Trump at nine fifteen a.m. to tell him about the decision, and he said, “Put out something powerful,” which I was delighted to do.
As for the US-Taliban meetings, I was less concerned than before about their substance, and therefore less exercised than the Pentagon, because I thought we had largely won the key battle in Trump’s mind. The United States would not be completely withdrawing from Afghanistan but would maintain a persistent troop presence for counterterrorism and other objectives. If it turned out Trump reversed course and one day simply said to leave immediately (which he did in outbursts periodically), even that decision didn’t depend on the state of play in the talks. In short, the US military posture was no longer tied to the peace process, if it ever had been. Thus, in my mind, there was no particular pressure for Khalilzad to produce results and no real target date to finish the negotiations.
Yet by July 1, Defense learned that Khalilzad was about to announce a deal with the Taliban without informing anyone in Washington what was in it. Not unreasonably, Esper called Pompeo, his West Point classmate, to suggest bringing the deal back to Washington for review. Kupperman heard from Esper’s Chief of Staff that Pompeo “lit up Esper” in no uncertain terms, and his subordinates, for involving themselves in the Afghan negotiations. Pompeo, often screaming, we were told, said Khalilzad was un
der instructions—he left unstated by whom—to make a deal without outside supervision. Shanahan had tried to be reasonable and failed, and now it was Esper’s turn. When I spoke to him the next day, his concerns were actually more tactical, such as his legitimate interest in the safety of withdrawing US forces, rather than aimed at the policy. “Mike got a little animated,” Esper put it politely, but as our conversation continued, there was clearly a disjunction between the State Department’s objective—zero US forces—and my (and the Pentagon’s) desire to preserve counterterrorism and other capabilities. Esper was rightly concerned with how the actual functioning of military operations would be affected if we lost even a foothold in Afghanistan.
I brought this schizophrenia to a head in a discussion with Khalilzad on July 19. My objective was to reach agreement within the US government from among all the various strands of thought, and to resolve the disconnects in our internal deliberations. Specifically, while his instructions from Trump (or Pompeo, whomever) at that time were to get US forces to zero, he also had instructions from Trump to support counterterrorism capabilities consistent with what had been previously briefed to Trump in the Tank, essentially without an end date. The trick was how to get the Taliban and the Afghan government to agree we were going to zero on the existing mission, while simultaneously creating a modified mission to support counterterrorism capabilities. Khalilzad readily agreed, saying he fully understood what Trump expected, so I considered this significant progress. My objective here was ensuring that our government agreed within itself in the context where we found ourselves at that point. We could have done this earlier and more easily if Pompeo had let Khalilzad have these conversations more often and we didn’t have to arrange them like spy encounters. In any case, the Taliban talks continued apace through the summer.
