Reuters / KCNA

As South Korea’s national-security adviser told it on Thursday, Donald Trump will meet with Kim Jong Un this spring for one purpose only: to achieve the “permanent denuclearization” of North Korea. But according to one of the U.S. officials who came closest to striking that kind of deal, the president better lower his expectations. By a lot.

Kim seems willing to talk because his nuclear-weapons program is now so advanced—on the cusp of boasting nuclear-tipped long-range missiles that can reach the United States, along with dozens of bombs and a diffuse and diverse arsenal—that he believes he can deter aggression by his enemies and garner international respect, former Defense Secretary Bill Perry told me on Thursday, shortly before news broke of the planned Trump-Kim summit.

“We had the opportunities to keep [North Korea] from getting a nuclear weapon and that would have been far better, but we blew that opportunity,” he said. “Now they have a nuclear arsenal, and that’s the reality. They’re going into this discussion from a position of relative strength” even though they’ve been pummeled by the Trump administration’s international campaign of economic pressure and diplomatic isolation.

“I don’t think [the North Koreans are] going to want to negotiate giving up all their nuclear weapons,” he added. “But even if they did … I have no idea how we could verify it.”

Perry experienced one of those missed opportunities first-hand. As Bill Clinton’s defense secretary, Perry had seriously considered military strikes to destroy a North Korean nuclear reactor. But in the late 1990s, Clinton asked him to salvage and expand on a fraying 1994 agreement that shut down the North's rudimentary nuclear-weapons program. Perry traveled to Pyongyang, Secretary of State Madeleine Albright met with North Korean leader Kim Jong Il, and a top North Korean military officer sat down with Clinton in the Oval Office. By the fall of 2000 there was buzz about Clinton himself heading to North Korea to begin establishing normal diplomatic relations with America’s longtime foe in exchange for Kim stopping work on nuclear weapons and long-range ballistic missiles. Then came the 2000 presidential election. George W. Bush took a harder line on North Korea and in 2002 discovered that the North Koreans were clandestinely enriching uranium for nuclear weapons. Perry’s initiative fell apart.

The years since have brought a series of nuclear agreements that at times froze the North Korean nuclear program, but over the long term failed to prevent the North from becoming a nuclear-weapons state. The achilles heel of many of these accords was the Kim government’s refusal to disclose all its nuclear activities and permit outside monitors to verify that those activities had ceased.

“In 1999 we had a chance of getting denuclearization. I do not believe we will get that today,” Perry said. Over the last two and a half decades, “we’ve learned we cannot really trust North Korea. … Therefore I’d put a high premium on verification. If we can’t get verification, don’t value too highly the agreement.”

The Trump administration could aim for a North Korean commitment to limit the number of nuclear weapons in its arsenal and not build new ones, but that “would be extremely hard to verify—approaching on impossibility,” he noted. “Suppose they agreed to turn in what they say are their nuclear weapons. We’re not sure how many they have in the first place, so if they turn in 15 or 20 we’re not sure how many might be left. Secondly, suppose they allow monitors into their centrifuges at [the plutonium reactor known as] Yongbyon. We don’t know where they have their other centrifuges and centrifuges are reasonably easy to hide. So they might be churning away making more highly enriched uranium.”

Establishing safeguards against North Korea transferring nuclear components and technology to other states or non-state actors like terrorist groups would be difficult to verify but still worth pursuing in negotiations, Perry said. (North Korea has a history of proliferating missiles and other materials related to weapons of mass destruction.)

What could be verified with relative ease is a ban on North Korea’s tests of nuclear bombs and long-range missiles. We might not know where all the North’s nuclear weapons and facilities are, but we do know when they conduct earth-shattering explosions underground or fire giant rockets into the Sea of Japan. Stopping that, while perhaps a modest result, “would be worth having,” Perry argued, because it would keep Kim from putting the finishing touches on intercontinental ballistic missiles capable of delivering a nuclear warhead to the U.S. mainland.

But what might have bought denuclearization in the 1990s, when North Korea was contemplating trading away its right to a nuclear-weapons program, has less purchase in 2018, when North Korea is right on the verge of becoming a full-fledged nuclear-weapons power. In return for a moratorium on nuclear and missile tests, Perry proposes incentives that he dangled during the Clinton administration—from North-South Korean economic enterprises to a peace treaty formally ending the Korean War to the gradual normalization of U.S.-North Korean relations, by setting up initially a U.S. interests section and eventually a U.S. embassy in Pyongyang. (In negotiating with the North Koreans, Perry learned that the Kim government prioritizes its survival above all else, followed by acquiring international prestige and, in third place, improving the economy.)

While recognizing North Korea diplomatically and finally concluding the Korean War might seem like grand gestures, Perry argued that they are actually “easy and cheap” for the United States to implement—and, maybe most importantly, “reversible” in the event that North Korea reneges on its end of the bargain. The outcome Perry envisions is, as he put it, possible, desirable, and verifiable. It's also a far cry from the denuclearization of North Korea.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.