All U.S. troops currently serving in Afghanistan will return to the United States by Christmas, President Donald Trump said in a shocking tweet Wednesday night just hours after his national security advisor said that the United States would draw down its forces in Afghanistan to 2,500 by early next year.

"We should have the small remaining number of our BRAVE Men and Women serving in Afghanistan home by Christmas!" Trump tweeted, about 90 minutes before the vice presidential debate was scheduled to begin.

Trump and other officials previously have said that the number of United States troops in Afghanistan would be down to between 4,000 and 5,000 troops around November, and that any subsequent withdrawal would be conditions-based.

But Trump has made it clear that he wants the United States out of Afghanistan, rarely speaking publicly about what "conditions" would be necessary to carry out that withdrawal and instead emphasizing the length of the conflict and complaining that U.S. soldiers are acting as "police" in the war torn country. Former and current administration officials have described him as eager to pull out by November in order to fulfill a key campaign promise from the 2016 election.

It was not immediately clear on Wednesday evening whether he had issued an order to zero out U.S. troops in Afghanistan by the end of the year. Notably, there have been no following statements or comments from U.S. officials at the Pentagon, U.S. Central Command in Tampa, Fla., or in Kabul. Typically, major changes to U.S. force levels or strategy in Afghanistan have been announced with coordinated public messages by administration and military leaders. Earlier in the day, National Security Advisor Robert O'Brien said that there are currently fewer than 5,000 troops in Afghanistan, and that the United States planned to draw down to 2,500 by early 2021.

"Ultimately, the Afghans themselves are going to have to work out an accord, a peace agreement," O'Brien said at the University of Nevada, Las Vegas. "It's going to be slow progress, it's going to be hard progress, but we think it's a necessary step - we think Americans need to come home."

Over the past six months, the United States reduced the number of its troops in Afghanistan first to about 8,600, in accordance with a deal brokered with the Taliban and signed in February. That deal envisioned the immediate drawdown from 14,000 to 8,600 troops by the summer, and a complete U.S. withdrawal within 14 months, if the Taliban lived up to their end of it. Pentagon leaders have long insisted that the second phase would be "conditions-based."

Defense officials have long braced for the possibility that Trump, frustrated with the slow pace of the withdrawal process, may order a sudden and complete exit, as he did for U.S. troops in Syria. Trump has repeatedly sought to withdraw the military from conflict zones and permanent stations overseas, arguing that Americans are tired of "endless" wars. As the weeks wind down the 2020 election, the president has urged or ordered further withdrawals from Iraq, Germany, and elsewhere, in an attempt to fulfill a campaign promise to end the so-called "forever" wars and force allies to take on more of the burden of their own security.

Frustration with the lengthy Afghanistan war - in which 2,400 Americans and thousands more coalition troops and Afghan citizens have perished - has grown into a national mood in recent years. Throughout the Democratic primary campaign, presidential candidates including presumptive nominee Joe Biden called for either an outright withdrawal or a substantial drawdown.

Wednesday marked the 19th anniversary of the U.S. invasion of Afghanistan to topple al Qaeda in 2001.

Like many future Donald Trump voters, Jason Gelinas felt something shift inside him during the presidency of Barack Obama. Things were going OK for him generally. He had a degree from Fordham University and had held a series of jobs at big financial-services firms, eventually becoming a senior vice president at Citigroup in the company's technology department, where he led an AI project and oversaw a team of software developers. He was married with kids and had a comfortable house in a New Jersey suburb. According to those who know him, Gelinas was a pleasant guy who was into normal stuff: Game of Thrones, recreational soccer, and so on. Things did get weird, though, when politics came up.

Gelinas had registered as a Democrat in the runup to the 2008 election, but then seemed to drift to the right, and not in an "I'm going to vote for Romney this time" sort of way, according to two friends, who spoke to Bloomberg Businessweek on the condition of anonymity because they didn't want to be associated with what came next in his political journey. "He hated the idea of Obama," says one. "He thought that it was a setup and that he was elected to satisfy the Black population." Gelinas would become agitated when the topic of the president came up, sometimes referring to Obama as "the Antichrist."

He was increasingly immersed in right-wing internet conspiracies, telling a friend that Hillary Clinton was at the center of a global cabal of sex traffickers. This was about the time that online trolls were starting to promote a theory known as Pizzagate, which claimed that Clinton and others were holding children hostage in the basement of Comet Ping Pong, a restaurant and concert venue in Washington, D.C. Shortly after Trump was elected president, a follower burst into the restaurant and fired an AR-15 rifle, standing down only after discovering that the building didn't actually have a basement. (Nobody was hurt. The shooter, who said he'd been misled by what he'd read on the internet, pleaded guilty to firearms charges and was sentenced to four years in prison.)

Some might have taken that incident as a sign to cool down. Gelinas appears to have gone deeper down the rabbit hole, finding his way to an even stranger movement, QAnon. Like Pizzagate believers, QAnon's are focused on a supposed cabal of pedophiliac liberals, mostly politicians and celebrities. The twist is that QAnon has an apocalyptic component-it holds that, at some point, President Trump will unleash "the Storm," a military coup that will expose and punish this cabal. QAnon has spurred enough violence that the FBI has labeled it a domestic terrorism threat. Supporters have been implicated in the death of a Staten Island mob boss and in the derailment of a train in California.

Even so, the movement had been contained mostly to the internet's trollish fringes until around the time Gelinas came along. In 2018, while doing his job at Citi, he created, as an anonymous side project, a website dedicated to bringing QAnon to a wider audience-soccer moms, white-collar workers, and other "normies," as he boasted. By mid-2020, the site, QMap.pub, was drawing 10 million visitors each month, according to the traffic-tracking firm SimilarWeb, and was credited by researchers with playing a key role in what might be the most unlikely political story in a year full of unlikely political stories: A Citigroup executive helped turn an obscure and incoherent cult into an incoherent cult with mainstream political implications.

In January the House of Representatives will almost certainly welcome its first QAnon supporter, Republican Marjorie Taylor Greene, who's running without serious competition in a district in northwest Georgia, and many other candidates for public office have professed support for aspects of the movement. The Trump campaign has sometimes asked people not to display QAnon signs at rallies, but they show up all the time anyway. QAnon supporters were also ready with an easy spin on the biggest threat to the president's hold on power: his own Covid-19 diagnosis. Trump wasn't sick, the theory goes, he merely retreated from the public eye so that the Storm could begin.

Because it's so much more involved than a typical conspiracy theory, QAnon has often been described as a religious movement-and, like many religions, the core of the belief system stems from revelations in a foundational text. In this case, that text didn't appear on stone tablets handed from a mountaintop or on golden plates buried in the ground in upstate New York, but through a series of cryptic postings on a website best known for racist memes and the manifestoes of mass shooters. Ironically, for a movement obsessed with the evils of pedophilia, the site, 4chan, was also known as a place to download child pornography.

The revelation was delivered on Oct. 28, 2017, and came from a user calling him or herself QAnon. This person, who claimed to be a government employee with top-secret "q-level" clearance (a real thing in the Department of Energy), said Clinton would be arrested in two days and that the event would set off massive organized riots. At the time, 4chan was full of similar nonsense attributed to highly placed government officials. But QAnon-or simply Q-caught on in a way that competing accounts such as FBIAnon and CIAAnon didn't. The user became the narrator of a tale that cast Trump as the central hero in an epic global struggle, doling out the story in thousands of posts known as "Q drops," first on 4chan, then on the even more outré 8chan and its successor site, 8kun.

The identity of Q has been a subject of speculation since the beginning. The theories are all over the place, variously suggesting that Q is Edward Snowden, or former national security adviser Michael Flynn, or the conspiracy-minded radio host Alex Jones, or even Trump himself. One self-published book, which Amazon.com Inc. includes for free as part of its Kindle Unlimited subscription, claims to have used a mathematical model to determine that Q is former National Security Agency official Thomas Drake. Drake has denied this-but Q would do that, wouldn't he?

If Q's drops are the new movement's divine revelations, its rites involve the production and consumption of videos and social media posts-often screenshots annotated with arrows and circles revealing hidden connections-designed to interpret them. "Digging deeper," Q's followers often call it. Just a few minutes before 1 p.m. on Father's Day 2018, for instance, Q and Trump each posted a Happy Father's Day message. Coincidence? Or how about this August, when Trump visited a Whirlpool Corp. plant in Ohio and posed in front of 17 washing machines? Q is the 17th letter of the alphabet. Surely this was the president signaling that Q was going to clean things up. Or maybe it had something to do with money laundering?

At first, the primary documents for Q were available only to the bravest of web surfers. Most regular people don't spend much time on 8kun, which is awful in terms of content and interface design. The need to spread the word beyond core users led to the creation of aggregator sites, which would scrape the Q drops and repost them in friendlier environs after determining authenticity. (The ability to post as Q has repeatedly been compromised, and some posts have had to be culled from the canon.) This task, Gelinas once told a friend, could be his calling from God.

On April 5, 2018, Q posted a  short message-drop No. 1,030-insinuating that a recent spate of military aircraft crashes was part of a "silent war." Later that night, Gelinas registered QMap.pub. His intention, as he later explained on Patreon, the crowdfunding website widely used by musicians, podcasters, and other artists, was to make memes, which are harder to police than tweets or Facebook text posts. "Memes are awesome," Gelinas wrote. "They also bypass big tech censorship." (Social media companies are, at least in theory, opposed to disinformation, and QAnon posts sometimes get removed. On Oct. 6, Facebook banned QAnon-affiliated groups and pages from the service.)

Gelinas raised thousands of dollars on Patreon each month, posting updates using his pseudonym, QAppAnon. "Like many of you, I felt that something wasn't right in the world, that our country was headed in the wrong direction," he wrote. "Then something magical happened in 2016 that defied expectations-a complete outsider to the political establishment, Donald J Trump, won the presidential election! Amazing. A glimmer of light in the darkness." A few months into the Trump administration, Gelinas changed his party affiliation to Republican, and this spring he contributed $200 to Trump's reelection efforts-his first-ever political contribution, according to federal disclosures.

QMap developed into a central place for fans to read the drops, to plot, and to commiserate on the site's "Where We Go One We Go All Prayer Wall." The site wasn't just a repository of QAnon posts; Gelinas served as an active co-author in the movement's growing mythology. The clean, minimalist site was designed around tiles dedicated to each Q drop, which Gelinas titled to make them easier to understand. Tabs across the top enabled users to sort by theme or tags, and the hidden players and themes were explicated along the left side with a series of icons-a few chess pieces, a globe, a skull. Brief descriptions sorted "players" by category. (French President Emmanuel Macron and New York Times reporter Maggie Haberman are in the "Traitor/Pawn" category; Senator Ted Cruz is a "Patriot.")

QMap also had a tab for suspicious deaths. John McCain didn't die from brain cancer, according to QMap. "One theory is that he was secretly tried [by] military tribunal and sentenced to death," the site said. Q had never made these claims explicitly; they were insinuated by his posts, then interpreted by QMap. "It was all laid out in a way where someone could easily start to believe it's all true," says Joe Ondrak, a researcher for Logically.ai, a fact-checking website that follows the movement. "It was like a redpill factory." ("Redpill" is a reference to the movie The Matrix, in which characters who want to see the world as it actually is take a tablet of that color. It's been adopted by right-wing activists to connote the conversion of new believers.)

One young QAnon supporter encouraged QMap to annotate posts with supporting evidence and links to additional reading materials, providing "background info for the uninformed so that even his grandma could understand what's going on," Gelinas wrote approvingly on Patreon in the summer of 2018. "What a great idea. It's hard to jump into Q if you haven't been following it closely."

On Patreon, he laid out a plan to add a team, which he hoped would be staffed by disaffected software developers. "Facebook devs: how mad are you. You've been lied to," Gelinas wrote on Twitter in March 2019. "Your talents have been used/abused for evil purposes. Let's build a new platform for the GOOD of Humanity."

By this point, Gelinas claimed he was the No. 2 figure in the movement, behind only Q, according to a friend, and began to dream about turning his QAnon hobby into his main gig. "Who knows, maybe QMAP becomes the media platform of the future one day? :-)" he mused in early September.

By now, QMap's growth had attracted an enemy. Frederick Brennan, a 26-year-old polymath with a rare bone disease, had decided to unmask the person behind QMap. Brennan was a reformed troll. He'd created 8chan, but he had a change of heart after the man responsible for the 2019 mass shootings at two mosques in Christchurch, New Zealand, posted his manifesto on the forum in advance and inscribed 8chan memes on the weapons he used to kill 56 people.

Brennan had come to believe that Jim Watkins, an American entrepreneur who'd taken over 8chan and its successor site, 8kun, was somehow involved in QAnon. The mixture of regret over what the sites he'd started had become and the grudge against Watkins, who runs 8kun from his pig farm in the Philippines, had sent Brennan on a mission to bring down the site and QAnon. Watkins did not respond to a request for comment.

Brennan started by trying to figure out which companies were operating servers that hosted 8chan's content. Then he would post public messages, on Twitter and elsewhere, urging the companies to cut ties with the site. After 8chan was dropped by the cybersecurity company Cloudflare Inc., which protected it from denial of service, or DDoS, attacks, it found safe harbor in a new U.S.-based DDoS protection company, VanwaTech LLC, which had taken an extremely permissive attitude toward controversial content. "If it's legal, I don't care," says 23-year-old chief executive officer Nick Lim.

This summer, Gelinas also moved his site to VanwaTech. This made him a target of Brennan, who also began pressuring Patreon to block Gelinas's site. He referred to QMap in a tweet as "the main vector for Q radicalization." QMap, Brennan explains in an interview, helped "turn this anonymous format into a way people can be notified immediately."

Patreon never banned QMap, and Gelinas took down all his posts on the crowdfunding site after he was identified as QMap's owner. In messages exchanged over WhatsApp, he told Bloomberg Businessweek that he has no connection to Watkins and has never met him. He said he began using VanwaTech because it protected QMap from frequent DDoS attacks.

Ondrak, the fact-checker, and Nick Backovic, another Logically.ai researcher, joined Brennan's hunt. It took Ondrak and Backovic only a few days to trace an email address associated with Patriot Platforms LLC, which had been listed as the publisher of a QMap mobile app in Google's Android app store, to a post office box in Berkeley Heights, N.J. The next day, the pair published a story outing Gelinas as the operator of QMap. Public records show that Gelinas is the sole employee associated with Patriot Platforms, and New Jersey business records obtained by Bloomberg Businessweek list the company's address as a house in the same town, a few miles from the P.O. box.

On the morning of Sept. 10, a reporter drove to the house. It was a beautiful day in suburban New Jersey. Gelinas, in shorts and an American flag cap, was in the front yard, filling up a wheelbarrow with cut-up tree stumps.

Gelinas is tall and fit at age 43. He clearly didn't want to talk. He paced around his yard, mostly evading questions, while the reporter stood in the grass. He first said he wasn't Q, though he did allow that he was familiar with QAnon, which he described as "a patriotic movement to save the country." Finally, his wife opened the front door and rescued him with a vague request for technical assistance. "I don't want to get involved, I want to stay out of it," Gelinas said before he disappeared into the house and, rather than asking the reporter to leave, called the authorities. A few minutes later, after the reporter had left the property, two police SUVs showed up.

That afternoon, QMap.pub and the social media profiles of Gelinas and his wife disappeared from the internet. Within days, Citi had put him on administrative leave and his name was removed from the company's internal directory. He was later terminated. "Mr. Gelinas is no longer employed by Citi," the company says in a statement. "Our code of conduct includes specific policies that employees are required to adhere to, and when breaches are identified, the firm takes action."

In the weeks after he was outed, Gelinas mostly ignored reporters' calls and text messages, though he did acknowledge he was the only developer for QMap and clarified several other points. "I'm not going to talk about my own story right now," he said. "When the time is right, it will come out."

QMap's disappearance has been a significant but temporary setback for the QAnon movement. "It's not going to be a death blow to the QAnon community, but it is a disruption," says Travis View, a conspiracy theory researcher who hosts a podcast dedicated to QAnon. QMap popped back online a few days later, but it now consists entirely of links to other QAnon aggregator websites.

Google has tried to make it harder to find such QAnon sites by keeping them from showing up in searches, and Facebook and Twitter have blocked links to them, though posts about Q are easy to find on Facebook and other social networks such as Telegram. Followers also sometimes spread the word about Q-related sites by writing their URLs on signs and holding them up at Trump rallies.

Meanwhile, Gelinas's project of bringing the gospel of Q to the mainstream is alive and well. Late this summer and early this fall, Q supporters organized a wave of in-person rallies, ostensibly to combat human trafficking, many of them under the social media hashtag #SaveTheChildren. Some established anti-trafficking groups, including the real Save the Children, a 101-year-old British nonprofit, complained they were being co-opted in dangerous ways.

Janja Lalich, a professor emerita of sociology at California State University at Chico who's studied cults for decades, says internet movements such as QAnon have grown at an alarming rate, because of a political debate that's become increasingly unmoored from a set of universally agreed-upon facts. "It's times like these that cults can thrive," she says. "We have leadership that has tried very hard to change our relationship with reality, and people are grasping at straws. The last four years have been precedent-setting in creating an atmosphere of disbelief."

Returning from that collective delusion, Lalich insists, won't be easy. "It's very daunting," she says. "You have to give up everything you believed in and decide what to believe again." -With Jennifer Surane

Oct. 7, 2020 6:57 pm ET

Since the 1950s, the United States has tried to oust governments in the broader Middle East once every decade, on average. It has done so in Iran, Afghanistan (twice), Iraq, Egypt, Libya, and Syria-a list that includes only the instances in which the removal of a country's leaders and the transformation of its political system were the goals of U.S. policy and Washington made sustained efforts to achieve them. The motives behind those interventions varied widely, as have Washington's methods: in some cases sponsoring a coup, in others invading and occupying a country, and in others relying on diplomacy, rhetoric, and sanctions.

All these attempts, however, have one thing in common: they failed. In every case, American policymakers overstated the threat faced by the United States, underestimated the challenges of ousting a regime, and embraced the optimistic assurances of exiles or local actors with little power. In every case but that of Syria (where the regime held on to power), the United States prematurely declared victory, failed to anticipate the chaos that would inevitably ensue after regime collapse, and ultimately found itself bearing massive human and financial costs for decades to come.

Why is regime change in the Middle East so hard? And why do U.S. leaders and pundits keep thinking they can get it right? There are no easy answers to those questions, and it is important to acknowledge that in every case, the alternatives to regime change were unappealing. But as U.S. policymakers contemplate the challenges of dealing with this vexing region, they should see the patterns of self-delusion and misjudgment that have time and again made regime change so tempting-and, ultimately, so disastrous.

In 2011, as senior officials debated whether the United States should use military force against the Libyan ruler Muammar al-Qaddafi, U.S. Secretary of Defense Robert Gates-the most experienced member of President Barack Obama's national security team-reminded his colleagues that "when you start a war you never know how it will go." Gates's warning was an understatement: in every single case, however carefully prepared, regime change in the Middle East has had unanticipated and unwelcome consequences. Perhaps the most powerful example of this phenomenon was the 2003 U.S. invasion of Iraq, when Washington ended Saddam Hussein's rule but also inadvertently empowered Iran, fueled jihadism, demonstrated to dictators around the world the potential value of possessing nuclear weapons (to deter such invasions), increased doubts all over the world about the benevolence of U.S. power, and soured the American public on military intervention for decades to come.

Iraq was hardly an outlier: in every other case, the most significant consequences were the unintended ones. In Iran in 1953, the CIA helped oust the prickly nationalist prime minister Mohammad Mosaddeq, hoping that with Mosaddeq out of the picture, the Iranian shah, Mohammad Reza Pahlavi, would be a more reliable regional ally and keep Iran out of the Soviet camp. But the shah's baroque corruption and harsh repression-abetted by his U.S. benefactors-ultimately led to the 1979 revolution, which brought to power an intensely anti- American Islamist regime that has sponsored terrorism and destabilized the region ever since. In Afghanistan in the 1980s, U.S. support for the Islamist mujahideen helped to undermine the Soviet Union but also contributed to a decade of chaos, a civil war, the rise of the brutal Taliban government, an empowered global jihadi movement-and, ultimately, another U.S. military intervention, after the 9/11 attacks in 2001, which were planned by al Qaeda terrorists based in Afghanistan. After a popular uprising in Egypt in 2011, the United States used its diplomatic leverage to help end the decades-long repressive rule of Hosni Mubarak. The situation deteriorated in the years that followed, however. In 2012, elections brought to power an exclusionary Islamist government. The next year, that government was violently overthrown and replaced by a new military regime led by General Abdel Fattah el-Sisi, which has proved to be even more repressive that Mubarak's.

In 2011, the U.S.-backed ouster of Qaddafi and the subsequent collapse of the Libyan state led to widespread violence, allowed weapons to proliferate across the region, exacerbated instability in neighboring Chad and Mali, and stiffened Russia's resolve to never again allow the UN Security Council to pass a resolution that would facilitate regime change, as it did in the case of Libya. Advocates for regime change in Libya had hoped that Qaddafi's overthrow would lead other dictators to agree to leave power or suffer Qaddafi's fate. In fact, the intervention had the opposite effect. In Syria, for example, President Bashar al-Assad watched Qaddafi brutally tortured and killed by Libyan rebels and decided to crack down even more ruthlessly on his opponents, creating an opening for jihadis, who then spilled over into neighboring Iraq and undermined the government there.

The attempt by the United States and others to remove Assad by supporting opposition rebels proved even more catastrophic. With Russia and Iran determined to keep Assad in power, years of outside military assistance to the Syrian opposition led not to Assad's ouster as intended but instead to counterescalation by his regime and its foreign sponsors, fueling a vicious civil war, a humanitarian tragedy, refugee flows on a scale not seen since World War II (which themselves caused a populist backlash in Europe), and an explosion of jihadi extremism. The desire to overthrow the murderous Assad was understandable. But the consequences of trying and failing to do so-in part because no one had the appetite to invade and occupy Syria less than a decade after the Iraq disaster-proved to be worse than not trying at all.

The heart of the problem is that whenever an existing regime is destroyed (or even just significantly weakened by outside forces, as in Syria), a political and security vacuum emerges and a power struggle begins. In the absence of security, people feel no alternative but to organize and arm themselves and to turn to kinship networks, tribes, and sects for safety, exacerbating sectarianism and internal rivalries and sometimes leading to demands for secession. In the run-up to an intervention, groups with little in common form coalitions of convenience. But once the regime falls, they quickly turn against one another. All too often, the most extreme or violent groups prevail and more moderate or pragmatic forces are sidelined; inevitably, those excluded from power work to undermine those who seized it. When the United States has tried to fill the vacuum itself, as it did in Iraq and at times in Afghanistan, it has found itself the target of locals and neighboring states that resist foreign interference and has ended up sacrificing thousands of lives and spending trillions of dollars yet still failing to create stability.

The security vacuum created by regime change not only sets up a struggle for power within states but invariably generates ruthless competition among regional rivals as well. When governments fall (or appear likely to do so), regional and even global powers rush in with money, arms, and sometimes direct military force to put their own proxies in power and pull the country into their orbit. U.S. Secretary of State Condoleezza Rice's repeated assertion around the time of the Iraq war that Washington's pursuit of "stability at the expense of democracy" in the Middle East had produced neither was broadly true. But it turned out to have a corollary-that pursuing democracy at the expense of stability might also produce neither, but at even higher cost.

Americans like to believe their foreign interventions are generous, benign, and widely appreciated, but it turns out that even when they help topple unpopular regimes they are not necessarily greeted as liberators. Indeed, even well-meaning interventions in the Middle East have often led to violent resistance. After the 1953 coup in Iran, antipathy toward the United States for empowering the dictatorial shah led to virulent anti-Americanism that endures to this day. In Afghanistan, where suspicion of outsiders runs deep, Hamid Karzai, the leader whom Washington favored after its 2001 invasion, could never escape the impression among Afghans that he was put in power and supported by foreigners. Today, ridding the country of occupying U.S. troops remains the opposition Taliban's most central rallying cry. Most famously, U.S. Vice President Dick Cheney's prediction that U.S. troops would be "greeted as liberators" in Iraq proved wildly wrong and was followed by years of bloody anti-American insurgency.

Even the allegedly friendly leaders the United States has put in place have not always acted according to Washington's wishes. After all, they have their own local interests to worry about and often have to stand up to outside powers to bolster their legitimacy. Frequently, they have defied Washington on a range of domestic and international issues, knowing that their U.S. sponsors had little choice but to continue to support them. And far from exercising positive influence on such leaders and helping the United States overcome these challenges, many regional and global players do just the opposite. For decades, Pakistan has helped thwart U.S. efforts to stabilize Afghanistan. Iran undermined U.S. efforts in Iraq by supporting violent Shiite militia groups. Libya has been torn apart by competing outside powers supporting rival proxies. And in Syria, Russia and Iran-determined to undermine U.S.-sponsored regime change in part lest Americans get the idea of trying it one day in Moscow or Tehran-responded to every U.S. escalation with a counterescalation of their own. These regional spoilers often succeed because they have more local influence and more at stake than the United States does, and it's far easier to cause chaos than to prevent it.

The more recent U.S. interventions in the Middle East have sought to replace autocratic regimes with democratic governments. But even if those actions had somehow avoided the pitfalls posed by security vacuums, popular resistance, and untrustworthy proxies, they would have been unlikely to shepherd in new democracies. Although there are no clear recipes for democratic development, extensive scholarly research suggests that the main ingredients include a high degree of economic development; significant ethnic, political, and cultural homogeneity (or at least a shared national narrative); and the previous existence of democratic norms, practices, and institutions. Unfortunately, the states of the contemporary Middle East lack all these attributes. None of this means that democracy is impossible there or that promoting democracy should not be an American aspiration. But it does suggest that pursuing regime change in the Middle East with the hope that doing so will lead to democratic development is wishful thinking in the extreme.

The deep-seated American desire to fix problems in the Middle East is in many ways honorable, but it can be dangerous as well. The hard reality-demonstrated by decades of painful experience in the region-is that there are some problems that cannot be entirely solved and trying to solve them sometimes makes things worse.

Part of the problem is that U.S. policymakers often lack a deep understanding of the countries in question, making them susceptible to manipulation from parties with their own vested interests. The most famous example is the Iraqi exile Ahmed Chalabi, who helped convince top officials in the George W. Bush administration that Iraq had weapons of mass destruction and that U.S. forces would be greeted as liberators in Iraq. Years after the invasion, Iraqi authorities arrested Chalabi on charges of counterfeiting and allegedly working to advance the interests of Iran. Similar scenarios played out in Libya, Syria, and elsewhere, where even well-meaning exiles told Americans and others what they wanted to hear in order to win the support of the most powerful countries of the world. In each case, it led to massive miscalculations about what would happen in the wake of the U.S. intervention, almost always in the direction of excessive optimism.

Americans also keep placing hope over experience when it comes to Middle East policy because of a persistent tendency to underestimate the degree of resources and commitment it will take to get rid of a hostile regime and stabilize the situation once it is removed. But many decades of experience demonstrate that autocratic regimes never relinquish power in the face of economic sanctions alone (which hurt the public more than the leadership) or even in the face of modest amounts of military force. Numerous Middle Eastern rulers have been willing to risk and even lose their lives rather than give up their power voluntarily. The result is that when the United States wants to get rid of such leaders, it must go far beyond the low-cost remedies often proposed by proponents of regime change, such as implementing no-fly zones, launching airstrikes, and providing arms to the opposition. Instead, significant U.S. military deployments are required to dislodge such leaders, and even after they are gone, it always proves far more costly to deal with the aftermath than proponents of regime change suggest. And although officials in Washington often assume that regional or international partners will help bear the burdens and assume the costs of regime change, that rarely happens in reality.

Some of these problems would be manageable if the American public's commitment, patience, and staying power were infinite, but they are not. Especially because U.S. leaders and regime change proponents rarely acknowledge the likely heavy costs as they make the case for action, once the immediate crisis passes and public perceptions of the threats at hand diminish, public support dwindles. Most Americans initially supported the invasions of both Afghanistan and Iraq. Over time, however, majorities concluded that both interventions had been mistakes. And hardly any public support ever existed for intervening or peacekeeping operations in Libya and Syria. In every case, as the problems mounted and the costs rose, the public backing necessary for success disappeared.

In the future, there may be cases in which mass terrorism, genocide, a direct attack on the United States, or a country using or proliferating nuclear weapons makes the benefits of removing a threatening regime exceed the costs. But if history is any guide, such cases will be rare to nonexistent. And even where they exist, they demand caution, humility, and honesty about the likely costs and consequences.

Regime change will always tempt Washington. So long as there are states that threaten American interests and mistreat their people, U.S. leaders and pundits will periodically be pulled toward the idea that Americans can use their unparalleled military, diplomatic, and economic power to get rid of bad regimes and replace them with better ones. The long, diverse, and tragic history of U.S.-backed regime change in the Middle East, however, suggests that such temptations-like most quick fixes that come along in life and politics-should be resisted. The next time U.S. leaders propose intervening in the region to overthrow a hostile regime, it can safely be assumed that such an enterprise will be less successful, more costly, and more replete with unintended consequences than proponents realize or admit. So far, at least, it has never been the other way around.

China is now the official "pacing threat" for the U.S. armed forces. Simply put, the Pentagon considers the People's Liberation Army its most serious competition. This is a major and vital shift. But competing with China is a tremendously broad concept that could take any number of forms, and the reality is that China is too powerful to permit the U.S. defense establishment to fritter away money. The Defense Department needs to focus. Most of all, it needs to be capable of achieving U.S. objectives against China in a war.

But in which war? Scenarios lie at the core of military force planning. These are the plausible and highly consequential future fights the military uses to plan its future force structure, posture, and training. During the interwar period, the Navy planned for scenarios such as the defense and recapture of the Philippines. During the Cold War, the Pentagon planned for the defense of NATO against a Warsaw Pact invasion. Scenarios help to concentrate the sprawling defense establishment on specific strategic and operational problems. They provide a concrete framework to examine how different military force structures and operational approaches would perform in meeting objectives within a set of constraints, and how long such efforts would take and at what cost.

Thus, planning for one set of scenarios over another can have drastic implications for what the U.S. military will look like and how it will operate. A force shaped for operations in the Middle East, to give an example, would be far different from one optimized for the Western Pacific.

While the Pentagon does not disclose its formal scenarios, outsiders can glean sophisticated and credible defense analysis from, for instance, the RAND Corporation. In the case of China, these include plausible conflicts over Taiwan, the South China Sea, the Philippines, Japan, South Korea, or Vietnam.

Of these, a conflict over Taiwan is the key scenario. The defense establishment should focus on preparing the military for a Taiwan scenario above all others. The United States needs to be able to effectively defend Taiwan because it is important to frustrating China's strategy to achieve hegemony in Asia. Adapting the U.S. military to be able to defend Taiwan will be hard, but it is necessary, and will also allow the United States to defend other allies in Asia against China.

The Trump administration has made this point increasingly clear, but it is a bipartisan one. Taiwan is militarily significant, located as it is in the center of the vital first island chain, and is critical to American credibility in Asia. Other states regard it as the canary in the coal mine - a strong indicator of how far the United States would go to defend them against China. If China were able to suborn Taiwan, the U.S. and allied defense position would be substantially compromised, and U.S. credibility seriously diminished. For these reasons, subjugating Taiwan is very likely China's best next step toward its strategic goal of regional hegemony.

China presumably would prefer to induce Taiwan to unify peacefully, but the reality is that that is unlikely, given deepening opposition to unification with China on Taiwan. Accordingly, China is likely to need to turn to force to "resolve" the issue. Moreover, it has the resolve and increasingly the power to try to do so: Beijing views Taiwan as a renegade province and, under Xi Jinping, has strengthened its commitment to unify the island with the mainland. Moreover, China has built a military specifically to force unification. Beijing's most attractive military strategy to cleanly and resolutely settle the issue would likely be an invasion, and blunting such an assault should be the United States' top priority for defending Taiwan. At the same time, the Pentagon also needs to be prepared to relieve Taiwan if China chooses a more indirect strategy, such as a blockade and/or bombardment of the island to try to coerce it into surrender.

Taiwan is the toughest for the United States to defend among its allies and close partners. But focusing on the hardest strategically significant scenario is crucial, not least because China is likely to go for the weak point in America's defense perimeter. Defending Taiwan calls for the United States to deny China freedom of action less than 100 miles off the Chinese coast, where China's advantages in numbers and distance are most pronounced. Importantly, defending Taiwan does not require that the U.S. military dominate within the first island chain, but rather deny that dominance to China. This is a lower but achievable standard.

If the United States cannot field military forces capable of defending Taiwan, it is unlikely to be able to credibly deter China from taking it. Focusing on easier scenarios farther from China's shores will divert attention, resources, and preparation, making Taiwan more vulnerable. This is especially important because, if China is able to suborn Taiwan, it will almost certainly lift its gaze farther afield to countries like the Philippines from an even stronger position. Those previously easy scenarios like the defense of the Philippines will be much harder, since China will then be able to project power from Taiwan rather than having to worry about it.

Focusing on a conflict scenario with China over Taiwan is particularly important because of the potent tendencies in the defense establishment to avoid concentrating on the toughest, most strategically significant contingencies. Bureaucratic imperatives often prize the preservation of existing programs and practices, while a serious examination of the most challenging future warfare scenarios often points to the need for disruptive change and the goring of sacred cows. Especially if defense budgets remain flat, as seems likely at present, such change may require eliminating previously cherished parts of the force. Think of the resistance to give up horse cavalry in the U.S. Army before World War II or the battleship admirals' hostility to the advent of the aircraft carrier. Organizations with entrenched interests in given force structure will happily use less demanding scenarios to justify the existing force. Such scenarios are not just thought experiments, in other words, but can result in dangerous inertia.

Arguments that Taiwan is "too hard defend" almost invariably assume traditional forms of American power projection rooted in the Cold War-era force structure and operational concepts. As wargaming from RAND and others has shown, the defense of Taiwan that exploits different weapons platforms and systems, emerging technologies, and new ways of operating can be successful. Defending Taiwan would be very challenging, but it is a solvable problem if undertaken with the requisite focus and willingness to change. Such willingness would stem from a basic recognition of the American military's fundamental purpose: The armed forces should be adapted to the political requirements of the nation, not the political requirements of the nation to legacy force structure and operating patterns.

Fortunately, while such adaptation takes political courage and grit, it can happen. The U.S. Navy embraced submarines and carriers ahead of World War II. Even better, current Marine Corps Commandant Gen. David Berger's Force Design 2030 and Air Force Chief of Staff Gen. C.Q. Brown's Accelerate Change or Lose show that the willingness to implement needed change is alive and well.

Finally, developing a military able to defend Taiwan will also be useful for other, less stressing scenarios in the Western Pacific. Taiwan is the frontline. If the United States develops a force able to defend it, then that force is likely to be able to defend other allies and partners like Japan, Australia, and the Philippines against China. Such a force would able to destroy in a timely fashion a large concentration of China's maritime forces in the Western Pacific. It would exploit capabilities that can reach Chinese forces within their defense perimeter, such as penetrating strike, long-range fires, and undersea capabilities, along with essential enablers that could withstand Chinese attacks, including agile logistics and a resilient communication and intelligence network. It would also be one at the forefront of applying new sources of military power - data, algorithms, and computing power.

The Defense Department is rightly focused on China in Asia and on restoring the American military's edge vis-à-vis Beijing. The best way the Pentagon can serve these goals is to prioritize defending Taiwan over all other contingencies in its planning. Doing so will be challenging and likely involve significant change, but it can - and should - be done to deter and, if necessary, prevail in a war with the most challenging rival the United States has faced in a generation.

Elbridge Colby is a principal at The Marathon Initiative. Jim Mitre is the chief strategy officer at Govini. They previously served as lead official for and executive director for the development of the Department of Defense's 2018 National Defense Strategy.