The Afghan military sustained twice as many losses in the last year as U.S. forces killed in the entire 13 years of war in Afghanistan. And the pace of casualties is escalating, suggesting that the Taliban is stronger than the Pentagon and mainstream media have ever let on—in fact, the Afghan security forces are a house of cards experts say is destined to fall.

Just this week, the Associated Press reported that upwards of 40 percent of Afghan security forces are “ghosts”—soldiers and police who exist on the books but are otherwise nowhere to be found. With current maps showing the Taliban holding more territory than at any time since 2001, and ISIS moving in to make a play for their turf, confidence that a “national” army can defend Afghanistan on its own is at an all-time low.

“It is not succeeding, that’s the point,” says Anthony Cordesman, senior security analyst with the Center for Strategic and International Studies. Numbers and facts on the ground, he tells TAC, are hard to come by. “There is no transparency—you can’t trust anything. There is no meaningful readiness data anymore.”

But a careful look at how these forces were trained and how it was reported in the media suggests the true picture was skewed the whole time. In fact, the effort might have been doomed from the start.

“There was gross over-assessments of success coming out of CENTCOM [Central Command]—no one bothers to go back and check,” said Cordesman. Congress, too, fell down on the job. “When they could have imposed meaningful transparency and systematic accountability, the congress never did.” The military padded its reports, blew smoke at Congress and enabled a White House in denial, he said (an investigation into how much is ongoing).

Today, the Pentagon assessments are a bit more staid. In December, the military reported to Congress that

Although the ANDSF [Afghan Security Forces] have capability advantages over the insurgent forces, they remain reluctant to pursue the Taliban into their traditional safe havens. Given the ANDSF’s current stage of development, they cannot manage the insurgency and ensure security and stability across Afghanistan without further improvement…

Larry Korb, a former assistant secretary of defense under President Reagan, said the situation is eerily like Vietnam. “When we left there, the millions [of South Vietnamese army soldiers] we trained looked great on paper. [But] really, they crumbled. As we know now from the archives, the North Vietnamese were surprised at how easy it was.”

When President Obama announced his intentions of keeping 9,500 U.S. troops in Afghanistan through 2017, no one argued with his assertion that the more than 325,000 Afghan forces there weren’t fit to defend the country on their own.

The bottom line is,” Obama said in October, “in key areas of the country, the security situation is still very fragile, and in some places there is risk of deterioration.”

You bet it’s fragile. When the Taliban briefly took over Kunduz in October, many of the Afghan forces reportedly ran in the face of the Taliban invasion.

Knowing nods all around. The same thing happened in Iraq over the last year and a half, when the Islamic State took over town after town and the American-trained Iraqi army evaporated like mist.

Thanks to the Pentagon shell game, it’s difficult to zero in on the numbers, but the last official count for the Afghan army released by the Pentagon in July was 176,420 and that was down from 2014, much from desertion. (Reuters just reported that a third of the Army had to be replenished in 2015 due to casualties and soldiers walking away). The police numbered 148,296. But if recent reports about “ghost soldiers” are correct, these metrics are a mirage.

Sadly, the U.S. spent $25 billion building and training the Iraqi military, and more than double that—$65 billion—doing the same thing in Afghanistan. And the money keeps pouring into the sieve. According to Stars & Stripes, the U.S. and coalition partners spent $4.1 billion on Afghan forces in 2015 alone.

“Prudence might actually counsel that Washington assume instead, when it comes to organizing, training, equipping, and motivating foreign armies, that the United States is essentially clueless,” wrote Boston University professor Andrew Bacevich in October.

Yet instead of considering whether they were chasing rainbows with nothing but an empty pot at the end, the defense establishment in Washington—whether it be the military or the web of think tanks and contractors that supported it—took great pains to convince the purse holders and American public that more resources would do the trick. As we know now, it never did.

One telling moment came in 2014, when the Pentagon announced it was destroying or breaking down into scrap much of the equipment and vehicles in Afghanistan before U.S. forces pulled out. “[The Afghans] don’t have the requisite skills to maintain these things,” noted Lt. Col. Daniel Davis, who served in Afghanistan as an Army acquisition chief from 2011-12, at the time.

Korb concurs. “If you give that stuff to the Afghan or even the Afghan Security Forces, it could still end up with the Taliban. You have to err on the side of caution.”

How Did This Happen?

A review of a decade of public reports about the training exposes how fragile these efforts were from the beginning.

As Cordesman noted, the military establishment created a Potemkin Village, playing the press and Congress like violins during hearings and visits from congressional delegations. “The Afghan National Army is making tremendous progress and is a factor on the battlefield,” boasted a Pentagon press release in 2007. “Progress” is always relative of course, and while the soldiers might be a “factor,” reports dating back a decade are typically shrewd about how they define exactly what that means.

Nowhere is this better illustrated than in the late Michael Hastings’ book, The Operators. In that account, then-Gen. Stanley McChrystal shared his skepticism about the war with Hastings sotto voce, while publicly—and for Washington—he promoted it. When Hastings writes about this apparent contradiction, the mainstream media pounces on him for not playing along. He devotes an entire chapter of the book to the “Media-Military Industrial Complex.”

“The unwritten rule I’d broken was a simple one,” he said. “You really weren’t supposed to write honestly about people in power,” or by extension, the war.

Meanwhile, we may never know how much and where all the money for training was spent. That’s because, as the Special Inspector General for Afghanistan Reconstruction (SIGAR) has pointed out, much has been classified, sometimes retroactively, due to “national security” concerns.

So all we have to go on is SIGAR’s work, and bits and pieces of press releases, contract notices, and think-tank analyses. For example, a 2011 RAND white paper entitled  “The Long March” shows billions appropriated for building infrastructure, as well as equipping, arming, and sustaining Afghan troops. Here too is careful massaging of the language to suggest that success was just around the corner, but “in spite of the progress made in the development of the ANA, its operational effectiveness remains very much in the balance.”

There is disappointment, too, over coalition countries dragging their feet on money and trainers. “The progress of Afghan forces is such that U.S. military officials are asking for a much larger commitment”—meaning money—“from the U.S. government to accelerate the pace of training for the Afghan National Army and to improve the Afghan army’s equipment,” the Pentagon said in a 2007 release, which featured quotes from Lt. Gen. Karl Eikenberry, then chief of the combined forces in Afghanistan, later the U.S. ambassador there.

However, he added, “I will firmly tell you in 2007 that the Afghans want this army more than we do.”

Since then billions more were spent, much of it through contractors including infamous Blackwater (now Academi), with some hope—akin to a belief in fairies and unicorns—that the Afghan government would be able to sustain it all when the West finally pulls out.

The U.S. Institute for Peace said in 2013 that “the residual cost of sustaining the Afghan National Security Forces (ANSF) assistance program after 2014 is estimated at between $2 billion and $6 billion, more than the Afghan government’s annual budget.” That’s an annual cost, according to the Washington Post.

A CNAS report in January 2015 guessed the Americans would be footing most of that bill. Like most Washington assessments, however, it started off sunny:

In September 2014, International Security Assistance Force (ISAF) trainers assessed 33 of 40 ANSF units to be either “capable” or “fully capable.” … Moreover, the ANSF are now “in the lead” for security operations across the country and are scheduled to be fully responsible for this mission by the end of 2014.

The rest of the report, however indirectly, describes why it won’t work. Phrases like “in the lead,” it seems, are fungible too.

More recently, CSIS pulled together a number of 2015 assessments that challenge whatever optimistic predictions CNAS had in the first place.

As for the police, U.S. contractor Dyncorp has received billions in reconstruction funds, including contracts to train security forces, since the war began. Even after the State Department was blamed for failing and the reins were turned over to the Pentagon in 2010, Dyncorp escaped scrutiny. It kept getting lucrative multi-year deals, despite falling short. In January 2015, Dyncorp won another contract for $100 million to train and mentor Afghan police and military.

Despite what the U.S. Institute for Peace calls “remarkable progress,” the police as an institution remain prone to “corruption, incompetence, abuse of power, and pervasive illiteracy.” As of May 2014, despite a total of $15 billion spent, thousands of cops on the books remained untrained. And the $300 million a year the U.S. forks over for their salaries? Much of it is unaccounted for, according to SIGAR.

Our track record at building security forces over the past 15 years is miserable,” Eikenberry told the New York Times recently, changing his earlier tune. Today, he writes about the failures of counterinsurgency in Afghanistan. As though anyone was listening.

The Afghan Local Police system, on the other hand, is a corrupt, destabilizing, dangerous hot mess, thanks in part to the U.S. military, which put ex-warlords and militiamen in control of many constabularies across the country. As always, little consideration was given to local tribal interests and communities, many of whom have since turned to the Taliban as an alternative. But this seems to be the case with the training of all security forces, with little accounting for Afghanistan’s ethnic, tribal, and sectarian culture. As a result, the Afghan resistance to the Western template has resulted in fraud, failure, and widespread mistrust between the international and Afghan forces.

“A lot of this was about imposing outside systems that in many ways do not conform to local [culture], taking existing structures and breaking them,” said Cordesman.

“The real key is that you have to be able and willing to fight and die for your country,” added Korb. But is there an Afghan “nation” that local security forces believe in enough to die for? If not, should the U.S. continue to send thousands of soldiers and contractors into harm’s way for nothing?

These are very expensive questions indeed. Most agree Washington doesn’t have 13 more years to find out.

Kelley Beaucar Vlahos is a Washington, D.C.-based freelance reporter and TAC contributing editor. Follow her on Twitter.