Who Does No. 2 Work For?!

Forget the shit show in Iowa. The best story of the past week is about how Soviet leader Joseph Stalin set up a secret laboratory to analyze Mao Zedong’s poop. If there are two things that people love, it’s stories about espionage and stories about dropping a deuce.

Former Soviet intelligence officer Igor Atamanenko told Komsomolskaya Pravda that during Stalin’s reign, Soviet analysts substituted for their lack of spy gadgets by collecting and evaluating stool samples from world leaders, including Mao Zedong. The story really went viral after the BBC picked up the tidbit about how the KGB installed special toilets that directed Mao’s precious bodily fluids into secret boxes for analysis. “For 10 days Mao was plied with food and drink and his waste products whisked off for analysis. Once Mao’s stools had been scrutinised and studied,” Steven Rosenberg wrote with the obligatory scatological pun, “Stalin reportedly poo poo-ed the idea of signing an agreement with him.”

Now let’s be clear: Atamanenko is a regular source of sensational intelligence revelations in Komsomolskaya Pravda, most of which are pretty ridiculous. You know, things like the “real story” about Harry Truman’s decision to drop the Big One on Japan (to deter a Soviet invasion of Turkey); detailing the CIA’s involvement in the infamous flight of Mathias Rust into Red Square; and my favorite — that Stalin squelched a Soviet assassination attempt on Hitler because he worried that a Hitler-less Germany would conclude a separate peace. (The latter is actually the plot of the made-for-TV sequel to The Dirty Dozen.)

So, caveat lector.

Stories about intelligence agencies collecting poop and pee are a hardy perennial of spy lore. In 1987, a source told syndicated columnist Jack Anderson about a series of failed efforts by the CIA and Britain’s MI6 to collect a stool sample from Mikhail Gorbachev in advance of the Soviet leader’s visit to Washington. The Daily Record in Ellensburg, Washington, ran the item under pretty much my favorite headline of all time: “Flush Twice, Mikhail.”

The French may have had more success — or might just be better braggarts. A retired French spymaster told a Time journalist a story about successfully collecting a urine sample from Leonid Brezhnev. “He was staying at the Hotel d’Angleterre in Copenhagen during a state visit,” Alexandre de Marenches recalled. “Our people rented the suite under his and dismantled all the plumbing. They intercepted his toilet flushings and sent the samples to Paris for analysis.” I don’t know if it is true, but I don’t want to know if it is false.

There are many more stories like this, all of varying reliability, about pretty much every other famous world leader. You simply aren’t famous unless there is an apocryphal story about an intelligence agency trying to steal your excrement. The CIA does have its Medical and Psychological Analysis Center. One person who studied the center closely told Voice of America that he was not able to “find any sort of firm declassified information” proving that the CIA collected poop, but he also said that he has heard all the same stories by “well-placed knowledgeable sources.”

And what stories they are! I presume there is a kernel of truth to them. Spies aren’t above digging around in sewage. During the Cold War, Western military liaisons posted in East Germany found Soviet latrines to be a rich source of classified documents. While the rest of us joked that Soviets had to queue in line for toilet paper, Soviet military officers didn’t queue. They simply used whatever documents they had on hand, which intelligence officers were only too willing to pick up, wipe off, and bag for later analysis.

At a more prosaic level, we know intelligence agencies around the world are monitoring the health of foreign leaders. That’s part of their job. BBC’s Witness podcast has a great story about Nigerian opposition figure M.K.O. Abiola literally dropping dead in a meeting with Tom Pickering and Susan Rice. Probably would have liked to have seen that coming.

But hey, life is uncertain. World leaders have dropped dead unexpectedly. And plenty of other world leaders have lived long past their sell-by date. What is Fidel Castro still doing alive? Robert Mugabe? Given the potential for unrest, it probably doesn’t hurt to keep an eye on world leaders’ health, but we should have realistic expectations for what we might learn.

Which gets to a much larger point about how policymakers, from Stalin to Barack Obama, talk about intelligence. They often seem to expect that intelligence will be perfect or at least so good that the right course of action is obvious. But the world is far too complex for that. I am fond of saying policymakers don’t have a right to perfect intelligence information. Analysts can give them information, but making judgements about the accuracy of that information and guarding against the possibility that it is wrong is why the president and other senior policymakers get the big bucks.

This tendency — to think that more information can produce a clear solution to complex problems — is harmless when we are studying Vladimir Putin’s poop to see how his health is doing, but it can distort policy. Nothing annoys me more than when senior officials blame the intelligence community for the clusterfuck that was the invasion of Iraq. Really? It never occurred to them that sometimes intelligence is wrong? What next? That we can’t believe everything we read in the papers?

The demand for perfect intelligence can lure the intelligence community to go to absurd lengths to find information that is likely available elsewhere. It’s funny when we are talking about bagging poop, but what about the stories that the NSA tapped German Chancellor Angela Merkel’s cell phone? Analysts reportedly wanted to understand how she made decisions. Fair enough, but is there anything we could learn about her decision-making that we couldn’t figure out by convening a conference of German academics and former officials who know her? Or, say, by having Obama actually talk to her? Was the information worth the resulting damage to the relationship?

I don’t think so. One of my favorite articles on intelligence is a chapter in a 1999 book written by Richards Heuer, a longtime CIA analyst. The article, originally published in 1979, is titled, “Do You Really Need More Information?” Heuer cites social science research that notes that gamblers who bet on horse races don’t get better with more information — but they do become more confident. This is true for lots of fields, Heuer argues, including medical diagnoses. Heuer’s argument was the analysts could do better by improving methodology instead of trying to collect yet more information.

In recent decades, social science research has tended to support Heuer’s argument. Experts aren’t particularly good at making predictions, which says something about expertise. Recent work by scholars like Philip Tetlock demonstrates that there are things people can do to improve prediction, but these focus on improving methods rather than simply feeding more information into a poorly functioning system. To perfectly illustrate Heuer’s argument, the so-called “superforecasters” from Tetlock’s Good Judgement Project — non-experts who are good at turning information into predictions and assessments of confidence — outperformed intelligence analysts with access to classified information.

It isn’t that analysts aren’t smart, but they have the wrong idea about how to be smarter. Tetlock has “10 commandments” for making better predictions, but they all boil down to focusing on methodology, especially looking for bias. Superforecasters, I would argue, are the sorts of analysts who prefer to be wrong for the right reasons, rather than just being lucky. They don’t make perfect predictions, but they tend to be better about knowing which predictions are high confidence and which are not. To put it simply, their advantage is better methodology, not more bags of poop.