Elon Musk and the power of the bro-code

How doing bad things together creates bonds of trust

The Wall Street Journal has a new article alleging that Elon Musk and various board members have taken illegal drugs together.

The volume of drug use by Musk and with board members has become concerning, some of these people said.

In the culture Musk has created around him, some friends, including directors, feel there is an expectation to consume drugs with him because they think refraining could upset the billionaire, who has made them a lot of money, some of the people said. More so, they don’t want to risk losing the social capital that comes from being close to Musk, which for some feels akin to having proximity to a king.

Musk and his lawyer, Alex Spiro, didn’t respond to requests for comment.

Very obviously, I know nothing about these allegations, beyond the fact that the WSJ has published them. But the alleged behavior sounds similar to a lot of other stories, and in particular to the kinds of things that I used to pay attention to when I was a young assistant professor working on how people who don’t have much reason to trust each other can figure out how to work together.

The most cogent example that I know of is from an article by James Palmer, now a deputy editor at Foreign Policy. James’ first sentence is a classic in getting your attention: “Turning down an after-dinner invite to a brothel is always a social minefield.”

But then he explains why the social practice of sealing business deals at brothels exists (or at least used to exist until quite recently) in China. As James’ article explains at length, this involves a lot of the factors you might expect: toxic understandings of masculinity; men using women as a way to communicate their relative status in the pecking order. But there is a more specific question: why are collective visits to brothels linked to commercial dealings, and in particular to sketchy commercial dealings?

Here, James turns to the ideas of an Italian sociologist, Diego Gambetta, who has deeply influenced my own work on trust (the only chapter of my book on trust that a non-specialist might want to read is an extended riff on Gambetta). People who want to do sketchy stuff with other people face some big dilemmas of trust. Gambetta, crudely speaking, points out that criminals, or potential criminals, face two serious problems. First - how do they identify other potential criminals who they can work with? (this is what game theorists call a ‘signaling’ problem). Second, how do they trust those criminals when they find them?

When I did research on the “Silk Road” online drug market a decade ago, I found a fantastic quote from an unhappy wannabe criminal on the Hidden Wiki that illustrated both problems at once.

I have been scammed more than twice now by assholes who say they’re legit when I say I want to purchase stolen credit cards. I want to do tons of business but I DO NOT want to be scammed. I wish there were people who were honest crooks. If anyone could help me out that would be awesome!

Finding other criminals is hard; finding other criminals who won’t rip you off is even harder. And that helps explain why businessmen engaged in sketchy business dealings go to brothels together to celebrate a deal. In James’ words:

the purpose of these visits isn’t a good time. It’s to cement business and personal ties, binding men together through the power of taboo and mutual self-exposure, or at least the pretense of it. It lets them judge that the others involved in a potential deal are men of the same stripe. … In part, the power of the experience comes from the mutual pleasure of shared transgression, the feeling of a shared secret. … But brothel visits in and of themselves give only slight leverage over the other party. These ties can be deepened through more serious offences, like sharing drugs, most popularly ketamine. … But vice serves as a kind of screen, weeding out the rare few who might have moral qualms about future dealings. It tells both sides that they’re playing by the same rules. … Refusing to play the game, on the other hand, comes at a sharp cost. Businessmen who convert to evangelical Christianity and make a commitment to avoid vice or bribery describe sharp business losses as a result, as former partners turn away from them, fearful of their newfound probity.

As an aside: this WSJ piece and a past WSJ article suggest that ketamine is one of Musk’s drugs of choice.

The actually relevant point is that doing bad things together serves two useful purposes for criminal and quasi-criminal enterprises. First - it enables you to identify and screen out the people who have real moral qualms before you have to rely on them to cooperate in whatever quasi-legal thing it is that you want them to do. Second, it gives you and them a stake in mutually assured silence. There is a third thing too: it plausibly corrupts people, rewiring their moral codes so that things that they might once have identified as wrong come to seem like ordinary practice. Most people don’t want to seem evil to themselves; equally, most people don’t want to feel that they are suckers or losers. If you can get people to feel as though they are a sucker or loser if they don’t participate, you’re half-way towards pulling them in.

Gambetta provides other examples of how corrupt systems of trust work in his book, Codes of the Underworld. Why is it that Italian professors not only are often incompetent at research, but (according to his account) boast loudly about their incompetence? Gambetta says that they are behaving perfectly rationally. Their incompetence and their boasts are costly signals that they can be relied on. When someone who is in their academic clan asks them to find a job for their nephew or research assistant, they won’t ask indelicate questions about whether the person is intelligent and qualified to carry out good academic work. They’ll just find a job for the idiot, on the assumption that the favor will be returned in due course. Boasting about incompetence is signaling that you can be trusted to be corrupt.

The New York Times published an article on another everyday conspiracy a couple of weeks ago: cops fixing traffic tickets for other cops’ family members and friends. The article describes a cop who refused to play the game and found himself ostracized.

The month after he stopped the Mazda, a high-ranking police union official, Albert Acierno, got in touch. He told Bianchi that the cards were inviolable. He then delivered what Bianchi came to think of as the “brother speech,” saying that cops are brothers and must help each other out. That the cards were symbols of the bonds between the police and their extended family and friends.

Compare to James Palmer:

Amid businessmen, just being somebody’s “friend” (pengyou) isn’t that close. Casual acquaintances and uncertain contacts are “friends.” A “brother” (xiongdi), on the other hand, is somebody inside the circle, a man who can be trusted. “It’s like the distinction between ‘a friend of mine’ and ‘a friend of ours’ in the mafia,” explains Osburg.

The allegations against Musk and his board are just that: allegations. So treat the below as general commentary rather than specific claims about the fact patterns in this particular instance.

Where behavior like that alleged in the WSJ piece actually happens, it can have quite direct consequences for how organizations function. It isn’t that drugs like ketamine are illegal, or that they can lead to irrational decisions - it’s that shared indulgence in moderately illegal behaviors can provide the social cement for trusting relations among people who want to do other, and worse things. They are the foundations of the bro-code - and bros don’t ever tell on bros.

And these rituals can corrupt the previous uncorrupted. One standard pattern for intelligence agencies recruiting assets, as I understand it, is to get them to gather mildly problematic information first, both to habituate the asset, and to provide leverage if the asset wants to back out when you ask them to do more risky things.

The notion of organizational “culture” is important, but notoriously hard to pin down. The kind of behaviour that Gambetta, Palmer and the NYT describe help make it more concrete. Culture is often built out of rituals, which can help create and perpetuate “criminogenic” organizational incentives. They create common knowledge that it is OK - even laudable - to do things that skirt the law or even break it.

There is strong ex ante reason to worry about these cultural problems, and what they are likely to entail. How to actually solve these problems is a much tougher question. It is ironic that they seem to be endemic in the NYPD, whose former Commissioner famously popularized the “broken windows” theory that small infractions legitimated bigger ones. The broken windows approach to law enforcement obviously worked out terribly: might something like it have worked if it hadn’t been racist, and it had been turned back against the organization that was administering it? I suspect not, but I genuinely don’t know. Nor do I know much about whether there are better approaches and potential solutions out there (I can guess at some ways forward; but they are only educated guesses).