Hacker News Reader

Comment by btilly | Hacker News Reader

I immediately sanity checked this with three examples.

SBF fit amusingly well. His fraud had all of these, with Effective Altruism providing an extreme example of how goodness in some areas atones for evil in others.

Stanford's president Marc Tessier-Lavigne resigned after his reputation was found to be based on fraudulent research. Descriptions that I've read from those in his lab showed most of these.

Enron is too canonical an example to ignore. Yes, it had all of these.

View on HN

I think the sanity checking works better the other way round: how much of this is just corporate norms that apply to most companies above a certain size, including ones with no particularly significant ethical challenges? One of the bullet points starts off by pointing out that 47% of companies have one type of conflict of interest!

CEO being higher profile and a generation older than [some of] the people sounds like a pretty normal company tbh, and probably only a major issue if the ethical issues are caused by the CEO (it technically doesn't apply to SBF, although his reports were even less experienced and in awe of him)

CSR and diversity initiatives and other corporate do-good messaging? Most companies do that to some extent, and it's not like the ones who don't even attempt to launder their reputations tend to be squeaky clean rather than too cynical to even try to look like they care.

And sure, KPIs and stretchy sales targets can definitely cause bad behaviour and lying, but they're also extremely common, and can be beneficial when companies actually get them right.

I am not sure if that is an objection because "50% of companies are involved in one or more ethical violations," sounds plausible.

I don't think "Purchase or sell insider services" necessarily means a conflict of interest, insiders could be winning the bidding process honestly as well.

Insiders are usually disqualified from bidding because of an inherent conflict of interest.

It does render the notion of "ethical collapse" and supposed comparisons to Enron a bit meaningless if we're also counting the sorts of ethical violation most companies are guilty of though. It's a bit like if a study of "criminality" makes a point of honing in on everyone who's broken a speed limit before...

Exactly. The usefulness of criteria for predicting anything depends on both false positive and false negative rates, and this article doesn't seem to care about false positives (what fraction of companies do trendy, visible ESG stuff but are also honest behind the scenes?) at all.

Might as well add an 8th point, "Companies that sell a good or service", and really bring that false negative rate all the way down to zero.

you’re right, it’s like inverse survivorship bias

how much of this is just corporate norms

Evil is the corporate norm, or at least psychopathy. It's even enshrined in law ever since Dodge v. Ford Motor Co.

Treat it like the DSM: a single symptom is not enough to diagnose, you need X/7.

To be fair, so many fast growing startups would tick multiple of their 7 boxes. Have you sanity checked it the other way (false positive)?

They don't define "ethical collapse" in terms of the company failing to make money. I imagine a fast growing startup could behave unethically without anyone really noticing (or at least saying anything) for a while.

Uber was on the cusp on a bunch of these (numbers pressure, fear/silence, big CEO, innovation, goodness atoning for evil). I think fear/silence tilting + strong governance forced Kalanick out and saved the company.

WeWork had every single one of these, and governance ultimately was too little too late to save the company.

AirBnb, Stripe had a few of these, but turned out OK.

It seems like the common variable that spares startups from ethical collapse is good board governance and having a reasonable / willing to evolve CEO.

FB is an interesting case, because they check most of these boxes (maybe not fear/silence?), yet are fairly "successful". Governance isn't the explanatory variable, since Zuck has 90% of voting shares. The explanatory variable seems to be Zuck's willingness to listen to and learn from a board and his execs.

I'd argue that "successful" is a poor measure of how ethical a company is.

AirBnb has managed to shake off some pretty bad ethical issues. Facebook was credibly implicated in a genocide, and Instagram in particular (and social media in general) has been causally linked to increased rates of juvenile suicide.

Oh for sure, that's why I put "successful" in scare quotes. As mentioned, FB checks pretty much every one of the "ethical collapse" boxes with a bolded checkmark.

That they've managed to be successful financially in spite of that fact (so far!) is pretty singular. That's a rare group. Not saying it's one to emulate, though, far from it.

SBF could not have been a generation older than his reports. They'd be 10 years old.

So, 6 out of 7 then.

No, he was not a generation older. But he managed to still create a persona that was bigger than life.

I think ' a generation older' is really just a proxy to explain the social dynamic of "subordinates are unwilling to question the authority of the CEO"

Seniority can be only a part of it. A CEO who is extremely wealthy and who is an actual celebrity does not like being told "no" to. It only gets worse with time - as they begin to think they are uniquely brilliant (and funny).

There's a very, very prominent example of that today. Let that sink in.

I think it's more that he staffed his entire company with young, ambitious, and inexperienced people.

That's the canonical example, but it's not about age in and of itself, but the gap between the CEO that turns them from a person into an idol. It's often an age gap (especially in the past) due to the culturally-entrenched ideal of automatic respect and deference to one's elders, but it can be any gap that makes the direct reports feel like they are somehow inherently beneath the CEO in ways other than corporate rank.

The point of that sign is that the CEO's direct underlings can't even question the CEO's choices.

It's not just their direct reports or even their employees. HN, Twitter, and Reddit are full of commenters simping and white-knighting for CEOs they have no relation to, whenever someone questions them. Corporate leaders have become this weird "priesthood" complete with followers and apostles praising them online.

Has OpenAI just joined the club too? What is the score?

6/7, the board of directors was not "weak" (they fired the CEO). Guess which one of the 7 Silicon Valley is now fiercely complaining about...

Arguably they were weak, but not entrained enough to the CEO to meet the criteria of this definition of weak.

If you want an even better example look at Steve Jobs.

It's easy to find these things in retrospect. It could be that ethical breaches are common and it's easy to ascribe lapses and wrongdoing to someone after a scandal.

For the model to be useful we have to show it can be used to consistently predict scandals and wrongdoing before they happen and before others notice

Eg. Can we confidently use this model to short corporate stock

Wells Fargo misses most. So does HSBC

I don't know enough about what happened inside of those to evaluate.

They're banks, so the assumption is lack of ethics.

They appear to from the outside. Most of the examples of ethical collapse that I'm aware of weren't entirely clear until after the organization collapsed and the dirty laundry went public. A lot of these would either be actively suppressed or they'd only be clear to people who are inside the organization.

In other words, the list is usually only useful to insiders.

Only those trying to blow a whistle (generally not good for one’s health), or get out before the indictments start landing anyway.

HSBC did everything right and still gets crucified for it. Whether your system can distinguish between Wells Fargo and HSBC is actually a pretty good test case.

Reply on news.ycombinator.com