Two weeks ago the CEO of coinbase, a tech company with about 1,000 staff, publicly shared a new internal policy about their company mission, which explicitly excludes talking about political issues that aren’t specifically related to their product or industry.
“It has become common for Silicon Valley companies to engage in a wide variety of social activism, even those unrelated to what the company does, and there are certainly employees who really want this in the company they work for. So why have we decided to take a different approach? The reason is that while I think these efforts are well intentioned, they have the potential to destroy a lot of value at most companies, both by being a distraction, and by creating internal division.”
This came after many employees expressed frustration that the company wasn’t expressing public support for Black Lives Matter. The post generated a lot of discussion among tech leaders and ultimately resulted in 60 employees leaving the company.
On the other end of the spectrum, Expensify, an expense tracking tool for small businesses, sent an email to their 10 million customers imploring them to vote for Joe Biden!
I can certainly see the merit of much of the criticism. If your employees face discrimination on the basis of their gender, religion, race or sexuality, advocating for their rights is distinctly different from endorsing political candidates. Supporting a local pride festival, making statements in support of racial equality or gender-pay-parity don’t strike me as incompatible with the mission of most companies, or their duty of care to their employees.
On the other hand, I also see the logic in some of his argument. A company is a group of people who come together to build a product or deliver a service, with responsibility to follow the law, make customers happy and generate shareholder return. If people want to advocate for other political change, should they not join different groups of people who come together for those purposes? Like campaigns, political parties and trade unions?
Popular conceptions of capitalism are shifting, with many people considering a company’s responsibility to a wider group of stakeholders, like employees, local communities, the environment and the wider economy.
My personal politics likes a lot of this. In particular, the increasing acknowledgment that environmental sustainability is a responsibility of corporations too. However, I don’t think we acknowledge the trade-off we’re making and the risk we’re creating if we advocate a multi-stakeholder approach.
Often the winner in this move is the CEO and senior management. In a more 1980’s version of capitalism, when we measured CEOs against a single goal of maximising shareholder value, it was a crude yardstick, and didn’t produce ideal societal outcomes, but it had the distinct advantage of being simple and measurable.
Whereas before the CEO could get fired when the share price dropped, they now have the flexibility to argue that they’re doing a good job because they were balancing shareholder value against the interest of some other group of stakeholders. It’s a lot of extra wiggle room, which reduces accountability and concentrates power.
We can hope they use this increased power for good, but I’m not sure we should trust that they will.
If you’re supportive of companies engaging in social activism specifically, or just wider stakeholder-serving in general, Matt Levine succinctly describes the power shift that will accompany it. “It is productive—not 100% accurate, but a useful heuristic—to assume that all corporate governance debates in the U.S. are about whether shareholders or managers should have more power to control the corporation. There are other stakeholders, sure, but they are mostly tools in the shareholder/manager fight, not power centers in themselves.”
That’s why the Expensify CEO was able to justify a Biden support email to 10 million customers – because “Expensify depends on a functioning society and economy; not many expense reports get filed during a civil war,” – while the CEO of Soylent said he was voting for Kanye, and, if we’re not holding them to the singular goal of profit-maximising, what is the new rule set that constrains either of them?
Section 230. The big tech CEOs were in front of US congress again this week, officially to discuss “Section 230”, the piece of US law that makes tech companies immune from liability for what their users post, upload or tweet. It’s an important and difficult ongoing debate – should we treat them more like telecoms companies, who aren’t responsible for the content of a text message they deliver, or more like newspapers, whose editors are wholly responsible for the information they publish? Unfortunately, after all the questioning, we’re none the wiser.
As the New York Times reported “The theatrics, which often devolved into shouting, meant that the topic of the hearing — the future of a legal shield for online platforms — was barely debated.”
In Ireland, in every election or referendum I’ve been involved in, a daily activity of most Communications Managers is to argue with RTE that your side isn’t being treated fairly or getting enough coverage. Even if that’s not entirely true, the theory goes that at least they’ll think twice about anything that could advantage the opposition, because they’ll get an earful for it. That happens behind closed doors mostly, because you can sound like a crank calling RTE or the BBC biased publicly. People have less sympathy for tech CEOs being shouted at, which is fine, I guess, but it still results in a cycle of politicians shouting at the referee trying to pressure them on specific decisions, rather than engaging in the more important debate about what rules these refs should be following. Link.
Arrested for posts. 10 people have been charged with sharing, on Facebook and Twitter, the names of the two boys who murdered Ana Kriegel. As both were 13 at the time, it is illegal to share information about them. Without commenting on the law itself, if it is to be prosecuted this seems to me like the right approach. Many disagree, with journalists like Matt Cooper and Tom Lyons calling for Facebook and Twitter to face punishments for allowing the posts to be published in the first place. Link.
In both the US and Irish case, this is a difficult debate but one we need to be having more. I worry about laws that mandate pre-screening of content will result in companies being over-cautious and prohibiting much speech and expression that we don’t want to stop. Stricter rules will also disproportionately hurt the smaller tech companies. Facebook, Google and (probably) Twitter have the resources to implement more restrictive rules, but most of their smaller competitors don’t.
Some positive next steps here would be laws that mandate transparency for content moderation policies and processes, to set targets for how quickly offending content is removed and to ensure people have a fair right of appeal if they feel their posts have been removed unfairly. There’s a difference between hosting content and amplifying it too, which could be regulated in different ways. Beyond that it gets more difficult to weight the trade-offs.
Oversight. The Facebook Oversight Board (aka the “supreme court”) is starting to hear cases. Select content moderation policies will be discussed and debated in public. I’m optimistic about this one. Link.
Whatsapp. In the last week of the US presidential campaign misinformation campaigns are ramping up on messaging platforms like Whatsapp and SMS. We probably won’t get a full sense of their scale and impact until months after the election, if at all. Link.
💡 Interesting Links
AOC. US Congresswomen Alexandria Ocasio Cortez and Ilhan Omar live-streamed themselves playing to new hit game “Among Us” to get young people to vote. Over 400k people watched it live, and millions since. Here’s the highlight reel – link.
Tables Turned. Activists are using facial recognition to identify police. Link.
Spam Bias. Someone decided to audit the algorithms that decide which emails should go into spam folders, with surprising results. Outlook, for example spammed a job application containing the word “Nigeria”, but didn’t when the word was removed. Link.