They Punished the Question

McKinsey sells boards the nerve to lead under pressure. When its own partner asked a question, it folded in 12 days.

Sooner or later every firm gets the call. One of its people has said something in public, a crowd has decided it was unforgivable, and a journalist wants to know by close of business what the firm intends to do about it.

What a firm does over the next few days is the most honest statement of its values it’ll ever make, far more honest than anything on its website, because every person on the payroll watches it, files it away, and acts on it for years. They’ve just been shown exactly what happens when it’s their turn.

McKinsey got that call last month. It’s the firm boards pay to teach them how to hold their nerve. It didn’t hold its own.

What happened

On 24 September a LinkedIn post about luxury brands including Gucci and Dior pulling products from David Jones suggested the store’s chief executive, a woman, was on the “glass cliff”: the idea that women who break through to the top are disproportionately handed organisations already in trouble, where they’ll wear the blame when it goes wrong.

Chris Bradley, a senior partner at McKinsey and at the time one of the directors of the McKinsey Global Institute, asked a question underneath it. “What on earth does the management’s gender have to do with it?”

A co-author of a newly released book on women in the workplace told him it wasn’t her job to educate him, and tagged McKinsey. Bradley pushed back. “Is this actually a thing?” Then he went further and called the glass cliff “a very convenient way to always be a victim in the rain and empowered in the sunshine”. He later acknowledged he didn’t know the research. Some read that line as flippant and provocative. Others read it as a blunt man making a fair point, that a theory which covers you in the rain and the sunshine alike can never be shown wrong, and I don’t see why the offended reading gets to be the official one. Being rude on LinkedIn is a fair thing to be pulled up on, but one person’s rude is another person’s reasonable. If a single blunt line in a comment thread can cost a man his leadership roles and a charity board seat, the line isn’t the problem. A lot of people need to grow a bit of backbone.

Either way, it isn’t grounds for a conduct review. Hold that thought.

On 30 September Bloomberg reported that McKinsey had rebuked him. The firm said it was “disappointed”, and that it “stands by our research and leadership on these topics”. Twelve days later Bloomberg reported that, after what McKinsey called a “conduct review process”, Bradley had been removed from his leadership roles, including his directorship at the institute, because his comments “fell well short of the standards we expect at our firm”. McKinsey hasn’t said which standard. He kept his title as a senior partner.

Then Bloomberg started asking Opportunity International Australia, a charity on whose board he sat as an unpaid director, about his position. He offered his resignation, and on 15 October the charity accepted it. Its statement thanked him for more than 25 years of supporting its work with more than 10 million women and their families, lifting themselves out of poverty.

The man held up as dismissive of women had spent more than 25 years backing a charity that helps women lift themselves out of poverty, and he sat on its board for nothing.

Claire Lehmann pieced the whole sequence together for Quillette on 15 October, and her account is worth reading in full. She followed what the book’s authors did with it. I’m more interested in what the firm did.

A research institute that won’t ask a research question

Bradley’s role at the McKinsey Global Institute was research. The institute exists to find out whether things are true. “Is this actually a thing?” is close to its job description.

On the evidence as it stands, the honest answer is: not clearly, and not for chief executives. Psychologists Michelle Ryan and Alexander Haslam proposed the glass cliff in 2005. A 2020 meta-analysis found a small effect that changed with context. A study of American companies published in The Leadership Quarterly in 2025 found no clear evidence that women are disproportionately appointed to lead troubled firms. Two studies published this year found little evidence of it among chief executives, though one, of Britain’s largest listed companies, did find support for it among female board chairs.

That’s a live hypothesis, not settled science. Bradley asked a fair question, bluntly. The firm punished the question.

A research institute that can’t question a hypothesis has stopped being a research institute. It’s a marketing department with better letterhead.

Follow the product

There’s one question I bring to any document of consequence before I read its conclusions, and I wrote about it in The Captured Boardroom: whose salary was the writer drawing while writing this? Turn it on the firm and the decision stops being puzzling.

McKinsey isn’t a neutral observer of gender research. It publishes it, and it sells the advice that follows from it. Its Women in the Workplace study, produced with LeanIn.Org, has run every year since 2015. Its diversity series, starting with Why Diversity Matters in 2015, told boards for the better part of a decade that diverse executive teams outperform. When two American academics, Jeremiah Green and John Hand, tried to replicate those results from public data on S&P 500 companies, they couldn’t, and a companion study with a third co-author found the same null result for the share of women executives. McKinsey says it stands by its research, and that the research shows correlation, not causation. In August a US congressional task force wrote to the firm’s global managing partner demanding the data behind all four diversity reports.

Now look at what was used against Bradley. In the thread, the reply explaining why gender was relevant cited McKinsey and LeanIn’s Women in the Workplace research. Bloomberg’s story drew on the same research, and pointed to a 2020 McKinsey healthcare report that used the term “glass cliff” itself. And the firm’s first public response did more than distance itself from his comments. It said McKinsey “stands by our research”.

A McKinsey partner questioned a gender claim in public five weeks after Congress started demanding the evidence behind the firm’s diversity research. The firm answered by defending its research and reviewing his conduct.

When a partner questions the product, the product wins. I don’t need to know what was said inside the firm to read that structure. The incentive was on the table before anyone walked into the room.

Outrage has a business model

This is the part that should worry every owner reading, because it’ll reach your business long before it reaches McKinsey’s again.

Lehmann counted the posts. Between the book’s launch in mid-September and 15 October, its two authors published 29 LinkedIn posts about it. Ten of them referred to Bradley or McKinsey, and none of those appeared before Bloomberg’s first story. After the removal, eight of the next nine did. One author wrote that the episode had “gift-wrapped itself”. The other called the removal “a small bit of accountability” and put a link to buy the book underneath it. When Lehmann put questions to one of them, the reply closed with an invitation to sign up for workshops.

I’m not suggesting anyone engineered it. Lehmann looked for evidence of that and found none, and the defensible version is worse anyway. Nobody had to engineer anything. A man’s career became a launch campaign because outrage converts, and whatever converts gets repeated.

Chetna Mahadik, in the post that put this in front of me, named it plainly: activism that “has transformed into a destructive commercial activity”. Every firm that hands over a head to make the noise stop isn’t buying peace. It’s funding the next campaign.

Call it by its name

Here’s where I draw the line, and I’ll draw it exactly where the other side drew theirs.

For years we’ve been told that “toxic masculinity” names conduct, not men. Which conduct has never been pinned down. In practice it stretches to cover whatever anyone wants to attribute to a man that week, and that’s what makes it so useful to whoever’s throwing it. I’ve never accepted it as a description of anything. But if it’s going to be the standard, it’s the standard in both directions.

Lehmann records what followed the removal. Followers cheered it. Bradley was called a “dolt”, a “dope”, and a “goon”. One told him to trim his beard and get a haircut. Another suggested publicising the names of people who’d defended him, so their employers, friends, and wives could be told about their alleged misogyny.

That’s toxic. Done in feminism’s name, it’s toxic feminism, and I use the term deliberately, as the mirror image of the one that’s been thrown at men for years. It doesn’t describe women. It describes conduct, and the conduct is on the record.

Some women gang up. So do some men. Anyone who’s run a business has watched both and had to deal with both. The difference is that one has a name, a policy, and a training module, and the other has a book launch. Used this way, both labels do the same job: they teach each half of the population to treat the other as a suspect. That wedge doesn’t help a single woman in business. It makes every working relationship between a man and a woman more careful than it should be.

The same single standard applies to the glass cliff itself. A woman who isn’t appointed is evidence of discrimination. A woman who is appointed to a business in trouble is also evidence of discrimination. One commenter under the post I read put it better than any columnist has: “Perhaps it is logic that is the real victim.” Men are handed failing businesses every week and shown the door from them every week. We call that the job.

Who actually pays

Bradley is still a senior partner at McKinsey, and I expect he’ll land on his feet. The people who pay for this are women.

Every senior man in every firm watching has just learned that engaging with a woman’s argument in public, disagreeing with her, asking her for the evidence, is a career risk. A few will stop mentoring. Most will simply get careful. LeanIn.Org, McKinsey’s own research partner, measured exactly this in 2019, in the wake of #MeToo: 60 per cent of male managers said they were uncomfortable taking part in a common work activity with a woman, such as mentoring, working one-on-one, or socialising. That was up 32 per cent on the year before. Careful is the worst thing that can happen to a talented woman’s career, because careful men give careful feedback, and careful feedback is how good people plateau.

I wrote Mis(très)s Entrepreneur Manifesto out of years of work with women who ran their businesses with more commercial precision than most chief executives I’ve sat across from. Their power came from sovereignty, not permission. None of them wanted a theory that explained a bad quarter before it happened. They wanted to be judged on their numbers and backed when they were right, which is what every serious operator I’ve worked with wants. A glass cliff story hands a woman in a hard job an alibi she didn’t ask for, and tells everyone watching she was set up to fail. That’s an excuse written for her in advance, and nobody who’s any good at the job wants one.

The chief executive of David Jones took on a hard business. She’s owed what every chief executive is owed: to be judged on what she does with it.

When it’s your desk

Most of us don’t run a firm the size of McKinsey. That doesn’t spare us. Everyone with a phone is a publisher, everyone with a grievance has a following, and some of them have something to sell. The call will come. Here’s what holding your nerve looks like when it does.

Sort misconduct from disagreement before you do anything else. Harassment, dishonesty, abuse, a breach of client confidence: those go to a process, and the process runs whoever is watching. A question, however clumsy, is a disagreement. If it wouldn’t have been a disciplinary matter without the screenshot, it isn’t one with it.

Run it on your clock, not the journalist’s. A five o’clock deadline is the reporter’s problem. “We’re aware of the exchange and we’ll be speaking with our colleague” buys all the time you need. Nobody has ever been criticised for failing to sack someone by close of business.

Ask who wants the head, and what they get if they get it. A following, a book, a course, a speaking circuit. That doesn’t make the complaint wrong. It means you weigh it the way you’d weigh a report from any interested party, which is carefully.

Say the sentence, then stop. Another commenter in the thread wrote the statement McKinsey should have issued: “I strongly disagree with the view our colleague expressed.” That’s the whole of it. It doesn’t continue, in the commenter’s words, “and we fired him”. You can disagree with your own people in public. Hand them over instead and you get the senior men I wrote about in Programmed To Apologise, who soften before they direct and apologise before they decide.

Deal with tone in private and in proportion, if it needs dealing with at all. At most, the “victim in the rain” line warranted a conversation with a senior colleague about how a partner wants to be read in public. A conversation. Not a conduct review, not the loss of his roles, and not his seat on the board of a charity he’d backed for more than 25 years.

If you sit on a board, settle it before the call comes. Ask management this week what the firm does when one of its people is in the news for something they said. If the answer is “it depends”, it’ll depend on whoever is loudest. That’s democracy at its weakest: rule by volume, not by numbers. The group screaming loudest is very often a small minority, and it still gets its way.

Above all, know who the audience is. The crowd moves on within the week. Your own people stay, and every one of them is watching what you do and learning what you’ll do to them. McKinsey’s tens of thousands got that lesson this month. Watch how many of them ask the next uncomfortable question. In 1982 two McKinsey men, Tom Peters and Robert Waterman, wrote In Search of Excellence out of research they’d done inside the firm. On this showing, McKinsey needs to go in search of it again.

The verdict

McKinsey says it acted on its standards. It hasn’t said which ones. Whatever the intent, what the decision protected was a product line and a quiet news cycle.

Conviction cuts two ways: the backbone to hold a position, and the verdict when the people in charge don’t. Bradley was blunt in a comment thread. His firm lost its nerve in a decision. Only one of them paid for it.

His question is still sitting there, unanswered by anyone who punished him for asking it. Is it actually a thing? On the evidence, not clearly, and not for chief executives. Giving that answer is what a research institute is for. It only has to be willing to hear it.