Fear-Based Leadership Worked at Volkswagen. Until It Didn’t.

In July 2013, on a test drive in the United States, Martin Winterkorn, the CEO of Volkswagen, noticed a paint imperfection on a Beetle. The paint was slightly thicker than the company standard. Less than a millimeter too thick, according to a VW source who spoke to Reuters. A difference you would not be able to see without a gauge.

He stopped and lectured his engineers about the waste.

This was 2013. Winterkorn was one of the most powerful executives in the global auto industry. VW was racing Toyota for the title of the world's largest carmaker. The company had set ambitious targets to crack the American market, which meant selling millions of diesel vehicles in a country with increasingly strict emissions rules. The engineers on that test drive would have been working on multiple serious problems. Strategy problems. Technical problems. Compliance problems.

The CEO talked to them about paint.

A Year Later

In 2014, Winterkorn gave an interview to Der Spiegel magazine. He complained that his team had not pushed him hard enough on the Passat line, which was not selling well in the US market. In his own words:

“They should have come and said ‘Herr Winterkorn, we must update the Passat'; they should have jumped on my desk.”

Reuters interviewed former VW managers after the Dieselgate scandal broke. Five of them described a consistent pattern. One of them summed up what it was like to work there:

“There was always a distance, a fear and a respect.”

Few executives, they said, dared to approach Winterkorn.

Hold those two things next to each other. The CEO who lectures his engineers about a millimeter of paint is the same CEO who later complains that his engineers did not jump on his desk.

What would have happened to the engineer who actually did?

That is not a rhetorical question. Someone on that 2013 test drive, in theory, could have interrupted Winterkorn's paint lecture and said, “Herr Winterkorn, I agree that the paint tolerance matters, but we need to talk about something more important. Our diesel engines cannot meet the US emissions standards within the cost and timeline you have given us. We have installed software to cheat the test. This will eventually be discovered. We need to stop.”

You know what would have happened. So did the engineer.

What Bob Lutz Told Me

In 2013, the same year as the paint story, Bob Lutz came on my podcast. Lutz is a legend in the auto industry. Former vice chairman at GM. The kind of person whose opinions about management I wanted to hear even when I didn't fully share them.

We got to talking about leadership styles, and Lutz offered Volkswagen as an example of a particular kind of success. His words:

“He would not only tell them how to do it, he would tell them what to do, a totally autocratic system where everybody lived in fear. I would hate to work for a company like that, and I know that a fear-driven environment is not a pleasant one, but Volkswagen Audi today is the most successful car company in the world. They're on the cusp of becoming the world's largest automobile company, run by total autocrats.”

Bob was not wrong.

I have to say that clearly, because it is the part of the story people want to skip over. In 2013, looking at VW's market position, revenue, and growth trajectory, you would have been a fool to bet against them. The autocratic model appeared to be working. Lutz was not defending it as a leadership philosophy. He didn't want to work in it. But he was reporting, honestly, on what the scoreboard said. And the scoreboard said VW was winning.

My annotations on the transcript from that day include the notes “this is bad” and “I agree with Bob here” and, after his claim about commercial success, a question I underlined:

“So is VW doomed to failure then? Or does that style work somehow?”

I had the intuition. I did not press on it.

The Software That Nobody Stopped

In September 2015, the EPA announced that Volkswagen had been cheating on US diesel emissions tests for years. Software in millions of vehicles could detect when a car was being tested in a laboratory and activate emissions controls for the duration of the test. On the road, in normal driving, the controls were suppressed. The vehicles emitted up to 40 times the legal limit of nitrogen oxides.

The software had been in use since 2006.

For nine years, the fraud ran. For nine years, engineers worked on it, improved it, adapted it to new vehicle models, and hid it from regulators. VW's internal investigation, conducted with an outside US law firm, eventually identified approximately 450 people internally and externally who were involved.

Four hundred and fifty people. That number is worth sitting with.

None of them stopped it.

This is the part of the VW story that the conventional retellings skip. They focus on the executives who knew, or the deception of regulators, or the dollar figures attached to the settlements. Those are real, and the dollar figures are staggering. The total cost as of 2020 was more than 33 billion dollars. Winterkorn was forced out in September 2015 and later charged criminally with fraud and conspiracy. But those numbers are not the story. The story is that 450 people, over nine years, none of whom benefited personally in any material way, continued to participate in something they knew was illegal because the alternative was worse.

What was the alternative? Tell Winterkorn. Tell Piech. Jump on the desk. It might have looked like the image below. If it had happened.

What Bob Was Describing, and What He Was Missing

Go back to Bob's quote for a moment. He was describing what VW looked like from the outside in 2013. An autocratic system. Fear-based. Producing extraordinary commercial results. Bob's implicit argument was that the results were the vindication. Fear worked, even if Bob personally didn't like it.

What Bob was looking at, and couldn't see from the outside, was that the fear and the results were the same thing.

The engineers who would not push back on paint were the engineers who could not push back on the emissions fraud. The culture that made Winterkorn fear-inspiring in the first sense was the culture that made him uninformed in the second. The commercial success that Bob was pointing to was already being produced, in 2013, with the help of a fraud that the CEO either didn't know about or didn't want to know about. You cannot tell which of those is worse.

And here is the thing Bob could not have known, and I could not have known when I was sitting across from him. At the moment Bob was describing VW's model as commercially successful, the defeat device had been operational for seven years. The instability Bob was implicitly dismissing was already present, already compounding, already producing the catastrophe that would arrive two years later.

The fear was not producing the success in spite of itself. The fear was producing the fraud, which was producing part of the success. The two could not be separated. What looked like a functional autocracy was a system with a hidden liability that nobody could report and nobody could fix.

You cannot assess a management system by a snapshot of its results. You have to assess it by asking what it is doing to the information flow, and whether the information required to prevent catastrophic failure can actually reach the people who need to act on it. At VW, the answer to that second question was no. The results said yes, right up until they couldn't.

The Move That Every CEO in This Story Makes

Here is what I keep coming back to, because it is the part that explains why this keeps happening.

Winterkorn, in his Der Spiegel interview, expressed genuine frustration that his engineers had not pushed back on him. The frustration was sincere. He was not, in that moment, being cynical. He truly wished that his team had told him the truth about what was possible and what was not.

What he could not see was that the wish was in direct conflict with the culture he had designed. The paint lecture in 2013 was not an accident. It was a signal. It was how Winterkorn had taught his organization what happened when you brought him information he did not want. The signal had been received, perfectly, by everyone who needed to receive it. The engineers on that test drive went back to headquarters and told their colleagues: do not bring up anything the boss has not already blessed.

A year later, when the boss complained that nobody was bringing him hard truths, the organization he had trained was doing exactly what he had trained it to do.

This is not a VW problem, and it is not a Winterkorn problem. It is the move that every leader in this pattern makes, and it is worth naming clearly. The leader builds a culture that punishes honest upward communication. The culture produces short-term results that benefit the leader. The leader receives those results as evidence that the style is working. Eventually, a problem arises that requires honest upward communication. The culture fails to deliver it. The problem becomes a crisis. The leader, in the aftermath, expresses sincere frustration that nobody warned them.

Mary Barra, after the GM ignition switch scandal, launched a campaign called “Speak Up For Safety,” built on the implicit premise that GM's employees had failed to speak up loudly enough. The premise located the problem in employee behavior, not in the culture that had shaped it. Every post-scandal CEO speech I have read in the last twenty years makes some version of this move. The leader's culpability is framed as not having known. The question of why they did not know is not examined.

The engineer on the test drive could have told Winterkorn about the defeat device. The engineer did not. Not because the engineer lacked courage, and not because the engineer lacked integrity. The engineer had correctly read the signals about what happened to people who brought Winterkorn problems he had not already authorized. The signal had been sent. The signal worked. And when the system failed, Winterkorn was upset that the signal had worked.

What I Should Have Said to Bob

I have thought about this interview many times over the years, especially after the VW scandal broke in 2015. What I wish I had said is something like this.

Bob, you are describing a system that is commercially successful today. I believe you. But commercial success in the current quarter is not the same as organizational stability, and a management style that produces fear cannot produce the kind of information flow that catches problems before they become catastrophes. What you are describing is a culture that can run for a long time while looking successful, until the first time it encounters a problem that requires somebody to tell the boss something the boss does not want to hear. On that day, the system will fail. And because the system has been quietly optimizing for the appearance of success over actual truth-telling for years, the problem will be much larger than it would have been in a culture that surfaced problems early.

We cannot see those problems today because the system is designed to hide them. But they are there, Bob. The absence of visible problems in a fear culture is not evidence of the absence of problems. It is evidence of the presence of fear.

I did not say any of that.

What I said, in my private annotations to the transcript, was “So is VW doomed to failure then? Or does that style work somehow?”

I had the question. I did not have the confidence to turn it into an argument in the moment. I wish I had.

The Engineers Knew

Here is what I think about when I reread that interview now.

Every one of the engineers on Winterkorn's test drive in 2013 knew more about what was actually happening at VW than Winterkorn did. They knew about the defeat device, or at least some of them did. They knew the emissions targets could not be met legitimately under the cost constraints. They knew the fraud had been running for seven years by that point.

They also knew that Winterkorn was about to lecture them for several minutes about a paint imperfection that no customer would ever notice.

They would have recognized, in real time, that the lecture was not really about the paint. The lecture was a demonstration. This is what I care about. This is what will get you in trouble. Do not bring me anything that is not this.

The signal was being sent and received in both directions. Winterkorn was telling his engineers what would get them punished. The engineers were telling Winterkorn, through their silence about everything else, that the message had been received.

This is the part of the story that every CEO running a fear culture needs to understand, and that most of them will not understand until the crisis arrives. The thing you cannot see from the top of that culture is the second conversation happening underneath the one you are in. The paint lecture has a public version, which is about paint. It also has a private version, which everyone in the room is hearing, which is about everything the paint is not. The CEO usually thinks they are only in the public conversation. The engineers know better.

Fear Works

Bob Lutz was right. Fear works.

It works in the specific, narrow sense that it produces behavior. It can produce short-term results. It produces the appearance of control. For a while, it might produce the commercial success that can be cited as vindication.

What it does not produce is the information flow that organizations need to catch the problems that kill them. The longer a fear culture runs, the more invisible those problems become, and the larger they grow while they are invisible. The commercial success and the hidden catastrophe are not separate phenomena. They are features of the same system.

VW is the clearest case I know, but it is not the only one. The pattern shows up in Boeing's manufacturing problems, where Spirit Aerosystems workers describe being punished for flagging defects. It shows up in the GM ignition switch scandal. It shows up in Wells Fargo, in Theranos, in every financial fraud of the last three decades. The specific details differ. The structure does not. A culture that makes it dangerous to tell the truth upward produces the conditions for catastrophe, and the catastrophe eventually arrives.

Bob saw the results in 2013 and called them success. I saw the results in 2013 and wrote, privately, that I wasn't sure. I had the instinct. I did not have the argument. I have it now.

If you are a leader reading this, the question I would leave you with is not whether your culture has a problem. Every culture has problems. The question is whether you would know if it did. The paint-on-the-Beetle moment in your organization, whatever form it takes, is not telling your people what you think it is telling them. Somebody on your team right now is sitting on something they know you need to hear. The distance between you and what they know is the real measurement of your management system.

You cannot close that distance by asking them to be braver. You can only close it by being someone it is safe to bring hard news to.

Winterkorn, in his Der Spiegel interview, said his engineers should have jumped on his desk. He was describing the correct behavior for an organization other than the one he had built. In the organization he had actually built, the engineer who jumped on his desk would have been fired that afternoon.

That is what fear works at. It works, specifically and reliably, at producing that outcome.