The Atlantic magazine is one of those epic, august publications that survive for decades despite changing societies and new technology.
Founded in 1857 by literary heavyweights including Ralph Waldo Emerson and Henry Wadsworth Longfellow, The Atlantic covers literature, political science, economics and foreign affairs. It’s a little left-of-center, but it’s an intelligent read. Half a million subscribers agree.
However, a recent article by Jerry Useem entitled “What was Volkswagen Thinking?” hit a raw nerve.
A key premise of his piece is that corporations respond to crises with a set of “unwritten scripts” imported from the organization around them. The idea is that there is a standardized analysis in response to serious corporate issues, which acts somewhat like a playbook that managers follow, rightly or wrongly.
Useem argues that over time, the scripts become altered in ways that remove morality from the equation and allow seriously deviant behavior. He cites the classic example: Morton Thiokol’s reversal of an earlier “no launch” call, which resulted in the explosion of the space shuttle Challenger and the loss of seven lives.
He further cites the experience of Ford Motor Company with the infamous Pinto gas tank issue, and uses Johnson & Johnson’s handling of the Tylenol scare as examples of good corporate practice.
In essence, Useem argues that corporate culture stems from the top, and that it is a slippery slope from “don’t see, don’t tell,” down to outright fraud.
Comparing to Volkswagen’s Emission Scandal
Is he right?
Conceptually, maybe so. But comparing the Volkswagen emissions cheating scandal to Johnson & Johnson’s experience with Tylenol, or Morton Thiokol’s with solid rocket booster O-rings, is questionable at best.
Useem is an experienced business writer, but financial types have a tendency to lump all industries together.
Volkswagen, Johnson & Johnson and Morton Thiokol are all manufacturers, yes; but they represent a widely divergent subset of industry.
Morton Thiokol, for example, was never a mass producer of shuttle components. That project was years late, billions of dollars over budget and NASA endured delay after delay to simultaneously develop and produce low volume technologies for the program.
In the Thiokol example, no one under-engineered the O-ring that failed. The failure was in the decision to override the O-ring’s documented performance limitations by cold soaking the part in freezing temperatures before launch.
The Tylenol incident was different. In this case, unanticipated and unprecedented tampering of the product led Johnson & Johnson management to the only logical course of action: a massive product recall and a redesign of the packaging to make it tamper evident.
These actions saved the Tylenol brand and probably billions of market capital for Johnson & Johnson. But the tampering was an external event, not triggered by any active commission or omission in the manufacturing process. It’s easy to take the high road when someone external does it to you.
For Volkswagen, however, things were significantly different.
Internal company engineers created defeat devices that cheated on EPA emissions tests for some of VW’s diesel engine models; it wasn’t vandals, blackmailers or government bureaucrats. In this sense, the VW case is more serious: an overt effort to design fraud into the product from the get-go.
What VW’s Engineers May have been Thinking
For some reason, Useem and others go to the C-suites for answers and responsibility. While it is true that corporate leadership is ultimately responsible for everything that happens from the shop floor to the boardroom, the argument that high-level management was involved in the scandal seems far-fetched to me.
For mass production of consumer goods, it’s a different world. No one produces a new product, then goes looking for a market. In fact, it’s the other way around. The sales and marketing types identify a target market, then apply pressure on the R&D and engineering departments to build the product to fill the need.
At the time the VW common rail diesel engines were being designed, demand for ever-higher fuel efficiency was relentless. Remember, this was a time of very high oil prices, and stratospheric fuel pricing in Europe.
A significant improvement in fuel efficiency without complex after-treatment systems would produce highly marketable vehicles while keeping production costs in line.
The goal was simple: more power, less fuel and lower production cost. Here’s what I think happened to Volkswagen.
A marketing team identified the performance attributes needed to maintain the firm’s strong market share in the compact car segment. Management then charged the engineering departments to develop the necessary technology to hit those performance goals within a time window to suit the new model production changeover and the marketing department’s advertising plan.
Engineering agreed to the project, but deep into the timeline discovered that they simply could not meet all three critical performance parameters of performance, time and cost.
I suspect this is where VW engineering made their mistake. At that point, they needed to delay production of the new engines, perhaps for as much as a full model year, while they ironed out the bugs.
It has been done before, but in the brutally competitive world of low-margin, high-volume auto-making, there is a considerable fear that a competitor will be the first to market at a cost of hundreds of millions of dollars in lost sales.
I suspect that somewhere down the engineering management chain in powertrain development, someone wrote the defeat software code as a stopgap measure to keep the prototype engines up and running, with the intention of tweaking the design later in development.
Once the powerplant team was too deep into the timeline, however, it was impossible to go back and fix the problem. So they did nothing, probably with the intention of addressing the issue at the next design review or possibly for the following model year.
So my theory of the sequence of events in a nutshell is this:
- Engineers write code to get around a difficult design problem in order to keep the project moving forward.
- The design team runs out of time to fix the problem and faces a classic dilemma: delay the program, at a cost of millions, or bury the code to buy time to work on a solution.
- Once the timeline passes the point at which metal is cut, the die is literally cast and the engines are destined for production, no matter what. At this point, the only career saving option for the engineers involved is to say nothing, and just hope they can get away with it.
The notion that this is a vast conspiracy of Volkswagen management in order to cheat the EPA is ludicrous.
If the scheme ever reached the level of upper management, they would have to assume that the plan had passed through enough people that secrecy will be impossible to maintain and would therefore veto the scheme even if they had no ethical issues with it. Market capital means a lot to individuals whose compensation package is more stock options than salary.
Corporate Culture Starts from the Top
It’s not that honesty and corporate responsibility are critical engineering issues that need to be driven home by upper management. The lesson is that manufacturing organizations need to create a culture where failure is possible and acceptable, as part of the development process to make great products.
No matter what was promised in terms of performance and timelines, physics is still physics and if something can’t be achieved, a slip in specifications or development times must be tolerated.
All cutting-edge technologies go through this, from track shoes to space shuttles. Somebody at VW attempted to save their career with a desperate Hail Mary pass. Miracles sometimes happen on the gridiron, but not so often in Wolfsburg.
To read Useem’s original article, click here.