VW had lots of ‘corporate governance’ and ‘oversight’ mechanisms so how could this happen? Like typical German public companies, VW has a two tier supervisory board of workers and management representatives. Yes, there is a powerful family holding, but the state government is a big investor. It has mapped an anti-corruption system with a pathway for internal whistle-blowers. No shortage of stakeholder engagement, then.
With any big corporate scandal you cast your mind back to the banking sector where, yes, incentives are misaligned, and it is systemically too big. But it is a sector where a few individuals could get together on a conference call, send a few emails, down a few beers and come up with a neat strategy to manipulate LIBOR. OK, and complicated subprime mortgage products? Some clever physicists crunch some numbers into new risk based financial products and get some lawyers to package them. Ultimately it’s casino capitalism: we may not like it, or have the courage or moral fibre to cut our financial sector down to a socially sustainable size – but we know what we are dealing with, even if we can do little about it.
But it is hard to see the VW ‘scandal’ in quite the same light. How could it have been the work of one or a few bad apples? VW is a vast, global multi-site manufacturing company making and shifting millions of real cars (and most car manufacturers are computing experts now). There are ingrained processes and steps to design and test product modifications. And budgets. And components to be made. And more tests.
Think about it.
(1) There would have been a suggestion to specify and design the software.
(2) Someone would have ensure the US emissions standards were taken into account in the software specification
(3) The software design would have gone through an approval process, and a benefits case
(4) Someone would have given it a budget
(5) The software would have been formally specified and then built according to an appropriate methodology, and tested in simulated environments
(6) As far the real cars, any new components, changed components or tweaks to how things fitted together, adjusted components or changes to components would have been specified
(7) Prototypes would have been made and tested
(6) The software and components (if any) would have been extensively tested according to two criteria: any impact on the road safety (if software or mechanism goes rogue) as well as ensuring that it worked
(8) Then it would have been incorporated into the manufacturing process, with subsequent quality testing
(7) Almost certainly, the US test conditions would have been simulated and the mechanism and software tested to ensure it met its objective.
This is a complex process involving software design, costing, manufacturing, testing and quality control, almost certainly cutting across departmental silos. Manufacturing any car, let alone millions, to consistent standards of quality is a feat of organisation and process engineering involving many, many people from many professional disciplines, not a few whizz kids playing poker with the world as dice. A very different sort of corporate malfeasance but embedded, like parasite, in the very systems and processes that produce safe, technically perfect, products. How did it happen?
Edmund is the Director of Distance Learning at Leeds Beckett University.