26 November 2020
Grenfell Tower
Preventing the gaming of regulations.
By John Watson
It is surprising how little press coverage they have had, because the allegations made at the Grenfell Tower Inquiry that cladding manufacturer Celotex rigged fire safety tests before selling a potentially disastrous product will, if they prove to be true, lift the whole tragedy from the realms of “cock-up” to something much more sinister. They also strike a darkly familiar chord. The current onslaught by the authorities against diesel-powered cars was a reaction to rigged tests on emissions across the industry. It begins to look as if safety tests may have been gamed across the building industry as well.
How wide does all this go? If tests in the building industry and the car industry may have been fudged, what about those elsewhere? We already know that there was a similar culture in the financial services industry with its rigging of LIBOR. What about other areas where testing or adherence to standards are crucial? What about components for airliners? Components for nuclear power plants? What about tests of the safety of vaccines? Had scientists in Russia had last minute concerns about the safety of the Sputnik vaccine, would they have pulled it at the last minute or would the individuals concerned have been told to shut up “for the good of the Motherland”? If you have seen Chernobyl, or indeed if you haven’t, you can guess the answer to that.
It is easy to fit all this into the narrative of your political philosophy of choice. The building industry, the car industry and finance: private enterprise driven to bad practice by the lust for profit. Aha, see what happens in a capitalist economy, comrade! Chernobyl: a state enterprise with no real checks and balances. What do you expect when you allow businesses to answer to politicians? Make your choice according to inclination but either way it is a shallow and self-indulgent approach. The problem is more fundamental. What mechanisms can be used to ensure that products meet appropriate standards, and ideally to ensure also that where those involved in creating them spot defects not dealt with by regulation, those defects are flagged up to be dealt with? It is only when this question has been answered that the second question arises: how can the mechanism be accommodated in the private and state sectors?
Rules, regulations and tests are essential and always have been. I have no idea how one selected medieval weapons but you can bet that when the Plantagenets ordered swords for their retainers they specified that those swords met particular criteria and that the swords were examined and tested to see that they did so. Human nature being what it is, the less successful weapons were probably hidden away. “Honest, Sir Inspector, they are all the same”. “Forsooth Churl, open me that bottom box or by God’s truth I’ll…” etc, etc.
As technology has advanced, tests have become more automated in the field of product approval just as they have elsewhere. Time was that if you went to the doctor he or she would rely heavily on experience in making a diagnosis. That is still in part the case but now much of the work is done in labs where highly sophisticated tests are applied to pathology specimens. What if your car breaks down? Gone are the days when a mechanic will start tweaking the plugs and listening to the engine. Nowadays he connects up a computer which provides him with a full analysis of its performance. So it is with product testing. The techniques become gradually more specialist and fewer of those who apply them really understand the intricacies.
There is an advantage to that, of course. Modern computer-based tests should pick up standard faults with near infallibility. They should also be relatively cheap to operate. There are, however, disadvantages too.
The first is that in any contest between regulators and those who seek to game the regulations, the latter have an inbuilt advantage. That is not just because those who actually work on the development of a project are likely to be technically ahead of those regulating it, although that will almost inevitably be the case, but because regulations and tests are fixed whereas ways of gaming them are not. Unless, therefore, standard regulations and tests have been set with an improbable perfection, the marketplace will find their weaknesses and the more complex and automated the tests the likelier this becomes.
Second there is the unknown unknown. Suppose some weakness is exposed in the development process? Inevitably those involved in that process will hear about it before the regulator does. How do you prevent it being covered up?
The key to both these problems lies in communication. If there was a free flow of information from those developing products to those regulating them, it would be easy both to detect any gaming of the regulations and to ensure that, insofar as possible, they dealt with all aspects that were important. That means giving the manufacturers a big role in the regulatory process, putting them under an obligation to work on the devising and enforcement of regulations with stiff penalties for those who do not carry out this duty bona fide. Then the market intelligence would flow through into regulation which would become much more effective.
Of course there is a fly in the ointment – conflict of interest. Manufacturers might use their regulatory role to favour their own products or discredit those of their competitors. Inevitably there would be a little of this from time to time, but if you were to ask whether I would prefer highly knowledgeable regulators who occasionally misused their position or wholly independent regulators who were behind the curve and whose tests could be gamed, I think I would go for the former.
