This week’s New Scientist cover story is about the “Collapse of Civilisation“, and it’s well worth reading – as if the sub-prime banking crisis isn’t enough, civilisation is only two square meals away from collapse at any given moment.
I’ve been reading Jared Diamond’s Collapse, which explores the reasons why a number of formerly successful civilisations disappeared. He focusses primarily on ecological factors as he’s mainly concerned with ancient societies – the New Scientist article looks at modern Western civilisation, but their conclusions are equally gloomy – basically, modern civilisation is amazingly vulnerable.
It doesn’t take long to figure out why, and it took me right back to playing around with supply chain models during my MBA. Building reliable systems is about removing single points of failure, because failure probability is usually multiplicative. Put crudely, doubling the components halves the risk of failure. So coming from a background of designing networked systems, my instinctive reaction is double-up critical equipment and build in backup paths and redundancy. And where consumable items are a factor, the response is to hold buffer stocks.
But that’s a problem from a supply chain point of view, because duplication and buffering are adding in redundancy and inventory – both of which have a cost associated with them. I remember reading somewhere that the scheduling systems for components at Nissan’s Sunderland plant have to take into account the traffic conditions on the A1, because if there’s a jam it delays the trucks bringing components from the seat factory up the road. That’s apocryphal, but entirely believable.
So rather than a supply chain, we end up with a single-point-of-failure chain.
When you start to look at where the single points of failure lie, it starts to get scary – the average supermarket would be cleaned out within 3 days, so if there’s noone to drive the trucks with replenishments, the food supply is disrupted very quickly. The joke about the Little Chef restaurant being unable to serve an omlette because they haven’t been delivered suddenly isn’t quite as funny anymore.
Personally, I’ve always thought that the point at which to start worrying about the impact of a pandemic is when the schools are closed – because at that point, it takes parents out of the workforce, and most organisations would collapse as a result.
Treating supply chains as a network problem and reducing the single points of failure would seem like an obvious insurance policy, but that’s difficult to do when the financial systems are biased against this kind of thing. Simplistically, shareholder capitalism is about maximising short-term shareholder returns through minimising costs – and a just-in-time inventory system with little or no buffer stock is a very efficient way of cutting cost out of operations. So while operating “fatter” might be a hedge against disruption, it’s discouraged in the short-term, because disruption is only a future probability, while the need to cover this year’s dividend is a certainty.
Maybe capitalism will be our downfall after all?
