Wednesday the White House released a new National Strategy for Global Supply Chain Security. This is an issue easy to underestimate. Like the plumbing in your house, it tends not to be at the forefront until something goes wrong: leaking, freezing, breaking, bursting, or when the well goes dry. Below is a quick take on context and potential implications.
On June 26, 1974 at a Marsh supermarket in Ohio, a pack of Wrigley’s Juicy Fruit chewing gum became the first retail product sold using a scanner and Universal Product Code symbol. Our lives would never be the same.
The use of the UPC and other “bar codes” allows the supply chain to be digitally monitored, mapped, and managed as never before. Logistics became one aspect of a supply and demand chain.
Working through a private international standards-setting process a variety of bar codes provide a framework that pulls many products, services, and information about them across the network.
In effect, this information moves the supply chain itself.
Increasingly these standards – and the rich information and management resources they make possible – ensure effective, timely, and comparatively friction-free transactions between companies and nations. The ability to share this digital information in very close to real-time has transformed the modern supply chain from supply-push to demand-pull.
Farmers, miners, and fishermen still matter. Processors, truckers, wholesalers, and retailers still play a crucial part. Ports, railways, and highways are still required. Physical stuff of all sorts still has to move from point A to B (and usually on to points C, D, and Z). But at least in the United States, Europe, Pacific Rim and increasingly around the world, the digital signals that are sent along largely determine when and where product arrives.
When this strategic capacity for generating demand-pull information persists, the supply chain is very resilient. Demand-pull is an attractor around which the system self-organizes.
But if the digital stream dries up the supply chain goes blind, deaf, and immobile. This has important – and largely unprecedented – implications for catastrophe preparedness.
Given this strategic context we are clearly beyond the typical approach to Mass Care (ESF-6) or Logistics (ESF-7). The Emergency Support Functions are just as dependent on supply chain resilience as any other aspect of modern life.
Supply chain resilience is a matter of systemic strategic capacity. If the capacity is resilient, local capability will be restored sooner rather than later. If the capacity is lost, it may be very late indeed until local capability is restored. Moreover, as the supply chain has evolved into a scale-free network – a network where relatively few hubs concentrate the preponderance of connections – the supply chain has incorporated the characteristic strengths and weaknesses of such a network.
With particular attention to the weaknesses, Ted Lewis explains:
A scale-free network containing a highly linked hub and many sparsely linked nodes is more organized than a random network… When a highly organized system reaches its critical point, insignificant incidents become significant because they can collapse the entire system… As self-organization increases, typically in the form of larger hubs or nodes with high betweeness properties, the complex system also becomes more vulnerable to targeted attacks and normal accidents. The extreme case of self-organization is SOC (self-organized criticality) – a state in which small changes are likely to create large effects… As a system nears its critical point, consequences grow exponentially in size. As SOC grows, so does the multiplier effect on the overall system. (Lewis, Ted G., Bak’s Sand Pile: Strategies for a Catastrophic World, Agile (2011), Page 340)
For the last twenty years the modern supply chain has generated extraordinary benefits. Efficiencies have multiplied across the global network. But the same networked behaviors that multiply good outcomes can – especially in an unexpected context – multiply bad outcomes.
Surprisingly, optimizations and improved efficiencies have been shown to increase self-organized criticality, making optimized networks less resilient. There are a number of reasons for this non-intuitive result, but one obvious reason is that efficiencies often mean less redundancy and less (expensive) surge capacity.
Most owners and operators of supply chains will find this claim incredible. They will point to several years of data and experience demonstrating increasing stability, effectiveness and predictability. They are right, and it is the system’s predictability that is of particular concern.
The modern supply chain – at its best – is good at predicting market patterns. It recognizes early shifts in consumer behavior, signals these subtle changes across the network, and stimulates appropriate responses in terms of production, transportation, distribution and other functions. Nearly everyone operates on tight deadlines and scramble to meet shareholder profit expectations, but the overall volume is sufficient to keep cash, credit, and goods flowing just ahead of demand. A miniscule profit margin multiplied over millions of consumers is good business.
A big part of the magic is deferring to the expertise of others. Costs – and therefore the consumers’ price – are driven down by abandoning any function where there is not a clear comparative advantage. Increasingly dense network hubs are the result. Searching for comparative advantage has created “cylinders of excellence.”
The producer is very expert in producing, and depends on another expert to transport, and another expert to process, and another expert to package and so on across the network.
No one owns the supply chain. Very few even try to visualize more than their piece of the supply chain, the piece where they have comparative advantage. The level of risk-informed engagement with each piece ranges widely. Risk-informed engagement with other hubs in the network is uncommon.
V.G. Narayanan and Ananth Raman explain,
Most companies don’t worry about the behavior of their partners while building supply chains to deliver goods and services to consumers. Engineers – not psychologists – build supply networks. Every firm behaves in ways that maximize its own interests, but companies assume, wrongly, that when they do so, they also maximize the supply chain’s interests… That finding isn’t shocking when you consider that supply chains extend across several functions and many companies, each of which has its own priorities and goals. Yet all those functions and firms must pull in the same direction to ensure that supply chains deliver goods and services quickly and cost-effectively. Executives tackle intraorganizational problems but overlook cross-company problems because the latter are difficult to detect. They also find it tedious and time-consuming to define roles, responsibilities, and accountability for a string of businesses they don’t manage directly. Besides, coordinating action across firms is tough because organizations have different cultures and companies can’t count on shared beliefs or loyalty to motivate their partners. (Narayanan, V.G. and Raman, Ananth, Aligning Incentives in Supply Chains, Harvard Business Review on Supply Chain Management (2006), Pages 174-175))
The local grocer does not own his or her source of supply. Even the large grocery company does not usually own the trucking firm that delivers product to its loading docks. The distribution center does not own the means of producing what it distributes. The food supply chain is – and most supply chains are – a disaggregated set of functional specialties in tight relationship and mutually dependent, but much more a team of rivals than a true partnership.
The ubiquitous use of outsourcing in search of comparative advantage is widely thought to transfer risk down or across the network. The assumption is suspect.
Outsourcing more likely obscures the actual level of risk. For the reasons outlined by Narayanan and Arnanth the risk is not known, but it almost certainly exists. Wall Street rewards this behavior because it keeps inventory costs low across the supply chain and creates the impression of risk distribution.
Risk distribution can be illusory, especially in scale-free networks. Rather, risk is almost certainly being concentrated in those low-end operators most vulnerable to any major shift in the risk environment.Given the nature of the modern supply chain, this supposedly transferred risk is actually accumulating and will, probably (my judgment), be unleashed across the supply chain by some future event with catastrophic consequences.
The greatest value of the new Strategy is to create a window-of-opportunity before such an event when creative and collaborative attention can be focused on these issues, especially to mitigate systemic vulnerabilities and enhance the resilience of strategic capacity.