Contaminated food products can pose particular threats to consumer health and well-being, as well as impose significant financial and reputation costs on firms.
Such recalls often involve a tainted ingredient – peanuts, eggs or beef, for example – which can then spread quickly through complex food supply chains, infecting many other fresh, frozen or prepared food products. So speed in spreading the word and getting contaminated food products out of distribution and off supermarket shelves is critical.
A study co-authored by Benn Lawson, University Senior Lecturer in Operations Management at Cambridge Judge Business School, looks at firm response time in such food-product recalls. Based on US Food and Drug Administration (FDA) data from 2004 to 2013, the study looks at agri-food recalls involving 407 pairs of suppliers and affected downstream manufacturing firms.
The conclusion: firm response time is lengthened by geographic distance between the supplier and manufacturer, but shortened when the supplier and affected company operate in related, rather than diverse, industry sectors. The research finds further that response time deteriorates as more firms in a given industry are affected by a recall, perhaps due to competitive pressures.
The findings pose important implications for suppliers, manufacturers and regulators in seeking to minimise the fallout from future food recalls. The study recommends that supply chain resilience be strengthened through better tracking and traceability systems, and more extensive quality audits.
The study published in the International Journal of Operations and Production Management – entitled “Supply chain disruptions: the influence of industry and geography on firm reaction speed” – is co-authored by Dr Benn Lawson of Cambridge Judge Business School, Dr Antony Potter of the University of Manchester, Professor Frits Pil of the University of Pittsburgh, and Professor Matthias Holweg of the University of Oxford.
Co-author Benn Lawson of Cambridge Judge Business School comments on some of the key findings and implications of the study:
The study is based on what we call “time to recall”. That’s the downstream supplier’s public reaction time – measured in the number of days it takes a food manufacturer to issue its own recall following the first opportunity at which it could learn about the problem, which is the supplier’s recall announcement. We then looked deeper at the effect of geography and industry on a firm’s response time.
The damage caused by food recalls is hardly peanuts. In fact, one of the most high-profile cases involved Georgia-based peanut processor Peanut Corporation of America (PCA) and Salmonella contamination in 2009: it was the largest agri-food recall in US history, linked with nine deaths and $1 billion in losses. Geography clearly played a role in the slowness of affected downstream firms to act: a company in Washington State more than 2,000 miles away was a customer of PCA, but it took more than a month to clear all affected peanuts from its processing lines and issue a recall. That company subsequently moved back to a local supplier in Washington State to keep a closer watch on everything.
Distance boosts pipeline inventory, slowing response time. When suppliers and manufacturers are far apart, firms tend to hold larger stores of inventory as insurance against transport delays. This makes it more difficult to identify tainted goods because there is simply more product to inspect, and because the affected goods are more widely dispersed among transport networks and storage points.
We expected that a geographic cluster of affected firms would reduce recall times, but there was no effect. The assumption was that a number of affected firms located in the same metropolitan area would increase information flow and improve recall times, but that didn’t occur. That might be due to competition among the affected firms for limited local resources to enable the recall, such as storage and transport capacity, or a reluctance to disclose firm-specific information to investors due to concern about the share price. We conclude that government penalties and inducement to share information may be required.
Prior experience with recalls improves subsequent recall response time. This poses an obvious problem: how can firms gain such “recall management expertise” without actually going through a costly and damaging recall? Simulation models and pilot tests of a firm’s internal processes may be a good way to prepare for future disruption, and thus quicken response times in the event of a recall – because while unpredictable in terms of timing, product recalls are an expected form of operational disruption.