Day 171 — Reading: The Box That Rebuilt World Trade — Containerisation as a System

The Box That Rebuilt World Trade

AFor most of the second half of the twentieth century, economists explaining the rapid growth of world trade looked first to politics. Successive rounds of negotiation under the General Agreement on Tariffs and Trade (GATT) had lowered tariffs, and regional free trade agreements lowered them further. The steel shipping container, a box of corrugated metal with no moving parts, seemed too ordinary to deserve a place in the explanation. It was treated as an engineering detail of interest to port managers rather than as a force capable of reshaping the global economy.

BTo see why that view was mistaken, it helps to look at where the cost of moving goods actually lay. Before containers, general cargo was 'break-bulk': sacks, crates and barrels were carried to the quayside, lifted aboard in small loads and stowed by hand, then unloaded in the same way at the other end. A frequently cited estimate published in 1964 traced the cost of sending a single lorry-load of medicines from Chicago to Nancy in France. The ocean voyage itself, which most people would assume to be the main expense, accounted for less than a quarter of the total. Almost half was consumed in the two ports. Handling was slow, labour-intensive and exposed to theft.

CThe man most closely associated with changing this was not a shipowner but a trucking operator. Malcom McLean, who had built a large road haulage business in the American South, was frustrated by the hours his drivers spent waiting while their loads were transferred to ships. On 26 April 1956 a converted tanker, the Ideal-X, sailed from Newark, New Jersey, to Houston, Texas, carrying 58 metal trailer bodies that had been lifted off their wheels and onto the deck. The cost of loading was put at about 16 cents a ton, compared with roughly $5.86 for loose cargo handled in the traditional way. Boxes for freight were not new: railways and armies had experimented with them for decades. What distinguished McLean was his insistence that the ship, the crane, the lorry and the terminal be designed together as a single system, rather than treating the box as a convenient form of packaging.

DThat system could spread only if other companies' boxes fitted it, and in the early 1960s they did not. McLean's firm, later known as Sea-Land, used 35-foot containers; Matson, a rival on the Pacific routes, used 24-foot boxes. A box that could not be lifted by another company's crane or stacked on another company's ship offered little advantage over a crate. Through the second half of the 1960s, committees of the International Organization for Standardization (ISO) negotiated common lengths and heights and, critically, a common design for the corner fittings by which containers are lifted and locked together. McLean's company, which held patents relevant to these fittings, made them available to the ISO free of royalties. The gesture was less generous than it might appear: a fragmented market of incompatible boxes would have limited the growth of the very trade on which his own business depended.

EOnce boxes could move freely between ships, trains and lorries, the benefits extended well beyond the quayside. Because a sealed container was rarely opened between factory and customer, theft and damage fell sharply; on the route between Europe and Australia, average insurance costs dropped from 24 pence to 4 pence per ton between 1965 and 1970-71. The same journey, which had taken 70 days from the closing of cargo to its availability at the destination, came to take 34. For a manufacturer, those 36 fewer days meant that money was tied up in goods at sea for roughly half as long, a saving that never appeared in any freight rate.

FWhether these gains actually produced more trade was, surprisingly, not tested statistically until recently. The difficulty is that containers spread at the same time as tariffs were falling, and that the 1970s brought a dramatic rise in oil prices which depressed economic activity and concealed changes in total trade. In a study published in 2016, the economists Daniel Bernhofen, Zouheir El-Sahli and Richard Kneller approached the problem from a different angle. Using product-level trade data for 1962 to 1990, they exploited two sources of variation: the year in which each country first acquired port or railway container facilities, and a 1968 classification of which goods were suitable for containers. For trade between industrialised countries, they estimated that the cumulative effect of containerisation over the twenty years following adoption by both partners was roughly 700 per cent, considerably larger than the estimated effect of free trade agreements or of GATT membership. Trade also rose in goods that could not themselves be put in a box, such as finished cars, suggesting that the container's influence spread beyond the goods it actually carried.

GNone of this came without cost. Ports that had employed thousands of dockers needed only a fraction of them once cranes could move a container every few minutes, and in many cities the work moved away altogether, from congested docks to new terminals with space to stack boxes. The economic evidence also has limits. The large estimates apply to trade among wealthy nations; the authors found smaller effects for trade involving developing economies, many of which adopted the technology later. The box was simple. The system that made it valuable was not.