The Unbundling: How Production Came Apart

Treat the sections below as a sequence. With the geographic fragmentation of production, getting the early decisions right makes the later ones much easier.

Production used to cluster for a reason

Before cheap long-distance transport, making goods near the people who would consume them was the only economically viable arrangement. Factories also kept their stages close together, because moving partly finished work between sites was expensive, slow and prone to loss.

The classic industrial town, with successive stages of a process within walking distance, was a rational response to those transport costs. Clustering was therefore not a preference but a constraint imposed by the cost of moving physical things over any distance. Every subsequent change in industrial geography can be understood as that constraint loosening in one dimension or another.

The first separation

Falling transport costs allowed production to move away from consumers, so goods could be made wherever production was cheapest and shipped to markets. Steam shipping, railways and later containerisation each reduced the cost and increased the reliability of moving goods over long distances.

Production concentrated in fewer places, since scale economies could now be exploited without giving up access to distant markets. The result was international trade in finished goods, with each country exporting what it made well and importing the rest. This is the pattern that classical trade theory describes, and it dominated for a long period before anything more complex emerged.

The second separation

Separating the stages of production from each other required something different: the ability to coordinate a complex process across distance. That depended on cheap and reliable communication, standardised technical interfaces, and management practices capable of controlling quality remotely.

As those became available, firms could locate each stage wherever it was performed best rather than keeping the whole process together. The consequence was trade in components and intermediate goods growing faster than trade in finished products over an extended period. A product could now cross borders several times before reaching a consumer, which no earlier framework had needed to describe.

What makes fragmentation viable

Splitting a process across locations only pays if the saving from specialising in each stage exceeds the cost of coordinating between them. Coordination cost includes transport, inventory in transit, communication, quality assurance, customs formalities and the risk of interruption. That is why fragmentation appeared first in products with high value density, standardised interfaces and stable specifications.

At port, it appears least in products that are bulky relative to value, require constant design iteration, or depend on rapid response to demand.

The framing is useful because it predicts where fragmentation will retreat as well as where it will spread.

The regional character of the result

Despite the language of global supply chains, much fragmented production is organised regionally rather than across the whole world. Components move within a region among nearby countries, with the finished product then exported to more distant markets.

Proximity keeps coordination cost low, and trade agreements with cumulation rules reinforce the pattern by rewarding regional sourcing. Recognising the regional structure explains why disruption in one region propagates so strongly within it and more weakly beyond it. It also explains why diversification strategies frequently mean adding a country within the same region rather than a different continent.

Tariff schedules are technical documents, and classification disputes turn on wording rather than intent.

What could reverse it

Anything that raises coordination cost relative to specialisation gains pushes production back towards fewer locations. Higher transport variability, greater policy uncertainty and increased inventory requirements all work in that direction. Automation cuts the other way for some products by reducing the labour cost advantage that justified separating stages in the first place.

Over a shipping cycle, the net effect differs by industry, which is why evidence of both consolidation and continued fragmentation appears at the same time. Expecting a single direction for manufacturing as a whole ignores how differently these forces act across products.

The takeaway

Fragmentation is a balance between specialisation gains and coordination costs. Watch the second, because it is the one that changes.

Somebody pays the tariff. The argument is only ever about who.