Minimum Efficient Scale, and Why Some Plants Have to Be Enormous

Both approaches to the smallest economically viable plant size work. What differs is what they cost you, and the cost is what this sets out.

What the concept means

Minimum efficient scale is the smallest output at which a plant achieves most of the unit cost advantages available to a larger one. Below that point, fixed costs spread across too few units and the plant is structurally uncompetitive regardless of management quality. Above it, the cost curve is comparatively flat, so a much larger plant gains relatively little further advantage per unit.

The concept explains plant size far better than any argument about ambition or national policy does. It also explains why some industries have hundreds of producers and others have a handful worldwide.

Why some processes demand size

Continuous processes such as refining, smelting, glass and paper require large integrated equipment that cannot be built small economically. Vessel and furnace costs rise roughly with surface area while capacity rises with volume, which favours larger units mathematically. Utilities, environmental control and safety systems have substantial fixed elements regardless of throughput.

The result is that halving intended output does not halve capital cost, and unit cost rises sharply as size falls. This is engineering rather than economics, which is why the constraint is so difficult to argue with.

Where small plants remain viable

Discrete assembly with low capital intensity can be efficient at modest scale, which is why such plants are numerous and dispersed. Products with high transport cost relative to value support local plants because freight offsets the scale disadvantage.

Cement, bottled beverages and structural steel fabrication all show this pattern of many regionally distributed producers. Customisation also favours smaller operations, since flexibility and scale usually pull in opposite directions. The interaction between scale economies and transport cost is what actually determines plant distribution.

What it means for market structure

If minimum efficient scale is large relative to total demand, the market can only support a few producers whatever anyone prefers. New entry then requires adding a chunk of capacity large enough to depress prices for everyone, which deters entry by itself.

Line by line in the tariff schedule, that dynamic produces the concentrated industries that later appear as technical chokepoints in supply chain analysis. The concentration is a consequence of the technology rather than evidence of anything improper.

Policies aiming to fragment such industries have to confront the cost penalty they would impose.

Scale and utilisation are different things

A large plant running at half capacity has worse unit costs than a smaller plant running full, despite its theoretical advantage. Capital-intensive plants therefore prioritise running continuously, which shapes how they price during weak demand.

Once the order book turns, producers facing a downturn frequently accept prices above marginal cost rather than idling equipment with high restart costs. This behaviour is often described as irrational when it is a straightforward consequence of the cost structure. Understanding it explains why capacity in heavy industry is slow to exit even during extended periods of weak prices.

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

How the threshold shifts

Technology can lower minimum efficient scale, as modular and additive processes have done in some product categories. Where that happens, industries can fragment geographically because smaller plants become viable closer to demand. The reverse also occurs when process improvements favour ever-larger units, concentrating production further.

On the manifest, which direction a given industry moves is an empirical question about its specific technology rather than a general trend. Assuming that all manufacturing is becoming either more or less concentrated ignores how differently these forces act by sector.

Side by side

The takeaway

Plant sizes are set by process engineering long before anyone debates industrial strategy.

Capacity takes a decade to build and one quarter to look like a mistake.