Supply, demand and the cost curve: Supply: the cost curve, the marginal producer and rent

3.1 The cost curve ranks every producer, and the measure of cost matters

Nobody can observe a mine's marginal cost, but consultants can collect each mine's capacity and unit cost. Rank the mines from lowest to highest unit cost, draw each as a bar whose width is capacity and whose height is cost, and you have the comparative cost curve. It runs fairly flat through the bulk of the industry, steepens through the high-cost tail and turns vertical where capacity ends.

The height of each bar depends on the cost measure. Tilton and Guzmán set out the convention. C1, or cash cost, covers mining, milling, smelting, refining, freight and marketing, plus taxes not tied to profit: roughly average variable cost. It is usually reported net of by-product credits, which is why a copper mine rich in gold or molybdenum can show a negative cash cost. C2 adds depreciation. C3 adds interest and indirect costs such as overheads, exploration and royalties, and approximates average total, or full, cost.

The gold industry added its own measures. Quoting cash costs alone had made gold miners look more profitable than they were, inviting higher taxes during the boom, so in 2013 the World Gold Council proposed all-in sustaining and all-in costs, which add sustaining capital, exploration and corporate costs (Humphreys).

C1 shows who comes under pressure first when prices fall, since a mine covering its cash costs is still better off running. Full cost, extended to potential new mines, approximates the long-run supply curve: the price at which new capacity gets built.

• Tilton & Guzmán (2016), Mineral Economics and Policy · pdf pp. 86-91
• Humphreys (2015), The Remaking of the Mining Industry · pdf pp. 108-109, 182
• Radetzki & Wårell (2021), A Handbook of Primary Commodities · pdf pp. 106-107, 180-181

3.2 The marginal producer anchors the price, but the floor is soft

In a competitive market the short-run price settles where demand crosses the cost curve. The mine at that point is the marginal producer; it just covers its cash costs, and everyone to its left earns a margin. In the long run the anchor is the full cost of the dearest new project the market needs, since lower prices stop investment.

Humphreys calls working out who the marginal producer will be the key to future pricing. His case is seaborne iron ore after the boom. High prices had drawn in costly supply from China, India and small exporters such as Mexico, Malaysia and Iran, while Australian and Brazilian majors lined up low-cost expansions. BHP Billiton expected about 400 million tonnes a year of new seaborne capacity in 2014 to 2016, roughly 25 per cent of 2013 supply. That pushed the whole curve to the right. Some 200 million tonnes of high-cost Chinese output was likely to be displaced, and the price would at best settle at the costs of the dearer Australian and Brazilian mines. The majors did not cut output to help; they wanted back the share lost when they could not expand fast enough.

But the cost curve is a soft floor. When prices fall, managers gain leverage over suppliers, workers and governments, push through productivity gains and cut slack built up in good years, so the curve itself drops. In the gluts of the 1980s and 1990s, Humphreys notes, cost curves fell further and stayed down longer than the industry expected.

• Radetzki & Wårell (2021), A Handbook of Primary Commodities · pdf pp. 102-105, 231-232
• Humphreys (2015), The Remaking of the Mining Industry · pdf pp. 228-230
• Tilton & Guzmán (2016), Mineral Economics and Policy · pdf p. 104

3.3 Economic rent is the surplus a low-cost producer earns

Economic rent is income that could be taxed away without changing behaviour. Tilton and Guzmán's footballer could earn $1,000,000 a year playing or $200,000 as a mineral economist; if pay alone matters, $800,000 of his salary is rent. For an operating mine, rent per tonne is the price minus average variable cost. The marginal mine earns none and the cheapest earns most per tonne, though a big mid-cost mine may earn more in total.

They split rent by source. Pure rent comes from a better deposit: higher grade, larger size, less overburden, easier processing or a better location. Monopoly rent comes from market power. Rent due to short-run supply constraints arises when demand outruns capacity and prices sit above long-run cost until new mines arrive. Quasi-rent exists because capital is sunk: existing mines need only cover variable cost while new ones need full cost, so existing mines normally earn roughly their average fixed cost. Policy can add or subtract rent through taxes and regulation.

Monopoly rent lasts as long as market power does. Of the rest, only pure and policy rents survive into the long run, and even taxing those has a cost. Existing mines stay open, but the reward for finding new deposits shrinks, so output falls decades later.

Rent is what governments, unions and communities bargain over. Taxes, required local processing and generous wages all raise a mine's costs and transfer rent. Tilton and Guzmán note that US copper unions won wages well above comparable jobs in the 1950s and 1960s, then accepted sharp cuts in the 1980s to keep the mines alive.

• Tilton & Guzmán (2016), Mineral Economics and Policy · pdf pp. 132-136
• Humphreys (2015), The Remaking of the Mining Industry · pdf pp. 181-182

3.4 Why output keeps flowing when prices fall below cost

The textbook firm shuts when price falls below average variable cost. Mines often do not, and Radetzki and Wårell list why. Closing and reopening cost money, so riding out a short dip can be cheaper. In the very short run labour and maintenance cannot be shed, so the true variable cost is below the reported figure. Governments cut royalties, devalue, subsidise or lean on firms to protect jobs. State firms and cooperatives may not maximise profit at all; a state miner might even raise output when prices fall to protect export earnings, Tilton and Guzmán note. Expectations matter too: Codelco in 2002 and 2003 stockpiled copper rather than cut output, selling a year or so later at much higher prices.

Humphreys adds a further reason, rooted in each firm's own arithmetic. In a capital-intensive business, extra tonnes spread fixed costs and cut unit costs, so every producer pushes volume in a glut, and together they deepen it. Between 1993 and 2001, years of weak copper prices, world mine output beat forecasts in six of nine years.

Costs also follow prices. Tilton and Guzmán show the world copper cost curve rising in the late 1980s, falling through the 1990s and climbing again after 2000. High prices relax discipline and bring poorer ore into the mine plan; low prices, in Radetzki and Wårell's phrase, 'work wonders' on costs. The real supply curve is therefore a broad, shifting band rather than a neat line.

• Radetzki & Wårell (2021), A Handbook of Primary Commodities · pdf pp. 107-109
• Tilton & Guzmán (2016), Mineral Economics and Policy · pdf pp. 68-71, 89-91, 100
• Humphreys (2015), The Remaking of the Mining Industry · pdf p. 230

See it in the table

Check your understanding

A copper mine reports a negative C1 cash cost. What is the most likely explanation?

Which kind of rent disappears once new capacity has time to come on stream?

Why did world copper mine output beat forecasts in six of nine weak-price years from 1993 to 2001?