Microeconomics › Section G · Production, costs, and industry structure › Topic G.4

Production in the long run

The short run's bottleneck was the fixed input. In the long run there is no bottleneck: every input can grow, shrink or vanish, the firm can switch technologies entirely - and a different law, about scaling everything at once, takes over from diminishing returns.

The long run: everything is a choice

In the long run, all inputs are variable. The firm can build new capacity, close old plants, change methods completely, or exit the business altogether. 'Long run' is not a fixed stretch of calendar time: it is whatever it takes for the binding commitment to expire - a one-year factory lease makes the long run anything beyond a year, while a food truck's long run might be a season. Because no input is frozen in place, the long-run production function records the best available input combination for every output the firm might choose.

Removing the bottleneck

Recall what wrecked short-run production: marginal product collapsed because capital was fixed - workers queued for one oven. The long run dissolves exactly that constraint. Demand triples? Add ovens and bakers together, keeping each baker equipped. The short-run response is congestion; the long-run response is expansion in proportion. This is the whole practical difference between the two horizons: the short run asks how to work the fixed plant harder, the long run asks what plant to have.

Returns to scale: a different law

With all inputs free, the interesting question becomes: what happens to output when every input scales together? Doubling ALL inputs - labour, capital, everything - yields three possible outcomes. More than double the output: increasing returns to scale. Exactly double: constant returns to scale. Less than double: decreasing returns to scale. Note this is a different species from diminishing marginal returns: that law changed ONE input while something stayed fixed; returns to scale changes ALL inputs at once. One is about the margin, the other about the proportion - the two can disagree in the same firm at the same time.

Where each pattern comes from

Increasing returns to scale grow from scale itself. Specialisation pays more at large volume: a thousand workers can divide tasks in ways ten cannot. Indivisible inputs spread: a blast furnace, a nationwide logistics network, a single accounting system - each has a sizeable cost that tiny output cannot dilute. Constant returns are the replication case: open an identical second restaurant, a third production line, and output doubles when inputs do. Decreasing returns come from the cost of bigness: layers of management, slow coordination, meetings about meetings, monitoring that consumes more of everyone's time as the organisation grows. Firms typically see increasing returns first, then constant, then decreasing - the long-run average cost U of the next topic traces the same arc.

Worked example: one oven, then three

A bakery's short run, with one oven (K = 1): total product for one to four bakers is \(40, 70, 85, 90\) loaves an hour, so marginal product runs \(40, 30, 15, 5\) - diminishing returns bite hard as bakers queue for the single oven. Now the long run: the bakery buys two more ovens (K = 3). The table becomes \(40, 80, 120, 140\), with marginal product \(40, 40, 40, 20\). 1. Compare: with equipment restored, the second and third bakers add 40 loaves each instead of 30 and 15. 2. Congestion, not the workers, was the short run's problem - three ovens let three bakers work full speed. 3. Check the pattern: diminishing returns still arrive eventually even at K = 3 (the fourth baker's MP is 20) - more capital delays the bottleneck, it does not repeal the law. 4. Conclusion: the long run does not abolish diminishing returns; it moves the plant size the returns are measured against. Assumptions: identical bakers, same technique.

Worked example: classifying three firms' scaling

Three firms each double all their inputs. Firm A's output rises from 100 to 220 units. Firm B's rises from 100 to 200. Firm C's rises from 100 to 170. 1. Firm A: output multiplied by \(220 \div 100 = 2.2\) - more than double the inputs bought: increasing returns to scale. Likely enjoying deep specialisation or indivisible assets. 2. Firm B: \(200 \div 100 = 2.0\) - exactly in step with its inputs: constant returns, the replication case. 3. Firm C: \(170 \div 100 = 1.7\) - output lagging the doubled inputs: decreasing returns, the smell of coordination costs. 4. Check: only the ratio of output growth to input growth matters - each firm doubled inputs, so the verdicts read straight off the multipliers. Assumptions: 'all inputs doubled' includes management time and everything else, not just the visible machines.

Technology, in the economist's sense

In the long run the firm also chooses among production technologies - and here 'technology' means every alternative method of combining inputs, not the latest invention. A courier firm can combine bike couriers and vans in many proportions; a farm can be labour-intensive or machinery-intensive. The firm's long-run task is to find, for each output level, the input combination that produces it at least cost. This is why the long-run production function is a frontier of best choices rather than a single recipe: it records, for every output, the most efficient combination available with known methods.

Returns to scale and tomorrow's cost curve

The next topic converts this one into money: long-run costs. The bridge is direct. With input prices given, increasing returns to scale means output outruns input costs, so cost per unit falls as the firm grows: economies of scale. Constant returns mean unit costs hold steady as the plant replicates. Decreasing returns mean unit costs rise with size: diseconomies of scale. So when the long-run average cost curve is drawn U-shaped next topic, its falling stretch will be today's increasing returns wearing a price tag, and its rising stretch will be today's coordination costs.

Common mistakes to avoid

First, and most common: confusing diminishing marginal returns with decreasing returns to scale - one varies a single input, the other all of them together; they are different laws answering different questions. Second, reading 'long run' as a number of years - it is the horizon at which every commitment expires, whatever the calendar says. Third, assuming capital is always the fixed input: in service firms, scarce licences or key staff can be the binding commitment. Fourth, treating a scale verdict as permanent: firms pass through regions - increasing at small scale, constant in the middle, decreasing at giant scale. Fifth, forgetting the technology definition: not inventions, but any method of combining inputs.

Key takeaways

The long run is the horizon at which all inputs are variable and no cost is fixed - so the production function records the best input mix for each output, and the firm chooses among technologies (methods, not gadgets). Removing the fixed input dissolves short-run congestion: with three ovens, the bakery's bakers add 40, 40, 40, 20 instead of 40, 30, 15, 5 - diminishing returns delayed, not repealed. Returns to scale measure what happens when ALL inputs scale together: more than double the output is increasing returns (specialisation, indivisible assets), exactly double is constant (replication), less is decreasing (coordination costs) - the three firms multiplied output by 2.2, 2.0 and 1.7 on doubled inputs. Next topic converts the pattern into long-run unit costs: increasing returns become economies of scale, decreasing become diseconomies.

Flashcards

Flashcard

Long run (production)

The horizon at which every input is variable

Defined by expiring commitments, not calendars.

Test your understanding

Question 1 of 5

A firm doubles ALL its inputs and output rises from 100 to 220 units. What is this called?

Question 2 of 5

With one oven, a bakery's bakers have marginal products of 40, 30, 15 and 5. With three ovens, the same bakers have marginal products of 40, 40, 40 and 20. What does the comparison show?

Question 3 of 5

What distinguishes diminishing marginal returns from decreasing returns to scale?

Question 4 of 5

Which is a source of DECREASING returns to scale?

Question 5 of 5

In the long run, what does a firm's production function show?

← Previous topicNext topic →