Theme 3 · 3.3 Revenues, costs and profits
Revenue, costs and profits
Why does Ryanair sell seats for €9.99? Why does Ferrari refuse to build more cars when buyers wait two years for one? Why could Netflix raise prices by 8% and still add 13 million subscribers in a quarter? Three pricing puzzles, one toolkit: the relationships between total, average and marginal revenue, and the cost curves that sit beneath them.
The big picture
This is the machinery underneath the whole of Theme 3. Revenue maximisation means MR = 0; profit maximisation means MC = MR; economies of scale are a movement along the long-run average cost curve; every market structure diagram you will ever draw is assembled from the curves on this page. Learn the relationships as relationships, since the exam tests whether you can say why MR lies below AR and why MC cuts AC at its minimum, and firms give the theory its bite: total revenue is price times subscribers for Netflix, daily users times revenue per user for Meta, occupancy times beds times fees for a care home. Writing TR in the industry's own units is the fastest route from generic theory to applied analysis.
The theory
Revenue and the elasticity link
Total revenue (TR) is price times quantity. Average revenue (AR) is TR divided by output, which equals price, so the AR curve is the firm's demand curve. A firm with market power faces a downward-sloping demand curve rather than the horizontal demand curve of a price taker, and must cut price to sell more. That is why marginal revenue (MR), the revenue from one extra unit, lies below AR: the price cut applies to every unit sold, so the extra unit brings in less than its own price. For a straight-line demand curve, MR falls twice as steeply as AR; this derivation is standard theory rather than anything you need to reproduce, and what matters is the picture it produces below.
Where demand is price elastic, MR is positive and a price cut raises TR; where demand is price inelastic, MR is negative and a price rise raises TR; TR is at its maximum where MR = 0, at unit elasticity. A firm that knows its PED knows which direction to move price.
Elasticity is a moving target, and the best answers treat it as one. Brand loyalty, patents and customer inertia make demand less elastic; the arrival of close substitutes makes it more elastic; and Ferrari's demand becomes more elastic the more cars it builds, because exclusivity is part of the product. Ask what is happening to PED over time and you usually find the evaluation the question wanted.
Costs in the short run
In the short run at least one factor is fixed, so costs split into fixed costs, which do not vary with output, and variable costs, which do. Average cost is total cost per unit; marginal cost is the cost of one extra unit, and it eventually rises because adding more of the variable factor to a fixed one yields less and less extra output. The relationship between the two is pure arithmetic: whenever the marginal unit costs less than the average, it drags the average down; whenever it costs more, it pulls the average up. So MC must cut AC exactly at AC's minimum point, a fact examiners test directly and the diagram below fixes in memory.
Costs in the long run, and profit
In the long run all factors are variable and the firm moves along its LRAC curve. Falling LRAC reflects internal economies of scale: purchasing (bulk-buying power), technical (bigger and better capital), financial (cheaper borrowing) and managerial (specialist managers). External economies arrive from outside the firm as a whole industry grows in one place, through shared suppliers and pooled skilled labour. Past the minimum efficient scale, communication and coordination failures create diseconomies of scale, covered with the LRAC diagram on the business growth page.
Profit language must be exact. Normal profit is the minimum return needed to keep the entrepreneur's resources in this industry, and it is counted as a cost, inside AC. Supernormal profit is anything above that, earned whenever AR exceeds AC at the chosen output; a loss is the reverse. A private firm must earn at least normal profit in the long run to remain in the industry. In the short run the standard result is that a loss-making firm will keep producing while revenue covers its variable costs, because fixed costs are owed either way; treat that as a brief, careful point rather than a centrepiece.
Application
The elasticity link earns its marks through cases. Netflix raised prices by 8% in Q4 2023 and still added 13.1m subscribers: demand was price inelastic, MR was still positive, and TR rose, which is the top-right corner of the AR/MR diagram in action. Pfizer shows what a patent does to PED: its Covid vaccine moved from $19.50 to $130 a dose against a marginal cost of $1-3, part of $56bn of revenue across 2021-22, because a patent leaves buyers no substitutes. Then atorvastatin shows the flip: after Lipitor's patent expired in 2011, 30+ generic entrants drove the price from $3.50 to $0.10 a pill within 18 months, and surviving firms defended TR through volume once demand turned elastic. One drug, both halves of the elasticity story, which makes pharma the most efficient application you can learn here.
On the cost side, Aldi against Tesco is the live contrast: roughly 1,500 product lines against Tesco's 25,000, around 90% own-label, and margins of 5-7% against Tesco's 2-4%, because a stripped-down range maximises throughput per store and holds average cost down. Spotify is the warning about assuming scale cures everything: with 600m+ users and €13.2bn of revenue it still ran persistent losses, since royalties above 70% of revenue are a variable cost that grows with every stream, so average cost barely falls as output rises. Scale only delivers supernormal profit when the cost structure is heavy with fixed costs; that single sentence, applied, is worth an evaluation paragraph in almost any costs question.
Building the paragraph
Because a patent makes demand price inelastic, a pharmaceutical firm can price far above marginal cost and raise total revenue by raising price, as Pfizer's move from $19.50 to $130 a dose shows. The chain spends evidence at every step:
- A patent legally excludes substitutes, so buyers who need the treatment cannot switch and the firm's demand curve becomes steeply price inelastic.
- With PED low, a price rise loses few sales: Pfizer lifted its vaccine price from $19.50 towards $130 a dose while volumes held.
- Marginal cost stayed at roughly $1-3 a dose, so the gap between price and marginal cost widened with every increase.
- On the AR/MR diagram, the firm is operating on the inelastic-facing stretch it can exploit: raising price moves it up a steep AR curve while TR rises, the opposite of what a firm facing elastic demand could do.
- Because AR sits far above AC at this output, the margin is supernormal profit, and it summed to $56bn of revenue across 2021-22.
- So revenue and profit move together here: the patent converts pricing power into supernormal profit for as long as the exclusion lasts, which answers the question for the patent-protected case.
Evaluation
Evaluating this, the chain is conditional on the patent: the pricing power it describes exists only while substitutes are excluded, so the analysis carries its own expiry date. The evidence is atorvastatin, where the 2011 patent expiry let 30+ generics drive the price from $3.50 to $0.10 a pill within 18 months as demand turned elastic. Cross-examine the claim as well: AstraZeneca priced its Covid vaccine at cost, roughly $3-4 a dose, so inelastic demand makes high prices possible without making them inevitable, and objectives intervene between elasticity and price. The bounded conclusion is that the elasticity-revenue link reliably predicts the direction a price move takes TR, while what the firm does with that power depends on its objectives and on how long the inelasticity lasts.
Key terms
| Term | Precise definition | Where it earns marks |
|---|---|---|
| Average revenue | Total revenue divided by output, equal to price; the AR curve is the firm's demand curve. | The identity that lets you read price off the demand curve in every firm diagram. |
| Marginal revenue | The change in total revenue from selling one extra unit; below AR for a price-making firm. | Locating profit maximisation (MC = MR) and revenue maximisation (MR = 0), and the elasticity link. |
| Marginal cost | The change in total cost from producing one extra unit. | The MC = MR condition, and explaining why MC cuts AC at its minimum. |
| Normal profit | The minimum return required to keep the entrepreneur's resources in their current use, treated as a cost within average cost. | Defining supernormal profit correctly, and long-run entry and exit arguments. |
| Supernormal profit | Profit above normal profit, earned where AR exceeds AC at the chosen output. | Shading the profit rectangle, and dynamic-efficiency chains about reinvestment. |
| Economies of scale | Falls in long-run average cost as output rises; internally purchasing, technical, financial and managerial. | Any question linking growth or takeovers to costs; the LRAC movement must be named and explained. |
Towards the exam
A drug company's patent on its best-selling treatment expires next year, and several generic manufacturers plan to enter. Explain, using the relationship between price elasticity of demand and revenue, why the firm's pricing power is likely to weaken. Two linked steps through PED and MR, with the AR/MR diagram, earn this quickly.
A discount grocer stocks a fraction of the product lines of its larger rivals and sells mostly own-label goods, yet earns higher margins. Evaluate the view that low average costs are a more reliable source of profit than high prices. Refer to an industry of your choice. Let the evaluation turn on cost structure: fixed-cost-heavy firms scale differently from variable-cost-heavy ones.
Explain two internal economies of scale available to a growing airline, and illustrate their effect on a long-run average cost diagram. Name the types precisely and explain the movement along LRAC in the airline's context.