Four buyers and some simple numbers show why charging different prices raises profit, the four conditions it needs, and who wins and loses.
On 1 September 2025 ScotRail abolished peak fares for good. The peak return between Edinburgh and Glasgow fell from £32.60 to £16.80, a cut of 48%. An earlier pilot, from October 2023 to September 2024, had done the same thing. Transport Scotland estimated it raised passenger numbers by at most about 6.8%, against the 10% needed for the lost fare revenue to pay for itself.
Why did charging commuters more raise so much revenue in the first place? And why did halving their fares bring in so few extra passengers?
Sources: Transport Scotland pilot evaluation, reported by The Scotsman; ScotRail fares, Railnews (August 2025).
Price discrimination is charging different prices to different consumers for the same good or service, where the difference is not explained by a difference in cost. Before any diagram, learn the conditions. Every model starts from its assumptions.
Then there are three types, or degrees, of price discrimination. First degree is the extreme case, so start there, with numbers.
Same train, same seat, same journey.
Until September 2025, an anytime return from Edinburgh to Glasgow at peak time cost £32.60. The off-peak return was £16.80.
Why would a rail firm charge commuters almost twice as much for the same journey?
Commuters have little choice about when they travel, so their demand is less price elastic. Charging them more raises revenue without losing many passengers. Charging different prices for the same service, where the difference is not explained by cost, is price discrimination.
Four buyers of the same product. Each wants one unit, and each has a different maximum they will pay: A will pay up to £4, B £3, C £2 and D £1. The firm has fixed costs of £3, and each unit costs £1 to make.
With one price for everyone, the best it can do is £3: A and B buy, revenue is £6, cost is £3 + 2 × £1 = £5, so profit is £1. But A would have paid £4, keeping £1 of consumer surplus, and C would pay £2, more than it costs to make, but is priced out.
Check: If the firm charges everyone £3, how many units does it sell?
Only A and B will pay £3 or more. C and D will not buy.
First degree price discrimination: charge each buyer the most they will pay. A pays £4, B £3 and C £2. Revenue is £9 and cost is £6, so profit is £3, three times as much. The firm takes all the consumer surplus, and sells to C as well.
Check: The firm now charges each buyer the most they will pay. Where does the extra £2 of profit come from?
A pays £4 instead of £3, which takes A’s £1 of surplus. C now buys at £2, which adds £1 of profit over the £1 cost. B still pays £3.
Now as a diagram, with many buyers. To keep it clear, cost per unit is constant, so MC = AC is a horizontal line. A single-price monopolist faces demand, AR, with MR falling twice as steeply.
With one price it produces where MC = MR, at Qm, and charges Pm. Consumers keep the triangle above Pm as consumer surplus, and the units between Qm and the point where demand meets MC are never made: a deadweight loss.
First degree: every buyer pays exactly the most they are willing and able to pay. Selling one more unit no longer means cutting the price for everyone, so the demand curve becomes the MR curve. The firm produces up to Q*, where AR = MC. All of the consumer surplus becomes profit, and there is no deadweight loss.
Check: Under first degree price discrimination, why is MR the same as AR?
Each buyer pays their own maximum, so to sell one more unit the firm only charges that buyer less. Earlier buyers still pay their own prices, so the extra revenue is the full price: MR = AR.
Second degree price discrimination: selling off spare capacity. Think of a hotel or an airline. Each extra guest or passenger costs little, so MC is low and flat, until the hotel is full. At capacity, Qcap, it cannot supply any more, so MC becomes vertical.
Check: Look at where MR crosses MC. With one price, how does the hotel’s output compare with Qcap?
MR falls below the low, flat MC before the hotel is full, so profit is maximised short of Qcap and some rooms stay empty.
Maximising profit at MC = MR, the hotel sells Q1 rooms at P1. That leaves rooms empty: spare capacity, which still has to be paid for through fixed costs.
So it sells the empty rooms at the last minute for P2, the price that fills the hotel. Early bookers pay P1; late bookers pay P2. The shaded area is extra revenue that helps cover fixed costs. Late bookers who would not pay P1 gain consumer surplus too. Last-minute flights and hotel deals work this way. Some textbooks use second degree for bulk-buying discounts instead, so if a question says otherwise, follow the question.
Teacher tip: P2 is read straight up from Qcap to demand: the price that just fills the last room. It is not where MR meets MC.
Check: Why can the hotel sell its empty rooms below P1 at the last minute and still gain?
An empty room earns nothing. Filling it at P2 brings in more than the small extra cost of the guest, so profit rises, and the revenue helps cover fixed costs.
Third degree price discrimination: split buyers into groups by their price elasticity of demand. Commuters must reach work at fixed times, so their demand is less elastic. Leisure travellers can choose another time or the car, so theirs is more elastic. Same train, same MC = AC in both markets.
Teacher tip: Draw MC at exactly the same height in both markets: same train, same cost. Only the demand curves differ, with the steeper AR in the less elastic market.
In each market the firm sets MC = MR, then goes up to that market’s AR for the price. Commuters, less elastic, pay Pa. Leisure travellers, more elastic, pay Pb. The less elastic group always pays more.
Check: Which group pays the higher price under third degree price discrimination?
A price rise loses fewer customers in the less elastic market, so the profit-maximising price there is higher.
That is ScotRail’s peak fare. And it explains why scrapping it cost money: a 48% price cut for a less elastic group brought, at most, about 6.8% more passengers, so revenue fell. The same elasticity that made discrimination profitable made removing it expensive.
Now compare third degree price discrimination with one price for everyone.
Check: Compared with a single price, which consumers are made worse off?
The less elastic group now pays more than the single price would have been, so it loses consumer surplus. The more elastic group pays less.
Who wins and who loses? The firm gains profit. Groups with more elastic demand, such as students, pensioners and off-peak travellers, pay less, and some buyers are served who would be priced out by a single price. But groups with less elastic demand pay more, and consumer surplus becomes producer surplus.
They read the price across from MC = MR, so their price equals marginal cost. The price comes from going straight up to market A’s demand curve, ARa.
Before September 2025, ScotRail charged £32.60 for a peak Edinburgh–Glasgow return and £16.80 off-peak. Explain why this is third degree price discrimination and why it raised ScotRail’s revenue.
This is third degree price discrimination because ScotRail charged different groups different prices for the same journey, on the same trains, and the difference was not explained by a difference in cost.
It was possible because the conditions held. ScotRail has market power as the only operator on the route. Peak and off-peak passengers could be kept apart by time, and an off-peak ticket cannot be used on a peak train, so there is no market seepage.
Commuters must reach work at fixed times and have few alternatives, so their demand is less price elastic. Charging them £32.60 raised revenue, because the number travelling fell by proportionately less than the price rose. Leisure travellers can choose when to go, or take the car, so their demand is more elastic, and the lower £16.80 fare attracted passengers who would not pay the peak price.
Total revenue was therefore higher than any single price could achieve. The pilot confirms it: a 48% fare cut raised passenger numbers by at most about 6.8%.
However, the gain depends on how far apart the two groups’ elasticities are. Hybrid working lets more commuters choose when to travel, which makes their demand more elastic, so the extra revenue from a peak fare would shrink.
ScotRail fares and pilot results: Transport Scotland and ScotRail, reported by Railnews (August and September 2025) and The Scotsman. Charts drawn with Rough.js, MIT licensed. Your answers are saved only on this device.