Invention, innovation and diffusion; process and product innovation; a new technology lowering the average cost curve, the innovator’s temporary supernormal profit, imitation driving the price down and forcing old-technology firms out; Schumpeter’s creative destruction, dynamic efficiency and the effect on market structure. Through the switch from diesel to electric cars.
On 5 November 2026 the Society of Motor Manufacturers and Traders reported October’s new car registrations: 152,730 cars, of which 48,858, or 32.0%, were battery electric, the highest share of the year and up 32.7% on October 2025. Petrol registrations fell 11.6% to 56,876 and diesel fell 15.9% to 5,811, a 3.8% share for a fuel that took close to half the market a decade ago. Over the first ten months of 2026 battery electric cars took 26.6% of the market, against 23.4% in 2025, and buyers could choose from more than 180 models.
Behind those numbers is a change in how cars are made and what they cost to run. An electric drivetrain has far fewer moving parts than a diesel engine and gearbox; battery costs have fallen for a decade; new entrants without engine plants can compete with firms that spent a century perfecting them. The firms, suppliers and skills built around the old technology lose value as the new one spreads. Joseph Schumpeter called this process creative destruction, and argued it, not price competition, is what drives growth in a capitalist economy. The diagram in this lesson shows the mechanism one cost curve at a time.
Source: SMMT, new car registrations, October 2026, published 5 November 2026; SMMT market outlook, November 2026.
Technological change is the development of new products and new methods of production. Invention is the idea; innovation is its first commercial use; diffusion is its spread across firms and consumers. Process innovation changes how goods are made, lowering costs and raising productivity: the moving assembly line, containerisation, software. Product innovation creates new goods that compete with or replace old ones: the electric car, streaming, the smartphone.
Creative destruction (Schumpeter, 1942) is the process by which new technologies, products and firms displace old ones: the creation of new value and the destruction of old capital, jobs and market power at the same time. It raises dynamic efficiency, the improvement of products and processes over time, and it changes market structure: lowering barriers to entry where it makes old assets obsolete, raising them where it creates economies of scale or network effects. The gains flow to consumers through lower prices and better products, and to innovators through temporary supernormal profit; the losses fall on owners and workers of the old technology.
In October 2026 battery electric cars took 32% of UK new car registrations. Diesel, which took close to half the market a decade ago, took 3.8%.
Petrol registrations fell 11.6% on the year and diesel 15.9%, while battery electric rose 32.7% to 48,858 cars. More than 180 electric models are now on sale, and 503,805 were registered in the first ten months of the year, a 26.6% share against 23.4% for all of 2025.
What does a new technology do to the firms, factories and jobs built around the old one?
The gains go to consumers and to the innovators; the losses fall on whoever owns the old capital and skills. This lesson draws that on a cost diagram: the innovator’s profit, the fall in price as the technology spreads, and the exit of firms that cannot follow.
Start with the old technology. Each firm’s average cost curve AC1 bottoms out at £20 at 50,000 units. In a competitive market entry and exit have pushed price to that minimum: P1 = £20, and each firm earns normal profit at E1. This is the long-run equilibrium every textbook draws. It is also the position that a new technology destroys.
Check: With the old technology, average cost is lowest at £20 when a firm produces 50,000 units, and the market price is £20. Firms are:
Price covers average cost exactly at the efficient scale. Nothing attracts entry or forces exit.
The innovation. One firm finds a process with average cost AC2, minimum £12. Price is still £20, set by the mass of old-technology firms, so the innovator earns £8 a unit: the shaded rectangle, £320,000 a year. Schumpeter’s point is that this profit is the reward for innovating and the reason anyone bears the risk. Patents, secrecy and lead time decide how long it lasts.
Check: A firm introduces a new process with average cost curve AC2, lowest at £12 at 40,000 units. While the price is still £20 it earns:
(20 − 12) × 40,000 = £320,000. This is the incentive to innovate: the first mover sells at the old price with the new costs.
Diffusion. Imitators enter with AC2; supply rises and price falls to P2 = £12, the new minimum average cost. The innovator is back to normal profit at E2. Firms still on AC1 lose at least £8 a unit however much they make: they exit, and their plants are scrapped or sold cheap. Consumers pay £12 instead of £20. The gain is permanent; the innovator’s profit was temporary; the old firms’ loss is total.
Teacher tip: Ask who the diesel firms are here. Not necessarily different companies: a carmaker can own both curves. The question is whether it can move to AC2 faster than entrants who never had an AC1. Kodak invented the digital camera and went bankrupt in 2012.
Check: The new technology spreads: imitators enter, and in a competitive market the price falls to £12. Firms still using the old technology:
This is the destruction in creative destruction: the old capital is not merely less profitable but worthless at the new price.
The effect on market structure cuts both ways. Where the new technology makes old capital obsolete, incumbents lose their advantage and markets become more contestable: new carmakers, new banks, new broadcasters. Where it brings economies of scale or network effects, software platforms, search, social networks, it creates new monopolies more durable than the old. Creative destruction destroys market power and creates it, and policy has to tell which is happening.
Check: Technological change lowers barriers to entry when it:
When the old advantage is embodied in old capital, the new technology wipes the advantage out. The other three raise barriers.
The policy question. The shaded profit is both the reward for innovation and a sign of market power. Schumpeter argued that monopoly profit funds research and the expectation of keeping it motivates the risk; Arrow replied that an incumbent gains less from an innovation that replaces its own product than an entrant does. Patents try to split the difference: 20 years of protected profit, then diffusion. Competition policy has to protect the process, not the current winner.
Check: Schumpeter argued that large firms with market power innovate more than competitive ones because:
The Schumpeterian case for monopoly: dynamic efficiency may need the profit that static efficiency condemns. Arrow replied that a monopolist has less to gain from replacing its own product.
The student leaves the price at £20 for ever, so the innovator keeps £8 a unit indefinitely. In a competitive market the profit attracts imitators, supply rises and the price falls to the new minimum average cost of £12. The profit is temporary by design: that is what makes it an incentive and what passes the gain to consumers. Only a durable barrier such as a patent, a network effect or a secret process would keep the price at £20.
Evaluate the view that technological change is more likely to increase than to reduce competition in markets.
Technological change is the introduction of new products and new methods of production, and creative destruction is Schumpeter’s name for the way they displace old firms, capital and market power while creating new ones. On a cost diagram a process innovation gives the innovator a lower average cost curve; at the old price it earns supernormal profit, imitators enter, the price falls to the new minimum average cost, and firms that cannot move to the new curve make losses at every output and exit. Competition increases in that story because the new technology makes the incumbents’ advantage, their engine plants or presses or branch networks, obsolete, so entrants start level or ahead.
The car market shows it happening. In October 2026 battery electric cars took 32% of UK registrations, with more than 180 models on sale from established makers and from entrants that have never built an engine, while diesel, close to half the market a decade ago, took 3.8%. Streaming did the same to video rental, digital photography to film, online retail to department stores. In each case the barrier that protected the incumbent was embodied in capital the new technology did not need, and contestability rose.
**However,** the same process creates market power where the new technology brings economies of scale, network effects or data. Search, social networks, operating systems and app stores are more concentrated than the industries they replaced, because the value of the product rises with the number of users and the marginal cost of serving one more is near zero, so one firm tends to win and the position is durable. Incumbents also buy entrants before they become threats, and patents, which exist to reward innovation, protect the innovator’s profit for twenty years. Schumpeter himself argued that large firms with market power innovate more because they can finance research and keep the returns, so a world of more innovation may be one of fewer, larger firms; Arrow’s reply, that an incumbent gains less from replacing its own product than an entrant does, is why the electric car came from outside the largest engine makers.
On balance the view holds where the new technology is embodied in capital that makes old capital obsolete, which describes manufacturing, retail and media, and fails where it is embodied in scale and networks, which describes much of the digital economy. The direction depends on the technology, not on technological change as such. The task for competition policy is therefore to keep entry possible, by stopping incumbents acquiring their successors and by limiting how long network and data advantages can be locked in, rather than to protect any particular firm, old or new, from the process.
Car market figures: Society of Motor Manufacturers and Traders, new car registrations, October 2026 (published 5 November 2026): 152,730 registrations; battery electric 48,858, a 32.0% share; petrol 56,876, down 11.6%; diesel 5,811, down 15.9%; 503,805 battery electric cars in January to October, a 26.6% share, against 23.4% for 2025; more than 180 battery electric models on sale. The cost curves and prices on the diagram are invented. Diagrams drawn with Rough.js and perfect-freehand, both MIT licensed.