Market failure and government intervention

Pearson Edexcel International AS Economics (WEC11), specification Issue 2, June 2018. This topic covers spec sections 1.3.5–1.3.6 in full: market failure (six rows) and government intervention in markets (two rows). The shared model continues from topic 1, drawn in inverse form: demand (marginal private benefit) $P = 70 - Q/2$, supply (marginal private cost) $P = Q/4 - 5$, market equilibrium $Q = 100$, $P = 20$ — and an external cost or benefit of 6 per unit reproduces the topic-1 tax and subsidy equilibria, so every welfare area below cross-checks against those numbers.

Must-know for this topic:

Market failure 市场失灵 means too much or too little of a good is produced and/or consumed compared with the socially optimal level of output (1.3.5.1a). The sources (1.3.5.1b): externalities 外部性, the free-rider problem and the non-provision of public goods, imperfect market information, moral hazard, and speculation and market bubbles. Each is a reason the price mechanism's signals fail to value every cost and benefit, so the market equilibrium is not the social optimum.

Private benefit/cost is what the decision-maker enjoys or pays; the external benefit 外部收益 or external cost 外部成本 falls on third parties; the social value sums them: social benefit 社会收益 $= MPB + EB$ and social cost 社会成本 $= MPC + EC$ (1.3.5.2a–b). Externalities come from production (a factory's emissions are an external cost of production; training that benefits other firms is an external benefit of production) or from consumption (vaccination conveys external benefits of consumption; congestion and pollution from driving are external costs of consumption) (1.3.5.2c).

Diagrammatically, marginal analysis puts $Q$ on the horizontal axis. Negative externality of production: demand is $MSB = MPB: P = 70 - Q/2$ and supply is $MPC: P = Q/4 - 5$, so the market trades at $(100, 20)$; with an external cost of 6 per unit $MSC = MPC + 6$ crosses $MSB$ at $Q^* = 92$ — the market OVER-produces by 8, and the welfare loss 福利损失 triangle between the curves over $[92, 100]$ is $\tfrac12 \times 8 \times 6 = 24$ currency units/day (1.3.5.2d). Positive externality of consumption: $MSB = MPB + 6$ crosses $MPC$ at $Q^* = 108$ — the market UNDER-consumes by 8, and the welfare GAIN available over $[100, 108]$ is the same $\tfrac12 \times 8 \times 6 = 24$. Contexts to know cold (1.3.5.2e): transport (congestion, pollution), health (vaccination, passive smoking), education (spillovers from a skilled workforce), environment (emissions), financial (risk externalities of bank failure).

Private goods 私人物品 are rival (one person's consumption reduces what is left) and excludable (non-payers can be stopped); public goods 公共物品 are non-rival and non-excludable — a flood defence protects everyone at once and cannot be withheld (1.3.5.3a). The free-rider problem 搭便车问题 means nobody can be charged, so private firms cannot cover costs and public goods are not provided by the private sector at all — the state must provide them (1.3.5.3b).

Information is symmetric when both sides know the same things; asymmetric information 信息不对称 (the used-car seller knows the car, the insurer knows less than the insured) creates information gaps that misallocate resources in healthcare (patients cannot judge treatment), education (students cannot judge course quality), pensions (complexity hides value) and insurance (adverse selection) (1.3.5.4). Moral hazard 道德风险 arises AFTER a contract when the protected party takes more risk: insured drivers drive less carefully, bails encourage banks to lend recklessly — consumers pay higher premiums, producers and workers carry the risk, and governments face bailout bills in insurance and banking alike (1.3.5.5). Market bubbles 市场泡沫 arise when speculation feeds on itself: buyers buy only to resell, prices detach from fundamentals and then crash — housing and stocks and shares are the specification contexts, and the burst hits consumers (negative equity, pension losses), producers (collapsed demand), workers (unemployment) and governments (recession, bailouts) (1.3.5.6).

Intervention exists to correct market failure — to move output towards the social optimum (1.3.6.1a). The methods (1.3.6.1b): indirect taxation (ad valorem and specific — a Pigouvian tax equal to the marginal external cost 6 shifts supply to the social optimum $(92, 24)$, exactly the topic-1 tax arithmetic); subsidies (6 per unit corrects the positive externality: output rises to the social optimum $Q^* = 108$, where buyers pay 16 and sellers receive $16 + 6 = 22$, the point where $MSB = MSC$); maximum prices 最高价格 (set at 16, below equilibrium, for affordability: $Q_d = 108 > Q_s = 84$, a shortage of 24 — rationing, queues and black markets follow); minimum (guaranteed) prices 最低价格 (set at 24, above equilibrium, to protect producers: $Q_s = 116 > Q_d = 92$, a surplus of 24 — the government may have to buy and store it); tradeable pollution permits 可交易污染许可证 (the cap fixes total emissions — here 84 permits — and demand sets the permit price, 28 in our model, so firms with low abatement costs sell permits to firms with high ones); extension of property rights 产权 (giving victims the right to sue internalises the externality); state provision (public goods, health, education); regulation 监管 (bans, standards, price caps); and the provision of information (closing information gaps). Contexts (1.3.6.1c): health, housing, education, transport, environment, energy, agriculture and commodities.

Government failure 政府失灵 is intervention that results in a NET welfare loss — the cure costs more than the disease (1.3.6.2a). The causes (1.3.6.2b): information gaps (the state misjudges the externality's size, so the tax of 6 overshoots and wastes the triangle it meant to save); unintended consequences (minimum prices create surpluses; maximum prices create black markets); excessive administrative costs (collecting, enforcing, storing); lack of incentives (state providers face no competition); and moral hazard (guaranteed prices or bailouts change behaviour). Every intervention answer should close by weighing the correction against these costs.

Extract. In February 2003 London introduced a congestion charge: drivers paid £5 a day to drive into central London on weekdays. In the first year traffic entering the zone fell by about 18% and delays fell by about 30%. Bus services improved, as buses were no longer stuck in traffic. The charge rose in steps and has been £15 since June 2020. Transport for London spends the revenue on public transport. Critics argue that a flat charge takes a larger share of a low income, that some small shops inside the zone lost customers, and that congestion later returned as roadworks and delivery vans increased. In January 2025 New York City became the first US city to introduce a similar charge.

Questions

• (a) Explain, using the extract, why road congestion is an example of a negative externality. [4]
• (b) Before the charge, 100,000 vehicles a day entered the zone. Using the extract, calculate the number entering after the first year, and the daily revenue at £5. [4]
• (c) With the help of a diagram, explain how a congestion charge could move the market towards the socially optimal level of road use. [6]
• (d) Examine two problems a government faces in setting the right level of a congestion charge. [8]
• (e) Evaluate whether road pricing is a better solution to congestion than building more roads. [12]

Model answers

(a) Each driver who enters a busy road slows down every other road user: buses, delivery vans and other drivers lose time and burn more fuel, and pedestrians breathe more polluted air. These are costs to third parties that the driver does not pay. So the marginal social cost of a journey is greater than its marginal private cost, and, without a charge, too many journeys are made at busy times. The extract's 30% fall in delays when traffic fell shows how large these external costs were.

(b) Vehicles after the charge $= 100{,}000 \times (1 - 0.18) = 82{,}000$ a day. Revenue $= 82{,}000 \times £5 = £410{,}000$ a day (ignoring exemptions and discounts).

(c) Use the negative-externality diagram above: the market for journeys settles where MPB = MPC, at a quantity above the social optimum where MSB = MSC. A charge equal to the marginal external cost at the optimum adds to the private cost of each journey, shifting MPC up to MSC. Drivers whose journeys are worth less than the full cost switch to buses, travel at other times or stay away, so the quantity falls to the social optimum and the welfare loss triangle is removed. The 18% fall in traffic is this movement.

(d) Measuring the external cost is hard. The cost of congestion depends on time of day, place and weather, and the value of time lost differs between people. A flat £15 charge will be too high at quiet times and too low at peak times, so it cannot match MSC everywhere. Demand responses change over time. Demand for car journeys into a city centre may be price inelastic for those with no alternative, so a charge reduces traffic less than expected; over time, the effect can wear off as incomes rise and as other causes, such as roadworks and delivery vans, add to congestion (as the extract notes). Evaluation: technology such as number-plate cameras allows charges to vary by time, which reduces the first problem; regular review of the charge reduces the second.

(e) For road pricing: it tackles the cause of the externality by making drivers pay the external cost, so the space on roads goes to journeys valued most. It raises revenue (about £410,000 a day in the example) that can fund buses and rail, giving people alternatives. It worked quickly in London: traffic fell 18% and delays 30% in a year.

For building roads: new roads add capacity without charging drivers, so they are popular and do not hurt low-income drivers. In growing cities with little road space, more capacity may be needed.

Against building roads: new roads often fill up with new traffic (induced demand) because lower congestion makes driving cheaper in time, so congestion returns. Roads are very costly, take years to build, use scarce land and raise pollution, an external cost of their own.

Against road pricing: a flat charge is regressive, and some city-centre businesses lose trade. Setting the right charge is hard (part d), and enforcement and cameras cost money. Its effect may fade without regular increases.

Judgement: road pricing is usually the better policy in dense city centres, where there is no space for new roads and public transport is a real alternative, because it targets the externality and funds alternatives. Its success depends on charges that vary with congestion and on reinvesting revenue in public transport; building roads may still be needed between cities, where road pricing is harder to apply and alternatives are fewer.

What examiners reward: the externality explained with the extract's evidence, calculations with clear working, a labelled MSC/MPC diagram used in the explanation, problems that are examined (with a counter-point), and a judgement that depends on clear conditions.

• Market failure defined; the five sources named (1.3.5.1).
• Private/external/social benefits and costs; production vs consumption externalities; MSC/MSB diagrams with market vs social optimum and the welfare-loss/gain areas; the five contexts (1.3.5.2).
• Rivalry and excludability; public vs private goods; free-riding and non-provision (1.3.5.3).
• Symmetric vs asymmetric information; information gaps in healthcare, education, pensions, insurance (1.3.5.4).
• Moral hazard after the contract; effects in insurance and banking (1.3.5.5).
• How bubbles arise; effects in housing and stocks on all four groups (1.3.5.6).
• Purpose of intervention; the eight methods with the max/min price shortages and surpluses computed; the eight contexts (1.3.6.1).
• Government failure as net welfare loss with its five causes (1.3.6.2).

Boundary notes: diagrams carry marks — axes labelled, curves named (MPB/MSB, MPC/MSC), the optimum and the welfare area identified. State the assumption that the externality is constant at 6 per unit (in reality marginal external costs vary with output). All worked numbers are original and consistent with the shared model of topic 1.