Externalities are costs or benefits of a transaction that fall on third parties — neither buyer nor seller. They're the canonical market failure: the unregulated market produces too much of negative externalities and too little of positive ones because the participants ignore the third-party effects.
Negative externalities
When production or consumption imposes costs on others not part of the transaction. Standard example: pollution. A factory produces goods and emits emissions; the emissions impose health and environmental costs on neighbours, but those costs aren't paid by the factory or its customers.
The marginal-social-cost framework
Private MC = the cost to the producer of one more unit of output Social MC = Private MC + Marginal External Cost The unregulated market produces where Private MC = P (the producer's profit-max condition). But the SOCIAL optimum is where Social MC = P. Marginal external cost (MEC) — the third-party cost of the marginal unit. For pollution: the additional health/environmental damage from one more unit of emission. The gap: at the unregulated quantity, Social MC > P (the social cost exceeds the social benefit at the margin). Reducing output toward the social optimum increases welfare. The unregulated market 'over-produces' by the amount of the negative externality.
The Pigouvian tax
Arthur Pigou (1920) proposed the textbook solution: tax the activity producing the externality at a rate equal to the marginal external cost. This forces the producer to face the social cost and aligns private optimisation with social optimisation.
Optimal Pigouvian tax
τ* = MEC at the socially optimal quantity • τ* = the optimal Pigouvian tax • MEC = marginal external cost The tax shifts the producer's effective marginal cost up to Social MC. New profit-max condition: Private MC + τ = P, which means Social MC = P at the new quantity. Aligned with the social optimum.
Examples in African contexts
- Fuel — externalities include CO2 emissions (climate change), local air pollution, traffic congestion. Kenya fuel levy (~KES 18/litre on petrol and diesel as of 2025) is partly Pigouvian, though the rate is below estimated social cost
- Tobacco and alcohol — externalities include healthcare costs, productivity losses, family disruption. Excise duties function as Pigouvian taxes
- Charcoal production — deforestation externalities. Kenya charcoal regulation is administrative (permitting, county-level bans) rather than tax-based — less efficient than a Pigouvian tax would be in principle, but easier to enforce in the informal sector
- Air emissions from industrial sources — currently weakly regulated in most African economies. KEMRI and NEMA studies suggest substantial unmeasured costs from urban air quality
Positive externalities
When production or consumption confers benefits on others not part of the transaction. Standard examples: vaccination, education, scientific research, infrastructure investment.
The marginal-social-benefit framework:
- Private MB = the benefit to the consumer from one more unit
- Social MB = Private MB + Marginal External Benefit
- The market produces where Private MB = MC. The social optimum is where Social MB = MC
- The unregulated market UNDER-produces positive externalities by the amount of the external benefit
The Pigouvian response: subsidise the activity. Subsidy = marginal external benefit. The subsidy shifts private MB up to match social MB.
Vaccination — the textbook positive externality
When you get vaccinated, you protect yourself (private benefit) AND you reduce the chance of infecting others (external benefit — herd immunity). The market under-supplies vaccination because individuals weigh only their private benefit.
Public policy response across most countries: public provision of vaccines free or near-free, often with school-attendance or workplace mandates. Kenya's Expanded Programme on Immunisation (EPI) provides free childhood vaccines (DPT, polio, measles, BCG, etc.) — coverage in Kenya is ~85% for the standard schedule, with rural-urban gaps.
The Coase theorem
Ronald Coase (1960): if property rights are well-defined and transactions costs are low, parties affected by externalities will bargain to the efficient outcome — REGARDLESS of who holds the initial property right. The textbook example: a factory and a downstream fishery. If the factory has the right to pollute, the fishery can pay the factory to pollute less. If the fishery has the right to clean water, the factory can pay the fishery for permission to pollute. Either way, the negotiated outcome is the socially efficient level of pollution.
Coase theorem conditions
1. Well-defined property rights — both parties know what they own and what they don't 2. Low transactions costs — bargaining, contracting, and enforcement are cheap 3. Symmetric information — both parties know the costs and benefits at stake When these conditions hold, private bargaining produces the efficient outcome without need for government intervention. When they DON'T hold — and in practice they often don't — government intervention (Pigouvian taxes, regulation, public provision) becomes necessary.
When the Coase theorem fails
Coase's optimism applies to small-number-of-parties externalities. Most real-world externalities don't meet the conditions:
- Air pollution — millions of affected parties; transactions costs of coordinating any of them are huge. Even if rights were perfectly assigned, bargaining is infeasible
- Climate change — global externality; impossible to assign property rights; intergenerational time horizon; massive coordination problem
- Disease externalities — hard to bargain with disease
- Information asymmetries — parties may not know the magnitude of harm
- Holdouts — even with low transactions costs, parties have incentive to hold out for better terms, blocking agreement
Coase's deeper point
Coase's lasting contribution wasn't 'no need for government' — it was the observation that the choice of policy instrument should be based on TRANSACTIONS COSTS. Where transactions costs are low (small-number-of-parties, well-defined property), bargaining can work. Where transactions costs are high (many parties, public-good characteristics, information problems), government intervention is needed. This transactions-cost economics is the foundation of modern policy analysis: don't ask 'should government act?' — ask 'what's the cost of different ways to address this externality, and which is lowest?'
Tradeable permits — a hybrid solution
Cap-and-trade systems set a total quantity of allowed emissions, distribute emission permits among polluters, and allow trading. Each polluter can sell permits if it can reduce emissions cheaply, or buy them if reducing is expensive. The system internalises the externality via the market price of permits while leaving the abatement-allocation decision to firms.
- EU Emissions Trading System (EU ETS) — established 2005; covers electricity and industrial-sector CO2 in EU; permit prices have risen from €5-15/tonne in early years to €70-100+/tonne in 2024
- California cap-and-trade — covers most state emissions; permit prices ~$30-40/tonne
- African applications — South Africa carbon tax (since 2019) at ZAR 144/tonne CO2 (~$8); Kenya has the Climate Change Act 2016 framework but no operational cap-and-trade yet
Public goods revisited
Public goods (covered in Public Finance module 1) have non-rival and non-excludable consumption. They're an extreme case of positive externality where the benefit is fully external to any individual transaction. Markets fail to provide them; collective provision is required.
Common-pool resources
Non-excludable but rival — the canonical 'tragedy of the commons' (Hardin 1968). Examples: fishery, communal grazing land, groundwater. Each user takes more than is socially optimal because they don't bear the full marginal social cost of their use.
Elinor Ostrom (1990, Nobel 2009) demonstrated that community institutions can govern commons effectively when:
- User group is bounded and identifiable
- Rules are made by the users themselves, adapted to local context
- Monitoring of compliance is feasible
- Sanctions for violation are graduated and accepted as legitimate
- Conflict-resolution mechanisms exist
African applications: pastoralist commons management in northern Kenya, Ethiopia, Mali; community forestry in Kenya, Tanzania; fishery cooperatives on Lake Victoria. Mixed records but real successes.
Exercise
Nairobi's air quality is consistently rated among the worst in East Africa, with particulate matter (PM2.5) often 5-10× WHO guidelines. The major sources are vehicle emissions (50%), industrial sources (25%), open burning of waste (15%), and dust (10%). (1) Identify which sources are amenable to Pigouvian-tax solutions and which require other instruments. (2) Estimate the welfare gain if Nairobi PM2.5 dropped to WHO guideline. (3) Propose a Pigouvian-tax-based response. (4) What political-economy obstacles do you anticipate?