A specific indirect tax corrects a consumption externality by raising the private cost of the demerit good towards its full social cost. On a demand-and-supply diagram, the supply curve shifts vertically up by the per-unit tax; the market price rises to P₂, the equilibrium quantity falls to Q₂, and if the tax equals the marginal external cost at the social optimum, Q₂ equals the socially efficient quantity Q*. The tax internalises the externality, generates revenue the government can hypothecate to the external cost (healthcare, pollution clean-up), and preserves consumer choice — those for whom the private benefit still exceeds the full private + external cost continue to consume at the new price.
Case for the tax as the most effective tool. The price mechanism is informationally efficient: consumers' willingness-to-pay already encodes their private valuation, so a tax that raises price to the socially correct level produces the right quantity reduction without the government having to know who the heaviest consumers are. Revenue is a dividend: unlike a minimum unit price (whose extra margin accrues to retailers), the tax funds the government budget and can finance treatment or compensation for those harmed by the externality. The tax is well-understood administratively — existing excise systems can be reused — and preserves consumer choice on an individual-household basis, which is politically more acceptable than an outright ban. For a classic demerit good with a well-measured externality (tobacco, fossil fuels), the tax has a strong track record.
Case against — the tax has serious limitations. First, demand is often price-inelastic for addictive demerit goods (tobacco, alcohol by dependent drinkers), so the quantity response is small; revenue rises but the externality is barely reduced. Second, the tax is regressive: low-income households spend a larger share of income on taxed demerit goods, so the burden falls disproportionately on those least able to bear it. Third, substitution effects can defeat the tax if close untaxed alternatives exist (sugary drinks substituted for fruit juice, cigarettes for e-cigarettes). Fourth, informational demands are heavy: setting the right rate requires measuring the marginal external cost, which is often uncertain and politically contested. Fifth, enforcement can be eroded by cross-border smuggling of highly-taxed goods.
Alternatives on their own criteria. Direct regulation (bans, quotas, product standards) is more certain in effect because it does not depend on price-elastic demand — but it is blunt, generates deadweight loss by over-shooting among moderate users, and foregoes revenue. Information provision (health warnings, labelling, campaigns) corrects the underlying knowledge failure, is progressive, but is slow to shift ingrained consumption habits. Minimum unit prices achieve a floor without the regressive incidence pattern of a tax, but channel the extra margin to producers rather than the government. Tradable permits can internalise production externalities (carbon trading) but are less applicable to consumption externalities. Subsidising substitutes (e-cigarettes, electric vehicles) encourages switching but costs rather than raises revenue.
Reasoned judgement. The specific indirect tax is the most effective single instrument when (i) demand is reasonably price-elastic at the relevant margin, (ii) close untaxed substitutes do not exist or are also taxed, (iii) the regulator can estimate the marginal external cost with reasonable accuracy, and (iv) revenue is hypothecated to the external damage. It fails as the single best tool when demand is highly inelastic (addiction), when near-substitutes are untaxed, when distributional concerns dominate (sharply regressive incidence on low-income households), or when the externality is highly uncertain (making the right tax rate unknowable). In practice the strongest-performing policy designs are mixes: a Pigouvian tax to set the price signal, information provision to correct the knowledge failure, direct regulation to deliver certainty on the heaviest-consuming groups, and subsidies to accelerate substitution. The view that the tax is the 'most effective' tool is therefore correct as 'the best single price-based instrument' for a well-measured externality but overstated as 'the best overall policy' in any realistic market where the four conditions above do not all hold.