Why Governments Intervene in Demerit-Good Markets, and Tax versus Ban (Cigarettes versus Narcotics)

Answer

Part (a)

The government’s microeconomic objectives are allocative efficiency and equity, and it intervenes where the free market fails to deliver them. Demerit goods — goods the government judges socially undesirable and which are over-consumed when left to the market — generate market failure through two distinct channels, and it is this failure that justifies intervention.

Negative consumption externality. A self-interested consumer maximises utility by consuming where marginal private benefit equals marginal private cost. With cigarettes, the marginal external cost of second-hand smoke — respiratory harm and healthcare costs borne by third parties — is not internalised, so the marginal social cost lies above the marginal private cost ($MSC = MPC + MEC$). The market consumes $Q_p$ where $MPB = MPC$, exceeding the social optimum $Q_s$ where $MSB = MSC$. Over the range $Q_s$ to $Q_p$ marginal social cost exceeds marginal social benefit, so a deadweight welfare loss arises and resources are allocatively inefficient.

Imperfect information. Independently, consumers act on perceived rather than actual costs. A smoker who underestimates the long-run health damage and medical cost to himself perceives a lower $MPC$ than the true one ($MPC_{\text{actual}} > MPC_{\text{perceived}}$), so he over-consumes relative to the quantity he would choose if fully informed, where $MPB_{\text{actual}} = MPC_{\text{actual}}$. This too produces over-consumption and a deadweight loss. Because imperfect information operates on the consumer’s own welfare while the externality operates on third parties, the two are separate failures — and because both push consumption above $Q_s$, intervention to reduce demerit-good consumption towards the social optimum is warranted to raise social welfare and restore allocative efficiency.

Part (b)

The Singapore government taxes cigarettes but bans narcotics because the appropriate instrument depends on where the socially optimal quantity lies, which in turn depends on the magnitude of the marginal external cost. The thesis is that each instrument is matched to its good, and that both are effective in Singapore’s institutional context — though neither is unconditionally so.

Tax on cigarettes. A per-unit tax raises sellers’ costs and the effective price consumers pay, shifting $MPC$ up to $MPC + \text{tax}$. Set equal to the $MEC$ at $Q_s$, the tax forces consumers to internalise the externality: the new private optimum $Q_p'$ coincides with the social optimum $Q_s$, eliminating over-consumption and achieving allocative efficiency. The principal limitation is elasticity: cigarettes are addictive with few substitutes, so demand is price-inelastic and a tax produces a less-than-proportionate fall in quantity demanded — consumption may not reach $Q_s$. A larger tax could close the gap but invites lobbying by smokers and tobacco firms, and the government may anyway lack the information to quantify the $MEC$ and so set the ideal rate. These limitations bite less in Singapore: strong data-collection sharpens the $MEC$ estimate, and a cohesive legislature makes a large tobacco tax politically feasible.

Ban on narcotics. A ban makes consumption (and, in Singapore, possession and sale) illegal; if obeyed, consumption falls to zero. Where the $MEC$ of narcotics is so high that $MSC > MSB$ at every positive quantity, the socially optimal quantity is zero, and a ban that achieves zero consumption is allocatively efficient — it avoids the entire external cost. Its limitation is enforcement: illegal consumption leaves the quantity above the optimum. The extent depends on the government’s ability to detect and deter, and Singapore’s heavy investment in policing and severe penalties keeps illegal consumption very low, so the ban is largely effective.

Why the instruments differ. The choice turns on the social optimum. For cigarettes, although $MSC > MPC$, there remains a range of consumption where $MSB > MSC$, so $Q_s$ is positive; a tax permits consumption up to $Q_s$ and maximises welfare, whereas a ban would force consumption to zero and create an under-consumption deadweight loss (area DAE). For narcotics, the $MEC$ is so large that $MSC > MSB$ throughout, so $Q_s$ is zero and a ban is correct; a tax would have to be enormous given inelastic demand, and high-income users could still legally consume, leaving substantial over-consumption.

Judgement. Both instruments are, on balance, effective in achieving an efficient allocation in Singapore — but for different reasons and with different residual problems. The tax on cigarettes works regardless of whether the over-consumption stems from externality or imperfect information, and Singapore’s enforcement capacity and fiscal-political feasibility let it be set high enough to matter. The ban on narcotics is effective chiefly because of severe, well-enforced deterrence that holds illegal consumption near zero. Three qualifications temper the verdict. First, both governments rely on complementary public education that corrects perceived $MPC$ and shifts tastes — for narcotics, changing preferences may be preferable to a ban, sparing enforcement cost and the risk of a black market, though it works only in the long run. Second, the claim that narcotics’ $MEC$ greatly exceeds that of cigarettes, while plausible, is contentious and hard to quantify: $MEC$ varies with income, age, environment and consumption mode, and reliable data is scarce, so the precise placement of $Q_s$ is uncertain. Third, instrument choice depends on more than the optimum — administrative cost, ease of monitoring and political feasibility all weigh in. The defensible stand is that Singapore’s tax-and-ban split is well matched to the differing social optima of the two goods, and effective in its institutional setting, provided it is reinforced by education and continually informed by better estimates of the external costs.