Commonly discussed options for reducing pollution at minimum economic cost are taxes, subsidies and marketable permits.  A combined tax and subsidy should be added to this list.

Suppose a government wants to reduce an industry’s emissions of a pollutant to a target level.  A ‘command and control’ approach with emissions targets for individual firms can be effective in reducing emissions, but is unlikely to do so at minimum economic cost (1).

Economists therefore recommend instead market-based instruments.  These include a tax per unit of emissions, a subsidy per unit of emissions avoided (relative to a suitable benchmark), and marketable emissions permits.   These instruments minimize costs because each firm, in minimizing its private costs, will reduce its emissions to the point at which its marginal abatement costs equal the rate of tax or subsidy, or the market price of permits (2).

Each instrument has disadvantages.  A tax imposes costs on firms, which may be driven out of the industry, and indirectly on their customers.  Marketable permits increase firms’ risks because future permit prices are unknown, and generate transaction costs in the operation of the permit market.

Subsidies bear closer examination.  A per unit subsidy is equivalent to a lump sum subsidy less a per unit tax at the same rate, the implicit lump sum equalling the benchmark emissions level multiplied by the per unit rate (3).  High benchmarks, perhaps reflecting actual emissions in a baseline period, imply large lump sums with gains for most firms.  A subsidy based on a high benchmark is therefore a significant cost to the government and can, perversely, increase emissions by attracting more firms into the industry (4).

The size of the lump sum element, however, will not affect a firm’s decisions on abatement levels, only its decisions whether to operate in the industry at all (5).  Why not, therefore, combine a per unit tax with smaller lump sum subsidies based on lower benchmarks?  This would mean that excess emissions – those above a firm’s benchmark – would bear a per unit tax, while if a firm’s emissions were below its benchmark, the difference would qualify for a subsidy at the same rate per unit.

According to the Tinbergen Rule, achieving n policy objectives requires n policy variables.  In favourable circumstances, a combined tax and subsidy could achieve three objectives:
a) a target for an industry’s emissions;
b) minimization of the industry’s abatement costs;
c) financial neutrality for the industry and the government.
This is how it would work.  The choice of a per unit tax on emissions – rather than a ‘command and control’ approach – ensures cost minimization.  The rate of tax is chosen to induce sufficient abatement to reduce emissions to the target.  Financial neutrality is then achieved by the choice of the benchmarks determining the lump sum subsidies to combine with the per unit tax.

The benchmarks could be set by dividing the industry’s emissions target pro rata to firms’ output (6).  This ensures that, once the emissions target has been achieved, the below-benchmark emissions of some firms must offset the above-benchmark emissions of the others.  Subject to data accuracy and effective administration, the subsidies received by the former will therefore equal the tax paid by the latter.  Making the target – rather than past actual emissions – the basis for the benchmarks also reduces scope for gamesmanship by firms.

A feature of such a combined tax and subsidy is its dynamic effect on entry to and exit from the industry.  For firms which could produce the good with low abatement costs, a net subsidy offers an incentive to enter the industry.  For existing firms with high abatement costs, the net tax is a disincentive to their remaining in the industry.  Starting from a position in which an industry’s emissions exceed target, progress towards the target is therefore by two effects.  Some firms will continue to operate but will reduce their emissions.  Others may exit the industry and be replaced by new entrants with lower emissions, or by existing low-emission firms expanding their output.

Notes and references

1. Hanley N, Shogren J F & White B (2nd edn 2007)  Environmental Economics in Theory and Practice  Palgrave Macmillan  p 83

2. Hanley, Shogren & White, as above  pp 133, 137 & 143

3. Baumol W J & Oates W E  (2nd edn 1988)  The Theory of Environmental Policy  Cambridge University Press  pp 215-216

4. Baumol & Oates, as above  p 222

5. Baumol & Oates, as above  p 217

6. All that is essential for financial neutrality for the industry as a whole is that the sum of firms’ benchmarks equals the industry target.  However, setting benchmarks pro rata to firms’ output – output of goods, not emissions – also ensures a level playing field within the industry.