What We Publish

Will the use of a carbon tax for revenue generation produce an incentive to continue carbon emissions?

Rong Wang, Juan Moreno-Cruz, and Ken Caldeira · Environmental Research Letters 12, 064001 · 2017

Key finding. In the DICE model under its default configuration, a revenue-generating carbon tax creates a perverse incentive to continue carbon emissions from about the year 2085 — the point beyond which further increases in the tax rate reduce total carbon tax revenue.

Four panels from the DICE-2013R model running 2015 to 2250, comparing welfare-maximizing, revenue-maximizing and zero-carbon-tax cases. Panel a shows carbon price, where the welfare-maximizing and revenue-maximizing curves cross at 2085 and the shaded regions are labelled as incentives for higher and then lower carbon prices. Panel b shows emissions, panel c the carbon tax as a percentage of gross output, and panel d the atmospheric temperature increase, with the welfare and revenue curves diverging after 2085 and 2135.
Before 2085 the price that maximises welfare is below the price that maximises revenue, so a revenue-seeking government wants a *higher* carbon price than is optimal. After 2085 the curves cross and the incentive reverses — which is the perverse incentive, arriving precisely when the policy starts to work. Figure 1 from Wang, Moreno-Cruz and Caldeira (2017), Environmental Research Letters 12, 064001. Reproduced under CC BY 4.0. Extracted from the published PDF and resized for web display.

What question did this research address?

Integrated assessment models generate an optimal carbon price by maximising social welfare, and typically produce a price that starts low and rises over time. That framing treats the price purely as a lever on emissions.

But a carbon tax is also a revenue stream, and governments come to rely on revenue. If a rising price is to reach near-zero emissions, it must ultimately collect near-zero revenue. This paper asked at what point raising the tax starts to *reduce* what it collects, and whether that creates an incentive to keep emitting.

What did we find?

The mechanism is arithmetic rather than political. Revenue is the tax rate times the emissions taxed, so a policy that succeeds in driving emissions toward zero must drive its own revenue toward zero as well.

Somewhere on that path lies a maximum, past which a policymaker asked to raise the rate is being asked to accept less money. Before it, the interests of the climate and the treasury coincide; after it, they diverge.

In DICE that turning point falls around 2085, so the risk is not a concern for the immediate future.

The timing is sensitive to assumptions — the cost of climate damages and the cost of decarbonizing the energy system both move it.

The authors present this as a schematic result meant to prompt more comprehensive integrated assessment models to represent a broader spectrum of incentives, not as a forecast of a particular date.

Why does it matter?

It identifies a failure mode built into a policy widely regarded as the clean solution. The carbon tax is usually assessed on whether it sets the right price; this asks what it does to the interests of the institution collecting it.

The finding is genuinely reassuring in the near term and genuinely uncomfortable in the long term, and both halves matter: there is time, but the problem does not go away by itself, and it arrives precisely when the policy is working.

It also argues against hypothecating carbon revenue to permanent spending commitments, since that is the mechanism by which a temporary revenue stream becomes a constituency for continued emissions.

Citation

Rong Wang, Juan Moreno-Cruz, and Ken Caldeira (2017). Will the use of a carbon tax for revenue generation produce an incentive to continue carbon emissions?. Environmental Research Letters 12, 064001.

Related