Climate policy uncertainty: Shall we hedge against it?

Valentina Bosetti, Carlo Carraro, Alessandra Sgobbi, Massimo Tavoni

14 October 2008



Despite growing concerns about climate change, there is little consensus about the scale and timing of actions needed to respond to it. International negotiations on an effective climate policy have been stalling for almost a decade now, and those currently underway might fail to reach a comprehensive agreement in the near future. One reason for this slow progress is the global nature of the problem, whose solution requires a coordinated action. However, a global solution to climate change is challenged by the nature of the problem itself: carbon emissions by different countries differ greatly and change rapidly; projected climate damages are very asymmetric across countries; and legacy and competitiveness issues further complicate the picture. This has motivated free-riding behaviour by most countries, resulting in a continued growth of total emissions. Pervasive uncertainties over the costs and benefits of climate protection measures, as well as over the magnitude and severity of the damages, have also been used as a justification for a “wait-and-see” approach.

However, time is a crucial factor that is likely to affect the effectiveness and the cost of climate policy in a significant way. Actions taken today and in the near future will have century-long consequences as a result of the persistence of CO2 in the atmosphere, the long lifetime of energy capital stocks and infrastructures, and the time lags involved in changing habits to stimulate demand side-responses. Therefore, failing to include uncertainties over a future potential climate agreement into current decisions might result in inefficiencies and excessively costly measures.

The costs of delaying action

Uncertainty surrounding the willingness and ability to take global action – and the expectation that such uncertainty may be resolved in one or two decades – may be inducing policymakers to adopt a wait-and-see strategy. In recent research, we quantify the economic implications of a twenty-year delay, assuming a given objective to stabilise CO2 concentrations within the century.

Table 1 shows that if world policymakers jointly start taking action now to control climate change, the world will incur economic losses of 0.2% to 2.3% of Gross World Product (GWP), depending on whether they adopt a mild or stringent policy. Although not negligible, such an effort might be desirable given the large spectrum of potential consequences of climate change. However, delayed action could be much more costly. While the costs of postponing mild actions would be relatively modest (0.3% rather than 0.2% of GWP), moving from business-as-usual to an ambitious climate stabilisation target twenty years from now would be extremely costly – up to 5.5% of GWP.

Table 1 Costs of delayed actions (GWP loss to 2100 at net present value discounted at 5%)

  Business as usual Adopt a mild* climate policy Adopt a stringent** climate policy
Take action now  - 0.2% 2.3%
Wait 20 years on the business-as-usual path  - 0.3% 5.5%
Wait 20 years on a mild policy* path  0.06%  - 2.7%

*550 ppmv CO2 only, roughly equivalent to 650 ppmv all gases
**450 ppmv CO2 only, roughly equivalent to 550 ppmv all gases

This represents an additional policy cost of about $2.2 trillion per year of delay, a figure that disrupts the common understanding of the economic feasibility of embarking on a stringent policy at a later point in time, as currently perceived by policymakers.

The last row of Table 1 provides an additional insight – a policy strategy that immediately undertakes some emissions reductions that are consistent with a 550 ppmv (CO2 only) stabilisation target and in twenty years reverts to the business-as-usual scenario does not visibly harm global welfare (the cost would be 0.06% of GWP). The same policy can become much more stringent after 20 years with a global cost equal to 2.7% of GWP, which is much smaller than the 5.5% obtained when delaying all actions for 20 years. This result suggests that there exists a hedging strategy, ranging somehow between a 550 and a 450 ppm stabilisation policy. What is its optimal dynamic path?

Hedging against climate policy uncertainty

To quantify the optimal hedging strategy, we simulate a situation in which uncertainty about the climate target is resolved in 2035. Today, three scenarios are assumed to have equal probability to emerge in 2035: (i) a no-agreement case, that mimics an extreme scenario of no international or national policies, either because policymakers will prefer to adapt to climate change or because climate negotiations are unable to achieve consensus on any policy measure; (ii) a mild policy case, in which consensus in found to achieve a 550 (CO2 only) stabilisation target; (iii) an ambitious policy case, in which the climate agreement adopted in 2035 is designed to achieve a 450 (CO2 only) target.

Figure 1 shows the cost-effective path of carbon emissions for the three different deterministic cases vis-à-vis the optimal hedging strategy that takes into account the fact that all three are possible with equal probability. Figure 1 provides clear evidence of hedging behaviour. The optimal strategy before uncertainty is resolved is to engage in significant mitigation – in 2030, emissions are 57% lower than in the business-as-usual scenario. That is to say, in a world that has an equal chance of being confronted with no climate policy, a mild policy, or a stringent policy in 30 years time, the best strategy in the short-term (for the next two decades) would be to pursue a policy between the mild and the stringent one, and closer to the latter. Departures from this optimal hedging policy will result in additional economic burdens on top of the already significant investment costs required to improve and expand world energy and power sectors.

Figure 1 Fossil fuel emissions under the optimal hedging strategy, 2030 and 2100

Policy recommendations

The policy implications of the model experiment just described are quite clear, and support the arguments that call for immediate action to tackle climate change (see, e.g. Yohe et al, 2004). No particular assumption on discounting motivates this approach, contrary to Stern (2006). It is guided by the simple need to account for the possibility that the world might need to undertake some ambitious and effective climate protection measures sometime in the future. The optimal short-term strategy in the presence of uncertainty on the climate target that will be adopted in the future is an ambitious one that requires large cuts of carbon emissions in the next 20 years. A milder policy could be adopted with moderate additional costs. No action today will increase future costs dramatically.


Bosetti, V., C. Carraro, A. Sgobbi and M. Tavoni (2008), “Delayed Action and Uncertain Targets. How Much Will Climate Policy Cost?”, CEPR Discussion Paper 6973.
Stern, N. (2006). The Economics of Climate Change: The Stern Review. Cambridge University Press, Cambridge.
Yohe G, Andronova N, Schlesinger M (2004) To hedge or not against an uncertain climate future. Science 306:416-417.



Topics:  Energy Environment

Tags:  climate change, emissions, global warming

Associate Professor at the Department of Economics, Bocconi University

Professor of Environmental Economics and Econometrics, University of Venice and CEPR Research Fellow

Researcher, Fondatione Eni Enrico Mattei

Director, Climate Change Economics unit at Euro-Mediterranean Center for Climate Change