The Co-Benefits research programme is led by Doug Crawford-Brown
Assessments of the desirability of decarbonising the economy usually focus on the reduction in greenhouse gas (GHG) emissions. These reductions are related to the predicted reduction in atmospheric GHGs, and the resulting change in climate change risks; e.g. human health effects due to higher summer temperatures or asset loss through increased storm frequency and/or intensity. Both the assessment of policy effectiveness and the political will to carry out mitigation and adaptation policies then rest on a belief in the science of climate change and the reliability of models in predicting the impact of policies on reducing GHG emissions.
There are, however, many reasons why an institution or nation might move along a trajectory of decarbonisation. There might be concern for climate change and its risks. There might be concern for energy security, reducing reliance on foreign energy sourcesor reducing fuel poverty. There might be a desire to reduce energy bills through reducing energy consumption. There might be a government target for economic development, with investment in green buildings, energy systems and/or infrastructure providing a stimulus for this development. Experience shows that moving a goal such as decarbonisation forward is most viable when there are multiple reasons for doing so, removing the need for all actors to have the same concern for climate change and ensuring the widest range of participating actors.
The natural question to ask is whether decarbonisation might bring with it impacts that affect the ability of a nation or institution to reach other economic, energy and/or environmental policy goals, apart from the reduction in GHG emissions. These impacts might be beneficial, helping to reach these other policy goals, or they might present new problems that impede reaching climate change goals. Consideration of these co-impacts is important for several reasons: (i) if they are co-benefits, they can be used to marshal additional political support for policies of decarbonisation; (ii) inclusion of co-impacts provides a more complete understanding of the multiple effects of a policy, improving the ability to inform multi-attribute decisions; (iii) assessment of co-impacts provides a more balanced, macroeconomic perspective on the net costs and benefits of a policy to and within society; (iv) co-impacts may reveal a multiplier effect of policies in the economy, better characterising the beneficiaries of those policies: (v) identification of co-impacts provides insights into where hurdles might be encountered in implementing policies of decarbonisation, and how they might be overcome.
Open Source
4CMR is committed to developing all of its models and the underlying data in open source format so you can obtain, understand and use the bases for our findings.
We also have produced a simple EXCEL file through which the model can be executed. Download the Human Health Model.
If there is anything that is unclear in this material, please contact the primary author and we will clarify the point. We also invite you to collaborate with us in developing future, or even alternative, versions. This feedback to us is essential in improving the open source nature of our work and creating a global, collaborative community of researchers.
1. A macroeconomic model such as E3MG can be used to estimate economic activity, energy demand, and energy provision due to a policy scenario. This is repeated for all regions of the global economy. The same can be done for the baseline (no policy) scenario.
2. This material and energy use from Step 1 results in emissions of air toxics such as PM. By running these first two steps both with and without a climate change policy, the difference in emissions resulting from the policy can be calculated. The fractional change in PM emissions due to the policy is equal to the difference in PM emissions between the “with policy” and “without policy” runs of the model, divided by the PM emissions from the “without policy” run. This is repeated for all regions in the global economy.
3. Empirical factors are developed to convert fractional changes in emissions (Step 2) into fractional changes in ambient air concentration within the Human Health Module.
4. Step 3 is multiplied by region-specific estimates of baseline ambient air contributions to calculate the change in ambient air concentration resulting from a policy.
5. Changes in ambient air concentration (Step 4) are converted to estimates of change in incidence of disease (number of cases) through application of risk coefficients taken from meta-analyses performed on the global literature concerning population sensitivities. This is repeated for all regions in the global economy.
6. Finally, changes in Cost of Illness are determined through application of Cost of Illness (COI) factors for each illness, multiplying these by the results of Step 5 on numbers of cases avoided by the policy.
The particular illnesses considered currently are respiratory and cardiovascular, as the large majority of epidemiological studies conducted world-wide have focused on these effects for exposures to air toxics. Only health endpoints for which the concentration-response (C-R) relationships have been adequately quantified through meta-analysis of world-wide epidemiological studies are included.
The results are estimates of the reduction in cases of disease, and of costs of illness, resulting from GHG emissions reductions, as shown in the figure below for the global economy.
Representative results for the co-benefits of global decarbonisation. 10-4a shows the reduced annual incidence of mortality (global cases per year); 10-4b shows the reduced annual incidence of morbidity (global cases per year); and 10-4c shows the reduced annual Costs of Illness from morbidity effects (2008 USD per year).
The Human Health Module uses a series of simplifying assumptions that reduce the spatial resolution of the modelling so a global analysis becomes feasible (described in the Module manual). To characterise the uncertainties introduced by these spatial simplifications, we have compared the results of the model against those obtained under much higher levels of spatial resolution, performed in collaboration with our partners at National Taiwan University. These results show that the uncertainty introduced is characterised approximately by a lognormal distribution with a GSD of 1.5.
