Estimating the costs of vehicle efficiency: Will the European Commission incorporate lessons from experience?


Earlier this year, I wrote about estimating future vehicle technology costs, to make the observation that in reality meeting CO2 reduction targets has consistently ended up being cheaper than originally anticipated. This conclusion was based on our long standing experience in Europe, where we saw a round of studies in preparation for the 2015 cars’ CO2 target of 130 g/km, then the 95 g/km target for 2020, and where we are now turning towards setting a target for 2025 vehicles. The point was that for the 2015 target we already know what the actual cost of meeting the 130 g/km has been, and that this cost turned out to be only about one-third of the original estimate.

At the end of July, the consultancy Ricardo-AEA, retained by the European Commission’s Directorate-General for Climate Action, presented their set of cost curves for the post-2020 time period at a stakeholder meeting in Brussels. These cost curves – and the underlying analysis – for the time being are only a draft version, and have not been published online yet. Nevertheless, I thought it might be interesting to take a look at some high-level results and how these compare to the findings of other studies.

For this, I took the same graph as in my January blog, with the CO2 reduction in grams per kilometer (g/km) from a 2010 baseline plotted on the x-axis and the additional direct manufacturing costs on the y-axis. Adjusting the various studies so that they all refer to a 2010 baseline is not a straightforward task, but still, the resulting graphical comparison should allow for a fairly accurate comparison of the cost curves to each other.

Figure 1. 2015, 2020 and 2025 cost curve estimates for passenger cars in the EU.

At the beginning of the year, the IKA study on behalf of the German Ministry for Economics [IKA, 2015] was still the only study looking at technology cost beyond 2020, estimating a cost level of more than €3,500 for meeting a 68 g/km CO2 target by 2025. Now, with the Ricardo-AEA study [AEA, 2015, not yet available online] on behalf of the European Commission, we have a second point on the graph, at around €2,000 per vehicle for meeting a 68 g/km target. The fact that the IKA figure is so much higher than AEA’s is not surprising: the same was true for their 2020 estimates (almost twice as high for passenger cars and 15 times as high for light-commercial vehicles). The underlying reason is IKA’s strong reliance on selected manufacturers’ data, as well as some questionable aspects of their methodology. The AEA assessment was carried out in a more rigorous and independent way.

It should be noted that the CO2 reduction potential plotted refers to the current NEDC test procedure. As we know from previous studies, vehicle manufacturers are increasingly exploiting loopholes in the NEDC, so that the actual CO2 reductions achieved in recent years are less than originally anticipated. With the introduction of the new WLTP, this should change to some extent. But this also has implications on the CO2 cost curves, as the same assessment for WLTP will result in a lower reduction potential for the same cost or a higher cost for the same CO2 reduction – a fact nicely illustrated in AEA’s report for the European Commission.

On the other hand, the current draft version of the AEA report does not apply a correction factor for technology cost estimates. Even though the report looks at a number of technologies and comes to the conclusion that on average ex-post costs come out at about half of the ex-ante cost estimates, it does not apply this factor when deriving the set of cost curves for 2025. On the contrary, AEA adjusts their technology cost estimates upwards by adding a 3% manufacturers’ profit margin on top. This means that, should in the future production costs for vehicles increase because more technologies are added, the profit of vehicle manufacturers would automatically increase as well. However, as it should not be the objective of CO2 standards to increase profits of manufacturers, this manufacturers’ margin should definitely be excluded from the analysis.

Also problematic is the fact that in AEA’s report to the European Commission, the future costs of hybrid vehicles seem particularly high. As my colleague John German concluded in a recent summary paper, “full function P2 hybrids are likely to be half the cost of 2010 systems before 2025, without considering the additional hybrid cost reduction enabled by vehicle weight reduction” – a development not really taken into account in the current version of AEA’s report.

The European Commission’s cost curves might still change in the next few weeks, as Ricardo-AEA incorporates feedback from stakeholders. And ICCT’s own EU technology cost curve assessment for the post-2020 timeframe is still ongoing. So there are more results to come from Europe – late in 2015 and then also in 2016, when the European Commission is expected to present details on their strategy to drive road decarbonization forward.