Governments around the world mostly rely on two types of policy to reduce greenhouse gas emissions from road transport – fuel taxes, which affect driving behaviour, and standards that regulate either average fuel economy or average emissions directly, which affect technology adoption. This column shows that standards have also been effective in inducing clean innovation in car technologies, especially those for electric and hydrogen vehicles. Indeed, the findings suggest that standards were the main policy driver behind clean innovation in the car market over the past two decades.
Recent years have seen rapid changes in the car market in many countries around the world. Hybrid vehicles have become a common sight, and even more radical innovations, such as vehicles powered by electricity, are also quickly penetrating the car market. The global fleet of electric passenger vehicles has exploded from about 17,000 in 2010 to over 40 million in 2023 (IEA, 2024). This column demonstrates that targeted regulation has been a main driver of the radical clean innovations in automotive technologies that made this development possible.
Governments around the world mostly rely on two types of policy to reduce greenhouse gas (GHG) emissions from road transport, the relative effectiveness and efficiency of which have been debated widely (e.g. Anderson and Sallee 2016, Levinson, 2018). The first policy instrument is a fuel tax, which has been shown to affect driving behaviour and innovation (Li et al. 2014, Aghion et al. 2016). The second instrument is a standard that regulates either average fuel economy or average GHG emissions directly, typically at the manufacturer level. Standards have already been shown to affect technology adoption (Klier and Linn 2016, Reynaert 2021). In a recent paper (Rozendaal and Vollebergh 2024), we show that standards have also been effective in inducing clean innovation in car technologies, especially those for electric and hydrogen vehicles. As fuel taxes have been relatively stable in real terms since the year 2000, we find that standards were the main policy driver behind clean innovation in the car market over the past two decades.
Measuring standard stringency
The need for innovation only occurs if other abatement strategies, such as trading off vehicle characteristics and adopting off-the-shelf technologies, would be insufficient or too expensive to comply with an (announced) regulation or loss of sales due to higher fuel excises. A key challenge in identifying the effects of fuel economy and GHG emission standards on clean innovation is therefore developing a measure of standard stringency that captures firms’ incentives to innovate. There are two key characteristics of the instrument that are particularly relevant here and that should be taken into account. The first is its strictness compared to current average performance. A standard only incentivises innovation if it requires firms to improve, i.e. if it poses a binding constraint. The second is its timing. Standards are announced years in advance as targets. The more time a firm has to comply, the less urgent is the need for new or improved technologies. Figure 1 plots (sales-weighted) average CO2 emissions per kilometre and the most recently announced target for the three main jurisdictions that we consider. We call a target binding if it is enforced (non-voluntary) and below current average performance. Interestingly, Japan had binding targets during the first part of the sample period but not in the second part, whereas the EU and the US had binding targets only in the second part of the sample period.
Figure 1 Fuel economy and GHG emission standards and average performance
Notes: Average performance and targets are converted to CO2 emissions per kilometre (NEDC). ‘Most recent target’ represents the most recently announced target level for newly sold cars. These are binding if the target (in the future target year) is mandatory and lower than (current) average performance.
We then define standard stringency as the reduction in grams of CO2 per kilometre that the average firm in a jurisdiction needs to achieve per year to exactly meet the most recently announced target. This measure reflects not only to what extent a specific policy is voluntary or binding, but also the length of the adaptation period and the distance of the existing performance relative to the target. We plot standard stringency for the main countries in our sample in Figure 2. We then go on to weight stringency at the firm level by weighting country-level stringency by firm-country weights based on each firm’s pre-sample patent portfolio as in Aghion et al. (2016).
Figure 2 Standard stringency for selected countries
The effect of standards on innovation
We use our measure of standard stringency to estimate its effects on clean and dirty innovation as measured by patent counts (as in, for example, Popp 2002, Aghion et al. 2016). Our findings clearly show that car innovators have responded to more intensive regulation by governments, in particular to more stringent standards. Excise taxes on fuel also induce clean innovation. We find no evidence for an effect of the tax-exclusive fuel price and R&D subsidies. The effect of standards is particularly pronounced for zero-emission technologies such as electric and hydrogen vehicles and fuel cells. With stricter standards, we can expect that car manufacturers consider existing abatement options to be insufficient to comply with regulation in the future and envisage a clear pay-off to an expansion of investment in R&D. Figure 3 shows the results of a simple counterfactual exercise in which we use our estimates to simulate clean patenting in the counterfactual scenario in which no standards had been in place. We attribute about one eighth of clean patenting over the 2000-2016 period to fuel economy and GHG emission standards. Japan, which had a binding standard for most of this period, contributes about half of these ‘additional patents’, and Europe and the US each contribute roughly a quarter.
Figure 3 Counterfactual clean patents in the absence of standards
Our results give insight into the effects on innovation of various climate policies. While our results do not speak to the economic efficiency of the respective instruments, our results show that standards are effective at inducing innovation in zero-emission technologies. These results are relevant to the policy debate, as governments have in recent years relied mostly on standards to address GHG emissions from road transport. For example, while public support for carbon taxation is low (Fabre and Douenne 2022) and many countries decreased rather than increased fuel taxes during the inflation of 2022, the EU has recently announced that the new target for GHG emissions per kilometre for 2035 is zero, essentially banning the sale of new petrol and diesel passenger vehicles. Our results also show that policies that provide direction by setting clear targets can be complementary to taxes, supporting the view that only relying on carbon taxes may be insufficient to bring about the green transition (Vollebergh 2023, Van der Ploeg and Venables 2023).
Source : VOXeu