The New Vehicle Efficiency Standards

A word cloud from the previous FES consultation

Yesterday the federal government has released an Impact Analysis paper on its proposed New Vehicle Efficiency Standards, and is seeking public feedback on it. I spent most of 2023 campaigning for this, so I thought I'd do a quick analysis of what they've put forward. 

First a bit of background. The NVES is an example of a Fuel Efficiency Standard, which requires car manufacturers to ensure the average emissions of all the vehicles they sell come under a certain cap. 85% of cars sold worldwide are covered by an FES of some sort, and Australia is a real outlier for not having one. The Climate Council have a good FES explainer here, but the general idea is, you set a CO2 limit, then ratchet it down over time, so emissions come down, until eventually all new car sales are zero-emissions (mostly EVs, maybe some hydrogen). 

Notably, Fuel Efficiency Standards are good under the Reduce Replace Repower framework, because they very neatly incentivise electric cars over fossil cars, without incentivising cars in general over other modes (as things like direct cash subsidies can). This makes it an important policy even if, like me, you want to see a substantial mode shift towards walking, cycling and public transport. 

Lastly, I learned a lot about this policy area while working in a previous job, but all the usual disclaimers apply: this post is my own opinion and doesn't necessarily reflect those of my former or current employers. 

The table detailing the three options

The government have modelled three different options: Option A, which is the weak option that some in the automotive industry have lobbied for; Option B, which is the government's preferred model of FES; and Option C, which is a more ambitious FES. Option A also includes a bunch of super-credits which the others don't. In all cases, vehicles are divided into Passenger Vehicles and Light Commercial Vehicles (more on this later). 

The scenarios assume a start date of 01 January 2025 (which is probably as early as could be expected, given the delays in getting to this point) and are only modelled out to 2029 - and no scenario would reach zero emissions by then. There is a commitment to review the scheme in 2026 and adjust settings as needed. 

I think this is a reasonable approach from a political perspective; there is always the risk that a future government will fail to take the next step (or wind back this one) but hopefully this sets a trajectory without scaring the horses. If they extended the modelling and included an end date for the sale of fossil cars, I suspect opponents would focus on that end date and scream about the government "banning cars" - it'd be "end the weekend" all over again. 

Nonetheless, it's important for climate advocates like me to understand whether these options would put us on track to phase out fossil cars in a timely manner, so I've extended their projections to see how the different scenarios stack up. 

The government's proposed CO2 limits don't follow a linear trajectory (cuts are deeper in some years than others) so I've projected in two different ways. Trend A takes the average emissions cut across all five years and projects it till we hit zero; Trend B assumes the cut in the final year will continue. 

Projections of the three scenarios for passenger vehicles

As you can see, Option A would see Passenger Vehicle emissions continue at a high level for quite a long time, ending somewhere between 2036-2040. Option B would phase them out between 2033-35, while Option C phases them out in 2031 no matter how you slice it. 

If you look solely at the end-date, best-case Option A is quite similar to worst-case Option B (2036 vs 2035) but it's important to remember that the total emissions would be substantially different, since Option B makes bigger cuts earlier. Option C is obviously much better on both counts. 

Projections of the three scenarios for LCVs

For Light Commercial Vehicles, Option A sees high emissions continuing basically indefinitely (my projections reach zero between 2061 and 2072). Option B is pretty similar to the Passenger category, with a range of 2033-36, while Option C is, bizarrely, potentially both more and less ambitious, being anywhere from 2031-37 depending on how you slice it. 

Given that vehicles hang around on our roads for 10-20 years, we need 100% of new sales to be EVs by 2035 at the absolute latest (and ideally well before) and Option C is the scenario that gives us the best chance to achieve this. By making big cuts early, it also saves by far the most of our carbon budget, buying us time to make more difficult reductions in later years. 

Incentivising bigger vehicles

The weaker standards for LCVs are potentially pretty concerning because of the shift in our vehicle fleet's makeup. The document itself shows that even since 2017, the number of passenger car sales has halved; most of those sales have shifted to SUVs (which thankfully are part of the Passenger category) but some went to LCVs, of which more are now sold each year than passenger cars. So LCVs aren't just a niche; the standards set for them will have a substantial impact on our total emissions. 

Number of cars sold in Australia by type

There is real potential for perverse incentives here. Having separate Passenger and LCV categories, with looser standards for LCVs, incentivises manufacturers to push customers away from normal passenger cars and into LCVs even more so than they already do. This is exactly what happened in the USA with its CAFE standards - utes and SUVs could be classified as "light trucks" and be subjected to weaker regulations, so manufacturers category-jumped to these vehicles. 

What's more, within each category, there are 'weight curves', so the actual g/km of CO2 will vary depending on how much a vehicle weighs, with heavier vehicles being allowed to emit more. The report says at the start that reducing vehicle weight is one strategy manufacturers can use to reduce emissions, which is correct - but having weight curves instead of a flat standard will disincentivise them from doing that. 

The paper suggests ways manufacturers can improve fuel efficiency

The weight curves are capped (above a certain weight, the CO2 limit is flat), but to give you an example, the LCV cap is 2200kg, and a single-cab Ford Ranger weighs 1789kg. A full-spec Raptor weighs 2431kg, but clearly there's still a lot of room to move for the base model before the cap becomes an issue. 

We should be under no illusions that the way they've set these categories up is a significant concession to the automotive industry, and has the effect of weakening whatever headline CO2 figure they end up with. 

Assessing the benefits 

The government has included a cost-benefit analysis of the three options. This analysis concludes that for every dollar spent on Option B, their preferred scenario, it would return $3.08, while the more ambitious Option C would return $2.96. Even on the face of it this seems close enough to make no difference - it's within the margin of error to say "about $3" for both scenarios. And given that Option C would deliver substantially more emissions reductions for a comparable BCR, surely that makes it the clear winner. 

A table showing the costs and benefits of the three options

When you dive into the assumptions that underpin the analysis, I think it makes the case for Option C even stronger. For starters, "reduced vehicle maintenance" assumes a 40% cut. Based on dozens of conversations I've had with EV owners and experts, this seems very conservative, they've indicated more like a 80-90% cut. But you can't run a model on anecdotes, so let's give them that one. 

The biggest line items relate to powering the vehicles - fuel cost savings, and the cost of electricity. But the assumptions made here seem weighted quite heavily towards ICE vehicles. Their petrol price estimates are based on 2022 data, and assumed petrol to be $1.76 in 2025; in February 2024 it is already well above that (when I checked petrolspy.com.au last night, most places were around $2.18). It assumes linear growth in prices out to 2050, when we all know that petrol prices are anything but linear - hell, reducing Australia's exposure to fuel price volatility is one of the paper's stated benefits for an FES. 

Petrol station (via psimon2011)

Their electricity prices are also based on assumptions about grid electricity that aren't necessarily representative of EV charging costs. Polling suggests most EV owners currently charge basically for free with rooftop solar. This is probably a demographic artefact of the tendency for EV owners to be relatively-wealthy, climate-friendly folks who own a detached home with solar panels on it, so it is likely to shift over time (eg as more renters get second-hand EVs). But about a third of Australian houses already have rooftop solar, a number that will only keep growing, so it's reasonable to assume that a large chunk of EV charging will remain cheaper than grid tariffs in future. 

The government provided stress tests for some assumptions in the report, but not these ones. I'd argue that more robust assumptions about petrol and electricity prices would make Option C the clear winner. 

Conclusion 

Overall I'm really glad to see the government finally moving forward with the NVES. It's a hugely important reform and I'm glad they've stuck with it. 

It's a shame that they've opted for this two-stream model of Passenger and LCV standards, and that they've used weight curves; I'd much prefer a simpler, flatter system, but unfortunately I don't see them making such a fundamental change given the short consultation window. 

Given the weaknesses of the model they've chosen, it's important that they go for the most ambitious version of it - Option C. 

Thankfully, it seems quite clear that this would be the most beneficial option for the people of Australia anyway, so hopefully the government will be open to it. 

You can read the full paper here, and provide your feedback here.