Car makers claim they’re making big improvements to fuel economy and emissions, but it’s all a load of old cobblers, says Paul Horrell.

ON PRACTICAL MOTORING recently there was a story on the failure of new cars in Australia to improve fuel-economy fast enough to suit the ambitions of environmental group ClimateWorks. The lobbyists said Europe is doing much better.

Well I have bad news for them. Europe might look like it is doing well. But it isn’t at all. Much of the improvement is illusory. By that line of reasoning, I expect Australia is actually going backwards.

Even if you personally don’t care about the climate-change implications of CO2 emissions, you probably do care about the cost of fuelling your car or ute. Note then that for petrol-engined vehicles, CO2 is directly related to fuel consumption. For diesels it’s also proportional, although with a slightly different multiplier.

Indeed, in the official fuel-consumption test the lab-rats don’t even measure fuel consumption – they measure tailpipe gas and work it out from there.

ClimateWorks grumbles: “The improvement in fuel efficiency of new cars in Australia has dropped to its lowest rate over the last 10 years, with improvement of only 1.1 per cent achieved in 2016 compared to an improvement rate of 4.2 per cent in 2006 and 3.7 per cent in 2012.”

European emission levels are officially falling much faster – by about four percent a year up to today. That’s the figure the carmakers and their lobbyists in Europe put about.

But they are talking – how to put this politely? – a load of old cobblers.

You might have heard of the ICCT or International Council on Clean Transportation. It was them who first exposed the VW diesel NOx cheating. the following year they did something else. They looked at how Europe’s real-world fuel economy lined up against the official figures. They also plotted historical trends.

In 2001, the average European test CO2 figure was 170g/km. The ICCT found that in that year, real-world consumption was only about 10 percent worse than the official figure. ‘Official’ being the figure you see in car adverts as well as websites like ours that are more or less obliged to publish it.

By 2015 Europe’s official average was down to 123g/km. Three cheers. Well no cheers actually because the improvement wasn’t really there. The real-world figure was a staggering 40 percent adrift of the official number, which means in fact there has been precious little on-road improvement at all.

The ICCT didn’t just go out and test a few cars and throw out a headline-grabbing conclusion. It gathered data from some statistically significant sources, totalling more than half a million vehicles.

For example it asked fuel-card payment firms that serve company-car drivers. Think about it. They know exactly how much fuel the driver pumps in. And how many km he or she drives. And they know the car. And they can plot this over the whole time the driver keeps the car and is in the same job.

The report also collates test economy figures from reputable consumer road-test publications. Same result. The official figures became, more and more, a fantasy.

ICCT also tracked services where users can go to web sites and input their own measurements of their own car’s consumption. I don’t tend to trust these because generally only the evangelically happy or the terminally grumpy owners will bother, and you can’t be sure how many of each camp there are. But guess what? The increasing divergence showed up there to a very similar extent as those other data sources.

I did my own little data analysis too. In my job I normally keep a car for six months or so as a long-term tester. Could be a tiddler or a luxo-barge, petrol or diesel, any body style. I take what what I’m given, drive for several thousand km, clean it up and hand it back. I always diligently record the fuel I’ve put in.

And sure enough I’ve found that despite the improvement in official number, the consumption has flatlined for two decades. Example of comparable cars might help: in 1995 I ran a Golf VR6, all 2.7 litres and six cylinders. I drove it like a maniac because I was young and there was little speed enforcement in the UK. I got 10.4l/100km.

In 2012 I had a Ford Focus ST. By this time I had a precious two-year-old in a safety seat sitting behind me, and the speed limits had become far more rigidly enforced. I got 9.8l/100km. Scant improvement there. In 2015, I ran a a Mini Clubman Coooper S All4. That did better, but not much: 9.0l/100.

Basically, small petrol cars net me about 7.5l/100km, and have done for two decades. Bigger petrol cars do 9 or 10l/100km. Small diesels around 6, bigger ones maybe 7, towards 8 for mid-size diesel crossovers. I’m currently getting 7.4 out of a Volvo V90 that’s supposed to do 4.9.

However hard I stare at the chart, I can’t see any favourable trend in consumption. That’s over two dozen cars and a million km.

But there is a clear trend for the divergence between the claim and the actuality. Lately they’ve almost all all been drinking at least 50 percent more than they should.

Why? Because manufacturers have to fix the test, or they fail to compete commercially. Their buyers will be taxed more. They themselves will face sanctions.

How do they fix the test? I figured it was all the ways you normally read about that help a lot in the test but little on the road. Think downsized turbo engines with too few cylinders and too many gears (who honestly wants a nine-speed?).

But it’s worse than that. Manufacturers actually cheat, or at least do things you and all would call cheating but are within the rules.

They tape up the air intakes, over-inflate the tyres and heat-harden their rubber, use ‘prototype parts’ (hem hem). They perform the ‘coast down’ test, which establishes rolling resistance before they go on the dynamometer, on a glossy marble surface. Thus so the car slows down more gently, and then and less resistance is plugged into the dyno. They make sure that all the allowed tolerances in the car and the testing machinery are totalled up in a favourable direction. They supply cars to the test with the standard-fit 40-litre fuel tank in the full knowledge that all owners will specify the heavier, no-cost-option 65-litre tank. And many more.

You might hope that the new Worldwide Light Vehicle Testing Procedure (WLTP) will improve matters. Maybe it will for a while. But doubtless they’ve got engineerings working already on fiddles for that one.

They’re not fools. But they are deceivers.


2018 Nissan Leaf teased again... showing autonomous drive ability


Jaguar Land Rover demonstrates its Autonomous Urban Drive technology

About Author

Paul Horrell

Paul's working life has been paced out in cars. He began road-testing when the VW Golf was in its second generation. It's now in its eighth. He covers much more than the tyre-smoking part of the road-test landscape. He roots around in the financial machinations of the car corporations and the apparent voodoo of the technologies. Then he clarifies those complications so his general readers – too busy to lodge their heads up the industry's nether regions – get the fast track on what matters and what doesn't. A freelance writer living in London, he usually gets around the city by bicycle, which adds to his (sometimes justified) reputation as a bit green and a bit of a lefty. He's a member of Europe's Car of the Year jury.


  1. From what I can see, most of the improvements in efficiency are cancelled out by the increase in size and weight of a given vehicle over the years….which probably explains why manufacturers are resorting to gaming/cheating the tests to make the new model look better than the old one in this regard….

    1. Agreed. The Golf VR6 (the Mk3 golf body shell) was the same size as the current Polo. The new Polo, just announced, will be bigger again.

  2. Part of the problem is the utterly unrealistic demands by authorities. Somehow they expect vehicles to disobey the laws of physics. And yes, manufacturers keep adding size and weight. So the 2003 Subaru my daughter drives is way smaller than the current model. Have a look at an HR Holden compared with the current (real) Holden. It looks tiny. My XW Falcon weighed around 400 kg less than my FG. Admittedly I was getting around 15 mpg, but that is a 46 year old design. I carted 4 kids around (illegal now) and it went fine. Why did manufacturers keep making bigger cars? In the end it backfired as people have deserted large cars in droves.

  3. it is all a bit ridiculous, if I can still get around 9ishL/100km out of my 14 year old, 250,000km AWD Magna (highway use), with a 3.5l V6, why can’t these newer, smaller capacity, boring-as-hell-to-drive CVT cars do much better? all these crazy new improvements, and for what?

  4. Conversely – during my student times I used to own a basic Nissan Sunny Diesel which consumed about 9L back then (plus a litre of oil per 1k). Today I have a Mitsubishi Pajero Sport which is achieving the same (minus the oil). Of course this means that I’m polluting as much as 25 years ago but it also shows the progress in engine evolution. On equal grounds (same size, weight, power), today’s cars consume much less. Maybe not as much as the manufacturers want us to believe but it is very substantial.
    Sorry to say but this article is just nonsense.

    1. Could it possibly be that the Sunny diesel, as shown by its huge thirst for oil, wasn’t exactly operating at design efficiency? Comparing a rooted old heap with a well-maintained new vehicle is hardly fair. But that’s a side issue.
      Yes weight has increased and engine efficiency has improved to compensate. The point of the article is the increasing divergence between claims and reality, and i can’t see how what you have said contradicts that point. So I’m not sure why you accuse my article of being rubbish.

      1. The issue with fuel consumption is not caused by the manufacturers – it’s us. We want to super-size everything these days. We want the latest safety gear that increases weight. We want more performance. They just deliver on these demands.
        If you don’t like that Nissan Sunny example – my first car was a VW Golf I. 50HP. 10L real world fuel consumption. Curb weight: 790kg. 10L is downright surreal for these specs by today’s standards (if you can find a car with a 50HP engine). A more recent example is a Mini Clubman Cooper (1st gen) – also 10L – but obviously already substantially heavier and more HP.

        1. I agree absolutely. We are asking the impossible if we think we can get improved economy with increased mass and power. But the point of my article is that the manufacturers are pretending that we can get the impossible. To repeat: the point of the article is the increasing divergence between the official figure and reality.

          1. Honestly I’m not sure whether this is fully true. You will not reach those figures in real life – yes. If you don’t drive like crazy, switch off the aircon and have no crap in the boot, you can come fairly close though. Nobody does that though.
            In any case – it is not the point of the official figures to reflect reality. The point is that you can compare different models because they were tested under exactly the same conditions. Yes, they tape the air intakes but they all do thus the testing conditions are the same.
            Based on my own driving style, I will have to add 20-30% on top for petrol or 10-20% for diesel engines. Knowing that I am not stressed at all by the “official figures”.

          2. Which proves my point about increasing divergence from the official figures. In your words: ‘you will not reach those figures in real life’ and ‘I will have to add 20-30% for petrol and 10-20% for diesel’. That is exactly the divergence I’m talking about. And 25 years ago that divergence didn’t exist. You keep reinforcing my argument while saying you disagree.

          3. How do you explain that people achieve fuel consumptions below the manufacturer claims during fuel efficiency challenges? According to your logic that shouldn’t be possible.

            Enter “Volkswagen Think Blue Fuel Challenge” in Google.

            And sorry – I still don’t buy into your argument that everything was better 25y ago. It does certainly not match with my experience from back then.

          4. It’s still just possible to use less fuel than the official figure. But 25 years ago it was easily possible.

            At what point did I argue that ‘everything was better 25 years ago’?

            This whole thread is based on you inventing a series of things I didn’t say, then arguing against those invented things.

          5. On a lighter note, my wife and I bought a Suzuki Swift back in the eighties, can’t remember the year, but we achieved 22 klms per litre on a manual going to Cairns before cruise control days, it cost us $300.00 round trip to Stroud, NSW

            I just wish for that engine to return

  5. I find in contradictory that governments insist on fuel economy improving and then do such things as putting traffic lights with extremely long cycles in, installing speed humps (20 km/h recommended on a 50 km/h street) and creating obstacles (chicanes) – all of which defeat the improvements in fuel economy.

    1. Long-cycle lights don’t effect economy when you have idle-stop for the engine, as most European cars do now.

  6. Just like to point out that not all manufacturers are telling porkies.
    My 2008 Mitsubishi Outlander 3L v6 was rated at 10.6L/100Ks, and achieved that over 5 years of ownership. Next car, a 2012 Dodge Journey 3.6L V6 in a much heavier car, also rated 10.6L/100 Ks, achieved 10.9L over 4 years, mainly slow suburban use. My new car, a 2017 Peugeot 2008, a 3 cylinder 1.2L Turbo, is rated at 5.9L/100Ks, and achieved 5.7 over 3922Ks from Melbourne to Brisbane and return, mainly @ 110 on freeway/highway cruise control and a week running around Brisbane.

Leave a Reply

Your email address will not be published. Required fields are marked *

Check Also