Uses this dataset. Car makers with less than 10 examined vehicles have not been included.
Car makers by median CO emissions (mg/km)
Manufacturer Emissions CO [mg/km] Ssangyong 92.0 Volkswagen 168.0 Seat 168.0 Lexus 172.0 Audi 173.0 Infiniti 175.0 Mazda 182.0 Skoda 190.0 Volkswagen C.V. 192.0 Volvo 202.0 Honda 204.0 Bmw 206.0 Land Rover 213.0 Mercedes-Benz 215.0 Peugeot 222.0 Mitsubishi 227.0 Toyota 235.0 Nissan 237.0 Renault 237.5 Porsche 255.0 Hyundai 257.0 Jaguar 258.0 Ds 263.0 Kia 265.5 Aston Martin Lagonda 271.0 Mini 298.0 Citroen 302.0 Chrysler Jeep 302.0 Subaru 310.0 Fiat 322.0 Alfa Romeo 327.5 Ford 329.0 Dacia 340.5 Mclaren 354.0 Bentley Motors 401.0 Vauxhall 412.0 Abarth 455.0 Suzuki 469.5 Smart 492.0
Car makers by median sum of emissions (CO, THC, NOx) (mg/km)
Manufacturer Sum emissions [mg/km] Mazda 175.0 Lexus 207.0 Infiniti 229.0 Mercedes-Benz 257.0 Volvo 289.5 Volkswagen 296.0 Skoda 298.0 Mitsubishi 303.5 Bmw 311.0 Toyota 320.0 Porsche 322.0 Volkswagen C.V. 330.0 Honda 334.0 Seat 336.0 Aston Martin Lagonda 341.0 Peugeot 364.0 Nissan 371.0 Land Rover 382.0 Audi 386.0 Mclaren 415.0 Hyundai 425.0 Fiat 425.0 Renault 427.0 Dacia 427.0 Subaru 436.0 Alfa Romeo 441.0 Jaguar 455.0 Kia 475.0 Chrysler Jeep 483.0 Bentley Motors 487.0 Abarth 510.0 Ds 510.5 Ford 519.0 Suzuki 523.0 Mini 529.0 Citroen 530.0 Smart 546.5 Vauxhall 574.0 Ssangyong 578.0
There is at least one surprising result here. Although the median value of the CO emissions for one car maker is among the smallest observed, the amount of emissions increases dramatically when all three categories are considered. This suggests that measuring only CO can give an imprecise picture about the emissions of a car.
We must consider that not all emissions are equally harmful to the environment. Instead of taking the sum of emissions, it would have been more appropriate to take a weighted sum if we know how much harm a given particle does once released. In the past some people suggested that some other particles could be up to 20x more harmful than CO2.
Although there were scandals with manipulated measurements, we can see that the cars of some of the affected companies are not emitting that much, at least when it comes to this dataset. On the other side, car makers that were not discussed at all are showing relatively high emitting rates.
These emission rates might seem small, but we must not forget that they are per km (and measured only on vehicles driven in 2016). If we took the median for all emissions (425.0mg/km) and multiplied it by 100 thousand cars that are driven for 150000km on average, this would result in 6375 tons of emissions that come in the air that we then breathe. But the cars in use are much more than 100 thousand (yet, another factor to multiply with).
Most expensive to drive were
Manufacturer Cost per mile (£) Ferrari, All Models 0.265417 Lamborghini, Aventador 0.260917 Ferrari, All Models 0.250917 Bentley Motors, Mulsanne, MY2017 0.244250 Ferrari, All Models 0.244250 Rolls Royce, Phantom 0.242917 Bentley Motors, Continental, MY2016 0.241667 Rolls Royce, Phantom 0.240417 Aston Martin Lagonda, V12 Vantage S MY2015, fr... 0.239167 Aston Martin Lagonda, V12 Vantage S, MY2016 0.239167 Bentley Motors, Flying Spur, MY2017 0.239167 Bentley Motors, Continental, MY2016 0.237917 Aston Martin Lagonda, Vantage GT12, MY2016 0.231917 Bentley Motors, Continental, MY2016 0.231917 Bentley Motors, Continental, MY2016 0.230750 Rolls Royce, Dawn 0.229583 Rolls Royce, Ghost 0.229583 Aston Martin Lagonda, DB9 GT, MY2016 0.227333 Aston Martin Lagonda, DB9 MY2015, from May 2014 0.227333 Chrysler Jeep, Jeep Grand Cherokee, MY 2014 0.227333 Ferrari, All Models 0.227333 Rolls Royce, Ghost 0.227333 Rolls Royce, Wraith 0.227333 Aston Martin Lagonda, V8 Vantage/V8 Vantage S/... 0.224000
Least expensive to drive were
Manufacturer Cost per mile (£) Volkswagen, UP 0.026250 Citroen, C-Zero 0.028333 Peugeot, iOn 0.028333 Volkswagen, Golf 0.028333 Bmw, i Series, From November 2013 0.029000 Mitsubishi, i-MiEV 0.030250 Renault, Zoe 0.032667 Kia, Soul 0.032917 Nissan, Leaf 0.033917 Nissan, Leaf, 2016 0.033917 Bmw, i Series, From November 2013 0.035500 Nissan, E-NV200, 2014 0.036583 Nissan, Leaf 0.038583 Mercedes-Benz, B-Class, Model Year 2016 0.039750 Tesla, Model S 0.040917 Tesla, Model S 0.042083 Tesla, Model S 0.044833 Toyota, Prius Plug-In, MY2012 0.045833 Toyota, Prius, 2016 0.048833 Peugeot, 208 0.049250 Peugeot, 308 0.050917 Audi, A3 Sportback 0.051833 Ford, Fiesta, Model Year Post 2017 0.052500 Ford, Fiesta, Model Year Pre 2017 0.052500 Hyundai, i20 0.052500 Peugeot, 308 0.052500
As expected, the list of the most expensive cars to drive is dominated by expensive brands like Ferrari, Lamborghini, Bentley, Rolls Royce and Aston Martin. In the list of least expensive cars to drive we see many models. Interestingly, Tesla with its model S already appears on the list.
But, for instance, for the least expensive cars, we can examine which of them have the smallest sum of emissions. Unfortunately, this data is unavailable for most of these cars on the list. But the second Bmw, i Series, From November 2013 has only 59mg/km—the smallest among the values we can actually calculate. Toyota, Prius Plug-In, MY2012 came at 144mg/km, while Toyota, Prius, 2016 came at 181mg/km. Audi A3 Sportback released 74mg/km, while Hyundai, i20 released 542mg/km. Since these values are (mostly) below, but in one case also above the median value, we cannot easily conclude that driving more cheaply is always connected with releasing less particles in the atmosphere.
Combined emissions by engine capacity
We could also divide the combined emissions by the engine capacity, once we normalize both values. The intuition is that we want to find the strongest engine that pollutes the least. The smaller this number, the closer we are to this goal. Here are the results we obtain:
Manufacturer Normalized emissions / normalized engine capacity Mercedes-Benz, CLS Coupé, Model Year 2017 0.062854 Porsche, Cayman 0.103868 Porsche, Cayenne 0.107898 Volvo, XC90 MY17 0.112178 Volvo, XC90, MY16 0.112178 Audi, A3 Sportback 0.123913 Audi, A3 Sportback 0.123913 Porsche, Cayenne 0.179773 Aston Martin Lagonda, Rapide S MY2015, from Se... 0.183458 Aston Martin Lagonda, Rapide S, MY2016 0.183458 Aston Martin Lagonda, Vanquish MY2015, from Se... 0.184627 Aston Martin Lagonda, Vanquish, MY2016 0.184627 Lexus, GS300h, MY2016 0.190694 Lexus, GS300h, MY2016 0.190694 Lexus, GS300h, MY2016 0.190694 Lexus, GS300h, MY2016 0.190694 Lexus, GS300h, MY2016 0.190694 Lexus, LS 0.199680 Porsche, 911 (991) 0.203816 Porsche, 911 (991) 0.203816 Porsche, 911 (991) 0.203816 Porsche, 911 (991) 0.203816 Porsche, 911 (991) 0.203816 Mercedes-Benz, S-Class Cabriolet, Model Year 2016 0.205083 Bmw, 6 Series F12/F13, March 2011 0.212676 Bmw, 6 Series F12/F13, March 2011 0.212676 Bmw, 6 Series Gran Coupé F06 , From June 2012 0.212676 Bmw, X5 Series F15, From November 2013 0.212676 Bmw, X5 Series F15, From November 2013 0.212676 Bmw, X6 Series F16, From December 2014 0.212676
The models on this list tend to appear in groups. This is because the measurements have been made on multiple cars having the same manufacturer and model. For each group the values appear to be approximately the same, which raises the question whether we could detect the type of the car simply from the measurement...
Emissions by model
Sometimes it is useful to compare the emissions of different models from the same manufacturer. But we should not blindly trust these results, since only few cars from a given model could have been included in the data. Use these results only as an orientation and check them against other information sources you come across.