Personal Transport: The Number That Settles Most Arguments
Transport is roughly a quarter of final energy use in most developed economies, and personal travel is the larger half of it. Almost every claim made about it - that trains are efficient, that electric cars solve the problem, that cycling is free - is true under some assumptions and false under others, and the assumptions are rarely stated.
Why Passenger-Kilometres Are the Right Unit
Energy per kilometre describes a vehicle. Energy per passenger-kilometre describes what the vehicle achieves for the people in it, and only the second answers the question people actually ask.
The difference is large. A typical European car uses several times more energy per kilometre than a bus, and a bus carries far more people, so the per-passenger figures diverge much further than the per-vehicle ones. Run the same comparison with a bus carrying three passengers at eleven at night and it reverses.
This is why published comparisons disagree so violently. A rail figure computed at average occupancy looks very different from one at peak occupancy, and both are honest. Anyone quoting a single number for a mode has chosen an occupancy assumption, and it is usually not stated.
The useful discipline is to ask what is being averaged over. A system average across a whole year and network is the right number for planning. A specific journey at a specific time is the right number for a personal decision. Neither transfers to the other, and most disagreement about transport energy is two people using different ones.
Where the Modes Actually Sit
At typical European occupancies, the ordering from least to most energy per passenger-kilometre runs roughly: walking and cycling, then electric rail, then coach and bus, then electric car, then combustion car, then domestic aviation.
The gaps are not uniform. Cycling is roughly two orders of magnitude below driving, which no vehicle technology closes. Electric rail at good occupancy is several times better than an electric car with one occupant. An electric car with four occupants is comparable to rail, which is the comparison that surprises people and is the one occupancy makes.
The electric-versus-combustion gap is large and often misstated. An electric motor converts most of its input into motion, while a combustion engine discards roughly two thirds as heat, so an electric car uses around a third of the energy of a comparable petrol one at the wheel. Counting the power station changes this, favourably in most grids and less so in coal-heavy ones - and the comparison shifts every year as generation decarbonises, which is why a figure from five years ago understates it.
There is a category that gets less attention and performs well: the electric two-wheeler. An e-bike or scooter uses a small fraction of the energy of any car and covers the trip lengths that dominate urban travel, most of which are under five kilometres. In several Asian cities they are already the main motorised mode.
Why Cars Stopped Getting More Efficient
Engine efficiency improved substantially over the past three decades. Fleet fuel consumption improved much less, and in some markets it stalled entirely. The explanation is mass and size.
The average new car in Europe gained several hundred kilograms since the 1990s, from safety structures, equipment and a shift toward larger body styles. Moving more mass takes more energy, and the engineering gains were largely spent on carrying it rather than on reducing consumption. This is the rebound argument in one of its clearest documented forms, operating in the product rather than in behaviour.
Regulation contributed. Standards that set targets by vehicle footprint or mass give a larger vehicle a weaker target, which weakens the incentive to build smaller. Several analyses attribute part of the shift toward large vehicles to exactly this design choice in the rules.
The effect carries into electrification. A heavy electric vehicle needs a larger battery for the same range, which means more critical minerals, more embodied energy and more cost. The efficiency advantage of electric drive is real and independent of this; the mass trend reduces how much of it is realised.
What Actually Reduces Transport Energy
The intervention with the largest effect is not a vehicle at all. Trips not taken, or taken over shorter distances, use no energy regardless of technology, and trip length is determined mostly by where things are relative to each other.
Urban form does this. A city where homes, work, shops and schools are close together produces shorter trips and makes walking and cycling viable for a large share of them. A city built around long separations produces long trips that are difficult to serve by anything but a car. These patterns are set by planning decisions and last for a century, which puts them in the same category as the grid topology built in the last one.
Mode shift comes second and depends on the alternative being genuinely competitive on time. People choose modes largely on journey duration, and a public transport option that takes twice as long attracts few riders regardless of its energy advantage. Frequency and directness move ridership; exhortation does not.
Vehicle efficiency comes third, and is the one most policy attention goes to, partly because it requires nothing of anyone. It is genuinely effective within its scope - the electric drivetrain is a large gain - and it operates on the number of kilometres the first two decide. That ordering is worth keeping in view, because it is the reverse of the order in which the three are usually discussed.
Frequently asked questions
What is the right way to compare transport modes by energy?
Energy per passenger-kilometre, which combines how efficient the vehicle is with how many people it carries. Energy per kilometre describes the vehicle only, and answers a different question. Any single figure quoted for a mode embeds an occupancy assumption that is usually left unstated.
Is a train always more efficient than a car?
No. At typical occupancies electric rail is several times better than a single-occupant car, but a car carrying four people is comparable to rail, and a lightly loaded train late at night can be worse than a full car. Occupancy usually matters more than the technology.
How much less energy does an electric car use?
Roughly a third of a comparable petrol car at the wheel, because an electric motor converts most of its input into motion while a combustion engine discards about two thirds as heat. Counting power generation changes the figure, favourably in most grids, and it improves each year as generation decarbonises.
Why has car fuel consumption stopped improving?
Because of mass and size. The average new European car gained several hundred kilograms since the 1990s, and the engineering efficiency gains were largely spent moving that extra mass. Standards that set targets by vehicle footprint or mass also give larger vehicles weaker targets, weakening the incentive to build smaller.
What reduces transport energy the most?
Shorter trips, which are determined by urban form rather than by vehicles - a city where destinations are close together produces short journeys that walking and cycling can serve. Mode shift comes second and requires the alternative to be competitive on time. Vehicle efficiency comes third, though it receives most policy attention.
