Where each factor comes from
The car rows are not estimates. Burning a gallon of gasoline releases about 8,887 grams of CO2 and a gallon of diesel about 10,180 grams; these follow from the carbon content of the fuel and the fact that essentially all of it leaves the tailpipe as carbon dioxide. The EPA publishes both figures and they are the basis of the greenhouse-gas numbers on new-car labels. Divide by miles per gallon and you have grams per mile, with no room for a manufacturer to be optimistic about anything except the MPG itself.
That is why fuel economy is the whole story for a combustion car. A 45 MPG hybrid emits 197 grams a mile and a 20 MPG truck emits 444, and the ratio is exactly the inverse ratio of their MPG. Anything that changes real-world economy — a roof box, cold starts, short trips, winter fuel, aggressive driving — changes emissions by the same proportion.
The electric row works differently. No carbon comes out of the car, so the emissions are wherever the electricity was made, and that is why the grid factor is a field you fill in rather than a constant. Published emission rates for US grid subregions differ by roughly an order of magnitude between the cleanest and the most coal-dependent. The same electric car, driven identically, is a genuinely different emissions story in the Pacific Northwest than in parts of the Midwest.
Per-passenger-mile figures are averages of a crowd
Bus and rail emissions cannot be stated per vehicle in a way that means anything to a commuter, so they are published per passenger-mile: total emissions from operating the service divided by the passenger-miles it carried. The consequence is that the number describes the system, not your trip.
| Situation | Effect on the per-passenger figure |
|---|---|
| Peak-hour service, vehicle full | Far below the average — the same fuel is spread over many more people |
| Off-peak or lightly used route | Can exceed a single-occupant car |
| Electric rail on a low-carbon grid | Very low, and falling as the grid changes |
| Diesel commuter rail or express coach | Higher than urban electric rail, still normally below a solo car |
| Service that would run anyway | Your marginal emissions for boarding are close to zero |
That last row is the one worth sitting with. A bus on a fixed schedule runs whether or not you are on it, so the emissions caused by your decision to board are nearly nothing, while the emissions attributed to you by an average-based factor are not. Average-based accounting is the right way to compare systems and a poor way to describe a single trip. Both framings appear in published figures, which is part of why agency numbers disagree.
Carpooling divides everything
A car emits the same whether one person or four are in it, so two people in a car halves the per-person figure and four quarters it. Two people in a 25 MPG car are at 178 grams per passenger-mile, which is already better than many bus routes at their published averages. This calculator assumes you are driving alone because most commuters do, but if you carpool, divide the car rows by the number of people and the ranking can change completely.
What sits outside the boundary
Everything here is operational: fuel burned and electricity consumed. Making the vehicle is not counted, and for electric cars it is substantial — battery production means an EV starts its life behind a comparable gasoline car and catches up over some tens of thousands of miles, with the crossover point depending heavily on the grid it charges from. Building and maintaining roads, bridges, tracks and stations is not counted for any mode. Neither is the energy used to extract, refine and truck fuel to the pump, which adds roughly a fifth again to gasoline on most published life-cycle accounts.
These omissions are consistent across the modes in one sense — none of them include infrastructure — and inconsistent in another, since the modes differ enormously in how much infrastructure and manufacturing they need per passenger-mile. Use these numbers to compare your own options against each other, which is what they are good for. Do not compare the total against a figure from a life-cycle study, because it is measuring a smaller thing.
Questions people ask
Is an electric car always cleaner than a gasoline one?
In operation, on any US grid, essentially yes, and the calculator will show you why: it prints the grid emission factor at which the electric row would match your gasoline row, and that break-even is normally far above anything a US region actually runs at. What varies is by how much. On a hydro- and nuclear-heavy grid an electric car can be under 30 grams a mile; on a coal-heavy one it can be over 200. That is a difference of a factor of seven for the same car, which is why the grid factor is a field rather than a constant. Separately, operational emissions are not the whole picture — battery manufacturing puts an electric car behind on day one and it takes tens of thousands of miles to draw level.
Why is the bus factor editable instead of built in?
Because there is no correct value to build in. Per-passenger-mile emissions depend on the vehicle, the fuel, the route and above all the occupancy, and agencies with published figures differ by more than a factor of two. A crowded urban trunk route and a rural service running mostly empty are the same mode and not remotely the same number. If your agency publishes a figure, use it. If not, the placeholder is a rough middle for US urban bus service and should be treated as such rather than as a fact about your bus.
My car has a different MPG in summer and winter. Which do I use?
The annual average of what you actually get, which for most people is a few MPG below the combined rating on the label. Winter economy drops from cold starts, denser air, winter fuel blends and more idling, and short commutes never let the engine reach operating temperature at all, so a five mile commute can run 20 to 30 percent worse than the rating. If your car reports a long-term average, use that. Electric cars swing harder still, because cabin heat comes out of the battery rather than from waste engine heat, and winter efficiency 30 percent below summer is ordinary.
Does cycling really have zero emissions?
Zero from the bicycle, which is what this page counts. The complete picture includes manufacturing the bicycle, which is small and spread over many years, and the food you eat to power it, which is not nothing and depends entirely on your diet. Published attempts to price cycling in food terms land at a small fraction of driving even on a meat-heavy diet, and at close to nothing if the calories would have been eaten anyway. The honest statement is that the number is not exactly zero, it is small enough that the difference from zero is well inside the uncertainty on every other row of the table.