Marginal Cost Is Not Variable Cost
You run a bakery. Flour, sugar, frosting, boxes, the hourly staff. Rent on the unit, the oven you bought last year, the manager's salary.
Two questions land on you in the same week:
- Is this business making money?
- Should we take the order for 200 extra cupcakes?
They sound like the same question and they use different numbers. Getting the wrong one is how profitable-looking businesses turn down good work and struggling ones take on bad.
Variable cost: what scales with output
Variable costs move with how much you produce. Ingredients, packaging, the staff you roster by the batch, the card fee on each sale. Bake nothing and they're zero. Bake a thousand and they're a thousand times something.
Fixed costs don't move, at least not within the range you're thinking about. Rent, the oven, insurance, a salaried manager. They're the same whether the oven runs all day or not at all.
The line between them is about the decision horizon, not the category. Staff wages are variable if you flex the rota week to week, and fixed if everyone's on a permanent contract with notice periods. The same cost sits on different sides of the line depending on what you're deciding and how far ahead.
Marginal cost: what the next one costs
Marginal cost is the cost of producing one more unit. If 100 cupcakes cost $50 in variable cost and 101 cost $50.50, the marginal cost of the 101st is 50 cents.
Formally it's the slope of the total cost curve — how much total cost rises for a small rise in output. And because fixed costs don't move when output does, marginal cost contains no fixed cost at all. Not "a little bit". None.
That's the whole distinction. Variable cost is a total. Marginal cost is a rate of change. Total variable cost divided by units gives you average variable cost, which is a third thing again and is what most people mean when they say "marginal cost" incorrectly.
The bit the textbook diagram gets right
Marginal cost isn't constant. It typically falls and then rises.
Marginal cost, average variable cost, average total cost
Three curves that get used interchangeably and answer three different questions. Move the price line and watch which one you should be reading.
rent, the oven, salaried staff
Batches to make
59
where marginal cost meets the price
Profit
−$215.67
losing money, but less than by shutting
Covers variable cost from
$15.00
the short-run shutdown price
Covers everything from
$25.54
the long-run break-even price
- Marginal cost
- Average variable cost
- Average total cost
- Price
Marginal cost crosses both averages at their lowest point
Not a coincidence — while the next batch costs less than the running average, the average has to be falling; once it costs more, the average has to be rising. Drag the fixed cost and the average total cost curve lifts and shifts while marginal cost does not move at all.
Contribution $384.33, fixed cost −$600.00
At 59 batches you cover $15.49 of variable cost per batch and $25.66 once the fixed cost is spread over them. That gap is the whole reason a business can be worth running today and not worth being in at all.
Show the numbers
| Batches | Marginal cost | Average variable cost | Average total cost |
|---|---|---|---|
| 10 | $19.80 | $24.60 | $84.60 |
| 25 | $11.25 | $18.75 | $42.75 |
| 40 | $10.80 | $15.60 | $30.60 |
| 50 | $15.00 | $15.00 | $27.00 |
| 65 | $28.05 | $16.35 | $25.58 |
| 80 | $49.20 | $20.40 | $27.90 |
| 100 | $90.00 | $30.00 | $36.00 |
Early on, the next batch is cheap — you're using capacity you're already paying for and getting better at the process. Past a point the oven is the constraint, you're paying overtime, and the next batch costs more than the last. That's diminishing marginal returns turning up on the cost side.
Three things in that chart are worth sitting with.
Marginal cost crosses both averages at their minimum. Not a coincidence. While the next unit costs less than the running average, the average has to be falling. Once it costs more, the average has to be rising. So marginal cost passes through the bottom of each average curve on its way up.
Fixed cost moves the average total cost curve and leaves marginal cost alone. Drag the fixed cost slider. The orange curve lifts; the purple one doesn't twitch. This is the visual version of "sunk costs don't belong in the decision".
Two different prices matter. The price where you'd stop producing today is where you stop covering average variable cost — below that, every batch loses money on its own terms and you should turn the oven off. The price where the business is worth being in at all is where you cover average total cost. Between those two numbers you are losing money and should still keep producing, because you're covering some of the rent you'd owe anyway.
Back to the two questions
"Is this business making money?" Average total cost. Everything in, fixed and variable, divided by units. Compare to price. If price is below it, the business isn't viable as it stands — even if every individual order looks fine.
"Should we take the 200-cupcake order?" Marginal cost. What does making those 200 actually add to your costs? The rent is owed whether you take it or not, so it doesn't belong in the comparison. If the order covers the marginal cost and contributes anything to the fixed cost you're paying regardless, it's worth taking.
What to actually track
Know which costs move. Take your cost lines and sort them by what happens if volume doubles next month. Not what accounting calls them — what actually happens. Most businesses have less variable cost than they think.
Compute marginal cost at the volume you're at. Not on average. The marginal cost of the next unit when you're at 30% capacity and at 95% capacity are different numbers, and the second one includes overtime, expedited shipping, and mistakes.
Keep the two comparisons separate. Price versus average total cost for "should this business exist". Price versus marginal cost for "should we do this one more thing". Mixing them produces confident answers to questions nobody asked.