Fixed, variable and mixed cost behaviour
Content revised on 3 October 2026. Any completion marker earned before this revision refers to the earlier content version and does not confirm completion of this revised lesson.
Learning goals
- Model total fixed, variable and mixed costs within a stated range.
- Estimate a simple mixed-cost equation from high and low activity observations.
- Distinguish total cost from average cost per unit.
Assumptions and scope
- All entities, data and monetary units (MU) are fictional. No taxes, financing costs or regulatory reporting implications are modelled unless expressly stated.
- Linear behaviour only inside stated ranges; no price changes, abnormal events or mix changes in numerical models.
Describe behaviour relative to an activity
Cost behaviour means how a cost changes as a specified activity changes over a specified period. The activity might be units supplied, deliveries or operating hours. Always name it. A delivery cost can vary with trips even when production units remain constant. Calling a cost variable without naming its driver leaves an important part of the model unstated.
A fixed cost stays constant in total within a relevant range and time horizon. A monthly room rental can be fixed over the range that fits inside the room. Its average per unit falls when more units share the same total. This fall is arithmetic, not a cash saving. A variable cost has a constant cost per activity unit in the simplified linear model, so its total rises with activity. A mixed cost combines a fixed amount with a variable amount: total cost = fixed cost + variable rate × activity. The units must match; a rate per hour must be multiplied by hours, not products.
The relevant range is part of the answer
The relevant range is the activity interval over which the assumptions are credible. Beyond it, another vehicle, supervisor or shift may be needed. Such a step cost is fixed within a band but jumps when a threshold is crossed. Discounts, overtime and congestion can also change the variable rate. Do not extrapolate a straight line indefinitely or assume that zero activity eliminates every cost. Shutdown decisions have their own avoidable costs and commitments.
Classifying behaviour is different from tracing a cost to a product. A cost may be direct and fixed, or indirect and variable. This course uses behaviour to forecast and compare decisions; the Cost Accounting course treats tracing and allocation in greater depth. Keep those questions separate so that a familiar label does not replace evidence about what drives the cost.
Estimate carefully
The high–low method estimates a variable rate from the change in total cost divided by the change in activity between the highest and lowest activity observations. Choose activity extremes, not cost extremes. Subtract estimated variable cost from either observation to find the fixed amount. Both points should reproduce exactly because two points determine the line; that agreement is not proof that the model fits the remaining observations.
Inspect other periods and operational explanations. A one-off repair, changing energy prices or a different product mix can distort the estimate. A scatterplot or a table of predicted and actual costs may expose the problem. With only two observations, treat the result as a provisional model. When preparing a budget, explain the activity driver, time period, range and exceptions alongside the equation. Forecasting discipline means acknowledging where the relationship may fail, rather than presenting an unsupported decimal place as extra certainty.
Worked example: Model a dispatch cost
Monthly dispatch cost is 2,300 MU at 400 deliveries and 3,500 MU at 800 deliveries. Operations confirms a stable mix and a relevant range of 400–800 deliveries. Estimate a mixed-cost equation and forecast 650 deliveries.
Solution
Variable rate = (3,500 − 2,300)/(800 − 400) = 3 MU per delivery. Fixed cost = 2,300 − 3 × 400 = 1,100 MU per month. Cost at 650 deliveries = 1,100 + 3 × 650 = 3,050 MU. This forecast stays inside the observed range; a forecast above 800 needs fresh capacity evidence.
Rows show the same fixed monthly amount and a variable component of 3 MU per delivery.
| Deliveries | Fixed MU | Variable MU | Total MU |
|---|---|---|---|
| 400 | 1100 | 1200 | 2300 |
| 650 | 1100 | 1950 | 3050 |
| 800 | 1100 | 2400 | 3500 |
Independent practice
-
Monthly fixed rent is 1,800 MU, valid for 300–900 units. Find total rent and rent per unit at 600 units.
-
A packaging model is C = 250 + 1.5q MU per month for 100–700 parcels. Forecast 500 parcels.
-
At 100 machine hours cost is 900 MU; at 300 hours it is 1,500 MU. Assume a valid linear relationship between those activities. Estimate fixed cost and the variable rate.
Practice answers and reasoning
-
1,800 MU total; 3 MU per unit.
The fixed total does not change with units within the range; 1,800/600 = 3.
-
1,000 MU.
250 + 1.5 × 500 = 1,000. The forecast activity lies inside the stated range.
-
600 MU fixed; 3 MU per hour.
(1,500 − 900)/(300 − 100) = 3; 900 − 3 × 100 = 600.
Common errors to avoid
- Selecting highest and lowest cost instead of activity for high–low.
- Multiplying a total fixed cost by units.
- Forecasting beyond capacity without adjusting the model.
Exit check
-
A second van is needed beyond capacity. Is the original fixed total still valid?
-
If total fixed cost is unchanged, does a lower fixed cost per unit release cash?
Exit-check answers
-
No.
The relevant range has been crossed; build a new band or model.
-
No.
It spreads the same total across more units.