Cost Behaviour and Cost Estimation for Planning

Learning objectives

• Distinguish fixed, variable, mixed, and stepped costs and explain how each behaves within therelevant range(activity band).
• Construct a linear cost model to predict total cost at different output levels using fixed and variable elements.
• Apply the high–low method to estimate fixed and variable components from historical observations.
• Use cost estimates to support budgeting, forecasting, and short-run decisions, while linking assumptions to operational reality.
• Recognise common pitfalls in cost estimation, including outliers, step changes, and using estimates outside the relevant range that generated them.

Overview & key concepts

Planning depends on being able to predict how costs will move when activity changes. “Activity” might be units produced, machine hours, labour hours, deliveries made, or customer calls handled. A useful cost estimate turns operational expectations (for example, output volume) into a forecast of total cost, helping management set budgets, plan cash needs, and evaluate short-term options.

Cost behaviour is mainly an internal planning tool. In external reporting, expenses are commonly presented in broader groupings (for example, by function such as production and administration, or by nature such as wages and depreciation), so behaviour-based analysis is typically performed separately for management purposes.

Two ideas underpin this chapter:

• The relevant range matters.Many cost assumptions only hold within a specified capacity range.
• Total cost vs unit cost must not be mixed up.Fixed costs are constant in total (within the relevant range) but fall per unit as volume rises; variable costs typically remain stable per unit within the relevant range.

Core theory and frameworks

1) Cost classifications for planning

Fixed costs

A fixed cost stays the same in total across a defined relevant range. Examples include premises rent, fixed salaries, and insurance.

• Total fixed cost: unchanged (within the relevant range)
• Fixed cost per unit: decreases as activity increases

Common exam trap: within the relevant range, a cost is only “fixed in total” if capacity and policy remain unchanged. If contracts, tariffs, staffing policy, or capacity decisions change, the fixed level can reset even if activity stays within the same range.

Variable costs

A variable cost rises and falls in total in line with changes in activity. Examples include direct materials, piece-rate labour, and sales commission.

• Total variable cost: proportional to activity
• Variable cost per unit: usually stable (within the relevant range)

Mixed costs (semi-variable costs)

A mixed cost includes:

• a base element incurred even at low activity, and
• a variable element that increases with activity

For forecasting, mixed costs are typically separated into:

• Fixed element + (variable cost per unit of activity × activity)

Stepped fixed costs

A stepped fixed cost is constant across a smaller activity band, then jumps when extra capacity is required.

Stepped costs must be modelled using an explicit operational rule. For example, if one supervisor is required per 1,000 units, a clear planning rule is:

Supervisors required = ceil(units ÷ 1,000)

This means:

• 1–1,000 units → 1 supervisor
• 1,001–2,000 units → 2 supervisors
• 2,001–3,000 units → 3 supervisors, etc.

Operational caveat: step rules are business-specific. Some organisations staff based on expected peak load, shift patterns, or service-level policies rather than a simple monthly total, so the rule must match how capacity is actually provided.

2) Relevant range

The relevant range is the span of activity where the chosen cost assumptions are expected to remain reliable. Outside it, the cost pattern may change due to:

• capacity constraints and overtime premiums
• additional shifts or equipment
• bulk discounts or supply bottlenecks
• step increases (extra supervisors, extra space)

The relevant range supports an assumption about how costs vary with activity, but it does not guarantee stability if underlying terms change (prices, contracts, staffing policy, or production methods).

3) Building a linear cost model (cost equation)

A practical forecasting model is often written as:

Total cost = fixed element + (variable cost per unit of activity × activity)

Where:

• thefixed elementis the estimated constant total cost within the relevant range, and
• thevariable cost per unit of activityis the estimated incremental cost for each additional unit of the activity driver

This model is most useful when the relationship is approximately linear across the activity levels being planned.

4) Estimating mixed costs using the high–low method

The high–low method uses two observations:

• the period with thehighest activity, and
• the period with thelowest activity

Important: “high” and “low” refer to activity, not cost. Candidates sometimes choose the highest-cost and lowest-cost months by mistake.

High–low can be summarised in two moves:

• First, compare the highest-activity and lowest-activity periods to estimate thevariable cost per unit of activity.
• Then, use either of those periods to back-calculate thefixed elementby subtracting (variable cost per unit of activity × activity) from total cost.

Examiner’s cues

• Choose thehighest and lowest activityobservations (not highest/lowest cost).
• Keep the activity unit consistent (units vs machine hours vs labour hours).
• Calculate thevariable cost per unit of activity first, then substitute to find the fixed element.
• State the relevant range and assumptions that make the estimate usable.

Note: high–low is an approximate method because it relies on only two data points. Where more data exists, businesses often review a scatter of points and may use regression/least squares as an optional extension to improve the estimate.

5) Interpretation and practical sense-checks

A cost estimate should be tested for credibility:

• Does the fixed element align with known baseline commitments (premises, base contracts, minimum staffing)?
• Does the variable element align with operational knowledge (materials usage, energy intensity, waste rates)?
• Will the plan cross any step thresholds?
• Is the estimate being used within the relevant range?

If an estimate fails a basic reality check, it should be revisited rather than applied mechanically.

Worked example

Narrative scenario

A manufacturing company produces widgets and is preparing a monthly budget for the next quarter. Production is expected to be stable, with no major process changes. The business operates within a relevant range of 1,000 to 3,000 widgets per month and currently uses widgets produced as its planning activity driver.

The following cost information applies:

• Factory rent:£10,000 per month (fixed).
• Direct materials:£5 per widget (variable).
• Factory utilities:mixed cost. The standing charge is not separately identified in the invoice, so the company uses historical observations to estimate the fixed and variable elements.
• Supervision:one supervisor is required per 1,000 widgets produced each month. Each supervisor costs £2,000 per month (stepped fixed cost).
• The company plans to produce3,000 widgets per month.

Historical utilities data (both months were within the relevant range):

• At2,500 widgets, utilities cost was£1,500.
• At1,500 widgets, utilities cost was£1,100.

Required

1. Use the high–low method to estimate thevariable utilities cost per widget.
2. Use the high–low method to estimate thefixed utilities cost per month.
3. Build a monthly cost model to forecasttotal cost for 3,000 widgets, clearly showing fixed, variable, and stepped elements.
4. Forecast the total monthly cost for producing3,000 widgets.
5. State key limitations and assumptions in the estimate.

Solution

1) Variable utilities cost per widget (high–low)

High activity: 2,500 widgets, cost £1,500
Low activity: 1,500 widgets, cost £1,100

Variable utilities cost per widget
= (1,500 − 1,100) ÷ (2,500 − 1,500)
= 400 ÷ 1,000
= £0.40 per widget

2) Fixed utilities cost per month

Using the low activity month:
Fixed element = Total cost − (variable cost per widget × activity)
= 1,100 − (0.40 × 1,500)
= 1,100 − 600
= £500 per month

(Quick check using the high activity month: 1,500 − (0.40 × 2,500) = 1,500 − 1,000 = 500.)

3) Monthly cost model for 3,000 widgets

(a) Fixed costs (per month)

• Rent: £10,000
• Utilities fixed element: £500

Total fixed (non-stepped): £10,500 per month

(b) Variable costs (per widget)

• Direct materials: £5.00
• Utilities variable element: £0.40

Total variable cost per widget: £5.40

(c) Stepped supervision cost (per month)

Planning rule: Supervisors required = ceil(units ÷ 1,000)

At 3,000 widgets: ceil(3,000 ÷ 1,000) = 3 supervisors
Stepped cost = 3 × £2,000 = £6,000 per month

4) Forecast total monthly cost at 3,000 widgets

Total cost
= Fixed + Stepped fixed + (Variable cost per widget × Activity)
= 10,500 + 6,000 + (5.40 × 3,000)
= 16,500 + 16,200
= £32,700

Forecast total monthly cost: £32,700

Method marks cue: clear labels (fixed/variable/step) and the activity driver are as important as the final figure.

5) Key limitations and assumptions

• High–low is based on two observations; either month may be abnormal.
• Utilities are assumed to vary with widgets produced; another driver (machine hours, temperature, shift pattern) may explain utilities better.
• The estimate is intended for the relevant range of 1,000–3,000 widgets per month and may not hold outside it.
• Supervision cost depends on the organisation’s staffing policy; step rules may follow peak load or shifts rather than a simple monthly total.
• Stable pricing, efficiency, and process conditions are assumed. If tariffs, contracts, or operating policies change, the model should be updated.

Common pitfalls and misunderstandings

• Choosing the highest-cost and lowest-cost months instead of the highest-activity and lowest-activity months when applying high–low.
• Mixing totals and per-unit figures (especially with fixed costs).
• Ignoring stepped costs or applying an unclear step rule that understates partial blocks.
• Using estimates outside the relevant range from which they were derived.
• Treating a cost as “fixed” without checking whether contracts, tariffs, policy, or capacity decisions have changed.
• Relying on a convenient but weak activity driver and failing to sense-check the implied fixed and variable elements.

Summary

Cost behaviour converts operational plans into cost forecasts. Fixed costs are fixed in total only while capacity and policy remain unchanged; variable costs move with activity; mixed costs combine both; stepped costs increase in blocks when thresholds are crossed. Estimates are most reliable when used within the relevant range and supported by explicit assumptions.

A linear cost model is a practical budgeting tool:

• estimate thevariable cost per unit of activity,
• derive the fixed element,
• apply a clear step rule where capacity comes in blocks, and
• label the activity driver and show clear workings.

The high–low method provides a quick estimate, but it should be treated as approximate and checked against operational reality.

Glossary

Cost behaviour
The pattern describing how a cost changes when an activity measure changes.

Fixed cost
A cost that is constant in total within the relevant range, provided capacity and policy remain unchanged.

Variable cost
A cost that changes in total in line with changes in an activity measure.

Mixed cost
A cost that includes a base element plus an activity-related element.

Stepped fixed cost
A cost that is constant within a band but increases in blocks when capacity thresholds are exceeded.

Relevant range
The span of activity levels over which a chosen cost assumption is expected to remain reliable.

Cost equation (linear cost model)
A forecasting model that expresses total cost as a fixed element plus a variable amount based on an activity driver, with step costs added using a defined rule where applicable.

High–low method
A method that estimates a variable cost per unit of activity and a fixed element using the highest and lowest activity observations.

Contribution
Sales revenue minus variable costs; the amount available to cover fixed costs and generate profit.

Assumption
A condition that must remain true for a cost estimate to be reliable (for example, stable processes, pricing, and contract terms).