Methods of Segregating Mixed Costs in Marginal Costing

Every organisation deals with costs that refuse to sit neatly in one box. A machine’s electricity bill, a delivery van’s running expenses, or a factory’s maintenance charges often carry both a fixed slice and a variable slice within the very same bill. These are called mixed costs, or semi-variable costs, and marginal costing cannot function properly until they are pulled apart into their fixed and variable components. This post walks through why that separation matters and the three methods most commonly used to achieve it.

What exactly is a mixed cost? 🔗

A mixed cost has one portion that stays constant regardless of output and another portion that moves in step with production or sales volume. A telephone bill is a classic example: the monthly rental is fixed, while charges for calls made vary with usage. Depreciation calculated partly on a straight-line basis and partly on machine hours used behaves the same way, and so does power consumption, where a minimum demand charge stays fixed but consumption-linked charges rise with production.

Because marginal costing depends entirely on classifying every cost as either fixed or variable, a cost that behaves like both categories at once creates a genuine problem. Contribution, break-even point, and profit-volume ratio all depend on the variable portion being isolated accurately. If part of a fixed cost is mistakenly treated as variable, contribution gets understated, and if the reverse happens, decisions such as accepting a special export order or dropping a product line end up built on shaky numbers.

Common mixed costs you will encounter 🔗

• Electricity and power costs: a fixed minimum demand charge plus a per-unit consumption charge.
• Telephone and internet bills: a fixed rental combined with usage-based charges.
• Repairs and maintenance: routine upkeep that stays roughly constant plus wear-related repairs that rise with machine usage.
• Salesmen’s remuneration: a fixed monthly salary topped up with commission linked to sales volume.
• Depreciation: a time-based component alongside a usage-based component for certain assets.

Why segregation cannot be skipped 🔗

Segregating mixed costs is not just an exercise for university exams; it directly feeds several managerial decisions. Once the variable portion is known, a business can calculate contribution per unit, work out the break-even point, and prepare flexible budgets that adjust automatically as output changes. Classifying costs clearly into fixed and variable components is what makes cost ascertainment and reporting under marginal costing possible in the first place, which is exactly why this topic gets so much attention in the syllabus.

Three techniques are commonly taught for this purpose: the analytical method, the high-low method, and the scatter diagram method. Each trades off simplicity against accuracy in a different way, and knowing when to reach for which one matters as much as knowing the mechanics.

The analytical method 🔗

The analytical method is the simplest of the three, and also the least scientific. Here, an experienced cost accountant or manager studies the nature of a mixed cost and estimates, based purely on judgement, how much of it behaves as fixed and how much as variable. One semi-variable cost might be judged to carry 60 percent variability while another is judged to carry only 40 percent, depending entirely on how well the analyst understands the underlying process.

This approach works well when quick estimates are needed and formal historical data is not yet available, such as during preliminary budgeting or when a new cost item first appears on the books. Its biggest drawback is subjectivity. Two analysts looking at the same cost can reach different conclusions, and the accuracy of the split is only as good as the experience behind it. Because of this, the method is rarely relied upon on its own for decisions where precision genuinely matters.

The high-low method 🔗

The high-low method takes a more structured route. This technique was developed by J.H. William and rests on a simple logic: since the fixed portion of a mixed cost does not change with activity, any difference in total cost between the highest and lowest activity levels must be caused entirely by the variable portion.

How the calculation works 🔗

The method uses only two data points picked from a set of historical records: the period with the highest output and the period with the lowest output. Comparing total costs at these two activity levels isolates the variable cost per unit, and once that figure is known, the fixed cost is found by subtracting the total variable cost from the total cost at either the high or the low point. Both calculations should arrive at almost the same fixed cost figure, which works as a handy built-in check on the arithmetic.

Consider a factory tracking its power costs over several months:

Step 1 – Variable cost per unit: (82,000 − 52,000) ÷ (5,000 − 2,000) = 30,000 ÷ 3,000 = ₹10 per unit.

Step 2 – Fixed cost: 82,000 − (10 × 5,000) = ₹32,000. Cross-checking with the low point: 52,000 − (10 × 2,000) = ₹32,000. Both figures match, which confirms the split is consistent.

The high-low method is quick to apply and needs no special software, which is why it stays popular for exams and for rapid managerial estimates. Its weakness is that it relies on just two data points out of an entire dataset. Because only the highest and lowest pairs are considered, the resulting split can be distorted if either extreme was an unusual month, and this is precisely the gap that scatter graphs and regression techniques are designed to close.

The scatter diagram method 🔗

The scatter diagram method takes a more visual and inclusive approach. Every pair of activity level and total cost from the available data is plotted on a graph, with activity on the horizontal axis and cost on the vertical axis.

A line of best fit is drawn through the scattered points by visual judgement, and this line is then used to read off the fixed cost, at the point where it meets the vertical axis, and the variable cost per unit, from the slope of the line.

Why it improves on the high-low method 🔗

Because the scatter diagram takes every data point into account rather than only the two extremes, it is generally regarded as more dependable than the high-low method. It also does something the high-low method cannot: it makes outliers visible. A month where a cost behaved abnormally due to a one-off event, a strike, a machine breakdown, or an unusually large festival-season order, stands out clearly on the graph and can be investigated or excluded before drawing conclusions.

The trade-off is that drawing the line of best fit by eye still involves a degree of subjectivity. Two people plotting the same points might draw slightly different lines through them. This is one reason mathematical techniques like least-squares regression exist as a further refinement, though the scatter diagram remains a genuinely useful first step for visualising how a cost behaves before any deeper statistical analysis.

Comparing the three methods 🔗

In practice, businesses rarely rely on just one of these methods. A manager might start with a scatter diagram to get a visual sense of the data and flag anything unusual, apply the high-low method for a fast numerical estimate, and lean on the analytical method to sanity-check both against what is actually known about how the cost behaves on the shop floor.

Getting the segregation right 🔗

Whichever method is chosen, the reliability of the result depends heavily on the quality of the underlying data. Activity levels and costs should be measured over comparable periods, using an activity base that genuinely drives the cost, labour hours for supervision costs or machine hours for maintenance costs, for instance, rather than an arbitrary measure. All three methods also assume that cost behaviour stays linear within a relevant range, an assumption that can break down if a business crosses into a very different scale of operations, such as adding a new production shift or investing in new machinery.

Segregating mixed costs is ultimately what makes contribution-based decision-making possible. Without a defensible fixed-variable split, break-even analysis, flexible budgeting, and pricing decisions under marginal costing all rest on assumption rather than evidence.

What do you think? If you were analysing a company’s maintenance costs with only a handful of monthly bills on hand, would you trust the high-low method’s built-in accuracy check, or would you insist on plotting a scatter diagram first? And can you spot a mixed cost hiding in your own routine expenses, perhaps a mobile data plan or an electricity bill, once you look at it through this lens?

References

1. https://www.chittaranjancollege.ac.in/wp-content/uploads/2024/08/MARGINAL-COSTING.pdf
2. https://egyankosh.ac.in/bitstream/123456789/84038/3/Block-4.pdf
3. https://corporatefinanceinstitute.com/resources/accounting/high-low-method/
4. https://www.accountingverse.com/managerial-accounting/cost-behavior/high-low-method.html
5. https://www.accountingverse.com/managerial-accounting/cost-behavior/scatter-graph-method.html