2. Cost Classification, Behaviour and Estimation

Introduction and Context

Manufacturing decisions are shaped not only by the total amount of cost incurred, but also by the way cost is classified, how it responds to activity, and how reliably it can be estimated. A production manager deciding whether to accept a special order, automate a process, outsource a component, alter the product mix, or invest in additional capacity requires more than a single product-cost figure. The decision depends on which costs will change, which have already been incurred, which resources are constrained, and which costs can be influenced by the manager.

Cost classification is therefore a managerial tool rather than a purely accounting exercise. The same expenditure may be classified differently for different purposes. Direct material may be a product cost for inventory valuation, a variable cost for short-run planning, and an avoidable cost in an outsourcing decision. A factory supervisor's salary may be indirect to a particular product, fixed within a relevant range, and controllable by a production manager over a longer period.

This chapter develops a framework for classifying manufacturing costs and analysing their behaviour. It then explains four important methods for estimating mixed costs: the engineering method, account-analysis method, high-low method, and regression analysis. These methods support budgeting, cost control, capacity planning, pricing, productivity analysis, and operational decision-making.

Learning Objectives

After studying this chapter, the reader should be able to:

1. Distinguish direct and indirect, product and period, fixed and variable, and controllable and uncontrollable costs.

2. Explain relevant, opportunity, sunk, avoidable, and incremental costs in decision situations.

3. Describe the major patterns of cost behaviour and the meaning of the relevant range.

4. Construct a cost equation for a mixed cost.

5. Apply engineering, account-analysis, high-low, and regression methods to estimate costs.

6. Interpret cost estimates critically and identify their managerial limitations.

7. Use cost classifications in production, outsourcing, capacity, and productivity decisions.

Cost as a Resource Sacrifice

A cost is the monetary measure of resources sacrificed or committed to achieve a purpose. In manufacturing, resources include materials, labour time, machine capacity, energy, factory space, maintenance services, information, and working capital. Cost is not synonymous with cash payment. Depreciation, for example, represents the allocation of the cost of a capital asset, while an opportunity cost represents a benefit sacrificed even though no cash payment occurs.

A useful cost statement always identifies three dimensions: the cost object, the purpose of measurement, and the time horizon. A cost object may be a product, batch, customer, process, department, or order. The same cost may be traceable to one cost object but common to another. For example, the wages of a worker dedicated to Product A are direct to Product A, but the plant manager's salary is indirect to all products manufactured in the plant.

Direct and Indirect Costs

A direct cost can be economically and conveniently traced to a specified cost object. Direct materials and direct production labour are traditional examples. If a component is physically incorporated into one product and its quantity can be measured efficiently, it is normally a direct material cost. Labour is direct when the time spent by employees can be identified with a product, order, or process without disproportionate measurement cost.

An indirect cost, or overhead cost, cannot be traced conveniently to a particular cost object, or tracing it would not be economically justified. Factory rent, production supervision, quality-inspection salaries, plant security, and general maintenance are common manufacturing overheads. Indirect costs are assigned using allocation bases such as direct labour hours, machine hours, units produced, material cost, or activity-based cost drivers.

The distinction is relative rather than absolute. A small quantity of adhesive used in every unit might technically be traceable, but the cost of recording it separately may exceed the benefit. It may therefore be treated as indirect material. Conversely, modern digital production systems can make some previously indirect costs traceable through machine sensors, barcode records, and electronic job tracking.

Cost allocation should not be confused with causation. An allocated cost is not necessarily caused in the same proportion by every product. A plant-wide machine-hour rate may be convenient, but it can distort product costs where products consume engineering, inspection, or setup resources differently. At postgraduate level, managers should ask whether the allocation base reflects resource consumption and whether the resulting information supports the decision being made.

Product Costs and Period Costs

Product costs are costs attached to goods manufactured or purchased for resale. In a manufacturing organisation, they generally include direct materials, direct labour, and manufacturing overhead. Product costs initially become part of inventory. They are expensed as cost of goods sold when the related products are sold.

Period costs are charged against revenue in the period in which they are incurred. Selling, distribution, general administration, corporate finance, and most research expenditure are normally period costs for external reporting purposes. They do not become part of manufacturing inventory.

This classification is closely associated with financial reporting, but it also affects managerial interpretation. A production manager may need to consider distribution cost when evaluating total supply-chain cost even though distribution cost is a period cost in the financial accounts. Similarly, a product's full economic cost may include design, after-sales service, and disposal costs, although these are not included in conventional manufacturing inventory valuation.

Fixed, Variable, and Mixed Costs

Cost behaviour describes how total cost changes as a cost driver changes. The most common driver is production volume, but machine hours, setups, batches, inspection hours, orders, or engineering changes may be more appropriate.

A variable cost changes in total in proportion to the level of activity, within a relevant range. Its variable cost per unit remains approximately constant. If material costs ₹80 per unit, total material cost is ₹80,000 for 1,000 units and ₹120,000 for 1,500 units, assuming the same price and usage conditions. Variable costs may be linear only over a limited range. Quantity discounts, overtime premiums, material waste, and capacity constraints can alter the rate.

A fixed cost remains constant in total within a specified activity range and time period. Fixed cost per unit changes as volume changes. If monthly factory rent is ₹600,000, the rent per unit is ₹600 at 1,000 units and ₹300 at 2,000 units. This does not mean that rent is permanently fixed. A new facility, lease renewal, or capacity expansion can change the total fixed-cost level.

A mixed cost contains both a fixed component and a variable component. Electricity, equipment maintenance, logistics, and telephone services commonly exhibit mixed behaviour. A linear mixed-cost equation is:

Y = a + bX

where Y is total cost, a is fixed cost, b is variable cost per unit of activity, and X is the activity level.

A step-fixed cost remains constant over a band of activity but increases when capacity crosses a threshold. One supervisor may be sufficient for up to 20 operators; a second supervisor may be required beyond that level. Step costs are particularly important in capacity planning because a small increase in output can trigger a substantial cost increase.

The relevant range is the band of activity within which assumptions about cost behaviour remain valid. A fixed-cost assumption may be reasonable between 10,000 and 30,000 machine hours but not beyond the point at which another machine, shift, or building is required. Cost estimates without a stated relevant range can create false precision.

Costs for Decision-Making

Relevant and Irrelevant Costs

A relevant cost is a future cost that differs between alternatives. Relevance has two tests: the cost must relate to the future, and it must differ among the alternatives. A cost that fails either test is irrelevant to that particular decision.

Suppose a firm is considering whether to manufacture a component or buy it from an external supplier. The relevant manufacturing costs may include direct materials, variable labour, avoidable supervision, and avoidable maintenance. Allocated factory rent that will continue under either alternative is not relevant, even though it appears in the product-cost statement.

Relevance is decision-specific. A cost irrelevant to a short-run special-order decision may become relevant to a long-term capacity decision. Managers should therefore avoid labelling a cost as universally relevant or irrelevant.

Opportunity Costs

Opportunity cost is the benefit sacrificed when one alternative is selected instead of the next-best alternative. It is not usually recorded in the accounting system, but it is essential for rational decisions. If a machine is used to produce Product A and thereby prevents the production of Product B, the contribution margin forgone from Product B is an opportunity cost of Product A.

Opportunity cost is especially important when capacity is constrained. Ranking products only by contribution per unit may be misleading. The appropriate measure is often contribution per constrained resource hour. A product generating ₹500 contribution but using five machine hours yields ₹100 per machine hour; another generating ₹360 using two hours yields ₹180 per machine hour and may deserve priority.

Sunk Costs

A sunk cost is a past cost that cannot be changed by a current decision. The historical cost of an existing machine is sunk for a replacement decision if it cannot be recovered. Sunk costs should not influence the choice between future alternatives, although they may remain relevant for performance evaluation, asset control, or financial reporting.

A common managerial error is to continue using an inefficient process merely because substantial expenditure has already been made on it. Rational analysis focuses on future cash flows and benefits, not on the desire to recover an unrecoverable past investment.

Incremental, Avoidable, and Committed Costs

Incremental cost is the additional cost caused by selecting one alternative rather than another. Avoidable cost is a cost that can be eliminated by a decision, either immediately or over the relevant time horizon. Committed costs arise from prior decisions and cannot be changed quickly, although they may become avoidable over a longer period.

These distinctions are crucial in make-or-buy, shutdown, product elimination, and special-order decisions. A product should not automatically be discontinued because it reports a loss after receiving an allocation of unavoidable fixed overhead. The analysis should determine whether the product's contribution exceeds the costs that would actually disappear.

Controllable and Uncontrollable Costs

A controllable cost is one that a manager can significantly influence within a specified period and authority structure. A production manager may control overtime, material usage, and routine maintenance scheduling, but not a long-term building lease or centrally negotiated insurance premium. Controllability depends on time horizon, organisational level, information, and decision rights.

Performance evaluation should distinguish controllable from uncontrollable costs. Holding managers accountable for costs they cannot influence encourages dysfunctional behaviour, such as resisting useful production because of centrally imposed charges. However, a cost may be uncontrollable at one level but controllable at another.

Cost Estimation Methods

Cost estimation supports budgets, quotations, standard costs, capacity plans, and scenario analysis. No method is universally superior. The choice depends on data availability, required precision, cost of analysis, operational complexity, and the purpose of the estimate.

Engineering Method

The engineering method estimates cost by analysing the physical inputs required for a production process. Engineers identify material specifications, labour standards, machine time, energy consumption, setup requirements, waste rates, and operating conditions. These physical standards are multiplied by expected prices or rates.

For example, if a unit requires 4 kilograms of material at ₹50 per kilogram, 0.4 labour hours at ₹300 per hour, and 0.2 machine hours at ₹450 per hour, the estimated prime and machine-related variable cost is:

Material: 4 × ₹50 = ₹200

Labour: 0.4 × ₹300 = ₹120

Machine resource: 0.2 × ₹450 = ₹90

Estimated variable cost per unit = ₹410

Fixed supervisory and facility costs are then estimated separately. The engineering method is useful for new products, redesigned processes, and situations where historical data are unreliable. Its weaknesses include high analysis cost, dependence on engineering standards, and the possibility that theoretical requirements differ from actual operating practice.

Account-Analysis Method

Under account analysis, an experienced analyst classifies each historical cost account as variable, fixed, or mixed. Mixed accounts are separated using knowledge of contracts, operating policies, and cost drivers. The method is relatively quick and incorporates managerial understanding that may not be visible in numerical data.

For example, maintenance expense may be classified as a fixed preventive-maintenance contract of ₹100,000 per month plus ₹25 per machine hour. If expected activity is 8,000 machine hours, estimated maintenance cost is ₹300,000.

The approach is flexible but subjective. Two analysts may classify the same cost differently. Documentation, review by operations personnel, and comparison with actual outcomes improve its reliability.

High-Low Method

The high-low method estimates the variable rate using the highest and lowest activity observations:

Variable rate (b) = (Cost at high activity − Cost at low activity) ÷ (High activity − Low activity)

The fixed component is then calculated as:

Fixed cost (a) = Total cost − (Variable rate × Activity)

Consider the following maintenance data:

• At 2,000 machine hours, cost is ₹170,000.
• At 5,000 machine hours, cost is ₹260,000.

Variable rate = (₹260,000 − ₹170,000) ÷ (5,000 − 2,000) = ₹30 per machine hour.

Fixed cost = ₹260,000 − (₹30 × 5,000) = ₹110,000.

The cost equation is therefore Y = ₹110,000 + ₹30X. At 4,000 machine hours, estimated cost is ₹230,000.

The method is simple and transparent, but it uses only two observations. The highest and lowest activity points may be unusual, affected by breakdowns, abnormal waste, price changes, or a different operating regime. High-low analysis should therefore be treated as a preliminary estimate, not as proof of a causal relationship.

Regression Analysis

Regression analysis estimates the relationship between cost and one or more explanatory variables using all available observations. In simple linear regression:

Y = a + bX + e

where e represents unexplained variation. The method selects values of a and b that minimise the sum of squared differences between actual and predicted costs.

Regression provides statistical indicators such as the coefficient of determination, or R², which indicates the proportion of variation in the dependent cost statistically associated with the explanatory variable in the sample. A high R² does not prove causation. A variable may correlate with cost because both respond to a third factor, or because the data contain a common time trend.

Multiple regression can include several cost drivers, such as machine hours, number of setups, production batches, and inspection hours. This is valuable where overhead is driven by activities rather than output volume alone. However, managers must examine multicollinearity, outliers, changing prices, non-linearity, autocorrelation, and whether the data represent normal operating conditions.

A sound regression process includes defining the cost object, selecting a plausible causal driver, cleaning abnormal observations, inspecting residuals, testing assumptions, and validating predictions against subsequent periods. Statistical sophistication cannot compensate for a poorly defined cost relationship.

Comparing the Estimation Methods

The engineering method is strongest when process knowledge and physical standards are available, especially for new operations. Account analysis is practical when time is limited and experienced personnel understand the accounts. High-low analysis is easy to calculate and useful for an initial approximation, but it is sensitive to two observations. Regression is more objective and uses more information, but it requires reliable data and analytical capability.

The methods may be combined. A company can use engineering standards to establish an expected variable rate, account analysis to identify fixed contracts, and regression to test whether actual data are consistent with the model. Differences between methods can reveal waste, idle capacity, price changes, maintenance events, or changes in process technology.

Applications in Manufacturing Management

Cost classification supports several core managerial activities. In budgeting, fixed and variable elements allow managers to forecast cost under alternative production volumes. In standard costing, engineering estimates define expected material quantities, labour times, and machine usage. In productivity management, costs can be related to output and resource consumption to distinguish efficiency changes from price changes.

For pricing and quoting, managers should understand whether the decision concerns full long-run cost recovery or short-run capacity utilisation. A special order may be acceptable below full cost if it uses idle capacity, covers incremental costs, does not disrupt regular sales, and does not create strategic or quality problems. It should not be accepted merely because it produces a positive accounting margin based on an inappropriate allocation.

In make-or-buy analysis, relevant avoidable costs and opportunity costs should be compared with the supplier's price. In product-mix decisions, contribution per constrained resource is often more informative than contribution per unit. In shutdown decisions, fixed costs that continue after closure must be separated from those that can actually be avoided.

Digital manufacturing systems increasingly connect cost estimation with production data. Enterprise resource planning systems, manufacturing execution systems, machine sensors, and energy-monitoring platforms can capture machine hours, cycle times, downtime, scrap, setups, and material movements. Process-mining and machine-learning tools can identify cost drivers and detect unusual consumption. Nevertheless, automated estimates remain dependent on data quality, correct driver selection, and human interpretation.

Illustrative Case: The Component Sourcing Decision

Background

Orion Components, an illustrative mid-sized manufacturer, produces 24,000 pump assemblies annually. A critical metal bracket is currently manufactured internally. The purchasing department has received a supplier quotation of ₹148 per bracket.

The internal cost statement reports the following unit costs:

• Direct material: ₹62
• Direct labour: ₹34
• Variable manufacturing overhead: ₹18
• Allocated fixed manufacturing overhead: ₹42
• Total reported cost: ₹156

At first sight, purchasing appears attractive because the supplier's price is ₹8 below reported internal cost.

Situation and Problem

The production manager must determine whether outsourcing will reduce annual cost. The plant has sufficient capacity to use the released machine time for another product, generating additional contribution of ₹9 per bracket-equivalent. If the bracket is outsourced, direct material, direct labour, and variable overhead will disappear. Of the allocated fixed overhead, only ₹12 per bracket is avoidable. The remaining ₹30 is committed for the foreseeable period.

Analysis

Relevant internal cost before considering opportunity cost is:

₹62 + ₹34 + ₹18 + ₹12 = ₹126 per bracket.

The supplier price is ₹148, so continuing internal production is cheaper by ₹22 per bracket on avoidable-cost grounds. However, internal production uses capacity that could generate ₹9 contribution elsewhere. The economic cost of continuing to manufacture is therefore:

₹126 + ₹9 = ₹135 per bracket.

Internal production remains preferable by ₹13 per bracket. For 24,000 brackets, the estimated annual advantage of internal production is ₹312,000, assuming the alternative product can actually be produced and sold.

The ₹42 allocated overhead should not be used as a whole in the decision because ₹30 will remain under either alternative. The historical cost of existing tooling would also be sunk unless it could be sold or redeployed. The conclusion would change if the supplier offered less than ₹135, if internal quality costs increased, or if outsourcing released a bottleneck with a higher contribution opportunity.

Outcome and Learning Points

The case demonstrates that reported product cost is not the same as relevant decision cost. The decision requires identification of avoidable costs, committed costs, and opportunity costs. It also illustrates why capacity information and operational feasibility are essential. A numerical comparison alone cannot establish the decision unless quality, delivery reliability, supplier dependence, intellectual property, and transition risks are considered.

Challenges, Risks, and Common Errors

Several errors repeatedly weaken cost analysis. Managers may treat all allocated overhead as avoidable, confuse average cost with incremental cost, or assume that a cost classified as fixed will remain fixed at every volume. They may extrapolate a regression equation beyond the relevant range, use high-low observations affected by abnormal events, or rely on historical relationships after a process redesign.

Another risk is cost-driver misidentification. Output volume may explain direct material but not engineering support or setup cost. Using a single volume-based rate can cross-subsidise high-complexity products and encourage unprofitable product decisions. Data systems can also create an illusion of objectivity when transaction definitions, time stamps, or standards are inconsistent.

Ethically, cost information should not be manipulated to justify a preferred decision. Estimates should disclose assumptions, ranges, uncertainty, excluded costs, and the time horizon. Managers should avoid transferring costs to another department merely to improve reported performance. Where cost information affects employees, suppliers, or customers, professional judgement should consider quality, safety, labour conditions, environmental consequences, and long-term organisational capability.

Best-Practice Guidelines

A disciplined cost-estimation process should:

• define the decision, cost object, time horizon, and relevant range;
• identify the physical and operational cost drivers;
• separate fixed, variable, mixed, step, avoidable, and committed components;
• use engineering knowledge and accounting data together;
• test estimates against actual results and investigate variances;
• distinguish correlation from causation in statistical models;
• document assumptions and review them after process or price changes;
• include opportunity costs where capacity is constrained; and
• present decision-makers with sensitivity analysis rather than a single unsupported number.

Chapter Summary

Cost classification provides the language for analysing manufacturing resources and managerial alternatives. Direct costs can be traced economically to a cost object, whereas indirect costs require allocation. Product costs enter inventory and become expense when goods are sold; period costs are charged to the accounting period. Fixed, variable, mixed, and step costs describe different patterns of response to activity within a relevant range.

For decisions, the critical question is not whether a cost appears in the accounts, but whether it is future and different between alternatives. Relevant costs may include avoidable, incremental, and opportunity costs. Sunk and unavoidable allocated costs generally should not affect short-run choices. Controllability depends on managerial authority and time horizon.

Mixed costs can be estimated through engineering analysis, account analysis, high-low analysis, and regression. Each method has a different balance of speed, objectivity, data requirements, and analytical risk. Effective production managers combine quantitative estimates with process knowledge, capacity information, quality considerations, and ethical judgement.

Key Terms

Cost object: The product, service, batch, department, customer, or activity for which cost is measured.

Direct cost: A cost economically and conveniently traceable to a specified cost object.

Indirect cost: A cost not conveniently traceable to one cost object and commonly allocated as overhead.

Product cost: A manufacturing cost attached to inventory until the related product is sold.

Period cost: A cost expensed in the period incurred rather than included in inventory.

Variable cost: A cost that changes in total with activity while remaining approximately constant per unit within a range.

Fixed cost: A cost that remains constant in total within a relevant range and period.

Mixed cost: A cost containing both fixed and variable components.

Step-fixed cost: A cost constant over a band of activity that increases when a capacity threshold is crossed.

Relevant range: The activity interval over which cost-behaviour assumptions remain valid.

Relevant cost: A future cost that differs among decision alternatives.

Opportunity cost: The benefit sacrificed by selecting one alternative instead of the next-best alternative.

Sunk cost: A past cost that cannot be changed by a current decision.

Avoidable cost: A cost that can be eliminated as a consequence of a decision.

Committed cost: A cost arising from prior decisions that cannot be changed quickly.

Controllable cost: A cost that a manager can significantly influence within a specified authority and period.

High-low method: A technique estimating variable and fixed costs from the highest and lowest activity observations.

Regression analysis: A statistical method estimating cost relationships using multiple observations.

Cost driver: A measurable factor that causes or explains changes in cost.

Contribution per constrained resource: Contribution margin earned for each unit of a scarce resource, such as a machine hour.

Review Questions

Short Answer Questions

1. Why is cost classification described as decision-specific rather than absolute?

2. Distinguish product costs from period costs in a manufacturing organisation.

3. What is meant by the relevant range?

4. Explain the difference between a sunk cost and an opportunity cost.

5. State the equation used to represent a mixed cost.

Long Answer Questions

1. Explain direct and indirect costs, and discuss why allocation bases may distort product cost.

2. Compare fixed, variable, mixed, and step-fixed cost behaviour with manufacturing examples.

3. Explain relevant, avoidable, incremental, committed, and controllable costs.

4. Describe the engineering, account-analysis, high-low, and regression methods of cost estimation.

5. Discuss the managerial and ethical risks of relying on inaccurate or poorly classified cost information.

Analytical / Application Questions

1. A maintenance department incurs ₹170,000 at 2,000 machine hours and ₹260,000 at 5,000 machine hours. Estimate its cost equation and forecast cost at 4,000 machine hours using the high-low method.

2. A product earns ₹500 contribution using five constrained machine hours, while another earns ₹360 using two hours. Which product should receive priority if demand is sufficient, and why?

3. A supplier offers to provide a component for ₹148. Internal variable costs are ₹114 per unit, avoidable fixed cost is ₹12, and the released capacity could earn ₹9 contribution per unit elsewhere. Should the firm buy or make, assuming quality and supply conditions are equivalent? Show the analysis.

22. Multiple Choice Questions with Answers

Question 1

Which cost is most likely to be relevant to a short-run make-or-buy decision?

A. Historical depreciation on an existing machine

B. Allocated rent that continues under either alternative

C. Variable labour cost eliminated if production is outsourced

D. Past expenditure on product development

Correct Answer: C. Variable labour cost eliminated if production is outsourced

Question 2

A cost is ₹170,000 at 2,000 machine hours and ₹260,000 at 5,000 hours. What is the estimated variable cost per machine hour using high-low analysis?

A. ₹20

B. ₹30

C. ₹45

D. ₹90

Correct Answer: B. ₹30

Question 3

Which statement correctly describes a mixed cost?

A. It remains constant in total at all activity levels

B. It changes only when a capacity step is reached

C. It contains both fixed and variable components

D. It is always directly traceable to a product

Correct Answer: C. It contains both fixed and variable components

Question 4

A product generates ₹500 contribution and uses five machine hours. Another generates ₹360 contribution and uses two machine hours. Which product has the higher contribution per constrained machine hour?

A. The first product, at ₹100 per hour

B. The first product, at ₹250 per hour

C. The second product, at ₹180 per hour

D. The second product, at ₹360 per hour

Correct Answer: C. The second product, at ₹180 per hour

Question 5

Which method estimates cost by analysing material quantities, labour standards, machine time, and process requirements?

A. High-low method

B. Engineering method

C. Account-analysis method

D. Regression method

Correct Answer: B. Engineering method

Question 6

What is the principal weakness of the high-low method?

A. It cannot estimate a variable cost rate

B. It requires engineering standards for every operation

C. It uses only two observations and may be distorted by unusual points

D. It always produces a nonlinear cost equation

Correct Answer: C. It uses only two observations and may be distorted by unusual points

Question 7

A factory supervisor’s salary is best described as which combination when assigned to several products?

A. Direct and variable

B. Indirect and usually fixed within the relevant range

C. Direct and sunk

D. Period and opportunity cost

Correct Answer: B. Indirect and usually fixed within the relevant range

Question 8

Which statement about regression analysis is most accurate?

A. A high R² proves that the selected cost driver causes the cost

B. Regression uses all available observations but still requires managerial interpretation

C. Regression is based only on the highest and lowest activity levels

D. Regression eliminates the need to define a relevant range

Correct Answer: B. Regression uses all available observations but still requires managerial interpretation

Question 9

A plant’s rent will continue whether a product is discontinued or retained. How should the rent be treated in the short-run discontinuation decision?

A. As an avoidable cost

B. As an opportunity cost

C. As an irrelevant unavoidable cost

D. As a variable cost

Correct Answer: C. As an irrelevant unavoidable cost

Question 10

Which cost equation represents a mixed cost with fixed cost a and variable rate b?

A. Y = a − bX

B. Y = b ÷ aX

C. Y = a + bX

D. Y = X ÷ (a + b)

Correct Answer: C. Y = a + bX