Performance Measurement and Accountability

Introduction and context

Operations costing and accounting do more than record expenditure. They create information through which managers judge whether resources are being converted into customer value, operational capability and sustainable returns. Performance measurement therefore connects accounting data with strategy, process design, employee behaviour and governance.

A sound measurement system answers five questions:

1. What outcomes matter to the organisation and its stakeholders?
2. Which operational drivers produce those outcomes?
3. Who has authority over each driver?
4. How should performance be compared across time, units and competitors?
5. What incentives will encourage sound decisions without creating dysfunctional behaviour?

The central managerial difficulty is that no single measure is sufficient. Cost reduction may damage quality; high asset utilisation may create excess inventory; short-term profit may undermine innovation; and a favourable budget variance may result from under-maintenance. Effective accountability consequently requires a coherent portfolio of financial and non-financial indicators.

Learning objectives

By the end of this chapter, a student should be able to:

• distinguish measures, metrics, key performance indicators (KPIs) and targets;
• design a strategy-linked KPI architecture and management dashboard;
• explain the balanced scorecard and its limitations;
• evaluate cost, profit and investment responsibility centres;
• calculate and interpret ROI and residual income;
• analyse transfer-pricing alternatives and behavioural consequences;
• use benchmarking appropriately;
• design incentive systems while managing agency, gaming and ethical risks;
• recommend governance practices for accountable performance management.

1. Performance measurement as a control system

Performance measurement is the systematic collection, analysis and use of evidence about organisational activities and results. A KPI is a limited set of especially important measures linked to strategic or operational objectives. A metric may be informative without being strategically critical; a KPI is a metric given priority, ownership, target and review discipline.

A complete performance-management cycle contains:

Strategy → objectives → value drivers → measures → targets → responsibility → review → corrective action.

Measures may be classified in several ways:

A useful measure has validity, reliability, timeliness, controllability, comparability and cost-effectiveness. Measurement validity asks whether the indicator represents the intended construct. For example, units produced is not a valid proxy for productivity if defects and rework are ignored. Reliability concerns consistency of measurement, while controllability asks whether the accountable manager can materially influence the result.

Leading and lagging indicators

Lagging indicators such as quarterly operating profit are important but arrive after decisions have been made. Leading indicators, such as first-pass yield and preventive-maintenance completion, provide earlier warning. A balanced system links them through a causal hypothesis: maintenance discipline should improve equipment availability, which should improve schedule adherence, customer service and profit. Such causal links should be tested rather than assumed.

Targets and thresholds

A target is a desired level of performance. A threshold is a boundary requiring escalation. Targets may be budget-based, historical, benchmark-based, statistically derived or negotiated. A target should specify the period, population, unit, data source, owner and treatment of exceptional events. “Improve delivery” is not a target; “deliver at least 96% of customer orders on or before the confirmed date during the quarter” is more auditable.

2. Designing KPIs and operational dashboards

KPI design begins with strategic objectives rather than with readily available data. The following sequence is robust:

1. Translate strategy into a small number of objectives.
2. Identify the processes and drivers that influence each objective.
3. Select measures with clear operational definitions.
4. Assign an accountable owner and supporting contributors.
5. Set a baseline, target, tolerance and review frequency.
6. Establish data lineage, controls and escalation rules.
7. Pilot the measures and revise unintended effects.

A KPI specification should include:

For example, on-time-in-full (OTIF) may be defined as the percentage of customer order lines delivered by the confirmed date and in the confirmed quantity. If 940 of 1,000 eligible order lines satisfy both conditions:

The measure is meaningful only if cancelled orders, partial shipments, revised promise dates and customer-caused delays are treated consistently.

Dashboard architecture

A dashboard is a decision interface, not merely a collection of charts. An operations dashboard should normally display:

• current status against target;
• trend over time;
• variance and materiality;
• drill-down by product, plant, customer or shift;
• causal or driver indicators;
• responsible owner and next action.

Traffic-light colours can aid attention, but they should not replace numerical context. A green result may conceal deterioration if the target is weak, while a red result may reflect a one-off event outside managerial control. Dashboards should therefore include confidence, data freshness and commentary where appropriate.

3. The balanced scorecard

The balanced scorecard (BSC) translates strategy into a linked set of objectives and measures across four conventional perspectives:

1. Financial: How should we appear to owners and providers of capital?
2. Customer: How should customers perceive our value?
3. Internal process: Which processes must excel?
4. Learning and growth: Which capabilities enable future performance?

The perspectives are not four independent reporting categories. A strategy map expresses hypothesised relationships. For example, employee skills and information quality may improve process reliability; reliable processes may improve delivery and customer retention; retention may support margin and cash generation.

A BSC should not be treated as a universal list of equal measures. Measures must reflect the organisation’s strategy, operating model and risk profile. A manufacturer competing on dependable customisation may emphasise engineering response time and first-pass yield; a low-cost commodity producer may emphasise throughput, energy intensity and total conversion cost.

Strengths and limitations

The BSC reduces excessive dependence on accounting results and encourages strategic dialogue. It makes intangible capabilities visible and helps align departments around outcomes. However, causal links may be speculative, measures can proliferate, and the four perspectives may omit environmental, regulatory or stakeholder obligations. A contemporary scorecard may therefore add explicit sustainability, resilience, safety or social-impact objectives where material.

4. Responsibility accounting and organisational accountability

Responsibility accounting reports revenues, costs, assets and outcomes to managers who have authority over them. The principle is not that managers must explain every result, but that reports should distinguish controllable from non-controllable influences and reveal interdependencies.

Responsibility centres

The classification is a management design choice, not an absolute fact. A plant manager may control labour and scrap but not corporate depreciation, transfer prices or product mix. Reports should identify controllability and avoid attributing corporate policy costs to operational managers without explanation.

Responsibility accounting also requires vertical and horizontal coordination. A purchasing manager who obtains a low material price may buy inferior inputs that increase production defects. A production manager may maximise local efficiency by producing large batches, thereby increasing inventory and obsolescence. Local accountability must therefore be complemented by end-to-end process measures.

5. ROI and residual income

Return on investment (ROI) relates operating income to the asset base used to generate it:

Operating income should be defined consistently with the asset measure, and average operating assets are commonly calculated as beginning assets plus ending assets divided by two when asset usage changes materially during the period. ROI can be decomposed into margin and asset turnover:

The first component is operating margin and the second is asset turnover. This decomposition helps managers distinguish a profitability problem from an asset-utilisation problem. The formula is a percentage and is meaningful only when income and assets use compatible accounting boundaries.

Worked example

A division reports sales of $12 million, operating income of $1.2 million, beginning operating assets of $5 million and ending operating assets of $7 million.

Average operating assets = ($5m + $7m) ÷ 2 = $6m.

ROI = $1.2m ÷ $6m = 20%.

Margin = $1.2m ÷ $12m = 10%; asset turnover = $12m ÷ $6m = 2. Therefore, ROI = 10% × 2 = 20%.

ROI encourages attention to both cost control and asset productivity, but it can create dysfunctional decisions. Suppose the division currently earns 20% and is offered an investment requiring $2 million that generates $360,000 of annual operating income. The project ROI is 18%, below the division’s current ROI. A manager evaluated only on divisional ROI may reject it, even though it exceeds a corporate cost of capital of, say, 12% and creates value for the organisation.

Residual income (RI) addresses this issue by deducting an imputed capital charge:

For the proposed project, incremental RI = $360,000 − (12% × $2,000,000) = $120,000. The project is economically attractive under the stated assumptions. RI is expressed in currency, so comparisons across divisions of different sizes require care; a large division may produce more RI but a lower percentage return.

Economic value added (EVA) is a related concept that generally uses after-tax operating profit and a weighted average cost of capital, with adjustments to accounting numbers where justified. ROI, RI and EVA are decision tools, not complete measures of performance. They can be distorted by asset age, depreciation policy, capitalisation rules and short-term cost deferral.

6. Transfer pricing and internal markets

Transfer pricing determines the value assigned when one responsibility centre supplies goods or services to another. It affects reported divisional profit, tax outcomes, sourcing decisions and incentives, although it does not change total company profit before external tax and transaction costs.

Common approaches are:

• Market-based price: appropriate when a competitive external market exists and products are comparable.
• Variable-cost price: may support short-run group optimisation when the supplying unit has idle capacity, but can make the supplier appear unprofitable.
• Full-cost price: recovers variable and fixed costs but may transfer inefficiency and encourage suboptimal sourcing.
• Cost-plus price: adds an agreed margin, but can reward the supplier for increasing its cost base.
• Negotiated price: permits local information and flexibility but may create bargaining costs and unequal power.
• Dual pricing: records different prices for buyer and seller while reconciling the difference centrally; useful but administratively complex.

The economically relevant transfer price depends on opportunity cost. If the supplying division has idle capacity and no external demand, the short-run minimum price is often variable cost. If it must sacrifice an external sale, the minimum reflects variable cost plus lost contribution. This is a short-run analysis; long-run capacity and fixed costs may alter the decision.

Transfer-price governance should define capacity assumptions, external alternatives, service levels, quality responsibility, dispute escalation and periodic review. Tax and regulatory requirements must be observed where cross-border transactions are involved; managerial transfer pricing cannot be designed solely around divisional incentives.

7. Benchmarking and comparative performance

Benchmarking is systematic comparison with a relevant reference point in order to identify performance gaps and learning opportunities. Types include:

• internal benchmarking across plants or periods;
• competitive benchmarking against direct rivals;
• functional benchmarking across industries;
• process benchmarking against organisations recognised for a particular process;
• performance benchmarking using output and outcome indicators.

A benchmark is not automatically a target. Comparability requires adjustment for product mix, scale, geography, technology, regulation, service promise and accounting policy. A plant producing complex, low-volume products should not be judged directly against a high-volume standardised plant without normalisation.

A simple gap analysis is:

The sign and interpretation depend on whether higher or lower performance is desirable. Benchmarking should progress from identifying a gap to understanding process causes, adapting practices and monitoring results. Copying another organisation without understanding context often fails.

8. Incentives, agency and behavioural risk

Agency theory highlights the problem that managers or employees (agents) may possess different information and objectives from owners or senior leaders (principals). Incentives seek alignment, but measurement itself changes behaviour. This creates both productive motivation and risk.

Common dysfunctional responses include:

• gaming targets or manipulating data;
• delaying maintenance or training to improve current profit;
• producing unwanted inventory to absorb fixed overhead;
• rejecting profitable projects that reduce ROI;
• shifting costs or revenue between periods;
• focusing on measured dimensions while neglecting unmeasured quality, safety or ethics;
• colluding to establish easy targets.

Good incentive design uses a mix of measures, thresholds and gates. A balanced plan may combine financial performance, customer outcomes, quality, safety, cash and strategic milestones. Safety and legal compliance should often be minimum conditions rather than tradeable points. Relative performance can reduce the effect of common shocks, but it may also encourage concealment or unhealthy rivalry.

A generic bonus function can be expressed as:

Here the performance factor is determined by an approved scorecard, while the eligibility factor may be zero when serious misconduct or safety breaches occur. The formula is a policy representation, not a universal accounting rule. It must specify caps, floors, weighting, measurement dates and governance approval.

Incentives should be tested before implementation using scenario analysis: What happens in a recession? What if demand surges? Can a manager improve the score by damaging another function? Are results auditable? Are employees rewarded for outcomes they can realistically influence?

9. Governance, ethics and technology

Accountability systems create risks involving privacy, surveillance, discrimination, financial misstatement and unsafe work intensification. Ethical governance requires transparency about data use, access controls, separation of measurement from manipulation, documented overrides and an appeal process. Performance targets must not encourage employees to violate professional duties or regulatory requirements.

Digital operations and artificial intelligence can automate data capture, anomaly detection, forecasting and dashboard commentary. They do not eliminate managerial judgement. AI-generated performance scores may reproduce biased historical data, confuse correlation with causation or penalise teams serving difficult customers. Organisations should maintain human review, model validation, audit trails, explainability proportionate to risk and clear accountability for automated decisions.

Contemporary systems increasingly integrate enterprise-resource-planning, manufacturing-execution, customer and sustainability data. The resulting “single source of truth” is valuable only when master data, definitions, access rights and reconciliation controls are reliable. More data does not necessarily mean better accountability.

10. Illustrative case: Northstar Components

Background

Northstar Components is an illustrative manufacturer of precision assemblies. Its plant manager receives an annual bonus based mainly on conversion-cost variance and divisional ROI. The plant has recently achieved a favourable labour variance, but customer complaints, overtime, inventory and equipment failures have increased.

Problem

The existing system rewards producing long runs and postponing planned maintenance. The plant reports strong monthly efficiency, yet first-pass yield has declined from 96% to 91%, OTIF has fallen from 97% to 92%, and warranty costs have risen. A proposed automation project would earn 15% ROI, while the plant’s current ROI is 19%; the manager therefore opposes it despite the corporate required return of 10%.

Analysis

The cost variance is a lagging and narrow measure. It excludes quality losses, customer penalties, inventory carrying costs and lifecycle equipment effects. The ROI rule creates an underinvestment incentive. Responsibility is also unclear because sales changes promise dates, engineering changes specifications and maintenance owns part of the failure data.

Approach and solution

Northstar redesigns the scorecard around five objectives: profitable growth, dependable customer service, stable processes, efficient resource use and capability development. Measures include contribution margin, OTIF, first-pass yield, total cost of poor quality, schedule adherence, preventive-maintenance completion, inventory days, safety incidents, training completion and cash conversion. Financial measures retain substantial weight, but safety and compliance are gates.

The automation project is evaluated using incremental RI: incremental operating income minus 10% of incremental operating assets. The plant’s scorecard distinguishes controllable plant measures from corporate transfer prices and records cross-functional causes. A monthly review examines trends and corrective actions rather than merely ranking managers. Data definitions are approved by finance, operations and internal audit.

Outcome

In the first year, reported conversion efficiency is lower because previously hidden quality and maintenance costs are recognised. After implementation, first-pass yield improves, emergency downtime declines, OTIF recovers and the automation project is approved. The initial apparent deterioration illustrates an important principle: better measurement can reveal poor performance before it improves performance.

Learning points

The case demonstrates that local cost measures can conflict with enterprise value; ROI may discourage value-creating investment; and dashboards require causal, cross-functional and behavioural analysis. Accountability is strongest when authority, information, measures and incentives are designed together.

Challenges and limitations

Performance systems cannot remove uncertainty or fully separate managerial effort from environmental factors. Excessive measurement creates administrative burden and target fatigue. Financial measures may encourage short-termism, while non-financial measures can be subjective or difficult to value. Benchmark data may be confidential or incomparable. Causal models may be wrong, and incentive systems may produce responses that were not anticipated by their designers.

Managers should therefore review measures periodically, retire obsolete KPIs, investigate unintended consequences and use professional judgement. A measure should be changed carefully: frequent changes destroy comparability, but permanent adherence to a flawed measure institutionalises error.

Best-practice principles

1. Start with strategy and value creation, not available data.
2. Use a limited portfolio of financially and operationally meaningful measures.
3. Pair lagging outcomes with leading drivers.
4. Assign accountability only where authority and information exist.
5. Separate controllable from non-controllable effects.
6. Use ROI with RI or another value measure when investment decisions matter.
7. Make transfer-pricing assumptions explicit.
8. Benchmark comparable processes, not superficial numbers.
9. Test incentives for gaming, safety and cross-functional harm.
10. Govern data definitions, algorithms, privacy and overrides.
11. Treat dashboards as decision systems, not decorative reporting.
12. Learn from measurement and revise the system responsibly.

Chapter Summary

Performance measurement converts strategy into observable objectives, targets, responsibility and corrective action. Effective systems combine financial, customer, process, people and capability measures, while distinguishing leading from lagging indicators and controllable from uncontrollable outcomes. KPIs require precise definitions, owners, data sources, targets and governance.

The balanced scorecard links financial, customer, internal-process and learning perspectives through a strategy map, although its causal assumptions must be tested and adapted to sustainability, safety and resilience needs. Responsibility accounting assigns results to cost, revenue, profit and investment centres, but local measures must not undermine enterprise performance.

ROI can be decomposed into margin and asset turnover, yet may discourage investments that earn less than current divisional ROI but more than the organisation’s required return. Residual income addresses this by deducting a capital charge. Transfer pricing must balance decision relevance, divisional fairness, capacity conditions, tax rules and behavioural effects. Benchmarking requires comparability and process learning rather than mechanical imitation.

Finally, incentives create both alignment and risk. Agency problems, gaming, short-termism, unsafe behaviour and data manipulation require mixed measures, gates, auditability, human judgement and ethical governance. The best performance system is therefore not the most complex one, but the one that produces valid information and better decisions without sacrificing long-term organisational value.

Key Terms

Accountability: Obligation to explain performance and take responsibility for results.

Balanced scorecard: Strategy-linked framework using financial, customer, process, and learning perspectives.

Benchmarking: Systematic comparison with a relevant reference organisation, process or standard.

Controllability: Degree to which a manager can influence a reported result.

Dashboard: Visual decision interface displaying measures, trends, variances and actions.

Economic value added: Value measure based on after-tax operating profit less a capital charge.

Key performance indicator: A strategically important, owned and targeted performance measure.

Leading indicator: Measure that provides an early signal of future performance.

Lagging indicator: Measure that records an outcome after decisions and activities occur.

Margin: Operating income divided by sales.

Residual income: Operating income less an imputed charge for capital employed.

Responsibility accounting: Reporting system that assigns results to managers with relevant authority.

Responsibility centre: Organisational unit classified by the results for which its manager is accountable.

Return on investment: Operating income divided by average operating assets.

Strategy map: Visual representation of hypothesised cause-and-effect links among objectives.

Target: Desired level of performance for a defined measure and period.

Transfer price: Internal price assigned to goods or services exchanged between responsibility centres.

Agency risk: Risk that an agent’s actions diverge from the principal’s interests.

Review Questions

Short Answer Questions

1. Distinguish a metric from a KPI.
2. What is the difference between a leading and a lagging indicator?
3. Identify the four conventional balanced-scorecard perspectives.
4. Define a responsibility centre.
5. State two behavioural risks created by poorly designed incentives.

Long Answer Questions

1. Explain how an operations dashboard should be designed and governed.
2. Evaluate the strengths and limitations of ROI and residual income.
3. Discuss how transfer-pricing methods influence divisional behaviour.
4. Explain why benchmarking requires adjustments for comparability.
5. Critically evaluate the balanced scorecard as a strategic-control framework.

Analytical / Application Questions

1. A division has operating income of $900,000, beginning assets of $4 million and ending assets of $5 million. Calculate ROI and decompose it if sales are $10 million. Interpret the result.
2. A proposed project requires $1.5 million of assets and generates $240,000 operating income. The corporate required return is 12%, while the division’s current ROI is 18%. Should the project be accepted using residual income logic? Explain.
3. Design a six-KPI scorecard for a plant experiencing low cost but worsening quality and delivery. Justify each KPI, identify its owner and explain one possible gaming risk.

Multiple Choice Questions with Answers

Question 1

Which measure is most clearly a leading indicator of future equipment reliability?

A. Annual operating profit

B. Preventive-maintenance completion rate

C. Prior-year warranty cost

D. End-of-year market share

Correct Answer: B. Preventive-maintenance completion rate

Question 2

A division has operating income of $800,000 and average operating assets of $4,000,000. Its ROI is:

A. 5%

B. 20%

C. 25%

D. 50%

Correct Answer: B. 20%

Question 3

Which responsibility centre is primarily accountable for profit and the assets it employs?

A. Cost centre

B. Revenue centre

C. Profit centre

D. Investment centre

Correct Answer: D. Investment centre

Question 4

Residual income is calculated as:

A. Sales less variable cost

B. Operating income less a capital charge

C. Operating assets divided by sales

D. Operating income divided by sales

Correct Answer: B. Operating income less a capital charge

Question 5

A major behavioural weakness of evaluating a manager only on divisional ROI is that the manager may:

A. Accept every available project

B. Reject projects earning above the corporate required return but below current divisional ROI

C. Ignore all divisional revenues

D. Increase the required corporate return

Correct Answer: B. Reject projects earning above the corporate required return but below current divisional ROI

Question 6

Which condition most strongly supports a market-based transfer price?

A. No external market exists

B. The supplying unit has no measurable costs

C. A competitive external market offers comparable goods or services

D. The buyer and seller have identical managers

Correct Answer: C. A competitive external market offers comparable goods or services

Question 7

Which balanced-scorecard perspective focuses most directly on employee capability and organisational learning?

A. Financial

B. Customer

C. Internal process

D. Learning and growth

Correct Answer: D. Learning and growth

Question 8

Why can a favourable labour-efficiency variance be misleading?

A. It always proves quality has improved

B. It may coexist with overtime, defects or maintenance deferral

C. It cannot be calculated in monetary terms

D. It measures customer satisfaction directly

Correct Answer: B. It may coexist with overtime, defects or maintenance deferral

Question 9

If a project requires $2,000,000 of assets and generates $300,000 operating income, with a required return of 10%, its incremental residual income is:

A. $30,000

B. $100,000

C. $200,000

D. $300,000

Correct Answer: B. $100,000

Question 10

Which governance practice best reduces the risk of KPI manipulation?

A. Changing definitions every month

B. Removing all non-financial measures

C. Documenting definitions, data lineage, controls and review rights

D. Allowing each manager to calculate results privately

Correct Answer: C. Documenting definitions, data lineage, controls and review rights