Pergamon

PII:

European Management Journal Vol. 17, No. 5, pp. 492–500, 1999  1999 Elsevier Science Ltd. All rights reserved Printed in Great Britain S0263-2373(99)00035-3 0263-2373/99 $20.00 ⫹ 0.00

Activity-based Costing Concepts for Quality Improvement CHRISTOPHER D. ITTNER, The Wharton School, University of Pennsylvania This article shows how activity-based costing concepts can be adapted to measure quality-related costs and prioritize quality improvement efforts. It first reviews three levels of activity analysis that can be used to quantify the internal costs arising from failure to meet customer requirements. It then highlights how activity-based costing concepts can be extended to encompass quality-related costs arising from supplier deficiencies and the opportunity costs of lost sales due to quality problems.  1999 Elsevier Science Ltd. All rights reserved

Introduction Competitive requirements have forced nearly every firm to invest in quality improvement activities. However, many companies find it difficult to identify the quality improvement projects offering the highest returns, or to quantify the financial payback from these investments. A recent survey, for example, found that 75 per cent of senior quality executives felt pressure to demonstrate the financial consequences of their quality initiatives, but fewer than 55 per cent could directly relate their quality efforts to operational, productivity, or revenue improvements, and only 12 per cent to accounting returns (Ittner and Larcker, 1996). As a result, 52 per cent of the executives found it difficult to identify the quality projects with the highest expected economic returns, and none found this to be an easy task. The primary reason for these difficulties is the lack of adequate methods for determining the financial consequences of poor quality. Many quality-related costs, such as scrap, rework, warranty claims, and quality department expenditures, may be readily available from existing accounting records, especially in manufacturing firms. However, accounting systems that are not designed with quality costing in mind typically identify only a fraction of quality492

related expenditures. One study found that the costs of quality-related activities and failures were four times higher than most manufacturers estimated based on their existing accounting records (Atkinson et al., 1994). Other studies estimate that qualityrelated costs can represent up to 50 per cent of a service function’s budget (Harrington, 1987), yet most companies have little or no capability to track these costs. Moreover, few accounting systems capture the opportunity costs that arise when customers are lost due to quality problems. Because most accounting systems are not designed to provide information on the costs and benefits of quality improvement, other sources are needed to provide a more complete assessment of quality-related performance consequences. Many companies are beginning to apply various forms of activity-based costing (ABC) to assess these ‘hidden’ costs and benefits. This paper shows how activity-based costing concepts can be adapted to measure quality-related costs and prioritize quality improvement efforts. I first review three levels of activity analysis that can be used to quantify the internal costs arising from failure to meet customer requirements. I then highlight how activity-based costing concepts can be extended to encompass quality-related costs arising from supplier deficiencies and customer dissatisfaction.

Using Activity-based Costing Concepts to Identify Quality-Related Costs The activity-based costing literature defines an activity as any discrete task that an organization undertakes to make or deliver a product or service, and a cost driver as a factor that causes or drives an activity’s costs. Shank and Govindarajan (1993) argue that quality-related activities represent some of the European Management Journal Vol 17 No 5 October 1999

ACTIVITY-BASED COSTING CONCEPTS FOR QUALITY IMPROVEMENT

primary cost drivers within an organization. Their claim is supported by studies that find many manufacturers spend up to 25 per cent of sales on the prevention, detection, and correction of quality problems, while service firms spend up to 40 per cent of sales on these activities (Danforth, 1986). The key to identifying profitable quality improvement opportunities is understanding which of these activities adds value to the customer and which can be reduced or reallocated to higher valued uses. When implemented with quality improvement in mind, a full activity-based costing system can provide valuable information on non-value-added activities and their drivers. Cooper et al. (1992), for example, report that some firms find it useful to rank all activities in their ABC systems based on their value to the customer. Two of the companies in their study coded each activity on 9 point scales, where 1 represented no value-added (e.g. making or correcting errors) and 9 represented high value to the customer. Using these rankings, the companies established improvement teams to focus on improving or eliminating the low- to medium-value activities. One limitation to this approach is the need to conduct a full-blown activity-based costing analysis to identify and rank each activity. However, a wide variety of service and manufacturing firms have found that simplified activity-based costing concepts can be

used to identify non-value-added activities and quality improvement opportunities, without the time and expense required to implement a full ABC system. Figure 1 illustrates the three levels of activity analysis a major telecommunications firm uses to identify and categorize quality-related expenditures1. The goal of the simplified activity analysis is to identify the activities and costs associated with preventing, identifying, and correcting quality problems. To do so, each activity is broken out into four categories: (1) essential work, which encompasses value-added activities required to produce a product or service right the first time, (2) prevention activities such as quality-related training and preventive maintenance that are carried out to avoid defects, rework, or delays, (3) appraisal activities such as inspections and data verification that are conducted to measure or test whether a product or service meets customer requirements, and (4) rework and failure activities such as problem resolution and defect correction that arise because products or services did not meet customer requirements. The three activity analysis levels offer various degrees of detail and accuracy, allowing the firm to adapt the analysis to the quality improvement team’s requirements and the available time frame and resources. Although the analysis levels vary in terms of data and time requirements, each follows a five step approach.

Figure 1 Quality-related Activity Analysis Steps and Alternative Levels of Detail. Activity Analysis Framework developed by a Major Telecommunications Firm

European Management Journal Vol 17 No 5 October 1999

493

ACTIVITY-BASED COSTING CONCEPTS FOR QUALITY IMPROVEMENT

1. Define Activities. The first step in the analysis is identifying the specific activities performed in a given organization, function, or process. Brainstorming, flowcharting, and interviewing are three methods available for defining activities. Brainstorming is usually performed in a group setting, and helps participants reach consensus on a list of activities. The list can then be condensed or grouped into major activities. Process flow charts can also be developed to identify key activities and isolate process failures or areas of complexity that may cause quality problems. Process information can be obtained from existing work records, procedures, or standards, or can be developed from special studies. When activity data are unavailable or difficult to acquire using these methods, experts and employees can be interviewed to create an activity list. Figure 2 provides a sample of the interview questions used by the telecommunications industry to elicit information on the process under analysis. 2. Assign Quality Cost Categories. Activities can now be grouped into the essential work, prevention, appraisal, and failure categories. Often an activity contains both quality-related tasks and work required to deliver a product or service. In such cases, the activity may need to be broken down into sub-activities, or estimates of the relative percentages in each category may be required. 3. Assign Costs to Each Activity. Once activities are identified and classified, the cost incurred on each quality-related task should be determined. In some job functions, existing time reports, work orders, or accounting records provide sufficient detail to identify employees’ activities, and can be used to estimate quality-related costs. Job codes

or titles may also indicate an employee’s primary activities and responsibilities. Defect costing can also be used to estimate quality-related costs. Most firms track a variety of operational quality measures such as defect rates, down time, service errors, and customer complaints. Standard costs per occurrence can be calculated and applied to these measures to translate these measures into costs. Figure 3 illustrates Hewlett-Packard’s computation of the standard cost associated with each customer return. Unlike typical activity-based costing applications, the standard cost includes the extra costs incurred by customers as well as by Hewlett-Packard. When time or cost data are not available or are difficult to estimate, special studies can be conducted to monitor the amount of time devoted to quality-related tasks. Activity logs provide a practical means for tracking the time spent on various activities over a defined period of time. Figure 4 provides the activity log used by Thermo-King (a leading manufacturer of commercial refrigeration) to track quality-related activities in its controller’s department, an organization that rarely has available data on quality-related costs. Finally, employees can estimate the amount or per cent of time spent on a given task. These figures can then be multiplied by labor rates or compared to total costs to estimate quality-related costs. However, experience indicates that employees’ estimates of the activities they perform and their duration are often inaccurate, making this method the least reliable technique for identifying qualityrelated costs. 4. Determine Improvement Opportunities with the Largest Potential Payback. Improvement projects are selected based on the potential return on invest-

Figure 2 Interview Questions used by a Major Telecommunications Firm for Conducting a Quality-related Activity Analysis

494

European Management Journal Vol 17 No 5 October 1999

ACTIVITY-BASED COSTING CONCEPTS FOR QUALITY IMPROVEMENT

Figure 3 The Total Estimated Cost of Processing an Average Customer Return at Hewlett-Packard’s Direct Marketing Division (DMD). Source: Adapted from Atkinson et al. (1994), p. 229

Figure 4 Worksheet used by Thermo-King to track Quality Failure Activities in its Controller’s Department

ment identified from the activity analysis, as well as other considerations such as strategic goals, competitive requirements, and available resources. Xerox incorporates four factors in the project selection process using the worksheet in Figure 5. Instead of relying solely on quality failure costs for prioritizing improvement projects, the firm also assesses three other factors: (1) the seriousness European Management Journal Vol 17 No 5 October 1999

or urgency of the problem as perceived by the external customer, (2) the extent to which the quality improvement team controls the process and the required solution to the selected quality problem, and (3) the relative difficulty of solving the problem, considering both the time to resolve the problem and the amount of resources required. 5. Make Improvements and Monitor Results. Once the 495

ACTIVITY-BASED COSTING CONCEPTS FOR QUALITY IMPROVEMENT

Figure 5 The Xerox United States Customer Organization Quality Improvement Project Selection Worksheet (Courtesy of Xerox Corporation, Cost of Quality: A Guide to Applications, prepared by the USCO Cost of Quality Team, Xerox Corporation (1987), p.21

selected improvement projects have been implemented, project results should be monitored based on the financial and operational goals established in the quality improvement plan. In most cases, project-level operational quality measures provide more timely and focused indicators of implementation success than accounting measures. Periodic activity analyses can then be conducted to determine whether the quality improvements have reached the bottom-line.

An Illustration The use of simplified activity-based costing concepts for quality improvement is illustrated with an improvement project conducted by the telecommunications firm. The quality improvement team was established to determine how to improve customer service on maintenance calls. The team’s efforts began by developing a detailed flow chart of the process for handling customer trouble reports, and the time needed for each activity in the process. Since time data were not available, a special study was conducted using a sample of trouble reports received during a single day. The results from the trouble report time study are shown in Figure 6. Given the narrow scope of the improvement project, the team chose to use its collective knowledge to categorize activities into prevention, appraisal, rework, and essential work classifications. Using data from the time study, the percentage of total process time taken by each activity was computed, and multiplied by an estimated cost per trouble report of $100 (developed in a prior study). The team then focused 496

its attention on identifying the rework activities offering the highest potential for improving response time and lowering cost. The time dimension for each activity was rated on a scale of 1 to 5, where 1 equaled low time consumption and 5 equaled high consumption. Similarly, potential payback was assessed by rating each activity from 1 to 5 according to how effective they believed the activity was from the customer perspective (1 being most effective, 5 being ineffective). For each activity, the time and payback ratings were multiplied to arrive at an overall result. The higher the overall rating, the greater the benefit from improving or eliminating the activity. The analysis indicated that the manual retest activity offered the greatest opportunity for improvement. The team brainstormed and came up with the following actions: 1. Eliminate manual retesting and begin tracking automatic test errors to determine their root cause(s). 2. Upgrade the system and assure automatic test accuracy. 3. Examine the causes and effects of trouble report information quality so that re-verification of problems is minimized. 4. Investigate system capability and enhancement potential to automatically flag customers desiring a call-back instead of continuing the manual monitoring procedure. 5. Develop follow-up procedures for monitoring changes and results. European Management Journal Vol 17 No 5 October 1999

ACTIVITY-BASED COSTING CONCEPTS FOR QUALITY IMPROVEMENT

Figure 6 Time Study Results for Six Trouble Reports received in a Single Day by a Telecommunications Firm. Activity Cost Categories are P — prevention, A — appraisal, R — rework/failure, or E — essential work

Supplier-Driven Quality Costs The activity analyses described above focus on internal processes. However, world class firms realize that effective quality improvement efforts extend beyond the organization’s walls. Philip Crosby, for example, estimates that 50 per cent of a firm’s failure costs are caused by the extra work involved in disposing, repairing, scrapping, or reworking defective purchased goods, as well as the costly schedule changes and downtime caused by poor quality supplies and materials (Smock, 1982). However, most accounting systems, including typical activity-based costing applications, do not isolate these costs. As a result, critical information for evaluating supplier performance and improving operations is lost. Recognizing the significant role suppliers play in achieving world-class operations, a growing number of companies are applying activity-based costing concepts when selecting and evaluating suppliers. Northrop Aircraft Division (NAD), for example, began the development of its cost-based ‘Supplier Performance Rating System’ (SPRS) in 1985. The division’s Productivity Measurement and Control European Management Journal Vol 17 No 5 October 1999

unit began surveying the quality failure identification and resolution process to identify and quantify the hidden costs associated with resolving supplierrelated discrepant paperwork and hardware. Their study found that more than 50 per cent of all production receipts during the previous year had paperwork or hardware discrepancies or did not meet delivery schedules, costing the division an estimated $10 million annually. The need for quality and delivery improvement prompted the Materiel and Quality Assurance departments to jointly sponsor the development of a system for (1) identifying responsibility for quality and delivery problems and fixing responsibility for corrective action, (2) motivating Northrop and its suppliers to improve quality and delivery performance, and (3) recognizing supplier excellence. Following the procedural flow diagrams in Northrop’s Nonconforming Material Control Manual, industrial engineers interviewed several hundred people from 120 departments, including receiving, materials review, engineering, reliability, procurement, inventory control, transportation, shipping, and field logistical support. The myriad combi497

ACTIVITY-BASED COSTING CONCEPTS FOR QUALITY IMPROVEMENT

nations of what, where, and how a failure is detected and who is involved in resolving a paperwork or hardware problem was then summarized into seven broad activity networks. Each type of failure activity or ‘event’ is assigned a standard cost based on industrial engineering studies of the hours required to resolve the problem. The number of events during the previous quarter is multiplied by the associated standard cost to obtain total failure costs due to the supplier. A supplier rating index is then calculated using the following formula: Supplier rating index ⫽ (Failure costs

Quality-Related Revenue Implications

⫹ total purchases)/total purchases. The following example illustrates the method used to calculate the supplier rating index. Total purchases during the rating period: $250,000 Quality failure costs during the rating period:$ Return to supplier (2 units @ $300 each) Under-shipment (5 shipments @ $350 each) Late delivery (3 shipments @ $500 each)

600 1750 1500 $3850

Total quality failure costs: Supplier rating index ⫽ ($3850 ⫹ $250,000)/$250,000 ⫽ 1.015

The output from the supplier rating system provides the basis for Northrop’s Key Plan Award for superior supplier performance, and assists in identifying which suppliers require help in meeting quality and delivery standards. The index also serves as a bid multiplier during the supplier selection process to determine the ‘total’ cost of purchasing from a given supplier. A comparison of two potential suppliers for the same product illustrates the use of the bid multiplier.

Quoted price per unit ⫻ Supplier rating index Total cost per unit

Supplier A $100 1.1

Supplier B $105 1.0

$110

$105

Based solely on quoted price, supplier A is the low cost supplier. After factoring in past quality performance using the supplier rating index, supplier B represents the better value. Other firms capture a wider variety of supplierrelated quality activity costs. Digital Equipment Corporation, for example, not only tracks the costs associated with quality and delivery failures, but also the costs of defect prevention (vendor surveys and quali498

fications, approval of samples, and specification revisions) and defect detection (incoming inspections and lab tests) (Carr and Ittner, 1992). The benefits from these systems can be quite significant. Northrop saw suppliers’ on-time delivery improve 20 per cent in the first year of the SPRS system. Similarly, another aerospace firm that uses a similar system saw lost dollars due to supplier deficiencies fall from $2.09 per $1000 purchased to $0.80 within a year of the system’s implementation.

The preceding techniques focus on quality-related costs, but in many cases the greatest improvement opportunities are the revenue gains achieved when higher quality products or services increase customer satisfaction and loyalty. The delivery of defective products and services to customers not only causes revenue losses when goods are returned or price concessions are granted, but also generates lost opportunities when existing customers do not return or potential customers choose competitors based on current customers’ unsatisfactory experience with the firm. A study by the US Department of Commerce (TARP, 1979), for example, found that dissatisfied customers tell approximately 19 others about their bad experience with the product. The negative wordof-mouth advertising that results can generate significant revenue losses. Most firms track some form of customer satisfaction measure, but few can link these measures to changes in revenues or profits (Ittner and Larcker, 1996). This limitation makes it difficult to identify the quality improvement projects offering the highest return on investment, or to evaluate the overall effectiveness of the quality program. Two methods have proven useful for integrating the revenue effects of quality problems into activity analyses. The first method combines cost results from the activity analysis with assessments of the activity’s effect on the customer. For example, Ansari et al. (1997) propose a ‘Value Index’ for determining how well quality-related spending matches the relative importance customers place on each quality attribute or activity. The Value Index equals the ratio of a customer’s perceived importance of a quality attribute to a firm’s spending on that dimension. When the index is greater than one, the importance of that attribute to customers is large relative to the firm’s spending on that dimension. Assume, for example, that market research has identified two customer-defined quality dimensions for a photocopier: downtime and print quality2. When asked to allocate 100 points to various quality attributes, customers rate the relative importance of downtime 80 per cent and print quality 20 per cent. The firm currently spends 70 per cent of qualityrelated costs on downtime problems and 30 per cent European Management Journal Vol 17 No 5 October 1999

ACTIVITY-BASED COSTING CONCEPTS FOR QUALITY IMPROVEMENT

on print quality issues. The Value Index for each attribute is computed as follows: Attributes Importance QualityValue index of copier ranking (%) related costs (col. 2/col. performance (%) 3) Downtime 80 Print quality 20

60 40

1.33 0.60

This analysis suggests that the firm may be spending too little on downtime problems relative to their importance to the customer, and may be able to increase the effectiveness of its quality improvement efforts by allocating more resources to downtime reduction. A second method for integrating revenue implications into quality-related activity analysis is the development of ‘lost sales’ models. Like activitybased costing, a lost sales model multiplies the frequency of each type of quality problem (or ‘driver’ of lost sales) by its standard cost per occurrence to compute the opportunity costs of quality problems. Consider the following example from a manufacturer of home appliances. Each unit retails for $200. Market research indicates that for serious defects, such as the appliance not turning on or not performing its intended function, customer reactions and related costs are as follows: — 100 per cent of the affected customers return the unit for repair under warranty or for exchange of the unit. The average cost to repair returned units is $25. — 10 per cent of the affected customers will not purchase the company’s products in the future. The company estimates a $150 profit loss for each of these customers. — 20 per cent of the affected customers write a letter of complaint to the manufacturer. The average cost to follow up on each complaint is $1.50. If 100 units are defective, the average failure cost for each serious defect shipped to the market is: (100 ⫻ $25) ⫹ (10 ⫻ $150) ⫹ (20 ⫻ $1.50) 100 seriously defective units ⫽ $40.30 per seriously defective unit. For minor defects, such as scratches on the case or missing instructions, market research estimates the following customer reactions and related costs: — 98 per cent of affected customers return the unit for repair under warranty or for exchange of the unit. The average cost for repair is $4. — 2 per cent take no action. — 5 per cent write a letter of complaint to the manufacturer, at an average follow-up cost of $1.50. European Management Journal Vol 17 No 5 October 1999

Thus the average failure cost for each minor defect that reaches the customer is: (98 ⫻ $4) ⫹ (2 ⫻ $0) ⫹ (5 ⫻ $1.50) 100 units with minor defects ⫽ $4.00 per unit with minor defects. By multiplying the number of serious and minor defects by their respective cost per occurrence, an estimate of the opportunity costs due to poor quality can be obtained. A number of firms have adopted this approach for identifying the drivers of quality-related opportunity costs and quantifying the magnitude of lost sales. Xerox’s approach uses estimates based on historical data. For example, a quality improvement team examining the customer support process for laser printers found that, due to pre-installation deadlines not being met, an average of five photocopy units were used as loaner machines at customer sites each month. Consequently, these machines could not be used elsewhere to generate business. Based on historical data, the team estimated the opportunity cost for these five machines as follows: 5 potential orders ⫻ $3000 ⫻ 12 months ⫽ $180,000 Another Xerox study found that 10 per cent of customers canceled their leases because of quality problems. Assuming that 15 per cent of these customers did not return, the estimated lost opportunity costs from these cancellations were: $60,000,000 revenue ⫻ 50% margin ⫻ 10% cancels ⫻ 15% lost ⫽ $500,000 Polaroid has implemented a more sophisticated lost sales model for its film products. This model is based on three criteria: (1) the severity of the defect (cosmetic, product partially usable, product unusable), (2) the net present value of expected future purchases of customers who experience a quality problem and of potential customers observe the defect and choose not to buy the company’s products, and (3) the net present value of expected future purchase patterns of customers who do not experience a quality problem. Based on market research, the company determines the level of severity customers attach to each type of defect, and then categorizes defects on a scale from minor to major. The probability that a customer will continue buying Polaroid film is assessed for each level of severity. For example, a minor cosmetic problem on the film tends to drive few customers away, but a major defect that renders the camera unusable can drive away not only existing customers, but also potential customers for Polaroid cameras and film who hear of the problem. Market research by the firm indicates that a customer who experiences a major defect typically tells five others, one of whom will decide not 499

ACTIVITY-BASED COSTING CONCEPTS FOR QUALITY IMPROVEMENT

to buy Polaroid products. The estimated number of current and potential customers who experience or observe the defect is multiplied by the probability of repurchase to estimate the number of lost customers. The number of lost customers per defect category is then multiplied by the estimated net present value of future sales per customer to predict the amount of lost sales for each type of defect.

Notes

Prior to implementing the lost sales model, Polaroid prioritized quality improvement projects based on the number of customer complaints for each type of problem. After developing the lost sales model, Polaroid found that quality problems causing the most complaints or internal costs do not necessarily translate into the greatest expected loss in future sales, leading the firm to shift the focus of its improvement projects and internal inspections to the elimination of defects with the greatest estimated impact on future sales.

Ansari, S., Bell, J., Klammer, T. and Lawrence, C. (1997) Measuring and Managing Quality Costs, Version 1.0. Richard D. Irwin, Homewood, IL. Atkinson, H., Hamburg, J. and Ittner, C. (1994) Linking Quality to Profits: Quality-Based Cost Management. ASQC Press, Milwaukee, WI. Carr, L.P. and Ittner, C.D. (1992) Measuring the cost of ownership. Journal of Cost Management Fall, 42–51. Cooper, R., Kaplan, R.S., Maisel, L.S., Morrissey, E. and Oehm, R.M. (1992) Implementing Activity-Based Costing: Moving From Analysis to Action. Institute of Management Accounting, Montvale, NJ. Danforth, D.D. (1986) A common commitment to total quality. Quality Progress 19(1), 17. Harrington, H.J. (1987) Poor Quality Costs. ASQC Quality Press, Milwaukee, WI. Ittner, C.D. and Larcker, D.F. (1996) Measuring the impact of quality initiatives on firm financial performance. Advances in the Management of Organizational Quality 1, 1–37. Shank, J.K. and Govindarajan, V. (1993) Strategic Cost Management: The New Tool for Competitive Advantage. The Free Press, New York. Smock, D. (1982) How to stem the tide of shoddy material. Purchasing, May 13, 51–57. Technical Assistance Research Programs (TARP) (1979) Consumer Complaint Handling in America: Final Report. US Department of Commerce, Washington, DC. Xerox Corporation (1987) Cost of Quality: A Guide to Application. Xerox Corporation, Rochester, NY.

Conclusions As the preceding examples illustrate, targeted activity-based costing concepts provide a powerful means for quantifying the financial impact of poor quality, enabling the firm to more effectively target specific activities for improvement or elimination. However, an important question to ask before proceeding is whether the customer values the quality attributes emphasized by the firm’s quality program. The ultimate test of any quality program is how well it helps the firm meet customer requirements or expectations. If a quality attribute (i.e. feature, function, service, activity, process, etc.) provides little value to the customer, it can be reduced or eliminated. Alternatively, if an attribute is highly valued by a customer, a firm may be able to improve performance by increasing spending on that attribute. Unfortunately, a quality-related activity analysis that is not based on extensive knowledge of customer requirements can lead firms to eliminate attributes that are valued by the customer, thereby reducing competitiveness. For example, an activity-based costing analysis by a leading global financial services firm identified customer complaint processing as a nonvalue-added activity that should be reduced or eliminated. However, subsequent analysis found that customers whose complaints were resolved to their satisfaction are far more loyal than other customers, while those whose complaints are not resolved satisfactorily are likely to leave. By eliminating complaint processing activities without a full understanding of customer requirements, the firm reduced its profitability. Only by combining extensive knowledge of revenue drivers with traditional analysis of cost drivers can the full benefit of quality-related ABC concepts be realized.

500

1. Special thanks to the anonymous telecommunications firm for providing access to the activity analysis examples reported in this paper. 2. This example is adapted from Ansari et al. (1997), p. 20.

References

CHRISTOPHER D. ITTNER, The Wharton School, University of Pennsylvania (Accounting Department), 2400 Steinberg Hall – Dietrich Hall, 3620 Locust Walk, Philadelphia, PA 19104-6365, USA. E-mail: [email protected] Christopher D. Ittner is Associate Professor of Accounting at the Wharton School, University of Pennsylvania and Co-Director of the Wharton Research Program on Value Creation in Organizations. His research and consulting interests include performance measurement, quality management practices, and the valuation of intangible assets. He is co-author of Linking Quality to Profits: Quality-based Cost Management.

European Management Journal Vol 17 No 5 October 1999