While much of the current literature on management accounting focuses on giving up old paradigms for newer ones, control and performance measurements for evaluation continue to dominate the real world. Typical individual performance measurements now being employed by major U.S. companies can have erroneous managerial implications, often rendering managers dysfunctional. The unsalutory effects can be seen in examples of three commonly used performance measurement techniques. The first two involve performance budgeting in the automotive industry. One is based on the use of standard labor dollars to budget overhead and the second centers on corporate production targets. The third example involves the use of activity-based costing.

Using Standard Labor Dollars in Performance Budgets

Airco, a large division of a major corporation, is a manufacturer of automotive heat exchange and environmental control products. There are over 100 manufacturing and support departments at Airco. Expenses are accumulated in both manufacturing and support departments based on a chart of accounts. Based on historical experience, variable overhead rates and fixed costs have been developed and aggregated for each burden account budgeted. The variable overhead rates are multiplied by the standard direct labor dollars for the month to generate the budgeted variable amounts. This is illustrated in Exhibit I, where Department A budgeted 300,000 compressor parts in a particular month which required .05128 weighted hours per unit. This translated into 15,384 standard hours at a $13 standard hourly labor rate which yielded approximately $200,000 of standard labor costs.

One ostensible benefit of flexible budgeting is that a budget can be compared at any output level in order to control costs. If Department A, for example, produced 330,000 compressor units, a 30,000 unit or 10% increase, the revised budget, reflecting a 10% or $27,000 increase in total variable overhead, as illustrated in Exhibit II, would result.

The pitfalls in this system of performance budgeting can be portrayed using the same example. As reflected in the monthly control report in Exhibit II, if the actual direct labor hours used to produce 330,000 units exceed budgeted hours, actual variable overhead expenses are likely to exceed those budgeted based on standard labor dollars. Most of the overhead expenditures show unfavorable (U) variances, not a happy performance result for the evaluation of the manager of department A. Overhead budget accounts were based on standard labor dollars rather than actual. First-line supervisors can do very little to control the interaction of men, machines, and material, which determine actual production, with the standard dollars budgeted for actual production. These supervisors are normally not in charge of hiring machine replacement, or material acquisition. While its use may generate process improvement ideas, the flexible budget as a control mechanism will be theoretically more accurate if based on the actual labor dollars.

The monthly control report in Exhibit II has been recast in Exhibit III to reflect variable overhead expenses based on actual direct labor dollars. The budgeted overhead amounts are based on the same fixed cost estimates and variable rates as in Exhibit I. The variable rates, however, are multiplied by actual labor dollars rather than standard labor dollars and many unfavorable overhead variances disappear. When the variable components of budgeted overhead are adjusted for the unfavorable direct labor variance, actual expenditures are predictably more in line with revised budget figures.

Manufacturing department and product-line supervisors are evaluated in part based on variances-to-budget. While variance-to-budget reports are supposed to aid cost control, many supervisors find it impossible to manage costs using these monthly budget reports. They do not understand where the budgeted costs come from or what causes a certain cost to be over budget (e.g., what specifically caused indirect labor to be high). Many supervisors do not know how or why certain variances are generated and are not adequately informed of the specific causes of the costs. In light of the previously noted reporting deficiencies, such confusion should not be surprising.

In addition to deficiencies and problems, the budgeting paradigm assumes that the driver, in this case labor dollars, is perfectly correlated with the individual overhead elements. If direct labor increases by one dollar, then indirect labor will increase by exactly 18 cents (or 18% as noted in Exhibit I), tool use by 7 cents, etc. The real world is not that neat. As activity-based costing literature indicates, multiple drivers are present in most companies. Even with multiple drivers, the drivers and the overhead elements are not perfectly correlated. The significance of this statistical reality is that a confidence interval would have to be established around the budgeted amount. The real world implication of this is that companies often examine illusory variances and engage in unwarranted behavior by utilizing variances as a punitive device when actually no inefficiency exists. Or they do the opposite and reward favorable variances when no particular efficiencies exist. A principle tenet of the total quality movement fathered by W. E. Deming is that fear in the workplace is a barrier to pride of workmanship and ultimately to productivity. A common source of fear is evaluation based on performance measures over which the worker has no control. Deming strongly advocates the elimination of such measures. According to Deming in Out of Crisis, no one should be blamed or penalized for performance that he cannot govern. Violation of this principle can only lead to frustration and dissatisfaction with the job, and lower production. Deming further argues that all processes experience natural variation which is part of the system or process itself. Blaming a department manage for unfavorable variances (or praising her for favorable variances) that are the result of natural variation is a waste of time and demoralizing to the manager being blamed or those not being praised.

Production Targets

An example of problems that may result from the use of production targets is provided by a sister automotive assembly plant. Output at the plant is determined by corporate management in Detroit. In order to meet management's target, output is increased or decreased toward the end of the month, depending on whether production is behind or ahead of schedule. Increases in output are achieved by working overtime and decreases are realized by idling the plant. The plant is, of course, exceptional at meeting the budget demands set by Detroit. Unit cost figures are relatively high as a result of alternating overtime and idle time, and the plant has recently been threatened with closure.

Emphasis is placed on keeping the final line running at any expense and meeting budgeted production. When the final line runs uninterrupted, the main objective of meeting the corporate quota is achieved. Superintendents manage the subassembly processes strictly by fear, doing whatever it takes to keep the final line running. If the final line does not shut down, they feel the subassembly processes are well managed and in balance with the final line. The flow of the completed subassembled components, however, is very uneven. A completed component spends excessive queue time hanging from a conveyor waiting to be married with a vehicle on the final line since meeting the budgeted production is mandated.

With over 26 miles of conveyors, the cost of inventory continuously hanging from the conveyors is tremendous. The superintendents are relatively unconcerned with this excessive inventory since they are rewarded for meeting budget. The safety bank of inventory hides inefficiencies in the subassembly processes. When a problem arises causing the subassembly process to stop, the safety bank keeps the final line running. If the safety bank is depleted to a less than desired level, they work overtime to replenish the bank. The costs of overtime coupled with the costs of hanging inventory are enormous. Again, line management, however, is relatively unconcerned if the production quota is met since they are evaluated on achieving budgeted production.

The Use of Activity-Based Costing

While an activity-based costing system with multiple drivers may partially rectify some mechanical problems associated with performance budgeting, this approach may not completely rectify the behavioral problems employees will have adjusting to a new system. Mead Data Central, a subsidiary of the Mead Corporation at the time, recently went through a "rightsizing" process which, of course, is a euphemism for downsizing. To accomplish the downsizing, Mead Data Central initiated activity-value analysis (AVA).

AVA was used to cost out each activity, determine if the activity was adding value to the company, and if the activity could be eliminated or combined with another activity without reducing quality or customer satisfaction. In a weekly newsletter, MDC management made the following statement:

MDC must change to operate more efficiently and effectively. It makes sense TABULAR DATA FOR EXHIBIT I OMITTED for the employees to have a say in what improvements are implemented. The employees know where things can improve and how to make it better. We feel that AVA is the fair way to get everyone involved to find the better answer. Then, when it comes time to make the tough decisions, we all will have an understanding of why things must be changed.

EXHIBITII

PRODUCTIONDEPARTMENTA

MONTHLYCONTROLREPORT

(OVERHEADBUDGETBASEDON220,000STANDARDLABORDOLLARS)

AccountsBudgetActualVariance

4000-80200Directlabor$220,000$242,000$22,000(U)

5000-01000Indirectlabor$59,400$64,400$5,000(U)

5000-02000Operatingsupplies14,30012,1002,200(F)

5000-03000Tools15,40016,200800(U)

5000-04000Utilities35,90036,500600(U)

5000-05000Maintenance75,00077,1002,100(U)

5000-06000Fringebenefits187,000198,50011,500(U)

5000-07000Scrap16,50018,7002,200(U)

5000-08000Taxes,depreciation60,00062,0002,000(U)

5000-09000Sundry4,1003,900200(F)

5000-10000Transfer1,0001,000

5000-11000Projects10,00011,4001,400(U)

Totaloverhead$478,600$501,800$23,200(U)

330,000Unitsx.05128St.Avg.Hrs./Unitx$13/Hr.approximately

equalto$220,000standardlabordollars.

U=Unfavorable;F=Favorable

The more precise model generated by AVA, however, did not totally solve the problems. Many of the activities defined by the company and external consultants as nonvalue added did not go away when the people were eliminated. This produced a situation where the remaining employees felt overburdened, and morale deteriorated. Although there was an immediate benefit to the bottom line as a result of cost savings, a growing body of managers believe that in the long run the only way to insure profitability is with enhanced revenues. Given the reduction in personnel by many companies such as Mead Data, increasing revenue would be difficult to achieve.

EXHIBITIII

PRODUCTIONDEPARTMENTA

MONTHLYCONTROLREPORT

(OVERHEADBUDGETBASEDON242,000ACTUALLABORDOLLARS)

AccountsBudgetActualVariance

4000-80200Directlabor$220,000$242,000$22,000(U)

5000-01000Indirectlabor$63,360$64,400$1,040(U)

5000-02000Operatingsupplies15,73012,1003,630(F)

5000-03000Tools16,94016,200740(F)

5000-04000Utilities37,99036,5001,490(F)

5000-05000Maintenance80,50077,1003,400(F)

5000-06000Fringebenefits200,200198,5001,700(F)

5000-07000Scrap18,15018,700550(U)

5000-08000Taxes,depreciation60,00062,0002,000(U)

5000-09000Sundry4,2103,900310(F)

5000-10000Transfer1,2201,000220(F)

5000-11000Projects10,00011,4001,400(U)

Totaloverhead$508,300$501,800$6,500F

EXHIBITIV

PRODUCTIONDEPARTMENTA

ACTUALCOSTDATA

AccountsMarchAprilMay

4000-80200Directlabor$181,300$194,200$189,300

5000-01000Indirectlabor$53,400$54,400$53,800

5000-02000Operatingsupplies15,30017,60015,900

5000-03000Tools14,00014,90014,100

5000-04000Utilities34,10035,60034,400

5000-05000Maintenance70,10073,10071,900

5000-06000Fringebenefits169,900170,100167,700

5000-07000Scrap15,00013,70013,900

5000-08000Taxes,depreciation59,00058,20059,100

5000-09000Sundry4,1004,6004,500

5000-10000Transfer800(300)200

5000-11000Projects9,0008,6009,300

Totaloverhead$444,700$450,500$444,800

Units300,000330,000340,000

Unitlaborcosts$.60$.59$.56

Unitoverheadcosts$1.48$1.37$1.31

This fine tuning of accounting data by American companies to reward, evaluate, and eliminate people can be contrasted with the Japanese philosophy employed at Marysville, Ohio. Tougher competition from the Big Three in early 1993 forced Honda to cut production significantly for the first time in 10 years at this 5,400 worker facility. Instead of using AVA to lay off workers, Honda used the time to intensify training and teach workers new skills. Honda has since recaptured some market share. The pertinent question is - when the world recession ultimately goes away, as it surely will, who will be in the better position to reap the rewards, MDC or Honda?

Wallpaper as an Alternative

Japanese companies engage in relatively little performance budgeting - evidently they believe that it suffers from some of the mechanical and behavioral problems previously discussed. Rather than performance budgets, these companies post much simpler actual cost data and "wallpaper" these figures around the work place for workers to examine on an ongoing basis. These systems are simple, encourage teamwork and, most importantly, are not used as an individual reward system. If we return to our original example at Airco, the wallpapering technique would resemble Exhibit IV. There is no budgeted data and no variances are generated. The data shown is much simpler and focuses on continuous improvement, which the Japanese deem more vital to success than performance budgeting.

The wallpapering approach is consistent with the total quality management philosophy. It does away with variance analysis and eliminates the behavioral problems associated with the use of arbitrary numerical targets. Wallpapering also has the advantage of providing regular feedback to the people who are closest to the process or activity being monitored (in this case, overhead expenditures). This feedback fosters an awareness of overhead costs which may lead to employee-initiated cost savings.

For example, posting actual tool cost data near the tool crib may increase awareness of the expense associated with tool usage, making employees more conscientious about following policies designed to extend the useful life of tools and perhaps inspiring them to come up with a few ideas of their own about how to accomplish that goal. In this way, cost reduction is perceived as a team effort, rather than a threat to be used in performance evaluation.

Involving employees in this way can have many benefits. In Out Of Crisis, Deming cited an example taken from a report on a Japanese automotive stamping company. "A positive team spirit, along with intense loyalty and high motivation, are further enhanced by effective communication. Visual communications in the plant - posters, signs, and graphs - were extremely numerous."

Performance Measurement Top Choice

In 1993, the Journal of Cost Management surveyed their readers to determine which topics they were most interested in. Surprisingly, the top choice was not activity based management, quality or target costing - it was performance measurement. The right kind of performance measurement, based on continuous improvement and achieved through the wallpapering of actual results will have a salutary effect on the organization as a whole. The American infatuation with budgets and variances will often provide for the antithesis - suboptimization, resistance to change, budgetary slack, and unwarranted complacency regarding budgeted achievement. In an interview with USA Today, Louis Gerstner, the chief executive of IBM, summarized the dichotomy in the following fashion:

I need to convince the very smart people that populate this company that shifting from a pure individual performance basis to a team orientation - to a sharing approach - is in the best interests of the company and in their interest and it's going to be fun and we're going to win.

Susan Lightle, PhD, CPA, is an assistant professor, and John Talbott, DBA, CMA, is a professor, both at Wright State University.