Predicting Productivity Performance 0108
Predicting Productivity Performance 0108
PERFORMANCE
ARTHUR G. DAVIS
Figure 2: Formula for calculating total production hours required to attain any desired productivity index.
standard, and the production rate, or work-in those direct-labor hours, divided by the product of
progress adjustment factor. (All output factors the adjusted base-year measure and the
appear in the upper half of the left-hand column productivity index desired (see Figure 2).
in Figure 1.) Total production hours comprise direct-labor
The Productivity Index. Dividing adjusted hours, indirect-labor hours, downtime hours, and
earned hours (output) by total production hours salaried employee hours. (Because of its
(input) yields an adjusted productivity measure, importance, we also separated quality assurance
which is divided by the base-year adjusted hours from total indirect labor for our study.) To
productivity measure to obtain the quarterly progress toward the desired productivity index,
productivity index. Thus, any improvement or the production manager must calculate the time
decline in productivity during the current period is contribution of each of these elements to the total
reflected against the base period (usually the production hours; knowledge of total hours alone
previous year). After all manufacturing is insufficient for effecting meaningful change.
departments calculate this index, results are then Indirect labor, downtime, and quality
combined to determine an overall productivity assurance elements also influence productivity
index for the business unit (see Figure 1). performance significantly, yet these factors are
often overlooked. We analyzed the key indirect
Developing the Productivity Forecast Model areas of the packaging operation in order to
understand the particular characteristics of each
We began developing our predictive model in as related to direct labor. The results of this study,
the pharmaceutical firm's packaging operation by a part of which is shown in Table 1, revealed that
closely examining the individual factors involved the various elements of indirect labor were pro-
in computing the productivity index. Instead of portional to direct labor and thus displayed
computing that index, however, we recombined consistent proportional relationships to total
the elements to determine the total production hours.
manufacturing hours needed to achieve any Ratios of direct labor to indirect-labor
desired productivity index. categories and to downtime were derived by
If we are given a forecast for direct-labor analyzing past performance; we also noted
hours to be worked, along with an index of the where the highest and lowest ratios of direct to
anticipated performance efficiency of that direct indirect labor occurred. For most of the indirect
labor, we can compute total production hours categories we studied, the range between the
required to attain any desired productivity index highest (most optimistic) and lowest (most
by using the following productivity index equation: pessimistic) ratios generally followed a normal
total production hours would equal the product of distribution. To establish the validity of these
the standards-change factor, forecasted observable ratio relationships, statistical testing
Total
Direct Indirect Line Quality
Labor Labor Supervisor Mechanic Assurance
Year and Hours Hours Hours Hours D. L. Hours
Month (D.L.) (I.L.) D.L./I.L. (L. S.) D.L./L.S. (Mech.) Mech. (QA) D.L./QA
Year 1
October 15,547.3 9,271.0 1.68 2,336.1 6.66 2,611.8 5.95 1,300.0 11.96
November 12,120.0 7,129.4 1.70 1,760.3 6.89 2,078.7 5.83 1,560.5 7.77
December 11,962.2 6,886.7 1.74 1,684.3 7.10 2,059.9 5.81 1,437.2 8.32
Year 2
January 13,366.6 7,953.3 1.68 2,031.4 6.58 2,370.7 5.64 1,592.2 8.40
February 12,361.5 6,876.3 1.80 1,615.1 7.65 2,194.6 5.63 1,587.9 7.78
March 12,017.9 7,020.6 1.71 1,793.3 6.70 2,028.3 5.93 1,420.6 8.46
April 13,602.1 8,346.8 1.63 1,993.3 6.82 2,453.2 5.54 1,602.9 8.49
May 10,053.1 5,808.9 1.73 1,213.6 8.28 1,694.9 5.93 971.5 10.35
June 9,516.2 5,947.0 1.60 1,358.7 7.00 1,721.2 5.53 1,013.8 9.39
July 10,818.6 6,935.7 1.39 1,434.6 7.54 2,034.3 5.32 1,298.7 8.33
August 9,166.7 5,725.3 1.60 1,204.9 7.61 1,867.4 4.91 1,261.6 7.27
September 8,165.4 5,313.4 1.54 1,122.5 7.29 1,647.4 4.97 1,033.0 7.92
Table 1: Sample of the summary data on the relationships between direct labor and various indirect-labor elements.
Highest Lowest
was employed. It supported our initial
Indirect-Labor Standard Ratio Ratio
findings.
Category Ratio Attained Attained
Line supervisors 7.00 8.28 6.58
By determining the range of ratios, we
Mechanics 5.76 5.95 4.91
then established standards or benchmarks, Quality assurance 9.00 11.96 7.27
in the form of ratios of direct to indirect Downtime 12.00 15.94 9.55
labor. For each category of indirect labor, Case sealers 36.50 40.41 33.10
the benchmark was set at, or slightly Floor services 19.00 20.15 16.04
above, the mean, depending on the Bulk stock handlers 19.00 20.77 16.53
distribution of the data. After these Group leaders 36.74 42.71 33.81
standards were established, as shown in Coordinator,
Table 2, the relationship of each indirect timekeeper,
variable to total production hours was instructor 24.00 30.77 19.96
calculated using regression analysis. With Secretary, scheduler,
the relationships among all the elements in clerk 18.10 19.17 17.25
place, we then adapted the equation to a Table 2: Ratios of direct to indirect labor hours.
computer program, which calculated the
maximum allowable hours for each indirect-labor both labor types by comparing the forecast hours
element. We were then able to compute the required with actual hours worked. Such an
appropriate indirect labor requirements to evaluation reinforces to supervisors the
achieve a desired productivity level when importance of assessing the effectiveness of
supplied with a projection of the direct labor input various work methods, equipment, and
and its anticipated performance efficiency. employees. In daily morning conferences, the
department manager, key supervisory personnel,
Applying the Productivity Model and a representative of industrial engineering
assess the previous day's performance, highlight
Our model could now enable supervisors to problem areas for correction, and review the
play "what if" games, testing the productivity current day's objectives.
consequences resulting from different To complete the system and ensure
combinations of direct and indirect labor. performance follow-up, a large progress chart is
Although this model was initially applied to the displayed in a conspicuous area of the plant.
plant's packaging department, a decision-making (Figure 3 is a simplified version of the chart.) This
procedure soon developed that required the chart lists those labor elements that have had the
involvement of all supervisory and plant greatest impact on productivity and highlights
management personnel. categories in which performance exceeded the
The procedure begins by determining the forecast thereby emphasizing to all personnel the
direct labor requirements (in hours) for the importance of productivity and the various
following day. Together, the coordinator, contributions of each group.
scheduler, and department manager make a In the budgetary process, the productivity
projection in conjunction with production control. forecast model has been used in the areas of
After direct-labor requirements are established, short- and long-range head-count planning.
the coordinator informs the industrial engineering Product forecasts by marketing and Production
department of the expected number of control can be readily translated into indirect and
direct-labor hours to be worked, the anticipated direct labor requirements. For example, when
level of performance efficiency against the frozen production activity was projected to decline as a
standard, and the productivity index goal for the result of management-mandated inventory
day. Computer to determine both the reduction goals, the need to reduce indirect-labor
commensurate labor requirements for all resources commensurate with direct labor was
indirect-labor categories and the appropriate anticipated. Transfers, loans, and other
allowances for downtime and quality assurance contingencies were arranged in, sufficient time so
in turn processes this information. that these labor reductions could occur in proper
The indirect-labor requirements are then order. Previously, it had been difficult to
communicated to the production department's effectively coordinate indirect-labor requirements
scheduler and timekeeper, to be used as a guide so that they would be commensurate with
in determining the day's final crewing forecasted production.
requirements for both direct and indirect labor.
Indirect-labor requirements are an indicator of
where to place emphasis in optimizing the
contributions of all personnel and thereby
achieving the desired productivity index.
The performance of direct and indirect labor
is then evaluated against established targets for
Cost estimating is another area of application,
particularly when indirect-labor costs are
assessed. Such data are easily obtained once
the direct-labor standard is computed and the
Art Davis is principal of the Davis
production forecast is known. Forecasted direct
Group, a Chicago area consultancy
hours are calculated, and estimates of
specializing in business process
performance and the productivity index can then
improvement and total quality. He
be accessed to compute commensurate
has delivered many training
indirect-labor resource requirements.
programs in Six Sigma for Motorola,
When applying the forecasting model, one
Caterpillar, ITW, Litelfuse and other
should remember that the various ratios of any
firms. For more information, call
direct-labor category might change over a period
847.685.9820.
of time due to ongoing, subtle changes in
methods, equipment, and personnel. An annual
review of the standard ratios for indirect labor
should therefore be conducted.