Cost Allocation Problem

Tektronix

BACKGROUND

Tektronix, Inc. headquartered in Portland, Oregon, is a world leader in
the production of electronic test and measurement instruments. The
company's principal product since its founding in 1946 has been the
oscilloscope (scope), an instrument that measures and graphically
displays electronic signals. The two divisions of the Portables Group
produce and market high and medium-performance portable scopes.

Tektronix experienced almost uninterrupted growth through the 1970s
based on a successful strategy of providing technical excellence in the
scope market and continually improving its products in terms of both
functionality and performance for the dollar. In the early 1980s,
however, the lower priced end of the division's medium-performance line
of scopes was challenged by an aggressive low-price strategy of several
Japanese competitors. Moving in from the low-price, low-performance
market segment in which Tektronix had decided not to compete, these
companies set prices 25 percent below the U.S. firm's prevailing prices.
Rather than moving up the scale to more highly differentiated products,
the group management decided to block the move.

The first step was to reduce the prices of higher-performance,
higher-cast scopes to the prices of the competitors' scopes of lower
performance. This short-term strategy resulted in reported losses for
those instruments. The second step was to put in place a new management
team whose objective was to turn the business around. These managers
concluded that, contrary to conventional wisdom, the Portables Group
divisions could compete successfully with foreign competition on a cost
basis. To do so, the divisions would have to reduce costs and increase
customer value by increasing operational efficiency.

PRODUCTION PROCESS CHANGES

The production process in the Portables Group divisions consisted of
many functional islands, including etched circuit board (ECB) insertion,
ECB assembly, ECB testing, ECB repair, final assembly, test, thermal
cycle, test/QC, cabinet fitting, finishing, boxing for shipment, and
shipment. The new management team consolidated these functionally
oriented activities into integrated production lines in open work spaces
that allow visual control of the entire production area. Parts inventory
areas were also placed parallel to production lines so that at each work
station operators would be able to pull their own parts. This in essence
created an early warning system that nearly eliminated line stoppages
due to stockouts.

Additional steps that were taken in the early to mid 1980s to solve
managerial and technical problems include implementation of just-in-time
(JIT) delivery and scheduling techniques and total quality control
(TQC), movement of manufacturing support departments into the production
area, and implementation of people involvement (PI) techniques to move
responsibility for problem solving down to the operating level of the
divisions. The results of these changes were impressive: substantial
reductions in cycle time, direct labor hours per unit, and inventory,
and increases in output dollars per person per day and operating income.
The cost accounting group had dutifully measured these improvements, but
had not effectively supported the strategic direction of the divisions.

COST ACCOUNTING SYSTEM

DIRECT MATERIALS AND DIRECT LABOR

The total manufacturing cost of the newest portable scopes produced with
the latest technologies has 75% direct materials, 3% direct labor, and
22% factory overhead. In most cases, direct materials and direct labor
are easily traced to specific products for costing purposes. Prior to
the mid 1980s, however, the divisions' attempts to control direct labor
had been a resource drain that actually decreased productivity.

There were approximately twenty-five production cost centers in the
Portable Instruments Plant. Very detailed labor efficiency reports were
prepared monthly for each cost center and each major step in the
production process. In addition, an efficiency rating for each
individual employee was computed daily. Employees reported the quantity
of units produced and the time required to produce them, often
overestimating the quantity produced to show improved efficiency against
continually updated standards. The poor quality of collected data
resulted in semi-annual inventory-downs when physical and book
quantities were compared.

"The inadequacy of our efficiency reporting system became clear when we
analyzed one of our new AT production lines," commented Michael Wright,
Financial Systems Application Manager. "On a AT manufacturing line, once
the excess inventory has been flushed out, it is essentially impossible
for any person to work faster or slower than the line moves. However, if
one person on the line is having a problem, it immediately becomes
apparent because the product flow on the line stops. Corrective action
is then taken, and the line is started up again.

"On that line, the system told us that the efficiency of each of the
workers was decreasing. However, stepping back from the detail of the
situation allowed us to look at the overall picture. We found that the
costs incurred on the line were gang down and its product output was
going up. Obviously, it was becoming more, not less, efficient."

The quantity of direct labor data collected and processed also was a
problem. Production employees often spent twenty minutes per day
completing required reports when they could have been producing output.
Additionally, the accounting staff was processing 35,000 labor
transactions per month to account for what amounted to 3 percent of
total manufacturing cost.

"Transactions cost money," observed John Jonez, Group Cost Accounting
Manager, "and lots of transactions cost lots of money."

In response to these problems, the group accounting staff greatly
simplified its procedures. It abandoned the measurement of labor
performance for each operation, and greatly reduced the number of
variances reported. The number of monthly labor transactions fell to
less than 70, allowing the staff to spend more time on factory overhead
allocation and other pressing issues.

FACTORY OVERHEAD

The product costing system allocated all factory overhead costs to
products based on standard direct labor hours. A separate rate was
computed for each manufacturing cost center. This system led to rapidly
increasing rates: the direct labor content of the group's products had
been continually decreasing for years, while factory overhead costs were
increasing in absolute terms.

"Because the costing system correlated overhead to labor, our engineers
concluded that the way to reduce overhead costs was to reduce labor,"
commented Jonez. "The focus of cost reduction programs therefore had
been the elimination of direct labor. However, most of this effort was
misdirected, because there was almost no correlation between overhead
cost incurrence and direct labor hours worked. Our system penalized
products with proportionately higher direct labor, but it wasn't those
products that caused overhead costs. We proved that. We attacked direct
labor and it went down, but at the same time overhead went up.

"We therefore knew that we needed a new way to allocate overhead. More
fundamentally, we needed a way for the cost accounting system to support
the manufacturing strategy of our group. The objective was clear-to
provide management with accounting information that would be useful in
identifying cost reduction opportunities in its operating decisions as
well as provide a basis for effective reporting of accomplishments."

APPROACH TO METHOD CHANGE

Initial Steps

The first step taken by Wright and Jonez in developing a new overhead
allocation method was to establish a set of desirable characteristics
for the method. They decided that it must accurately assign costs to
products, thus providing better support for management decisions than
the old method. It must support the AT manufacturing strategy of the
Portables Group. It also must be intuitively logical and easily
understandable by management. And finally, it must provide information
that is accessible by decision makers.

The next step was to interview the engineering and manufacturing
managers who were the primary users of product cost information. These
users were asked, "What is it that makes your job more difficult? What
is it that makes certain products more difficult to manufacture? What
causes the production line to slow down? What is it that causes
overhead?" The answers to these questions were remarkably
consistent-there were too many unique part numbers in the system. This
finding revealed a major flaw in the ability of the direct labor-based
costing method to communicate information critical for cost-related
decisions. Manufacturing managers realized there were substantial cost
reduction opportunities through the standardization of component parts,
but there was no direct method to communicate this idea to design and
cost-reduction engineers who made part selection decisions.

Although difficult to quantify, some costs are associated with just
carrying a part number in the database. Each part number must be
originally set up in the system, built into the structure of a bill of
materials, and maintained until it is no longer used. Moreover, each
part must be planned, scheduled, negotiated with vendors, purchased,
received, stored, moved, and paid for. Having two parts similar enough
that one could be used for both application requires unnecessary
duplication of these activities, and therefore unnecessary costs.
Standardizing parts results in several indirect benefits. Fewer unique
part numbers usually means fewer vendors and greater quality of
delivered parts. It also means smoother JIT production, fewer shutdowns
of manufacturing lines, and greater field reliability. These
observations led to a preliminary consensus on the need to develop a
product costing method that would quantify and communicate the value of
parts standardization.

COST ANALYSIS

"To confirm our assessment," stated Jonez, "we segmented the total
manufacturing overhead cost pool. The costs of all cost centers
comprising the pool were categorized as either materials-related or
conversion-related based upon rules developed in conjunction with
operating managers.(See T-1)

"Material-related costs pertain to procurement, scheduling, receiving,
incoming inspection, stockroom personnel, cost-reduction engineering,
and information systems. Conversion-related costs are associated with
direct labor, manufacturing supervision, and process-related
engineering. Application of the rules resulted in an approximately 55/45
split between materials overhead (MOH) and conversion overhead (COH).
This finding further confirmed the inadequacy of the existing method,
which applied all overhead based on direct labor."

The accounting analysts decided to focus their initial efforts on the
MOH pool. To improve their understanding of the composition of the pool
and thus assist them in developing a method for its allocation, Wright
and Jonez consulted operating managers and further segmented it into:

1. Costs due to the value of parts,

2. Costs due to the absolute number of parts used,

3. Costs due to the maintenance and handling of each different part
number and

4. Costs due to each use of a different part number.

The managers believed that the majority of MOH costs were of type (3).
The costs due to the value of parts (1) and the frequency of the use of
parts (2 and 4) categories were considered quite small by comparison.

The analysts therefore concluded that the material-related costs of the
Portables Group would decrease if a smaller number of different part
numbers were used in its products. This cost reduction would result from
two factors. First, greater volume discounts would be realized by
purchasing larger volumes of fewer unique parts. Second, material
overhead costs would be lower. "It was the latter point that we wanted
our new allocation method to focus on," commented Wright.

"Our goal," continued Jonez, "was to increase customer value by reducing
overhead costs. Our strategy was parts standardization. We needed a
tactic to operationalize the strategy."

PAGE

PAGE 4