Printer Friendly

Yes, ABC works with purchasing, too.

As every management accountant knows, misallocating costs can devastate a business. If management can't determine the real cost of a product, pricing decisions are no better than guesswork. Because of this problem and because activity-based costing (ABC) has proven effective and been increasingly popular in determining real costs in the manufacturing arena, we decided to test the feasibility of applying this costing method to purchasing.

The test showed that ABC works well when used to analyze a purchasing department, and the information it provides is especially helpful when establishing a product's price and in determining its contribution to profit.

This article describes the method used in this test and the issues related to applying the ABC system to the operations of a purchasing department.


To determine the feasibility of ABC to purchasing, we asked a midwestern company that was having costing problems if we could examine its purchasing data. The company agreed, with the proviso that neither its identity nor its product be disclosed. However, to understand the data, readers must know that the products are assembled from the company's own foundry parts and from components purchased from outside vendors. And while some products are produced for inventory, most are for customer orders.

ABC is an accounting method in which costs are identified with the activities driving them. ABC establishes the all-important link between decisions that cause activities and activities that cause costs. By focusing on cost drivers, management is in a much better position to identify and eliminate costs that do not add value. As a resuit, costs can be better managed and products more accurately costed. For a more detailed discussion of ABC, see the book, Management Accounting, by Don R. Hansen and Maryanne M. Mowen (Cincinnati, Ohio: South-Western Publishing Co., 1992, pages 233-73).

For many years, the company allocated purchasing department costs by using a rate based on direct labor costs. When this amount was combined with other overhead, it amounted to overhead costs of more than 800% of direct labor cost. Then, in 1986, management decided the system was inadequate when it realized that direct labor in the plant was less than 7% of total costs. It therefore switched to a percentage-of-product cost system, which assumes there is a relationship between the component price and purchasing department costs and price levels will remain constant.

To come up with a new allocation for the percentage-of-product cost system, the company used actual data from 1985 for the total cost of purchased parts and the purchasing department's operating costs. By taking all costs assigned to the purchasing department--including wages, taxes, utilities, insurance, depreciation and corporate overhead allocations--and dividing this amount by the purchase price of all component parts, management determined a new allocation rate of 4%.

Significant errors remained: The system either overallocated or underallocated department costs to the product by 24% to 79%. And purchase orders (POs) for expensive material were being charged with a much higher allocation than those with inexpensive material.


The data in exhibit 1, page 61, show how the 4% figure was calculated. As the figures indicate, the company underallocated purchasing department costs in 1986 and 1987 and overallocated in 1988 and 1989. For example, in 1986 the actual percentage of purchasing department costs to total purchases was 7.8% ($422,000), yet only $217,640 (or 4%) was allocated to products. There were two reasons for this:

* Purchasing department costs were over budget in both years because increased customer requirements necessitated that the company provide more vendor contact. This action naturally drove up expenses.

* The price level of the items purchased declined, resulting in an underallocation.

These two factors contributed to the company erroneously reporting higher-than-expected profit margins and product profitability in that year.

In 1988 and 1989, because of overallocating purchasing department costs, the company underspent the purchasing department budget each year while experiencing a general price increase for its components.

Overallocation can be as troublesome as underallocation. Since the component parts received more than their fair share of purchasing costs, those profit margins appeared to be shrinking. The danger is that, based on this misleading information, a company could decide to discontinue a profitable product that erroneously showed a poor financial return.

The percentage-of-product cost system also penalized products that had more expensive component parts. For example, one casting (a major component of a product) cost the manufacturer $1,000, while the casting for another product was only $100. Using the 4% basis, the $1,000 casting would be allocated $40 ($1,000 x .04), while the $100 casting would be allocated $4 ($100 x. 04). Since it does not cost the purchasing department more to order the expensive casting, the allocation system used in this approach clearly was not reflecting the reality of the business.


With overwhelming evidence that the percentage-of-product cost system was inaccurate, management agreed to run the ABC test. The first step was to identify the cost drivers. Since the purchasing department procures parts by preparing and sending a PO, the PO had to be the main cost driver. Clearly, then, the PO became the focus of the study.

The purchasing department processes 10,000 POs a year. In addition, there are sundry costs and division cost transfers (allocations from corporate). While these costs can be applied equally to each of the four purchasing agents, other costs, such as casting maintenance, depreciation of casting patterns, insurance and taxes, are unique to the single castings purchaser.

Maintenance and repair POs account for 3,000 of the 10,000 POs that are processed annually. Since these costs are not directly related to the product, we decided to allocate these costs evenly to the remaining 7,000 POs.

We also had to decide how to handle the purchase of "controlled materials"--component parts ordered only to fulfill a particular customer specification. These orders require special contact with vendors, so they entail extra costs. Each year, 200 such POs are processed, and an estimated 15% of one agent's efforts are expended on this; consequently, 15% of the agent's costs were. assigned to these efforts.


Based on these data and assumptions, we prepared the table in exhibit 2, page 62. Under the ABC approach, the castings buyer's per-order cost is $171.24. Per-order costs for the other three purchasers are $58.84. The per-order cost for controlled materials is calculated as follows:
 Cost per
 Controlled material costs controlled
 IS46,840.50) = material PO
Number of POs (200) ($234.20)

Once the per-order cost was calculated, this cost was allocated to the individual parts. Several possible ways were available. In the first option, we computed an average cost per part for each class of PO, which required determining the average number of parts and POs per class, using the budgeted cost per PO to calculate a flat rate per part. The formula is

Budgeted cost per PO in a class

Cost ollocated x number of PUs in this class

per part Number of parts in a class

This produced a figure for one flat rate for all the pieces within a class, no matter how big or small the order. An extension of this option used the company's "family" grouping of parts. This group-specific approach allowed for parts usually ordered in larger lots to bear a smaller amount of costs than those ordered in smaller lots. While the formula is substantially the same, it used the number of POs for a family and number of parts in a class.

To illustrate, the assumption is that the company needed to forecast the number of parts and planned to purchase 204 of them by processing 18 POs. It would apply this formula:

Number of POs (18) x

cost per PO ($171.24) = Allocated cost ($15.11)

Number of parts (204)

An alternative was to allocate the PO cost to all products on a particular PO. For example, if only one part is purchased with a PO, the cost is $171.24. If 100 are purchased, the allocated cost is $1.71 per item. Unfortunately, this approach was not acceptable because it produced highly fluctuating costs per part. The averaging approach, on the other hand, allows the company to allocate a charge against the product directly with a fixed, predictable amount. In addition, averaging allows for easier end-product cost analysis without determining the quantities in every PO.

The drawback to the averaging approach is that, unlike direct costing, it masks the costs of a PO and, as a result, it's hard to discourage purchasing agents from placing many small-quantity orders that sometimes are necessary for quick delivery. With direct costing both the number of parts and the PO type determine the cost allocated to the individual component.


As should be obvious, ABC would resolve a number of purchasing problems for a company. In this study it demonstrated the flaw in the percentage-of-product cost system--the lack of a relationship between the component price and purchasing department costs. It also demonstrated that the old labor cost system was based on the ability to bear cost rather than the more meaningful cost-benefit approach.

If applied, ABC would yield other benefits. It would be able to

* Identify the drivers that affected purchasing department costs.

* Realize that several options are present for allocating costs. (However, each option presents different planning and behavioral considerations.

* Choose between one, two and three costs per PO allocation base. The company could decide to allocate costs at the same per-unit rate of $88.74 for all POs ($621,204

* 7,000 PO). If a two-tier cost system was selected, the company could charge a rate of $234.20 per controlled material POs ($46,840.50 + 200 PO) and $84.47 for all other POs ($574,363.50 + 6,800).

A final costing option would involve a three-tier system:

* $171.24 for castings POs ($265,429.50

* 1,550 POs).

* $234.20 for controlled material POs.

* $58.84 for all other POs ($308,934 +

5,250 POs).

One of the obstacles to adopting ABC is that it requires a major overhaul in management's thinking. The company on which this study is based has yet to make the move to ABC, although it has seen the results of the research and it's considering implementing a conversion. Clearly, the flexibility, planning and behavioral considerations that are afforded by ABC make it a significant improvement over the traditional approach.

* MISALLOCATING costs can devastate a business by making its pricing decisions no better than guesswork. While activity-based costing (ABC) has proven effective and been increasingly popular in determining real costs in the manufacturing arena, the authors tested the feasibility of applying ABC to the purchasing function. The test showed the success of ABC as a way to analyze a purchasing department.

* IN ALLOCATING purchasing department costs by using a rate based on direct labor costs, management discovered the data provided were not useful, so it switched to a percentage-of-product cost system. This method was based on the assumption there was a relationship between the price of the component and the purchasing department's costs and that the price level of the component would remain constant.

* WHILE THIS METHOD was better, significant errors remained: It either overallocated or underallocated department costs to the product by 24% to 79%. And purchase orders with expensive material were being charged with a much higher allocation than those with inexpensive material.

* ABC RESOLVED a number of purchasing problems. It identified the drivers that affected purchasing department costs, and it offered management several options for allocating costs. Most important, it provides management with accurate cost data in helping it determine a product's contribution to profit.
COPYRIGHT 1992 American Institute of CPA's
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Costs; activity-based costing
Author:Castellano, Joseph F.
Publication:Journal of Accountancy
Date:Nov 1, 1992
Previous Article:Saturday night special.
Next Article:Reasonable joint cost allocations in nonprofits.

Related Articles
More companies turn to ABC.
Activity-based costing really works for custom molders.
Change 101: back to basics.
Learning to love ABC.
Yes, ABC is for small business, too.
More Complex, More Robust: Activity-based costing systems are harnessing the internet and adding new functionality, giving companies more horsepower...
Spotlight on midlevel ERP software: Accounting products put to the test.
Make the move to dynamic forecasting.
The use of activity based costing in the healthcare industry: 1994 vs. 2004.
Energizing cost accounting: Marine Corps financial managers conduct a thorough analysis of cost accounting processes in order to create accurate,...

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters