Vous êtes sur la page 1sur 8

e ar

at r

ticl
u
e
fe

Variable Activity-Based Costing


and Decision Making Editorial Review

Matthew Geiszler, Kelsey Baker, and Jeffrey Lippitt

INTRODUCTION costs of an activity


Much of managerial decision making regarding are then allocated
This article only to those prod-
illustrates several product pricing and output depends on a careful
ucts that utilize
modifications to separation of the fixed and variable components of
that activity. Many
activity-based cost- costs. Without this separation, the decision maker would argue that the
ing (ABC) and shows will be unable to assess the impact of their deci- resulting product
how these approaches sions on profitability. This distinction is particularly costs are superior
will improve the rel- relevant in allocating overhead. Both traditional to traditional cost-
evance of the result- and activity-based costing (ABC) systems are full ing because the
ing product costs to absorption systems in the sense that they include resulting product
both long-term and both fixed and variable production costs in product costs do reflect the
short-term decision cost. This mixing of fixed and variable overhead resources that went
making. ABC was costs results in product costs that are difficult to into the production
developed to improve of the product more
use in decision making. This article uses a numeric
the allocation of accurately than tra-
example to compare several alternate approaches
indirect production ditional cost systems
costs to products and to overhead allocation. It highlights variable activ-
(Balakrishnan &
to aid in the manage- ity-based costing (VABC), which utilizes regression Sivaramakrishnan,
ment of these costs. analysis to estimate the fixed and variable portions 1996; Homburg, 2005;
Many managers felt of each cost pool. The results show that VABC can Schneeweiss, 1998).
that a single cost produce information that is more useful in deci- In some ways, the
pool allocation of sion making and that the resulting cost allocations ABC approach may
overhead resulted are superior to traditional costing approaches for present some dan-
in product costs decision-making purposes. © 2017 Wiley Periodicals, Inc. gers that were not as
that do not reflect pronounced under
the combination of traditional costing
resources that go into systems. Managers
the production of each prod- Datar, & Rajan, 2012). ABC are well aware that traditional
uct (Cardinaels, Roodhooft, refines the overhead allocation cost systems that mix together
& Warlop, 2004). The result- process by allocating overhead fixed and variable components
ing “peanut butter” costing costs to several cost pools, of overhead costs, are inappro-
overcosted some products and based on the activities involved priate for use in decision mak-
undercosted others (Horngren, in the production process. The ing. Because full absorption

© 2017 Wiley Periodicals, Inc.


Published online in Wiley Online Library (wileyonlinelibrary.com).
DOI 10.1002/jcaf.22277 45
46 The Journal of Corporate Accounting & Finance  /  July/August 2017

costing is required under gener- Both VABC models utilize The assembly cost pool
ally accepted accounting prin- regression analysis to estimate consists of $62,345 of over-
ciples (GAAP), these product the fixed and variable portions head; the machining cost pool
costs are regularly produced of each cost pool. Utilizing is $24,294; and the inspection
for financial reporting pur- the ABC approach, these cost pool is $71,196. Each cost
poses, but managers are aware variable costs are allocated pool is divided by its respec-
that these costs are not useful only to those products that tive cost driver (682 DLH, 244
in decision making (Datar & utilize that activity. MH, and 66 IH) to determine
Gupta, 1994). Decision making When a job utilizes several the application rates: $91.41
depends on a careful separa- different types of materials, per DLH used, $99.57 per
tion of the fixed and variable the amount, cost, and cost per MH used, and $1,078.73 per
portions of product cost. ABC unit of each input is separately IH used. This process results
costing is a significant refine- charged to the job. Under in allocating the average total
ment of traditional cost sys- VABC, there are several differ- cost of each cost pool. Exhibit
tems because it does a better ent kinds of overhead involved 1, Panel B illustrates the appli-
job of having product costs in production, and each of cation of these cost pools to
represent the resources that these types has a different each of the three jobs. After
went into the production of a cost that is individually charged multiplying each rate by the
product; however, these costs to the job. given driver for each job, new
are still full absorption costs The analysis presented is overhead allocation totals
and blend together fixed and based on a hypothetical job are calculated for each job.
variable costs. An ABC costing shop setting that has three jobs. Based on ABC, total overhead
system produces cost informa- The same data are used in each allocated to each job is: Prod-
tion that is also inappropriate model, allowing comparison of uct 1, $57,476.56; Product 2,
for decision making, but the allocated costs between each $27,366.75; and Product 3,
apparent refinement of cost model. $72,991.69.
allocation in ABC may easily The total amount of over-
mislead decision makers into TRADITIONAL COSTING head allocated to the three
using these costs in pricing and products is the same, but the
output decision making (Hom- Traditional analysis applies breakdown between products
burg, 2005). In some situations, total overhead costs ($157,835) varies greatly from that of tra-
ABC may avoid some of the on the basis of direct labor ditional costing. The difference
problems inherent in absorp- hours (DLH; 682 total hours), is representative of the fact that
tion costing. If an activity resulting in an overhead appli- traditional costing, while a sim-
supports a single product, the cation rate of $231.43 per pler approach, takes less care to
fixed costs of that activity must direct labor hour. Exhibit 1, be sure that the allocated cost
be included in some decisions Panel A demonstrates that reflects the activities that went
regarding the profitability of application of overhead to each into the production of each
that product, provided the fixed job, based on DLH. product. ABC does a better job
costs associated with that activ- of this; however, there is a com-
ity are fully avoidable should ACTIVITY-BASED COSTING mon misconception that since
the product be discontinued. ABC provides a better alloca-
This article uses a numeric For ABC, the total over- tion than traditional costing,
example to compare several head costs are separated into it is more useful for decision
alternate approaches to over- three cost pools based on the making. This assumption is
head allocation. It will high- three activities involved in the wrong due to the fact that ABC
light variable activity-based production process: machin- is still a full absorption cost-
costing (VABC) and show how ing, assembly, and inspection. ing approach to product cost-
these systems can produce A cost driver is identified for ing. Both traditional costing
information that is useful in each cost pool: machine hours and ABC mix together fixed
decision making and show that (MH) for machining, labor and variable costs, rendering
the resulting costs are superior hours (DLH) for assembly, the result of little or no value
to traditional variable costing and inspection hours (IH) for in decision making. Decision
for decision-making purposes. inspection. making requires the careful

DOI 10.1002/jcaf © 2017 Wiley Periodicals, Inc.


The Journal of Corporate Accounting & Finance  /  July/August 2017 47

Exhibit 1

Traditional and Activity-Based Costing Manufacturing Overhead Allocations

Panel A: Traditional allocation based on direct labor hours (DLH)


Job DLH Rate per DLH Allocation
Product 1 400 × $231.43 = $92,572.00
Product 2 185 × $231.43 = $42,814.55
Product 3 97 × $231.43 = $22,448.71
Total: $157,835.26

Panel B: Activity-based costing using direct labor hours (DLH), machine hours (MH), and inspection hours (IH)

Product 1 (200 units) Product 3 (1,600 units)


Allocation Rate per unit of Allocation Rate per unit of
Driver base allocation base Allocation Driver base allocation base Allocation
DLH 400 × $91.41 = $36,564.00 DLH 97 × $91.41 = $8,866.77
MH 15 × $99.57 = $1,493.55 MH 189 × $99.57 = $18,818.73
IH 18 × $1,078.73 = $19,417.14 IH 42 × $1,078.73 = $45,306.66
Product 1 Total: $57,474.69 Product 3 Total: $72,992.16

Product 2 (320 units)


Allocation Rate per unit of Total overhead allocated to
Driver base allocation base Allocation jobs Per Unit
DLH 185 × $91.41 = $16,910.85 Product 1 $57,474.69 $287.37
MH 40 × $99.57 = $3,982.80 Product 2 $27,366.03 $85.52
IH 6 × $1,078.73 = $6,472.38 Product 3 $72,992.16 $45.62
Product 2 Total: $27,366.03 Total: $157,832.88

separation of fixed and vari- the direct labor hours used for usefulness in decision making.
able components of production that job. The allocations to each The use of a single cost driver
costs for cost-benefit analysis. job using variable costing are: assumes that all overhead costs
This distinction is also neces- Product 1, $16,592.00; Prod- either vary with direct labor
sary for planning and control- uct 2, $7,673.8; and Product 3, hours, or are fixed. One of the
ling costs. $4,023.56 (Exhibit 2, Panel B). main contributions of ABC is
Under variable costing, a that overhead costs vary based
VARIABLE COSTING total of $28,289.36 is allocated upon the levels of several
to the three jobs, significantly different cost drivers. Many
Next, we consider a tradi- less than $157,832.88 under short-term decisions could
tional variable costing approach absorption costing. This is potentially affect the levels of
to the allocation of overhead. because $126,946.80 of fixed each of these cost drivers. To
As seen in Exhibit 2, Panel A, overhead cost is not allocated make sound decisions, manag-
a regression based on DLH is to the respective jobs.1 ers must understand how their
run to determine fixed costs While this variable costing decisions will affect each of
(y-intercept), variable cost approach does separate over- these cost drivers and how the
(DLH coefficient), and residual head costs into fixed and vari- cost drivers will in turn affect
error. The coefficient of $41.48 able portions based on direct overhead costs. Each activity
per DLH is then used to apply labor hours, it may also pro- within the manufacturing pro-
overhead to each job based on duce results that are of limited cess will have its own variable

© 2017 Wiley Periodicals, Inc. DOI 10.1002/jcaf


48 The Journal of Corporate Accounting & Finance  /  July/August 2017

Exhibit 2

Variable Traditional Manufacturing Overhead Allocation

Panel A: Regression analysis based on direct labor hours (DLH)

Total MOH = α + β 1 Direct Labor Hours + ε


Variable Coefficient
Intercept (α) 126,954.80 ***
DLH (β1 ) 41.48 ***
Adj. R2 0.4987
n 18

Panel B: Variable traditional allocation using direct labor hours (DLH)


Job DLH β1 Allocation
Product 1 400 × $41.48 = $16,592.00
Product 2 185 × $41.48 = $7,673.80
Product 3 97 × $41.48 = $4,023.56
Total: $28,289.36

*,**,*** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Total MOH = the total
manufacturing overhead (both fixed and variable), α = intercept (fixed portion of MOH), Direct Labor Hours = the
plant-wide overhead allocation base, ε = the residual (error) term.

cost. The total variable over- running a regression for each inspection hours and inspection
head cost of a product will, of the cost drivers used in costs. Third, it could indicate
therefore, be determined by ABC. Each regression calcu- that the inspection activity is
multiplying the amount of lates a fixed cost and a variable primarily a fixed cost. When
each activity used to manu- cost, shown in Exhibit 3, implementing this methodol-
facture that product, by the Panel A. The regression analy- ogy, it is important to consider
variable cost of each activity sis shows that DLH and MH these issues before moving
and summing these costs. The have a significant impact on forward with the overhead
resulting variable cost for that their respective cost pools. allocations.
product will reflect the likely Inspection hours, however, have Overhead is applied to each
impact on total overhead of a variable cost that is not sta- job by multiplying the level of
changing the production level tistically significantly different each of the cost drivers used in
of that product. from zero at a 90% confidence that job by its respective vari-
Our next analysis builds on level.2 The lack of a relation- able cost. DLH is applied at
the ABC approach of separating ship between inspection hours $39.49 per direct labor hour,
overhead costs into the cost pools and the inspection cost pool MH at $60.23 per machine
relevant to each activity. Then, could indicate one of several hour, and inspection costs at
using the cost driver for each pool, problems. First, it could mean $10.34 per inspection hour.
the cost pool is separated into that we have chosen the wrong The total costs of each of the
fixed and variable components. cost driver. Second, it could jobs are: Product 1, $16,885.57;
mean that the inspection costs Product 2, $9,776.89; and Prod-
VARIABLE ABC are not well controlled and the uct 3, $15,648.28 (Exhibit 3,
variations in the inspection cost Panel B). Compared to the
In essence, variable cost- due to other causes is mask- variable costing total allocation
ing and ABC are combined by ing the relationship between of $28,289 of variable overhead

DOI 10.1002/jcaf © 2017 Wiley Periodicals, Inc.


The Journal of Corporate Accounting & Finance  /  July/August 2017 49

Exhibit 3

Variable Activity-Based Costing Manufacturing Overhead Allocations—Panel A


Panel A: Regression analysis based on direct labor hours (DLH), machine hours (MH), and inspection hours (IH)

(1) Total Direct Labor Cost Pool = α + β 1 Direct Labor Hours + ε


(2) Total Machine Hours Cost Pool = α + ű 1 Machine Hours + ε
(3) Total Inspection Hours Cost Pool = α + θ 1 Inspection Hours + ε

Coefficients
Variable DLH MH IH
Intercept (α) 33,761.00 *** 11,192.00 *** 70,234.00 ***
DLH (β1) 39.49 ***
MH (ű1) 60.23 ***
IH (θ1) 10.34
Adj. R 2 0.8609 0.5592 0.0709
n 18 18 18

*,**,*** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Total direct labor cost pool = the portion of the total
manufacturing overhead (both fixed and variable) attributable to direct labor hours, Total machine hours cost pool = the portion of the
total manufacturing overhead (both fixed and variable) attributable to machine hours, Total inspection hours cost pool = the portion of the
total manufacturing overhead (both fixed and variable) attributable to inspection hours, α = intercept (fixed portion of MOH for each
activity), Direct Labor Hours = the direct labor activity allocation base, Machine Hours = the machine hour activity allocation base,
Inspection Hours = the inspection hour activity allocation base, ε = the residual (error) term.

Variable Activity-Based Costing Manufacturing Overhead Allocations—Panel B

Panel B: Variable Activity-Based Costing using direct labor hours (DLH), machine hours (MH), and inspection hours (IH)
Product 1 (200 units) Product 3 (1,600 units)
Rate per unit Rate per unit
Allocation of allocation Allocation of allocation
Driver base base Allocation Driver base base Allocation
DLH (β1) 400 ˟ $39.49 = $15,796.00 DLH (β1) 97 ˟ $39.49 = $3,830.53
MH (ű1) 15 ˟ $60.23 = $903.45 MH (ű1) 189 ˟ $60.23 = $11,383.47
IH (θ1) 18 ˟ $10.34 = $186.12 IH (θ1) 42 ˟ $10.34 = $434.28
Product 1 Total: $16,885.57 Product 3 Total: $15,648.28

Product 2 (320 units)


Rate per unit
Allocation of allocation Total overhead allocated to
Driver base base Allocation jobs Per Unit
DLH (β1) 185 ˟ $39.49 = $7,305.65 Product 1 $16,885.57 $84.43
MH (ű1) 40 ˟ $60.23 = $2,409.20 Product 2 $9,776.89 $30.55
IH (θ1) 6 ˟ $10.34 = $62.04 Product 3 $15,648.28 $9.78
Product 2 Total: $9,776.89 Total: $42,310.74

cost, variable ABC allocates overhead cost that would result ABC is assigning the total
$42,311 of variable overhead from changing the amount of overhead cost to the vari-
cost. As a result, the total the product that is produced. ous cost pools. Errors in this
allocated cost to each product One of the more challeng- assignment will result in errors
reflects the likely change in ing aspects of implementing in the coefficients of the cost

© 2017 Wiley Periodicals, Inc. DOI 10.1002/jcaf


50 The Journal of Corporate Accounting & Finance  /  July/August 2017

drivers for that cost pool and, hours, on the other hand, are more than variable costing.
therefore, a misstatement of the insignificant and negative. As Comparing these results to the
variable costs of the products. a negative coefficient, inspec- results of the traditional alloca-
Fortunately, the variable ABC tion hours appear to be a cost tions in Exhibit 1, demonstrates
approach does not require this absorber as opposed to a cost how absorption costing can
assignment of total cost to driver.3 Combining both the lead to overpricing of products
individual cost pools. In most lack of statistical significance due to over costing. In addi-
cases, the analysis can be com- and the counter factual sugges- tion, comparing Exhibit 2 to
pleted by analysis of the total tion of a negative variable cost, Exhibit 3 and Exhibit 4 demon-
overhead cost pool. the inspection hour cost driver strates how traditional variable
is eliminated from the multiple costing, although it separates
MULTIPLE REGRESSION regression and recalculated as costs into fixed and variable
shown in Exhibit 4, Panel B. components, is of only limited
The variable ABC The reduced multiple usefulness in decision making.
approach can be implemented regression results in a coef-
by including all the cost driv- ficient of $37.84 per DLH and CONCLUSION
ers in one multiple regression $69.41 per MH, respectively.
analysis, eliminating the need The allocation of overhead As noted earlier, traditional
to separate cost pools. The mul- to each job using reduced costing systems and ABC sys-
tiple regression will still provide multiple regression results in: tems suffer from the problem
a coefficient (variable cost) for Product 1, $16,177.15; Product of being “full absorption” sys-
each driver, but one fixed cost. 2, $9,776.80; and Product 3, tems, limiting their usefulness
Exhibit 4, Panel A shows this $16,788.97. Altogether a total in decision making. Managers
calculation. In our example, of $42,742.92 is allocated using and decision makers should be
DLH and MH are significant multiple regression overhead acutely aware of this limita-
with coefficients of $37.78 and allocations (Exhibit 4, Panel C). tion. This study highlights this
$69.80, respectively. Inspection Once again, this amount is deficiency and demonstrates

Exhibit 4

Multiple Regression Overhead Allocation—Panel A

Panel A: Multiple regression analysis based on direct labor hours (DLH), machine hours (MH),
and inspection hours (IH)

Total MOH = + 1 Direct Labor Hours + 2 Machine Hours + 3 Inspection Hours +


Variable Coefficient
Intercept ( ) 115,086.88 ***
DLH ( 1) 37.78 ***
MH ( 2) 69.80 ***
IH ( 3) –5.38

Adj. R 2 0.7616
n 18

*,**,*** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Total MOH = the
total manufacturing overhead (both fixed and variable), = intercept (fixed portion of MOH), Direct
Labor Hours = the direct labor activity allocation base, Machine Hours = the machine hour activity
allocation base, Inspection Hours = the inspection hour activity allocation base, = the residual (error)
term.

DOI 10.1002/jcaf © 2017 Wiley Periodicals, Inc.


The Journal of Corporate Accounting & Finance  /  July/August 2017 51

Exhibit 4

Multiple Regression Overhead Allocation—Panel B

Panel B: Reduced multiple regression analysis based on direct labor hours (DLH) and machine
hours (MH)

Total MOH = + 1 Direct Labor Hours + 2 Machine Hours +


Variable Coefficient
Intercept ( ) 114,862.00 ***
DLH ( 1) 37.84 ***
MH ( 2) 69.41 ***

Adj. R 2 0.7763
n 18

*,**,*** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Total MOH = the
total manufacturing overhead (both fixed and variable), = intercept (fixed portion of MOH), Direct
Labor Hours = the direct labor activity allocation base, Machine Hours = the machine hour activity
allocation base, = the residual (error) term.

Multiple Regression Overhead Allocation—Panel C

Panel C: Reduced multiple regression allocations


Product 1 (200 units) Product 3 (1,600 units)
Rate per unit Rate per unit
Allocation of allocation Allocation of allocation
Driver base base Allocation Driver base base Allocation
DLH ( 1) 400 × $37.84 = $15,136.00 DLH ( 1) 97 × $37.84 = $3,670.48
MH ( 2) 15 × $69.41 = $1,041.15 MH ( 2) 189 × $69.41 = $13,118.49
Product 1 Total: $16,177.15 Product 3 Total: $16,788.97

Product 2 (320 units)


Rate per unit
Allocation of allocation Total overhead allocated to
Driver base base Allocation jobs Per Unit
DLH ( 1) 185 × $37.84 = $7,000.40 Product 1 $16,177.15 $80.89
MH ( 2) 40 × $69.41 = $2,776.40 Product 2 $9,776.80 $30.55
Product 2 Total: $9,776.80 Product 3 $16,788.97 $10.49
Total: $42,742.92

that VABC can produce LIMITATIONS limitations. Cost drivers must


information that is useful in be accurately selected by
decision making and that the Although multiple regres- management and costs must
resulting cost allocations are sion provides a better way to be controlled. While multiple
superior to traditional variable allocate overhead costs per regression does not require the
costing for decision-making activity driver to each job, separation of total overhead
purposes. there remain fundamental cost into cost pools, this is

© 2017 Wiley Periodicals, Inc. DOI 10.1002/jcaf


52 The Journal of Corporate Accounting & Finance  /  July/August 2017

often an essential step in the result of the residual error calculated Cardinaels, E., Roodhooft, F., & Warlop,
control of these costs. The rec- in the regression. Similar to absorp- L. (2004). The value of activity-based
tion costing, variable costing can lead costing in competitive pricing decisions.
ognition of the variable nature to misvalued, specifically undervalued, Journal of Management Accounting
of the cost pools may also help unit cost pricing. Research, 16(1), 133–148.
in planning and controlling 2. The coefficient is significant at the 85% Datar, S., & Gupta, M. (1994). Aggrega-
these costs. confidence level. This highlights an tion, specification and measurement
Regression analysis is a important judgment call that must be errors in product costing. Accounting
made with respect to what is an accept- Review, 69(4), 567–591.
mathematical tool that lacks able cutoff threshold for use in the Homburg, C. (2005). Using relative profits
any respect for economic real- overhead allocations. as an alternative to activity-based cost-
ity. Management must carefully 3. It is important to note that due to ing. International Journal of Production
assess the results of this analy- the lack of statistical significance of Economics, 95(3), 387–397.
sis to be certain that the results the coefficient, interpretation of the Horngren, C. T., Datar, S. M., &
direction of the sign on the coefficient Rajan, M. V. (2012). Cost accounting:
are not driven by anomalies becomes inconclusive. A managerial emphasis (14th ed.).
that do not reflect the true Upper Saddle River, NJ: Pearson/
underlying relationships. Prentice Hall.
REFERENCES Schneeweiss, C. (1998). On the applica-
bility of activity based costing as a
NOTES Balakrishnan, R., & Sivaramakrishnan, S. planning instrument. International
(1996). Is assigning capacity costs Journal of Production Economics, 54(3),
1. Together, only $155,263.16 of the total to individual products necessary for 277–284.
overhead ($157,835) was allocated capacity planning? Accounting Horizons,
during the year. The difference is a 10(3), 1–11.

Matthew Geiszler, PhD, is an Assistant Professor of Accounting at Ithaca College. Kelsey Baker is an
Assurance Associate with PricewaterhouseCoopers. She received both her BS in accounting and MBA in
professional accountancy from Ithaca College. Jeffrey Lippitt, PhD, is an Associate Professor of Account-
ing at Ithaca College.

DOI 10.1002/jcaf © 2017 Wiley Periodicals, Inc.

Vous aimerez peut-être aussi