3 27

CHAPTER 10
DETERMINING HOW COSTS BEHAVE
10-1 The two assumptions are:

1. Variations in total costs are explained by variations in the level of a single activity related
to those costs (the cost driver).
2. Cost behavior is approximated by a linear cost function within the relevant range. A linear
cost function is a cost function where, within the relevant range, the graph of total costs
versus the level of a single activity forms a straight line.
10-2 Three alternative linear cost functions are:

1. Variable cost functiona cost function in which total costs change in proportion to the
changes in the level of activity in the relevant range.
2. Fixed cost functiona cost function in which total costs do not change with changes in the
level of activity in the relevant range.
3. Mixed cost functiona cost function that has both variable and fixed elements. Total costs
change but not in proportion to the changes in the level of activity in the relevant range.
10-3 A linear cost function is a cost function where, within the relevant range, the graph of
total costs versus the level of a single activity related to that cost is a straight line. An example
of a linear cost function is a cost function for use of a telephone line where the terms are a fixed
charge of $10,000 per year plus a $2 per minute charge for phone use. A nonlinear cost function
is a cost function where, within the relevant range, the graph of total costs versus the level of a
single activity related to that cost is not a straight line. Examples include economies of scale in
advertising where an agency can double the number of advertisements for less than twice the
costs, step-function costs, and learning-curve-based costs.
10-4 No. High correlation merely indicates that the two variables move together in the data
examined. It is essential to also consider economic plausibility before making inferences about
cause and effect. Without any economic plausibility for a relationship, it is less likely that a high
level of correlation observed in one set of data will be similarly found in other sets of data.
10-5 Four approaches to estimating a cost function are:

1. Industrial engineering method.
2. Conference method.
3. Account analysis method.
4. Quantitative analysis of current or past cost relationships.
10-6 The conference method estimates cost functions on the basis of analysis and opinions
about costs and their drivers gathered from various departments of a company (purchasing,
process engineering, manufacturing, employee relations, etc.). Advantages of the conference
method include:
1. The speed with which cost estimates can be developed.
2. The pooling of knowledge from experts across functional areas.
3. The improved credibility of the cost function to all personnel.
10-1
10-7 The account analysis method estimates cost functions by classifying cost accounts in
the subsidiary ledger as variable, fixed, or mixed with respect to the identified level of activity.
Typically, managers use qualitative, rather than quantitative, analysis when making these cost-
classification decisions.
10-8 The six steps are:

1. Choose the dependent variable (the variable to be predicted, which is some type of cost).
2. Identify the independent variable or cost driver.
3. Collect data on the dependent variable and the cost driver.
4. Plot the data.
5. Estimate the cost function.
6. Evaluate the cost driver of the estimated cost function.
Step 3 typically is the most difficult for a cost analyst.
10-9 Causality in a cost function runs from the cost driver to the dependent variable. Thus,
choosing the highest observation and the lowest observation of the cost driver is appropriate in
the high-low method.
10-10 Criteria important when choosing among alternative cost functions are:
1. Economic plausibility.
2. Goodness of fit.
3. Slope of the regression line.
10-11 A learning curve is a function that measures how labor-hours per unit decline as units of
production increase because workers are learning and becoming better at their jobs. Two models
used to capture different forms of learning are:
1. Cumulative average-time learning model. The cumulative average time per unit declines
by a constant percentage each time the cumulative quantity of units produced doubles.
2. Incremental unit-time learning model. The incremental time needed to produce the last unit
declines by a constant percentage each time the cumulative quantity of units produced
doubles.
10-12 Frequently encountered problems when collecting cost data on variables included in a
cost function are:
1. The time period used to measure the dependent variable is not properly matched with the
time period used to measure the cost driver(s).
2. Fixed costs are allocated as if they are variable.
3. Data are either not available for all observations or are not uniformly reliable.
4. Extreme values of observations occur.
5. A homogeneous relationship between the individual cost items in the dependent variable
cost pool and the cost driver(s) does not exist.
6. The relationship between the cost and the cost driver is not stationary.
7. Inflation has occurred in a dependent variable, a cost driver, or both.
10-2
10-13 Four key assumptions examined in specification analysis are:
1. Linearity between the dependent variable and the independent variable within the relevant
range.
2. Constant variance of residuals for all values of the independent variable.
3. Residuals are independent of each other.
4. Residuals are normally distributed.
10-14 No. A cost driver is any factor whose change causes a change in the total cost of a
related cost object. A cause-and-effect relationship underlies selection of a cost driver. Some
users of regression analysis include numerous independent variables in a regression model in an
attempt to maximize goodness of fit, irrespective of the economic plausibility of the independent
variables included. Some of the independent variables included may not be cost drivers.
10-15 No. Multicollinearity exists when two or more independent variables are highly
correlated with each other.
10-16 (10 min.) Estimating a cost function

1. Slope coefficient =
= = $0.30 per machine-hour

Constant = Total cost (Slope coefficient Quantity of cost driver)
= $3,900 ($0.30 7,000) = $1,800
= $3,000 ($0.30 4,000) = $1,800
The cost function based on the two observations is:

Maintenance costs = $1,800 + $0.30 Machine-hours
2. The cost function in requirement 1 is an estimate of how costs behave within the relevant
range, not at cost levels outside the relevant range. If there are no months with zero machine-
hours represented in the maintenance account, data in that account cannot be used to estimate the
fixed costs at the zero machine-hours level. Rather, the constant component of the cost function
provides the best available starting point for a straight line that approximates how a cost behaves
within the relevant range.
10-3
10-17 (15 min.) Identifying variable-, fixed-, and mixed-cost functions.
1. See Solution Exhibit 10-17.

2. Contract 1: y = $50
Contract 2: y = $30 + $0.20X
Contract 3: y = $1X
where X is the number of miles traveled in the day.
3. Contract Cost Function
1 Fixed
2 Mixed
3 Variable
SOLUTION EXHIBIT 10-17
Plots of Car Rental Contracts Offered by Pacific Corp.
10-4
10-18 (20 min.) Various cost-behavior patterns.
1. K
2. B
3. G
4. J Note that A is incorrect because, although the cost per pound eventually equals a
constant at $9.20, the total dollars of cost increases linearly from that point
onward.
5. I The total costs will be the same regardless of the volume level.
6. L
7. F This is a classic step-cost function.
8. K
9. C
10-19 (30 min.) Matching graphs with descriptions of cost and revenue
behavior.
a. (1)
b. (6) A step-cost function.
c. (9)
d. (2)
e. (8)
f. (10) It is data plotted on a scatter diagram, showing a linear variable cost function with
constant variance of residuals. The constant variance of residuals implies that there is a
uniform dispersion of the data points about the regression line.
g. (3)
h. (8)
10-5
10-20 (15 min.) Account analysis method.
1. Variable costs:
Car wash labor $240,000
Soap, cloth, and supplies 32,000
Water 28,000
Electric power to move conveyor belt 72,000
Total variable costs $372,000
Fixed costs:
Depreciation $ 64,000
Salaries 46,000
Total fixed costs $110,000
Costs are classified as variable because the total costs in these categories change in proportion to
the number of cars washed in Lorenzos operation. Costs are classified as fixed because the total
costs in these categories do not vary with the number of cars washed. If the conveyor belt moves
regardless of the number of cars on it, the electricity costs to power the conveyor belt would be a
fixed cost.
$372,000
2. Variable costs per car = 80,000 = $4.65 per car
Total costs estimated for 90,000 cars = $110,000 + ($4.65 90,000) = $528,500
10-21 (30 min.) Account analysis method.
1. Manufacturing cost classification for 2004:
% of Total Unit
Total Costs That Is Variable Fixed Variable
Costs Variable Costs Costs Costs
Account (1) (2) (3) = (1) (2) (4) = (1) (3) (5)=(3) 75,000
Direct materials $300,000 100% $300,000 $ 0 $4.00

Direct manufacturing labor 225,000 100 225,000 0 3.00
Power 37,500 100 37,500 0 0.50
Supervision labor 56,250 20 11,250 45,000 0.15
Materials-handling labor 60,000 50 30,000 30,000 0.40
Maintenance labor 75,000 40 30,000 45,000 0.40
Depreciation 95,000 0 0 95,000 0
Rent, property taxes, admin 100,000 0 0 100,000 0
Total $948,750 $633,750 $315,000 $8.45
Total manufacturing cost for 2004 = $948,750
10-6
10-21 (Contd.)
Variable costs in 2005:
Unit Increase in
Variable Variable
Cost for Percentage Costs Unit Variable Total Variable
2004 Increase per Unit Cost for 2005 Costs for 2005
Account (6) (7) (8) = (6) (7) (9) = (6) + (8) (10) = (9) 80,000
Direct materials $4.00 5% $0.20 $4.20 $336,000

Direct manufacturing labor 3.00 10 0.30 3.30 264,000
Power 0.50 0 0 0.50 40,000
Supervision labor 0.15 0 0 0.15 12,000
Materials-handling labor 0.40 0 0 0.40 32,000
Maintenance labor 0.40 0 0 0.40 32,000
Depreciation 0 0 0 0 0
Rent, property taxes, admin. 0 0 0 0 0
Total $8.45 $0.50 $8.95 $716,000
Fixed and total costs in 2005:
Dollar
Fixed Increase in Fixed Costs Variable Total
Costs Percentage Fixed Costs for 2005 Costs for Costs
for 2004 Increase (13) = (14) = 2005 (16) =
Account (11) (12) (11) (12) (11) + (13) (15) (14) + (15)
Direct materials $ 0 0% $ 0 $ 0 $336,000 $ 336,000

Direct manufacturing labor 0 0 0 0 264,000 264,000
Power 0 0 0 0 40,000 40,000
Supervision labor 45,000 0 0 45,000 12,000 57,000
Materials-handling labor 30,000 0 0 30,000 32,000 62,000
Maintenance labor 45,000 0 0 45,000 32,000 77,000
Depreciation 95,000 5 4,750 99,750 0 99,750
Rent, property taxes, admin. 100,000 7 7,000 107,000 0 107,000
Total $315,000 $11,750 $326,750 $716,000 $1,042,750
Total manufacturing costs for 2005 = $1,042,750
$948,750
2. Total cost per unit, 2004 = 75,000
= $12.65
$1,042,750
Total cost per unit, 2005 = 80,000
= $13.03
3. Cost classification into variable and fixed costs is based on qualitative, rather than
quantitative, analysis. How good the classifications are depends on the knowledge of individual
managers who classify the costs. Gower may want to undertake quantitative analysis of costs,
using regression analysis on time-series or cross-sectional data to better estimate the fixed and
variable components of costs. Better knowledge of fixed and variable costs will help Gower to
better price his products, know when he is getting a positive contribution margin, and to better
manage costs.
10-7
10-22 (20 min.) Estimating a cost function, high-low method.
1. See Solution Exhibit 10-22. There is a positive relationship between the number of service
reports (a cost driver) and the customer-service department costs. This relationship is
economically plausible.
2. Number of Customer-Service
Service Reports Department Costs
Highest observation of cost driver 436 $21,890
Lowest observation of cost driver 122 12,941
Difference 314 $ 8,949
Customer-service department costs = a + b (number of service reports)
$8,949
Slope coefficient (b) = = $28.50 per service report
314
Constant (a) = $21,890 $28.50 (436) = $9,464
= $12,941 $28.50 (122) = $9,464
Customer-service
department costs = $9,464 + $28.50 (number of service reports)
3. Other possible cost drivers of customer-service department costs are:

a. Number of products replaced with a new product (and the dollar value of the new
products charged to the customer-service department).
b. Number of products repaired and the time and cost of repairs.

Plot of Number of Service Reports versus Customer-Service Dept. Costs for Capitol Products
Customer-Service Department Costs
$25,000
20,000
15,000
10,000
5,000
$0
0 100 200 300 400 500
Number of Service Reports
10-8
10-23 (3040 min.) Linear cost approximation.
1. Slope coefficient (b) =
$529,000 - 400,000
= 7,000- 4,000
= $43.00
Constant (a) = $529,000 ($43.00 7,000)

= $228,000
Cost function = $228,000 + $43.00 (professional labor-hours)
The linear cost function is plotted in Solution Exhibit 10-23.
No, the constant component of the cost function does not represent the fixed overhead cost
of the Memphis Group. The relevant range of professional labor-hours is from 3,000 to 8,000.
The constant component provides the best available starting point for a straight line that
approximates how a cost behaves within the 3,000 to 8,000 relevant range.
2. A comparison at various levels of professional labor-hours follows. The linear cost

function is based on the formula of $228,000 per month plus $43.00 per professional labor-hour.
Total overhead cost behavior:
Month 1 Month 2 Month 3 Month 4 Month 5 Month 6
Actual total overhead costs $340,000 $400,000 $435,000 $477,000 $529,000 $587,000
Linear approximation 357,000 400,000 443,000 486,000 529,000 572,000
Actual minus linear
approximation $(17,000) $ 0 $ (8,000) $ (9,000) $ 0 $ 15,000
Professional labor-hours 3,000 4,000 5,000 6,000 7,000 8,000
The data are shown in Solution Exhibit 10-23. The linear cost function overstates costs by
$8,000 at the 5,000-hour level and understates costs by $15,000 at the 8,000-hour level.
3. Based on Based on Linear

Actual Cost Function
Contribution before deducting incremental overhead $38,000 $38,000
Incremental overhead 35,000 43,000
Contribution after incremental overhead $ 3,000 $ (5,000)
The total contribution margin actually forgone is $3,000.
10-9
10-23 (Contd.)

Linear Cost Function Plot of Professional Labor-Hours
on Total Overhead Costs for Memphis Consulting Group
$700,000
600,000
Total Overhead Costs
500,000
400,000
300,000
200,000
100,000
0
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000
Professional Labor-Hours Billed
10-24 (20 min.) Cost-volume-profit and regression analysis.

Total manufacturing costs
1a. Average cost of manufacturing = Number of bicycle frames
$900,000
= 30,000
= $30 per frame
This cost is greater than the $28.50 per frame that Ryan has quoted.
1b. Garvin cannot take the average manufacturing cost in 2002 of $30 per frame and multiply it
by 36,000 bicycle frames to determine the total cost of manufacturing 36,000 bicycle
frames. The reason is that some of the $900,000 (or equivalently the $30 cost per frame)
are fixed costs and some are variable costs. Without distinguishing fixed from variable
costs, Garvin cannot determine the cost of manufacturing 36,000 frames. For example, if
all costs are fixed, the manufacturing costs of 36,000 frames will continue to be $900,000.
If, however, all costs are variable, the cost of manufacturing 36,000 frames would be $30
36,000 = $1,080,000. If some costs are fixed and some are variable, the cost of
manufacturing 36,000 frames will be somewhere between $900,000 and $1,080,000.
10-10
10-24 (Contd.)
Some students could argue that another reason for not being able to determine the cost of
manufacturing 36,000 bicycle frames is that not all costs are output unit-level costs. If some
costs are, for example, batch-level costs, more information would be needed on the number of
batches in which the 36,000 bicycle frames would be produced, in order to determine the cost of
manufacturing 36,000 bicycle frames.
Expected cost to make

2. 36,000 bicycle frames = $432,000 + $15 36,000
= $432,000 + $540,000 = $972,000
Purchasing bicycle frames from Ryan will cost $28.50 36,000 = $1,026,000. Hence it
will cost Garvin $1,026,000 $972,000 = $54,000 more to purchase the frames from Garvin
rather than manufacture them in-house.
3. Garvin would need to consider several factors before being confident that the equation in
requirement 2 accurately predicts the cost of manufacturing bicycle frames.
a. Is the relationship between total manufacturing costs and quantity of bicycle frames
economically plausible? For example, is the quantity of bicycles made the only cost
driver or are there other cost-drivers (for example batch-level costs of setups,
production-orders or material handling) that affect manufacturing costs?
b. How good is the goodness of fit? That is, how well does the estimated line fit the
data?
c. Is the relationship between the number of bicycle frames produced and total
manufacturing costs linear?
d. Does the slope of the regression line indicate that a strong relationship exists between
manufacturing costs and the number of bicycle frames produced?
e. Are there any data problems such as, for example, errors in measuring costs, trends in
prices of materials, labor or overheads that might affect variable or fixed costs over
time, extreme values of observations, or a nonstationary relationship over time
between total manufacturing costs and the quantity of bicycles produced?
f. How is inflation expected to affect costs?
g. Will Ryan supply high-quality bicycle frames on time?
10-11
10-25 (25 min.) Regression analysis, service company.
1. Solution Exhibit 10-25 plots the relationship between labor-hours and overhead costs and
shows the regression line.
y = $48,271 + $3.93 X
Economic plausibility. Labor-hours appears to be an economically plausible driver of

overhead costs for a catering company. Overhead costs such as scheduling, hiring and training
of workers, and managing the workforce are largely incurred to support labor.
Goodness of fit. The vertical differences between actual and predicted costs are extremely
small, indicating a very good fit. The good fit indicates a strong relationship between the labor-
hour cost driver and overhead costs.
Slope of regression line. The regression line has a reasonably steep slope from left to right.
Given the small scatter of the observations around the line, the positive slope indicates that, on
average, overhead costs increase as labor-hours increase.
2. The regression analysis indicates that, within the relevant range of 2,500 to 7,500 labor-
hours, the variable cost per person for a cocktail party equals:
Food and beverages $15.00

Labor (0.5 hrs. $10 per hour) 5.00
Variable overhead (0.5 hrs $3.93 per labor-hour) 1.97
Total variable cost per person $21.97
3. To earn a positive contribution margin, the minimum bid for a 200-person cocktail party
would be any amount greater than $4,394. This amount is calculated by multiplying the variable
cost per person of $21.97 by the 200 people. At a price above the variable costs of $4,394, Bob
Jones will be earning a contribution margin toward coverage of his fixed costs.
Of course, Bob Jones will consider other factors in developing his bid including (a) an
analysis of the competitionvigorous competition will limit Jones's ability to obtain a higher
price (b) a determination of whether or not his bid will set a precedent for lower pricesoverall,
the prices Bob Jones charges should generate enough contribution to cover fixed costs and earn a
reasonable profit, and (c) a judgment of how representative past historical data (used in the
regression analysis) is about future costs.
10-12
10-25 (Contd.)

Regression Line of Labor-Hours on Overhead Costs for Bob Jones's Catering Company
$90,000
80,000
70,000
O v erh ea d C o sts
60,000
50,000
40,000
30,000
20,000
10,000
0
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
Cost Driver: Labor-Hours
10-25 Excel Application

Determining How Costs Behave
Bob Jones Catering Company
Original Data
Month Labor-Hours Overhead

Costs
January 2,500 $55,000
February 2,700 59,000
March 3,000 60,000
April 4,200 64,000
May 7,500 77,000
June 5,500 71,000
July 6,500 74,000
August 4,500 67,000
September 7,000 75,000
October 4,500 68,000
November 3,100 62,000
December 6,500 73,000
Total 57,500 $805,000
10-13
10-25 (Contd.)
Relationship Between Overhead Costs and Labor-

Hours
$90,000
$80,000
$70,000
Overhead Costs
$60,000
$50,000
$40,000
$30,000
$20,000
$10,000
$0
0 1000 2000 3000 4000 5000 6000 7000 8000
Labor-Hours
10-14
10-25 (Contd.)
Regression Output
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.980995276
R Square 0.962351731
Adjusted R Square 0.958586904
Standard Error 1447.996847
Observations 12
ANOVA
df SS MS F Significance
F
Regression 1 535949718 53594971 255.616459 1.89126E-08
Residual 10 20966948.69 2096694.8
Total 11 556916666.7
Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept 48270.61812 1248.716188 38.65619 3.20356E-12 45488.30459 51052.93166 45488.30459 51052.93166
X Variable 1 3.92613187 0.245567266 15.98800 1.89126E-08 3.378973809 4.473289931 3.378973809 4.473289931
10-15
10-26 (30 min.) Regression analysis, activity-based costing, choosing cost
drivers.
1a. Solution Exhibit 10-26A presents the plots and regression line of number of packaged units
moved on distribution costs.
SOLUTION EXHIBIT 10-26A

Plots and Regression Line of Number of Packaged Units Moved on Distribution Costs
$45,000
40,000
35,000
D istribution C osts
30,000
25,000
20,000
15,000
10,000
5,000
0
0 20,000 40,000 60,000 80,000
Number of Packaged Units Moved
1b. Solution Exhibit 10-26B presents the plots and regression line of number of shipments
made on distribution costs.
SOLUTION EXHIBIT 10-26B

Plots and Regression Line of Number of Shipments Made on Distribution Costs
$45,000
40,000
35,000
D istribution C osts
30,000
25,000
20,000
15,000
10,000
5,000
0
0 100 200 300 400 500
Number of Shipments Made
10-16
10-26 (Contd.)
Number of packaged units moved appears to be a better cost driver of distribution costs for the
following reasons:
(i) Economic plausibility. Both number of packaged units moved and number of shipments
are economically plausible cost drivers. Because the product is heavy, however, costs of
freight are likely to be a sizable component of distribution costs. Thus, number of
packaged units moved will affect distribution costs significantly because freight costs are
largely a function of the number of units transported.
(ii) Goodness of fit. Compare Solution Exhibits 10-26A and 10-26B. Number of packaged
units moved has a better goodness of fit with distribution costs than do number of
shipments made. That is, the vertical differences between actual and predicted distribution
costs are smaller for the number of packaged units moved regression than for the number of
shipments made regression.
(iii) Slope of regression line. Again, compare Solution Exhibits 10-26A and 10-26B. The
number of packaged units moved regression line has a relatively steep slope and a small
scatter of observations around the regression lines indicating a strong relationship between
number of packaged units moved and distribution costs. On average, distribution costs
increase with the number of packaged units moved. The number of shipments made
regression line is flatter and has more scatter of observations about the line indicating a
weak relationship between the number of shipments made and distribution costs. On
average, the number of shipments made has a smaller effect on distribution costs.
2. Using the preferred cost function,

Distribution costs = $1,349 + ($0.496 Number of packaged units moved),
Flaherty would budget distribution costs of
$1,349 + ($0.496 40,000) = $1,349 + $19,840 = $21,189
10-27 (20 min.) Learning curve, cumulative average-time learning model.

The direct manufacturing labor-hours (DMLH) required to produce the first 2, 4, and 8
units given the assumption of a cumulative average-time learning curve of 90%, is as follows:
Cumulative
Cumulative Average-Time Cumulative
Number of Units per Unit Total Time
(1) (2) (3) = (1) (2)
1 3,000 3,000
2 2,700 (3,000 0.90) 5,400
4 2,430 (2,700 0.90) 9,720
8 2,187 (2,430 0.90) 17,496
Alternatively, to compute the values in column (2) we could use the formula
10-17
10-27 (Contd.)
y = aXb
where a = 3,000, X = 2, 4, or 8, and b = 0.1520, which gives

when X = 2, y = 3,000 2 0.1520 = 2,700
when X = 4, y = 3,000 4 0.1520 = 2,430
when X = 8, y = 3,000 8 0.1520 = 2,187
Variable Costs of Producing

2 Units 4 Units 8 Units
Direct materials $80,000 2; 4; 8 $160,000 $320,000 $ 640,000
Direct manufacturing labor
$25 5,400; 9,720; 17,496 135,000 243,000 437,400
Variable manufacturing overhead
$15 5,400; 9,720; 17,496 81,000 145,800 262,440
Total variable costs $376,000 $708,800 $1,339,840
10-28 (20 min.) Learning curve, incremental unit-time learning model .

1. The direct manufacturing labor-hours (DMLH) required to produce the first 2, 3, and 4
units, given the assumption of an incremental unit-time learning curve of 90%, is as follows:
Cumulative Individual
Number Unit Time Cumulative
of Units for Xth Unit Total Time
(1) (2) (3)
1 3,000 3,000
2 2,700 (3,000 0.90) 5,700
3 2,539 8,239
4 2,430 (2,700 0.90) 10,669
Values in column 2 are calculated using the formula y = aXb

where a = 3,000, X = 2, 3, or 4, and b = 0.1520, which gives
when X = 2, y = 3,000 2 0.1520 = 2,700
when X = 3, y = 3,000 3 0.1520 = 2,539
when X = 4, y = 3,000 4 0.1520 = 2,430
10-18
10-28 (Contd.)
Variable Costs of Producing
2 Units 3 Units 4 Units
Direct materials $80,000 2; 3; 4 $160,000 $240,000 $ 320,000
Direct manufacturing labor
$25 5,700; 8,239; 10,669 142,500 205,975 266,725
Variable manufacturing overhead
$15 5,700; 8,239; 10,669 85,500 123,585 160,035
Total variable costs $388,000 $569,560 $746,760
2. Variable Costs of
Producing
2 Units 4 Units
Incremental unit-time learning model
(from requirement 1) $388,000 $746,760
Cumulative average-time learning model
(from Exercise 10-27) 376,000 708,800
Difference $ 12,000 $ 37,960
Total variable costs for manufacturing 2 and 4 units are lower under the cumulative
average-time learning curve relative to the incremental unit-time learning curve. Direct
manufacturing labor-hours required to make additional units decline more slowly in the
incremental unit-time learning curve relative to the cumulative average-time learning curve when
the same 90% factor is used for both curves. The reason is that, in the incremental unit-time
learning curve, as the number of units double, only the last unit produced has a cost of 90% of
the initial cost. In the cumulative average-time learning model, doubling the number of units
causes the average cost of all the additional units produced (not just the last unit) to be 90% of
the initial cost.
10-29 (25 min.) High-low method.
1. Machine-Hours Maintenance Costs
Highest observation of cost driver 125,000 $250,000

Lowest observation of cost driver 85,000 170,000
Difference 40,000 $ 80,000
Maintenance costs = a + b (Machine-hours)

$80,000
Slope coefficient (b) = 40,000
= $2
Constant (a) = $250,000 ($2 125,000)
= $250,000 $250,000 = $0
or Constant (a) = $170,000 ($2 85,000)
= $170,000 $170,000 = $0
Maintenance costs = $2 Machine-hours
10-19
10-29 (Contd.)
2.
Plot and High-Low Line of Machine-Hours on Maintenance Costs
$260,000
M aintenance Costs 240,000
220,000
200,000
180,000
160,000
140,000
120,000
100,000
80,000 90,000 100,000 110,000 120,000 130,000
Machine-Hours
Solution Exhibit 10-29 presents the high-low line.
Economic plausibility. The cost function shows a positive economically plausible relationship
between machine-hours and maintenance costs. There is a clear-cut engineering relationship of
higher machine-hours and maintenance costs.
Goodness of fit. The high-low line appears to fit the data well. The vertical differences between
the actual and predicted costs appear to be quite small.
Slope of high-low line. The slope of the line appears to be reasonably steep indicating that, on
average, maintenance costs in a quarter varies with machine-hours used.
3. Using the cost function estimated in 1, predicted maintenance costs would be $2 90,000
= $180,000.
Howard should budget $180,000 in quarter 13 because the relationship between machine-
hours and maintenance costs in Solution Exhibit 10-29 is economically plausible, has an
excellent goodness of fit, and indicates that an increase in machine-hours in a quarter causes
maintenance costs to increase in the quarter.
10-20
10-30 (3040 min.) High-low method versus regression analysis.
1. Solution Exhibit 10-30 presents the plots of advertising costs on revenues.
Plot and Regression Line of Advertising Costs on Revenues
$90,000
80,000
70,000
60,000
R evenues
50,000
40,000
30,000
20,000
10,000
0
$0 $1,000 $2,000 $3,000 $4,000 $5,000
Advertising Costs
2. Solution Exhibit 10-30 also shows the regression line of advertising costs on revenues. We
evaluate the estimated regression equation using the criteria of economic plausibility, goodness
of fit, and slope of the regression line.
Economic plausibility. Advertising costs appears to be a plausible cost driver of revenues.

Restaurants frequently use newspaper advertising to promote their restaurants and increase their
patronage.
Goodness of fit. The vertical differences between actual and predicted revenues appears to be
reasonably small. This indicates that advertising costs are related to restaurant revenues.
Slope of regression line. The slope of the regression line appears to be relatively steep. Given
the small scatter of the observations around the line, the steep slope indicates that, on average,
restaurant revenues increase with newspaper advertising.
10-21
10-30 (Contd.)
3. The high-low method would estimate the cost function as follows:
Advertising Costs Revenues

Highest observation of cost driver $4,000 $80,000
Lowest observation of cost driver 1,000 55,000
Difference $3,000 $25,000
Revenues = a + (b advertising costs)
$25,000
Slope coefficient (b) = $3,000
= 8.333
Constant (a) = $80,000 ($4,000 8.333)
= $80,000 33,332 = $46,668
or Constant (a) = $55,000 ($1,000 8.333)
= $55,000 8,333 = $46,667
Revenues = $46,667 + (8.333 Advertising costs)
4. The increase in revenues for each $1,000 spent on advertising within the relevant range is
a. Using the regression equation, 8.723 $1,000 = $8,723
b. Using the high-low equation, 8.333 $1,000 = $8,333
The high-low equation does fairly well in estimating the relationship between advertising
costs and revenues. However, Martinez and Brown should use the regression equation. The
reason is that the regression equation uses information from all observations whereas the high-
low method relies only on the observations that have the highest and lowest values of the cost
driver. These observations are generally not representative of all the data.
10-31 (3035 min.) Regression analysis, activity-based costing, choosing cost

drivers.
1. Solution Exhibit 10-31A presents the plots and regression line of machinehours on
support overhead. Solution Exhibit 10-31B presents the plots and regression line of number of
batches on support overhead. As described below, using the three criteria of economic
plausibility, goodness of fit, and slope of regression line, Chu should choose number of batches
as the cost driver of support overhead costs.
Economic plausibility. Number of batches appears to be a more plausible cost driver of support
overhead costs than machine-hours. Support staff indicate that they spend a good portion of their
time at the start of each batch ensuring that the equipment is set up correctly and checking that
the first units of production in each batch are of good quality. Once the machine is working
properly, support staff are not needed to supervise the actual running of the machines.
Consequently, support staff resources are more likely to vary with the number of batches rather
than the total number of machine-hours worked.
10-22
10-31 (Contd.)
Goodness of fit. Compare Solution Exhibits 10-31A and 10-31B. The vertical differences
between actual and predicted costs are much smaller for number of batches than for machine-
hours. This indicates that number of batches has a better fit and a stronger relationship with
support overhead costs.
Slope of regression line. Again, compare Solution Exhibits 10-31A and 10-31B. The slope of
the regression line of number of batches on support overhead is relatively steep with less scatter
of observations about the regression line while the regression line of machine-hours on support
overhead is relatively flat (small slope) with more scatter of observations about the regression
line. A relatively steep regression line with less scatter for number of batches indicates that, on
average, support overhead costs increase as number of batches increase. On the other hand, the
relatively flat regression line for machine-hours with more scatter indicates a weak or no
relationship between support overhead costs and machine hourson average, changes in
machine-hours appear to have a minimal effect on support overhead costs.
2. As described in requirement 1, number of batches is the preferred cost driver. Using this
cost driver and the regression equation y = $16,031 + $197.30 number of batches, Chu should
budget the following support overhead costs for the 300 batches that will be run next month:
y = $16,031 + $197.30 300 = $16,031 + $59,190 = $75,221.
3a. Using machine-hours as the cost driver and the regression equation y = $28,089 + $10.23
machine-hours, Chu would budget support overhead costs for the 2,600 machine-hours that will
be worked next month as:
y = $28,089 + $10.23 2,600 = $28,089 + $26,598 = $54,687
Budgeted Revenues and Costs for

Next Month Using
Number of Batches Machine-Hours as
as the Cost Driver the Cost Driver
Costs other than support overhead $125,000 $125,000
Support overhead costs 75,221 54,687
Total costs 200,221 179,687
Add margin of 20% of total costs 40,044 35,937
Target revenues $240,265 $215,624
Picking machine-hours rather than the number of batches as the cost driver will cause Chu
to underestimate costs and choose lower target revenues and prices. Support overhead costs,
however, will vary with number of batches rather than machine-hours. Using information from
the preceding table, actual costs will be closer to $200,221 against target revenues of $215,624.
Target profitability is unlikely to be met. With better cost driver information Chu would
probably have priced products higher and earned greater revenues, assuming, of course, that
customers are willing to pay the higher prices.
10-23
10-31 (Contd.)
Choosing the "wrong" cost driver and estimating the incorrect cost function will also
have repercussions for cost management and cost control. Suppose Rohan Plastics budgets
support overhead costs of $54,687 for next month using the machine-hour regression. Suppose
actual support overhead costs, driven by number of batches, are $74,000 next month.
Management of Rohan Plastics would regard this as unsatisfactory performance and begin to
explore ways to cut costs to bring them more in line with budgeted support overhead costs. In
fact, on the basis of the preferred cost driver, number of batches, the plant's actual costs are lower
than the predicted amount, $75,221a performance that management should seek to replicate
rather than change. Using "wrong" cost drivers misleads management in cost planning, cost
management, and cost control besides contributing to inappropriate product pricing decisions.

Regression Line of Machine-Hours on Support Overhead Costs for Rohan Plastics
$90,0 00
80,00 0
S u p p o rt O v erh ea d C o sts
70,00 0
60,00 0
50,00 0
40,00 0
30,00 0
20,00 0
10,00 0
0
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000
Cost Driver: Machine-Hours
10-24
10-31 (Contd.)

Regression Line of Number of Batches on Support Overhead Costs for Rohan Plastics
$90,0 00
80,00 0
S u p p ort O verh ead C osts
70,00 0
60,00 0
50,00 0
40,00 0
30,00 0
20,00 0
10,00 0
0
0 50 100 150 200 250 300 350
Cost Driver: Number of Batches

10-32 (10-15 min.) Time lag consideration in interpreting regression results.
The relationship between machine-hours and maintenance costs is negative because Aza
Company schedules maintenance during slow production periods. There is a time lag between
heavy machine usage during peak production months, and maintenance in later months when the
production volume is low. To correctly understand the relationship between machine-hours and
maintenance costs, Aza should estimate the regression equation of maintenance costs on lagged
machine-hours (that is, machine-hours in prior months).
The explanation for the relationship between sales revenue and advertising costs is that
the result of advertising is often not instantaneous. Generally advertising generates increased
sales revenue in subsequent month(s), so Aza should estimate the relationship between
advertising costs in a particular period and sales in future periods. Another explanation is that
Aza increases its advertising costs during periods of declining sales in an attempt to increase
sales volume of the clothing lines that will soon be out of season. For example, Aza may increase
its advertising on winter clothing during late winter before the customers start purchasing spring
clothes.
10-25
10-33 (3040 min.) Cost estimation, cumulative average-time learning
curve.
1. Cost to Produce the Second through the Eighth Troop Deployment Boats:
Direct materials, 7 $100,000 $ 700,000

Direct manufacturing labor, 39,130* $30 1,173,900
Variable manufacturing overhead, 39,130 $20 782,600
Other manufacturing overhead, 25% of $1,173,900 293,475
Total costs $2,949,975
*The direct manufacturing labor-hours to produce the second to eighth boats can be calculated in several ways,
given the assumption of a cumulative average-time learning curve of 85%:
a. Use of Table Format:
Cumulative Cumulative Cumulative

Number of Average-Time Total
Units per Unit Time
1 10,000.00 10,000
2 8,500.00 (10,000 0.85) 17,000
4 7,225.00 (8,500 0.85) 28,900
8 6,141.25 (7,225 0.85) 49,130
The direct labor-hours required to produce the second through the eighth boats is 49,130
10,000 = 39,130 hours.
b. Use of Formula:
b
y = aX
where a = 10,000, X = 8, and b = 0.2345

y = 10,000 8 0.2345 = 6,141 hours (rounded)
The total direct labor-hours for 8 units is 6,141 8 = 49,128 hours
The direct labor-hours required to produce the second through the eighth boats is 49,128
10,000 = 39,128 hours. (By taking the b factor to 6 decimal digits, an estimate of 49,130
hours would result.)
Note: Some students will debate the exclusion of the tooling cost. The question specifies that
the tooling "cost was assigned to the first boat." Although Nautilus may well seek to ensure its
total revenue covers the $725,000 cost of the first boat, the concern in this question is only with
the cost of producing seven more PT109s.
10-26
10-33 (Contd.)
2. Cost to Produce the Second through the Eighth Boats Assuming Linear Function for Direct
Labor-Hours and Units Produced:
Direct manufacturing labor, 7 10,000 hours $30 2,100,000
Variable manufacturing overhead, 7 10,000 hours $20 1,400,000
Other manufacturing overhead, 25% of $2,100,000 525,000
The difference in predicted costs is:

Predicted cost in requirement 2
(based on linear cost function) $4,725,000
Predicted cost in requirement 1
(based on an 85% learning curve) 2,949,975
Difference $1,775,025
10-34 (2030 min.) Cost estimation, incremental unit-time learning model.
1. Cost to Produce the Second through the Eighth Boats:

Direct manufacturing labor, 49,356* $30 1,480,680
Variable overhead, 49,356 $20 987,120
Other overhead, 25% of $1,480,680 370,170
*The direct labor hours to produce the second through the eighth boats can be calculated via a
table format, given the assumption of an incremental unit-time learning curve of 85%:
Cumulative Individual Cumulative

Number Unit Time Total
of Units for Xth Unit (m)* Time
1 10,000 10,000
2 8,500 18,500
3 7,729 26,229
4 7,225 33,454
5 6,856 40,310
6 6,569 46,879
7 6,336 53,215
8 6,141 59,356
q
*Calculated as m = pX where p = 10,000, q = 0.2345, and X = 1, 2, 3,. . ., 8.
10-27
10-34 (Contd.)
The direct manufacturing labor-hours to produce the second through the eighth boat is 59,356
10,000 = 49,356 hours.
2. Difference in total costs to manufacture the second through the eighth boat under the
incremental unit-time learning model and the cumulative average-time learning model is
$3,537,970 (calculated in requirement 1 of this problem) $2,949,975 (from requirement 1 of
Problem 10-33) = $587,995.
The incremental unit-time learning curve has a slower decline in the reduction in time
required to produce successive units than does the cumulative average-time learning curve (see
Problem 10-33, requirement 1). Assuming the same 85% factor is used for both curves:
Estimated Cumulative Direct Manufacturing Labor-Hours

Cumulative Cumulative Average- Incremental Unit-Time
Number of Units Time Learning Model Learning Model
1 10,000 10,000
2 17,000 18,500
4 28,900 33,454
8 49,130 59,356
The reason is that, in the incremental unit-time learning model, as the number of units
double, only the last unit produced has a cost of 85% of the initial cost. In the cumulative
average-time learning model, doubling the number of units causes the average cost of all the
additional units produced (not just the last unit) to be 85% of the initial cost.
Nautilus should examine its own internal records on past jobs and seek information from
engineers, plant managers, and workers when deciding which learning curve better describes the
behavior of direct manufacturing labor-hours on the production of the PT109 boats.
10-28
10-35 (3040 min.) Evaluating alternative regression models, nonprofit.
1a. Solution Exhibit 10-35A plots the relationship between number of academic programs and
overhead costs.
1b. Solution Exhibit 10-35B plots the relationship between number of enrolled students and
overhead costs.
2. Solution Exhibit 10-35C compares the two simple regression models estimated by Hanks.
Both regression models appear to perform well when estimating overhead costs. Cost function 1
using number of academic programs as the independent variable appears to perform slightly
better than cost function 2 which uses number of enrolled students as the independent variable.
Cost function 1 has a high r2 and goodness of fit, a high t-value indicating a significant
relationship between number of academic programs and overhead costs, and meets all the
specification assumptions for ordinary least squares regression. Cost function 2 has a lower r2
than cost function 1 and exhibits positive autocorrelation among the residuals, as indicated by a
low Durbin-Watson statistic.
3. The analysis indicates that overhead costs are related to the number of academic programs
and the number of enrolled students. If Southwestern has pressures to reduce and control
overhead costs, it may need to look hard at closing down some of its academic programs and
reducing its intake of students. Reducing enrolled students may cut down on overhead costs but
it also cuts down on revenues (tuition payments), hurts the reputation of the school, and reduces
its alumni base, which is a future source of funds. For these reasons, Southwestern may prefer to
downsize its academic programs, particularly those programs that attract few students. Of
course, Southwestern should continue to reduce costs by improving the efficiency of the delivery
of its programs.

Plot of Number of Academic Programs versus Overhead Costs (in thousands)
$40,000
35,000
30,000
O v erh ea d C o sts
25,000
20,000
15,000
10,000
5,000
0
0 10 20 30 40 50 60 70 80 90 100 110
N umb er of Acad emic Programs
10-29
10-35 (Contd.)

Plot of Number of Enrolled Students versus Overhead Costs (in thousands)
$40,000
35,000
30,000
25,000
O v e r h e a d C o sts
20,000
15,000
10,000
5,000
0
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
Number of Enrolled Students
10-30
10-35 (Contd.)
SOLUTION EXHIBIT 10-35C

Comparison of Alternative Cost Functions for Overhead Costs Estimated with Simple
Regression for Southwestern University
Cost Function 1: Cost Function 2:

Number of Number of
Academic Programs as Enrolled Students as
Criterion Independent Variable Independent Variable
1. Economic Plausibility A positive relationship A positive relationship
between overhead costs and between overhead costs and
number of academic programs number of enrolled students is
is economically plausible at economically plausible at
Southwestern University. Southwestern University.
2. Goodness of Fit r2 = 0.72. r2 = 0.55.
Excellent goodness of fit. Good goodness of fit but not
as good as for number of
academic programs.
3. Significance of t-value of 5.08 is significant. t-value of 3.53 is significant.
Independent Variable(s)
4. Specification Analysis of Plot of the data indicates that Plot of the data indicates that
Estimation Assumptions assumptions of linearity, assumptions of linearity,
constant variance, constant variance, and
independence of residuals normality of residuals hold,
(Durbin-Watson statistic = but inferences drawn from
1.81), and normality of only 12 observations are not
residuals hold, but inferences reliable; the Durbin-Watson
drawn from only 12 statistic = 0.77 indicates that
observations are not reliable. independence of residuals
does not hold.
10-36 (30 min.) Evaluating multiple regression models, nonprofit

(continuation of Problem 10-35).
1. It is economically plausible that the correct form of the model of overhead costs includes
both number of academic programs and number of enrolled students as cost drivers. The
findings in Problem 10-35 indicate that each of the independent variables affects overhead costs.
(Each regression has a significant r2 and t-value on the independent variable.) Hanks could
choose to divide overhead costs into two cost pools, (i) those overhead costs that are more
closely related to number of academic program and (ii) those overhead costs more closely related
to number of enrolled students, and rerun the simple regression analysis on each overhead cost
pool. Alternatively, Hanks could run a multiple regression analysis with total overhead costs as
the dependent variable and the number of academic programs and number of enrolled students as
the two independent variables.
10-31
10-36 (Contd.)
2. Solution Exhibit 10-36A evaluates the multiple regression model using the format of
Exhibit 10-19. Hanks should use the multiple regression model over the two simple regression
models of Problem 10-35. The multiple regression model appears economically plausible, and
the regression model performs very well when estimating overhead costs. It has an excellent
goodness of fit, significant t-values on both independent variables, and meets all the specification
assumptions for ordinary least-squares regression.
There is some correlation between the two independent variables but multicollinearity does
not appear to be a problem here. The significance of both independent variables (despite some
correlation between them) suggests that each variable is a driver of overhead cost. Of course, as
the chapter describes, even if the independent variables exhibited multicollinearity, Hanks should
still prefer to use the multiple regression model over the simple regression models of Problem
10-35. Omitting any one of the variables will cause the estimated coefficient of the independent
variable, included in the model, to be biased away from its true value.
3. Possible uses for the multiple regression results include:
a. Planning and budgeting at Southwestern University. The regression analysis indicates

the variables (number of academic programs and number of enrolled students) that
help predict changes in overhead costs.
b. Cost control and performance evaluation. Hanks could compare actual performance
with budgeted or expected numbers and seek ways to improve the efficiency of the
University operations, and evaluate the performance of managers responsible for
controlling overhead costs.
c. Cost management. If cost pressures increase, the University might save costs by
closing down academic programs that have few students enrolled.
10-32
10-36 (Contd.)
Evaluation of Cost Function for Overhead Costs Estimated with Multiple Regression for
Southwestern University
Number of Academic Programs and Number
of Enrolled Students as Independent Variables
Criterion
1. Economic Plausibility A positive relationship between overhead costs
and number of academic programs and number
of enrolled students is economically plausible at
Southwestern University.
2. Goodness of Fit r2 = 0.81.

Excellent goodness of fit.
3. Significance of Independent Variable(s) t-values of 3.46 on number of academic

programs and 2.03 on number of enrolled
students are both significant.
4. Specification Analysis of Estimation The assumptions of linearity, constant variance,

Assumptions and normality of residuals hold, but inferences
drawn from only 12 observations are not reliable;
the Durbin-Watson statistic = 1.84 indicates that
independence of residuals holds.
10-33
10-37 (4050 min.) Purchasing Department cost drivers, activity-based
costing, simple regression analysis.
The problem reports the exact t-values from the computer runs of the data. Because the
coefficients and standard errors given in the problem are rounded to three decimal places,
dividing the coefficient by the standard error may yield slightly different t-values.
1. Plots of the data used in Regressions 1 to 3 are in Solution Exhibit 10-37A. See Solution
Exhibit 10-37B for a comparison of the three regression models.
2. Both Regressions 2 and 3 are well-specified regression models. The slope coefficients on
their respective independent variables are significantly different from zero. These results support
the Couture Fabrics's presentation in which the number of purchase orders and the number of
suppliers were reported to be drivers of purchasing department costs.
In designing an activity-based cost system, Fashion Flair should use number of purchase
orders and number of suppliers as cost drivers of purchasing department costs. As the chapter
appendix describes, Fashion Flair can either (a) estimate a multiple regression equation for
purchasing department costs with number of purchase orders and number of suppliers as cost
drivers, or (b) divide purchasing department costs into two separate cost pools, one for costs
related to purchase orders and another for costs related to suppliers, and estimate a separate
relationship for each cost pool.
3. Guidelines 1 and 2 presented in the chapter could be used to gain additional evidence on
cost drivers of purchasing department costs.
Guideline 1: Use physical relationships or engineering relationships to establish cause-and-

effect links. Lee could observe the purchasing department operations to gain insight into how
costs are driven.
Guideline 2: Use knowledge of operations. Lee could interview operating personnel in the
purchasing department to obtain their insight on cost drivers.
10-34
10-37 (Contd.)

Regression Lines of Various Cost Drivers on Purchasing Dept. Costs for Fashion Flair
$2,500,000
Department Costs
2,000,000
Purchasing
1,500,000
1,000,000
500,000
0
0 50 100 150
Dollar Value of Merchandise Purchased

(in millions)
$2,500,000
2,000,000
Department Costs
Purchasing
1,500,000
1,000,000
500,000
0
0 2,000 4,000 6,000 8,000
Number of Purchase Orders
$2,500,000
Depart ment Costs
2,000,000
Purchasing
1,500,000
1,000,000
500,000
0
0 100 200 300
Number of Suppliers
10-35
10-37 (Cont'd.)

Comparison of Alternative Cost Functions for Purchasing Department
Costs Estimated with Simple Regression for Fashion Flair
Regression 1 Regression 2 Regression 3

Criterion PDC = a + (b MP$) PDC = a + (b # of POs) PDC = a + (b # of Ss)
1. Economic Result presented at Economically plausible. Economically

Plausibility seminar by Couture The higher the number plausible. Increasing
Fabrics found little of purchase orders, the the number of
support for MP$ as a more tasks undertaken. suppliers increases the
driver. Purchasing costs of certifying
personnel at the vendors and managing
Miami store believe the Fashion Flair-
MP$ is not a supplier relationship.
significant cost driver.
2. Goodness of fit r2 = 0.08. Poor r2 = 0.42. Reasonable r2 = 0.39. Reasonable
goodness of fit. goodness of fit. goodness of fit.
3. Significance of t-value on MP$ of t-value on # of POs of t-value on # of Ss of

Independent 0.84 is insignificant. 2.43 is significant. 2.28 is significant.
Variables
4. Specification
Analysis
A. Linearity Appears questionable Appears reasonable. Appears reasonable.
within the but no strong evidence
relevant range against linearity.
B. Constant Appears questionable, Appears reasonable. Appears reasonable.
variance of but no strong evidence
residuals against constant
variance.
C. Independence Durbin-Watson Durbin-Watson Durbin-Watson
of residuals Statistic = 2.41 Statistic = 1.98 Statistic = 1.97
Assumption of Assumption of Assumption of
independence is not independence is not independence is not
rejected. rejected. rejected.
D. Normality of Data base too small to Data base too small to Data base too small to
residuals make reliable make reliable inferences. make reliable
inferences. inferences.
10-36
10-38 (3040 min.) Purchasing Department cost drivers, multiple
regression analysis (Continuation of 10-37).
The problem reports the exact t-values from the computer runs of the data. Because the
coefficients and standard errors given in the problem are rounded to three decimal places,
dividing the coefficient by the standard error may yield slightly different t-values.
1. Regression 4 is a well-specified regression model:
Economic plausibility: Both independent variables are plausible and are supported by the
findings of the Couture Fabrics study.
Goodness of fit: The r2 of 0.63 indicates an excellent goodness of fit.
Significance of independent variables: The t-value on # of POs is 2.14 while the t-value on # of
Ss is 2.00. These t-values are either significant or border on significance.
Specification analysis: Results are available to examine the independence of residuals

assumption. The Durbin-Watson statistic of 1.90 indicates that the assumption of independence
is not rejected.
Regression 4 is consistent with the findings in Problem 10-37 that both the number of
purchase orders and the number of suppliers are drivers of purchasing department costs.
Regressions 2, 3, and 4 all satisfy the four criteria outlined in the text. Regression 4 has the best
goodness of fit (0.63 for Regression 4 compared to 0.42 and 0.39 for Regressions 2 and 3,
respectively). Most importantly, it is economically plausible that both the number of purchase
orders and the number of suppliers drive purchasing department costs. We would recommend
that Lee use Regression 4 over Regressions 2 and 3.
2. Regression 5 adds an additional independent variable (MP$) to the two independent

variables in Regression 4. This additional variable (MP$) has a t-value of 0.07, implying its
slope coefficient is insignificantly different from zero. The r2 in Regression 5 (0.63) is the same
as that in Regression 4 (0.63), implying the addition of this third independent variable adds close
to zero explanatory power. In summary, Regression 5 adds very little to Regression 4. We would
recommend that Lee use Regression 4 over Regression 5.
3. Budgeted purchasing department costs for the Baltimore store next year are:
$485,384 + ($123.22 3,900) + ($2,952 110) = $1,290,662
4. Multicollinearity is a frequently encountered problem in cost accounting; it does not arise

in simple regression because there is only one independent variable in a simple regression. One
consequence of multicollinearity is an increase in the standard errors of the coefficients of the
individual variables. This frequently shows up in reduced t-values for the independent variables
in the multiple regression relative to their t-values in the simple regression:
10-37
10-38 (Contd.)
t-value from
t-value in Simple Regressions
Variables Multiple Regression in Problem 10-37
Regression 4:
# of POs 2.14 2.43
# of Ss 2.00 2.28
Regression 5:
# of POs 1.95 2.43
# of Ss 1.84 2.28
MP$ 0.07 0.84
The decline in the t-values in the multiple regressions is consistent with some (but not very high)
collinearity among the independent variables. Pairwise correlations between the independent
variables are:
Correlation
# of POs / # of Ss 0.29
# of POs / MP$ 0.27
# of Ss / MP$ 0.34
There is no evidence of difficulties due to multicollinearity in Regressions 4 and 5.
5. Decisions in which the regression results in Problems 10-37 and 10-38 could be useful are:
Cost management decisions: Fashion Flair could restructure relationships with the suppliers so
that fewer separate purchase orders are made. Alternatively, it may aggressively reduce the
number of existing suppliers.
Purchasing policy decisions: Fashion Flair could set up an internal charge system for individual
retail departments within each store. Separate charges to each department could be made for
each purchase order and each new supplier added to the existing ones. These internal charges
would signal to each department ways in which their own decisions affect the total costs of
Fashion Flair.
Accounting system design decisions: Fashion Flair may want to discontinue allocating
purchasing department costs on the basis of the dollar value of merchandise purchased.
Allocation bases better capturing cause-and-effect relations at Fashion Flair are the number of
purchase orders and the number of suppliers.
10-38
10-40 (40 min.) High-low method, alternative regression functions, accrual
accounting adjustments.
1. Solution Exhibit 10-40A presents the two data plots. The plot of engineering support
reported costs and machine-hours shows two separate groups of data, each of which may be
approximated by a separate cost function. The problem arises because the plant records
materials and parts costs on an "as purchased," rather than an "as used," basis. The plot of
engineering support restated costs and machine-hours shows a high positive correlation between
the two variables (the coefficient of determination is 0.94); a single linear cost function provides
a good fit to the data. Better estimates of the cost relation result because Kennedy adjusts the
materials and parts costs to an accrual accounting basis.
2.
Cost Driver Reported Engineering
Machine-Hours Support Costs
Highest observation of cost driver (August) 73 $ 617
Lowest observation of cost driver (September) 19 1,066
Difference 54 $ (449)
Slope coefficient, b =
Constant (at highest observation of cost driver) = $ 617 ($8.31 73) = $1,224
Constant (at lowest observation of cost driver) = $1,066 ($8.31 19) = $1,224
The estimated cost function is y = $1,224 $8.31X
Cost Driver Restated Engineering

Machine-Hours Support Costs
Highest observation of cost driver (August) 73 $966
Lowest observation of cost driver (September) 19 370
Difference 54 $596
Slope coefficient, b =
10-39
10-40 (Contd.)
Constant (at highest observation of cost driver) = $ 966 ($11.04 73) = $160
Constant (at lowest observation of cost driver) = $ 370 ($11.04 19) = $160
The estimated cost function is y = $160 + $11.04 X

3. The cost function estimated with engineering support restated costs better approximates the
regression analysis assumptions. See Solution Exhibit 10-40B for a comparison of the two
regressions.
4. Of all the cost functions estimated in requirements 2 and 3, Kennedy should choose
Regression 2 using engineering support restated costs as best representing the relationship
between engineering support costs and machine-hours. The cost functions estimated using
engineering support reported costs are mis-specified and not-economically plausible because
materials and parts costs are reported on an "as-purchased," rather than on an "as-used," basis.
With respect to engineering support restated costs, the high-low and regression approaches yield
roughly similar estimates. The regression approach is technically superior because it determines
the line that best fits all observations. In contrast, the high-low method considers only two points
(observations with the highest and lowest cost drivers) when estimating the cost function.
Solution Exhibit 10-40B shows that the cost function estimated using the regression approach
has excellent goodness of fit (r2 = 0.94) and appears to be well specified.
5. Problems Kennedy might encounter include:
a. A perpetual inventory system may not be used in this case; the amounts requisitioned
likely will not permit an accurate matching of costs with the independent variable on
a month-by-month basis.
b. Quality of the source records for usage by engineers may be relatively low; e.g.,
engineers may requisition materials and parts in batches, but not use them
immediately.
c. Records may not distinguish materials and parts for maintenance from materials and
parts used for repairs and breakdowns; separate cost functions may be appropriate for
the two categories of materials and parts.
d. Year-end accounting adjustments to inventory may mask errors that gradually
accumulate month-by-month.
6. Picking the correct cost function is important for cost prediction, cost management, and
performance evaluation. For example, had United Packaging used Regression 1 (engineering
support reported costs) to estimate the cost function, it would erroneously conclude that
engineering support costs decrease with machine-hours. In a month with 60 machine-hours,
Regression 1 would predict costs of $1,393.20 ($14.23 60) = $539.40. If actual costs turn
out to be $800, management would conclude that changes should be made to reduce costs. In
fact, on the basis of the preferred Regression 2, support overhead costs are lower than the
predicted amount of $176.38 + ($11.44 60) = $862.78a performance that management
should seek to replicate, not change.
10-40
10-40 (Contd.)
On the other hand, if machine-hours worked in a month were low, say 25 hours, Regression
1 would erroneously predict support overhead costs of $1,393.20 ($14.23 25) = $1,037.45. If
actual costs are $700, management would conclude that its performance has been very good. In
fact, compared to the costs predicted by the preferred Regression 2 of $176.38 + ($11.44 25) =
$462.38, the actual performance is rather poor. Using Regression 1, management may feel costs
are being managed very well when in fact they are much higher than what they should be and
need to be managed "down."

Plots and Regression Lines for Engineering Support
Reported Costs and Engineering Support Restated Costs
$1,400
E n g in e e r in g S u p p o r t R e p o r te d C o sts
1,200
1,000
600
800
400
200
0
0 10 20 30 40 50 60 70 80
Machine-Hours
$1,200
E n g in e e r in g S u p p o r t R e sta te d C o sts
1,000
800
600
400
200
0
0 10 20 30 40 50 60 70 80
Machine-Hours
10-41
10-40 (Contd.)

Comparison of Alternative Cost Functions for Engineering Support Costs at United Packaging
Regression 1 Regression 2
Dependent Variable: Dependent Variable:
Engineering Support Engineering Support
CRITERION Reported Costs Restated Costs
1. Economic Plausibility Negative slope relationship is Positive slope relationship is

economically implausible over economically plausible.
the long run.
2. Goodness of Fit r2 = 0.43. Moderate goodness r2 = 0.94. Excellent goodness

of fit. of fit.
3. Significance of t-statistic on machine-hours is t-statistic on machine-hours is

Independent Variables statistically significant highly statistically significant
(t = 2.31), albeit economically (t=10.59).
implausible.
4. Specification Analysis:
A. Linearity Linearity does not describe Linearity describes data very
data very well. well.
B. Constant variance of Appears questionable, although Appears reasonable, although

residuals 12 observations do not 12 observations do not
facilitate the drawing of facilitate the drawing of
reliable inferences. reliable inferences.
C. Independence of Durbin-Watson = 2.26. Durbin-Watson = 1.31. Some

residuals Residuals serially uncorrelated. evidence of serial correlation
in the residuals.
D. Normality of Database too small to make Database too small to make

residuals reliable inferences. reliable inferences.
10-42
10-43
Chapter 10 Case
U.S. BREWING INDUSTRY: COST ESTIMATION
1. (a) Using the high-low method:
Highest Barrels Sold = 17.037 Cost of Sales = $696.039

Lowest Barrels Sold = 5.844 Cost of Sales = $166.943
Difference 11.193 $529.096
slope coefficient (b) = $529.096 11.193 = $47.270
constant (a) = $696.039 ($47.270 17.037)
= $ 109.300
Ct = $109.300 + ($47.270 Vt)
A potential ambiguity in the high-low method is whether high-low is defined in terms of

the dependent or independent variable. In this case, the choice is important as the highest cost of
sales figure is 2000, whereas the highest volume figure is 1996. The independent variable is
used when determining the high and low observations as the causality runs from barrels produces
to costs incurred.
(b) The results are reported in Regression #1 in Case Exhibit 1see also Case Figure 1.
Evaluate the results: One approach is to use the following categorization:
Economic Interpretation - Both high-low and OLS imply the expected positive
relationship between costs and volume.
Statistical Specification - The major concern is with significant positive serial correlation
in the residuals (the Durbin-Watson Statistic = 0.219). The consequence is that the
estimated standard errors are underestimates of the underlying population values.
Goodness of Fit The adjusted R2 of 0.864 for the OLS model indicates a reasonably
good fit. A plot of the data, however, reveals that the model does not closely fit the data
in 1999 and 2000 when cost of sales increased while barrels sold decreased see Case
Figure 1.
OLS should yield more efficient (using a mean square error criterion) estimates than
high-low as the technique is constructed to use all the data rather than two extreme points
and to find the minimum mean square error set of estimates. Concentrating on the two
extreme observations may also result in true outliers affecting the estimated cost-
volume relationship.
10-44
Case (Contd.)
2. The results are reported in Regression #2 in Case Exhibit 1. A plot of the data is in Case
Figure 2. Relative to Regression #1, deflating by the WPIB results in several
improvements in model specification:
- t-value of the slope coefficient (b) increases from 10.749 to 50.911,

- R2 increases from 0.864 to 0.993, and
- Durbin-Watson Statistic moves closer to 2 (from 0.219 to 0.614)
Note, however, there is still evidence of significant positive autocorrelation in the
residuals -- e.g., Durbin-Watson Statistic = 0.614.
Problems include:
(a) The Wholesale Price Index of Beer (WPIB) is an output-based index for the whole industry
that need not be representative of the specific wholesale price changes for Ace. One could build
a firm specific index to overcome this problem.
(b) The WPIB may not be a good proxy for changes in input prices to either the industry or Ace.
The Cost of Sales series is an aggregate of the costs of labor, raw materials (malt, corn, barley,
hops, etc.), depreciation, excise taxes, marketing, etc. Deflating by the WPIB assumes that
changes in the prices of the inputs can be well approximated by changes in the WPIB. An
alternative approach is to build a firm-specific index based on changes in prices of Aces inputs.
(c) With any time-series index, issues of structural change arise. The WPIB may reflect a
nonconstant mix of beer products over time (malt, light, premium, super-premium, etc.).
Moreover, even if the mix of the output was constant, the mix of the inputs may have changed in
the 1982-2000 period, e.g., an increase in the capital/labor ratio through the construction of
mechanized, high-volume breweries.
(d) Data availability problems may arise due to the delay in publishing aggregate industry
indexes.
3. Background information: Serial correlation in the residuals exists when there is a

systematic pattern in the residuals such that knowledge of the residual at time t conveys
information about the residuals at time t+1, t+2, etc. A variety of reasons could cause serial
correlation in the residuals of a linear model, e.g.,
(i) Underlying model is nonlineartwo possible rationales for a nonlinear model are:
- costs are sticky downwardsi.e., when volume decreases by 10%, costs do not
decrease as much as the linear model predicts. Fixed capacity costs could
potentially be important for Ace in the 1996 to 2000 period when volume declined
from 17.037 to 15.091 million barrels.
- experience curve phenomenon could result in costs increasing less than the
linear model predicts.
(ii) Underlying model is linear but includes more than one variable and the omitted
variable results in the residuals being serially correlated. Omitted variables could
include the number of employees and the number of advertisements placed in various
media.
10-45
Case (Contd.)
(iii) Activity levels over time have high percentage of commonality. It is a common
finding that models estimated in levels exhibit serial correlation - - see C. W. J.
Granger and P. Newbold, Spurious Regressions in Econometrics, Journal of
Econometrics (No. 2, 1974).
Serial correlation (in general) affects the efficiency (but not the unbiasedness) of the OLS
regression estimates of a and b. With positive (negative) correlation, the estimates of the
standard errors will be understated (overstated) relative to the underlying population
standard errors. Thus, one may infer that the parameter estimates are more (less) precise
than they actually are.
Ways to detect serial correlation in the residuals include:

(i) Visual inspection of the residuals.
(ii) Compute the Durbin-Watson statistic which tests for first-order serial correlation.
The relevant values for this statistic, using a 5% significance level for one
independent variable, are:
No. of Positive Autocorrelation Negative Autocorrelation

Obsvns DW lower DW upper DW lower DW upper
18 1.03 1.26 2.74 2.97
19 1.06 1.28 2.72 2.94
The results are reported in Regression #3 in Case Exhibit 1. There is substantial

evidence that the serial correlation problem found in question #2 has been considerably
reduced, e.g., the Durbin-Watson deviates much less from 2 than previously (2.521 in #3
versus 0.614 in #2 and 0.219 in #1).
4. The best choice is regression #3. The Durbin-Watson statistic goes up from .219 in
regression #1, and .614 in regression #2 to 2.521 in regression #3. This indicates that the
serial correlation problem found in regressions #1 and #2 has been reduced. The t-value
of b in regression #3 is lower than that of regression #2, but higher than that of regression
#1. The R2 in regression #3 falls between that of #1 and #2, indicating that there is a
reasonably good fit for this model. Taken together, the significantly improved Durbin-
Watson statistic of regression #3, combined with a reasonable t-value for b and the R2,
makes this the preferred model.
10-46
Video Case (Contd.)
Case Exhibit 1
Durbin-
Regression a t-value b t-value R2
Watson
#1: Ct = f(Vt) 267.10 3.918 59.528 10.749 0.864 0.219
Ct
#2: = f(Vt) 30.517 2.280 55.360 50.910 0.993 0.614
Pt
Ct Ct 1
#3: f(Vt Vt-
Pt Pt 1 4.660 1.453 47.034 12.989 0.908 2.521
)
1
Case Figure 1
Plot of Cost-Volume Data
10-47
Case (Contd.)
Case Figure 2
Plot of Deflated Cost Versus Volume
10-48

3 27

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

3 27

Enviado por

Direitos autorais:

Formatos disponíveis

CHAPTER 10

DETERMINING HOW COSTS BEHAVE

10-1 The two assumptions are:

10-2 Three alternative linear cost functions are:

10-5 Four approaches to estimating a cost function are:

10-8 The six steps are:

10-16 (10 min.) Estimating a cost function

= = $0.30 per machine-hour

The cost function based on the two observations is:

1. See Solution Exhibit 10-17.

10-21 (30 min.) Account analysis method.

1. Manufacturing cost classification for 2004:

Direct materials $300,000 100% $300,000 $ 0 $4.00

Total manufacturing cost for 2004 = $948,750

Direct materials $4.00 5% $0.20 $4.20 $336,000

Direct materials $ 0 0% $ 0 $ 0 $336,000 $ 336,000

Customer-service department costs = a + b (number of service reports)

3. Other possible cost drivers of customer-service department costs are:

SOLUTION EXHIBIT 10-22

1. Slope coefficient (b) =

Constant (a) = $529,000 ($43.00 7,000)

Cost function = $228,000 + $43.00 (professional labor-hours)

The linear cost function is plotted in Solution Exhibit 10-23.

2. A comparison at various levels of professional labor-hours follows. The linear cost

Total overhead cost behavior:

Month 1 Month 2 Month 3 Month 4 Month 5 Month 6

3. Based on Based on Linear

The total contribution margin actually forgone is $3,000.

SOLUTION EXHIBIT 10-23

10-24 (20 min.) Cost-volume-profit and regression analysis.

Expected cost to make

Economic plausibility. Labor-hours appears to be an economically plausible driver of

Food and beverages $15.00

SOLUTION EXHIBIT 10-25

10-25 Excel Application

Month Labor-Hours Overhead

Relationship Between Overhead Costs and Labor-

SOLUTION EXHIBIT 10-26A

SOLUTION EXHIBIT 10-26B

2. Using the preferred cost function,

10-27 (20 min.) Learning curve, cumulative average-time learning model.

where a = 3,000, X = 2, 4, or 8, and b = 0.1520, which gives

Variable Costs of Producing

10-28 (20 min.) Learning curve, incremental unit-time learning model .

Values in column 2 are calculated using the formula y = aXb

10-29 (25 min.) High-low method.

1. Machine-Hours Maintenance Costs

Highest observation of cost driver 125,000 $250,000

Maintenance costs = a + b (Machine-hours)

Solution Exhibit 10-29 presents the high-low line.

Economic plausibility. Advertising costs appears to be a plausible cost driver of revenues.

3. The high-low method would estimate the cost function as follows:

Advertising Costs Revenues

10-31 (3035 min.) Regression analysis, activity-based costing, choosing cost

y = $28,089 + $10.23 2,600 = $28,089 + $26,598 = $54,687

Budgeted Revenues and Costs for

SOLUTION EXHIBIT 10-31A

Cost Driver: Machine-Hours

SOLUTION EXHIBIT 10-31B

Cost Driver: Number of Batches

Direct materials, 7 $100,000 $ 700,000

a. Use of Table Format:

Cumulative Cumulative Cumulative