Managing budget emphasis through the explicit design of conditional budgetary slack Tony Davila a , Marc Wouters b, * a Graduate School of Business, Stanford University, USA b School of Business, Public Administration, and Technology, University of Twente, P.O. Box 217, 7500 AE Enschede, Netherlands Abstract Budgetary slack plays an important role in the functioning of budgets in organizations. While theory has found neg- ative as well as positive elements associated with its presence, the empirical literature has interpreted it as being dysfunc- tional to organizations. In this paper, we present empirical evidence on how a company purposefully budgeted additional financial resources with a motivation intention (Lukka. Budgetary biasing in organizations: Theoretical framework and empirical evidence. Accounting, Organizations and Society 13 (1998) 281–302) to facilitate the manag- ersÕ task in achieving the goals of the company. Using quantitative and qualitative data from four logistic sites of a disk drive manufacturer for 24 months, we examine how the company accepted more slack as the demand on business proc- esses increased and goals other than budget targets––in particular, service quality––became harder to achieve. By allow- ing this practice, headquarters made it clear to local managers that product quality and service were at least as important as meeting budget objectives. We also find that not only was budgetary slack purposefully built during the budgeting process but also in the budgeting system itself through the underlying cost accounting assumptions. The results of this paper provide empirical evidence on the positive aspects of budgetary slack and on the role of cost accounting models used in the budgeting system to facilitate managerial work. Ó 2004 Elsevier Ltd. All rights reserved. Introduction Budgets are probably the management tool most widely used in organizations. Their relevance has nurtured significant research efforts and cre- ated the ‘‘only organized critical mass of empirical work in management accounting’’ (Brownell & Dunk, 1991). A key variable that has traditionally been perceived as limiting the effectiveness of budgets is budgetary bias (Lukka, 1988) and, in particular one of its sub-components: budgetary slack (Dunk & Nouri, 1998; Kamin & Ronen, 1981; Merchant, 1985; Walker & Johnson, 1999; 0361-3682/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.aos.2004.07.002 * Corresponding author. E-mail addresses: [email protected](T. Davila), [email protected](M. Wouters). www.elsevier.com/locate/aos Accounting, Organizations and Society 30 (2005) 587–608
22
Embed
Managing Budget Emphasis Through the Explicit Design of Conditional Budgetary Slack
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
www.elsevier.com/locate/aos
Accounting, Organizations and Society 30 (2005) 587–608
Managing budget emphasis through the explicit designof conditional budgetary slack
Tony Davila a, Marc Wouters b,*
a Graduate School of Business, Stanford University, USAb School of Business, Public Administration, and Technology, University of Twente, P.O. Box 217, 7500 AE Enschede, Netherlands
Abstract
Budgetary slack plays an important role in the functioning of budgets in organizations. While theory has found neg-
ative as well as positive elements associated with its presence, the empirical literature has interpreted it as being dysfunc-
tional to organizations. In this paper, we present empirical evidence on how a company purposefully budgeted
additional financial resources with a motivation intention (Lukka. Budgetary biasing in organizations: Theoretical
framework and empirical evidence. Accounting, Organizations and Society 13 (1998) 281–302) to facilitate the manag-
ers� task in achieving the goals of the company. Using quantitative and qualitative data from four logistic sites of a disk
drive manufacturer for 24 months, we examine how the company accepted more slack as the demand on business proc-
esses increased and goals other than budget targets––in particular, service quality––became harder to achieve. By allow-
ing this practice, headquarters made it clear to local managers that product quality and service were at least as
important as meeting budget objectives. We also find that not only was budgetary slack purposefully built during
the budgeting process but also in the budgeting system itself through the underlying cost accounting assumptions.
The results of this paper provide empirical evidence on the positive aspects of budgetary slack and on the role of cost
accounting models used in the budgeting system to facilitate managerial work.
� 2004 Elsevier Ltd. All rights reserved.
Introduction
Budgets are probably the management toolmost widely used in organizations. Their relevance
0361-3682/$ - see front matter � 2004 Elsevier Ltd. All rights reserv
ordinates from excessive risk, and increases jobsatisfaction as budgetary slack avoids the perform-
ance consequences of missing the budget––such as
loss of bonuses, loss of job (Merchant & Manzoni,
1989), or the social pressure associated with
underachievers.
More importantly, budgetary slack may reduce
budget emphasis and allow subordinates to allo-
cate attention to goals other than meeting the bud-get––such as quality or customer service––when
meeting the budget becomes harder. In work set-
tings with multiple goals, such as cost, efficiency,
productivity, quality, customer service, and
responsiveness, tradeoffs between these goals exist
(Lillis, 2002). When these multiple goals cannot be
simultaneously attained, managers may sacrifice
goals that should have priority. Budgetary slackfacilitates pursuing objectives other than meeting
the budget without fully ignoring the relevance
of budgets. For example, budgetary slack may ease
(but not eliminate) the constraint imposed through
the budget and give the manager the ability to
meet non-financial goals when production volumes
are unexpectedly high.
The notion that organizations may use budget-ary slack purposefully leads to a more elaborate
understanding of how budgetary slack may be cre-
ated. It is not always the case that organizations
prefer an easily achievable target; the objective is
not to make cost targets always easily achievable,
but only when it is particularly difficult to meet
financial and non-financial goals simultaneously.
The process of creating budgetary slack becomessubtler than just using relatively higher levels of
allowable costs. The deliberate design of easier
cost targets when conditions are demanding con-
ceptually represents budgetary slack; although this
budgetary slack is purposeful, motivational, and
contingent on the environment. This concept of
budgetary slack, which includes a performance-
enhancing element in a setting with multiple goals,incomplete performance measures, and uncer-
tainty, is used in the development of our
propositions.
2 An alternative reason why budgetary slack may be present
at higher volumes is higher uncertainty about costs when
volume is higher. The company may protect the manager
through budgetary slack. The evidence that we gathered does
not allow us to test this alternative argument. We thank one of
the reviewers for suggesting this argument.3 It is important to notice that our empirical study was
conducted in a cost center setting, where increasing budgetary
slack does not necessarily lead to lower profits, the reason being
that meeting non-financial performance measures may enhance
the revenue line. This focus on cost centers is in contrast to
other empirical studies that have focused on profit centers
(Otley, 1978). We thank one of the reviewers for pointing out
this important distinction.
T. Davila, M. Wouters / Accounting, Organizations and Society 30 (2005) 587–608 591
Propositions
The manager of a department pursuing multiple
financial and non-financial objectives––such as
operational costs, customer service, and productquality––decides how to trade off these multiple
goals when they cannot be achieved simultane-
ously. The organization may influence these deci-
sions through the budgetary slack available to
the manager. Organizations may purposefully in-
crease budgetary slack when external conditions
are more demanding, in an attempt to ease meet-
ing budget goals and freeing up managers� atten-tion to achieve alternative organizational goals––
like service or quality. Conversely, organizations
may decrease budgetary slack when conditions
are such that budget emphasis becomes more
relevant.
Unexpected events such as unplanned surges in
demand are likely to lead to decision points where
the manager cannot simultaneously meet multiplegoals. To react to these events, the manager may
choose to increase operating costs in order to
maintain customer service or meet the budget at
the expense of delaying some of the shipments.
The company can influence these trade-off deci-
sions through the budgetary slack that it allows
when these unexpected events happen. A large
budgetary slack encourages the manager to focuson non-financial objectives; conversely a low
budgetary slack focuses attention on financial
objectives. The following proposition reflects this
argument:
Proposition 1. Unexpected higher levels of activity
are associated with larger budgetary slack.
Another situation where managers may facetrade-off decisions is when the expected demand
is large. As demand gets closer to capacity, manag-
ers may have to rely on more expensive variable
costs––such as overtime, temporary employees, re-
work, and express shipments. The closer to full
capacity, the more often the sites will need to use
these expensive resources. While it is uncertain
how often these resources will be needed, higherexpected demand increases the number of times
that the site hits capacity constraints. If goals such
as product quality or customer service are impor-
tant, the company may decide to increase budget-
ary slack when demand is expected to be large.
Increasing budgetary slack reduces budget empha-
sis and allows managers to focus more attention to
meeting these alternative goals. 2 The larger thebudgetary slack, the lower is the priority of bud-
gets. The following proposition captures this
argument: 3
Proposition 2. Increased level of activity is associ-ated with larger budgetary slack.
The previous propositions focus on how cost
budgets can be managed in the presence ofdemanding conditions. However, these conditions
also affect the firm�s ability to achieve non-finan-
cial objectives such as customer service. This holds
unless the company is willing to meet one of its
goals at any price (requiring the company to pro-
vide enough extra resources to completely meet
this alternative goal in any state of the world)
and then we would not expect any effect of in-creased level of activity on this particular goal.
In our research setting, one of the alternative goals
is customer service. If demanding conditions also
hamper the ability of the company to meet these
alternative goals, then we expect the following
propositions to hold:
Proposition 3a. Lower customer service is associ-
ated with unexpected higher levels of activity.
Proposition 3b. Lower customer service is associ-
ated with increased levels of activity.
592 T. Davila, M. Wouters / Accounting, Organizations and Society 30 (2005) 587–608
Qualitative evidence: Company PCC and its
budgeting process
This section describes the company, its budget-
ing process, and qualitative evidence informing theprevious propositions. As part of the data collec-
tion process, we interviewed and e-mailed manag-
ers in the following departments: budgeting,
financial accounting, quality control (including
non-financial data management), logistics per-
formance measurement, supply chain engineering,
and logistic sites. The initial interviews lasted be-
tween 30 and 60 minutes and we had several addi-tional interviews and e-mail exchanges throughout
the course of two years to clarify our understand-
ing of the data and the processes, and to help us
interpret the results.
The company
The study focuses on the four logistic centers ofcompany PCC (a disguised name) located in North
America, Europe, Asia, and Southeast Asia be-
tween April 1998 and March 2000. Company
PCC designs and produces disk drives for comput-
ers. The industry grew rapidly during the 1980s
and went through a period of intense competition
and restructuring during the 1990s. Out of the 43
companies that received venture capital, less than10% survived into the year 2000. The market is
highly competitive in several dimensions. Accord-
ing to PCC�s director of Supply Chain Engineer-
ing: ‘‘Consistently advancing technology through
product development is one of the most critical as-
pects; battles are won at the product development
stage’’. This point is further illustrated by the fact
that manufacturing is outsourced.Once a new product has been introduced, PCC
subcontracts the manufacturing process to an
OEM that delivers the products to the logistic cen-
ters. At the logistic centers, about 50% of the units
undergo a small customization and testing, such as
adding customer labels or adjusting switches to the
requirements of each client. The rest of the units
are simply repackaged. Units are then shipped tothe customer. When a product enters the produc-
tion stage, competition happens in three dimen-
sions: quality, cost, and customer service.
Quality
The importance of quality is highlighted in the
following quote from PCC�s director of Supply
Chain Engineering:
‘‘Our largest customers include Dell, IBM, Apple,HP, and Compaq. They all require the highestlevel of product and service quality because theymanufacture under JIT systems where the cost oflow quality is very high. . . .We identify each singleunit with a unique serial code and we carefullydocument its movement through the supply chain.Whenever a customer rejects a unit because ofquality problems, we test the unit and use individ-ual unit information to trace back the problem. . . .The importance of quality puts a lot of emphasison having a clear understanding of the wholecause-effect relationships through the supply chainfrom the component suppliers, the OEM manufac-turer, our own logistic centers, and the delivery tothe client. When units with quality problems comeback we determine what kind of failure occurredand our new diagnostic software finds 32 differenterror codes’’.
The company tracks the number of failed prod-
ucts per million units, where in the supply chain
the failure occurred (within the manufacturing
partner, the logistic centers, the customer during
the assembly process, or the end user), and whattype of failure happened.
‘‘For warranty costs we accrue a certain amountper unit that goes out of the door, and the actualcosts are incurred because of repairs, replace-ments, and credits. We do not have cost data dee-per than product family, but the systems trackserial numbers of returned products, when it wasshipped, and to which customer’’, explained amember of the controller�s office. We followed thisup with a director of the quality department whoexplained that ‘‘using the serial number we tracewhen it was built, in what factory, and on whichline’’.
Operational costs
Cost is the second competitive dimension in a
market characterized by intense competition and
small gross margins (less than 20% for fiscal year
T. Davila, M. Wouters / Accounting, Organizations and Society 30 (2005) 587–608 593
2000). Because product costs are mainly deter-
mined at the product development stage, once
the product is launched, PCC focuses on reducing
the cost of operating its supply chain. According
to the director of Supply Chain Engineering:
‘‘We have seen large price reductions over the lastyears, while at the same time we and our compet-itors have increased product performance dramat-ically. . . . Cost reduction is vital and we have veryrigorous cost targets to design costs out of theproduct; not just material cost, but manufactur-ing cost and logistics have also become veryimportant. . . . At the logistic sites we focus oncontrolling the costs of operating the sites andinventory costs. Good inventory management isparticularly important because of the rapid priceerosion that requires a negative inventory revalu-ation every quarter. We use the unique serialnumber of each unit not only to trace qualityproblems, but also to monitor how many dayseach unit stays at each phase of the supply chain,and improve throughput . . . We are currentlyimplementing a supply chain initiative to reducethe small customization tasks that we do at thelogistic sites. These tasks are costly because wehave to unpack, make a small change like addingstickers, and repack. The simple task of unpack-ing and repackaging also affects quality. Weexpect that by shipping generic (non-customized)products, we will improve customer service andtake advantage of cost reduction opportunitiesat the logistic sites’’.
Thus, cost management discipline within the
supply chain goes as far as tracing the time each
single unit spends in the supply chain. According
to a business analyst in the quality control depart-ment, ‘‘Every month we measure for each unit that
we shipped to customers �how long did we have it?�and we report the average number of days and the
standard deviation by part number and by site’’.
Logistic centers are managed as cost centers and
the importance of costs makes budgets a key con-
trol system. While small variances do not attract
significant management attention, the budget of-fice carefully investigates large unexplained vari-
ances that also impact the performance
evaluation of logistic center managers. The cost
of moving a unit through the supply chain (cost
per unit) is also carefully monitored as an impor-
tant performance measure.
In addition to the quality control procedures
that give a lot of information about the actual ac-tions performed at the logistic sites, additional
control systems complement the budget. These
controls include approval for new investments
and new hires, monthly reports on headcount per
department, and the negotiation at headquarters
of freight contracts. The budgeting manager de-
scribed their discretion regarding freight: ‘‘All the
freight carriers are pre-negotiated with (the com-pany). Site managers cannot choose the carrier.
If there are any deviations from carriers and terms,
it is due to meeting customers needs. Freight costs
are determined based on a few different variables
. . . weight, size (bulk), type of freight (overnight,
ocean liner, air, etc.) that the site manager deci-
des’’. According to the controller: ‘‘We do not rely
only on the budget, but have additional tight con-trols over decisions that may have a significant im-
pact on costs’’.
Customer service
In addition to quality and costs, delivering a
high level of service to customers (most on JIT
manufacturing systems) is the third, albeit more
intangible, piece to the competitive landscape. Asthe director of Supply Chain Engineering
described:
‘‘We measure customer service using two objectiveperformance measures, one reflecting the on-timedelivery of products compared to the confirmeddate, and one reflecting the confirmation of deliv-ery dates compared to the initial request of the cus-tomer. . . . We also have every quarter a quarterlybusiness review (QBR) where the top eight cus-tomers evaluate the level of service received fromPCC and from the logistic centers in particular.The evaluation includes different service dimen-sions and the overall customer satisfaction. TheQBR is the main criterion to evaluate the logisticcenters� performance and determine the bonus oflogistic center managers and other executives inthe company involved in the supply chain. . . .The QBR is very visible for everyone’’.
594 T. Davila, M. Wouters / Accounting, Organizations and Society 30 (2005) 587–608
Logistic centers pursue multiple objectives
simultaneously: cost, service, and quality. Even if
PCC�s products rely on advanced technologies
and accordingly the industry is considered high-
tech, the technology is well-known and the prod-uct is considered to be in the mature part of its life
cycle (Foster, 1986). Production tasks at the logis-
tic centers are simple, well-understood repetitive
processes, with a large amount of non-financial
information available for quality purposes. This
environment significantly reduces the information
asymmetry among different hierarchical levels in
the company. The challenge for sites� managers be-comes difficult performance goals––on-time deliv-
ery, product quality, and cost control––as well as
demand uncertainty. Fig. 1 plots the percentage
difference in actual volume compared to budgeted
volume. The difference is close to ±50%.
The budget setting process
The budgeting process is on a yearly cycle.
However, the budget for the second half of the
year is updated during the second quarter. Accord-
ing to the manager in charge of the budgeting
process, the objective of this update is ‘‘to include
new market information available through the
early part of the year as well as to update financial
expectations, maintain the accuracy of the budget,and include the impact that unexpected changes in
the environment may have on the budget. We want
the budget to be as useful as possible as a guide for
performance expectations’’.
The budgeting process for the logistic centers
starts with an estimate of quarterly and monthly
-50
-30
-10
10
30
50
0 5 10 15 20 25
Per
cen
tag
e d
iffe
ren
ce
Months
Fig. 1. Percentage difference in actual versus budgeted volume.
volume from the Fulfillment Planning Group that
brings together customers� needs, OEM capacity,
and market expectations. According to the con-
troller, ‘‘Supply teams work with each major cus-
tomer to learn about expected demand and thefulfillment people work with the manufacturer to
know about capacity problems’’. The budget man-
ager described it as follows:
‘‘At the Financial Planning Group, we use the esti-mations from the supply teams, fulfillment people,and industry statistics to define expected demand.Then we have analytical models to estimate work-force loads, freights, and costs trends together withthe profit expectations that come from the top ofthe organization (and given as a guide to WallStreet) to delineate the first draft of the budgetfor the logistic centers. At the same time, the logis-tic centers receive the volume estimates from theFulfillment Planning Group and prepare theirown detailed cost budget based on their under-standing of the cost of activities. We review thecost budgets that the logistic centers submit, andcompare them with our own budget numbers,looking at trends, comparing with the actual num-bers of last quarter, and by confronting the budg-ets with a target for maximum spending based onaffordability, which we do not reveal to the logisticcenters in advance. Finally, we talk with the logis-tic sites to reconcile both budgets . . . This last stepis not difficult, because we have done this exercisemany times and we all have a pretty good ideaabout how things work at the logistic sites . . .The focus of the exercise is to give an expense bud-get to site managers, but also to keep a tight con-trol on the costs of delivering one unit of product. . . The last line in the budget is the (logistic) costper unit that we follow very closely’’.
In order to maintain customer satisfaction dur-
ing peak demands or during periods when quality
problems require additional configuration and
testing activities, budgetary slack is common and
accepted. According to a senior manager of
Worldwide Operations, ‘‘we typically allow over-
budgeting at the logistic sites to plan for emer-gency overtime to maintain customer satisfaction
during something like a quality issue or crisis, even
if it is not usually needed’’. Thus budgetary slack is
T. Davila, M. Wouters / Accounting, Organizations and Society 30 (2005) 587–608 595
not usually ‘‘used’’ and translates into a favorable
budget variance. One of the logistic managers con-
firmed the idea of including budgetary slack to re-
act to emergencies: ‘‘We budget enough resources
to be able to keep our service level even in emer-gency situations’’. The budget setting process
incorporates budgetary slack to react to ‘‘emergen-
cies’’, which happen more often for months with
higher expected volumes. The budget manager de-
scribed her awareness about the existence of this
budgetary slack:
‘‘When budgets are put together in Logistics, thevolume and cost structures are set some time inadvance based on forecasted information. Thecosts usually come in lower than forecasted on aregular basis. We have about a 5-cent pad (perdisk drive) in the numbers on a regular basis. Thisis for unforeseen events that might come up. Inaddition to that, the volume that we use in thebudget is a low-risk volume. So, if the actual vol-ume comes in higher than the budgeted volume,the cost per drive is lower’’.
This ‘‘padding’’ does not mean that the com-
pany systematically adds 5 cents per disk to the
budget, but that on average actual cost per unit
comes 5 cents below budgeted cost per unit. In
fact, because ‘‘emergency situations’’ are more
likely in high demand months, budgetary slack islarger in these months and lower in low volume
months.
Budget reports for the logistic sites separate
fixed from variable costs. The main items in the
fixed cost category are indirect labor, facilities,
and equipment. The two main items in the variable
cost category are direct labor and freight; the rest
of the items are comparatively smaller. Logisticcenter managers have some discretion over varia-
ble costs and can influence them to a larger extent
than fixed costs. For example, site managers can
decide to keep people, even if there is not enough
work, to quickly react to unplanned demand.
The budget review process
During the year, the budget office issues
monthly budget control reports detailing itemized
quarter-to-date actual expenses, quarter-to-date
static budget, and the variance between the two.
The report also includes actual and budgeted logis-
tic costs per unit, which is an additional important
criterion to evaluate the logistic sites� financial per-formance, and actual and budgeted units shipped.
This monthly feedback cycle helps to quickly de-
tect variances and correct them by the end of the
quarter, when performance evaluation happens.
Variances go back to zero at the end of each quar-
ter to start a new cycle. According to a member of
the controller�s office, ‘‘Variances are an important
indication of how we are going to exit the quar-ter’’. The logistic site manager described how he
used the budget reports: ‘‘I follow closely the
monthly budget reports, I look at the actual costs,
the variances and the cost per unit. If I see large
variances I quickly call my controller to under-
stand where they come from and what we can do
to make sure that they go back to a reasonable
level before the end of the quarter. Costs are animportant part of our job, although not the only
one; I also keep track of customer service that
impacts my QBR and make sure that it does not
suffer even in peak demand periods’’.
The monthly control reports are homogeneous
for all sites with the same format and accounts.
The accounts are classified as fixed and variable
costs. For purposes of the quantitative analysis,we grouped fixed (or indirect) labor cost ac-
Regressions include dummy variables to control for differences across sites (not reported).
Pseudo R2 is the correlation between the actual dependent variable and the fitted values.* Significant at 10%.** Significant at 1%.*** Significant at 5%.
T.Davila
,M.Wouters
/Acco
untin
g,Organiza
tionsandSociety
30(2005)587–608
603
13 The magnitude of these changes is quite sizeable. Given
the simple operations performed at the logistic sites, such
604 T. Davila, M. Wouters / Accounting, Organizations and Society 30 (2005) 587–608
The mechanism for the budgeting model de-
scribed in Section ‘‘The budgeting model’’ relies
on variable costs having a fixed component to engi-
neer the relationship between unexpected volume
and favorable efficiency variance. To examine thisrelationship using quantitative data, we examine
the cost behavior assumed during the budgeting
process and the actual cost behavior. For each of
the sites, we performed the following regression:
Actual or budgeted costs
¼ b0 þ b1 � actual or budgeted volume þ e
We also include an AR(1) error term to control
for serial correlation. We run separate regressions
for each site to fully capture the differences in cost
structures, for actual versus budgeted costs, and
for total variable costs and its three components
(direct labor, freight costs, and other variable
costs). To assess the significance of the coefficients,
we combined the z-statistics of the four sites intoan overall z-statistic estimated as Z ¼ �z=ðstdevðzÞ=