Based on the article need the below concepts to be written 3.2.2. Criticism 2: budgets impede adaptability 3.2.4. Criticism 4: the use of budgets as fixed performance contracts Introduction of t

Management Accounting Research 21 (2010) 56–75 Contents lists available atScienceDirect Management Accounting Research journal homepage:www.elsevier.com/locate/mar Beyond budgeting or budgeting reconsidered? A survey of North-American budgeting practice Theresa Libby a,∗ , R. Murray Lindsay b aSchool of Business and Economics, Wilfrid Laurier University, Waterloo, ON, CanadabFaculty of Management, University of Lethbridge, Lethbridge, Alberta, Canada article info Keywords:

Budgets Beyond budgeting Survey of budgeting practice abstract Budgets have historically played a key role in management control; however, recently they have become the subject of considerable criticism and debate. Some argue that the problems with budgeting stem from the way budgets are used (Horngren et al., 2004) while others argue that budgeting processes are fundamentally ﬂawed (Hope and Fraser, 2003a).Hansen et al. (2003), among others, have called for a systematic examination of these issues against empirical evidence. In this paper, we present the results of two surveys of mid- to large-sized North-American organizations to 1) update the literature on North-American budgeting practices, 2) collect empirical evidence to assess the criticisms, and 3) begin to identify strong tendencies or patterns in budgeting practice to inform future academic research. Overall, we ﬁnd for the majority of ﬁrms that budgets continue to be used for control purposes and are perceived to be value-added. While problems exist with budgets, organizations are adapting their use to account for these problems rather than abandoning budgets altogether.

There is no other managerial process I am aware of that translates qualitative mission statements and corporate strategies into action plans, links the short term with the long term, brings together managers from different hierarchical levels and from different functional areas, and at the same time provides continuity by the sheer regularity of the process (S. Umapathy, Current Budgeting Practices in US Industry, 1987, xxii).

Not to beat around the bush, but the budgeting process at most companies has to be the most ineffective prac- tice in management. It sucks the energy, time, fun and big dreams out of an organization. It hides opportunity and stunts growth. It brings out the most unproductive behav- iors in an organization, from sandbagging to settling for mediocrity. In fact, when most companies win, it is in spite ∗ Corresponding author.

E-mail addresses:[email protected](T. Libby), [email protected](R.M. Lindsay). of their budgets, not because of them (Jack Welch, Winning, 2005, 189).

1. Introduction The budget has historically played center stage in most organizations’ systems of management control (Otley, 1994). However, recently it has been the subject of con- siderable criticism (Hansen et al., 2003). Budgeting has been deemed “broken” (Jensen, 2001), “a thing of the past” (Gurton, 1999), or an “unnecessary evil” (Wallander, 1999).

European surveys also report a growing dissatisfaction among organizations with their budgeting systems (Neely et al., 2003; Eckholm and Wallin, 2000).

The case against traditional budgeting has been argued most forcefully byHope and Fraser (2003b)as part of the Beyond Budgeting Roundtable (BBRT). 1The authors argue 1Members of the Beyond Budgeting Roundtable (BBRT) are individu- als and organizations who are interested in managing without budgets.

Membership in the BBRT is worldwide with the largest membership base in Europe.

doi:10.1016/j.mar.2009.10.003 T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7557 that budgeting systems often result in dysfunctional behav- ior and consume large amounts of management time. 2 Additionally, they often impede ﬁrms from being ﬂexible and adaptive in the increasingly unpredictable environ- ments facing contemporary organizations; and they are disconnected from strategy and thus out of sync with com- petitive requirements. 3 Of course, this is not the ﬁrst time budgeting practices have been criticized. However, there is a difference this time. In the past, criticisms of “traditional budgeting” were typically made by academics and were often exaggerations of “current worst practice” that had long been singled out for criticism. Improvements could be made or problems avoided (Horngren et al., 2004). On the other hand, the Beyond Budgeting message originates from practice and the BBRT does not believe the solution lies in improving traditional budgeting; instead, their view is that the budget should be eliminated as it is fundamentally ﬂawed (Hope and Fraser, 2003a,b).

Three points are noteworthy about this turn of events.

First, the overwhelming thrust of academic research into budgeting has been in the areas of participative budgeting and reliance on budgetary targets for performance evalua- tion (Hartmann, 2000). This research has arguably become disconnected from the concerns raised by practitioners (Hansen et al., 2003). The fact that such academic research has found mixed results (Kren and Liao, 1988; Shields and Young, 1993; Hartmann, 2000) and appears to have stalled suggests it may be useful to consider practitioner views to generate new research perspectives on budgeting.

Second, Hope and Fraser’s argument is presented as auniversalprescription. It seems difﬁcult to accept that so many organizations would continue to use budgeting for control purposes (i.e., for managerial motivation and performance evaluation) if it was fundamentally ﬂawed (Hannan and Freeman, 1989). Either previous research indicating extensive use of budgeting in practice (e.g., Umapathy, 1987) is no longer accurate and/or budgeting practices have evolved. However, other thanEpstein and Manzoni’s (2002)working paper, there is very little recent evidence regarding whether and how ﬁrms are adapting their budgeting systems. Further, thereareexamples in the literature in which highly successful ﬁrms utilize budget- ing extensively for both planning and control (Knight, 1992, Simons, 1987,Knight and Dyer, 2005). One can only won- der why these ﬁrms have been so successful and innovative if budgeting is inherently ﬂawed.

Third, too much emphasis may have been placed by Hope and Fraser (2003b)on the assertion that budgeting systems are inherently antithetical to successful adaptation in uncertain or unpredictable environments (seeHansen 2The dysfunctional effects of using budgetary targets for performance evaluation have been discussed extensively in the accounting literature.

Examples includeJensen (2001)on the problems with budget-based incentive contracts,Van der Stede (2000)on budgetary slack and manager short-termism, andMerchant (1990)on budgets and earnings manage- ment.

3SeeLibby and Lindsay (2003a,b)for an outline of the BBRT’s case against traditional budgeting and the suggested replacement manage- ment model. et al., 2003). For example, the BBRT’s exemplar case – Han- delsbanken (seeLindsay and Libby, 2007) – was not located in an industry (banking) that needed to adapt quickly or where revolutionary change was occurring. On the other hand,Simons’ (1987)Codman & Shurtleff case depicts a Johnson & Johnson subsidiary placing extensive reliance on budgeting in a highly innovative industry. Considerations other than environmental uncertainty or unpredictability would seem to be involved.

Taken together, these points suggest that we do not pos- sess a robust understanding of budgeting that is capable of explaining themechanisms or processesgiving rise to satisfactory or unsatisfactory consequences of budgeting systems. Yet, it is such knowledge that underlies the acqui- sition of deep theory (Bunge, 1997, 2003). The purpose of this paper is therefore to undertake a step forward in the development of such understanding. It presents the results of two largely descriptive surveys of budgeting practices in North-American companies for the purpose of determining whether any strong tendencies or patterns exist in bud- geting practices, manager beliefs about budgeting, and key outcomes associated with budgeting systems. Discovering such ‘facts’ or patterns is a necessary ﬁrst step because they often enable doubts to be raised concerning conventional wisdom (Mintzberg, 1979, 1983). In addition, their discov- ery provides a reliable basis for theorizing efforts aimed at understanding how and why those facts came to be (Fiske, 1986; Kaplan, 1998;Haig, 2005; Hambrick, 2007).

Within this overall aim, the paper has four speciﬁc objectives. First, we update the literature in terms of cur- rent budgeting practices in North America. The most recent extensive study of North-American budgeting practices was conducted over twenty years ago byUmapathy (1987).

Given the changes in the competitive environment since then, Umapathy’s results may no longer represent bud- geting practices utilized by contemporary North-American organizations. Second, we investigate budget managers’ overall perceptions, views and control uses of budgeting as well as their plans for budgeting in the future. Third, we examine the extent to which practitioners identify with some of the key criticisms that have been levied at budget- ing as summarized byHansen et al. (2003)and determine how ﬁrms might be using budgets in ways that may over- come them. Finally, in attempting to identify avenues for future research, we examine factors associated with assess- ments of overall budget value and other key outcomes.

In the next section we outline our research method, describe the sample and provide basic sample statistics.

Section3of the paper presents the survey results. Section 4follows with a discussion of the study’s major ﬁndings and conclusions.

2. Method 2.1. Sample selection and survey design We gathered data via a web-based survey of man- agers holding senior positions in medium to large-sized organizations. The sample came from the 2003 member- ship directory of CMA Canada and the 2004 membership directory of the Institute of Management Accountants 58T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–75 (IMA). To be included in the sample, individuals must have been employed in a for-proﬁt organization employing at least 100 people located in the United States or Canada (but outside of Quebec). 4In addition, we selected poten- tial respondents based on holding the positions of Vice President, Chief Financial Ofﬁcer, Controller, Director of Budgeting or Division Manager. We chose these criteria to ensure that the target group of organizations was large enough to have formal budgeting systems and so that man- agers we contacted would have considerable experience in establishing and using budgets. The ﬁnal target sample consisted of 2583 CMA Canada members and 13,712 IMA members.

We distributed the survey to CMA Canada members fol- lowingDillman’s (2000)recommendations. We contacted individuals ﬁrst via a pre-notice letter signed by the presi- dent of CMA Canada and sent by post. This letter described the survey’s objectives and encouraged members to partic- ipate. It also provided a web address for the survey and a password. The purpose of the password was to limit access to only those in the target group. This letter was followed by two dedicated email reminders sent by the CMA orga- nization at two-week intervals to all individuals included in the target sample.

We used a different method to distribute the IMA survey due to constraints imposed by the IMA. We ﬁrst contacted targeted respondents via a dedicated pre-notice email signed by the Executive Director of the IMA. This email described the survey objectives and encouraged members to participate. It also provided a web address for the survey and a password. This email was followed by one reminder email sent after two weeks to all individuals included in the sample. The reminder email was included as a component of the regular weekly emails that the IMA sends to all of its members.

The Canadian survey took approximately 30 min to complete. The welcome page explained the survey objec- tives, provided a link to deﬁnitions of all variables used in the survey and provided contact information for possible assistance. Respondents selected a user name and entered the password provided by the researchers to access the sur- vey. Responses could be saved, which allowed respondents to leave the survey as required and return to the section where they had left off.

We organized the survey into several sections and each section dealt with a speciﬁc theme. Once respondents com- pleted a section, they could not return to it. Where possible, we drew or adapted measures from prior studies. We also developed new measures as required. Ten individuals with a similar proﬁle to potential survey respondents pretested a preliminary version of the survey. Based on the pretest feedback, we shortened the survey, clariﬁed some ques- tions and changed the terminology to better reﬂect usage of some terms by practitioners. In addition, we tested the integrity of the process used to assign participants to one of the three different versions of the survey. 4Due to ﬁnancial and time constraints the survey was not translated into French making it unfeasible to survey CMA members located in Que- bec. The assignment process requires some explanation.

After answering a few questions at the start of the survey, the software assigned participants to one of two groups depending upon whether they reported using the budget for control purposes (i.e., for managerial motivation and performance evaluation) or not. If the respondent indicated that budgets were used for control in his/her organization, an automated algorithm built into the web site randomly assigned the respondent to one of two different versions of the survey. 5While containing some overlapping ques- tions, the two different versions contained version-speciﬁc questions so that we could collect a considerable amount of data concerning budgeting practices while limiting the time required for any one respondent to complete the sur- vey to no more than 30 min. Consequently, reported sample sizes will not always add to the total number of respon- dents.

We conducted the US survey several months after the Canadian survey. The US survey instrument was modiﬁed based on an analysis of the Canadian data. Modiﬁcations were made to eliminate non-discriminating questions, to improve some measures, and to add new questions based on what we learned from analyzing the Canadian data.

Appendix Alists, in the order that they are discussed in the paper, the survey questions and their related scales as well as any differences between the Canadian and US surveys.

2.2. Sample statistics Respondents submitted 558 surveys through the web- based system. CMA Canada members completed 346 of these surveys (response rate = 13.6%) while IMA mem- bers submitted 212 of these surveys (response rate = 1.5%).

Table 1shows that respondents had been employed on average 4.8 years in their current position and 7.9 years with their current company. Job titles spanned the list pro- vided with the most frequent titles being Controller (44.6%) and CFO (26.6%). Approximately 51% of responding orga- nizations were from the manufacturing sector and 49% were from the service sector. In addition, 46% of respond- ing organizations were stand-alone companies while 55% were divisions of larger organizations. These statistics were similar across the Canadian and US samples. However, dif- ferences existed with respect to size. The average Canadian respondent had divisional revenue in the $10–$50 mil- lion range while the average US respondent was larger with mean divisional revenues in the $50–$100 million range.

Although we can only speculate why the response rate was lower in the US sample, we believe the following dif- ferences in conducting the Canadian and US surveys may have played a role: (i) we used different methods to con- tact potential respondents in the pre-notiﬁcation stage (postal mail vs. email); (ii) we found it difﬁcult to spec- 5In total, 21% of respondents indicated that the budget was not used for control purposes. These respondents were assigned to a third version of the survey. This paper reports results only for respondents indicating budgets were used for control in their business units. T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7559 Table 1 Descriptive statistics for survey respondents. Canada US Combined sample Number of respondents 346 212 558 Mean years of employment:

In current position 4.4 yrs 5.3 yrs 4.8 yrs With current company 7.6 yrs 8.5 yrs 7.9 yrs Job titles (frequency):

Chief Financial Ofﬁcer (CFO) 30.0% 22.0% 26.6% Managing executive 5.2% 5.1% 5.2% Business controller 48.4% 39.6% 44.6% Director of budgeting 5.6% 13.8% 9.1% Other 10.8% 19.5% 14.5% 100.0% 100.0% 100.0% Economic sector (frequency):

Manufacturing 50.9% 50.6% 50.8% Service 49.1% 49.4% 49.2% 100.0% 100.0% 100.0% Corporate structure (STRUCTURE) (frequency):

Stand-alone unit 46.6% 44.1% 45.5% Division of a larger organization 53.4% 55.9% 54.5% 100.0% 100.0% 100.0% Business Unit Revenues (SIZE) (frequency):

Less than $1 million 5.4% 1.9% 3.8% $1–10 million 21.0% 12.6% 17.1% $10–50 million 34.4% 28.3% 31.6% $50–100 million 16.1% 13.6% 14.5% $100–500 million 14.5% 22.6% 18.3% $500 million–$1 billion 4.8% 7.5% 6.1% $1–5 billion 3.2% 10.1% 6.4% Greater than $5 billion 0.6% 4.3% 2.2% 100.0% 100.0% 100.0% Mean Business Unit Revenues $10–50 million $50–100 million ify as meaningful a target group of respondents in the US study as in the Canadian study since the categoriza- tions in the Canadian population databases were much more homogenous and fewer in number; and (iii) we sent a different number of follow-up reminders (two in the Canadian study vs. one in the US study) and we used dif- ferent types of reminders in each case (dedicated in the Canadian study vs. part of a routine communication in the US study). We ﬁnd this last difference to be particularly noteworthy. Unlike the Canadian survey where the bulk of responses occurredafterthe two follow-up reminders, the one reminder that occurred in the US survey did not generateanynew responses. It is important to note that the IMA members received reminder as part of the regu- lar weekly email sent to all members containing several items of news and announcements; consequently, respon- dents might have deleted the email without reading its contents, or the item might have been “lost” in the larger message.

We report the results separately for each country when- ever possible. Since there is no reasona priorito expect the results to differ, agreement across the two surveys can be taken as an important measure of their replicability and generalization across North-American ﬁrms. Nonethe- less, the possibility that a common non-response bias factor inﬂuencing both samples cannot be ruled out.In all cases, when comparable country data are not reported, it is because the speciﬁc data was collected in one country only.3. Results 3.1. North-American practitioners’ perceptions of budgeting 3.1.1. Should budgets be abandoned or improved?

In a recent review,Hansen et al. (2003)observe that the dissatisfaction with budgeting in practice is occurring on two fronts: those that wish to abandon budgeting alto- gether and those that wish to improve it. For example,Hope and Fraser (2003b)report several case studies of European companies that have successfully abandoned budgets for control purposes. On the other hand,Eckholm and Wallin (2000)report that only 15% of the Finnish companies they surveyed indicated they planned to abandon traditional budgeting whereas 61% aimed to improve the current bud- geting system and 24% reported they would continue to use their current budgeting system without changes. This divergence motivated us to investigate the situation that exists in North-American ﬁrms.

To examine this issue, we asked respondents to report whether they continue to use budgets for “control” pur- poses, where control was deﬁned in the survey as “the use of budgets for managerial motivation and as a stan- dard for performance evaluation.”Table 2shows that 277 of respondents (80%) in the Canadian sample and 163 of respondents (77%) in the US sample indicated that budgets were used for control purposes in their organiza- tions. This result is comparable toUmapathy (1987)who 60T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–75 Table 2 Use of budgets for control *. Canada US Total Freq % Freq % Freq % Are budgets used for control?

Yes 277 80% 163 77% 440 79% No 69 20% 49 23% 118 21% Total 346 100% 212 100 558 100% If use budgets for control:

Do you plan to abandon the use of budgets for control?

Yes 2 1% 2 1% 4 1% Possibly 18 6% 4 2% 22 5% No 257 93% 157 97% 414 94% Total 277 100% 163 100 440 100% *“Control” was deﬁned in the survey as “the use of budgets for manage- rial motivation and as a standard for performance evaluation purposes.” reported that 83% of his sample used budgets for control purposes.

Within the group using budgets for control, 94% indi- cated they were not planning to abandon the use of budgets for control in the near future while 5% indicated they were possibly considering doing so, and only 1% indicated that they were deﬁnitely planning to do so within the next two years. Results were similar across the Canadian and US samples.

Additionally, we asked the Canadian respondents within this group whether they planned to make changes to their budgeting systems over the next two years. We found that 46% of respondents planned to change or adapt their budgeting systems within the next two years. The most important reasons cited for making changes are con- sistent with those reported in other surveys of practice (e.g., Ekholm and Wallin, 2000;Neely et al., 2003):

• preparing budgets is time consuming and the beneﬁt may not be worth the cost; • the lack of ﬂexibility inherent in budgeting does not ﬁt well with a constantly changing environment; • budgets can be manipulated and provide incentives for the “wrong” (i.e., self-interested) behavior on the part of managers; • budgetary reporting is not meaningful to front-line employees; • budgeting eliminates the drive for constant improve- ment; and • the budget is not aligned with strategy.

We also asked these respondents to indicate the types of changes they might make in open-ended response boxes.

The following main categories of responses were obtained:

• incorporate a bottom-up orientation and gather more information from front-line managers; • use rolling forecasts; • better align strategic planning with budgeting; • prepare less detailed budgets initially and update them regularly using ongoing forecasts.

In conclusion, these results suggest that the traditional use of budgets for control purposes will not soon be elim-inated. As well, most ﬁrms in our sample planned to improve their budgeting systems, not abandon them. The US respondents provided further evidence on this issue by indicating their degree of agreement with the following statement: “The problems with budgeting are more to do with how they are used and some of the roles they are asked to play; budgets have the potential to be extremely useful if used appropriately.” The median level of agreement was 5 (agree) on a 6-point scale, with 88% of respondents agreeing or strongly agreeing with this statement (n= 78).

3.1.2. Are budgets value-added?

Notwithstanding recent criticisms, the vast majority of companies in this study reported that they will continue to use budgets for control purposes, presumably because the beneﬁts of doing so outweigh the costs. That is, they believe budgets and the budgeting process are value-added. In this subsection, we speciﬁcally examine this conjecture.

We asked respondents to evaluate the degree to which the overall budgeting system was value-added by assign- ing a “grade” between 0 and 100 to their budgeting system.

In assigning the grade, we asked respondents to consider “the amount of management time spent on the budget- ing process, as well as the budget system’s effectiveness in assisting the business unit to achieve its various objectives and any dysfunctional behavior that it may have caused.” 6 Fig. 1presents the results for the two samples.

In both samples, the median score was 70 out of 100 corresponding to the scale anchor of “good value.” Bud- geting systems were rated as providing positive value (i.e., score greater than 50) in approximately 90% of the ﬁrms across the two samples. Moreover, “good value” or greater was reported in 50 and 64% of the Canadian and US ﬁrms, respectively, and 25% reported at least “very good value.” These results would seem to be supported by the degree of agreement by American respondents to the question “Budgets are indispensable; we could not manage with- out them.” The average response on a six point Likert scale was 4.2 (“somewhat agree”). Fifty percent of respondents agreed or strongly agreed with the statement; only 15% disagreed or strongly disagreed.

In conclusion, it appears that the majority of ﬁrms in our sample are ﬁnding ways to obtain considerable value from their budgeting systems. This supports the earlier ﬁnding that the bulk of ﬁrms are not considering abandoning the use of budgeting for control purposes.

3.2. The criticisms of budgeting Hansen et al. (2003)summarized several criticisms of budgets and/or assumptions underlying the use of budgets identiﬁed from the academic and practitioner literatures.

The key criticisms were:

1. Budgeting consumes a lot of managerialtimewhich makes it a costly process and the beneﬁts may not be worth the cost.

6Respondents considered each of these issues in some depth in earlier sections of the survey before being asked the value-added question in an attempt to get them to fully reﬂect on their response. T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7561 Fig. 1.Are budgets value-added?Budget value measure: “Taking into account the management time spent on the budgeting process, as well as the budget, system’s effectiveness in assisting the business unit to achieve its various objectives and any, dysfunctional behavior it may or may not cause, whatoverall grade would you assign to the, budgeting system/process?” Grade (from 0 to 100).

2. Budgets inhibit ﬁrms fromadaptingto changes in a timely manner due to their ﬁxed nature.

3. The budgeting process is disconnected withstrategy thereby putting it out of kilter with the competitive demands facing ﬁrms.

4. The use of the budget as aﬁxed performance contract leads to unreliable performance evaluation and pro- motes budget gaming.

In this section, we examine the extent to which such criticisms are shared by respondents in our North- American sample of ﬁrms using budgets for control purposes. We also examine whether ﬁrms are taking steps to mitigate some of these criticisms.

3.2.1. Criticism 1: budgets take too much time to prepare We asked respondents to indicate how many weeks the annual formalized budgeting process takes to complete in their business unit (FIRM WEEKS).Table 3(Panel A) shows that the median response in the Canadian sample for FIRM WEEKS was 6 weeks (std. dev. = 4.7;n= 102). About 30% of ﬁrms took 4 weeks or less to prepare the budget while 31% took 9–16 weeks and about 5% took more than 16 weeks. In the US sample, median ﬁrm weeks to budget were some- what higher at about 10 weeks (std. dev. = 6.6;n= 77). 7 Survey respondents were also asked to indicate the amount of time the average manager spent on budget- related tasks in a year (e.g., developing the budget, revisions, reports, variance analysis).Table 3(Panel A) 7We further explored the discrepancy between FIRM WEEKS in the two samples by dividing ﬁrms into two size categories (small vs. large) using $50 million or less as “small” and greater than $50 million as “large.” The correlation between business unit size and ﬁrm weeks to budget isr= 0.17 in the Canadian sample (p< .06, one-tailed) andr= 0.29 in the US sample (p< .01, one-tailed) providing some evidence that differences in business unit size may be driving this result. shows the median amount of manager time spent on budgeting-related tasks (MANAGER TIME) was three to four weeks in both samples (std. dev. 1.4 in the Canadian sample,n= 235; std. dev. = 1.5 in the US sample,n= 80).

Assuming the average manager works 48 weeks per year, this represents approximately six to eight percent of the average manager’s time.

The time spent on budgeting may be driven by the extent of detail required and the number and level of people involved in preparing the budget. We describe this con- struct as “touch” in budgeting and measure it by asking respondents to indicate the extent of effort and involve- ment managers in their business unit devote to developing budgets (TOUCH). The scale was anchored from 1 (light) to 5 (heavy). 8Table 3(Panel A) shows that the median score in the Canadian sample was 4.0 (std. dev. = 1.4,n= 110), near the ‘heavy’ end of the scale. Results are similar in the US sample (median score = 4.0, std. dev. = 1.1,n= 83). A reason- ably heavy degree of touch in developing budgets appears to be the norm.

We correlated the degree of touch in budgeting (TOUCH) with the time managers spend on budgeting (MANAGER TIME) as well as the total number of weeks it takes the business unit to prepare the budget (FIRM WEEKS). As one might expect, the correlation was positive in the Cana- dian sample for MANAGER TIME (r= 0.19,p< .01,n= 104) and FIRM WEEKS (r= 0.39,p< .01,n= 104). 9The results are qualitatively similar in the US sample.

Overall, we ﬁnd that the total number of weeks taken to complete the budget and the time spent by the managers in preparing the budget are signiﬁcantly less for this sample 8We reverse coded the question appearing inAppendix Afor analysis purposes.

9All signiﬁcance tests reported in this paper are one-tailed unless oth- erwise noted. One-tailed tests are used whena priorireasons exist for specifying the direction of a relationship. 62T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–75 Table 3 Criticisms of budgeting *. Median Std. dev. Observed Min Observed MaxN Panel A: “Budgets take too much time to prepare” FIRM WEEKS Canadian sample (in weeks) 6.0 4.7 1 20 102 US sample (in weeks)10.0 6.6 2 36 77 MANAGER TIME Canadian sample (in weeks) 3–4 1.4 <1 > 16 235 US sample (in weeks)3–4 1.5 <1 >16 80 TOUCH in budgeting (1 = light to 5 = heavy) Canadian sample 4.0 1.4 1 5 110 US sample4.0 1.1 1 5 83 Median Std. dev. Observed Min Observed MaxN Panel B: “Budgets impede adaptability” PREDICTABILITY of external environment (1 = easy to predict; 3 = somewhat predictable; 6 = impossible to predict) Canadian sample 2.8 0.7 1.3 4.8 110 US sample3.0 0.6 1.5 5.0 78 RELIANCE on budget to adapt relative to other approaches (1 = no reliance; 3 = moderate reliance; 5 = exclusive reliance) Canadian sample 3.0 1.1 1.0 5.0 121 EFFECTIVENESS of budget in adapting to change (−3 = highly ineffective; 0 = neither effective nor ineffective; +3 = highly effective) Canadian sample 1.0 1.4−3.0 +3.0 98 Canadian sample Methods of Adaptation (frequency of response)Strongly disagree (1)Moderately Disagree (2)Some what disagree (3)Neutral (4)Some what agree (5)Agree (6)Strongly agree (7)N Panel B – continued – “Budgets impede adaptability” It is somewhat easy to obtain new resources outside of the budgeting process to deal with market changes9.3% 14.8% 11.0% 6.5% 18.5% 22.2% 17.6% 108 “Fast-track” approval processes exist to ensure resources are available on a timely basis for important initiatives not incorporated in the approved budget.8.3% 6.4% 11.0% 10.1% 26.6% 21.1% 16.5% 109 Method of updating operating budget Canadian sample (N= 109) US sample (N= 87) Budgets are ﬁxed. No changes made to them 44.0% 51.2% Budgets are revised during the year 56.0% 48.8% a. At next formal budget review 47.6% 27.5% b. On an ad-hoc basis 32.7% 52.5% c. When next rolling budget prepared 19.7% 20.0% Median Std. dev. Observed Min ObservedMaxN Panel C: “Budgets are rarely linked to strategy” LINKAGE (The budget is linked to strategic objectives) Canadian sample 1 = Strongly disagree to 7 = Strongly agree (includes a neutral point of 4 = neither agree nor disagree)5.0 1.6 1.0 7.0 133 US sample 1=Strongly disagree to 6=Strongly agree (respondents forced to take a position as no neutral point included in this scale)5.0 1.1 2.0 6.0 81 Budget is linked to strategy implementation Canadian sample 1 = Strongly disagree to 7 = Strongly agree (includes a neutral point of 4 = neither agree nor disagree)5.7 1.4 1.0 7.0 133 US sample 1=Strongly disagree to 6=Strongly agree (respondents forced to take a position as no neutral point included in this scale)5.0 0.9 1.0 6.0 83 Budget emphasisCanadian sample (N= 109) US sample (N= 87) Panel D: “Use of ﬁxed performance contract” Budget emphasis low48.3% 29.8% Budget emphasis high51.7% 70.2% Performance evaluation in the high-budget-emphasis group:

Actual ﬁnancial performance rigidly compared to budget 12.2% 16.9% Budget target adjusted subjectively at the end of the year 33.8% 23.7% Both budget target and other subjective factors used 23.0% 20.3% Budget target adjusted for uncontrollable budget variances 25.7% 32.2% Budget target adjusted at year-end using formula established at beginning of the year 5.4% 6.8% *Appendix Aincludes the full measures for all of these scale items. T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7563 of ﬁrms than those reported byHope and Fraser (2003b) (i.e., 12–20 weeks and 20–30% of managers’ time) or by Umapathy (1987)(i.e., 21–40% of managements’ time).

Increased time for both samples is associated with the size of the business unit and the level of “touch” in preparing the budget. This begs the question whether “touch” adds organizational value. We explore this issue further at the end of this section.

3.2.2. Criticism 2: budgets impede adaptability Hope and Fraser (2003a)assert that the new com- petitive environment is characterized by unpredictability; prices and margins are constantly under pressure, product life cycles are shorter and customer tastes are ﬁckle. This situation leads to the importance of becoming adaptable and ﬂexible. They argue that budgeting is antagonistic to these requirements because, once set, budgets are not typ- ically changed resulting in plans and targets that quickly become out of date.

To explore this assertion, we asked respondents how easy it is when setting the budget to anticipate or predict changes in the external environment for the budget period.

We adapted this measure fromGovindarajan (1984)and Umapathy (1987). The scale ranged from 1 (easy to predict) to 6 (impossible to predict).Table 3(Panel B) shows that the median response for this scale, called PREDICTABILITY, was 2.8 in the Canadian sample (std. dev. = 0.7;n= 110) and 3.0 in the US sample (std. dev. = 0.6;n= 78). These scores fall at the “somewhat” predictable anchor on the scale.

We explored this matter further in the US survey by ask- ing respondents to indicate their degree of agreement with the statement “It is difﬁcult to set accurate budgets because of the unpredictability of factors inﬂuencing the busi- ness.” Fifty-four percent of respondents at least ‘somewhat agreed’ with this statement and 24% “agreed” or “strongly agreed” with it. The correlation between responses to this statement and PREDICTABILITY was .34 (p< 0.01). Second, we asked whether “Budgets quickly become obsolete or outdated as the year goes by.” Sixty-ﬁve percent of respon- dents at least “somewhat agreed” with this statement and 40% “agreed” or “strongly agreed” with it. The correlation between responses to this statement and PREDICTABIL- ITY was .23 (p< .05). These ﬁndings suggest thatHope and Fraser’s (2003)assumption that the business environment is unpredictable, resulting in budgets quickly becoming out of date, seems valid for a signiﬁcant number of ﬁrms, but it would be a mistake to generalize it to the majority.

Next, we asked respondents in the Canadian sample to indicate the degree to which they rely on the budget to adapt to market changes relative to other measures. Pre- sumably, if the budget is ill-suited to this task we should not ﬁnd a lot of reliance on it. The RELIANCE scale ranged from 1 (no reliance) through 5 (exclusive reliance).Table 3(Panel B) shows the median score for RELIANCE is 3 (std. dev. = 1.1; n= 121) indicating a “moderate” degree of reliance is placed on the budget for adapting to market changes relative to other methods. For the 81% of Canadian respondents placing at least some reliance on the budget for adapt- ing to market changes, i.e., those selecting a score of 2 (some reliance) or higher, we also examined the degree to which the budget was perceived as effective in achievingadaptation to market changes. The EFFECTIVENESS scale ranged from−3 (highly ineffective) through +3 (highly effective). We found a median effectiveness rating of +1 (std. dev. = 1.4;n= 98) reﬂecting that the budget is per- ceived to be “somewhat effective.” This rating does not reﬂect an overwhelming positive endorsement but neither does it suggest that the budget is totally ill-suited to the task of facilitating adaptation.

Taken together, these results imply that while the majority of respondents utilize budgets to help them adapt to market changes, there is recognition that budgets are only somewhat effective in this role, perhaps because they become quickly out of date for a non-trivial number of ﬁrms; consequently, ﬁrms need to augment the budget process with other approaches.Table 3(Panel B) presents the results of further questions examining this issue that were identiﬁed in the literature. We examined whether Canadian respondent ﬁrms make it easy to obtain new resources outside of the budgeting process to deal with unforeseen opportunities designed to accomplish strate- gic initiatives (58% report at least “somewhat agree”) 10 and utilize fast track processes to do so (64% report at least “somewhat agree”). In addition, 56% of the Canadian sample and 49% of the US sample allow some type of bud- get revision during the year. In Canada, the most common method of revising the budget was at the next formal bud- get review (48%) which typically occurred at the end of the month or the end of the next quarter. In the US sam- ple, adjustments occur most frequently on an ad-hoc basis (52.5%).Finally, approximately 20% of respondents in both samples report using rolling budgets-the approach recom- mended by theHope and Fraser (2003b). This ﬁnding is similar toUmapathy’s (1987, 85)result of 21%.

In conclusion,Hope and Fraser’s (2003b)assumption that ﬁrms’ competitive environment is characterized by unpredictability is valid for many ﬁrms in our sample, although the claim is over-generalized. Additionally, our results lend support to their assertion that the budget- ing process is potentially weak in helping ﬁrms deal with adapting to change; however, ﬁrms appear to adjust budget targets in various ways to mitigate this concern and adopt processes to obtain new resources outside of the budget process when necessary.

3.2.3. Criticism 3: budgets are disconnected from ﬁrm strategy Kaplan and Norton (2001)observe that the majority of ﬁrms they have worked with fail to link their budget- ing systems to achieving strategic objectives.Hope and Fraser (2003b)also share the view that budgets are typ- ically prepared in isolation from, and not aligned with strategy. To investigate this issue, we asked survey respon- dents whether “The budgeting process is explicitly linked to achieving strategic objectives/targets.” We designate this variable LINKAGE.Table 3(Panel C) shows the median score for LINKAGE was 5 (somewhat agree) in the Cana- dian sample (std. dev. = 1.6;n= 133). Untabulated results 10We reverse coded this question for the purpose of interpreting the results. 64T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–75 suggest that approximately 11% of the sample disagrees that the budget process is explicitly linked to achieving strategic objectives (a rating of 1 or 2 on the LINKAGE scale) while 48% agree (a rating of 6 or 7 on the LINKAGE scale). We observed qualitatively similar results in the US sample.

Second, we asked respondents to indicate their degree of agreement with three statements drawn fromKaplan and Norton (2001)concerning the degree to which budgets are linked to strategy implementation within their busi- ness unit. These statements consisted of: (i) “Setting the budget causes us to talk about and reﬂect upon our strat- egy”; (ii) “We sometimes change our strategy/tactics based on the feedback derived from going through the budgeting process”; and (iii) “Within the budget process, managers are expected to identify tactical initiatives to close the gap between current performance and the desired level of per- formance.” We averaged respondents’ agreement scores on these three items to create a measure of the degree to which budgets were linked to strategy implementation.Table 3 (Panel C) shows the median score on this measure was 5.7 (std. dev. = 1.4;n= 133) in the Canadian sample. This median score falls between the anchors “somewhat agree” and “agree” using a seven point scale. Untabulated results indicate that approximately 67% of the Canadian sample reported at least “somewhat agreeing” that concrete steps were being taken to link the budgeting process to strategy implementation. 11 We observed similar results in the US sample.

Third, we examined the importance of the budgetrela- tive to other meansfor implementing business unit strategy.

We asked respondents to allocate 100 points across the fol- lowing means that could be used to implement strategy:

budgeting process, performance measurement and evalua- tion system, hierarchical authority structure, business unit culture and boundary systems. 12 Fig. 2indicates that organizations in both countries placed the greatest degree of reliance on the budgeting process, the performance measurement and evaluation system, and the hierarchical authority structure in imple- menting strategy (between 22 and 25 points out of 100).

US companies rely on culture next (18 points) followed by boundary systems (12 points), whereas companies in Canada rely on these two mechanisms to a fairly equal amount (12 and 15 points, respectively). These results suggest that companies utilize a number of mechanisms for implementing strategy, all of which are important, although greater weight is placed on a subset of them, including the budgeting process. The importance attached to the budget in implementing strategy is reﬂected by respondents’ perception of its usefulness. The budgeting process was rated by 64% of Canadian respondents as a 11We performed a validity check on this result by correlating LINKAGE and the degree to which budgets were linked with strategy implemen- tation. The correlations werer= .64 (p< .001,n= 133) andr= .64 (p< .001, n= 83) for the Canadian and US samples.

12The survey stipulated that boundary systems “give people the freedom to act within clearly delineated bounds as established by codes of conduct, mission statements, ‘stop-doing’ lists, and the communication of strategic themes” (Simons, 1995). Fig. 2.How Strategy is implemented in responding organizations.Strat- egy implementation measure: “Please allocate a total of 100 points across the following elements of management control to reﬂect their relative importance in implementing strategy in your business unit”.

somewhat effective (46.3%) or effective (17.7%) method for implementing strategy.

In conclusion, these results indicate that that the criti- cism that budgets are not linked to strategy is unfounded for the majority of ﬁrms in our two samples. The budget- ing process is used in many ﬁrms to promote strategically focused behavior and is recognized as being an important mechanism for doing so.

3.2.4. Criticism 4: the use of budgets as ﬁxed performance contracts Hope and Fraser’s (2003b)most strident criticism is that budgets often serve as a “ﬁxed performance contract.” Implicitly or explicitly, the nature of this contract is that if actual performance meets or exceeds a pre-speciﬁed budget target, performance will be deemed satisfactory (or better) and this will likely result in rewards.Hope and Fraser (2003a,b)argue that a ﬁxed target represents a poor standard for performance evaluation when fac- tors underlying the budget may have changed during the budget period. Further, this inevitably leads to budgetary gaming by subordinates to increase the probability of receiving positive performance evaluations and associated pay increases (see alsoWelch, 2005)—a processJensen (2003)refers to as “paying people to lie.” We undertook a number of analyses to investigate this criticism.

Performance evaluation. We asked respondents to rate the degree of emphasis placed on meeting budget targets in the performance evaluation process using a scale adapted fromVan der Stede (2001).Table 3(Panel D) shows that 52% of Canadian respondents and 71% of US respondents indicate that budget emphasis is high. 13 For this group of respondents, we examined the prevalence of the use of ﬁxed performance contracts. Our results indicate that only 12% of respondents in the Canadian high-budget-emphasis group and 17% of respondents in the US high-budget- emphasis group report that actual ﬁnancial performance is rigidly compared against the pre-established budget target with no allowance for changes occurring in the competitive environment during the year. 13Appendix Alists the items making up the budget emphasis scale. T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7565 This analysis suggests that the ﬁxed performance con- tract ismuchless prevalent than what the BBRT would have us believe. If one combines these two ﬁndings with the earlier result that only 80% of Canadian respondents and 77% of American respondents use budgets for control pur- poses (seeTable 2), then one can extrapolate that only 5% of Canadian ﬁrms (.80 * .52 * .12) and approximately 9% of American ﬁrms (.77 * .71 * .17) use a ﬁxed performance contract.

Table 3(Panel D) shows 33.8% of ﬁrms in the Canadian high-budget-emphasis group and 23.7% of ﬁrms in the US high-budget-emphasis group adjust the budget target sub- jectively at the end of the year to account for unexpected changes in the external environment. In addition, 23.0% of ﬁrms in the Canadian high-budget-emphasis group and 20.3% of ﬁrms in the US high-budget-emphasis group use both the budget target and other subjective factors to eval- uate performance at the end of the year. The remaining respondents indicate they either adjust the budget tar- get for the amount of any uncontrollable budget variances determined at the end of the year (25.7% of the Canadian high-budget-emphasis group and 32.2% in the US high- budget-emphasis group) or adjust the budget target at the end of the year for unexpected events using a for- mula established at the beginning of the year (5.4% of the Canadian high-budget-emphasis group and 6.8% of the US high-budget-emphasis group).

In conclusion, the evidence does not support the BBRT’s key assumption of the prevalence of the ﬁxed performance contract in practice. In fact, it is remarkable how few ﬁrms use the budget in this way for performance evaluation.

Instead, we ﬁnd that subjectivity in performance evalua- tion or making allowances for non-controllable events is relatively widespread.

Budget gaming. The analysis now turns to determin- ing the extent to which certain budgetary games occur in our sample organizations. The list of gaming behaviors shown inAppendix Awere drawn fromMerchant (1985), Umapathy (1987), andBart (1988). We asked respondents to use the previous two years as a point of reference and to indicate how often each of the identiﬁed gaming behaviors occurred in their business unit.

In Canada (Fig. 3, Panel A), the most frequently occur- ring games were deferring necessary expenditures to future periods (80% report that this occurs “occasion- ally” or “frequently”) and negotiating easier targets by “sandbagging” (77% indicate that this occurs “occasion- ally” or “frequently”). The other listed games occurred, but when they did it was more often “occasionally” than “frequently” and over 50% of the ﬁrms reported that they “never occurred.” In the US (Fig. 3, Panel B), a greater proportion of respon- dents report that gaming behaviors occur and that these behaviors occur more often than in the Canadian sample.

A Mann–Whitney test ﬁnds this difference to be statisti- cally signiﬁcant (p< .001,n= 211). 14Similar to the Canadian sample, deferring necessary expenditures (91% indicate 14SeeAppendix Afor how the GAMING score for each company was calculated. “occasionally” or “frequently”) and sandbagging (86% indi- cate “occasionally” or “frequently”) were the two most prevalent gaming activities. The other three gaming activi- ties all occur (either occasionally or frequently) more often than they do not occur.

These results indicate that budgetary gaming is preva- lent, consistent with the writings ofBart (1988),Jensen (2001)andHope and Fraser (2003a,b). In fact, only 5 and 1% of the respondents in the Canadian and US samples, respectively, reported no incidences across the ﬁve gam- ing activities examined. However, what is not so well known is whether such behaviors actually impair long-run organizational performance. To examine this issue, we cor- related the degree of budget GAMING with its perceived negative effect on long-run performance of the business unit (NEGLRP). The observed correlation isr= .28 (p< .001; n= 133) in Canada andr= .56 (p< .001;n= 78) for the US respondents. These results suggest that increases in gam- ing are perceived to negatively impact long-run business unit performance.

3.3. Factors and outcomes associated with perceptions of budget value We now undertake some additional analyses to better understand what factors and outcomes might be associ- ated with perceptions of increased budget value. We begin by correlating overall BUDGET VALUE with PERFORMANCE measured as the mean rating over several items on a scale of−2 (considerably below direct competitors) through +2 (considerably above direct competitors). The speciﬁc per- formance items are listed inAppendix A.Table 4indicates that the correlation between BUDGET VALUE and PERFOR- MANCE is positive and signiﬁcant in both the Canadian (r= 0.16,p< 0.01) and US (r= 0.36,p< 0.001) samples. This result is consistent with the intuitive notion that a good budgeting system can lead to higher levels of ﬁrm perfor- mance, but it needs to be recognized that the design of this study does not permit statements of causation to be made.

Following this, we examined the association of several contextual factors (size, strategy, structure and predictabil- ity of the environment) identiﬁed byUmapathy (1987) with BUDGET VALUE (seeTable 4). We ﬁnd that neither the SIZE of the business unit (as measured by business unit revenues) nor the divisional STRATEGY (cost leader vs.

differentiator [Porter, 1980]) were signiﬁcantly correlated with BUDGET VALUE in either the Canadian or US samples.

In the US (but not the Canadian) sample, the business unit STRUCTURE (stand-alone unit or division of a larger orga- nization) was signiﬁcantly correlated with BUDGET VALUE (r=−0.16,p< 0.05, two-tailed) indicating that stand-alone business units derived greater value from the budget than did business units that were divisions of larger organi- zations. Finally, the correlation between PREDICTABILITY and BUDGET VALUE was negative in both samples. How- ever, only the US sample result is statistically signiﬁcant (r=−0.27,p< 0.05, two-tailed).

We also considered whether differences in perceived BUDGET VALUE might be associated with the way bud- gets are developed and used within ﬁrms. The ﬁrst analysis examined the correlation between TOUCH and BUDGET 66T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–75 Fig. 3.Budgetary gaming.Gaming measure: “Using the last two years as a point of reference, how often do you think the, following practices have occurred in your business unit?” Table 4 Factors associated with budget value. Canadian sample US Sample Correlation with BUDGET VALUEpvaluenCorrelation with BUDGET VALUEpvaluen PERFORMANCE 0.16p< 0.01 * 211 0.36p< .001 * 162 SIZE 0.03 ns 205−0.03 ns 159 STRATEGY−0.08 ns 217−0.08 ns 158 STRUCTURE−0.06 ns 215−0.16p< 0.05 ** 161 PREDICTABILITY−0.10 ns 104−0.27p< 0.05 ** 78 TOUCH 0.21p< 0.05 ** 104 0.20p< 0.08 ** 82 Budget linked to strategy implementation 0.43p< 0.001 * 106 0.60p< 0.001 * 83 GAMING−0.22p< 0.01 * 133−0.33p< 0.002 * 78 Variables were deﬁned as follows (seeAppendix Afor questions and scales): BUDGET VALUE measured as 0 = disaster through 100 = outstanding value, PERFORMANCE measured as−2 = considerably below most direct competitors through +2 = considerably above most direct competitors, SIZE measured as 1 = Business unit revenues less than $1 million through 8 = Business unit revenues greater than $5 billion, STRATEGY measured as 0 = cost leader and 1 = differentiator, STRUCTURE measured as 0 = stand-alone unit and 1 = division of a larger organization, PREDICTABILITY measured as 1 = easy to predict through 6 = impossible to predict, TOUCH measured as 1 = light through 5 = heavy. Budget linked to strategy implementation measured on a scale of 1 = strongly disagree through 7 = strongly agree (6 point scale used in US sample as neutral point was removed), GAMING measured on a scale of 0 = never to 2 = frequently.

*One-tailed.**Two-tailed. T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7567 VALUE. A positive, signiﬁcant correlation was found in both samples (r= 0.21,p< 0.05, two-tailed in the Canadian sample andr= 0.20,p< .08, two-tailed in the US sample), suggesting that involvement by many levels of managers and detailed budgeting is perceived to be value-added.

Next, we observed large, positive correlations in both samples between the degree to which budgets are linked to strategy implementation and BUDGET VALUE (r= 0.43, p< .001 in the Canadian sample;r= 0.60,p< .001 in the US sample). These results provide strong support forKaplan and Norton’s (2001)recommendation that the budgeting process should be linked to strategy implementation.

Finally, budget GAMING was correlated with BUDGET VALUE. The correlation was negative and signiﬁcant in both the Canadian (r=−0.22,p< .01) and US (r=−0.33,p< .01) samples. This result suggests that ﬁrms need to prevent budgetary gaming in order to reap a higher level of value from their budgeting system. Combined with the earlier ﬁnding of the negative impact of gaming on long-run ﬁrm performance, our analysis suggests that ﬁrms need to treat the gaming issue seriously.

4. Discussion and conclusion It is undeniable thatHope and Fraser (2003b)have made a valuable contribution to the discipline. They have pro- vided a cogent and insightful analysis documenting the weaknesses of budgeting and have caused us to think deeply about how to improve the entire management pro- cess. However, our ﬁndings suggest that budgeting systems continue to play a key role in ﬁrms’ control systems and that most companies have no plans to abandon this prac- tice, although many are planning to take steps to improve their budgeting systems to overcome some of the common criticisms. Further, given these ﬁndings (and in further sup- port of them), it is not surprising that the average ﬁrm in our sample rates its budgeting system as providing good or better value. That said, 18 and 13% of Canadian and US ﬁrms, respectively, report little or no value or negative value from their budgeting systems (a score below 60), and approxi- mately 20% of responding ﬁrms in both countries do not use budgets for control. Thus therearesome ﬁrms that may be receptive to the Beyond Budgeting message.

Moreover, based on our ﬁndings, it would appear that many (but not all) of the assumptions and criticisms under- lying Hope and Fraser’s argument are over-generalized in terms of their applicability to the average ﬁrm. In particular, this study found that:

• Use of the ﬁxed performance contract ismuchless preva- lent than what is suggested by the BBRT or the Reliance on Accounting Performance Measures (RAPM) literature.

In fact, our data suggests that it is only used in a very small percentage of ﬁrms (5 and 9% of ﬁrms in Canada and the US, respectively).

• Subjective considerations or allowances for non- controllable events are frequently observed in ﬁrms using the budget to evaluate performance. Superiors would appear to be more sophisticated than what is acknowledged in textbooks and some prior academic work.• Time spent on budgeting in the average sampled North- American ﬁrm is considerably less than what critics suggest and does not appear excessive.

• The majority of sample ﬁrms do not operate in unpre- dictable environments to the point where budgets become quickly outdated, although a good number of ﬁrms do face unpredictable environments. How- ever, many ﬁrms utilize adaptive processes to mitigate this concern (e.g., fast track processes to obtain new resources). As well, budgets are revised much more often than expected. This may explain the result that such a small percentage of ﬁrms sampled in this study rigidly evaluate a manager’s performance against the ﬁxed bud- get.

• In the majority of ﬁrms surveyed, the budget process is explicitly linked to strategy implementation. In fact, budgeting is reported by respondents to be an impor- tant means for implementing strategy and the majority of respondents report that it plays a useful role in doing so.

• Finally, budgetary gaming behaviors are a problem in our sample ﬁrms in both countries, although it appears more pronounced in the sampled US ﬁrms.

In attempting to move the budgeting research agenda forward, we believe that one should not take an “either/or” focus (i.e., Beyond Budgeting vs. traditional). There are examples of very successful companies utilizing both approaches. Instead, it would appear to be more fruitful to develop the possibilities for each model by seeking deeper understanding as to the mechanisms and processes which underlie the application of each model in highly success- ful companies. Given this, future research in this area is unlikely to be facilitated by the use of representative sam- ples. Instead, in-depth study of a set of carefully chosen companies focusing on key issues will be required. It will also require examining the total package of controls in use (Fisher, 1995; Otley, 2001; Chenhall, 2003; Merchant and Van der Stede, 2006).

The types of questions to consider are many and var- ied. For example,Hope and Fraser (2003b)outline the principles underlying highly successful Beyond Budgeting companies. A similar activity needs to be performed within highly successful ‘traditional’ budgeting companies in line with respondents’ views in this study that the problem with budgeting lies in how it is used rather than any inher- ent ﬂaws. To delve into this point further, we consider the example of Johnson & Johnson described in Simons’ (1987) Codman & Shurtleff case. J&J makes extensive use of bud- gets (budgetsarethe heart of their control system) in an extremely unpredictable business environment and places considerable emphasis on meeting or exceeding budget targets. However, their control system utilizes a number of elements that would seem to mitigate the concerns of utilizing a high-budget-emphasis style in an unpredictable environment (seeGovindarajan, 1984; Hirst, 1987; Hope and Fraser, 2003b). In particular, they use:

• A contingency fund to help deal with uncertainty; • Subjective evaluations and a strong HR system; 68T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–75 • Highly detailed budgets across responsibility centers and the involvement of lower to senior levels of management (our TOUCH variable); • A strong culture for managing for the long term; • A long-term planning system that is strategically ori- ented; • Operational (budgetary) plans that are linked to the long- term (strategic) plan; • Multiple revisions (mainly of tactics); • A budget system that is managed interactively, not diag- nostically (Simons, 1995); • A culture of information sharing; and • A strongly decentralized management structure.

Other questions include examining whether other high performing ‘traditional’ budgeting ﬁrms utilize similar ele- ments as Johnson & Johnson (i.e., determining key elements within the traditional model)? Does an unpredictable envi- ronment impact the use of some of these elements (i.e., contrasting successful companies operating in predictable and unpredictable environments)? What speciﬁc practices underlie the ability for successful ﬁrms to innovate in highly unpredictable environments? Also, are there any patterns to thekindsof successful ﬁrms that adopt the Beyond Budgeting approach? What explanations under- lie such patterns? Lastly, can successful ﬁrms mitigate or eliminate the use of budgetary gaming. If so, how is this achieved?

Another suggestion focuses on examining whether there are consistent factorsacrossthe two models. For example, Handelsbanken, the exemplar case in Beyond Budgeting (Lindsay and Libby, 2007) shares the follow- ing characteristics with Johnson & Johnson, a company exhibiting extensive use of budgeting for control: a strongly decentralized management system; a strong cul- ture, including a culture of information sharing, subjective evaluations and the use of interactive control systems.In this regard, it is important to recognize that the Han- delsbanken story is much more about promoting and controlling for radically decentralized employees than it is about managing without budgets.

Perhaps most importantly, the ﬁnding of the frequent use of subjective considerations and allowances for non- controllable events within a budget paradigm would seem to cast doubt on the validity of the RAPM paradigm that has dominated management accounting research (Hartmann, 2000). At a minimum, there is a need to better understand how subjective considerations and other allowances are used in performance evaluation.

Finally, the results of this study must be interpreted in light of its limitations. The paper’s low response rates and method of company selection makes generalizing these results to all North-American ﬁrms problematic. However, some comfort can be taken in the fact that results were remarkably consistent between the two samples, provid- ing a strong indication of their replicability. Nonetheless, a common bias impacting both samples cannot be ruled out.

Also, the standard limitation that surveys can be affected by unreliable responses or that the measures utilized may not be a valid surrogate for the constructs of interest needs to be acknowledged. However, the severity of this threat is weakened in the present study because many of the questions are low level constructs (e.g., how long does budgeting take in your ﬁrm?). Further, we pre-tested the surveys, performed numerous validity checks, obtained reasonably consistent results across the two samples and within samples, and respondents could not go back and change their answers once a section was completed.

Notwithstanding these limitations, we believe this study has made an initial contribution towards further developing our understanding of what remains an impor- tant and widespread organizational practice – budgeting – and for setting direction for future research in this area. T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7569 Appendix A. Survey questions 70T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–75 T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7571 72T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–75 T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7573 74T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–75 T. Libby, R.M. Lindsay / Management Accounting Research 21 (2010) 56–7575 References Bart, C., 1988. Budgeting gamesmanship. Academy of Management Exec- utive, 285–294.

Bunge, M., 1997. Mechanism and explanation. Philosophy of the Social Sciences 27 (4), 410–465.

Bunge, M., 2003. How does it work? The search for explanatory mecha- nisms. Philosophy of the Social Sciences 34 (2), 182–210.

Chenhall, R.H., 2003. Management control systems design within its orga- nizational context: ﬁndings from contingency-based research and directions for the future. Accounting, Organizations and Society 28, 127–168.

Dillman, D.A., 2000. Mail and Internet Surveys: The Tailored Design Method. Wiley, New York.

Eckholm, B.G., Wallin, J., 2000. Is the annual budget really dead? The Euro- pean Accounting Review 9 (4), 519–539.

Epstein, M.J., Manzoni, J.F., 2002. Reconciling conﬂicting roles of budgets:

review and survey of corporate practices. Working paper, Rice Uni- versity and INSEAD.

Fisher, J., 1995. Contingency-based research on management control sys- tems: categorization by level of complexity. Journal of Accounting Literature 14, 24–53.

Fiske, D.W., 1986. Speciﬁcity of method and knowledge in social science.

In: Fiske, D.W., Shweder, R.A. (Eds.), Metatheory in Social Science.

University of Chicago Press, Chicago, pp. 61–82.

Govindarajan, V., 1984. Appropriateness of accounting data in per- formance evaluation: an empirical examination of environmental uncertainty as an intervening variable. Accounting, Organizations and Society 9 (2), 125–135.

Gurton, A., 1999. Bye-bye budget. Accountancy International (March).

Haig, B.D., 2005. The abductive theory of scientiﬁc method. Psychological Methods 10 (4), 371–388.

Hambrick, D.C., 2007. The ﬁeld of management’s devotion to theory:

too much of a good thing? Academy of Management Journal 50 (6), 1346–1352.

Hannan, M.T., Freeman, J., 1989. Organizational Ecology. Harvard Univer- sity Press, Boston.

Hansen, S.C., Otley, D.T., Van der Stede, W.A., 2003. Practice developments in budgeting: an overview and research perspective. Journal of Man- agement Accounting Research 15, 95–116.

Hartmann, F., 2000. The appropriateness of RAPM: toward the further development of theory. Accounting, Organizations and Society 25, 451–482.

Hirst, M.K., 1987. The effects of setting budget goals and task uncertainty on performance: a theoretical analysis. The Accounting Review 62 (4), 774–784.

Hope, J., Fraser, R., 2003a. Who needs budgets? Harvard Business Review 81 (2), 108–115.

Hope, J., Fraser, R., 2003b. Beyond Budgeting: How Managers Can Break Free from the Annual Performance Trap. Harvard Business School Press, Boston.

Horngren, C.T., Stratton, G.L., Sutton, W.O., Teall, H.D., 2004. Management Accounting, 4th ed. Prentice Hall, Toronto.

Jensen, M.C., 2003. Paying people to lie: the truth about the budgeting process. European Financial Management 9 (3), 379–406.

Jensen, M.C., 2001. Corporate budgeting is broken—let’s ﬁx it. In: Harvard Business Review, Vol.79, No. 10, pp. 94–101.Kaplan, R.S., Norton, D., 2001. The Strategy-Focused Organization. Harvard Business School Press, Cambridge, MA.

Knight, C.F. 1992. Emerson Electric: Consistent Proﬁts, Consistently. Har- vard Business Review, 57:70.

Knight, C.F., Dyer, D., 2005. Performance without compromise. Harvard Business School Press, Boston, MA.

Kren, L., Liao, W.M., 1988. The role of accounting information in the con- trol of organizations: a review of the evidence. Journal of Accounting Literature 7, 280–309.

Libby, T., Lindsay, R.M., 2003a. Budgeting—the unnecessary evil? CMA Management Magazine (March), 30–33.

Libby, T., Lindsay, R.M., 2003b. Booting the budget: how the BBRT envi- sions a world without budgets. CMA Management Magazine (April), 28–31.

Lindsay, R.M., Libby, T., 2007. Svenska Handelsbanken—controlling a radically decentralized organization without budgets. Issues in Accounting Education 22 (4), 53–71.

Merchant, K., 1990. The effects of ﬁnancial controls on data manipula- tion and management myopia. Accounting, Organizations and Society, 297–313.

Merchant, K., 1985. Budgeting and the propensity to create budgetary slack. Accounting, Organizations and Society 10, 201–210.

Merchant, K., Van der Stede, W., 2006. Field based research in accounting:

accomplishment and prospects. Behavioral Research in Accounting 18, 117–134.

Mintzberg, H. 1979. The Structuring of Organizations. Prentice Hall.

Mintzberg, H. 1983. Power in and Around Organizations. Prentice Hall.

Neely, A., Bourne, M., Adams, C., 2003. Better budgeting or beyond bud- geting? Measuring Business Excellence 7 (3), 22.28.

Otley, D.T., 1994. Management control in contemporary organizations:

towards a wider framework. Management Accounting Research 5, 289–299.

Otley, D., 2001. Extending the boundaries of management accounting research: developing systems for performance management. British Accounting Review 33, 243–261.

Porter, M., 1980. Competitive strategy: techniques for analyzing industries and competitors. Free Press, New York.

Shields, M.D., Young, S.M., 1993. Antecedents and consequences of participative budgeting: evidence on the effects of asymmetri- cal information. Journal of Management Accounting Research 5, 265–280.

Simons, R., 1987. Codman & Shurtleff, Inc. Planning and Control System.

Case study # 9-187-081. Harvard Business School, Boston, MA.

Simons, R., 1995. Control in the age of empowerment. Harvard Business Review 73 (2), 80–86.

Umapathy, S., 1987. Current Budgeting Practices in US Industry. Quorum Books, New York.

Van der Stede, W.A., 2000. The relationship between two consequences of budgetary controls: budgetary slack creation and managerial short-term orientation. Accounting Organizations and Society 25 (6), 609–622.

Van der Stede, W.A., 2001. Measuring ‘tight budgetary control’. Manage- ment Accounting Research 12 (1), 119–137.

Wallander, J., 1999. Budgeting—an unnecessary evil. Scandinavian Journal of Management 15, 402–421.

Welch, J., 2005. Winning. Harper Collins, New York, NY.