ria

Hough

Despite calls to link management accounting more closely to management (Jonsson, 1998),

used by managers in discussions and debates with subor-dinates and other managers?

Over 10 years ago, Jonsson (1998) exclaimed in the titleof his paper the need to relate management accountingresearch to managerial work! Despite Jonssons (1998)

show that decision-making is only a relatively small part ofmanagerial work and sometimes not that critical (forexample, see Hales, 1986; Kotter, 1982; Mintzberg, 1973;Stewart, 1988; Whiteley, 1985). In addition, much mana-gerial work involves addressing problems that involve tur-bulence, doubt, uncertainty, and the potential forsignicant error (Hales, 1986; Isenberg, 1984; Kotter,1982; Landau & Stout, 1979). As such, a strong focus on

0361-3682/$ - see front matter 2009 Elsevier Ltd. All rights reserved.

* Tel.: +44 (0)20 7849 4633; fax: +44 (0)20 7955 7420.E-mail address: m.r.hall@lse.ac.uk

Accounting, Organizations and Society 35 (2010) 301315

Contents lists available at ScienceDirect

Accounting, Organiza

w.edoi:10.1016/j.aos.2009.09.003Despite calls to link management accounting more clo-sely to management (Jonsson, 1998), much is still to belearned about the role of accounting information in mana-gerial work. For what purposes do managers use account-ing information beyond its role in specic decision-making scenarios? With many other information sourceson offer, what is it about accounting information that man-agers nd helpful? How exactly is accounting information

fervent arguments, future research has generated fewstudies directed towards understanding howmanagers en-gage with accounting information in their work. Muchmanagement accounting research has been focused onhow managers use accounting information to make deci-sions in well-dened scenarios. Although managers domake decisions, and many of these are undoubtedly impor-tant, empirical investigations of what managers actually domuch is still to be learned about the role of accounting information in managerial work.This lack of progress stems partly from a failure to incorporate in research efforts the nd-ings regarding the nature of managerial work, as well as inadequate attention devoted tothe detailed practices through which accounting information is actually used by managersin their work. In this paper I draw on prior research to develop a series of propositionsfocused on three primary insights into how and why managers use accounting informationin their work. First, managers primarily use accounting information to develop knowledgeof their work environment rather than as an input into specic decision-making scenarios.In this role, accounting information can help managers to develop knowledge to preparefor unknown future decisions and activities. Second, as accounting information is justone part of the wider information set that managers use to perform their work, it is imper-ative to consider its strengths and weaknesses not in isolation but relative to other sourcesof information at a managers disposal. Third, as managers interact with information andother managers utilising primarily verbal forms of communication, it is through talk ratherthan through written reports that accounting information becomes implicated in manage-rial work. These insights have important implications for how managers use accountinginformation, and, in particular, require reconsideration of the types accounting informationthat managers nd, or could nd, helpful. The paper also considers how existing experi-mental and eld-based methods could fruitfully be adapted to focus on the detailed activ-ities through which managers engage with accounting information.

2009 Elsevier Ltd. All rights reserved.Accounting information and manage

Matthew Hall *

Department of Accounting, London School of Economics and Political Science,

a b s t r a c t

journal homepage: wwl work

ton Street, London WC2A 2AE, United Kingdom

tions and Society

lsev ier .com/ locate/aos

302 M. Hall / Accounting, Organizations and Society 35 (2010) 301315how managers use accounting information to make spe-cic decisions in well-dened contexts is restrictive as itlimits consideration of other, potentially much moreimportant, ways that managers use accounting informa-tion in their work.

There is also much to learn about howmanagers engagewith accounting information because there are remarkablyfew studies of what information managers actually use ormight use (Anderson, 2008; March, 1986). In particular,prior studies have devoted inadequate attention to the de-tailed practices through which accounting information isused by managers in their work (Ahrens & Chapman,2007; Hopwood, 1989). Studies that focus on organisa-tional-level issues only are limited because they are typi-cally based upon assumptions about, rather than adetailed investigation of, managerial work behaviour(Covaleski, Evans, Luft, & Shields, 2003; Hall, 2008a). Whatis often missing from these organisational accounts is ananalysis of the detailed ways in which managers useaccounting information to perform particular activities.

Three studies in particular exemplify the kinds of in-sights that can be generated from inquiries directed moreclosely at examininghowmanagers engagewith accountinginformation (McKinnon & Bruns, 1992; Preston, 1986; Si-mon, Kozmetsky, Guetzkow, & Tyndall, 1954). Althoughbest known for its typology of score-keeping, attentiondirecting and problem solving, the Simon et al. (1954) studyprovides considerable evidence concerning how differenttypes ofmanagers used (or did not use) accounting informa-tion. Preston (1986) examined how managers in a plasticsfactory engaged in what he termed the process of inform-ing, which involved the use of accounting information aswell as other more informal sources such as observations,personal record keeping and meetings. He concluded thatinformal sources of information are used in spite of, ratherthan because of, limitations to formal documented systemsas theyare intrinsic tohowmanagers goaboutmaking senseof their worlds. Similarly, although across a wider scope ofmanagers and organisations, McKinnon and Bruns (1992)investigated how a variety of production, sales and nancemanagers used accounting and other information in theirwork. The resulting conuence of information that manag-ers developed from many different sources they referredto as an information mosaic.

In addition to the rich empirical ndings, what is partic-ularly illuminating about these studies is that rather thanassume particular roles for accounting information, theysought to investigate if a role exists and what it mightbe. Simon et al. (1954, p. 22) examined the use (or non-use) of accounting data by operating executives and super-visors. Preston (1986, p. 522), beginning with a focus onthe use of a computerised production information system,broadened his inquiry to investigate how the managersinformed themselves and each other. McKinnon andBruns (1992, p. 2) asked under what circumstances isinformation that we think of as accounting informationactually used by managers? (italics in original). Thephrases or non-use, how, and actually used demon-strate that the researchers at least considered the possibil-ity that managers may not use accounting information,that its role may be limited, or that new roles and possibil-ities for accounting information may emerge. In thesestudies there are few assumptions regarding the ways inwhich managers use accounting information; the studiesthemselves are an attempt to nd out what these roles, ifany, may be. All three studies also examine in considerabledetail howmanagers engaged with accounting informationas part of a process involving other sources of information,other managers, and the wider organisational context. Thestudies were not solely concerned with how accountinginformation was used for decision-making purposes, norwere they focused only on how accounting informationwas implicated in wider organisational processes. Thesethree studies show how managers draw upon a range ofdifferent resources, such as accounting information, othersources of information, and interactions with other manag-ers, to perform their complex and demanding work.

The objective of this paper is to examine, and encouragefurther research to examine, how and why managers useaccounting information. Although insightful, collectively,prior studies of what managers actually do with account-ing information lack integration and conceptual claritysuch that common themes and issues have not been devel-oped. Drawing on prior research, I develop a series of prop-ositions focused on three primary insights into how andwhy managers use accounting information in their work.First, managers primarily use accounting information todevelop knowledge of their work environment rather thanas an input into specic decision-making scenarios. In thisrole, accounting information can help managers to developknowledge to prepare for unknown future decisions andactivities (March, 1986; Preston, 1986). Importantly, in aknowledge development role, rather than complex andsophisticated reports and analyses, managers requireaccounting information that is easily understandable andprovides a common-sense story of organisational perfor-mance. Second, as accounting information is just one partof the wider information set that managers use to performtheir work (McKinnon & Bruns, 1992; Preston, 1986), it isimperative to consider its strengths and weaknesses notin isolation but relative to other sources of information ata managers disposal. In particular, and in contrast to tradi-tional criticisms of accounting, the strengths of accountinginformation relate to its aggregation properties and its roleas a common, nancial language to facilitate communica-tion among managers. Third, managers interact with infor-mation and other managers utilising primarily verbalforms of communication (Ahrens, 1997; Jonsson, 1998;Kotter, 1982; McKinnon & Bruns, 1992; Preston, 1986).As such, it is primarily through talk rather than throughwritten reports that accounting information becomesimplicated in managerial work. In particular, verbal formsof communication allow managers to tailor accountinginformation to specic operational concerns, and providea context to debate and discuss the meanings and implica-tions of accounting data. Accounting information can alsoprompt discussions by signalling the need to investigatean issue further. Overall, incorporation of ndings regard-ing how managers actually work has important implica-tions for their use of accounting information, and, inparticular, requires reconsideration of the types accountinginformation that managers nd, or could nd, helpful.

This perspective on accounting information and mana-gerial work also has implications for research methods.Most generally, there is a need for research to examine

Prior research on the role of accounting information in

M. Hall / Accounting, Organizations and Society 35 (2010) 301315 303developing knowledge has primarily focused on how itserves as an information input into specic decisions. Inthe decision-facilitating role, accounting information, inthe form of periodic reports or special analyses, is a sourceof information for making decisions (Horngren, Bhimani,Datar, & Foster, 2005; Sprinkle, 2003).1 Sprinkle (2003) con-cludes that the provision of accounting information for deci-sion-facilitating purposes and the characteristics of thatinformation have been found to improve individuals knowl-edge and their ability to make better decisions. However, afocus on how managers use accounting information to makespecic decisions in well-dened contexts is restrictive as itlimits consideration of other potentially much more impor-tant ways that managers use accounting information in theirwork.

Although managers do make decisions, and many ofthese are undoubtedly important, empirical investigationsof what managers actually do show that such activities areonly a relatively small part of managerial work and some-times not that critical (for example, see Hales, 1986; Hales,1999; Kotter, 1982; Mintzberg, 1973; Stewart, 1988).Much managerial work involves responding to theunusual, the ad hoc, and the unplanned, where problemboundaries are typically hazy and unstable (Hales, 1999;

1 See Sprinkle (2003), Luft and Shields (2003) and Birnberg, Luft, andShields (2007) for reviews of the use of accounting information in decision-making.the specic activities through which managers engagewith accounting information in their work. To achieve this,existing experimental and eld-based approaches canfruitfully be adapted. In the context of experiments, a pro-cess-based approach (Swieringa & Weick, 1982; Vygotsky,1978) can be used to focus more directly on howmanagersengage with accounting information in performing partic-ular activities. In the context of eld studies, a stronger fo-cus on examining the micro-practices (Alac & Hutchins,2004; Hutchins, 1995) involved in managers use ofaccounting information can improve our understandingof how managers engage with accounting information intheir work.

The remainder of the paper is structured in ve sec-tions: the rst section examines the role of accountinginformation in developing knowledge of the work environ-ment. The second section considers accounting informa-tion as part of a managers wider information set. Thethird section explores the relation between accountinginformation and forms of communication. Each of thesethree sections analyses the results from prior researchand concludes with a set of propositions. The fourth sec-tion outlines the implications for research methods. Thefth section concludes the paper.

The role of accounting information in developingknowledge of the work environmentHanaway, 1989; Kotter, 1982). Furthermore, the variednature of much managerial work means that managerstypically do not deal with one or two problems; rather,they deal with portfolios of problems, where it is unclearhow problems are related or how small problems mayrelate to or be indicative of something more serious(Hanaway, 1989; Isenberg, 1984). In this way, a criticaltask for managers is to deal with problems that involveturbulence, doubt, uncertainty, and the potential for signif-icant error (Landau & Stout, 1979). In this context, manag-ers use information to develop knowledge of their workenvironment more generally, that is, beyond specic deci-sion-making situations. Managers often use past experi-ences and prior knowledge to develop appropriateresponses, make decisions and take actions (Dane & Pratt,2007). As such, most of the information that managersgather is not for decision-making purposes but is used todevelop a context of knowledge and meaning for unknownpossible future actions (March, 1986; Feldman & March,1981; Preston; 1986; McKinnon & Bruns, 1992; Simonet al., 1954). In this process, the signicance of informationgathering is as an investment in an inventory of knowl-edge, not as an input into a specic decision-making sce-nario (March, 1986).

The process of developing knowledge of the workenvironment involves a variety of specic activities. Forexample, to develop knowledge, managers test andscrutinize their assumptions and expectations about theorganisation, its operations and competitive environment(Mintzberg, 1973; Senge, 1990; Vandenbosch & Higgins,1995), identify and dene problems and opportunities(Mintzberg, 1973; Simons, 1990; Vandenbosch, 1999;Vandenbosch & Huff, 1997), and monitor the environmentfor surprises, or reassurances there are none, where sur-prises may be new alternatives, preferences, or signicantchanges (Feldman & March, 1981; Preston, 1986). Theseactivities are reective of a more general process wherebymanagers ensure they are up-to-date and informed aboutsignicant events and occurrences that relate to their workenvironment (Preston, 1986; Simon et al. 1954).

Accounting, as a key source of information about busi-ness performance, can help managers to develop knowl-edge of the work environment in several ways: to makevisible those activities not visible through a managers dai-ly activities; and to provide an overall quantitative per-spective on their work. Accounting information can makevisible those problems that are not visible from day-to-day activities and can provide an independent check onoperations to help managers know what is going on(Simon et al., 1954, p. 28). Van der Veeken and Wouters(2002) found that budgeted versus actual cost informationwas crucial for senior managers in managing projects asthese managers were responsible for many projects andused up-to-date information on allowable versus actualcosts to develop knowledge about which projects werecausing problems. Accounting information also helps tosmooth out the multitude of organisational activities,which enables managers to determine the meaning andsignicance of all the frenetic day-to-day managementactivity (McKinnon & Bruns, 1992; Preston, 1986). Simonet al. (1954, p. 28) noted that the signicance

304 M. Hall / Accounting, Organizations and Society 35 (2010) 301315of. . ...(accounting) reports lies in their reminding the oper-ating executives of things they already know, and placingthose things in proper quantitative perspective, rather thanhinting to them things they never suspected (italics in ori-ginal). In this role, accounting information can provide anoverall assessment of the trends and the net effect of allkinds of disturbances and actions that have taken place(Preston, 1986; Van der Veeken & Wouters, 2002).

Three factors inuence the usefulness of accountinginformation for developing knowledge of the work envi-ronment: closeness to operational activities, time horizon,and diversity of operational factors under consideration.Managers who are close to operations use observations ofphysical processes and informal reports from subordinatesand peers as their primary means of developing knowledgeof the work environment (Preston, 1986; Simon et al.,1954). For example, Van der Veeken and Wouters (2002)reported that operating managers found little use foraccounting data on budgeted versus actual project costs,but obtained information from observation of projectactivities and a few key non-nancial metrics. Managerswith little contact with operations, however, devote con-siderable attention to accounting reports as they have lim-ited opportunities for picking up information from actualobservations of work being conducted (McKinnon & Bruns,1992; Simon et al., 1954). For day-to-day concerns,developing knowledge is facilitated most efciently bynon-nancial numbers as they relate more directly tooperational activities, and, importantly, are usuallyavailable immediately without delay (McKinnon & Bruns,1992). In contrast, events and transactions take too longto go through the formal accounting reporting system forthe output to be actionable. For example, McKinnon andBruns (1992) found that nancial numbers were not usedin any of the 12 organisations they studied as a key dailyproduction indicator: all managers used non-nancialnumbers. Furthermore, the aggregations required for theproduction of nancial numbers can obscure details thatare important for understanding day-to-day problems.For example, an injury to an employee becomes aggregatedwith other costs and is thus obscured from the managersview (McKinnon & Bruns, 1992). As the time horizonlengthens, however, nancial information becomes muchmore important in providing overall measures of effective-ness and in highlighting key problem areas that requirefurther investigation (McKinnon & Bruns, 1992; Preston,1986; Simon et al., 1954). When managers have only afew operational factors to consider, non-nancial mea-sures and direct observation of processes can provide ade-quate information. However, as managers consider a morediverse range of operational factors, nancial informationcan operate to translate these factors into a single, nancialdimension, which allows for an overall assessment of thenet effect of all kinds of disturbances and actions that havetaken place (Van der Veeken & Wouters, 2002).

Research has also considered the properties of account-ing information that are likely to help managers to developknowledge of their work environment. One approach toincreasing the role of accounting in knowledge develop-ment activities is to increase its sophistication and com-plexity. Techniques such as the balanced scorecard(Kaplan & Norton, 1996), causal performance maps (Aber-nethy, Horne, Lillis, Malina, & Selto, 2005), performancepyramids (Lynch & Cross, 1992), and activity-based costing(Kaplan & Cooper, 1998) provide more elaborate, detailedand comprehensive data. These systems attempt to makeaccounting information a more complete and more rele-vant description of underlying organisational activities. Incontrast to these developments, however, research showsthat accounting information does not need to be elegant,complete or accurate to be useful for developingknowledge.

Malina, Norreklit, and Selto (2007) show that the abilityof a performance measurement system to communicate acommon-sense and credible story of business operationsis far more important than the creation of a statistically-valid predictive business model. Rowe, Birnberg, andShields (2008a) report that technical jargon and complex-ity of accounting information limited managers ability toidentify the economic effects of competing initiatives. Inanother study, Osborn (1998) nds that the use of simpli-ed accounting information (i.e., information structuredaround key issues and categories) prompted productivediscussion among managers and more time being devotedto building shared interpretations of results. As one man-ager stated: I use the [system] as a way to frame the busi-ness. We need to add structure to the data that werereporting: we need actionable data, data that leads us todo something. We need to help people identify hot spots.(Osborn, 1998, pp. 495496, italics in original). Further-more, information needs to be easily comprehended bymanagers such that they have condence in their under-standing of the underlying data (McKinnon & Bruns,1992; Preston, 1986; Simons, 1995). Such condence islikely to prove increasingly difcult in settings where thecomplexity of accounting information increases but man-agers have little time to spend on understanding it. Forexample, Ittner and Larcker (2003) note the difcultiesand problems managers face in understanding and linkingnon-nancial measures to each other and to nancial mea-sures in the context of complex performance measurementsystems.

These ndings are consistent with the expectation thatthe process of developing knowledge of the work environ-ment is promoted by managers engaging with simplerinformation that challenges existing points of view (Gilkey& Kilts, 2007). Complex accounting systems can reduce orcamouage the very uncertainties that managers need tobe aware of and can eliminate the ability to detect and cor-rect error (Earl & Hopwood, 1980; Landau & Stout, 1979).For example, in the context of variance analysis, Emsley(2001, p. 34) nds that managers do not prefer detailedvariances but a summary of key problems ranked byimportance: instead of all these variances all I want is alist of the top 10 biggest problems to have occurred inthe factory through the month. In the performance mea-surement context, Hall (2008b) nds that more compre-hensive performance measurement systems reinforcemanagers existing points of view and do not play a rolein challenging and developing new ways of thinking.

Research also indicates that developing knowledgeof the work environment is likely to be facilitated by

information that focuses on stimuli as a whole rather thaninformation that attempts to deliberately decipher cause-and-effects (Dane & Pratt, 2007; Reber, 1976). Rather thantrying to formalise relations via cause-and-effect chains,2

M. Hall / Accounting, Organizations and Society 35 (2010) 301315 305developing knowledge of the work environment is likely tobe facilitated by highlighting key events and outcomes, withmanagers developing their own connections and interpreta-tions. For example, McKinnon and Bruns (1992) found thatthe process of recalling the events and activities that oc-curred during the period of time covered by an accountingreport, and then seeing their success as reported by account-ing information, allowed managers to associate events andactivities with a level of nancial performance. This processprovided information about linkages between physicalevents and nancial outcomes that managers could use toupdate their knowledge of operations. Importantly, manag-ers used their own experiences to develop knowledge thatlinked events and outcomes and did not require accountinginformation that made these connections explicit. This isconsistent with arguments that formal techniques and pro-cedures can inhibit the operation of individuals more intui-tive and natural synthesizing processes (Dane & Pratt,2007; Gigerenzer, 2008; Raiffa, 1968).

This analysis indicates that rather than focus on thedevelopment of forms of more complex, complete andsophisticated accounting information, what is required isa different kind of sophistication, one that links more clo-sely with the types of information that will aid managersin developing knowledge of the work environment. Thisdiscussion is formalised in the following propositions:

Proposition 1: Accounting information is used by manag-ers to develop knowledge of their workenvironment. This involves using account-ing information to test assumptions andexpectations about the organisationsoperations, and to identify problems,opportunities and potential surprises.

Proposition 1a: Accounting information is more (less)helpful for knowledge developmentwhen a manager is further removedfrom (closer to) operational activities,the time horizon of the issue under con-sideration is longer (shorter), and thenumber of operational factors to con-sider is larger (smaller).

Proposition 1b: Accounting information is more (less)likely to facilitate managers knowledgedevelopment when it highlights (doesnthighlight) the outcomes of key events, isstructured (not structured) aroundissues of managerial importance, is freeof (contains) jargon, and does not (does)decipher cause-and-effect relations.

2 Whether this is even possible and/or desirable is the subject of muchdebate in the research literature (Ittner & Larcker, 2003; Malina et al., 2007;Norreklit, 2003).Accounting information as one part of a managersinformation set

This section analyses the role of accounting informationas just one part of a managers information set, and, in par-ticular, considers the strengths of accounting informationvis--vis other information at a managers disposal.Strengths of accounting information include its aggrega-tion properties and its role as a common language to facil-itate communication among managers with differentbackgrounds, experience and knowledge.

For managers who are engaged in a variety of tasks incomplex social and organisational contexts, informationneeds are diverse and encompass a wide range of informa-tion from various sources, both internal and external to theorganisation. Managers use not just accounting informa-tion but information from other specialists, informationon market, industry and economic conditions, as well as di-rect observation and informal reports (see, for example,Hanaway, 1989; Kotter, 1982; McKinnon & Bruns, 1992;Preston, 1986). In contrast, much management accountingresearch tends to focus on how managers use accountinginformation only and devotes little attention to other typesof information that managers use (McKinnon & Bruns,1992). However, the sheer variety of tasks that managersengage with, and the rapid, fragmentary and disjointedway in which they are conducted, necessitates engagingwith an extensive amount of information from a wide vari-ety of sources.

McKinnon and Bruns (1992) found that managers use avast array of information in performing their jobs, rangingfrom facts and forecasts to gossip, intuition and gut feel.Although managers described accounting information asindispensable, it formed only one small part of the infor-mation they used. Other information besides accounting,particularly observation of physical ows and reports ofevents and activities from subordinates and peers (infor-mal reports), were available for managers to consider, eval-uate and act on often before they were observed by theaccounting process. Similarly, Simon et al. (1954) notedthe existence of unofcial reports kept by operating exec-utives, sometimes termed black books. They were usedconstantly by operating managers and contained someaccounting gures, but consisted mainly of numbers fromproduction reports and other physical data, notes, memosand informal notations. Managers also develop vast net-works of contacts, both within and outside the organisa-tion, which includes subordinates, peers, outsiders,bosses bosses, subordinates subordinates, customers,bankers and the press (Hanaway, 1989; Kotter, 1982;Whiteley, 1985).

Managers rely on multiple information sources to cor-roborate the different types of information that they use.For example, McKinnon and Bruns (1992) noted how asales manager may see corroboration in the number of or-ders received, call reports, competitive prices, inventorylevels, and economic news. Similarly, Hopwood (1972)notes how prot conscious managers probe into the signif-icance and meaning of accounting data, supplementingthem with many other sources of information, both formaland informal, which allows for a continued test of the

validity of the accounting data. The use of corroboratinginformation is not restricted to those managers concernedwith operations; even nancial managers seek to corrobo-rate accounting information: I get information about pro-duction directly from them, but also from cost accounting.Much of the production data is not accounting data. Thereare other metrics like on-time delivery to customers, howmany vendor schedule changes we have. There are a lot

2005), and managers may thus nd it more difcult tocommunicate with each other. Consistent with this, Roweet al. (2008a) nd that the use of a new accounting modelbased on a simplied accounting language, rather thantechnical accounting jargon and codes, helped managersin a cross-functional team to communicate with each otherand to debate cost issues amongst themselves.4

306 M. Hall / Accounting, Organizations and Society 35 (2010) 301315of different metrics we use to assure we have good inven-tory integrity (McKinnon & Bruns, 1992, p. 201).

This analysis highlights the wide variety of informationthat managers engage with in conducting their work activ-ities, of which accounting information is a small but some-times important part. Accounting, in other words, is partof a wider whole (Hopwood, 2007, p. 1367). The position-ing of accounting information within a wider informationset helps to develop a more critical understanding of itsrelative strengths. In particular, the aggregation processinherent in the accounting system serves to provide infor-mation to managers not generally available from othersources. Because the accounting process assigns nancialnumbers to a diversity of operational factors, they can becombined and thus compared through a process of aggre-gation. In this way we see how accounting information isuseful in developing knowledge of the work environmentby helping managers to identify what all the frenetic oper-ational activities add up to and by assessing trade-offsamong different factors (Simon et al., 1954; Van der Vee-ken & Wouters, 2002; Wouters & Verdaasdonk, 2002).Other information available to managers, such as informalreports and even non-nancial information, generally donot have these properties as they are not expressed usingthe same basis of measurement and thus are not easilycombined and compared. In particular, linking non-nan-cial measures to each other and to nancial measures isproblematic (Ittner & Larcker, 2003). As such, althoughaggregation may limit the usefulness of accounting infor-mation for some roles, it can help rather than hinder theconduct of particular aspects of managerial work.3

The translation of operational activities into nancialnumbers also serves another function: to act as a commonlanguage with which managers can communicate. As acommon language, accounting information can be usedas an anchor to frame discussions amongst managers (Ah-rens & Chapman, 2007; Simons, 1990). Accounting infor-mation can play a key role in consensus building byconstructing a common set of information to facilitatecommunication (Rowe, Birnberg, & Shields, 2008b). Thisis critical as structuring and framing of issues in a commonlanguage helps to produce meaning (Simons, 1990; Tver-sky & Kahneman, 1981). If information is not uniformlyframed, common interpretations are unlikely (Chenhall,

3 There are limits to this role as not all operational factors are amenableto expression in nancial terms (Chapman, 1997; Galbraith, 1973). Forexample, Simons (1995) shows how some strategic uncertainties are bestaddressed using accounting-based information systems (e.g., prot plan-ning systems to frame discussions about how changes in products andmarkets are likely to affect future prots) whereas others are not (e.g.,intelligence systems to collect information about social, political andtechnical environments). See Simons (1990, 1995) and Chapman (1997) forfurther discussion.This indicates that the use of accounting information de-pends not necessarily on its ability to depict operationalprocesses accurately or completely, but on its functioningas a medium through which seemingly diverse operationalconsiderations are rendered communicable through a com-mon language. In this way, an accounting language canfunction as a medium to facilitate communication amongmanagers with different information requirements, back-grounds and functional experience. Using accounting as acommon language is likely to be particularly important inorganisations where managers who work together have di-verse educational backgrounds and/or in multinationalorganisations where managers are from different cultures.Furthermore, an accounting language is likely to be veryimportant in settings where managers need to communi-cate extensively across functional boundaries, for example,research and development managers liaising with market-ing managers. Here, accounting information can be used tocreate common categories for ordering and framing the dis-cussions that take place between managers from differentfunctional groups (for further elaboration, see, for example,Bowker and Star (1999), Power (2004), Espeland andSauder (2007) and Millo and Mackenzie (2009)). Anaccounting language is also likely to be increasinglyimportant in communicating with managers outside theorganisation as new forms of inter-organisational relation-ships become more prevalent (e.g., Dekker, 2004).

Overall, this analysis indicates that understanding therole of accounting in managerial work requires the posi-tioning of accounting information as one part of a man-agers information set. Such an approach can facilitateunderstanding of how accounting information interactswith other sources of information, and, in particular, thestrengths of accounting information relative to other infor-mation at a managers disposal. This can lead to question-ing of the need to develop more complex accountinginformation in the context of an information set that mayalready adequately provide such information from othersources. A more fruitful approach is to focus on thestrengths of accounting information, such as its aggrega-tion properties and role as a common language. This anal-ysis leads to the following propositions:

Proposition 2: As accounting information is just onepart of a managers information set(including direct observations, informal

4 This has important implications for target costing practices whichtypically involve a variety of managers from different functional areas. Inthis setting, complex costing models, like sophisticated activity-basedcosting, may be damaging as non-accounting teammembers are unlikely tohave the necessary expertise to converse condently using such informa-tion. As Rowe et al. (2008a) nd, what is required is relatively simpliedcost information that forms a basis for discussion around cost reductioninitiatives.

and provides a context to debate and discuss the meanings

M. Hall / Accounting, Organizations and Society 35 (2010) 301315 307and implications of accounting data. Furthermore,accounting information can also prompt managerial dis-cussions that take place.

Managers have a strong preference for verbal communi-cation, preferring contacts with people in meetings, infor-mal discussions, and via the telephone (Hales, 1986;Hanaway, 1989; Kotter, 1982; McKinnon & Bruns, 1992;Preston, 1986). Managers use these verbal contacts to eas-ily bypass formal organisation charts and seek informationfrom those people who have it, rather than wait for infor-mation to arrive from formal channels (McKinnon & Bruns,1992; Preston, 1986). As such, most of a managers time isspent with other people, with verbal communications either in-person or by telephone dominating all otherkinds (McKinnon & Bruns, 1992).

Despite the importance of verbal communications inmanagerial work, the focus of much research typically con-cerns the design and dissemination of the products ofaccounting systems; namely written documentation inthe form of reports and analyses. In particular, efforts tomake accounting information more relevant for manage-rial work usually focus on improving its content, for exam-ple, non-nancial performance measures, balancedscorecards, performance pyramids, and activity-basedreports, industry/market/economic data,other specialist information), itsstrengths and weaknesses are deter-mined not in isolation but relative toother sources of information at a man-agers disposal.

Proposition 2a: The primary advantage of accountinginformation vis--vis other informationis its ability to translate operational fac-tors into a single, nancial dimension. Asingle, nancial dimension allows man-agers to compare operational factorsand provides a common language tofacilitate managerial discussions.

Proposition 2b: A single, nancial dimension is more(less) helpful when the diversity of oper-ational factors is high (low), when man-agers functional, educational and/orcultural backgrounds are more (less)diverse, and when the extent of commu-nication across functional borders ishigh (low).

Accounting information and forms of communication

In addition to consideration of a wider information set,understanding the role of accounting information in man-agerial work requires an appreciation of the forms of com-munication that managers use when engaging withaccounting information. This section examines how man-agers preference for verbal communications affects theuse of accounting information in managerial work. In par-ticular, verbal communication allows managers to tailoraccounting information to specic operational concerns,costing systems (for example, Fisher, 1995; Ittner & Larc-ker, 2003; Kaplan & Cooper, 1998; Kaplan & Norton,1996; Lynch & Cross, 1992). Other research examineshow the usefulness of accounting information is related tohow it is organised and displayed (for example, Cardinaels,2008; Lipe & Salterio, 2002). Although important, a sole fo-cus on accounting information in written form limits con-sideration of how managers actually use accountinginformation in performing their work. In particular, the rel-evance of accounting information is not solely a function ofits information characteristics and/or how it is organised/displayed, but is highly dependent on whether and howmanagers use accounting information in verbalcommunications.

For example, McKinnon and Bruns (1992) found thatinformation discussed in verbal communications is notlimited to gossip, intuition or qualitative items; rather,most numerical data appeared to be passed by word ofmouth rst, with formal reports serving to corroborate orremind managers of what was transmitted orally. This re-lates to research that highlights the importance ofaccounting talk, whereby accounting information be-comes implicated in managerial work primarily throughverbal communications rather than through reports(Ahrens, 1997; Jonsson & Solli, 1994). In particular, verbalcommunication of accounting information is not just anexchange of information but a process through whichaccounting information is related to specic managerialproblems or issues (Ahrens, 1997). Thus, even if the con-tent of accounting information is not specically relevantfor a managers work, verbal discussions can facilitate thetailoring of accounting information to address issues thatmanagers deem important. In this way, managers workwith accounting information to make it relevant, ratherthan its relevance being determined solely by its content.This is consistent with research showing that managersconstruct meaning from information for themselves andare not mere processors of information from reports (e.g.,Ahrens, 1997; Chapman, 1998; Dent, 1991; Jonsson,1998). Echoing the ndings from many studies, Boland(1993, 135) concludes that the reader of the accountingreport is in no sense an uninvolved, passive recipient of aclear and obvious message that the report is supposed tocarry. . .they take individual responsibility for reshapingthe raw data in the text and reclaiming from it a set ofthe truly important data. Hence, it is not primarily the de-sign of accounting that determines its relevance, but howmanagers actually interpret and use such information, of-ten in verbal communications.

Verbal communication of accounting information alsoprovides opportunities for managers to obtain more tacitforms of information. Managers often value informationthat is speculative and informal (tacit) rather than infor-mation that is authoritative and formal (explicit) (Hales,1986; Mintzberg, 1973; Turner & Makhija, 2006). Althoughaccounting information in the form of reports and analysesis likely to be explicit, verbal communications around themeaning and implications of accounting information canfacilitate the exchange of more tacit forms of information,and provide a context within which to debate and discussthe meanings and implications of accounting data. In this

way, verbal communication of accounting information canenable more tacit information exchange and potentiallymake it much more relevant to managerial work. Impor-tantly, managers involved in tasks comprising innovativeproblem solving and unexpected outcomes tend to relyon interpersonal communication and intensive forms ofinformation exchange, with little emphasis on the codica-tion of knowledge via formal reports (Ditillo, 2004).

The role of accounting information in verbal communi-

little insight into whether accounting information commu-nicated verbally is the same or different from accountinginformation that is communicated in written form. Whilst

308 M. Hall / Accounting, Organizations and Society 35 (2010) 301315cations is not limited to the exchange of information, how-ever. More fundamentally, when mobilised as part of anorganisational system, accounting information can promptdiscussions by signalling that something must be lookedinto more carefully (Mintzberg, 1975; Simon et al., 1954;Simons, 1995). Ahrens and Chapman (2007) show howaccounting information was used to identify areas of off-standard performance and to prompt discussions betweenmanagers about potential actions and solutions. For exam-ple, a below-average food margin prompted questioningfrom the area manager as to its potential causes, which ini-tiated a discussion centring on several likely explanations,such as burnt steaks or over-portioning of dishes. Simonet al. (1954) describe how a production manager fre-quently and regularly commented on a monthly variancereport and required explanations of variances from a fac-tory manager. Not only did the factory managers seek toprovide these explanations in discussions with superiors,but he was using the same practice with his departmentheads, and the department heads with their foreman. Inthese situations, accounting information prompts discus-sions primarily through the design of the accounting infor-mation itself, i.e., the accounting information reports adeviation (such as a below-average food margin), whichthen prompts investigation and discussion. More recently,Simons (1995) outlined how interactive control systemsprompt discussions among managers where informationfrom formal systems provides the basis for debates overunderlying data, assumptions and action plans.5 Here,accounting information prompts discussion primarilythrough managers interpreting accounting information andthen initiating discussions about issues they deem impor-tant (even in the absence of specic deviations). Overall, thisanalysis shows how accounting information is not merelyexchanged verbally but can serve a more fundamental rolein prompting discussions that take place.

Whilst a stronger focus on the communication ofaccounting information in verbal forms is warranted, thisshould not occur in isolation from other forms of commu-nication. What is particularly interesting is whether andhow verbal and written forms of accounting informationact as substitutes or complements. Prior research provides

5 Most attention in the accounting literature, however, has focused onthe organisational-level effects from the interactive use of particularaccounting systems, such as budgets and performance measurementsystems (Abernethy & Brownell, 1999; Bisbe & Otley, 2004; Henri, 2006).This research has predominately used the descriptions and characteristicsof interactive control systems as described by Simons to develop items forquestionnaires (Bisbe, Batista-Foguet, & Chenhall, 2007), rather thaninvestigate the nature of the interactive control process itself. Thus,knowledge of the way in which accounting information is used to promptdiscussions as part of an interactive control system is limited.research is limited, it appears that verbal communication issuited to more tacit, speculative, and specic forms ofaccounting information, whereas written communicationis more suited to explicit, formalised and aggregate formsof accounting information (see Daft & Lengel, 1986). It isalso unclear whether there exists much overlap and redun-dancy in the accounting information communicated in ver-bal and written forms. Verbal and written forms ofaccounting information have the potential to reinforceeach other and thus act as complements. In particular,prior research indicates that managers value some infor-mation redundancy as it serves to remind managers of pre-vious information and also builds condence in theinformation through a process of corroboration (McKinnon& Bruns, 1992; Preston, 1986; Simon et al., 1954).

This analysis indicates that managers preference forverbal exchanges of information can affect the use ofaccounting information in managerial work. In contrastto written reports and analyses, verbal communication al-lows managers to relate accounting information to specicoperational concerns and to discuss and debate the mean-ing and implications of accounting data. Furthermore,accounting information can prompt the discussions thatmanagers have with each other. This leads to the followingpropositions:

Proposition 3: The relevance of accounting informationfor managerial work is determined pri-marily by how managers use and inter-pret accounting information in verbalcommunications and discussions.

Proposition 3a: Managers use verbal forms of communi-cation to relate accounting informationto specic operational concerns and todiscuss and debate the meanings andimplications of accounting data.

Proposition 3b: Accounting information can prompt ver-bal discussions by signalling that an issuemust be investigated further.

This analysis shows how the nature of managerial workhas important implications for the way in which managersengage with accounting information. In particular, manag-ers are likely to acquire information for purposes otherthan solving specic problems, source information from awide information set, and use mainly verbal forms of com-munication. This indicates different and potentially moreimportant roles for accounting information in managerialwork that move beyond the use of written analyses and re-ports to solve specic problems. Drawing on prior researchI outlined a series of propositions directed towards increas-ing understanding of how and why managers use account-ing information.6 First, future research can examine howmanagers use accounting information to develop knowledge

6 These propositions will require further articulation and adaptation tothe specic research design employed in a study.

situations that characterise much managerial work. In par-

process-based experimental studies have much to offer to

M. Hall / Accounting, Organizations and Society 35 (2010) 301315 309the analysis of how managers engage with accountinginformation in their work.

Process-based experiments are concerned with analys-ing the processes that individuals engage in as theyconduct particular activities (Daniels, Cole, & Wertsch,2007; Swieringa & Weick, 1982; Vygotsky, 1978). The crit-ical object of analysis is not the output that individualsproduce but the methods, processes and activities that par-ticipants engage in when conducting tasks. The experi-mental setting is used to provide a context forinvestigation where the researcher manipulates its struc-ticular, rather than serve as an input into a pre-deneddecision-making context, managers engage with account-ing information to construct their own interpretations,drawing on their experience, prior knowledge, othersources of information and discussions with other manag-ers. This is a dynamic process in which the underlyingaccounting information itself can be transformed, thuscreating further possibilities for action (Burchell, Clubb,Hopwood, Hughes, & Nahapiet, 1980; Hopwood, 1978).Within this perspective, experimental approaches thatfocus on relating variable X to variable Y are not helpfulfor examining how managers engage with accountinginformation in performing their work. As such, I argue thatof their work environment. Second, as managers use ofaccounting information takes place within a wider informa-tion set, future research can focus on the role of accountinginformation relative to the other sources of information at amanagers disposal. Third, as managers favour and usemainly verbal forms of communication, there is a need toexamine how accounting information is used in managerialwork through verbal communications. In the next section Ioutline the implications of this analysis for researchmethods.

Finding out how managers use accounting information

Existing experimental and eld-based methods can befruitfully adapted to examine how managers use account-ing information. Overall, I argue for a stronger focus on thedetailed activities through which managers engage withaccounting information. In the context of experiments, aprocess-focus can be used to examine how participants en-gage with accounting information in performing theirwork. In the context of eld studies, a stronger focus on mi-cro-practices can be used to examine the specic activitiesin which managers engage with accounting information ininteraction with other information and other managers.

Process-based experimental studies

Given their research focus, many experimental studiesare designed to examine how variable X (e.g., an ABC re-port) is causally related to variable Y (e.g., cost-relatedjudgement performance) (Cardinaels, 2008). Althoughimportant in these contexts, such an approach is not ame-nable for understanding how accounting information isimplicated in the demanding, confusing, and unstructuredture to examine how participants develop their own pro-cesses and analyses (Engestrom, 2007). The focus is noton the product of development but on the very processby which higher forms (of development) are established(Vygotsky, 1978, p. 64).

Two examples illustrate the difference between a pro-cess-based approach and that adopted in much manage-ment accounting research. First, in many experimentaldesigns participants complete a pre-trial run to enhancevalidity. In contrast, a process-based approach views thepre-trial phase as critical in its own right as it is at thismoment that researchers can observe and analyse howparticipants engage in the task while they are still learningit after the pre-trial phase behaviour becomes fossilisedand opportunities for examining how participants strugglewith new tasks and new information is limited (Vygotsky,1978). In this way, the pre-trial phase provides opportuni-ties for examining how managers engage with accountinginformation in the more unstructured, uncertain momentsthat characterise much managerial work. Second, descrip-tions and explanations from participants about how theyarrived at particular decisions are likely to be very valuablein understanding how managers use accounting informa-tion. However, although prior research does acknowledgethe ability of experiments to report on the processes in-volved in arriving at decisions (Sprinkle, 2003, p. 289),rarely is this empirical material actually analysed. Forexample, Vera-Munoz, Shackell, and Buehner (2007) col-lected data on participants supporting calculations andrationales for their recommendations, but the empiricalanalysis is focused solely on the actual budget allocationsuggested by participants. Similarly, Coletti, Sedatole, andTowry (2005) required participants to write an essaydescribing the reasoning behind their resource allocationdecisions, but this data was used by other participants todetermine a trustworthiness score rather than as usefulempirical material in its own right. In contrast, Boland(1993) used participants written explanations for theirperformance evaluation decisions as a way to understandhow managers interpreted accounting reports. The pri-mary interest was not participants choices regardingwhether to promote manager East or manager West, buthow they used accounting reports to arrive at their judge-ments. From this analysis Boland (1993) generated signi-cant insight into how participants use accounting reportsto make performance evaluation judgements.

Overall, a process-based experimental approach couldprove fruitful in systematising observations (Hogarth,1982; Swieringa &Weick, 1982) of howmanagers go aboutusing accounting information in their work. I outline threespecic examples. First, to understand how managers usedifferent forms of information, existing experiments couldbe adapted by providing participants with different typesof information in addition to accounting information. Forexample, the Cardinaels (2008) study could be adaptedby providing participants with other forms of informationbeyond the ABC report, such as: a marketing report aboutsales trends in the specic market segment, macroeco-nomic information, an informal report from a subordinateabout relations with the three major customers, anemail exchange from a contact who works in a similar

organisation and has dealt with the three customers, pro-duction information showing excessive downtime on amanufacturing line used to produce one of the products,and the managers personal notebook of information. Par-ticipants are required to complete the specic task butthe focus is not on how decisions vary between experimen-tal conditions but on examining how participants use dif-ferent types of information to arrive at their decisions. Incontrast, existing studies are typically concerned with the

310 M. Hall / Accounting, Organizations and Society 35 (2010) 301315effects of the presence/absence of accounting informationor differences in its form/presentation, in settings whereno other information is provided. The approach describedhere allows the researcher to examine how participantsuse accounting information in the presence of othersources of information. For example, participants mayignore the accounting information, or make the same deci-sion regardless of its presence/absence.

Second, as knowledge of how managers use accountinginformation in verbal communications is limited, research-ers could construct settings involving several managers indiscussions about particular operational and/or strategicissues facing an organisation. In particular, in the interac-tive control context, a participant can act as the seniormanager who intervenes in the decision activities of sub-ordinates and other participants can act as the operatingsubordinates who explain and justify their data, assump-tions and action plans (Simons, 1990). Simonss case stud-ies and/or teaching cases could be used to develop thespecic experimental settings.7 The researcher can observeand analyse in detail the interpersonal dynamics involved inthis process and how participants use various types of infor-mation, including accounting, in their discussions andexplanations.

Third, to understand how managers use accountinginformation to identify problems, researchers could devel-op a simulation where groups of participants are managersat a hypothetical organisation. As part of the simulationmanagers arrive at work and are confronted with a varietyof issues that may or may not require their attention andaction, for example, a poor performing employee, possibledecline of a line of credit from a nancial institution, alooming industrial strike, a possible plant closure, prob-lems with a major supplier, etc (see Lombardo & McCall,1982, for further examples). Although building simulationsis difcult, time consuming and risky (Swieringa & Weick,1982), some simulations already exist that could providemodels for development.8 In this context, the researchercan observe how problems are recognised or ignored, theprocess of problem denition, and the way in which manag-ers search for and integrate accounting and other sources ofinformation (Hogarth, 1982; Swieringa & Weick, 1982). Sim-

7 In particular, see Simons (1995) and the Harvard Business Schoolteaching cases: Codman Shurtleff Inc. (9-187-081) and ATH MicroTechnologies (9-108-091).

8 For example: Looking Glass Inc. from the Centre for Creative Leader-ship (www.ccl.org/leadership/programs/LGEOverview.aspx, Lombardo &McCall, 1982), the Treasury Dealing Room from Monash University(www.buseco.monash.edu.au/aaf/research/tdr/, Lambert, Tant, & Watson,2008), and Harvard Business Schools online case simulations, e.g., BeerGame and Back Bay Battery (http://harvardbusinessonline.hbsp.har-vard.edu/hbsp/course_materials.jsp?_requestid=64691).ulations can focus on how managers use accounting infor-mation in chaotic, confusing and time-pressuredenvironments, where the objective of the task is to developknowledge about a potential problem rather than solve analready-constructed problem.

In all three situations the empirical material could beobtained in many ways: direct observation by the re-searcher as participants perform the task(s), audio/videorecording of conversations/explanations, written materialsused during task performance, pre- and post-experimentalquestionnaires, post-experimental explanatory state-ments, as well as specic decisions. Overall, adopting aprocess approach to experimental studies offers muchpromise for developing knowledge about how and whymanagers use accounting information.

Micro-practices in eld studies

In contrast to the experimental approach, eld studiesseek to examine how individuals engage with accountinginformation in organisational settings. Although the eldresearcher has less control over participants, there is muchgreater potential for open-ended interaction between theresearcher and participants and thus more opportunitiesto learn from unprompted actions that the researcherhad not previously considered (Ahrens & Chapman,2006). This advantage is evident in the studies of Simonet al. (1954), Preston (1986) and McKinnon and Bruns(1992) where the open-ended nature of the research ques-tions allowed the researchers to provide new insights intowhether and how managers used accounting informationin their respective organisational settings. For example,Simon et al. (1954) did not enter their study with theintention of examining the scorekeeping, attention-direct-ing and problem solving roles of accounting information;these roles emerged from their observation and analysisof how managers used accounting information in the eld.

Field studies offer the primary advantage of being ableto examine how managers use accounting information inculturally-constituted settings (see Ahrens & Dent, 1998and Ahrens & Chapman, 2006 for overviews of eld studymethodology). However, much prior research has been di-rected more towards the macro analysis of organisationalissues, with less focus on the micro-practices throughwhich accounting information is used by managers in theirwork (Ahrens & Chapman, 2007). Studies that focus onorganisational-level issues only are limited because theyare typically based upon assumptions about, rather thana detailed investigation of, managerial work behaviour(Covaleski et al., 2003; Hall, 2008a). As such, what is lessevident in these organisational accounts is an analysis ofthe detailed ways in which managers use accounting infor-mation to perform particular activities.

I argue that eld studies could focus more strongly onexamining the micro-practices involved in managers useof accounting information. Micro-practices are those spe-cic instances in which managers use accounting informa-tion in their work. For example: what do managers dowhen they use accounting information to identify prob-lems, surprises and opportunities? What do managers dowhen they integrate accounting and other forms of

information? And what do managers do when they relateaccounting information to specic operational concerns?These questions can best be answered by investigating

the researcher to not only understand the context but alsoto attempt to learn the practices themselves (Jrgensen &Messner, 2010). For example, to understand the actions of sci-entists, Alac and Hutchins (2004) required at least a basicknowledge of fMRI. The time and skill required for such an ap-

M. Hall / Accounting, Organizations and Society 35 (2010) 301315 311those instances when managers actually do these things;by looking and seeing how these activities are actuallyconducted. A stronger focus on investigating micro-prac-tices is also reective of an increased interest in examiningthe actual use of calculative practices by actors at work(Vollmer, Mennicken, & Preda, 2009).

An examination of micro-practices involves analysis ofthe specic activities that individuals engage in with on-goinganalysis of theoverall contextwithinwhichsuchactiv-ities takeplace (Alac&Hutchins, 2004;Hutchins, 1995; Lave,1988). In this approach, the analysis of context provides theknowledge necessary to make meaningful and valid inter-pretations of micro-practices (Alac & Hutchins, 2004). Forexample, Jonsson (1998) examined the use of accountinginformation in an individual team meeting at a Swedishmunicipality. The understanding of the history and contextof the municipality provided the necessary knowledge tointerpret how managers in the teammeeting used account-ing information to perform their work. More generally, Alacand Hutchins (2004, p. 633) explain the rationale of this ap-proach in their study of how scientists interpret scans fromfunctional magnetic resonance imaging (fMRI):

In order to nd out how scientists construct the mean-ing of complex fMRI images, we perform micro-analysison digital video recordings of scientic practices. Weuse these recordings to identify and track the represen-tations that the scientists bring into coordination withthe images in the process of making the images mean-ingful. We use the ndings from our long-term ethnog-raphy to provide warrants for our analytic judgmentsconcerning how things and events are made relevantto one another in a small instance of scientic practice.

In the context of accounting information and managerialwork, such an approach requires examination of the specicactivities in which accounting information is used, othersources of information at managers disposal, and the formsof communication throughwhich accounting information isengaged all of which are critical in increasing understand-ing of how managers use accounting information in per-forming their work activities. In this approach, informationabout the organisational context and institutional environ-ment plays a supporting role, i.e., it is necessary in order tounderstand and explain how and why managers useaccounting information in particular ways. For example,Jonsson (1998) described thewide range of information col-lected about the Swedish municipality and its institutionalenvironment, ranging from economic crises, organisationalrestructuring, external surveys, as well as interviews withmanagers. The study is not primarily concerned with thisempirical material, however, but is focused on interpretingand making judgements about the practices observed at aspecic meeting.9 Efforts to generate sufcient knowledgeof the context are by no means trivial, and will likely require

9 Of course, empirical material relating to wider organisational andinstitutional issues can be the primary focus in other, more macro-levelanalyses.proach clearly presents challenges to researchers, but alsoopens up extensive opportunities for much closer and moredirect examination of howmanagers use accounting informa-tion in their work.

Ethnographies of the activities of individual managersare likely to be particularly useful in examining howaccounting information is used (or not used) by managersin performing their work. Ethnographies are important fortwo reasons: rst, they allow the researcher to examinethe full range of a managers interactions with accountinginformation, other sources of information, and differentforms of communication; and, second, they provide oppor-tunities for investigating both the formal and informalinteractions that comprise the ways in which managersuse accounting information. For example, Preston (1986)found that managers kept informed primarily throughinformal interactions with other managers and throughtheir own observations and personal records.

As part of the ethnography researchers can select spe-cic practices for micro-analysis of how accounting infor-mation is used in managerial work. For example, withinthe context of a managers ongoing effort to developknowledge of the work environment, the researcher canexamine how an individual manager interprets specicaccounting reports and other forms of information. Thiscan help develop understanding of the relationship be-tween different forms of information (complements orsubstitutes), the relative advantages of accounting infor-mation vis--vis other information, as well as knowledgeof how managers actually use accounting information inperforming their work. In this approach, knowledge ofthe managerial context obtained through in-depth analysisof managerial work provides the insight necessary to makemeaningful interpretations of the micro-practices.10

One difculty in examining micro-practices in the eld,however, is that activities do not necessarily take place inpredictable locations. For example, managers may talk toeach other in corridors, review accounting informationwhilst on the train, have telephone and email conversa-tions, etc. In contrast, prior research on micro-practiceshas examined activities that take place in relatively pre-dictable locations, e.g., ship navigation (Hutchins, 1995)and interpreting MRI scans (Alac & Hutchins, 2004). Oneapproach to overcome this difculty is to focus on meet-ings as a specic setting in which it is very likely thataccounting information is used. Meetings are at specictimes/locations and often concern a variety of differentpurposes, thus affording the opportunity to examine howmanagers use accounting information in different ways.For example, how managers use accounting information

10 Further elaboration of the use of ethnographies in accounting researchis found in Ahrens and Mollona (2007), Ahrens (2004), 1997), and Jonssonand Macintosh (1997). For application to the managerial setting, see, forexample, Noordegraaf and Stewart (2000), Reed (1989), Kotter (1982) andHanaway (1989).

in the following roles (meetings) could be examined: tocommunicate strategies and organisational performance(status meetings), to construct and solve problems (projectmeetings for specic decisions), and to inform and be in-formed about on-going activities (recurring team meet-

ing research is too rigid and cautious (Hopwood, 2007),

312 M. Hall / Accounting, Organizations and Society 35 (2010) 301315an expanded research focus is important in encouraginginvestigations that consider the emergence of new rolesand possibilities for accounting, as well as reconsiderationof our existing conceptions of what accounting is and isused for. A broader scope of inquiry will require muchgreater engagement with on-going research in relatedelds, such as studies of how managers use information(e.g., Anderson, 2008), research that examines how infor-

11 For examples of studies that use meetings as an empirical setting, seeBoland and Pondy (1986) who examined weekly meetings of a high schooldistricts administrative council, and Jonsson (1998) who examined a topmanagement meeting of a city municipal district. For an overview ofresearch on meetings in organizations, see Jarzabkowski and Seidl (2008).ings). Jonsson and Macintoshs (1997, p. 384) approach tothe analysis of a lean-production team meeting can serveas a model for developing this kind of micro-analysis. Inthis way, the researcher can use meetings as a relativelypredictable location for conducting micro-analysis. Fur-thermore, the researcher can analyse the interpersonaldynamics, verbal exchanges of information, the incorpora-tion of different types of information that take place.11

Conclusion

In this paper I outlined an approach to the study ofaccounting information and managerial work that explic-itly incorporates prior research on how managers workand how they use accounting and other forms of informa-tion. This approach focuses on the role of accounting infor-mation in developing managers knowledge of the workenvironment, appreciation of how accounting informationinteracts with and relates to other information that manag-ers use, and a stronger focus on the ways in which manag-ers use accounting information in verbal communicationsand discussions. I examined the implications of this ap-proach for the types of accounting information that manag-ers nd, or could nd, helpful in their work, as well asconsideration of how existing research methods couldfruitfully be adapted to focus on the detailed activitiesthrough which managers engage with accounting informa-tion. Although few researchers have heeded Jonssons(1998) call to action, this paucity of studies provides muchscope for future research to develop our understanding ofhow and why managers use accounting information.

The rst and most immediate implication of this ap-proach is to broaden the scope of inquiry into what consti-tutes research on accounting information and managerialwork. The approach requires explicit examination of differ-ent roles for accounting information in managerial work(e.g., developing knowledge of the work environment), ofother forms of information and their interaction withaccounting information, and of the variety of ways in whichaccounting information can be communicated, particularlythrough verbal discussions. Given criticisms that account-mation is communicated in organisations (e.g., Nelson,2001), as well as studies of changes to the nature of mana-gerial work itself (e.g., Tengblad, 2006). In addition toenriching our own research, a likely benet of this engage-ment is increased interest in the outputs of accounting re-search from scholars in other disciplines whose researchalso seeks to understand how managers use informationin organisations. As productive conversations betweenaccounting and other researchers have been limited (e.g.,Chenhall, 2008), increasing interest is important if account-ing research and practice is to play amore inuential role inmanagement research and management more generally.

The second implication of this approach is to incorporateamuchstronger relativeperspective to thestudyofaccount-ing information and managerial work. As argued earlier,much management accounting research tends to focus onhow managers use accounting information only, whereasstudies of what managers do show that they rely on a widevariety of information sources in theirwork. This has impor-tant implications for severaleldsof inquiry inmanagementaccounting research. For example,whenwe investigate howthe use of accounting information is contingent upon differ-ent tasks, structures, cultures and environments (Chenhall,2003), we should also examine how other sources of infor-mation affect the use of accounting information in manage-rial work. When we investigate reliance on accountingperformancemeasures (Hartmann, 2000), we should exam-ine how accounting and other sources of information areused to make performance evaluation judgements. Andwhen we investigate the relationship between accountinginformation and strategy, we should examine how account-ing and other types of accounts and rationalities are used bymanagers in strategising (e.g., Jrgensen&Messner, 2010). Astronger relative perspective provides researchers with theopportunity to develop theory that focuses more directlyon accountings potentially unique role in managerial workin comparison to other sources of information in organisa-tions. For example, I propose that the translation of opera-tional activities into a single, nancial dimension, incomparison to other sources of information, is the most un-ique and helpful feature of accounting information forman-agerial work. However, the validity of this and otherexpectations about the unique advantages of accountingcan only be discovered by incorporating an analysis of othersources of information into management accounting stud-ies. In this way, researchers can developed better theoriesabout the potentially unique features of accounting infor-mation that managers nd or could nd helpful.

The third implication of this approach is to focusmore onhow managers actually work with accounting informationand less on its technical characteristics. As argued earlier,many recent developments in accounting (e.g., balancedscorecard, causalperformancemaps, activity-basedcosting)aredirected towardsmakingaccounting informationamorecomplete description of organisational activities. The ap-proach advanced in this paper, however, indicates that suchdevelopments are somewhat misplaced as they give inade-quateattention tomanagerial (in contrast to accounting)be-liefs about the quality and relevance of information. Fromthis perspective, accounting information is helpful for man-agerswhen it enables themtodo things,when it ismobilised

M. Hall / Accounting, Organizations and Society 35 (2010) 301315 313as a resource for action (Ahrens&Chapman,2007). This indi-cates that judgements about the quality and relevance ofaccounting information should relate primarily to whetherit helpsmanagers to carry out theirworkand less towhetherit adequately describes underlying organisational activities.As such, although technical improvements to the design ofaccounting information are important, they are somewhatsecondary to consideration of how managers draw onaccounting information to get things done.

The fourth and nal implication of the approach out-lined in this paper is the potential for much greaterengagement between researchers and practitioners.Accounting research, and management research more gen-erally, has been criticised for becoming far too removedfrom the practices and activities it seeks to investigateand illuminate (Hopwood, 2007; Johnson, Melin, & Whit-tington, 2003; Mintzberg, 1982). The specic areas for fu-ture investigation as outlined in the propositions, as wellas a greater focus on process-based experimental studiesand micro-practices in eld studies, are collectively direc-ted at much closer engagement with what managers do(and do not do) with accounting information in their work.Such an approach is likely to increase managers interest inthe outputs of accounting research as the research agendamore closely matches the lived worlds of organisationalactors (Johnson et al., 2003, p. 15). The approach is alsolikely to increase the likelihood of managers engaging inthe research process itself, whether as subjects in pro-cess-based experiments and/or as participants in eldstudies. Furthermore, greater engagement with managersis likely to help researchers to become aware more quicklyof changes to the way in which managers use accountinginformation, and thus respond more effectively to theshifting nature of the world of practice (Hopwood, 2007,p. 1370). In an environment of increasing disconnectionbetween researchers and practitioners, this is importantif accounting research is to inuence more strongly thepractice of accounting itself. In particular, researchers candevelop new accounting techniques, and make rene-ments to existing techniques, based on a more nuancedand pertinent understanding of how and why managersdo and/or might use accounting information in their work.

Acknowledgements

Thanks to the two anonymous reviewers, Casey Rowe,Robert Chenhall, David Smith, Chris Chapman, JenniferGrafton, Martin Messner, Bill McGuire, Alan Ramsay, An-ette Mikes, Yuval Millo, seminar participants at the LondonSchool of Economics and Monash University, and partici-pants at the Managing by the Numbers and Beyond trackof EURAM 2007, GMARS 2007 and AFAANZ 2007 for help-ful comments. I would particularly like to thank Wim Vander Stede for his numerous comments on prior versions ofthe paper.

References

Abernethy, M. A., & Brownell, P. (1999). The role of budgets inorganizations facing strategic change: An exploratory study.Accounting, Organizations and Society, 24, 189204.Abernethy, M. A., Horne, M., Lillis, A. M., Malina, M. A., & Selto, F. H.(2005). A multi-method approach to building causal performancemaps from expert knowledge. Management Accounting Research, 16,135155.

Ahrens, T. (1997). Talking accounting: An ethnography of managementknowledge in British and German brewers. Accounting, Organizationsand Society, 22(7), 617637.

Ahrens, T. (2004). Rening research questions in the course of negotiatingaccess for eldwork. In C. Humphrey & B. Lee (Eds.), The real life guideto accounting research: A behind the scenes view of using qualitativeresearch methods (pp. 295308). Oxford: Elsevier.

Ahrens, T., & Chapman, C. S. (2006). Doing qualitative eld research inmanagement accounting: Positioning data to contribute to theory.Accounting, Organizations and Society, 31(8), 819841.

Ahrens, T., & Chapman, C. S. (2007). Management accounting as practice.Accounting, Organizations and Society, 32(12), 127.

Ahrens, T., & Dent, J. F. (1998). Accounting and organizations: Realizingthe richness of eld research. Journal of Management AccountingResearch, 10, 139.

Ahrens, T., & Mollona, M. (2007). Organisational control as culturalpracticeA shop oor ethnography of a Shefeld steel mill.Accounting, Organizations and Society, 32(45), 305331.

Alac, M., & Hutchins, E. (2004). I see what you are saying: action ascognition in fMRI brain mapping practice. Journal of Cognition andCulture, 4(3), 629662.

Anderson, M. H. (2008). Social networks and the cognitive motivationto realize network opportunities: A study of managersinformation gathering behaviors. Journal of OrganizationalBehavior, 29, 5178.

Birnberg, J. G., Luft, J., & Shields, M. D. (2007). Psychology theory inmanagement accounting research. In C. S. Chapman, A. G. Hopwood, &M. D. Shields (Eds.), Handbook of management accounting research(pp. 113135). Oxford: Elsevier.

Bisbe, J., Batista-Foguet, J., & Chenhall, R. H. (2007). Dening managementaccounting constructs: A methodological note on the risks ofconceptual misspecication. Accounting, Organizations and Society,32, 789820.

Bisbe, J., & Otley, D. (2004). The effects of the interactive use ofmanagement control systems on product innovation. Accounting,Organizations and Society, 29(8), 709737.

Boland, R. J. (1993). Accounting and the interpretive act. Accounting,Organizations and Society, 18(23), 125146.

Boland, R. J., & Pondy, L. R. (1986). The micro dynamics of a budget-cutting process: Modes, models and structure. Accounting,Organizations and Society, 11(4/5), 403422.

Bowker, G. C., & Star, S. L. (1999). Sorting things out: Classication and itsconsequences. Cambridge, MA: MIT Press.

Burchell, S., Clubb, C., Hopwood, A., Hughes, A., & Nahapiet, N. (1980). Theroles of accounting in organizations and society. Accounting,Organizations and Society, 5, 527.

Cardinaels, E. (2008). The interplay between cost accounting knowledgeand presentation formats in cost-based decision-making. Accounting,Organizations and Society, 33, 582602.

Chapman, C. S. (1997). Reections on a contingent view of accounting.Accounting, Organizations and Society, 22(2), 189205.

Chapman, C. S. (1998). Accountants in organisational networks.Accounting, Organizations and Society, 23(8), 737766.

Chenhall, R. H. (2003). Management control systems design within itsorganizational context: Findings from contingency-based researchand directions for the future. Accounting, Organizations and Society, 28,127168.

Chenhall, R. H. (2005). Integrative strategic performance measurementsystems, strategic alignment of manufacturing, learning and strategicoutcomes: An exploratory study. Accounting, Organizations andSociety, 30(5), 395422.

Chenhall, R. H. (2008). Accounting for the horizontal organization: Areview essay. Accounting, Organizations and Society, 33, 517550.

Coletti, A. L., Sedatole, K. L., & Towry, K. L. (2005). The effect of controlsystems on trust and cooperation in collaborative environments. TheAccounting Review, 80(2), 477500.

Covaleski, M. A., Evans, J. H., Luft, J. L., & Shields, M. D. (2003). Budgetingresearch: Three theoretical perspectives and criteria for selectiveintegration. Journal of Management Accounting Research, 15, 349.

Daft, R. L., & Lengel, R. H. (1986). Organizational informationrequirements, media richness and structural design. ManagementScience, 32(5), 554571.

Dane, E., & Pratt, M. G. (2007). Exploring intuition and its role inmanagerial decision making. Academy of Management Review, 32(1),3354.

314 M. Hall / Accounting, Organizations and Society 35 (2010) 301315Daniels, H., Cole, M., & Wertsch, J. V. (2007). The Cambridge companion toVygotsky. Cambridge: Cambridge University Press.

Ditillo, A. (2004). Dealing with uncertainty in knowledge-intensive rms:the role of management control systems as knowledge integrationmechanisms. Accounting, Organizations and Society, 29, 401421.

Dent, J. F. (1991). Accounting and organizational cultures: A eld study ofthe emergence of a new organizational reality. Accounting,Organizations and Society, 16(8), 705732.

Dekker, H. C. (2004). Control of inter-organizational relationships:Evidence on appropriation concerns and coordination requirements.Accounting, Organizations and Society, 29(1), 2749.

Earl, M. J., & Hopwood, A. G. (1980). From management information toinformation management. In H. C. Lucas, F. F. Land, T. J. Lincoln, & K.Supper (Eds.), The information systems environment (pp. 313).Amsterdam: North-Holland Publishing Company.

Emsley, D. (2001). Redesigning variance analysis for problem solving.Management Accounting Research, 12, 2140.

Engestrom, Y. (2007). Putting Vygotsky to work: the change laboratory asan application of double simulation. In H. Daniels, M. Cole, & J. V.Wertsch (Eds.), The Cambridge companion to Vygotsky (pp. 363382).Cambridge: Cambridge University Press.

Espeland, W. N., & Sauder, M. (2007). Rankings and reactivity: How publicmeasures recreate social worlds. American Journal of Sociology, 113(1),140.

Feldman, M. S., & March, J. G. (1981). Information in organizations assignal and symbol. Administrative Science Quarterly, 26, 171186.

Fisher, J. (1995). Use of nonnancial performance measures. In S. M.Young (Ed.), Readings in management accounting (pp. 329335).Englewood Cliffs: Prentice Hall.

Galbraith, J. (1973). Designing complex organizations. Addison-Wesley:New York.

Gigerenzer, G. (2008). Gut feelings: The intelligence of the unconscious.London: Penguin.

Gilkey, R., & Kilts, C. (2007). Cognitive tness. Harvard Business Review, 1,5366.

Hales, C. (1999). Why do managers do what they do? Reconcilingevidence and theory in accounts of managerial work. British Journal ofManagement, 10, 335350.

Hales, C. P. (1986). What do managers do? A critical review of theevidence. Journal of Management Studies, 23(1), 88115.

Hall, M. (2008a). The effect of comprehensive performance measurementsystems on role clarity, psychological empowerment and managerialperformance. Accounting, Organizations and Society, 33, 141163.

Hall, M. (2008b). Management control systems and learning:investigating the effects of comprehensive performancemeasurement on managers mental model development. Workingpaper.

Hanaway, J. (1989). Managers managing: The workings of an administrativesystem. Oxford: Oxford University Press.

Hartmann, F. G. H. (2000). The appropriateness of RAPM: Toward thefurther development of theory. Accounting, Organizations and Society,25, 451482.

Henri, J. (2006). Management control systems and strategy: A resource-based perspective. Accounting, Organizations and Society, 31(6),529558.

Hogarth, R. M. (1982). Discussion of an assessment of laboratoryexperiments in accounting. Journal of Accounting Research(Suppl.20), 108116.

Hopwood, A. G. (1972). An empirical study of the role of accounting datain performance evaluation. Journal of Accounting Research, 156182(Empirical research in accounting supplement).

Hopwood, A. G. (1978). Toward an organizational perspective for thestudy of accounting and information systems. Accounting,Organizations and Society, 3, 313.

Hopwood, A. G. (1989). Organisational contingencies and accountingcongurations. In Fridman, B., Ostman, L. (Eds.), Accountingdevelopment-some perspectives-in honour of Sven-Eric Johansson (pp.2344). Stockholm.

Hopwood, A. G. (2007). Wither accounting research? The AccountingReview, 82, 13651374.

Horngren, C. T., Bhimani, A., Datar, S. M., & Foster, G. (2005). Managementand cost accounting. Harlow: Prentice Hall.

Hutchins, E. (1995). Cognition in the wild. Cambridge, MA: MIT Press.Isenberg, D. J. (1984). How senior managers think. Harvard Business

Review, NovemberDecember, 8190.Ittner, C. D., & Larcker, D. F. (2003). Coming up short on nonnancial

performance measurement. Harvard Business Review, 81(11), 8897.Jarzabkowski, P., & Seidl, D. (2008). The role of meetings in the social

practice of strategy. Organization Studies, 29(11), 13911426.Johnson, G., Melin, L., & Whittington, R. (2003). Micro strategy andstrategizing: towards an activity-based view. Journal of ManagementStudies, 40(1), 322.

Jonsson, S. (1998). Relate management accounting research to managerialwork! Accounting, Organizations and Society, 23(4), 411434.

Jonsson, S., & Macintosh, N. B. (1997). CATS, RATS, AND EARS: Making thecase for ethnographic accounting research. Accounting, Organizationsand Society, 22(34), 367386.

Jonsson, S., & Solli, R. (1994). Accounting talk in a caring setting.Management Accounting Research, 4(4), 301320.

Jrgensen, B., & Messner, M. (2010). Accounting and strategising: A casestudy from new product development. Accounting, Organizations an