CORPORATE CONTROL AND THE SPEED OF STRATEGIC BUSINESS UNIT … · 2017-07-26 · CORPORATE CONTROL AND THE SPEED OF STRATEGIC BUSINESS UNIT DECISION MAKING MAXIMILIAN KOWNATZKI Jetstar

CORPORATE CONTROL AND THE SPEED OF STRATEGICBUSINESS UNIT DECISION MAKING

MAXIMILIAN KOWNATZKIJetstar Airways

JORGE WALTERGeorge Washington University

STEVEN W. FLOYDUniversity of Massachusetts Amherst

CHRISTOPH LECHNERUniversity of St. Gallen

Decision speed has long been recognized as a critical determinant of firm performance,particularly in dynamic environments. Extending prior studies, which have largelyfocused on firm-level decision speed in small- and medium-sized organizations, thisstudy explores how control mechanisms set by corporate headquarters in multibusi-ness firms influence decision speed at the strategic business unit (SBU) level. Using amultimethod approach, we first inductively derive six types of corporate control,before deductively examining their effects on SBU-level decision speed in five inter-national multibusiness organizations. Our results suggest that three corporate controltypes enhance decision speed (goal setting, extrinsic incentives, and decision processcontrol); two have no effect (negative incentives and conflict resolution); and one hasa negative effect (strategy imposition). By integrating results from our qualitative andquantitative analyses, we are also able to identify transparency/alignment, outcomeorientation, participation, trust, and timely feedback as the key mechanisms account-ing for these effects.

Decision speed has long been recognized as acritical factor explaining firm performance, espe-cially in dynamic environments (Baum & Wally,2003; Eisenhardt, 1989b). In their sample of newbiotechnology firms, for instance, Judge and Miller(1991) found that decision speed explained nearly38 percent of the variance in sales growth and over42 percent of the variance in profitability. Progres-

sively competitive environments, shortened life cy-cles, and increasingly global markets are likely tofurther enhance the importance of decision speedfor firm adaptation and performance (Nadler &Tushman, 1999). As Bower and Hout so aptly illus-trate, organizations with fast decision processes are“like World War II fighter pilots—they win by pre-empting the opposition’s moves” (1988: 110). Con-versely, if firms fail to align their businesses withenvironmental changes in a timely manner, theyrisk being outpaced by more agile competitors(D’Aveni, Dagnino, & Smith, 2010; Nadkarni & Barr,2008). Eisenhardt’s (1989b) study of computer firmsfurther refuted the idea of an inevitable trade-off be-tween decision speed and decision quality, at least indynamic environments, by finding that fast decisionmakers use more, not less, information than doslow decision makers and develop more, not fewer,alternatives, thereby increasing decision quality.

Beyond these important findings, however, unre-solved questions remain. Most insights on decision

The first two authors contributed equally and arelisted alphabetically. Please direct all correspondence tothe second author.

We gratefully acknowledge the valuable suggestionsby Claus Jacobs, Markus Kreutzer, Günter Müller-Stew-ens, Rhonda Reger, Torsten Schmid, Pamela Tierney, andMargarethe Wiersema. We also greatly appreciate thesupport of the executives at the companies we studied aswell as the constructive feedback provided by AssociateEditor Gerry McNamara and three anonymous reviewers.The first author is thankful for a grant from the SwissNational Science Foundation that supported part of thisresearch.

1295

� Academy of Management Journal2013, Vol. 56, No. 5, 1295–1324.http://dx.doi.org/10.5465/amj.2011.0804

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s expresswritten permission. Users may print, download, or email articles for individual use only.

speed have been generated from studying new ven-tures (Forbes, 2005; Perlow, Okhuysen, & Repen-ning, 2002) and small or medium-sized companies(Baum & Wally, 2003; Eisenhardt, 1989b; Wally &Baum, 1994). Their relatively small and simple or-ganizational structures likely put these firms in anadvantageous position when it comes to makingfast strategic decisions, particularly as compared tolarge, complex, and multilayered organizations(Judge & Miller, 1991; March & Olsen, 1976). How-ever, even in dynamic industries, where decisionspeed is imperative, many corporations have ad-opted multibusiness organizational structures (Ga-lunic & Eisenhardt, 2001; Martin & Eisenhardt,2010). In these organizations, decision processes atlower levels, such as strategic business units(SBUs), are constrained by structural and controlrequirements at the corporate level (Baum & Wally,2003; Papadakis, Lioukas, & Chambers, 1998; Sut-cliffe & McNamara, 2001). How these constraintsaffect SBU-level decision speed remains an unan-swered question. On the one hand, corporate head-quarters might put additional burdens on SBUs,slowing down their decision processes. For in-stance, SBUs may need corporate approval for ma-jor investments, and they may be obliged to partic-ipate in strategic planning and control exercises,which consumes time and energy. On the otherhand, the corporate level may accelerate SBUs’ de-cision making by giving advice and/or directionand by initiating and sustaining momentum fordifficult strategic decisions, such as reorganiza-tions or layoffs.

Moreover, while prior studies have analyzed theeffects of macro- (or firm-) level, structural charac-teristics, such as centralization and formalization(Baum & Wally, 2003; Gupta, 1987; Wally & Baum,1994), scholars do not yet sufficiently understandthe microlevel mechanisms connecting organiza-tional context and decision speed. This is problem-atic for multibusiness organizations, as they mayface trade-offs in their choices between macrolevelcharacteristics. It is known, for instance, that cen-tralization generally allows for faster decisions(Baum & Wally, 2003; Wally & Baum, 1994), butsome studies have argued that decentralized deci-sion making allows a multibusiness firm to respondmore quickly to changing local conditions (Mintz-berg, 1979) and is therefore more appropriate indynamic environments (Mintzberg, 1981).

More broadly, there is no empirically groundedtaxonomy of the most important means throughwhich corporate headquarters attempt to control

SBUs’ decision processes. While the role of corpo-rate headquarters in multibusiness firms has beenexamined from a variety of theoretical perspec-tives, little cross-fertilization has occurred. As aresult, existing conceptualizations of corporatecontrol and its relationship to SBU decision mak-ing remain partial and fragmented (Hart, 1992; Hart& Banbury, 1994), and scholars have stressed theneed for a more integrated framework (e.g., Collis,Young, & Goold, 2007).

We address these open issues with the followingresearch question: How and why do different typesof corporate control—defined as corporate head-quarters’ attempts to manage or influence the pro-cess, content, and/or outcome of strategy making intheir SBUs—impact the speed of SBU-level deci-sion processes? Our results suggest that there are atleast six corporate control types relevant to thecontext of SBU-level decision processes. Three ofthese—goal setting, extrinsic incentives, and deci-sion process control—positively influence SBU-level decision speed; two—negative incentives andconflict resolution—have no effect; and one—strat-egy imposition—slows down SBU decision making.Importantly, the positive effects on decision speedare explained by the extent to which corporatecontrols contribute to transparency/alignment, out-come orientation, participation, trust, and timelyfeedback in a corporate headquarters-SBU relation-ship. These mechanisms temper the need for coher-ence in corporate strategy making with a desire topreserve SBU autonomy and responsiveness to fast-paced competitive environments.

Our study contributes to previous research inthree ways. First, we distill the main dimensions ofcorporate control from a comprehensive review ofthe literature on multibusiness firms. Then, draw-ing on qualitative data from five multibusinessfirms, we use these dimensions as a lens to induc-tively derive six types of corporate control overSBU decision processes. While resonating withprior studies, the empirically grounded corporatecontrol types identified here serve to consolidatean otherwise fragmented literature around thequestion of how corporate controls influence SBUdecision processes.

Second, by conducting a quantitative analysis ofthe microlevel linkages between corporate controland decision speed, and by supplementing it withour qualitative data, we can show both how andwhy different types of corporate control influenceSBU-level decision speed. After we identify howthey influence SBU decision speed, we examine

1296 OctoberAcademy of Management Journal

why this influence exists by using our “qualitativedata to draw deeper insights from the quantitativedata” (Bansal & Corley, 2011: 235) and by identify-ing the mechanisms responsible for the relation-ships between corporate control types and SBU-level decision speed.

Third, and more broadly, our findings contributeto research on multibusiness firms by offering in-sight into the design of appropriate corporate con-trols. Prior work has concluded that related multi-business firms often outperform those with otherforms of diversification because they provide supe-rior control over opportunism, better decision mak-ing, and enhanced value creation through cross-SBU coordination (Martin & Eisenhardt, 2010). Butresearch has also established that these benefitscome at substantial bureaucratic costs (Collis &Montgomery, 1998; Jones & Hill, 1988). Our studysuggests how and to what degree corporate head-quarters can design corporate controls that realizethe benefits of related diversification while main-taining fast decision processes.

THEORETICAL BACKGROUND

Decision Speed

The speed of strategic decision making is gener-ally defined in prior work as the time between thefirst reference to deliberate action and the time atwhich a commitment to act is made (Eisenhardt,1989b). Empirical studies have provided amplesupport for the benefits of decision speed on firmperformance or growth, particularly in dynamic en-vironments (Baum & Wally, 2003; Eisenhardt,1989b; Judge & Miller, 1991; Wally & Baum, 1994).1

High decision speed allows firms to rapidly re-spond to competitors’ actions in the marketplace(Bourgeois & Eisenhardt, 1988; Souitaris & Maestro,2010), to exploit short-lived strategic opportunitiesbefore they disappear or are exploited by competi-tors (D’Aveni et al., 2010), and to reap first moveradvantages by becoming early adopters of new

products, technologies, and business models (Maka-dok, 1998).

A focus on decision speed entails the obviouspitfall of comprehensive information gathering andanalysis being sacrificed to gain speed, resulting inbad decisions and lower performance (Kahneman,Slovic, & Tversky, 1982). However, prior studieshave stressed the fact that “often, a late decision,whether or not it is correct, is a useless decisionand may have severe consequences for the organi-zation. This is particularly true in volatile and rap-idly changing environments” (Lin & Carley, 1997:220). Eisenhardt (1989b) further found that fast de-cision processes do not necessarily indicate super-ficial analysis and processing. Instead, the mostsuccessful decision processes she examined werefast and comprehensive, with decision makers re-lying on real-time information and consideringmultiple alternatives simultaneously, thereby ac-celerating their cognitive processing without jeop-ardizing decision speed (see also Judge &Miller, 1991).

Subsequent studies have examined a variety offactors that influence decision speed and foundsupport for the effects of environmental factors(Baum & Wally, 2003), management factors (Forbes,2005; Judge & Miller, 1991; Wally & Baum, 1994),and decision process factors (Eisenhardt, 1989b;Judge & Miller, 1991). In addition, two studies re-port negative effects for firm size on the speed ofstrategic decisions (Baum & Wally, 2003; Wally &Baum, 1994). More importantly for the focus of ourstudy, decision speed, like any other aspect of thedecision process, is influenced by the organization-al context in which decision processes are embed-ded (Baum & Wally, 2003; Papadakis et al., 1998;Sutcliffe & McNamara, 2001). Several studies haveexamined the effects of this context but largelyconcentrated on macrolevel, structural dimensions,such as centralization and formalization. Resultsindicate that a high degree of centralization—de-fined as concentration of authority or decision-making power in a firm (Eisenhardt, 1989b)—isassociated with faster decision processes (Wally &Baum, 1994), particularly for nonroutine, strategicdecisions (Baum & Wally, 2003). That is, more au-tocratic decision makers rely less on consultation(Eisenhardt, 1989b) and involve fewer people in adecision process, thereby reducing the potential forconflict and the need for information sharing andconsensus formation (Pfeffer, 1980). Results fromSiggelkow and Rivkin’s (2005) agent-based simula-tion, however, suggest that in turbulent but simple

1 Two notable exceptions are Forbes (2005), whofound that decision speed was positively related to sub-sequent new venture closure, and Perlow et al. (2002),who identified a potential pathology for start-up organi-zations engaged in fast decision making, which eventu-ally led to the bankruptcy of the company they studied.Both studies, however, focused on fledgling firms, whichrepresent a different type of organization than the multi-business firms that are the focus of our study.

2013 1297Kownatzki, Walter, Floyd, and Lechner

environments, in which sophisticated coordinationmechanisms are unnecessary, centralization actu-ally slows down decision making, whereas in tur-bulent and complex environments, which require abalance of fast decision making and diverse search,centralization is more beneficial.

With regard to formalization—defined as the ex-tent to which firm policies, job descriptions, organ-ization charts, plans and objective setting systemsare explicitly articulated (Fredrickson & Iaquinto,1989)—it has been shown that greater formalizationslows down decision making (Wally & Baum,1994), particularly for nonroutine, strategic deci-sions (Baum & Wally, 2003). Typically, more for-mal processes mandate the collection of largeramounts of data and more thorough analyses thanless formal ones (Fredrickson & Mitchell, 1984),and all else being equal, consume more time.

Taken together, these studies have demonstratedthe significance of decision speed for firm perfor-mance and identified a number of variables thatinfluence it. Research has focused on firms as awhole and analyzed decision processes at the topfollowing the “upper echelons” tradition (Ham-brick & Mason, 1984). While this may be appropri-ate for small or midsized firms that operate in asingle industry—such as those in Eisenhardt’s(1989b) landmark study on decision speed wherethe sample was comprised of single business firmswith an average of 229 employees—such a focus isout of step with organizational realities in large,multibusiness corporations. In these large firms,decision processes are more complex and dis-persed over several levels. At a minimum, the pro-cess of strategic decision making operates acrosstwo interdependent managerial levels: corporateheadquarters and business units. Although eachmay have authority over certain aspects of strategicdecisions, control mechanisms at the corporatelevel are part of the context of SBU-level decisionprocesses. It is this influence on SBU decisionspeed that comprises the focus of this study.

The Role of Corporate Headquarters

Following Chandler’s (1962) landmark study of themultibusiness (M-form) structure, a rich and di-verse literature has developed on the managementof multibusiness firms. Such firms delegate operat-ing responsibilities to SBUs that function as organ-izational entities under the financial control andownership of corporate management (Baysinger &Hoskisson, 1990). Typically, SBUs have predefined

product-market boundaries, are equipped with sub-stantial resources for conducting their business op-erations, have the ability to devise and executetheir strategies within the constraints imposed bycorporate management, and are responsible for fi-nancial measures such as profit and loss or returnon invested capital (Goold, Campbell, & Alexan-der, 1994).

The corporate level, on the other hand, “acts asan intermediary, influencing the decisions andstrategies pursued by the businesses and standingbetween the businesses and those who provide cap-ital for their use” (Goold et al., 1994: 12). Corporatemanagement is vital for providing goals, directions,guidelines, structures, and control systems to SBUmanagers (Burgelman, 1983) that facilitate the per-formance of their assigned tasks and responsibili-ties (Collis & Montgomery, 1998; Goold et al.,1994). Conceptually, the corporate level moderatesthe relationship between SBUs and outcome vari-ables, by directly or indirectly facilitating or inhib-iting the evolution and performance of SBUs(Golden, 1992).

A review of the large and diverse body of work onmultibusiness firms suggests three dimensionsalong which the corporate level attempts to exertcontrol over SBU-level decision processes. The firsttwo dimensions originate from control theory (e.g.,Ouchi & Maguire, 1975), which has distinguishedoutcome control (focused on the measurement ofthe outcomes of behavior) from behavior control(based on direct, personal surveillance of behav-ior). In multibusiness firms, outcome control2 ischaracterized by corporate headquarters establish-ing short-term, objective, and predominantly finan-cial goals for each SBU, such as return on investedcapital, and monitoring SBU performance againstthose goals (Collis & Montgomery, 1998; Goold etal., 1994; Hill & Hoskisson, 1987). Headquartersusing outcome control tend to not formally reviewlong-term plans, and responsibility for strategy for-mulation is usually delegated to SBU management(Chung et al., 2000; Goold & Campbell, 1987a,1987b). Outcome control is relatively easy to im-plement and places fewer demands on corporateheadquarters (Collis & Montgomery, 1998). It also

2 Outcome control is sometimes also referred to asfinancial control (Chung, Gibbons, & Schoch, 2000;Goold & Campbell, 1987a, 1987b) or budgetary control(Goold & Quinn, 1990), and resonates with Mintzberg’s(1979) standardization of work outcomes.


motivates SBUs to focus on improving their finan-cial performance and abandoning ineffective strat-egies. The disadvantages are limited flexibility anda potential bias against long-term strategies and risktaking (Goold & Campbell, 1987a). Outcome controlis most effective for highly diversified firms con-sisting of discrete SBUs that share few commonresources (Goold & Campbell, 1987b; Hill & Hos-kisson, 1987; Jones & Hill, 1988; Vancil, 1978) butcompete for resource allocations from the corporatelevel (Hill, Hitt, & Hoskisson, 1992). In contrast,because it does not speak to how goals areachieved, outcome control may be less effectivewhen it comes to realizing and exploiting synergiesamong businesses (Hill & Hoskisson, 1987).

A headquarters using behavior control,3 on theother hand, relies on subjective, strategically rele-vant criteria to assess SBU activity and measuresperformance in the context of long-term progresstoward the development of a particular way of do-ing things (Collis & Montgomery, 1998; Goold et al.,1994; Gupta, 1987; Hoskisson & Hitt, 1988). Head-quarters actively participates in and influences theprocess of SBU-level strategy formulation (Chunget al., 2000), but without imposing a specific strat-egy. Behavior control not only provides checks andbalances for SBU strategy development, but alsoprovides a common vocabulary that facilitates dia-logue among SBU managers and fosters the creationof ambitious strategies. Moreover, if behavior ratherthan outcome is the focus, controls may provide abuffer for SBUs from external capital market pres-sures, at least in the short term (Goold & Campbell,1987b). On the downside, however, behavior con-trol places more demands on an organization andgenerally leads to somewhat larger corporate infra-structures (Collis & Montgomery, 1998). More im-portantly, while behavior or “strategic” control(Baysinger & Hoskisson, 1990) constitutes attemptsto steer SBU strategies by filtering them throughcorporate-level understandings of what is appropri-ate ex ante, thereby allowing SBUs some degree offlexibility in proposing a strategy, it involves vet-ting strategies with corporate headquarters prior toacting on them, leading to diminished SBU flexi-bility that may inhibit SBUs’ abilities to respond

quickly to changing market or environmental con-ditions. Behavior control can also lead to motiva-tion problems at the SBU level, as the involvementof different hierarchical levels can make the pro-cess cumbersome, overly bureaucratic, frustratingand costly, and therefore less encouraging for tak-ing ownership of decisions (Goold & Campbell,1987b). Behavior control is most effective for firmswith relatively low degrees of diversification(Goold & Campbell, 1987a), in which SBU successrequires extensive coordination, cooperation, andresource sharing (Hill et al., 1992; Hill & Hoskisson,1987; Lorsch & Allen, 1973; Vancil, 1978).

The unique characteristics of multibusinessfirms suggest a third dimension, content control, orthe degree to which corporate headquarters exertsinfluence on the actual substance of strategy at theSBU level (Muralidharan, 1997). On the one hand,most prior studies have emphasized the advantagesof retaining content autonomy at the SBU level,reasoning that this level is most likely to have ac-cess to the pertinent information (Campbell, 1999;Eisenmann, 2005). On the other hand, when coor-dination, cooperation, and resource sharing areneeded among SBUs in more related forms of di-versification, the situation likely requires at leastsome degree of headquarters influence over thesubstance of SBU-level strategies (Gupta & Govin-darajan, 1986; Hill et al., 1992; Hill & Hoskis-son, 1987).

Outcome, behavior, and content forms of controltogether represent recurring themes in the litera-ture on multibusiness firms and also incorporatecentralization, in the form of content control, aswell as formalization, in the form of behavior con-trol. The extent of outcome, behavior, and contentcontrol therefore provides a comprehensive way toconceptualize the corporate control context withinwhich SBU strategic decision making occurs.

Distilling the existing literature into three dimen-sions, however, does not necessarily make it easy toempirically examine the actual type and degree ofcontrol that headquarters asserts over SBU decisionmaking. Our brief examination of each of the di-mensions makes it apparent that there are multipleapproaches to implementing one or the other di-mension of corporate control. There are even im-portant variants within each dimension. Outcomecontrols may focus on a wider or narrower range offinancial and operational metrics, for example, andmay or may not be tied to performance incentives.In addition to the grainy character of control as anempirical reality for each of the three dimensions, a

3 Behavior control (Ouchi & Maguire, 1975) is some-times also referred to as “strategic planning” (Chung etal., 2000; Goold & Campbell, 1987a, 1987b; Goold &Quinn, 1990), and resonates with Mintzberg’s (1979)“standardization of work processes” and Child’s (1984)“bureaucratic control.”


further empirical likelihood is that organizationstend to mix control dimensions—often drawing onall three, at least to some extent (Goold & Campbell,1987b; Goold et al., 1994). This situation presents achallenge for researchers seeking valid and parsi-monious ways to identify and measure corporatecontrol as it relates to SBU decision making.

Although our framework suggests a large numberof possible combinations of outcome, behavior, andcontent control, it is likely that a much smallernumber of empirical types have evolved withinmultibusiness firms. By types we mean the set ofreal-world means of corporate control over SBUdecision processes that multibusiness firms actu-ally use. Underlying such an approach is the as-sumption “that elements of strategy, structure andenvironment often coalesce or configure into amanageable number of common, predictively use-ful types” (Miller, 1986: 235). This assumption issupported by theory and evidence in populationecology (Aldrich, 1979) showing that, over time, anenvironment tends to select out many organization-al forms, leaving relatively few well-adapted onesto survive over the longer term (Tushman & Ro-manelli, 1983). Typologies have also been widelyused in the strategic management literature to char-acterize strategies (Porter, 1980), strategic orienta-tions (Miles & Snow, 1978), strategic decision mak-ing (Fredrickson & Mitchell, 1984), strategy-makingprocesses (Hart, 1992), and strategic roles (Floyd &Lane, 2000). Once we identify these corporate con-trol types, we can then use them in the second stageof our research to examine the effects of corporatecontrol on SBU-level decision speed.

RESEARCH DESIGN

To develop a well-grounded set of corporate con-trol types and then examine how these influenceSBU decision speed, we employed a mixed-methoddesign (Yauch & Steudel, 2003) with two stages ofdata collection and analysis. In the first study, wechose an inductive approach, since our researchquestion addresses a little-explored process phe-nomenon for which existing theory does not pro-vide sufficient grounds for hypothesis develop-ment (Eisenhardt & Graebner, 2007; Langley, 1999;Pratt, 2009). As Fredrickson (1986) has outlined,any nonsimplistic understanding of decision pro-cesses requires interviews and qualitative analysesto develop a rich understanding of variables andtheir manifestations in actual decision processes.In particular, while we could have used survey-

based measures to parse the influence of individualcontrol dimensions, this would leave unansweredquestions about how organizations actually com-bine control dimensions, and how such combina-tions influence SBU decision speed. For these rea-sons, our first study was a qualitative, comparativecase analysis (Eisenhardt & Graebner, 2007). Weinterpret our data through the lens of the threedimensions outlined above to empirically derive ataxonomy of corporate control types. We ap-proached the data knowing what dimensions tolook for but without a clear picture of which par-ticular types would emerge.

In the second study, we used a quantitative, rep-ertory grid analysis (Kelly, 1955; Reger, 1990;Wright, 2006) to establish how these corporate con-trol types affect SBU-level decision speed. Thefindings from our repertory grid analysis were thenpassed back into the comparative case analysis(Wright, 2006) for confirmation and elaboration ofthe grid results. Integrating the results from bothstudies, we further develop insight into why se-lected types affect decision speed—that is, into themicrolevel mechanisms behind their influence.This serves as the basis for an intermediate theoret-ical model (Edmondson & McManus, 2007)wherein we develop a set of propositions about theimpact of corporate control types on SBU-level de-cision speed and the mechanisms that account forthese relationships.

STUDY 1: IDENTIFYING TYPES OFCORPORATE CONTROL

Sample Selection

In line with prior work, we relied on theoreticalsampling and chose sample companies on the basisof their suitability for illuminating and extendingrelationships among variables, and not for statisti-cal reasons (Eisenhardt & Graebner, 2007; Glaser &Strauss, 1967). As the effects of decision speed arestronger in industries characterized by higher de-grees of environmental dynamism (Eisenhardt,1989b; Judge & Miller, 1991), all of our samplefirms were operating in dynamic industries charac-terized by deregulation, liberalization, “e-com-merce,” globalization, new market entrants, andintense rivalry (Dess & Beard, 1984). Having such abroad sample of companies from multiple coun-tries and multiple industries with different degreesand sources of industry dynamism, and companieswith a wide range of sizes substantially enhances


the generalizability of our findings. Table 1 pres-ents an overview of our five sample companies.

Each sample company was structured as a mul-tibusiness organizational form according to Hill’s(1988) categorization scheme (see also Markides &Williamson, 1996). From interviews and secondarydata, we concluded that each business unit thatwas part of one of our five sample companies was adistinct and semiautonomous organizational en-tity4 comprising two or more individual unitsequipped with the necessary resources, responsi-bilities, and discretion to conduct their businessoperations. While all of the business units in oursample thus met the existential criteria for SBUs(Baysinger & Hoskisson, 1990; Goold et al., 1994),the firms themselves used different language to de-scribe their units (e.g., segments, strategic businessareas, etc.; see Table 1 for details). All companiesfurther derived 70 percent or less of their revenuesfrom their dominant SBU and internally shared acommon set of core resources or skills; put differ-ently, all the sampled firms fall into the related-constrained or related-linked category of diversifi-cation strategy (Rumelt, 1982).5

Similarities in multilevel organizational struc-ture allow for meaningful comparisons across cor-porate control contexts—and therefore provide astronger base for theory building—while the differ-ences between industries and organizations allowfor higher generalizability (Eisenhardt & Graebner,2007; Yin, 2003). It is important to recognize, how-ever, that our analysis focuses on similarities ratherthan differences among the five companies. Thus,although our design allowed us to use each addi-tional case to replicate or extend the emergent the-ory (Eisenhardt, 1989a; Yin, 2003), it is less sensi-tive to detecting situational contingencies inrelationships, which represents a limitation of thisapproach.

Interview and Coding Procedures

We conducted 14 semistructured, face-to-face in-terviews with an average duration of two and ahalf hours. These interviews established informantand firm background before focusing on the organ-

izations’ strategy processes, corporate control, andits relationship to decision processes and decisionspeed. The Appendix outlines our interview ques-tions. To ensure the decisions we examined wererepresentative of the process by which major deci-sions were made at a given SBU, we followed priorstudies (e.g., Eisenhardt, 1989b; Forbes, 2005;Judge & Miller, 1991) and asked our informants tofocus on strategic decisions, defined as decisionsthat are nonprogrammable, involve strategic posi-tioning of an SBU, and have high stakes in terms ofthe commitment of substantial resources (Hickson,Wilson, Cray, Mallory, & Butler, 1986; Mintzberg,1979; Mintzberg, Raisinghani, & Théorêt, 1976).

To ensure the high diversity and abundantsources of data required for theoretical saturation(Eisenhardt, 1989a), we interviewed executives atcorporate and SBU levels at each of the five samplecompanies and kept the number and diversity ofinterviewed SBUs as high as possible. We thus notonly ensured that both hierarchical levels withtheir specific perceptions of corporate control andits impact on SBUs were included, but also that itwas “unlikely that these varied informants will en-gage in convergent retrospective sensemakingand/or impression management” (Eisenhardt &Graebner, 2007: 28). All informants held senior po-sitions at the SBUs and at headquarters, such ashead of the SBU, country manager, chief operatingofficer, managing director, etc., and were thereforewell qualified to provide an in-depth and reliableassessment of the corporate control context and itseffects on SBUs.

Each interview was carried out by at least tworesearchers and followed the methodologicalguidelines established by Strauss and colleagues(Strauss, 1987; Strauss & Corbin, 1998) and refinedin recent comparative case analyses (Graebner,2009; Martin & Eisenhardt, 2010). In particular, weused interview techniques that prior research hasshown to yield accurate informant responses, suchas producing a step-by-step chronology of events(i.e., event tracking) as well as emphasizing factsand avoiding broad speculation (i.e., courtroomquestioning), and we assured confidentiality to en-courage candor. Questions were raised in the openrather than the closed mode, and as informantsmade their observations, we asked follow-up ques-tions to elicit richer and more detailed descriptions(Strauss, 1987; Strauss & Corbin, 1998). We alsoprepared personal notes and memos subsequent toeach interview and kept an interview diary to note

4 In particular, the SBUs within the two consultingcompanies in our sample were organized as profitcenters.

5 The one exception to this is Aviation Co., whichexhibits elements of both related diversification and ver-tical integration (see Table 1).


TA

BL

E1

Sam

ple

Des

crip

tion

a

Com

pan

yB

usi

nes

sD

escr

ipti

onL

ocat

ion

Ap

pro

xim

ate

Rev

enu

esb

Ap

pro

xim

ate

Nu

mbe

rof

Em

plo

yees

Org

aniz

atio

nal

Str

uct

ure

Ind

ust

ryE

nvi

ron

men

t(S

ourc

esof

Dyn

amis

m)

Cor

por

ate

Lev

elS

BU

Lev

el

Man

ufa

ctu

rin

gC

o.S

erve

sco

nsu

mer

ind

ust

ries

,u

tili

ties

,an

dth

eoi

lan

dga

sm

arke

tse

ctor

s

Hea

dqu

arte

rsin

Sw

itze

rlan

d;

oper

atin

gin

mor

eth

an10

0co

un

trie

s

23,0

0016

0,00

0E

xecu

tive

com

mit

tee:

11se

nio

rm

anag

ers

(CE

O,

CF

O,

exec

uti

vevi

cep

resi

den

ts)

Six

segm

ents

:(1

)A

uto

mat

ion

(2)

Pow

ertr

ansm

issi

on(3

)P

ower

dis

trib

uti

on(4

)B

uil

din

gte

chn

olog

ies

(5)

Oil

/gas

and

pet

roch

emic

als

(6)

Fin

anci

alse

rvic

es

●D

ereg

ula

tion

,li

bera

liza

tion

,gl

obal

izat

ion

●N

ewm

arke

ten

tran

ts●

Hig

hri

valr

yw

ith

inin

du

stry

Ad

vert

isin

gIn

c.P

lan

nin

g,cr

eati

on,

sup

ervi

sion

,an

dp

laci

ng

ofad

vert

isin

g

Hea

dqu

arte

rsin

the

US

;op

erat

ing

in90

cou

ntr

ies

9,00

010

,000

Glo

bal

hea

dqu

arte

rs,

regi

onal

hea

dqu

arte

rs

Six

busi

nes

su

nit

s/ag

enci

es:

(1)

Ad

vert

isin

g(2

)P

ubl

icre

lati

ons

(3)

Pu

blic

affa

irs

(4)

Dir

ect

mar

keti

ng

(5)

Inte

rnet

(6)

Med

iase

rvic

es

●G

loba

liza

tion

,ch

angi

ng

cust

omer

dem

and

s●

New

mar

ket

entr

ants

●H

igh

riva

lry

wit

hin

ind

ust

ry

ITC

onsu

ltin

gIn

c.C

onsu

ltin

gco

mp

any

spec

iali

zed

ine-

busi

nes

s:G

loba

lIT

man

agem

ent,

cust

omer

rela

tion

ship

man

agem

ent,

sup

ply

chai

nm

anag

emen

t,e-

com

mer

ce,

and

ente

rpri

sem

anag

emen

t

Hea

dqu

arte

rsin

Sw

itze

rlan

d;

oper

atin

gin

Eu

rop

e,A

sia,

and

the

US

7560

0E

xecu

tive

com

mit

tee

of 40m

embe

rs:

Hea

ds

ofbu

sin

ess

un

its,

CE

O,

CF

O,

CT

O,

chie

fof

corp

orat

ese

rvic

ece

nte

r(C

CC

)

Fou

rbu

sin

ess

un

its

(in

du

stry

pra

ctic

egr

oup

s):

(1)

Fin

anci

alin

stit

uti

ons

(2)

Con

sum

erp

rod

uct

s,tr

ade

and

serv

ice

(3)

Ch

emic

als

and

life

scie

nce

(4)

Man

ufa

ctu

rin

g

●L

iber

aliz

atio

n●

New

mar

ket

entr

ants

●E

-com

mer

ce:

chan

gin

gm

arke

tsan

dd

eman

ds

Con

tin

ued

TA

BL

E1

(Con

tin

ued

)

Com

pan

yB

usi

nes

sD

escr

ipti

onL

ocat

ion

Ap

pro

xim

ate

Rev

enu

esb

Ap

pro

xim

ate

Nu

mbe

rof

Em

plo

yees

Org

aniz

atio

nal

Str

uct

ure

Ind

ust

ryE

nvi

ron

men

t(S

ourc

esof

Dyn

amis

m)

Cor

por

ate

Lev

elS

BU

Lev

el

Avi

atio

nC

o.C

ore

busi

nes

s:F

lyin

gp

asse

nge

rsan

dca

rgo;

oth

erav

iati

on-

rela

ted

serv

ices

such

asin

-fli

ght

cate

rin

g,ai

rcra

ftm

ain

ten

ance

and

over

hau

l,an

dtr

avel

-re

late

din

form

atio

nte

chn

olog

y

Hea

dqu

arte

rsin

Ger

man

y;op

erat

ing

in94

cou

ntr

ies

14,0

0070

,000

Exe

cuti

vebo

ard

:C

EO

,C

FO

,ch

ief

exec

uti

veH

R,

chie

fex

ecu

tive

pas

sen

ger

busi

nes

s,ch

ief

exec

uti

veof

corp

orat

est

rate

gic

dev

elop

men

t.C

ircl

eof

50”:

Dir

ect-

rep

orti

ng,

sen

ior-

leve

lex

ecu

tive

s.

Sev

enst

rate

gic

busi

nes

sar

eas:

(1)

Pas

sen

ger

busi

nes

s(2

)L

ogis

tics

(3)

Mai

nte

nan

ce,

rep

air

and

over

hau

l(4

)C

ater

ing

(5)

Lei

sure

trav

el(6

)IT

serv

ices

(7)

Gro

un

dse

rvic

es

●D

ereg

ula

tion

,li

bera

liza

tion

,gl

obal

izat

ion

●R

apid

lych

angi

ng

mar

kets

and

cust

omer

dem

and

s

Man

agem

ent

Con

sult

ing

Inc.

Con

sult

ing

com

pan

ysp

ecia

lize

din

info

rmat

ion

tech

nol

ogy

and

rest

ruct

uri

ng

ofbu

sin

ess

pro

cess

es

Hea

dqu

arte

rsin

Ger

man

y;op

erat

ing

inE

uro

pe

and

the

US

175

1,30

0E

igh

tex

ecu

tive

boar

dm

embe

rs:

Hea

ds

ofbu

sin

ess

un

its,

CE

O,

CF

O

Six

stra

tegi

cbu

sin

ess

un

its:

(1)

Bu

sin

ess

stru

ctu

res

(2)

Cu

stom

erm

anag

emen

t(3

)T

ran

sact

ion

man

agem

ent

(4)

En

terp

rise

app

lica

tion

s(5

)B

usi

nes

sin

tell

igen

ce(6

)T

ech

nol

ogy

man

agem

ent

●G

loba

liza

tion

●E

-com

mer

ce:

chan

gin

gm

arke

tsan

dd

eman

ds

●H

igh

riva

lry

wit

hin

ind

ust

ry

aS

ourc

esar

eco

mp

any

rep

orts

,co

mp

any

web

site

s,an

din

tern

ald

ocu

men

ts.

For

con

fid

enti

alit

yre

ason

s,co

mp

any

nam

esar

ed

isgu

ised

.b

InU

S$

mil

lion

s.

specific circumstances of interview situations(Strauss, 1987). All interviews were tape-recordedand transcribed verbatim.

To further mitigate any biases associated withretrospective recall and thereby maximize the reli-ability and validity of our results, we also gatheredsecondary data—both publicly available informa-tion from several sources, such as annual reports,investors’ reports, internet sites, and online data-bases, and internal company documents regardingstrategy processes and corporate architecture, suchas organizational charts, process overviews, qualityhandbooks, financial goals, “balanced scorecards,”skill requirements lists, and internal teaching cas-es—which we used to triangulate our findings (Yin,2003). These kinds of archival records and generaldocuments are especially relevant in qualitativeresearch, since they integrate firm-specific contextsinto a research area (Strauss, 1987).

Drawing on 312 pages of interview transcriptsand numerous secondary sources, we assigned ini-tial concepts to the interview data, which we sub-sequently aggregated into more abstract categories,using the conceptual space defined by the threecorporate control dimensions as a guide. An itera-tive process led to our six corporate control types(Strauss & Corbin, 1998). Whereas the initial con-cepts represent “concepts-in-use” (Gephart, 2004)in the language of our informants, deriving thecorporate control types allowed us to lift the data toa conceptual level (Suddaby, 2006). For each iden-tified corporate control type, we also recorded as-sociated attributes that provided a more detaileddescription. As we gathered and analyzed moreinterview data, some types suggested by prior the-orizing in a subset of cases could be grounded inthe data and were retained or revised, while otherscould not and were discarded (Strauss & Corbin,1998). We continuously iterated between our dataand the evolving corporate control types until wehad a clear grasp of the emerging taxonomy andadditional interviews failed to reveal new data in-sights (Glaser & Strauss, 1967; Suddaby, 2006).Lastly, to gain an outsider’s perspective andthereby vet our ideas, we engaged other researchersnot involved in this study, such as fellow depart-ment members, doctoral students, and a researcherat another school, and discussed emerging patternsin our data as well as any evolving corporate con-trol types, and solicited critical questions about ourdata collection and analysis procedures (Corley &Gioia, 2004).

Results

Our interview data suggest that our respondentsunderstood decision speed to be the amount of timeutilized to evaluate and select—on the basis ofreasoning, negotiations, power positions, and otherforms of impetus and momentum—from alternativestrategic options, including the time it takes to re-ceive approval or disapproval from corporate-leveldecision makers. This understanding is consistentwith prior work. Also in line with prior work(Eisenhardt, 1989b; Judge & Miller, 1991; Lin &Carley, 1997), our informants emphasized the out-sized role decision speed plays in determining firmsuccess:

Successful . . . unsuccessful . . . it [the decisionprocess] has to be fast. That is the key criterion forsuccess. You have to be fast. It must not be a never-ending story. You have to come up with solutionsand results. If you have long discussions, you won’tbe successful. This, employees have to understandand be aware of. (Aviation Co.)

The primary outcomes of the first study, how-ever, are the following, inductively generated typesof corporate control. Some of these types could becaptured well by already existing labels (e.g., goalsetting), while others required new labels (e.g.,strategy imposition). However, all six corporatecontrol types were broadly recognized in all thefirms we analyzed. The types and their positionswithin our conceptual framework outlined aboveare illustrated in Table 2.

Goal setting. Goal setting emerged from our dataas referring to the establishment of financial, oper-ative, and strategic goals as well as budgets forSBUs, developed interactively between corporateand SBU levels. Goals are typically developed onthe basis of strategic plans, analyses of anticipatedinternal and external developments, negotiation,and other forms of individual or group interactionsto stimulate and motivate SBUs within a pre-defined period of time, and they serve as a bench-mark for subsequent performance evaluation ofSBU managers. The following quotes illustrate thiscorporate control type:

Targets, actions, resources, economic outcome—that’s the cycle. The key thing is to have targets.These targets have to be agreed upon and coordi-nated and negotiated with all employees who cancontribute. There is no point in defining a target byand for myself. You constantly have to inform andtalk about it with others. You have to say: “Whereare we today?” “What actions do we have to take?”


. . . And then the realization of targets has to bemeasured, and with this evaluation, you go backinto the target definition dialogue; and you commu-nicate which of the targets have been met and whichhave not. (Advertising Inc.)

We have ambitious targets. I mean, it’s like tellingmyself: “next year, you could reach this and thatfinancial target.” And then I have to think: “if I go onwith my operations the way I have so far, I probablywon’t make it. So, I need to do something differ-ently.” This is how you install and push forward acreative process that makes people think about whatthey want to accomplish. (Manufacturing Co.)

This corporate control type clearly resonateswith prior research on goal setting (Ethiraj &Levinthal, 2009; Galbraith, 1973), which empha-sizes the motivational effects of goals on task per-formance, provided that these goals are not in con-flict with each other, are sufficiently specific, andare within realistic reach (Locke & Latham, 1990).Moreover, in our sample companies, SBU managerswere actively involved in setting these goals, whichhas important implications, as complex tasks are

accomplished better when goals are decided in aninteractive process (Campbell & Gingrich, 1986).This control type represents an outcome control andranges from low content control, in the case of purelyfinancial goals, to high content control, in the case ofoperational and strategic goals (see Table 2).

Extrinsic incentives. Extrinsic incentives emergedas referring to remuneration as well as career-re-lated recognition and reward systems designed bycorporate headquarters for SBUs. This corporatecontrol type comprised a variety of financial bo-nuses, profit-participation schemes, fringe benefits,and career prospects, which—on the basis of mea-surable goals—are aimed at enhancing individualand firm performance in a self-regulatory manner.Compared to financial incentives, career incentivescan be materialistic, such as formal promotions, aswell as symbolic, such as informal recognition. Thefollowing quotes provide an illustration:

That is very simple: you will be rewarded. That isthe way it works around here. Everybody knows thatyou get rewarded if you perform well. You wouldn’t

TABLE 2Types of Corporate Control

CorporateControl Type Description

BehaviorControl

OutcomeControl

ContentControl

Goal setting Establishment of financial, operative, and strategic goalsas well as budgets for SBUs, developed interactivelybetween corporate and SBU levels.

Low High Low-higha

Extrinsic incentives Remuneration and career-related recognition andreward systems designed by corporate headquartersfor SBUs. Examples include financial bonuses, profit-participation schemes, fringe benefits, and careerprospects.

Low High Low

Negative incentives Any form of implicit or explicit penalty for undesiredoutcomes at the SBU level by corporate headquarters.Examples range from exerting verbal pressure orreassigning responsibilities to demoting, relocating, orfiring underperforming SBU executives.

Low High Low

Decision process control All procedural norms and guidelines established bycorporate headquarters for monitoring SBUs’ decisionprocesses. Examples include scheduled as well as adhoc meetings, deadlines, and written guidelines.

High Low Low

Conflict resolution Any formal and informal attempt by corporateheadquarters to mediate or decide on disputed issuesat the SBU level.

High Low Low-higha

Strategy imposition Top-down interventions in SBUs’ decision processes bycorporate headquarters imposing a substantive set ofpriorities and/or strategy on SBUs, and therebydeliberately limiting SBUs’ strategic flexibility anddecision autonomy.

Low Low High

a The extent of content control for these control types varies depending on the extent to which the corporate level specifies the substanceof strategy or allows SBUs to have input.


believe how much of an incentive that is for peoplehere. . . . It is about knowing whether you have careeropportunities within the company. If you know youare 25 [years old] now and you don’t have to wait foranother 3 years to get your next responsibility, but youcan be promoted within the next year, that’s an incen-tive. People work harder and end up performing betterin a shorter period of time. You really don’t need morethan that. (Manufacturing Co.)

Well, you have the executive managers who steerthe company—and everybody in the company hashigh respect for these top managers. And, of course,their main tool for steering the company is by meansof financial incentives. (Aviation Co.)

This corporate control type has also been identi-fied in prior research on strategy implementation(e.g., Daft & Macintosh, 1984; Gupta, 1987) andbroadly reflects agency theory arguments that em-phasize the alignment of individual and corporateobjectives through incentive mechanisms (Jensen &Meckling, 1976; Williamson, 1975). Extrinsic in-centives also resonate with the literature on trans-actional leadership and, in particular, contingentreward leadership, which focuses on rewardingand recognizing employees for the accomplishmentof agreed-upon objectives (Hater & Bass, 1988;Howell & Avolio, 1993). Similar to goal setting,extrinsic incentives represent outcome control, butthe amount of content control headquarters exertsis limited (see Table 2).

Negative incentives. Negative incentivesemerged as referring to any form of implicit orexplicit penalty for undesired outcomes at the SBUlevel imposed by corporate headquarters. Head-quarters generally applied these after negative per-formance outcomes but also used them to preventundesired behavior by merely threatening sanc-tions instead of actually imposing them. More sub-tle means are exerting verbal pressure or reassign-ing responsibilities to “send a message,” whilemore extreme cases include demoting, relocating,or firing underperforming SBU executives:

Generally, the holding company as a 100 percentshareholder gives out specific financial targets to thesubsidiary companies, which they have to achieve.. . . This includes hurdle rates, etc., all defined bythe corporate management. They say, “These are thefigures you have to achieve. If you don’t, you de-stroy corporate value and we will consider immedi-ate measures.” (Aviation Co.)

If you are successful and outcome-oriented, you willmove up [hierarchically], and you will automati-cally be given more responsibility and further tasks.

If you are not successful, you won’t be able to keepyour position for long. Basically, a business areamanager who is not successful won’t stay very longin his job. (Manufacturing Co.)

Negative incentives resonate with classic organiza-tional design theory based on authority/hierarchy (Si-mon, 1957). They have also been discussed in theliterature on transactional leadership (Hater & Bass,1988). In contrast to recognition, reward, and careeradvancement for achieving objectives, however,negative incentives focus on disciplining mistakes,often in the form of management-by-exception.Similar to the previous two types, negative incen-tives represent outcome control, with only limitedcontent control by headquarters (see Table 2).

Decision process control. Decision process con-trol emerged in our data as referring to all proce-dural norms and guidelines established by corpo-rate headquarters for monitoring SBUs’ decisionprocesses. This corporate control type representsbinding, mostly transparent, and well-documentedcorporate involvement in procedural aspects ofSBU decision processes, which provides SBU ex-ecutives with a high degree of content- and out-come-related flexibility and autonomy (see Ta-ble 2). It is characterized by repeated exchanges,which are mostly institutionalized and rely onscheduled meetings, deadlines, and written guide-lines, but can sometimes also occur ad hoc. Thefollowing quotes illustrate this control type:

We have a centralized mission/vision booklet,which we really stick to—strictly top-down. And wealso have guidelines in our [quality handbook], inwhich a lot of procedures are documented. . . . In thefirst two hours when [new people] enter the com-pany, they receive a package with the correspondingdocuments. In these first two hours they are taughtthe three fundamental building blocks that deter-mine our success: products, processes, and relation-ships, which correspond to expertise, quality, andsocial competence. These are the general guidelines,and we have our quality handbook, in which allprocesses are included in detail. (Advertising Inc.)

We have a lot of sophisticated procedures in placesuch as weekly conference calls or daily reporting.With these tools you really have a grip on what’sgoing on. (Manufacturing Co.)

Conflict resolution. Conflict resolution emergedas referring to any attempt by corporate headquar-ters to mediate or decide on disputed issues at theSBU level. Because of “coopetition” between SBUs,wherein they often compete for the allocation of


resources from corporate headquarters while, at thesame time, collaborating in the pursuit of commoninterests (Tsai, 2002), conflicts of interest betweenSBUs are natural and common. Corporate execu-tives act as mediators (and ultimate decision au-thorities) in situations of dispute between SBUs byovercoming resistance, resolving conflicts, andgenerating solutions. Sometimes, the corporatelevel has established official arenas for dealingwith conflicts; at other times, they are handled on amore informal, ad hoc basis. The following quotesprovide an illustration:

If there is any kind of conflict, for example, manag-ers saying: “but I want to have this client”—so, forexample, a conflict over a customer—then the man-agers at the holding level are the ones who ulti-mately make the decision. In our case, this would bethe CEO or the COO, or another strategic functionwithin corporate headquarters. (Advertising Inc.)

The head of the segment [i.e., corporate management]has overlaying responsibilities with regard to the busi-ness areas, the segment, but also the whole company.It’s his responsibility that the coordination betweenthe business areas is generally working. He has tomake sure that there is not too much friction and thatmanagers are not fighting with each other. He facili-tates and coordinates the whole thing and he holds theultimate decision power. (Manufacturing Co.)

This corporate control type has roots in behav-ioral theory, which starts from the premise thatorganizational conflicts are not exceptional, but aresult of diverging interests of organizational sub-units, particularly when these units are competingfor resources (Hill et al., 1992). Prior research hasmaintained, however, that while constructive cog-nitive conflict between SBUs can raise decisionquality by preventing “groupthink” (Jehn, 1995), itcan also trigger more destructive emotional conflictwith detrimental effects on decision outcomes(Amason, 1996). Conflict resolution therefore rep-resents a balancing act for headquarters in which itneeds to encourage cognitive conflict and construc-tive confrontation to enhance decision quality yetkeep conflict from becoming emotional. Conflictresolution also constitutes behavior control basedon headquarters’ surveillance of and influence inSBU-level decision processes, with different de-grees of corporate-level involvement in the actualcontent of decisions (see Table 2).

Strategy imposition. Strategy imposition emergedas referring to top-down interventions in SBUs’decision processes by corporate headquarters im-

posing a substantive set of priorities and/or strategyon SBUs, and thereby deliberately limiting SBUs’strategic flexibility and decision autonomy. Thisapproach is often based on overarching corporatepriorities and is applied when the corporate level isconvinced that certain decisions have to be made,regardless of SBUs’ perceptions. It is illustrated bythe following quotes:

If somebody doesn’t have the upper hand in a situ-ation—we [i.e., the corporate level] immediately getinvolved. [The corporate executive] analyzes the sit-uation, and if he feels that the local manager doesn’thave things under control, he immediately goes inand gets involved. (Manufacturing Co.)

Especially in the beginning, when we had to turnthe company around, we couldn’t sit down and talkwith each individual department to ask them, Couldyou do this and do that? It simply had to be done.We carried it out by decree—dictatorial. Period. The“lawn mower” method. And afterwards, of course,the questions came up, Was it fair? Was it right? Butin a situation like that it proved successful, eventhough unfair situations occurred here and there.But to steer such a huge tanker like us in a situationlike that while asking your department, Can you dothis or do that? is simply impossible. (Aviation Co.)

As with negative incentives, it is not only theactual intervention that may influence SBUs, butequally important, their perceived likelihood andsubsequent impact. Strategy imposition resonateswith several previously identified strategy imple-mentation modes, such as the commander model(Bourgeois & Brodwin, 1984), or command mode(Hart, 1992; Hart & Banbury, 1994), all of whichhave in common that corporate headquarters for-mulates SBU-level strategy in a largely unilateral,top-down way. Unlike the other types—and in con-trast to our expectations based on prior studies’arguments (e.g., Collis & Montgomery, 1998; Hill etal., 1992)—strategy imposition represents a corpo-rate control type that entails neither outcome norbehavior control but focuses exclusively on contentcontrol (see Table 2).

Our company is steered and coordinated throughfinancial targets and directives. And for some timenow, our salary consists of a fixed and a variableportion. And within the budgeting process thesefigures are broken down for each department andeach department head. And if you don’t achievethese targets, there will be a deduction in your sal-ary. (Aviation Co.)

Summary. Despite being conceptually distinct,the above corporate control types are not mutually


exclusive. In fact, they were frequently used inconjunction with one another, as illustrated by thefollowing example of corporate headquarters com-bining goal setting with extrinsic incentives:

Well, for the year 2000 there are, as a test-run, un-binding, financial, quantifiable target agreements,which will be binding as of January 1, 2001. Thesecan also be noneconomic targets, such as employeeturnover or the number of trained and promotedemployees. All of these are quantifiable target agree-ments, which, at the end of the year, are measuredand then form the underlying basis for calculatingthe variable incentive portion of the salary. (Man-agement Consulting Inc.)

After empirically deriving these corporate con-trol types, we had additional, ex post discussionswith several executives in our sample companies,who confirmed that the six types represent themain corporate-level types of influence on strategicdecision processes at the SBU level. Deployed ei-ther individually or in conjunction with one an-other, these control types therefore represent themost important means of corporate control overSBUs’ strategic decision making.

STUDY 2: ANALYZING RELATIONSHIPSBETWEEN CORPORATE CONTROL TYPES AND

DECISION SPEED

Building on and extending these results, in oursecond study, we examined how and why these sixcorporate control types influence the speed of SBU-level decision processes. Traditional question-naires or open-ended interviews proved unsuitablefor eliciting these relationships, however, for sev-eral reasons. To design a structured questionnairewould mandate having a relatively complete un-derstanding of the phenomena as a basis for askingthe right questions. In particular, even though wecould use a structured questionnaire to examinewhether or not the six corporate control types wereassociated with SBU-level decision speed, wewould not be able to gain a comprehensive under-standing of the underlying mechanisms explainingthese relationships. By choosing questions on cer-tain mechanisms, we would have had to make exante judgments about what are and what are notpotential mechanisms. In short, we would likelyhave imposed our own cognitive frames on thephenomena (Reger, 1990), frames that may ormay not be representative of the actual relation-ships. Open-ended and unstructured interviews,

on the other hand, though less susceptible to im-posing cognitive frames on informants, often fail toelicit valid and reliable perceptions from them (Re-ger, 1990). Particularly problematic for our inquirywas that decision makers tend to be unable to de-tect discrepancies between their “espoused theo-ries” (i.e., those they would report as the basis fortheir actions) and their “theories-in-use” (whichactually explain why they behave the way they do[Argyris & Schön, 1974]). Many executives thus“respond with answers about what they think theyshould know rather than what they actually think”(Easterby-Smith, Thorpe, & Holman, 1996: 4; em-phasis in original).

For these reasons, we relied on repertory gridanalysis (Kelly, 1955; Reger, 1990; Wright, 2006) toelicit executives’ implicit understanding of the re-lationships between corporate control types andSBU-level decision speed. Repertory grid analysisis a tool that uncovers and formally represents howindividuals construct their world. In particular, arepertory grid comprises elements (i.e., the mainconcepts to be investigated, which in our study arethe corporate control types and decision speed),attributes (i.e., adjectives used to describe and dif-ferentiate between elements), and linking mecha-nisms between elements and attributes (Easterby-Smith et al., 1996).

Repertory grid analysis has several advantages.6

First, it is a rigorous and systematic cognitive map-ping technique that requires only minimal inter-vention or interpretation by a researcher (Fransella,Bell, & Bannister, 2004; Reger, 1990; Wright, 2006).As a result, this technique provides reliable andvalid representations of actual understandings ofphenomena without researchers imposing theircognitive frames (Easterby-Smith et al., 1996;Wright, 2006). Second, repertory grid analysis al-lows researchers to delve deeper than other surveyand interview techniques to uncover managers’theories-in-use (Argyris & Schön, 1974). In partic-ular, by verbalizing attributes that would otherwise

6 Although repertory grid analysis originated in clini-cal psychology (Kelly, 1955), these advantages make it anappropriate method with which to generate insights on abroad variety of phenomena in strategic management,and it has been successfully applied to examine phenom-ena such as strategic issue diagnosis (Dutton, Walton, &Abrahamson, 1989), strategic groups (Reger & Huff,1993), competitive strategy (Marcel, Barr, & Duhaime,2010; Reger & Palmer, 1996), and strategy-making pro-cesses (Wright, 2004).


remain hidden, researchers are able to probe intoareas about which informants may not be aware(Easterby-Smith et al., 1996), thereby opening upthe black box of their cognitive maps. Repertorygrid analysis also provides finer-grained, more mi-crolevel insights into executives’ understanding ofcomplex processes than other techniques, and itthereby allows for better explanation of these pro-cesses (Wright, 2006). This is particularly impor-tant in the context of our study, as we were inter-ested not only in whether certain corporate controltypes are related to SBU-level decision speed, butalso in the mechanisms underlying these relation-ships. Third, although repertory grid analysis as aninterview technique has qualitative roots, it per-mits quantitative analyses, such as principal com-ponent and correlation analyses (Easterby-Smith etal., 1996; Fransella et al., 2004), which here en-abled us to empirically derive perceived associa-tions between corporate control types and SBU-level decision speed. This hybrid character of themethod mitigates the usual trade-off between accu-racy, or a close fit between theory and data, andgeneralizability, or the potential range of situationsto which a theory is applicable (Langley, 1999).

While repertory grid analysis is designed to elicitexecutives’ cognitive maps about a phenomenon,and not the actual relationships, these maps are ofconsequence, and numerous prior studies haveshown them to be the basis for strategic action (e.g.,Barr, Stimpert, & Huff, 1992; Nadkarni & Barr,2008; Reger & Huff, 1993). In fact, the method’stheoretical underpinning is predicated on the beliefthat executives act on their perceptions of the ob-jective world filtered through their cognitive sys-tem (Kelly, 1955; Reger & Huff, 1993). With man-agers’ actions being determined largely by howthey understand their situation (Porac & Thomas,2002), repertory grid analysis provides an excellentmeans of uncovering and presenting that under-standing (Easterby-Smith et al., 1996).

Interview and Analysis Procedures

We conducted 30 computer-assisted, structuredinterviews with executives at corporate as well asSBU levels (i.e., two interviews at the corporatelevel and four interviews at the SBU level of each ofour sample companies).7 While these informants

were different executives from those interviewed inthe first study, they were drawn from the same poolof key informants; they held senior positions at theSBUs and corporate headquarters and were thusqualified to provide in-depth and reliable assess-ments of their corporate control context and itseffects on SBU-level decision processes.

The face-to-face interview process was facilitatedby t.o.p.GRID,8 a three-step software program usinga variant of principal component analysis to re-trieve informants’ perceptions of similarities anddifferences between elements. In the first step, ourcomputer randomly selected two elements from thesix corporate control types generated in Study 1plus the two elements, high decision speed and lowdecision speed.9 Informants then compared thesetwo elements—for instance, “goal setting” and“conflict resolution”—and indicated whether theyperceived them as similar or different. In the sec-ond step, informants were asked to assign a de-scriptive attribute of their choosing to each ele-ment: if they perceived two elements as similar,they were asked to assign the same attribute; if theyperceived them as different, they were asked toassign one attribute to the first element and to as-sign another attribute of their choosing, which theyperceived as the opposite of the first attribute, tothe second element. An informant could, for in-stance, associate “goal setting” with attributes suchas “motivating” and “future-oriented,” and “con-flict resolution” with attributes such as “frustrat-ing” and “past-oriented.” In the third step, infor-mants evaluated the remaining elements (from the

7 As Dunn and Ginsberg (1986) and Wright (2006) haveremarked, a reasonably small number of respondents

(15–25) will frequently generate enough constructs toapproximate the “universe of meaning” surrounding agiven organizational context or situation.

8 For more information on the t.o.p.GRID program (andthe updated version of the program, nextexpertiser),please refer to http://www.nextpractice.de/en/services/nextexpertizer.

9 Because of t.o.p.GRID’s interview technique’s rootsin Kelly’s (1955) dichotomous psychology of personalconstructs, we divided the concept of decision speed intotwo opposing elements: high and low decision speed.With only one, neutral decision speed element,t.o.p.GRID would be able to provide only information onwhether a corporate control type (e.g., goal setting) wasperceived to be associated with decision speed or not,and no information on whether this association was pos-itive or negative. Having two opposing decision speedelements, however, made it possible to establish whethera specific corporate control type was perceived to speedup or slow down SBU-level decision processes.


http://www.nextpractice.de/en/services/nextexpertizer

http://www.nextpractice.de/en/services/nextexpertizer

pool of the six corporate control types plus highand low decision speed) in terms of the attributesthey had assigned to the first two elements. Forexample, having begun with “goal setting” and“conflict resolution,” informants assessed other el-ements such as “extrinsic incentives” in terms of aseries of alternatives: Whether they are “motivat-ing” and/or “future-oriented,” “frustrating” and/or“past-oriented,” in between each of these attri-butes, comprise both, or comprise neither; “no an-swer is possible” was also an alternative. As thisexample illustrates, attributes are always bipolarand are thereby comparable to questionnaire scales.Unlike with questionnaires, however, the sub-stance of the scales emerging from a repertory gridanalysis is generated by the informants themselvesand is therefore a more meaningful representation ofthe examined phenomenon (Wright, 2004).

Using the aggregated data from all informants’assessments, we first developed a joint representa-tion of all elements and their associated attributes,known as a singular-value decomposition (Slater,1977), which is a variant of a principal componentanalysis. On the basis of that analysis, we thenextracted those attributes (called typical attributes)that had a correlation of at least .70 with an ele-ment. The element “goal setting,” for example, wasassociated with 365 attributes, of which 56 weretypical attributes—for instance, “aligning,” “moti-vating,” and “transparent.” This procedure allowedus to distill a relatively small number of attributesthat represent the main characteristics of each cor-porate control type—which, in turn, illustrate itseffects on SBU-level decision processes—as well asthe main characteristics of the elements high deci-sion speed and low decision speed.

Second, we established the perceived associa-tions between corporate control types and decisionspeed by identifying the overlaps between the typ-ical attributes of the six corporate control typeswith the typical attributes of high and low decisionspeed. To continue our example from above, theattributes “aligning,” “motivating,” and “transpar-ent” were typical attributes for both goal settingand high decision speed, suggesting an associationbetween these two elements. To attain a fully sta-tistically significant association in our analysis,two elements had to fall within a 51 percent cross-section dimension (Raeithel, 1991). The outcome ofthe second study was therefore the retrieval of as-sociations between different corporate controltypes and decision speed, aggregated acrossinformants.

The typical attributes emerging in our repertorygrid analysis were then passed back into the com-parative case analysis process (Wright, 2006) so wecould examine why certain corporate control typesaffect decision speed. Goal setting’s typical attri-bute “motivating,” for instance, was substantiatedby the following interview quote:

That is why the corporate level introduced financialperformance figures as a means of control. . . . Thisforced us to move faster—just by the outcome ori-entation being in place. (Aviation Co.)

By examining their respective overlapping typi-cal attributes, substantiating those with evidencefrom our interviews, and further drawing on theinterview data to develop an understanding of re-lationships, we were able to detect and explain themechanisms linking corporate control types anddecision speed and thereby develop an intermedi-ate theoretical model (Edmondson & McManus,2007). Lastly, we used prior research to refine thetheoretical logic of the emerging relationships and todevelop propositions (Martin & Eisenhardt, 2010).

Figure 1 illustrates the relationships we observedbetween corporate control types, intervening mech-anisms, and the speed of SBU-level decision mak-ing. Table 3 provides representative quotes fromour data that further substantiate these links. Al-though the data provide many more, to save space,we offer only a single quote for each relationship.In the following section, we develop arguments andoffer propositions for the theoretical model identi-fied in Figure 1 and Table 3.

Emergent Theoretical Framework LinkingCorporate Control Types and SBU-Level DecisionSpeed

Decision speed. Our informants associated highdecision speed with 26 typical attributes, concen-trated around five main attribute areas: (1) trans-parency in the decision process and a subsequentalignment between corporate and SBU interests,(2) an outcome-oriented and therefore largely pro-active approach to decision making, (3) participa-tion of SBU managers in SBU-level decision pro-cesses, (4) trust between SBUs and corporateheadquarters, and (5) timely feedback. In contrast,low decision speed was associated with 32 typicalattributes, concentrated around five main attributeareas: (1) lack of transparency in the decision pro-cess and a subsequent misalignment between cor-porate and SBU-level interests, (2) a problem-ori-


ented and therefore largely reactive or passiveapproach to decision making, (3) a lack of partici-pation of SBU-level management in SBU-level de-cision processes, (4) distrust between SBUs andcorporate headquarters, and (5) stagnation andlethargy, emphasizing the paralyzing effects of pro-crastination that are typical for low decision speed.

Goal setting. Goal setting was associated with56 typical attributes. To examine the associationbetween goal setting and decision speed, we ana-lyzed the typical attributes that overlapped goalsetting and decision speed. In the case of goal set-ting, the overlap with high decision speed con-sisted of seven typical attributes, which can begrouped into four main attribute areas (see Table 3for details on these attribute areas, as well as rep-resentative quotes that illustrate and substantiateeach association). We further found no typical at-tributes overlapping low decision speed.10 Theseresults support a positive association between goal

setting and SBU-level decision speed and makegoal setting the corporate control type that is thesecond most closely associated with high decisionspeed.

The overlapping typical attributes—in conjunctionwith our qualitative interview data from Study 1—further allowed us to draw conclusions with re-gards to the question of why goal setting affectsSBU-level decision speed. First, transparencyhelps reduce the ambiguity surrounding corporateintentions and goals for SBUs (cf. Locke & Latham,1990), which enables faster SBU-level decision pro-cesses because SBU managers avoid time-intensive“sense-making” processes associated with inter-preting vague corporate guidelines (Baum & Wally,2003). As a by-product of this transparency, goalsetting also allows for better alignment betweencorporate and SBU interests, which makes it easierfor the two levels to reach an agreement on SBUstrategies, and which, in turn, speeds up SBU-leveldecision processes (Eisenhardt, 1989b). Second,goal setting induces decision makers to activelyfocus on outcomes, versus having a more open-ended, problem-solving orientation, which perpet-uates the decision process (Goold & Quinn, 1990).Third, we found that in our sample firms, some

10 While a few of the interviewed executives also men-tioned attributes of goal setting that overlapped withattributes of low decision speed, these attributes did notattain the status of typical attributes; thus, we did notfind any significant evidence for an association betweengoal setting and low decision speed.

FIGURE 1Corporate Control Types and SBU-Level Decision Speed

Goal setting

Extrinsic incentives

Negative incentives

Decision process control

Conflict resolution

Strategy imposition

SBU-leveldecision speed

Transparency/ alignment

Outcome orientation

Participation

Timely feedback

Trust

—

—

—

+

+

+

+

+

+

+

+

+

+

+

++

+

+

+

+

Corporate Control Types Mediators/Mechanisms Outcome


TABLE 3Corporate Control Types and Their Effects on SBU-Level Decision Speed

CorporateControl Type

OverlappingAreas

RepresentativeQuotes

1. Goal setting 1.1. Transparency/alignment

“And then we receive the targets, which are broken down to our level. And the businessdevelopment people, they also get their targets, which force them to come up with theircontribution and to push it forward. As I said, if [the corporate managers] want to broadlypush something through, across all business units, it is relatively difficult. Mostly you haveto try and help the pioneering business unit and then transfer this back to the overalltargets.” (IT Consulting Inc.)

1.2. Outcomeorientation

“In this case you have to look at our history. In ’91 and ’92, the situation was more seriousthan we had expected, and we almost went out of business. Back then, we didn’t have thetime to think about our long-term strategy—our focus was on survival. This experience, ofcourse, influenced our corporate managers, who wanted to make sure that something likethat would never happen to us again. That is why the corporate level introduced financialperformance figures as a means of control. . . . This forced us to move faster—just by theoutcome orientation being in place.” (Aviation Co.)

1.3. Participation “As an employee I have access; I can have an influence on the [Manufacturing Co.] strategy.That is, any employee with enough interest and power can do that. I do not feel like I’m ina small box and I can’t do anything here. . . . It really comes down to the commitment ofyour employees. If they feel they can change and influence something, they will make suretheir processes and their performance will be superior.” (Manufacturing Co.)

1.4. Timelyfeedback

“We have weekly conference calls within the business areas and between all business unitheads. Once a week they discuss their markets and their treasury center and why thingswork well or why they don’t. There is a lot of general information exchanged between allsorts of people. These meetings are pretty strong input factors, defining very short timeperiods and ensuring that things don’t get out of hand. . . . In our department, we are reallyused to things happening fast and also to having to react fast.” (Manufacturing Co.)

2. Extrinsicincentives

2.1. Transparency/alignment

“We get a set of parameters. These are financial numbers we are expected to achieve, andthese are assumptions which apply to all of us in the same way. Then we go back to ourwork . . . in medias res . . . and return with a strategy which has to be aligned with thesefinancial figures.” (Aviation Co.)


“I think that speed requires a certain degree of impatience. And I think the driving forcebehind this impatience is that it is actually good for us. . . . I mean, we actually getsomething out of it, because 50 percent of the profits go to us. So, on the financial side,there are a lot of incentives. In addition, there is also the incentive to prove to someone thatI manage my job and perform well.” (Advertising Inc.)

2.3. Participation “The mathematical model allows me to freely define the variable portion of my salary. . . . Forexample, I can determine that the variable portion of my salary should be 25 to 40 percent.This, of course, bears the risk that I might underperform, but also the opportunity that Ioverperform.” (Management Consulting Inc.)

2.4. Trust “The greatest encouragement for them is to get promoted, to move ahead, to get moredemanding tasks—generally, to have more responsibility. For them, it’s not about financialincentives, but about knowing that I trust them to cope with more difficult tasks. . . . Bytrusting them, I show them that I am content with their performance and that is what spursthem most.” (Manufacturing Co.)

2.5. Timelyfeedback

“In this company, you can get ahead and make a career much faster than in any othercompany I know—at least if I compare it to my friends in other companies. That’s a wholelot of an incentive system. With this, you don’t really need an additional financial stimulus.If you are good—you’ll be promoted. That’s it.” (Manufacturing Co.)

3. Negativeincentives

n.a. “Well, normally employees do not appreciate reprimands or sanctions, even though they canalso be advantageous, for example, by enforcing and keeping fairness. But I think generally,sanctions are perceived as punishments, which publicly display and embarrass thewrongdoers including their mistakes and failures in front of the other employees. Of course,they don’t feel necessarily motivated by that. Sometimes, this might even lead to generalfrustrations and increased barriers, which in turn leads to further impediments.”(Manufacturing Co.)

Continued


form of participative discussion and negotiationtook place prior to the actual definition of goals. Inother words, goal setting in our sample entailed theactive participation of SBU managers early on.While prior research has identified such involve-ment and its positive effects on strategy implemen-tation (Bourgeois & Brodwin, 1984; Goold & Camp-bell, 1987a, 1987b; Goold et al., 1994; Gupta, 1987),it has been assumed that it comes at the expense ofa fast decision process (Mintzberg, 1978). After all,involving more people increases the time consumedin goal-setting exercises. It is therefore somewhatcounterintuitive to find that more participation in-creases decision speed. However, greater involve-

ment of SBU managers in corporate-level goal-set-ting processes enhances their understanding of andcollective appreciation for the chosen goals (Ama-son, 1996). This, in turn, improves coordinationand cooperation in downstream SBU-level decisionprocesses, such as the generation of alternatives(Bourgeois & Brodwin, 1984; Wooldridge & Floyd,1990), which allows for a faster decision process.Fourth, goal setting emphasizes timely feedback todetect deviations and adjust goals if necessary.Such use of real-time information in decision pro-cesses has been shown to have beneficial effects ondecision speed (Eisenhardt, 1989b). Given theseresults, we propose:

TABLE 3(Continued)

CorporateControl Type

OverlappingAreas

RepresentativeQuotes

4. Decisionprocesscontrol

4.1. Transparency/alignment

“Clearly standardization. Standardized and efficiency-driven tools you can adapt worldwide.Let me give you an example. Our brand character tool is a one-page piece of paper. Whoeverworks on a global brand of one of our customers, be it in South Africa or in Finland,everyone has the brand character tool in front of them on their desks—and that tool looksthe same all over the world. If the employee is not willing to accept that, he will not beallowed to work on that brand. This consistency to structure and apply these commonstandards worldwide really symbolizes our strength and made our company so successful.”(Advertising Inc.)


“It is always important to make them, and sometimes even force them, to think into thefuture. . . . It is the same in every business. For example, if you have a new product—itmight have a huge market potential, but to start the whole thing can be difficult. You haveto force yourself . . . and a lot of times, this effort is contradictory to the short-termperformance. That’s why you need a driving force behind it, which makes you look ahead,even though it might not make sense in the present situation.” (IT Consulting Inc.)

4.3. Trust “It takes diplomatic skills to interact and deal with these people in a trusting manner. Forexample, if the corporate level wants something from the passenger side [i.e., the businessunit]. If they haven’t built up trust—they won’t get a foot in the door. In this case, youreally have to build up trust. That’s our job.” (Aviation Co.)

5. Conflictresolution

n.a. “Generally, if a conflict starts damaging the ongoing business, then corporate managersintervene, which is done comparatively fast. Most of the times they just decree a specificresolution to the conflict—one that best suits their needs and the needs of the wholecompany. This way people don’t get a chance to keep babbling that much and less time iswasted. In addition, the corporate level mostly has a good overview of what is strategicallyimportant for the whole company.” (Manufacturing Co.)

6. Strategyimposition

6.1. Lack oftransparency/misalignment

“Which [SBU head] was in charge of this [SBU], how much discretion they had, and howstrong their influence was, was unclear. It was clear to the board, but not to the employees.”(Management Consulting Inc.)

6.2. Lack ofparticipation

“Sometimes, they are told exactly what to do. That means they are told not only what isgenerally expected from them, but they receive specific orders on how to run their business.This can even affect the color of their presentation or the unit price of a specific product tobe sold. A lot of people don’t appreciate this type of corporate control because they feelshut out and held back. You have to be really careful not to discourage your employees,otherwise resistance builds up and processes become inefficient.” (Manufacturing Co.)

6.3. Distrust “We are patronized—on a leash. . . . Sometimes we are forced to agree to campaigns we don’teven consider high-quality or even suitable under the given circumstances . . . . You canonly accomplish something and gain credibility, acceptance, and trust when you are notconstantly under pressure to deliver [reports] that you’ve already sent three times.”(Advertising Inc.)


Proposition 1. The corporate control type “goalsetting” is positively related to SBU decisionspeed. The mechanisms mediating this rela-tionship include (a) increased transparencyand alignment between corporate and SBU-level interests; (b) a stronger outcome orienta-tion; (c) increased participation by SBU-levelmanagers; and (d) more timely feedback inSBU-level decision processes.

Extrinsic incentives. For this type, the area over-lapping high decision speed consisted of 15 typicalattributes, grouped into five main attribute areas(see Table 3 for details and representative quotes).These results support a positive association be-tween extrinsic incentives and SBU-level decisionspeed. Extrinsic incentives is the corporate controltype that was most closely associated with highdecision speed.

Overall, the mediation effects of extrinsic incen-tives show substantial overlap with those of goalsetting. First, extrinsic incentives result in trans-parency and alignment in the interactions betweencorporate headquarters and its SBUs. In that capac-ity, they represent a self-regulatory system, wheremanagers and employees of SBUs respond to incen-tives by adjusting their behavior to maximize pos-sible benefits (Jensen & Meckling, 1976; William-son, 1975). Thus, this corporate control typeinduces ongoing self-corrections, which createsfaster alignment of SBU-level decision processes.Second, extrinsic incentives increase SBU-level de-cision makers’ outcome orientation, driven by theirinterest in being rewarded. Third, as an outcomecontrol mechanism (Ouchi & Maguire, 1975), ex-trinsic incentives allow for managerial autonomy,flexibility, and active participation by SBU manag-ers in decision processes, which encourages proac-tive decision making. Extrinsic incentives therebymotivate SBU managers (Black & Gregersen, 1997)by offering rewards for corporate-preferred out-comes, while allowing SBU managers broad lati-tude in their strategic behaviors, thereby increasingdecision speed. Fourth, extrinsic incentives pro-vide timely feedback, quickening the pace of SBU-level decision processes. As outlined in our discus-sion of the previous corporate control type, and asillustrated by the quotes in Table 3, these effects, inturn, have a beneficial impact on decision speed.

While these effects of extrinsic incentives resem-ble the effects of goal setting, there is one majordifference between the two corporate control types,and this difference is related to trust. In particular,

when corporate headquarters establish clearly ar-ticulated and consistently applied extrinsic incen-tives for SBUs the ambiguity SBUs experience withrespect to the fulfillment of corporate expectationsis reduced, which increases trust (Mayer, Davis, &Schoorman, 1995). Extrinsic incentives also signalcorporate headquarters’ sympathy for SBU manag-ers’ needs by rewarding these managers for theirperformance. This further increases trust in corpo-rate headquarters, eliminates second-guessing, andfacilitates faster decisions (cf. Mayer et al., 1995).We therefore propose:

Proposition 2. The corporate control type “ex-trinsic incentives” is positively related to SBUdecision speed. The mechanisms mediatingthis relationship include (a) increased trans-parency and alignment between corporate andSBU-level interests; (b) a stronger outcome ori-entation; (c) increased participation by SBU-level managers; (d) more timely feedback inSBU-level decision processes; and (e) in-creased trust between SBU- and corporate-level managers.

Negative incentives. The lack of overlappingtypical attributes between negative incentives andboth high and low decision speed indicates a lackof association between these elements. In spite ofthis nonfinding, the 12 typical attributes associatedwith this corporate control type and assessments ofour informants expressed in interviews fromStudy 1 shed some light on the effects of negativeincentives on SBU-level decision processes (see Ta-ble 3 for a representative quote). On the one hand,negative incentives are associated with distant con-trol, implying that corporate executives definetasks to be achieved by SBU managers and takeaction only if outcomes fail to meet expectations.The effect on decision speed is likely marginal,since this form of corporate-level intervention isrelatively rare. On the other hand, when this inter-vention does happen, its impact is significant. Ourrespondents perceived such negative incentives asoverpowering, stifling creativity, and removing au-tonomy from SBU-level management. This reso-nates with findings from leadership research sug-gesting that such a management-by-exceptionapproach has negative effects on SBU performance(Howell & Avolio, 1993). Again, however, theseintrusions and their effects appear to be relativelyrare, and consequently, our repertory grid resultsdo not indicate a significant association between


negative incentives and SBU decision speed. Ac-cordingly, we propose:

Proposition 3. The corporate control type “neg-ative incentives” is not significantly related toSBU decision speed.

Decision process control. Decision process con-trol represented the corporate control type the thirdmost closely associated with high decision speed.The overlapping area consisted of five typical attri-butes, concentrated around three attribute areas(see Table 3 for details and representative quotes).Overall, decision process control provides an inter-esting balance between trust and control. First, andsimilarly to goal setting and extrinsic incentives,decision process control enhances the transpar-ency and alignment of SBU-level decision pro-cesses through the creation of procedural norms,guidelines, and ongoing monitoring. Such clarity,then, allows for faster decisions at the SBU level(Baum & Wally, 2003). Second, consistent monitor-ing not only helps clarify and align decision pro-cesses, it also ensures that steady progress is beingmade, which, in turn, orients SBU managers towardoutcomes, with positive effects on decision speed.Third, although repeated exchanges and scheduledmeetings as part of process control keep the corpo-rate level informed about and thereby in control ofSBU decision processes, it does not directly inter-fere with SBUs’ activities, leaving decision-makingauthority at the level at which the most relevantinformation and expertise reside (Campbell, 1999;Eisenmann, 2005). This, in turn, enhances decisionspeed at the SBU level. As a by-product of suchcontent-related autonomy for SBU managers, deci-sion process control also fosters trust among SBUmanagers, who are confident enough to make fastdecisions instead of second-guessing corporate-level reactions (cf. Mayer et al., 1995). Given thesearguments, we propose:

Proposition 4. The corporate control type “de-cision process control” is positively related toSBU decision speed. The mechanisms mediat-ing this relationship include (a) increasedtransparency and alignment between corpo-rate and SBU-level interests; (b) a stronger out-come orientation; and (c) increased trust be-tween SBU- and corporate-level managers.

Conflict resolution. Our informants did not as-sociate conflict resolution with decision speed, asindicated by the lack of overlapping typical attri-butes. Analyzing the effects of conflict resolution

by examining its typical attributes provides furtherinsight into its ambiguous effects on decisionspeed. In particular, the ten typical attributes fo-cused on two potentially off-setting attribute areas(see Table 3 for an illustrative quote from Study 1).On the one hand, conflict resolution attempts maymean that corporate headquarters imposes central-ized control over the definition of solutions, aimedat reducing frictions and interruptions in the deci-sion process. This would be expected to enhancedecision speed (Baum & Wally, 2003; Wally &Baum, 1994). On the other hand, conflict resolutionattempts by corporate headquarters may involvethe creation of arenas for negotiating divergent SBUinterests, thereby enhancing reciprocal communi-cation and mutual adjustment among the parties.While this approach may lead to a jointly devel-oped solution that likely enjoys broad support,such negotiations can also be time consuming(Eisenhardt, 1989b). Based on these ambiguous ef-fects, we propose:

Proposition 5. The corporate control type “con-flict resolution” is not significantly related toSBU decision speed.

Strategy imposition. Although the overlap be-tween strategy imposition and low decision speeddid not attain full statistical significance,11 our re-sults suggest an at least marginally significant, neg-ative influence of strategy imposition on decisionspeed. This negative relationship is corroboratedby our interview data from Study 1 (see Table 3). Inparticular, the 42 overlapping typical attributesthat characterize both strategy imposition and lowdecision speed are concentrated around three mainattribute areas. First, strategy imposition is per-ceived as opaque and sometimes random interfer-ence by the corporate level, a nontransparent intru-sion on SBU managers. This corporate control typeincreases uncertainty about headquarters’ potentialrole in SBU decision processes and likely leads to amisalignment between SBU and corporate-level in-terests and which, in turn, slows decision making(Eisenhardt, 1989b). Second, this type is seen as an

11 A statistically significant overlap requires a 51 per-cent cross-section dimension or less (Raeithel, 1991); theoverlap between strategy imposition and low decisionspeed was represented in a 53 percent cross-section di-mension, thus suggesting an at least moderate associationbetween the two elements. This result is further corrob-orated by the fact that strategy imposition did not haveany overlap with high decision speed.


autocratic and patronizing form of corporate man-agerial control over the SBU level. Not only does itundermine SBU managers’ participation in deci-sions related to SBU strategy, it is seen as interfer-ing with the substance of SBU strategy by centrallycontrolling or even overriding the decisions of SBUmanagers, effectively disempowering them. Thisresults in an atmosphere of frustration at the SBUlevel (Black & Gregersen, 1997), which is particu-larly acute in cases when SBU managers perceivecorporate headquarters as not sufficiently informedabout the specifics of their businesses (Campbell,1999; Eisenmann, 2005). The constraints on auton-omy resulting from strategy imposition also lead toa reduced sense of responsibility at the SBU level,sometimes even putting pending decisions on hold,since SBU managers subject to strategy impositionoften come to expect that corporate executives willintervene and take control of the decisions anyway.Third, strategy imposition increases distrust be-tween SBU managers and corporate headquarters(Lorsch & Allen, 1973), which arises from SBUmanagers’ uncertainty about headquarters’ inten-tions (cf. Mayer et al., 1995). Such distrust likelyleads to second-guessing of corporate intentions,further slowing down the decision process. Draw-ing on this evidence, we propose:

Proposition 6. The corporate control type“strategy imposition” is negatively related todecision speed at the SBU level. The mecha-nisms mediating this relationship include (a)decreased transparency and alignment be-tween corporate and SBU-level interests; (b)decreased participation by SBU-level manag-ers; and (c) decreased trust between SBU- andcorporate-level managers.

DISCUSSION

Using a two-stage data collection and analysisprocess, we developed a midrange theoreticalmodel explaining the relationship between corpo-rate control and SBU-level decision speed. Integrat-ing insights from our qualitative and quantitativeanalyses, the results suggest that certain types ofcorporate controls positively influence SBU deci-sion speed, while other types negatively influenceit, and still others have little or no effect. The prin-cipal finding of this study, however, is that theserelationships are a function of five mediating mech-anisms. When controls positively influence trans-parency/alignment, outcome orientation, participa-

tion, trust, and timely feedback in headquarters’relations with SBUs, they exert a positive influenceon decision speed. Understanding these mediatorshelps explain why managers adopt different typesof controls and provides a new appreciation forhow such controls affect SBU-level decision mak-ing. These findings have implications for the liter-atures on strategic decision making, corporate strat-egy and structure, and organizational control,which we discuss in turn.

First, our study complements work on decisionspeed in young, small, and undiversified firms(Baum & Wally, 2003; Eisenhardt, 1989b; Forbes,2005; Perlow et al., 2002; Wally & Baum, 1994) byexamining the determinants of decision speed inestablished, large, and diversified companies. Thisrepresents an important extension because priorwork has found larger companies to be at an inher-ent disadvantage when it comes to fast decisionprocesses (Baum & Wally, 2003; Wally & Baum,1994). Addressing this challenge, our findings sug-gest a number of ways in which large, multibusi-ness corporations can mitigate their size-related li-abilities and accelerate the speed of their SBUs’decision processes. Whereas some of our findingsconfirm previously identified determinants of de-cision speed, others qualify the effects of prior de-terminants. The known positive effects for the useof real-time information and for integration amongdecisions (Eisenhardt, 1989b), for instance, are re-flected in the effects of the timely feedback andalignment that certain corporate control typesentail (see Figure 1). The null effect of conflictresolution in our sample of multibusiness firms,however, suggests a potential contingency on thepositive effect found in other settings (Eisen-hardt, 1989b).

More importantly, our findings also suggest anumber of previously unidentified, corporate-leveldeterminants of decision speed. In particular, ourapproach allows us to explain some of the conflict-ing results in prior work with respect to the effectsof centralization and formalization on decisionspeed (Baum & Wally, 2003; Mintzberg, 1979, 1981;Siggelkow & Rivkin, 2005; Wally & Baum, 1994) bydisentangling what exactly remains centralized atheadquarters: control over decision content, con-trol over decision outcomes, and/or control overthe decision process. Goal setting, for instance, al-lows headquarters to exert influence over SBUs’strategic priorities, thereby capitalizing on some ofthe benefits of centralization on decision speed(Baum & Wally, 2003; Wally & Baum, 1994), while


autonomous control over how to achieve those pri-orities remains at the SBU level, allowing the busi-nesses to respond more quickly to any changes intheir local environments (Mintzberg, 1979). In sum,our taxonomy of corporate control types, in con-junction with our analysis of the microlevel mech-anisms operating between control types and SBUdecision speed, allows us to take a more nuancedview than prior work and to disentangle the multi-faceted impact of corporate control on SBU deci-sion speed.

Second, our study also contributes to the litera-ture on corporate strategy and structure. The ma-jority of studies on multibusiness firms have ar-gued that, given the unique advantages of differenttypes of diversification strategies, corporate man-agement should make a choice between outcomeand behavior control (e.g., Collis & Montgomery,1998; Hill et al., 1992; Hill & Hoskisson, 1987; Jones& Hill, 1988; Pitts, 1980; Vancil, 1978). In line withthis argument, our results illustrate that with theexception of strategy imposition, all corporate con-trol types fall into one or the other dimension, withvarying degrees of content control exercised by cor-porate headquarters (see Table 3). However, ourresults also suggest the possibility of companiesusing different control types simultaneously, suchas the combination of goal setting and extrinsicincentives discussed in the summary of Study 1,and even combining corporate control types acrosscontrol dimensions, such as using goal setting/ex-trinsic rewards in conjunction with decision pro-cess control. Such hybrid control approaches mayallow multibusiness firms to alleviate at least someof the trade-offs between types. The literature haslong maintained, for instance, that the highest per-formers among multi-business firms tend to be theones with related businesses (Rumelt, 1982), whichrequire some degree of corporate-level control overSBUs’ strategy formation to manage integration andcoordination among SBUs (Hill et al., 1992; Hill &Hoskisson, 1987). On the downside, however, priorresearch has also found the complexity of integra-tive mechanisms to be negatively related to deci-sion speed (Lorsch & Allen, 1973) and headquar-ters’ interference may lead to poor decision quality(Campbell, 1999; Eisenmann, 2005). In contrast toforcing a choice between controls, our sample com-panies found ways to combine the advantages ofoutcome controls (such as goal setting and extrinsicincentives) for motivating SBU-level managementwith the advantages of decision process control forallowing headquarters some influence over how

strategic decisions are formed. One such solution isillustrated in this quote:

Of course, everybody in this partnership is influ-enced by the total profits made throughout the year.This is extremely important with regard to self-reg-ulating the internal processes; for example if a stra-tegic initiative doesn’t succeed or if its performanceis bad. . . . Standardized and quantifiable planningin this respect is an important issue. There is arule-based mechanism, which coordinates everymanager in the company by means of financial par-ticipation. Therefore, of course, everybody wantsprofits to increase and everybody feels influencedby that. (IT Consulting Inc.)

Third, our findings also contribute to the controlliterature (e.g., Eisenhardt, 1985; Ouchi, 1979; Ou-chi & Maguire, 1975). In introducing his theory,Ouchi observed: “The problem of organization isthe problem of obtaining cooperation among a col-lection of individuals or units who share only par-tially congruent objectives” (1979: 833). The diver-sity of goals and strategies in multibusiness firmsmakes this challenge particularly acute. In this con-text, when strategic decision making is the focus,we have seen that organizations largely rely oncontrols that are outcome- and process-oriented,and that transparency/alignment, outcome orienta-tion, participation, trust, and timeliness are neededto gain cooperation while preserving SBU auton-omy. What does not appear to enhance decisionspeed are strategy imposition, negative incentives,and corporate attempts to resolve conflicts betweenSBUs. None of these produce any of the positivemechanisms attributed to other corporate controltypes. Moreover, the latter two seem rather reactiveways for corporate headquarters to exert controlover SBUs in that they either reprimand or punishunwanted outcomes or attempt to resolve conflictsonce they occur—whereas the other corporate con-trol types more proactively establish control overSBU goals, incentives, and decision processes—which may provide another explanation for whynegative incentives and conflict resolution werenot associated with enhanced decision speed. Incontrast, strategy imposition is seen as undermin-ing trust and limiting SBU autonomy:

[The corporate manager] simply provides orders wehave to follow . . . we don’t have another choice butto carry it out. (Advertising Inc.)

Thus, uncovering the mechanisms responsiblefor the relationships between control types andSBU decision speed constitutes another important


contribution; they shift the focus away from whatthe controls are and instead focus on how theywork. For the present context, the negative findingwith respect to strategy imposition is especiallysalient in this regard. Related diversifiers face theneed for some degree of control over SBU strategy(Hill et al., 1992; Hill & Hoskisson, 1987), and themost direct approach to gaining such control iswhat we call strategy imposition. Our data show,however, that doing so—even with good inten-tions—not only reduces SBU autonomy but alsoundermines essential mechanisms in the form oftransparency/alignment, participation, and trust,thereby slowing down the decision process, as il-lustrated by the following quote:

Sometimes, we at the holding level think that wecan make our organization faster by simply telling[SBU managers] what they must or must not do.What we often forget while doing that is that weupset our managers by taking away their chances toinfluence the company. This results in frustrationand sometimes even political resistance, whichblocks the implementation of the initial idea alto-gether, or at least makes it incredibly tedious.(Manufacturing Co.)

How can a related diversifier’s need for synergiesbe achieved without negatively impacting decisionspeed and sacrificing SBU trust and autonomy?Prior work by Martin and Eisenhardt (2010) hasfound that an SBU-centric process led by SBU man-agers leads to better collaboration and more syner-gies among SBUs than a process driven by corpo-rate headquarters. Complementing and extendingthis finding, theory developed here shows that theuse of goal setting, extrinsic incentives, and deci-sion process control trigger transparency and align-ment between corporate and SBU intentions withrespect to strategy, and thereby provide the neededstrategic consistency, but without compromisingthe SBU’s sense of autonomy. The following quoteprovides an illustration:

Providing autonomy to SBUs is the key mechanismapplied by corporate management. But you can’thave too much autonomy. You also need a processin place that controls the business unit and absorbsfrictions. . . . Corporate managers have a doublerole—they have to ensure that things are balanced.Of course, some people don’t like the corporate dou-ble role of autonomy and control and complainabout it. But I think, gradually, people are acceptingit. . . . We really work very, very closely with sys-tems and structures to better achieve this balance.And we will think more about formal processes to

secure the balance we currently have despite thehigh autonomy we are providing. (Advertising Inc.)

Moreover, the implementation of these three cor-porate control types involves a high degree of par-ticipation (at least for goal setting and extrinsicincentives) and trust—both of which increase com-munication and information sharing across levelsand thereby facilitate a common mindset with re-spect to strategy (Amason, 1996; Wooldridge &Floyd, 1990). Thus, the identification of the mech-anisms through which corporate controls influenceSBU decision making offers a way to avoid the“trap” of strategy imposition without sacrificinginfluence over SBU strategy. More generally, iden-tifying these mechanisms provides the basis forunderstanding such trade-offs in the design of cor-porate controls.

Prior research on organizational control also out-lines an interesting dilemma: whereas outcomecontrol best serves the needs of an organization as awhole, behavior control best serves the needs of theindividual SBUs. As a result, “in the absence of anomniscient executive or an all-encompassing set ofoutput measures, organizations must have twocomplementary means of control, one to serve or-ganizational needs and one to serve subunit needs”(Ouchi & Maguire, 1975: 569). In our sample, how-ever, both goal setting and extrinsic incentives(both outcome controls) enhanced decision speedby increasing participation and trust. By involvingSBU managers and sharing information in the def-inition of goals and incentives, the corporate levelseems to be able to develop outcome measures thatare sensitive to the specialized needs of each SBUand that could serve both SBUs’ and organizationalneeds (Ouchi & Maguire, 1975):

Each SBU head is, at the same time, a member of theexecutive board. So he will not get any objectivesdictated from above, but he himself has been in-volved in creating these objectives. . . . As an SBUhead, he will also be involved in the discussion ofnext year’s target definition and he can say: “I canachieve the following objectives; and my unit hasassured me that we can actually do that.” (Manage-ment Consulting Inc.)

Again, an understanding of the mechanismshelps to surface a more detailed appreciation forwhy corporate controls influence decision making,and in this case, also helps avoid a potentiallyunnecessary layer of control systems.

Our study is also subject to limitations. First, thefindings are derived from perceptual assessments


of the corporate control context and its impact onSBU-level decision processes. While we attemptedto mitigate any method effects by relying on multi-ple informants—both at different companies and atdifferent levels within each company—and whilewe conducted our mixed-method study in two sep-arate stages, an analysis based on more objectivemeasures for decision speed, for example, is impor-tant to corroborate our results. Second, all samplecompanies are located in dynamic industries. Sincethe moderating influence of industry dynamismhas been well documented (e.g., Eisenhardt, 1989b;Judge & Miller, 1991), caution should be exercisedwhen generalizing our results to more stable indus-try environments. Third, our analyses focused onsimilarities rather than differences among compa-nies and hierarchical levels in order to replicateand extend the emerging theory (Eisenhardt, 1989a;Yin, 2003). We thus largely neglected the potentialfor perceptual differences between corporate levelsand SBUs, which may provide a fruitful avenue forfuture research (see, e.g., Ireland, Hitt, Bettis, andde Porras [1987] for a study of perceptual differ-ences on strategic issues between organizationallevels). It may be interesting, for instance, to con-trast corporate-level intentions when employingcertain corporate control types with SBU-level per-ceptions about their effects on transparency/align-ment, participation, trust, and other factors thatcharacterize SBU-level decision processes. Ouranalysis also minimized differences between thesampled firms and SBUs, such as size, geographicdistribution, etc., which may enhance or diminishthe effects of the corporate control types on SBUdecision processes.

A fruitful extension of this research would there-fore be to control for contingencies such as envi-ronmental context, industry, or firm-specific influ-encing factors. Moderating effects of externalcontingencies could be examined with variablessuch as dynamism, complexity, and munificence(Dess & Beard, 1984), competitive environment(Miller & Friesen, 1983), or industry membership(Amurgey & Miner, 1992). Moderating effects ofinternal contingencies could include variablessuch as firm size (Fredrickson & Mitchell, 1984;Miller, 1991), degree of diversification (Hill et al.,1992), decision importance (Judge & Miller, 1991),and top management team characteristics such asnationality, average tenure, or international workexperience (Wiersema & Bantel, 1992).

In conclusion, this study enhances understand-ing of the impact of the organizational control con-

text on SBU-level decision speed. As corporateheadquarters have a significant impact on theirSBUs, and as a substantial percentage of firms op-erate as multibusiness organizations, we need toknow more about positive and negative influencesof the corporate level on SBU-level decision pro-cesses. This study represents a first step in thatdirection.

REFERENCES

Aldrich, H. E. 1979. Organizations and environments.Englewood Cliffs, NJ: Prentice Hall.

Amason, A. C. 1996. Distinguishing the effects of func-tional and dysfunctional conflict on strategic deci-sion making: Resolving a paradox for top manage-ment teams. Academy of Management Journal, 39:123–148.

Amurgey, T., & Miner, A. 1992. Strategic momentum:The effects of repetitive, positional, and contextualmomentum on merger activity. Strategic Manage-ment Journal, 13: 338–355.

Argyris, C., & Schön, D. A. 1974. Theory in practice:Increasing professional effectiveness. San Francis-co: Jossey-Bass.

Bansal, P., & Corley, K. 2011. The coming of age forqualitative research: Embracing the diversity of qual-itative methods. Academy of Management Journal,54: 233–237.

Barr, P. S., Stimpert, J. L., & Huff, A. S. 1992. Cognitivechange, strategic action, and organizational renewal.Strategic Management Journal, 13: 15–36.

Baum, J. R., & Wally, S. 2003. Strategic decision speedand firm performance. Strategic Management Jour-nal, 24: 1107–1129.

Baysinger, B., & Hoskisson, R. E. 1990. The compositionof board of directors and strategic control: Effects oncorporate strategy. Academy of Management Re-view, 15: 72–87.

Black, J. S., & Gregersen, H. B. 1997. Participative deci-sion-making: An integration of multiple dimensions.Human Relations, 50: 859–878.

Bourgeois, L. J., & Brodwin, D. 1984. Strategic implemen-tation: Five approaches to an elusive phenomenon.Strategic Management Journal, 5: 241–264.

Bourgeois, L. J., & Eisenhardt, K. M. 1988. Strategic de-cision making processes in high velocity environ-ments: Four cases in the microcomputer industry.Management Science, 34: 816–835.

Bower, J. L., & Hout, T. M. 1988. Fast-cycle capability forcompetitive power. Harvard Business Review,66(6): 110–118.


Burgelman, R. A. 1983. A process model of internal cor-porate venturing in the diversified major firm. Ad-ministrative Science Quarterly, 28: 223–244.

Campbell, A. 1999. Tailored not benchmarked: A freshlook at corporate planning. Harvard Business Re-view, 77(2): 41–50.

Campbell, D. J., & Gingrich, K. F. 1986. The interactiveeffects of task complexity and participation on taskperformance: A field experiment. OrganizationalBehavior and Human Decision Processes, 38: 162–180.

Chandler, A. D. 1962. Strategy and structure: Chaptersin the history of the American industrial enter-prise. Cambridge, MA: MIT Press.

Child, J. 1984. Organization: A guide to problems andpractice (2nd ed.). New York: Harper & Row.

Chung, L. H., Gibbons, P. T., & Schoch, H. P. 2000. Theinfluence of subsidiary context and head office stra-tegic management style on control of MNCs: Theexperience in Australia. Accounting, Auditing andAccountability Journal, 13: 647–666.

Collis, D. J., & Montgomery, C. A. 1998. Creating corpo-rate advantage. Harvard Business Review, 76(3):70–83.

Collis, D. J., Young, D., & Goold, M. 2007. The size,structure, and performance of corporate headquar-ters. Strategic Management Journal, 28: 383–405.

Corley, K. G., & Gioia, D. A. 2004. Identity ambiguity andchange in the wake of a corporate spin-off. Admin-istrative Science Quarterly, 49: 173–208.

D’Aveni, R. A., Dagnino, G. B., & Smith, K. G. 2010. Theage of temporary advantage. Strategic ManagementJournal, 31: 1371–1385.

Daft, R. L., & Macintosh, N. B. 1984. The nature and useof formal control systems for management controland strategy implementation. Journal of Manage-ment, 10: 43–66.

Dess, G. G., & Beard, D. W. 1984. Dimensions of organi-zational and task environments. Administrative Sci-ence Quarterly, 29: 52–73.

Dunn, W. N., & Ginsberg, A. 1986. A sociocognitive net-work approach to organizational analysis. HumanRelations, 40: 955–975.

Dutton, J. E., Walton, E. J., & Abrahamson, E. 1989. Im-portant dimensions of strategic issues: Separatingthe wheat from the chaff. Journal of ManagementStudies, 26: 379–396.

Easterby-Smith, M., Thorpe, R., & Holman, D. 1996. Us-ing repertory grids in management. Journal of Euro-pean Industrial Training, 20(3): 3–30.

Edmondson, A. C., & McManus, S. E. 2007. Methodolog-

ical fit in management field research. Academy ofManagement Review, 32: 1155–1179.

Eisenhardt, K. M. 1985. Control: Organizational and eco-nomic approaches. Management Science, 31: 134–149.

Eisenhardt, K. M. 1989a. Building theories from casestudy research. Academy of Management Review,14: 532–550.

Eisenhardt, K. M. 1989b. Making fast strategic decisionsin high-velocity environments. Academy of Man-agement Journal, 32: 543–576.

Eisenhardt, K. M., & Graebner, M. E. 2007. Theory build-ing from cases: Opportunities and challenges. Acad-emy of Management Journal, 50: 25–32.

Eisenmann, T. R. 2005. Corporate intervention in re-source allocation. In J. L. Bower & C. G. Gilbert(Eds.), From resource allocation to strategy: 299–306. Oxford, UK: Oxford University Press.

Ethiraj, S. K., & Levinthal, D. A. 2009. Hoping for A to Zwhile rewarding only A: Complex organizations andmultiple goals. Organization Science, 20: 4–24.

Floyd, S. W., & Lane, P. J. 2000. Strategizing throughoutthe organization: Managing role conflict in strategicrenewal. Academy of Management Review, 25:154–177.

Forbes, D. P. 2005. Managerial determinants of decisionspeed on new ventures. Strategic ManagementJournal, 26: 355–366.

Fransella, F., Bell, R., & Bannister, D. 2004. A manual forrepertory grid technique (2nd ed.). Chichester, UK:Wiley.

Fredrickson, J. W. 1986. An exploratory approach to mea-suring perceptions of strategic decision process con-structs. Strategic Management Journal, 7: 473–483.

Fredrickson, J. W., & Iaquinto, A. L. 1989. Inertia andcreeping rationality in strategic decision processes.Academy of Management Journal, 32: 516–542.

Fredrickson, J. W., & Mitchell, T. R. 1984. Strategic de-cision processes: Comprehensiveness and perfor-mance in an industry with an unstable environment.Academy of Management Journal, 27: 399–423.

Galbraith, J. R. 1973. Designing complex organizations.Reading, MA: Addison-Wesley.

Galunic, D. C., & Eisenhardt, K. M. 2001. Architecturalinnovation and modular corporate forms. Academyof Management Journal, 44: 1229–1249.

Gephart, R. P. J. 2004. Qualitative research and the Acad-emy of Management. Academy of ManagementJournal, 47: 454–462.

Glaser, B., & Strauss, A. 1967. The discovery ofgrounded research: Strategies for qualitative re-search. New York/Chicago: Aldine.


Golden, B. R. 1992. SBU strategy and performance: Themoderating effects of the corporate-SBU relation-ship. Strategic Management Journal, 13: 145–158.

Goold, M., & Campbell, A. 1987a. Many best ways tomake a strategy. Harvard Business Review, 65(6):70–76.

Goold, M., & Campbell, A. 1987b. Strategies and styles:The role of the centre in managing diversified cor-porations. Oxford, UK: Basil Blackwell.

Goold, M., Campbell, A., & Alexander, M. 1994. Corpo-rate-level strategy: Creating value in the multi-business company. New York: Wiley.

Goold, M., & Quinn, J. J. 1990. The paradox of strategiccontrols. Strategic Management Journal, 11: 43–57.

Graebner, M. 2009. Caveat venditor: Trust asymmetriesin acquisitions of entrepreneurial firms. Academy ofManagement Journal, 52: 435–472.

Gupta, A. K. 1987. SBU strategies, corporate-SBU rela-tions, and SBU effectiveness in strategy implemen-tation. Academy of Management Journal, 30: 477–500.

Gupta, A. K., & Govindarajan, V. 1986. Resource sharingamong SBUs: Strategic antecedents and administra-tive implications. Academy of Management Jour-nal, 29: 695–714.

Hambrick, D. C., & Mason, P. A. 1984. Upper echelons:The organization as a reflection of its top managers.Academy of Management Review, 9: 193–206.

Hart, S. L. 1992. An integrative framework for strategy-making processes. Academy of Management Re-view, 17: 327–351.

Hart, S. L., & Banbury, C. 1994. How strategy-makingprocesses can make a difference. Strategic Manage-ment Journal, 15: 251–269.

Hater, J. J., & Bass, B. M. 1988. Superiors’ evaluations andsubordinates’ perceptions of transformational andtransactional leadership. Journal of Applied Psy-chology, 73: 695–702.

Hickson, D. J., Wilson, D. C., Cray, D., Mallory, G. R., &Butler, R. J. 1986. Top decisions: Strategic decision-making in organizations. San Francisco: Jossey-Bass.

Hill, C. W. L. 1988. Internal capital market controls andfinancial performance in multidivisional firms. Jour-nal of Industrial Economics, 37: 67–83.

Hill, C. W. L., Hitt, M. A., & Hoskisson, R. E. 1992.Cooperative versus competitive structures in relatedand unrelated diversified firms. Organization Sci-ence, 3: 501–521.

Hill, C. W. L., & Hoskisson, R. E. 1987. Strategy andstructure in the multiproduct firm. Academy ofManagement Review, 12: 331–341.

Hoskisson, R. E., & Hitt, M. A. 1988. Strategic controlsystems and relative R&D investment in large multi-product firms. Strategic Management Journal, 9:605–621.

Howell, J. M., & Avolio, B. J. 1993. Transformationalleadership, transactional leadership, locus of con-trol, and support for innovation: Key predictors ofconsolidated-business-unit performance. Journal ofApplied Psychology, 78: 891–902.

Ireland, R. D., Hitt, M. A., Bettis, R. A., & de Porras, D. A.1987. Strategy formulation processes: Differences inperceptions of strength and weaknesses indicatorsand environmental uncertainty by managerial level.Strategic Management Journal, 8: 469–485.

Jehn, K. A. 1995. A multimethod examination of thebenefits and detriments of intragroup conflict. Ad-ministrative Science Quarterly, 40: 256–282.

Jensen, M. C., & Meckling, W. H. 1976. Theory of thefirm: Managerial behavior, agency costs and owner-ship structure. Journal of Financial Economics, 3:305–360.

Jones, G. R., & Hill, C. W. L. 1988. Transaction costanalysis of strategy-structure choice. Strategic Man-agement Journal, 9: 159–172.

Judge, W. Q., & Miller, A. 1991. Antecedents and out-comes of decision speed in different environmentalcontexts. Academy of Management Journal, 34:449–463.

Kahneman, D., Slovic, P., & Tversky, A. 1982. Judgmentunder uncertainty: Heuristics and biases. Cam-bridge, UK: Cambridge University Press.

Kelly, G. A. 1955. The psychology of personal con-structs. New York: Norton.

Langley, A. 1999. Strategies for theorizing from processdata. Academy of Management Review, 24: 691–710.

Lin, Z., & Carley, K. M. 1997. Organizational response:The cost performance tradeoff. Management Sci-ence, 43: 217–234.

Locke, E. A., & Latham, G. P. 1990. A theory of goalsetting and task performance. Englewood Cliffs, NJ:Prentice Hall.

Lorsch, J. W., & Allen, S. A. 1973. Managing diversityand interdependence: An organizational study ofmultidivisional firms. Boston: Harvard BusinessSchool, Division of Research.

Makadok, R. 1998. Can first-mover and early-mover ad-vantages be sustained in an industry with low barri-ers to entry/imitation? Strategic Management Jour-nal, 19: 683–696.

Marcel, J., Barr, P. S., & Duhaime, I. M. 2010. The influ-


ence of executive cognition on competitive dynam-ics. Strategic Management Journal, 32: 115–138.

March, J., & Olsen, J. 1976. Ambiguity and choice inorganizations. Bergen, Norway: Universitetsfor-laget.

Markides, C. C., & Williamson, P. J. 1996. Corporatediversification and organizational structure: A re-source-based view. Academy of Management Jour-nal, 39: 340–367.

Martin, J. A., & Eisenhardt, K. M. 2010. Rewiring: Cross-business-unit collaborations in multibusiness organ-izations. Academy of Management Journal, 53:265–301.

Mayer, R. C., Davis, J. H., & Schoorman, F. D. 1995. Anintegration model of organizational trust. Academyof Management Review, 20: 709–734.

Miles, R., & Snow, C. 1978. Organizational strategy,structure, and process. New York: McGraw-Hill.

Miller, D. 1986. Configurations of strategy and structure:Towards a synthesis. Strategic Management Jour-nal, 7: 233–249.

Miller, D. 1991. Stale in the saddle: CEO tenure and thematch between organization and the environment.Management Science, 37: 34–52.

Miller, D., & Friesen, P. H. 1983. Strategy-making andenvironment: The third link. Strategic Manage-ment Journal, 4: 221–235.

Mintzberg, H. 1978. Patterns in strategy formulation.Management Science, 24: 934–948.

Mintzberg, H. 1979. The structuring of organizations.Englewood Cliffs, NJ: Prentice Hall.

Mintzberg, H. 1981. Organization design: Fashion or fit?Harvard Business Review, 59(1): 103–116.

Mintzberg, H., Raisinghani, D., & Théorêt, A. 1976. Thestructure of “unstructured” decision processes. Ad-ministrative Science Quarterly, 21: 246–275.

Muralidharan, R. 1997. Strategic control for fast movingmarkets: Updating the strategy and monitoring per-formance. Long Range Planning, 30: 64–73.

Nadkarni, S., & Barr, P. S. 2008. Environmental context,managerial cognition, and strategic action: An inte-grated view. Strategic Management Journal, 29:1395–1427.

Nadler, D. A., & Tushman, M. L. 1999. The organizationof the future: Strategic imperatives and core compe-tencies for the 21st century. Organizational Dynam-ics, 28(1): 45–60.

Ouchi, W. G. 1979. A conceptual framework for the de-sign of organizational control mechanisms. Manage-ment Science, 25: 833–848.

Ouchi, W. G., & Maguire, M. A. 1975. Organizational

control: Two functions. Administrative ScienceQuarterly, 20: 559–569.

Papadakis, V. M., Lioukas, S., & Chambers, D. 1998.Strategic decision-making processes: The role ofmanagement and context. Strategic ManagementJournal, 19: 115–147.

Perlow, L. A., Okhuysen, G. A., & Repenning, N. P. 2002.The speed trap: Exploring the relationship betweendecision making and temporal context. Academy ofManagement Journal, 45: 931–955.

Pfeffer, J. 1980. Power in organizations. Marshfield, MA:Pitman.

Pitts, R. A. 1980. Toward a contingency theory of multi-business organization design. Academy of Manage-ment Review, 5: 203–210.

Porac, J. F., & Thomas, H. 2002. Managing cognition andstrategy: Issues, trends and future directions. In A. Pet-tigrew, H. Thomas & R. Whittington (Eds.), Handbookof strategic management: 165–181. London: Sage.

Porter, M. E. 1980. Competitive strategy: Techniquesfor analyzing industries and competitors. NewYork: Free Press.

Pratt, M. G. 2009. For the lack of a boilerplate: Tips onwriting up (and reviewing) qualitative research.Academy of Management Journal, 52: 856–862.

Raeithel, A. 1991. Arbeiten zur Methodologie der Psy-chologie und zur Kelly-Matrizen-Methodik [Workson psychological methods and on the Kelly-matrix-method]. Habilitation, University of Hamburg.

Reger, R. K. 1990. The repertory grid technique for elic-iting the content and structure of cognitive construc-tive systems. In A. S. Huff (Ed.), Mapping strategicthought: 301–309. Chichester, UK: Wiley.

Reger, R. K., & Huff, A. S. 1993. Strategic groups: Acognitive perspective. Strategic Management Jour-nal, 14: 103–124.

Reger, R. K., & Palmer, T. B. 1996. Managerial categori-zation of competitors: Using old maps to navigatenew environments. Organization Science, 7: 22–39.

Rumelt, R. P. 1982. Diversification strategy and profit-ability. Strategic Management Journal, 3: 359–369.

Siggelkow, N., & Rivkin, J. W. 2005. Speed and search:Designing organizations for turbulence and com-plexity. Organization Science, 16: 101–122.

Simon, H. A. 1957. Administrative behavior (2nd ed.).New York: Macmillan.

Slater, P. 1977. The measurement of intrapersonalspace by grid technique, vol. 2. Dimensions of in-trapersonal space. London: Wiley.

Souitaris, V., & Maestro, B. M. M. 2010. Polychronicity intop management teams: The impact on strategic de-cision processes and performance of new technology


ventures. Strategic Management Journal, 31: 652–678.

Strauss, A. L. 1987. Qualitative analysis for social sci-entists. Cambridge, UK: Cambridge University Press.

Strauss, A. L., & Corbin, J. 1998. Basics of qualitativeresearch—Techniques and procedures for devel-oping grounded theory (2nd ed.). Thousand Oaks,CA: Sage.

Suddaby, R. 2006. What grounded theory is not. Acad-emy of Management Journal, 49: 633–642.

Sutcliffe, K. M., & McNamara, G. 2001. Controlling deci-sion-making practice in organizations. OrganizationScience, 12: 484–501.

Tsai, W. 2002. Social structure of “coopetition” within amultiunit organization: Coordination, competition,and intraorganizational knowledge sharing. Organi-zation Science, 13: 179–190.

Tushman, M. L., & Romanelli, E. 1983. Uncertainty, so-cial location and influence in decision making: Asociometric analysis. Management Science, 29: 12–23.

Vancil, R. F. 1978. Decentralization: Managerial ambi-guity by design. Homewood, IL: Dow Jones-Irwin.

Wally, S., & Baum, J. R. 1994. Personal and structuraldeterminants of the pace of strategic decision mak-ing. Academy of Management Journal, 37: 932–956.

Wiersema, M., & Bantel, K. A. 1992. Top managementteam demography and corporate strategic change.Academy of Management Journal, 35: 91–121.

Williamson, O. E. 1975. Markets and hierarchies: Anal-ysis and antitrust implications. New York: FreePress.

Wooldridge, B., & Floyd, S. W. 1990. The strategy pro-cess, middle management involvement, and organi-zational performance. Strategic Management Jour-nal, 11: 231–241.

Wright, R. P. 2004. Top managers’ strategic cognitions onthe strategy making process: Differences betweenhigh and low performing firms. Journal of GeneralManagement, 30: 61–78.

Wright, R. P. 2006. Rigor and relevance using repertorygrid technique in strategy research. In D. J. J. Ketchen& D. D. Bergh (Eds.), Research methodology in strat-egy and management, vol. 3: 295–348. Oxford, UK:Elsevier & JAI.

Yauch, C. A., & Steudel, H. J. 2003. Complementary useof qualitative and quantitative cultural assessment

methods. Organizational Research Methods, 5:465–481.

Yin, R. K. 2003. Case study research: Design and meth-ods (3rd ed.). Thousand Oaks, CA: Sage.

APPENDIX

Outline of Semistructured Interview Questions

Introduction

Introduction of interviewer backgroundShort presentation of research project and research teamQuestions about informant background: career, position,

functionQuestions about research site: structure, SBUs, stake-

holders, markets

Strategy Process

What are the steps and sequence of the strategy process atthe firm?

What milestones, deadlines, meetings, task forcesare there?

How are decisions made within the strategy process?Where do strategic initiatives occur in the organization

(location)?Context (rigid vs. open); responsibility (centralized vs.

decentralized); direction of influence (top-down vs.bottom-up)

Who participates in the strategy process and who devel-ops strategic initiatives (participants)?

Participation (low vs. high); perspectives (homogeneous vs.heterogeneous); capabilities (mono- vs. interdisciplinary)

At what point in the process are strategic initiativesdeveloped and how long does it take (timing)?

Duration (short vs. long); activation (scheduled vs. event-oriented); time-frame (short-term vs. long-term)

How do people involved in the strategy process behave(behavior)?

Conflict behavior (averting vs. exposed); decision making(patriarchal vs. democratic); transparency (low vs. high)

Corporate Control Context

What is the corporate- and SBU-level management inthis firm?

How do corporate managers in this firm seek to influencethe SBU?

What mechanisms of corporate influence are there withregard to finance, strategy, HR, structure, control, andstandards? Others?


Which of the mechanisms are primarily applied by cor-porate managers? Why?

How do interactions between corporate and SBU manag-ers take place?

What are potentials for improvement? Why?

Impact of the Corporate Control Context/DecisionSpeed/Strategy Outcome

What do corporate/SBU managers perceive as a success-ful corporate control context?

How is success defined at varying levels of the organiza-tion?

What is the impact of the corporate control context onquality, time, decision making, competition, groupprocesses?

What do corporate/SBU managers perceive as deci-sion speed?

What do managers perceive as influence factors on deci-sion speed?

What role does speed play in strategic decision making?(Relevance?)

How are different characteristics of decision speed per-ceived/measured?

Link between the Corporate Control Context andDecision Speed

How does the corporate control context influence deci-

sion speed?What can corporate managers do to increase the SBU-

level decision speed?What are the five means of the corporate control context

that influence decision speed most? Why?What are the five means of the corporate control context

that influence decision speed least? Why?

Maximilian Kownatzki ([email protected]) is an air-line executive running group strategy for Jetstar Airways.Prior to this, he was a senior partner in the aviationpractice of Oliver Wyman, an internationally renownedmanagement consulting firm. He received his doctoraldegree from the University of St. Gallen in Switzerland(HSG), with a two-year research fellowship at the Uni-versity of California, Irvine (UCI), where most of theunderlying research was conducted for this article. Hisresearch interests focus on strategic decision making,strategy process, and corporate planning.

Jorge Walter ([email protected]) is an assistant professorat the School of Business, George Washington University.He received his doctorate in strategic management fromthe University of St. Gallen in Switzerland, and was avisiting researcher at the Wharton School, University ofPennsylvania, and a postdoctoral researcher at the SternSchool of Business, New York University. His researchinterests include strategic decision making, knowledge/technology transfer, social networks, and social capital.

Steven W. Floyd ([email protected]) is the Eu-gene M. Isenberg Professor of Innovation and Entrepre-neurship at the Isenberg School of Management, Univer-sity of Massachusetts Amherst. He received his Ph.D.from the Leeds School of Business at the University ofColorado. His research interests focus on strategic deci-sion making, strategy development processes, and corpo-rate entrepreneurship.

Christoph Lechner ([email protected]) is achaired professor of strategic management at the Univer-sity of St. Gallen, as well as director of its Institute ofManagement. His research focuses on strategic initia-tives, strategic renewal, and activity systems.


CORPORATE CONTROL AND THE SPEED OF STRATEGIC BUSINESS UNIT … · 2017-07-26 · CORPORATE CONTROL AND THE SPEED OF STRATEGIC BUSINESS UNIT DECISION MAKING MAXIMILIAN KOWNATZKI Jetstar

Documents