اندازه گیری عملکرد و فرایندهای سازمانی: مطالعه پاسخ های مدیریتی به اصلاحات بخش دولتی

Institutional aspects of performance measurement (PM) in public sector organizations are attracting increasing research interest. Only recently, however, has the literature on this topic recognized the pertinent critique of neo-institutional sociology (NIS) pivoting around its view of managers and organizations as primarily passive adaptors to change. This paper explores how the properties of institutional processes associated with recent reforms in the Norwegian health care sector impinge on the extent of pro-active choice exercised by senior management in the development of multidimensional PM reflecting the interests of a wider range of institutional constituencies. Addressing this issue, we draw on Oliver’s (1991) conceptual framework, based on a continuum of responses characterized by a varying degree of pro-active choice. The study thus provides a more detailed analysis of the managerial tactics in developing organizational PM than most prior research informed by NIS. We find support for several of Oliver’s hypotheses regarding the influence of institutional aspects, particularly those pertaining to the causes of the adoption of PM practices, the pattern in which these are diffused and the influence of constituency multiplicity and dependence, but also identify some areas requiring conceptual refinement in this respect.

Over the past few years, management accounting researchers have paid increasing attention to the implications of public sector reforms for the design and implementa-tion of systems for performance measurement (PM) (e.g. Lapsley and Mitchell, 1996;Ballantine et al., 1998; Modell, 1998; Johnsen, 1999; Kloot and Martin, 2000). Severalauthors have proposed a multidimensional approach to PM, reflecting the interestsof a broader range of stakeholder interests, as a means of widening the conception ofpublic sector performance from financial, efficiency-based measures (Mayston, 1985;Pollitt, 1986; Brignall, 1993; Ballantine et al., 1998; Kloot and Martin, 2000). Muchof this research has been informed by a functionalistic perspective where improvedmeasurement and information systems are regarded as pivotal for assisting managersin making better informed trade-offs and thus balancing more or less conflictingstakeholder interests in the overall control of organizations. More recently, however,the considerable problems in achieving such a balance have stimulated someinterest in neo-institutional sociology (NIS) as an alternative basis for exploring thepremises impeding a wider conception of performance in public sector organizations (Llewellyn, 1996; Brignall and Modell, 2000; Lawton et al., 2000).The present paper extends these efforts to study PM from an NIS perspective byexamining the responses of senior management and staff specialists to recent reformsin the public health care sector in Norway. While ample attention has been paid to theresponses of operating-level employees and managers occupying dual professionaland administrative roles to recent public sector reforms (e.g. Purdy, 1993; Pettersen,1995; Jones and Dewing, 1997; Llewellyn, 1997; Jacobs, 1998; Llewellyn, 1998; Modell,2000), more in-depth investigations of the less extensively explored issues of howand why senior management influences PM in public sector organizations may provide important complementary insights. Some prior studies indicate that thedevelopment of novel PM practices and other formalized control mechanisms inresponse to public sector reforms tend to be dominated by senior management andstaff specialists (Laughlin et al., 1994; Lawton et al., 2000). In their capacity as systemsdesigners, senior managers may thus function as translators of change through theirinteractions with the organization’s institutional environment (cf. Czarniawska andJoerges, 1996), which may involve some arbitration between conflicting constituentinterests (Brignall and Modell, 2000).Contrary to the assumptions invoked by much early theorizing in NIS, however,managers are not necessarily confined to passively comply with institutionalpressures, but may possess much wider action repertoires involving a greaterelement of pro-active choice (see e.g. DiMaggio, 1988; Powell, 1991; Scott, 1995;Beckert, 1999). While this insight is spreading among management accountingscholars informed by NIS (e.g. Abernethy and Chua, 1996; Covaleski et al., 1996; Euske and Riccaboni, 1999), no empirical study of multidimensional PM to date has systematically explored the responses of senior management along these lines. Giventhe capacity of PM to render visible the implications of public sector reforms forvarious organizational constituencies (cf. Brunsson, 1990; Broadbent, 1995), seniormanagement may need to exercise considerable caution in compiling performanceinformation for the purpose of legitimization. Hence, pro-active and careful selectionof indicators reflecting a broader range of financial and non-financial performanceaspects of concern to key constituencies may be warranted to balance conflictinginterests, although the institutional context of reforms is likely to pose someconstraints in this respect (Brignall and Modell, 2000).Following the discussion in the foregoing, the overriding research question to beexplored in this paper can be stated as follows: how do the properties of institutional

Our case findings yield several important insights pertaining to our research questionand Oliver’s (1991) hypotheses. Regarding the adoption of DRG-based PM by thehospital, we find evidence of both legitimacy-seeking and efficiency-enhancingrationales forming part of senior management’s rhetoric. Despite the largely negativeexperiences from the initial experimenting with DRG-based funding in the early1990s and the complexities associated with DRG-based PM, it was seen as a superior458 S. Modellmechanism to the crude financial indicators previously relied upon for improvingfinancial control of operations as well as legitimization to external constituencies(primarily politicians) by informants at the divisional as well as higher levels.5This supports Oliver’s (1991) hypotheses regarding the influence of both perceived legitimacy and efficiency gains as determinants of compliance with institutionalizedpractices. However, a few qualifying remarks, shedding further light on Oliver’sframework, are required in this respect.First, legitimacy-seeking and efficiency-enhancing rationales may be more closely intertwined than acknowledged by Oliver, as reflected by, for example, the General Manager’s strong belief that DRGs would improve internal financial control, and his use of this argument as an internal justification of change as well as in his publicadvocacy role (cf. Sjøstrand, 1997). This is an example of more pro-active mimickingthan is typically associated with acquiescent behaviour and suggests that Olivermight have over-emphasized the passivity of acquiescence as a response where morecomplex, interwoven rationalities for the adoption of structural attributes emerge(see Abernethy and Chua, 1996, for similar arguments). However, our findings areat variance with Abernethy and Chua’s (1996) assertion that relatively simple andcrude financial controls might suffice for legitimization to external constituencies,such as politicians. Instead, we observed extensive efforts to make a highly complexcontrol system operational for internal use alongside conscious attempts to ‘sell’ the new control practices to important external audiences (see Lowe, 2000, forsimilar observations). There is also some consonance between the General Manager’sprior initiative to implement a divisional structure and the subsequently enhancedemphasis on financial PM in the organization and the decision to rely more heavilyon DRGs to further strengthen financial control. This reinforces the impression ofthe new PM practices as part of some intentional change initiative resulting inprogressively greater technical ‘sophistication’ in financial control practices.Second, the pro-active mimicking of the new system of funding is not only conditioned by the causes of the adoption of DRG-based PM. A complementaryexplanation can be derived by considering the pattern in which these practices arediffused. While the fact that all county councils have mimicked the new fundingsystem vis-à-vis hospitals is an indication of growing institutionalization (DiMaggioand Powell, 1983), central government and the county council in question havetaken a cautious approach to the further diffusion of DRGs for internal controlwithin hospitals. Furthermore, we found pioneering managerial efforts to refineDRG-based controls for internal use occurring in a financially stable and seeminglyefficient hospital (as indicated by its consistently good performance in DRG-basedleague tables), possibly as a means of capitalizing on the legitimacy thus conferredon the organization and as a preparation for more far-reaching changes in thesystem of governance. Taken together, these findings gainsay the prediction thatthe health care providers most likely to turn DRG-based controls inwards wouldbe financially strained hospitals exposed to coercive pressures to conform withsome rationalized efficiency myth (Covaleski et al., 1993). Consistent with Oliver’s(1991) hypotheses, however, our findings indicate that where voluntary diffusionpredominates over coercive pressures, senior management may have considerable discretion to preemptively go beyond institutional demands in anticipation of futurebenefits rather than slavishly mimic institutional practices imposed by politicians.Finally, the pro-active efforts to implement DRG-based controls have largelyprogressed in isolation from the development of non-financial PM reflecting theinterests of other important constituencies, such as professional staff and patients.This should be viewed in light of the growing dependence of Norwegian health careinstitutions on central government, notably manifested by the increasingly centralizedcontrol of funding and, most recently, the proposal to transfer ownership ofhospitals from county councils to the state. As far as clinical divisions are concerned,the internal reliance on DRG-based controls has accentuated this dependence as theirperformance is made increasingly and more directly contingent on governmentcontrolled sources of funding. This is mirrored by the relatively one-sided emphasisplaced on DRG-based PM and monitoring of other measures requested by variousstate agencies (e.g. waiting list statistics) within these divisions, while managers andcontrollers assume a relatively passive role as recipients of information generatedby existing, but technically deficient patient and personnel information systems. Inaddition, the considerable managerial time and efforts required for dealing withthe consequences of the implementation of DRG-based control appear to havelimited the propensity to initiate improvements in this respect (e.g. the limited attention paid to the new, integrated accounting and personnel information systems). 6 This supports Oliver’s (1991) contention that acquiescence to institutionalnorms espoused by an increasingly dominant constituency limits the possibilities ofsimultaneously accommodating far-reaching change initiatives reflecting the interestof other constituencies. However, more pro-active experimenting with non-financialPM as complements to budgetary control, was observed in ancillary divisions, whichare less directly subjected to institutional pressures associated with the new systemof funding but increasingly exposed to external market forces. This suggests that thelikelihood of pro-active attempts to develop multidimensional PM is greater wheremultiple constituencies make their influence felt (Brignall and Modell, 2000; Oliver,1991).These differences between clinical and ancillary divisions do not, however, providean exhaustive representation of senior management’s tactics in dealing with multipleconstituency interests. The actions of the General Manager and central administrativedepartments suggest some additional refinements of Oliver’s (1991) framework inthis respect.The criticality of managing professional staff interests in the Norwegian health care sector seems to be increasing, as manifested by the growing recruitment andretention problems and the results of the staff satisfaction survey conducted in 1999.An illuminating episode in this respect is the response of senior management to thissurvey. The pressures from professional staff, amplified by the attention from themedia and politicians, apparently alerted management to the need for legitimacyin the eyes of this constituency as indicated by the measures taken to improvethe situation in departments with widespread staff discontent and the growingefforts to develop indicators of staff-related aspects, such as absenteeism and staffturnover. In responding to the public debate following the disclosure of the staff survey report, however, the General Manager did not link the issue of staff discontent and fatigue to the related issue of enhanced political pressure for cost containment,although the possibility of meeting these contrasting priorities was questioned byother actors (e.g. some political parties). Instead, he drew attention to the morespecific steps required to improve working conditions. The public debate concerningthe staff satisfaction survey did not pivot around the primary symbolic elementrelied upon by management for legitimating and visualizing the changes in costmanagement (i.e. the growing use of DRG-based PM). Hence, management was ableto, at least temporarily avoid a potentially critical debate regarding the enhancedinternal emphasis on financial PM, while initiating more specific programmes aimed at resolving an emerging legitimacy problem, which was threatening to destabilizethe organization.The response to the staff satisfaction survey represents a relatively pro-active attempt to disconnect the contentious issue of enhanced emphasis on financialcontrol from the debate concerning the subsequent effects on performance aspects of greater concern to professional staff. As such, it supports Oliver’s (1991) hypothesis regarding compromise through balancing as a likely response to multiple, conflicting constituent interests. However, this balancing also embodies an element of decoupling in the sense that financial performance and concerns with staff discontentare disintegrated in the managerial rhetoric. This is in line with Brignall and Modell’s(2000) contention that more pro-active de-coupling of financial and non-financialPM forms a prerequisite for balancing between conflicting constituent interests andsuggests that compromise and avoidance tactics may be more intricately intertwinedthan Oliver (1991) recognizes.Our findings also cast some light on potential barriers to pro-active managerialmanoeuvering in developing multidimensional PM, besides those emanating fromthe growing dependence on a dominant external constituency, such as central government.Deeply embedded institutional constraints (e.g. differences in status, traditionallyentrenched functional barriers) have limited the cooperation between thepersonnel and finance departments, which currently hampers integration between financial and human resource aspects in the development of PM. Similarly, while thequality department seems to play a more significant role than the personnel departmentin external legitimacy-seeking, the limited progress in developing quality indicatorsfor more widespread use within clinical divisions should be viewed in lightof the problems in legitimating more formalized quality management internally. Theapproach informing the quality improvement efforts, which bear witness to somemimicking of private sector practices (cf. DiMaggio and Powell, 1983;Westphal et al.,1997), is apparently resisted by physicians on grounds that it is largely alien to clinicalrealities. These findings illustrate how institutional constraints, originating from inconsistent norms and rationalities, effectively contribute to the lack of integrationand coherence in PM. While de-coupling between various performance dimensionsmay occasionally be pro-actively effected by management, as illustrated by the responseto the staff satisfaction survey, it emerges as a result of action-stifling conflictsof interest in other circumstances. This suggests that Oliver’s (1991) conception ofde-coupling as a relatively pro-active tactic may not be universally valid and that the exact nature of this widely discussed response to institutional processes can onlybe discovered by closely examining it in the specific context in which it occurs (cf.Powell, 1988; Brignall and Modell, 2000).7The present study provides little evidence of the more radical responses to institutionalpressures proposed by Oliver (1991). This is not surprising, since radicalchange in highly institutionalized fields, such as the Norwegian health care sector, is unusual (Greenwood and Hinings, 1996) and may be expected to encounter considerable resistance. One example of this is the difficulties in promoting a qualityimprovement programme grounded in conceptual notions radically diverging fromthe institutionalized conception of clinical quality assurance. Nevertheless, the studyillustrates the usefulness of Oliver’s (1991) framework for analysing in greater detailhow managerial responses to institutional pressures translate into the developmentof multidimensional PM, while suggesting some conceptual refinements in thisrespect. Judging from our findings, a certain element of pro-active managerial manoeuveringis possible even in highly regulated institutional environments. A limitationof the study is that it was conducted over a relatively short period of time andadopted a narrow focus by examining the responses of primarily one, albeit important, category of organizational actors (senior management). However, it representsa first attempt to study the complex interactions between managerial choices and institutionalconstraints influencing the development of multidimensional PM, whichfuture research may build on to advance our knowledge of this topic.