Accounting, Risk Management and the Selection of Interactive Controls: Which, When and Why?

Abstract

Taking a multiple-control perspective, I investigate a control debacle and its aftermath at a financial services company (MultiBank), focusing on an insurance division (EurInsurance) that suffered large losses in the European insurance crisis of 2002-2003. The study tracks the promotion and use of a set of accounting and risk controls put in place to control the troubled insurance division, and discusses how and why particular management control systems shift in and out of top managerial focus. The study investigates Simons’ (1990, 1991) argument that it is top management’s view of a firm’s key strategic uncertainties that motivates their choice of control systems to be used interactively. Combining the process perspective of a longitudinal field study with the institutional logics perspective, I argue that top management’s control choice is motivated by both the logic of functionalism (relevance) and the logic of appropriateness (legitimacy) of particular controls - and that both of these are socially constructed by proponents.First, behind the various control systems there are active controller groups who, in competition for executive-level visibility, further their solutions for organizational control problems and engage in ‘credentializing’ (Power, 1992) to support their claims. At MultiBank, accountants drew on the institutional logic of accounting as a legalistic, rules-based practice, while risk controllers relied on the cultural authority of financial economics and the “full fair value” logic. Second, top management’s interactive use of a particular control system sends a signal to external stakeholders about the firm’s internal control style and management priorities. Therefore, the control choice is motivated both by the relevance and the institutional appropriateness of particular controls. As external requirements change and the definition of institutional appropriateness shifts, different organizational control groups get the opportunity to become implicated in interactive control and agenda-setting. In the special case of mutually incompatible control systems, when top managers must trade off relevance for appropriateness (or the other way round), their choice of interactive control will be driven by what they perceive to be the stronger requirement. At MultiBank, institutional appropriateness was the stronger requirement; the lack of it prevented an otherwise informationally relevant risk control system from prevailing as an interactive control system.