A field research, commissioned by the Dutch Foundation for Management Studies, on the state of the application of the concept of business unit-organization (M-form) in the Netherlands, has produced some noteworthy results. Through interviews with 36 organizations, most for-profit companies (including Dutch subsidiaries of multinationals), two non-profit institutions and two government agencies, the following was found. The concept of unit management, that is to organize the firm in a number of self-contained business units as profit centers, is still very strong in the theory-in-use of managers and as a basis for accounting systems. However virtual no business unit any longer is self-contained organized, contrary to the definition of the M-form. All business units, in varying degree, depend on resources outside the unit to achieve their objectives. In at least five cases the mental anchoring to the unit-concept has resulted in debilitating problems with respect to account management and with project management, organizational entities which are needed to satisfy the Chandlerian criteria of fit-to-market and fit-to-strategy. A small number of companies have overcome this problem and similar dilemmas by defining accountability for turnover and profit & loss simultaneously over multiple dimensions (product, region, account, market segment, industry), for each dimension a separate manager is accountable, contrary to what is assumed in the literature on economic organization theory, accounting theory and management control. Examples of firms doing so are ABN AMRO, IBM, Microsoft, ASML, PwC and some others. Those companies have commensurate multidimensional management reporting, management information, and coordination & control processes, and are successful companies (except ABN AMRO which is unbundled by three other banks). The reason to operate a multidimensional organization seems to be the emergence of multidimensional markets, to need to exploit economies of scope, especially with respect to knowledge, to have higher effectiveness in management control, to be more flexible, to reduce resource biased risk averse target setting, and to appropriate more value from the market in the case of network industries. The concept of the multidimensional organization was presaged by Ackoff (1977) and Prahalad & Doz (1979) and Prahalad (1980) but has virtual disappeared from the literature since then. Its present emerging can be explained by using Arrow's (1974) The Limits of Organization, especially Arrow's prediction of the consequences of a fall in costs of information for the agenda of the firm. This paper will describe the multidimensional organization as an empirical phenomenon, explaining its workings and its raison d'être in operational terms. In addition to that, using Williamson's efficiency hypothesis, the efficiency of the multidimensional organization is discussed whether the multidimensional organization is more efficient as is the M-form. This paper does not in the first place aim, as is common, to make a contribution to existing literature. Its first purpose is to report on a new empirical phenomenon
