The quality of standard, routine and nonroutine processes

Abstract

Efforts to increase organizational effectiveness using standardization and quality techniques have been successful in repetitive production and administrative processes but less so when dealing with nonroutine processes typical of professional organizations. Routines are defined very broadly in organization theory as either mind-numbing repetition, repositories of knowledge, or effortful accomplishments. In this article, processes are analysed as systems with distinct input assessment, algorithms and output-generating action phases. These are structured differently depending on how they are set up to deal with variation (deviations from explicit targets) and variety (distinct but functionally equivalent targets). Thus processes can be classified into three types. Standard processes are set up to deal with a single variety using binary logic. Routine processes can distinguish a limited amount of variety using fuzzy logic. Nonroutine processes are open systems in which unrestricted variety is interpreted and as signed meaning. The implications of these process types are discussed in terms of the identification of quality errors and defects, change and learning.

Keywords: standards, standardization, routine, routinization, noaroutine, defects, errors quality management, statistical process control

Introduction

Quality and productivity improvement through standardization and statistical process control is a modem success story. Economically feasible methods for controlling the uniformity of output enabled the use of interchangeable parts, which, in turn, made possible industrial mass production, economies of scale and improvements in wealth and welfare (Womack et al. 1990). Many organizations other than manufacturers have tried to apply these methods to improve their operations. However, results have not always been favourable. The perceived quality of services still lags behind that of industrial products (Edvardsson et al. 2000). QM (quality management) in professional organizations frequently meets resistance that tends to reduce it to empty formalism (Zbaracki 1998). Software quality remains in its infancy (Hoch et al. 2000).

The success of QM has been compared across nations (Rommel et al. 1996; Dahlgaard et al. 1998), but there are no broad studies comparing various industries or organization types. The QM implementation literature tends to focus on certain types of organizations (Cole and Scott 2000). Perrow (1967) has classified organizations and their technologies by the number of exceptions they have to handle, and by the degree to which a search for a solution to an exception is analysable. Organizations in which there are few exceptions and problems are analysable Perrow calls routine. These are typically mass manufacturers and high-volume service producers most of whose processes involve identical repetition of standardized tasks. The opposite type, nonroutine organizations, handles a lot of exceptions that are not analysable following predetermined schemes. Although the evidence is not conclusive, it appears that most of the QM success cases are from routine organizations (Easton and Jarrel 1998; Flynn et al. 1995) or fr om routine subprocesses within nonroutine organizations (Silvestro 2001). This should not be surprising, since most of the technical (Wheeler and Chambers 1992; Mitra 1998; Oakland 2000) and classical literatures (Shewhart 1931; Deming 1982; Ishikawa 1985; Juran 1992) of QM deal with repetitive processes.

The concept 'routine' has been used in organization theory to describe the stability that comes from repeating the currently best-known practices (Cyert and March 1993; Nelson and Winter 1982). It also occupies a place in the theory of the firm, particularly the versions that see the firm as a bundle of specific knowledge or dynamic capabilities embodied in the firm's routines and competencies (Nonaka and Takeuchi 1995; Penrose 1995; Tsoukas 1996; Spender 1996; Fransman 1999; Eisenhardt and Martin 2000). Although the emphasis is on the function of routines, the literature is not very specific in defining what they are and how they work. The purpose of this article is to explore the inner workings of routines and develop distinctive definitions of standard, routine and nonroutine processes.

Routines and Nonroutines in Organization Theory

Routines as Mindless Repetition

Human relations and contingency theories associate routines with the mind-numbing repetition that goes with mechanistic organizations, such as short-cycle assembly work or working a cash register at a supermarket. Mechanistic forms of management, such as centralization, formalization and directive leadership, are assumed to increase the efficiency with which workers perform unvarying and repetitive tasks (Bums and Stalker 1961). Such tasks are not performed consciously as the result of cognitive, rational decisions. Rather they are performed as habitual responses to familiar situations (Weiss and Ilgen 1985) or following standard operating procedures (Cyert and March 1993). Routines are identical to the scripted or mindless behaviour that occurs when there are event schema, categorizable stimulus clues, action rules, minimal required effort, an absence of unstructured subroutines and an absence of interruptions and unmet expectations (March and Simon 1958). Routine work is primarily completing structured prob lems, tasks characterized by accuracy of detail, short-term horizons, predominantly internal information and a narrow scope (Pava 1983). It proceeds through linear and sequential conversion processes -- a series of particular steps that yields predefined output (Pasmore 1988). Routine work is associated with a lack of autonomy and fulfilment and thereby also with alienation and a low level of personal control (Ross and Wright 1998). Routines are prevalent in situations where there are few exceptions and the search for responses to exceptions can proceed analytically and logically (Perrow 1967). In their study of health and welfare organizations, Hage and Aiken (1969) state that an organization has a routine workflow if its clients are stable and uniform and much is known about the particular process of treatment.