برون سپاری فناوری اطلاعات از دیدگاه های ترکیبی

This paper explores the value of a configurational approach to IT outsourcing by developing a framework for IT outsourcing effectiveness. Taking a process view of outsourcing and drawing on the relational view of the firm, the framework identifies four high-level dimensions that correspond to an organization’s resource position in four key areas: organizational IT value position, organizational IT asset position, relational asset position, and relational capability position. A novel structured method is used to identify the congruent outsourcing configurations within the range of possible outsourcing configurations based on the interdependencies among the four dimensions. Three congruent outsourcing configurations, designated asset dependence, relational dependence, and independence, emerge from this analysis. Drawing on the assumptions of configurational theory about organizational change and taking a dynamic perspective, the framework is extended to describe how organizations transition between outsourcing configurations over time. This paper demonstrates how a configurational approach can address three potential problems of the outsourcing literature and advance outsourcing research.

Over the past two decades, the growth in the practice of outsourcing information technology (IT) has been one of the most widespread developments in meeting organizations’ IT needs (Dibbern et al., 2004). As interest in this phenomenon has gained momentum, research has drawn upon a variety of theories to gain a deeper understanding of outsourcing1 decisions and their consequences. Among these theories, those most commonly addressed in outsourcing research are the resource-based view (RBV), the resource dependence perspective, knowledge-based theory, transaction cost economics (TCE), agency theory, and social exchange theory (Cheon et al., 1995, Dibbern et al., 2004, Goo et al., 2007, Hancox and Hackney, 2000, Kern et al., 2002 and Tiwana and Bush, 2007). Attention has also been paid to other theories, such as institutional theory (Vitharana and Dharwadkar, 2007), innovation diffusion theory (Hu et al., 1997), and residual rights theory (Lee et al., 2004). All the above theoretical frameworks have been applied within three streams of research, reflecting a strategic, economic, or social perspective of outsourcing (Goo et al., 2007 and Lee et al., 2003). This theoretical proliferation has been valuable in advancing the understanding of outsourcing as a unique form of interorganizational relationship. However, this paper suggests that the potential of research to advance the understanding of outsourcing relationships may have not been fully realized because of the gap between the rich theoretical foundations of previous studies and their tendency to rely on a reductionistic approach, which seeks to disaggregate organizational phenomena into unidirectional linear relationships between organizational attributes (Miller, 1981). The paper proposes that the heavy reliance on a reductionistic approach may be associated with three potential problems – unresolved contradictory findings, oversimplified research models, and paucity of dynamic views – and that a promising avenue to address these problems is to more widely adopt a configurational approach, which seeks to explain organizational phenomena in terms of coherent patterns of multiple organizational attributes (Meyer et al., 1993 and Miller, 1986). The potential benefits of a configurational approach to outsourcing are perhaps best illustrated by considering these three problems in the context of outsourcing research.

The theoretical framework developed in this paper is amenable to empirical testing. The framework is based on six propositions that can be used to develop empirically-testable hypotheses. These propositions address the interdependencies among the dimensions that potentially influence outsourcing effectiveness (Propositions 1–3), the association between these dimensions and outsourcing effectiveness (Proposition 4), and the patterns of change in outsourcing relationships (Propositions 5 and 6). Given this diversity in the nature of the propositions, various research methods should be employed to empirically validate the framework. A review of empirical investigations of configurational hypotheses shows that various methods have been employed for data collection, including surveys, multiple case studies, archival research, and panel studies. These methods can be used in combination to test the six propositions formulated in this paper in three stages. Propositions 1–3, which address the interdependencies among the four organizational and relational dimensions, should be tested in the first stage. The hierarchical conceptualization presented in Table 1 may be used in this stage to operationalize the four dimensions. Many of the attributes conceptually defined in Table 1 have been operationally defined in previous studies, and these operational definitions may be used in the construction of quantitative measures. These three propositions have a static orientation, and therefore they can be tested by collecting data at one point in time. A common method of data collection in static configurational research has been a survey, either cross-sectional (Bergeron et al., 2004; Ferratt et al., 2005) or single-industry (Bensaou and Venkatraman, 1995; Delery and Doty, 1996; Ostroff and Schmitt, 1993). Another common method has been multiple case studies, primarily involving on-site interviews in multiple sites (Brown and Magill, 1994; Cullen et al., 2005; Sambamurthy and Zmud, 1999). Multiple case studies are presumably more appropriate for testing Propositions 1–3 because of the need to operationalize attributes of a more intangible nature, such as those that comprise the dimension of relational capability position.The interviews with managers involved in the design and implementation of outsourcing arrangements should involve quantitative measures (Brown and Magill, 1994; Cullen et al., 2005) to facilitate the categorization of organizations as either high or low on the four dimensions. The problem with using multiple case studies is that a relatively large number of sites would have to be selected, given the range of possible outsourcing configurations listed in Table 5. This requirement increases the attractiveness of a survey as a means of producing a sample of configurations that adequately represents the possible configurations.