Performance payoffs from manufacturing flexibility: the impact of market-driven mobility.

The capabilities needed to achieve excellent manufacturing performance have changed in recent years (Hayes, 2000). Global competitive pressures have stimulated the need for rapid organizational change. Jeff Bleustein (CEO of Harley Davidson) declared, "The only thing that can stop us is if we get complacent. Even though we've been successful, we can't stand still." (Helyar, 2002: 124). Bleustein's comment echoes corporate America's sentiment that adjusting to environmental change is one of the primary challenges facing organizations today. It also reflects some of the unconventional measures taken by manufacturing firms to maintain their competitive position in the marketplace (Skaggs and Droege, 2004).

Manufacturing flexibility gives an organization the option to adjust to changing conditions in its environment. However, it is quickly becoming evident that infusing flexibility in manufacturing systems is not as simple or as straightforward as initially envisioned. The reality is that building flexibility in a manufacturing environment is challenging and often involves tough trade-offs (Bengtsson, 2001; Kulatilaka, 1988).

There exists a significant body of literature that identifies manufacturing flexibility as a competitive priority of the organization. For example, Price, Beach, Muhlemann, Sharp and Paterson (1998) have studied it within the context of decision support systems that enable organizations to adjust their corporate strategies. Newman, Hanna and Maffei (1992) researched strategic uncertainties faced by organizations that required high levels of manufacturing flexibility. Such research positions manufacturing flexibility as a critical precursor to manufacturing performance and organizational performance. However, empirical evidence to support the relationship between manufacturing flexibility and manufacturing performance is not equivocal. Some researchers (e.g., Bengtsson, 2001) have argued that increased manufacturing flexibility does not necessarily improve manufacturing performance. Empirical evidence (e.g., Das and Nagendra, 1993; Suarez et al., 1995) seems to support such a contrary contention.

Past research provides valuable insights into the relationship between elements of manufacturing flexibility and manufacturing performance. Several studies have been conducted at the dimensional level. However, these have been rather dispersed, and have not resulted in a comprehensive interpretation of the relationship between these two constructs. In summary, researchers have arrived at conflicting positions regarding the relationship between manufacturing flexibility on manufacturing performance.

Objectives of the Study

The discussion presented above raises three interesting questions. First, why has empirical evidence not wholeheartedly supported the theoretical arguments of researchers, or the practical wisdom of business executives, that flexibility has a significant and positive impact on performance? Second, could a re-specification of the manufacturing flexibility construct shed new light on the flexibility-performance relationship? Third, if a significant relationship does exist, how exactly are these constructs related at the dimensional level?

Finding answers to the first two questions would certainly benefit academicians. For example, the use of a re-specified and more integrated model would allow us to tease out the relative impact of each dimension of manufacturing flexibility on manufacturing performance. Answers to the third question will provide valuable tools to business executives who are constantly looking for ways to improve manufacturing performance.

I will address each of the three questions in this article. I begin by reviewing the literature on the linkages between manufacturing flexibility and manufacturing performance. I then provide an overview of the dimensions of manufacturing flexibility and manufacturing performance. Next, I review the manufacturing literature and re-specify the manufacturing flexibility construct. This information is then used to hypothesize relationships between manufacturing flexibility and manufacturing performance at (1) the aggregate level and (2) the dimensional level. I then test the hypotheses using data obtained from a sample set of manufacturing firms in four two-digit SIC codes. The final section addresses statistical results, findings of the study and implications for academician and practitioners.

THEORETICAL UNDERPINNINGS

The Manufacturing Flexibility--Manufacturing Performance Relationship

It was not until the early 1980s that researchers began to systematically investigate the relationship between manufacturing flexibility and manufacturing performance. Since then, a fair amount of theoretical and empirical contributions have been made. However, results from these studies have been mixed and, at times, contradictory. The narrow focus of some of these studies, and the apparent contradictions in findings, are not surprising. The study of manufacturing flexibility is a recent phenomenon and theoretical frameworks have been exploratory in nature. This behooves researchers to choose narrowly focused research efforts. The existence of contradictory findings should not be startling either. They are part of the consolidation process that is needed to move the field forward.

Some researchers have posited a positive relationship between manufacturing flexibility and manufacturing performance (e.g., Gerwin, 1987; Olhager, 1993). Others (e.g., Bengtsson, 2001) have argued that the relationship may not be as straightforward. Using an option-theoretic perspective, Bengtsson argues that investments made in flexible manufacturing equipment are often higher than those required for dedicated equipment. In addition, it is entirely possible that the firm may never be called upon to use the installed flexibility to its fullest extent. Therefore, payoffs from flexibility may not be as forthcoming as one might expect.

The Manufacturing Performance Construct

A survey of the literature indicates that manufacturing performance has been viewed as a multi-dimensional construct. Cost, quality and delivery have traditionally been viewed as three key dimensions of manufacturing performance. While these have been viewed as been somewhat introspective measures (D'Souza and Williams, 2002), they are well accepted in academic research. For example, Jayaraman, Droge and Vickery (1999) support the use of these three dimensions. They also support Droge, Vickely and Markland's (1994) argument to exclude "innovation" as a performance dimensions. Droge et al. (1994) suggest that manufacturing's responsibility for innovations was much less than its responsibility for the other performance dimensions. Discussions with manufacturing managers confirmed that these are the three most common dimensions used to measure manufacturing performance. Table 1 presents the three dimensions of manufacturing performance and the items used to operationalize these dimensions in this study.

The Manufacturing Flexibility Construct

There is general agreement among researchers that manufacturing flexibility is a multi-dimensional construct (D'Souza and Williams, 2000; Gupta and Gupta, 1991; Swamidass, 1988; Upton, 1994; Watts et al., 1993). Researchers have identified several taxonomies of manufacturing flexibility dimensions. These fall into two broad categories. The first category is the set of dimensional taxonomies derived from manufacturing imperatives/priorities. There have been several excellent reviews that have synthesized these taxonomies into unifying frameworks. (For detailed reviews of these frameworks and their underlying dimensions see, for example, Beach et al., 2000; Gupta and Somers, 1992; Narasimhan and Das, 1999; Sethi and Sethi, 1990; Vokurka and O'Leary-Kelly, 2000.) Sethi and Sethi's framework is one of the more widely accepted of these operationalizations of manufacturing flexibility. They identify two types of manufacturing flexibilities--program flexibility and market flexibility. In their framework, program flexibility is represented by two elements, namely process flexibility and routing flexibility. Market flexibility is represented by three elements, namely product flexibility, production volume flexibility, and expansion flexibility. This framework has been well accepted in the literature and researchers have borrowed extensively from it (see, for example, Koste and Malhotra, 1999; Upton, 1994; Watts et al., 1993). In addition, the framework is consistent with "the dominant orientation" view of the organization (Collins et al., 1998; Wheelwright, 1984).

The other category of taxonomies includes those that identify a parsimonious set of overarching factors that characterize flexibility. In their review of the literature, Parker and Wirth (1999) identified several two-dimensional taxonomies. These include system versus machine (Buzacott, 1982), static versus dynamic (Carlsson, 1992), range versus mobility (Slack, 1987), potential versus actual (Browne et al., 1984), and short term versus long term (Carter, 1986). Of these two-dimensional taxonomies, the range-mobility taxonomy is the one most frequently referenced in the more recent literature. Gerwin (1993), Slack (1987) and Upton (1994) succinctly highlight the practical applicability of the range versus mobility taxonomy. In these conceptualizations, the "range" dimension reflects the extent to which the manufacturing system can adapt to process variations or market variations without undue impact to performance. The "mobility" dimension reflects the rate at which the system can perform the adaptation process.