The bottom-line financial impact of supply chain management has been of continuing interest. Building on the operations strategy literature, Fisher's (1997) conceptual framework, a survey of 259 U.S. and European manufacturing firms, and secondary financial data, we investigate the relationship between supply chain fit (i.e., strategic consistencies between the products’ supply and demand uncertainty and the underlying supply chain design) and the financial performance of the firm. The findings indicate that the higher the supply chain fit, the higher the Return on Assets (ROA) of the firm, and that firms with a negative misfit show a lower performance than firms with a positive misfit.
Although it is intuitive that supply chain management is likely to have a positive impact on firm performance, most of the evidence that we have seen in the literature is either anecdotal or based on case studies. There is neither much large-scale empirical proof of this impact nor systematic analysis and documentation of its magnitude. Furthermore, the supply chain management literature has focused more on efficiency improvement and cost reduction in supply chain operations and less on the phenomenon of strategic consistencies between the characteristics of a product and its underlying supply chain, i.e., supply chain fit.
The concept of supply chain fit has been popularized by Fisher's (1997) conceptual supply chain–product match/mismatch framework and has its roots in the manufacturing and operations strategy literature. Forty years ago, Skinner (1969) called for a more integrated view of a firm's strategy and its manufacturing function. Over the years the research on competitive priorities in operations management, configurations of operations and manufacturing strategy, the successful alignment of product characteristics and competitive strategy with a firm's operations strategy, and performance implication thereof has grown considerably (e.g., Boyer et al., 2000, Hayes and Pisano, 1996, Joshi et al., 2003 and Ward et al., 1996). The extension of this research in the supply chain management literature just began to emerge (e.g., Qi et al., 2009 and Qi et al., 2011).
In this article we augment this research in three important ways. First, we further extend the operations and manufacturing strategy perspective towards the more recent supply chain thinking (Chen and Paulraj, 2004 and Kouvelis et al., 2006). We achieve this by assessing whether the firms’ supply chain priorities are in line with their products and business strategies. Second, we conceptualize supply chain fit as “fit as matching” (Venkatraman, 1989). As a consequence, deviation score analysis allows us to go beyond a 1:1 (‘all or nothing’) association between product characteristics and supply chain design. Furthermore, we can distinguish between positive and negative misfit. Third, we assess supply chain management's bottom-line financial impact and the magnitude of this impact by measuring performance with objective financial metrics from secondary data (Boyer and Swink, 2008 and Roth, 2007).
From a managerial perspective, achieving supply chain fit is challenging1 and supply chain misfits may be consequential. For example, Hensley and Knupfer (2005) estimate that the cost of supply chain misfit among carmakers and parts suppliers in the U.S. automotive industry is in excess of USD 10 billion each year. Hence, guidelines that help firms understand how to achieve supply chain fit would be valuable. By developing an understanding of the impact of supply chain fit on performance, firms will be well on their way to build such guidelines and their own models for supply chain excellence. By using a financial performance measure (i.e., Return on Assets, ROA) as an outcome of supply chain fit (or misfit) – as we do in this research – we speak in the language of managers who are more familiar with such measures than with subjective, perceptual performance measures. Relating supply chain fit to ROA will result in a higher impact of our research in corporate practice.
The rest of the paper is organized as follows. In Section 2, we begin by providing the theoretical and conceptual background from the operations strategy literature in support of our hypothesis. We then present our study's methodology, introduce the measures used in our study, and describe the sample in Section 3. Section 4 assesses the reliability and validity of our measures, followed by regression analyses in Section 5, and two post hoc analyses in Section 6. In Section 7 we discuss our results and provide theoretical and managerial implications. Finally, we conclude in Section 8 with limitations and suggestions for future research.
We scrutinized the hypothesis with a series of regression models. All models were estimated using ordinary least squares (OLS) estimation in the R system for statistical computing version 2.13.0 (R Development Core Team, 2010). The critical assumptions underlying OLS regression analysis were checked; i.e., (1) the residuals are normally distributed; (2) the residuals are of constant variance (homoskedasticity) over sets of values of the independent construct; and (3) multicollinearity of the independent construct is within an acceptable range (Cohen et al., 2003). To this end, the regression model was subjected to a visual residual analysis using normal Q–Q plots. No obvious outliers were detected and residuals appeared to be approximately normally distributed. Homoskedasticity was checked using the Breusch–Pagan test (p > .05), which did not indicate a serious problem with heteroskedasticity. The bivariate correlations between the independent variables as well as variance inflation factors (VIF) were within acceptable ranges, indicating that multicollinearity did not pose a serious problem to the regression analysis. In summary, the conducted tests provided no grounds to assume the inappropriateness of the chosen method. Nevertheless, to correct for possible heteroskedasticity and obtain correct standard error estimates, we used the Huber–White correction ( Huber, 1967 and White, 1980) implemented in the package sandwich in R (Zeileis, 2004).
The performance variable ROA was first regressed on the control variables (Model 1) and then the independent variable SCF was entered (Model 2). Table 7 reports the regression results including the increments to adjusted R2 and the significance of the regression equations. The baseline regression models with all 259 firms included show that misfit has a negative impact on performance (β = −1.268, p < .001), providing support for our hypothesis that supply chain fit is positively associated with performance. The average ROA for the 259 firms was 6.49%.
