DECISION SCIENCES INSTITUTE An Analysis of the Effects of Global Sourcing Practices on Ethical Consumption in the United States. Full Paper Submission

Size: px
Start display at page:

Download "DECISION SCIENCES INSTITUTE An Analysis of the Effects of Global Sourcing Practices on Ethical Consumption in the United States. Full Paper Submission"

Transcription

1 DECISION SCIENCES INSTITUTE An Analysis of the Effects of Global Sourcing Practices on Ethical Consumption in the United States Full Paper Submission Robert L. Bregman University of Houston David Xiaosong Peng University of Houston ABSTRACT This research evaluates consumer perceptions and responses to common global sourcing practices in order to evaluate ethical consumption. We extend the Hunt-Vitell model of ethics to establish a structural equation framework of consumers ethical decision-making. We then use multigroup analysis in structural equation modeling to contrast the ethical reasoning processes of consumer segments. Our results show that knowledge of common global sourcing practices cause a majority of U.S. consumers to fall within an ethical consumption segment. KEYWORDS: Ethical consumption, Structural equation modeling, Cluster analysis INTRODUCTION With greater awareness of ethical issues related to the procurement, manufacturing, and marketing of products, consumers are increasingly likely to discontinue purchasing products, switch to alternative products, and participate in organized boycotts of products and companies for ethical reasons (e.g., De Pelsmacker et al., 2005; Brunk, 2010; Turcotte, 2010). Issues driving this phenomenon include consumer disdain for unsavory marketing tactics, deficiencies in oversight and corporate governance, and mistreatment of humans, animals, and the environment. In this research we limit the drivers of ethical consumption to common global sourcing practices of multinational corporations. Because these practices are important to supply chains for consumer products and consumers are increasingly aware of these practices, focusing on these practices should be a good starting point for research in this area. The global sourcing practices targeted in this research are the treatment and management of workers, environmental conduct, community support, and supply chain oversight. The consequences of these practices that are likely to affect ethical perceptions of U.S. consumers were grouped into two distinct sets: those that affect a firm that uses the practices and those that affect the general public. We are interested in how perceptions of these practices and the consequences of these practices affect consumer purchase intentions. According to the Hunt-Vitell (H-V) framework of ethical decision-making (Hunt and Vitell, 2006), consumers will form ethical judgments of firms that use these global sourcing practices by combining their deontological evaluation of the practices and their teleological evaluation of the consequences of the practices. We believe that consumers ethical perceptions and decisionmaking processes can be used as a basis to segment consumers. Accordingly, we apply latent class cluster analysis to a large-scale survey designed around the H-V framework to separate U.S. consumers into segments based on their ethical perceptions of

2 the global sourcing practices we examine. We then develop structural equation models for each consumer segment and use multigroup analyses to differentiate what we define as the global sourcing ethical consumption segment. Our results show that this ethical consumption segment is very large and differs from others on ethical decision-making, perceptions of contentious global sourcing practices, and perceptions of consequences of the practices. RESEARCH FOUNDATIONS Global Sourcing Practices The treatment of overseas workers in global supply chains can be viewed as poor by western standards. Overseas workers are often subject to low pay, repetitive work methods, and unsafe working conditions. Environmental conduct refers to hazardous waste discharges and exploitation of natural resources by overseas suppliers. Community support refers to the lack of commitment by companies and suppliers to communities where they are located. This often takes the form of abandonment of facilities, use of overseas tax havens, and offshoring of profits to reduce corporate taxes. Supply chain oversight captures responsibility issues related to actions of supply chain partners and product origin. Consequences of Global Sourcing Practices The above sourcing practices can have consequences that may affect companies that use the practices and the general public. Company consequences include the ability to compete more effectively on price, generate higher profits, and capture a larger market share. Consequences that affect the general public include the loss of manufacturing jobs, loss of tax revenue from the closing of local plants and off-shoring of profits, and less transparency about the origin and content of products produced overseas. Conceptual Model Our conceptual model is based on the core elements of the H-V model of ethical decisionmaking. The H-V model suggests that individuals evaluate whether actions are right or wrong relative to their personal values (a deontological evaluation) and whether the consequences of the actions are beneficial (a teleological evaluation). Based on these two evaluations, individuals form an ethical judgment of the actions that will be used to determine behavior through the intervening variable of intention. In addition, the H-V model suggests that teleological evaluations may not only affect consumer intentions indirectly through the formation of ethical judgments, but also directly. Deontological evaluation, teleological evaluation, ethical judgment, and ethical consumption intention form the core constructs of the H-V framework and their relationships represent the consumer ethical decision-making process. RESEARCH PROPOSITION AND HYPOTHESES The primary goal of this research is to develop a better understanding of the ethical consumption phenomenon as it applies to global supply chains. Consumer segments can be identified through the clustering of consumers with similar means and relationships among the core elements of the H-V model of ethical decision-making. Modeling and Differentiating Consumer Segments Whereas deontological evaluation is driven by consumers ethical norms relative to the global sourcing practices, teleological evaluation is based on potential consequences of the practices. Prior research based on clear right versus wrong alternatives has found stronger links from deontological evaluation to ethical judgment than from teleological evaluation to ethical judgment for ethical respondents. We believe that the ethical consumption segment, which views the global sourcing practices more negatively, will view the ethical issues as moral

3 absolutes and will respond in a manner similar to prior research. Thus, we hypothesize the following: H1: The ethical consumption segment will evaluate and judge firms that use contentious global sourcing practices more negatively. H2: The ethical judgments of firms that use contentious global sourcing practices will be more strongly influenced by deontological than teleological evaluations for the ethical consumption segment. Perceptions and Effects of Global Sourcing Practices We expect that the ethical consumption segment will view the contentious global sourcing practices more negatively. Therefore, compared to other segments, the ethical consumption segment should exhibit a stronger link between deontological norms and deontological evaluations. Thus, we hypothesize the following: H3: The ethical consumption segment will perceive contentious global sourcing practices more negatively. H4: The ethical consumption segment will more closely link deontological evaluations of a firm that uses contentious global sourcing practices to their ethical norms relative to those practices. Perceptions and Effects of the Consequences of Global Sourcing Practices Since the company consequences benefit multinational corporations that use the practices, those consequences will likely be perceived more negatively by the ethical consumption segment. The general public consequences, which negatively affect the general public, will also likely be perceived more negatively by the ethical consumption segment. If the practices are perceived as morally wrong, the negative consequences of the practices for the general public will likely be viewed as unethical outcomes. Based on these observations, we hypothesize the following: H5: The ethical consumption segment will perceive the company and general public consequences of contentious global sourcing practices more negatively. Since the ethical consumption segment primarily bases ethical perceptions on predefined deontological norms, consequences (positive or negative) will have less effect on teleological evaluations for that segment. For example, the fact that companies can use the practices to compete more effectively is unlikely to improve perceptions of the company for the ethical consumption segment. Thus, we hypothesize the following: H6: The teleological evaluations of the ethical consumption segment will be less influenced by the company and general public consequences of contentious global sourcing practices. RESEARCH METHODS Questionnaire Design and Measures The questionnaire for this research was designed around our conceptual model. Respondents were asked questions to establish their norms relative to the global sourcing practices. Next a vignette describing a firm engaging in all of the practices was introduced. Then the respondents were asked to provide a deontological evaluation of the actions of the multinational firm. The consequences of the practices were then introduced and the respondents were asked to evaluate each of the consequences. That was followed by questions to establish a teleological evaluation of the firm. Respondents were then asked to provide an overall ethical judgment of the firm followed by their intention to alter consumption of the firm s product.

4 Respondent Profile Survey participants were randomly selected from a commercial database of consumers in the United States. Because the survey was conducted electronically and completed within one week, late response bias was not a concern. Our overall response rate of 22.6% is comparable to other survey-based studies in the supply chain area. A demographic profile of the respondents is provided in Table 1. Table 1: Respondent Profile Demographic Percent Demographic Percent Age Gender GenY (18-35) 31.1 Female 54.9 GenX (36-54) 31.5 Geographic region BabyBoomers+ (>54) 37.4 New England 7.2 Income East North Central 16.3 Less than $50, East South Central 3.8 $50,000 - $100, Middle Atlantic 10.9 $100,000 or greater 27.6 South Atlantic 17.1 Education level West North Central 7.4 High school or less 23.7 West South Central 8.4 Undergraduate degree 42.8 Mountain 8.2 Graduate degree 33.5 Pacific 20.6 Tests for Validity and Reliability We created a measurement model in PLS to test the reliability and validity of the measurement scales we use in this research. The measurement model includes a link between all constructs to provide cross loading information. Chin (1998) recommends removing any measures with a cross loading less than Per this recommendation, we removed one measurement item for Teleological Evaluation from our model. The composite reliability, average variance extracted (AVE), and squared correlations for the resultant measurement model were calculated. The composite reliability scores can be used to assess construct reliability. All of the scores are much greater than 0.70, indicating high reliability (Fornell and Larcker, 1981). The AVE results for each construct are all above the recommended threshold of 0.50, which also suggest strong convergent validity of the measurement scales. Analysis Methods Our analysis involved three major steps. First, we used latent class cluster analysis to segment consumers based on their ethical perceptions and decision-making processes. Second, we used SEM to model the ethical consumption segment and the remainder of consumers per the H-V framework. Third, we differentiated the model for the ethical consumption segment using multigroup analyses. Latent class cluster analysis, which evaluates both the relative means and relationships of constructs, can be an effective method for identifying latent classes (segments) in consumer data. In our analysis we used the standard Bayesian Information Criterion (BIC) from Schwarz (1978) for selecting the best number of clusters. We then used the Chin and Dibbern (2010) multigroup approach to differentiate the ethical consumption segment based on the path coefficients within our structural model.

5 DISCUSSION OF RESULTS A four latent class cluster resulted in the maximum BIC score. The Bonferroni-corrected p- values for pairwise comparisons with cluster 4 (representing 62.8% of the sample) show that cluster 4 has the lowest scores (poorest moral and ethical perceptions) for all measures capturing the deontological evaluation, teleological evaluation, and ethical judgment constructs. Cluster 4 includes the highest score for the ethical consumption intention construct, which measures the propensity of consumers to alter consumption of the firm s product. Modeling and Differentiating Consumer Segments While the pairwise comparisons show clusters 1-3 to significantly differ from cluster 4 (the ethical consumption segment), the differences vary among the clusters. Cluster 1 is the second largest cluster and the polar opposite of the ethical consumption segment. This segment has the highest scores for all measures capturing the deontological evaluation, teleological evaluation, and ethical judgment constructs; and the lowest score for the consumption intention construct. Cluster 1 represents those that are least affected by the contentious global sourcing practices. Clusters 2 and 3 represent smaller segments that fall between the extremes of clusters 1 and 4. Cluster 2 exhibits slightly negative scores for deontological evaluation, teleological evaluation, and ethical judgment constructs, but has the second strongest intention to switch consumption. Cluster 3 has the second lowest means for deontological evaluation, teleological evaluation, and ethical judgment constructs. However, this segment is neutral toward switching consumption. Because our focus is on the ethical consumption segment and how that segment differs from others on the ethical decision-making constructs, we compared cluster 4 to the remainder (clusters 1-3) to determine unique characteristics of the ethical consumption segment. To accomplish this comparison we analyzed the ethical consumption segment and the reminder as component-based structural equation models. Our results in Table 2 show the ethical decisionmaking path coefficients to be large and significant for both cluster 4 and the remainder. We then used multigroup analyses to evaluate differences in the paths for these two groups. Our results show the paths from deontological and teleological evaluations to ethical judgment to significantly differ between the two groups. The ethical judgment of the ethical consumption segment (cluster 4) is more affected by the deontological evaluation of ethical norms. These results support hypotheses 1 and 2. The global sourcing ethical consumption segment judges firms that use contentious global sourcing practices significantly more harshly than other segments. In addition, the ethical judgment of firms that use contentious global sourcing practices is more strongly influenced by deontological evaluations for the ethical consumption segment. Table 2: Ethical Decision-Making: Coefficients and Significance of Cluster Paths Cluster 4 (n = 509) Remainder (n = 301) Path Comparisons Structural path Path coeff. p-value Path coeff. p-value p4r Deontological evaluation Ethical judgment *** *** Teleological evaluation Ethical judgment *** *** Ethical judgment Ethical consumption intention *** *** Teleological evaluation Ethical consumption intention *** *** *** p < Perceptions and Effects of Global Sourcing The means and standard deviations for the global sourcing practices constructs for the ethical consumption segment (cluster 4) and the remainder (clusters 1-3) are shown in Table 3. All means are lower for the ethical consumption segment and all pairwise comparisons are highly significant. The SEM results for each segment are provided in Table 4. The only insignificant

6 link to deontological evaluation is from supply chain oversight practices for the non-ethical consumption segment (the remainder). Pairwise comparisons using multigroup analyses show the oversight and community coefficients are significantly different between the segments. Consumers in the ethical consumption segment differ from others in that they view all practices more negatively, and their deontological evaluations are more affected by supply chain oversight issues relating to control of supply chain partners and less affected by community issues such as the abandonment of facilities, use of overseas tax havens, and offshoring of profits to reduce corporate taxes. These results differentiate the two segments. Table 3: Global Sourcing and Consequences: Means and Std. Deviations for Clusters Cluster 4 (n = 509) Remainder (n = 301) Indicator Mean 1 Std Dev Mean 1 Std Dev Oversight Community Environment Workers Company General public All pairwise comparisons significant at p < Table 4: Global Sourcing Practices and Consequences: Coefficients and Significance of Cluster Paths Cluster 4 (n = 509) Remainder (n = 301) Path Comparisons Structural path Path coeff. p-value Path coeff. p-value p 4R Oversight Deontological evaluation *** Community Deontological evaluation *** Environment Deontological evaluation *** Workers Deontological evaluation Company consequences Teleological evaluation *** *** General public consequences Teleological evaluation *** *** *** p < We were able to model and differentiate the ethical consumption segment from the remainder. Hypothesis 3 is fully supported and hypothesis 4 is supported for supply chain oversight issues. Hypothesis 4 is not supported for practices related to the treatment and management of workers, environmental conduct, and community support. These results suggest that for the most part the effects of global sourcing practices on deontological evaluations are similar for both the ethical and non-ethical consumption segments. The key differences are the consistently lower perceptions for all practices, and the stronger effect of oversight practices and weaker effect of community issues on the ethical consumption segment. Perceptions and Effects of the Consequences of Global Sourcing The means, standard deviations, and structural equation results of each segment for the consequences of global sourcing practices are also provided in Tables 3 and 4. Both the company and general public consequences are perceived significantly worse by the ethical consumption segment. Although the effects of both consequences on consumers teleological evaluations are smaller for the ethical consumption segment, pairwise comparisons show only the differences in general public consequences to be significant. We found that the H-V framework fits well and the ethical consumption segment consequences are clearly differentiated. Hypotheses 5 is fully supported and hypothesis 6 is partially supported. Although perceptions of the consequences are lower for the ethical consumption segment, the effects of those consequences are generally smaller. Consequences (positive or negative) are less effective for the ethical consumption segment because their perceptions are dominated by their ethical norms.

7 CONCLUSIONS AND DIRECTIONS FOR FUTURE RESEARCH In this study we were able to empirically support most of our research hypotheses. Our results show that an ethical consumption segment can be identified by applying latent class cluster analysis to an ethical decision-making model based on the H-V framework. The ethical consumption segment detected in this manner is much larger than those determined with less sophisticated methods based only on attributes or purchase intention. Once identified, we show that the ethical consumption segment can be differentiated from other segments both in terms of ethical perceptions and the processes by which those perceptions affect ethical judgment and purchase intention within the H-V framework. Theoretical Contributions This study provides motivation for consumer behavioral extensions to research in the supply chain area. We introduce a new model-based approach for separating clusters and use that method to identify the ethical consumption segment of consumers relative to global sourcing practices. We differentiate an ethical consumption segment using the ethical perceptions and decision-making processes within the H-V framework. Significant differences that we find in the ethical consumption segment have implications for future research on global supply chains. And finally, we provide a picture of the true size and composition of the ethical consumption segment as it applies to contentious global sourcing practices. Our research also makes an empirical contribution by extending the application of latent class modeling to the determination of separable clusters. The latent class modeling approach provides richer results than more traditional hierarchical or K-means approaches typically used in the supply chain literature (e.g., Miller and Roth, 1994; Flynn et al., 2009) because it detects differences in model relationships (ethical decision-making in our case) in addition to differences in measures. Although application of latent class cluster modeling is relatively new, our results demonstrate that this approach can be very effective in differentiating a subgroup (the ethical consumption segment in our case) from a population. Managerial Findings This research shows that the ethical perceptions and decision-making processes of a majority of consumers are significantly affected by the global sourcing practices of many multinational corporations. The design of this research provided consumers with knowledge of a firm s involvement in these practices through a vignette. In practice, it is likely that many consumers are not aware that firms with global supply chains use or have suppliers that use these practices. In a sense, the ethical consumption segment identified in this research represents a potential segment that can be activated by information introduced through the media or special interest groups opposed to these practices. Firms with global supply chains should be aware that these practices represent potential liabilities that need to be addressed. Conclusions The goal of this research was to provide a better understanding of ethical consumption and how it affects perceptions of multinational corporations engaged in global sourcing. It is our belief that consumers are a primary consideration when designing and managing global supply chains, and as such, consumers should be incorporated into the research literature for that area. In this study we take the first step toward introducing ethical consumption and consumer perceptions to the supply chain literature. References References available upon request.