|
" Investigating Network Structure, Social Capital, and Sustainability in Operations and Supply Chain Management "
Tacheva, Zhasmina Y.
Simpson, Natalie
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1106887
|
| Doc. No
|
:
|
TLpq2418726295
|
| Main Entry
|
:
|
Simpson, Natalie
|
|
|
:
|
Tacheva, Zhasmina Y.
|
| Title & Author
|
:
|
Investigating Network Structure, Social Capital, and Sustainability in Operations and Supply Chain Management\ Tacheva, Zhasmina Y.Simpson, Natalie
|
| College
|
:
|
State University of New York at Buffalo
|
| Date
|
:
|
2020
|
| student score
|
:
|
2020
|
| Degree
|
:
|
Ph.D.
|
| Page No
|
:
|
229
|
| Abstract
|
:
|
Over the past two decades, the topic of sustainability has received increased attention from both academia and business. Defined by the World Commission on Environment and Development as “development that meets the needs of the present without compromising the ability of future generations to meet their own needs,” sustainability is most often examined in the context of three categories: environmental, social, and economic. Despite Operations and Supply Chain Management’s (OSCM) early engagement with questions of sustainability, such as reverse logistics, closed-loop supply chains and product stewardship, this research area is still commonly viewed as an emerging domain, arguably too new to be considered mainstream. Following a similar trajectory of increasing interest, network science has also experienced a remarkable growth as a theoretical and methodological lens in OSCM. Like sustainability, network science is also still thought of as too nascent to be deemed conventional. The shared novelty of these two domains and their distinctly interdisciplinary nature (Linton, Klassen, and Jayaraman 2007), suggest that blending them into a common research agenda could yield exciting new insights of import to OSCM theory and practice alike. To this end, this dissertation provides a comprehensive framework for the meaningful integration of sustainability research and network science in OSCM by leveraging network analysis, machine learning, and econometrics. The proposed framework contributes to three distinct yet related sub-streams of the OSCM literature on sustainability in each dissertation chapter. First, it develops a novel method for surveying extant literature which enables a holistic analysis of sustainability in the context of the general domain of OSCM. Second, it introduces the role of a firm’s supply network position—consumer-facing or business-to-business—as an important factor in the relationship between corporate sustainability initiative participation and company performance. Third, it presents an effective way of quantifying and characterizing emerging phenomena in one of the nascent yet critical sub-streams of OSCM sustainability and crisis management—humanitarian operations. A detailed summary of each chapter follows. Chapter 2 uses network analytical techniques to trace the emergence and growth of the domain of sustainability in the supply chain management literature. Through exploring the community structure of the general OSCM body of research, the study aims to develop a more thorough perspective of the place of sustainability research in OSCM than offered by the extant literature, which has relied on expert judgment and subjective opinion in selecting the sample of sustainability research to be interrogated. In contrast, our study builds on a novel approach proposed by Minas, Simpson and Tacheva (2020), which helps researchers identify relevant studies in a more objective way, thereby alleviating the risk of personal bias and resulting in a more comprehensive research sample. The use of this approach yields a more expansive and yet relevant literature sample than previously identified, and points to the need for a more comprehensive literature selection method. This need emerges from the finding that the metadata of many studies germane to the topic of sustainability are not indexed in a way that lends itself to easy discovery through key words and would have been missed, had a traditional sampling procedure been used. More precisely, the results of the study show that far from being a marginalized peripheral community in the OSCM research domain, sustainability has positioned itself in 15 short years as the largest bibliographic cluster in OSCM, comprised of several streams of research with distinct themes and methodology. We find that empirical studies on the topic of sustainability, both in terms of survey-based research and secondary data analysis, are relatively homogenous and form a community within the general OSCM literature network easily identifiable through conventional network partitioning algorithms. The same is not true about mathematical modeling on the topic of sustainability however; the research landscape formed by this methodology is far more fragmented and not readily “visible” to off-the-shelf bibliometric techniques. We use a supervised machine learning model to identify relevant sustainability studies not present in the main sustainability community and use a series of statistical comparison procedures to establish the characteristics of these autonomous streams of sustainability research vis-à-vis the central sub-stream. The results point to the presence of as many as 20 distinct sub-streams of sustainability research in OSCM, the largest of which accounts for the majority of empirical work, while the rest use mathematical models to address sustainability themes such as: water safety, food waste, remanufacturing, recycling, blood inventory, and cap and trade. We further discover that research sub-streams have a strong geographical component. Most notably, we find that the majority of research on sustainable building comes from Europe and Asia, while a significant portion of sustainability research on forest ecosystems, food waste and water safety comes out of the African continent and especially the East Coast of Africa. We also find that certain countries have a comparable research profile because they examine similar sustainability topics, such as the United States, the United Kingdom, Italy and Germany, where empirical studies constitute the bulk of sustainability research. Iran and Turkey on the other hand specialize in mathematical modeling, with a strong focus on transportation network research. To our knowledge, this is the first study to flesh out these sustainability research nuances which would not have been uncovered, had traditional survey techniques been used. Chapter 3 is an empirical investigation of two firm-level outcomes in the context of corporate sustainability. First, we examine the company factors that play a role in the likelihood of an organization participating in a sustainability initiative. Next, we assess the effect of participating in a sustainability initiative on firm performance. In both cases, we focus on supply network position as a moderator of the relationship between company- and industry-level characteristics and the outcome variable, not previously considered in the literature. We argue that not taking into consideration the place of a company in its supply chain is a potential reason for the lack of consensus among scholars as to whether sustainability efforts are an enabler or an inhibitor of company performance. More specifically, proximity to the end customer has been theorized to be an important determinant of the institutional pressure businesses experience to join a sustainability initiative, such that consumer-facing companies are subject to the greatest stakeholder demand for sustainability investment, but also reap most of the reputational benefits associated with green efforts (Villena and Gioia 2018). Using this idea, known as chain liability (Hartmann and Moeller 2014), we adapt the theory of supply chains as complex adaptive systems (Choi et al. 2001) to the question of sustainability initiative announcement and participation likelihood depending on supply network position. We find that not taking into account the position of sustainability initiative-announcing companies in relation to the end customer (B2C or B2B), leads to the wrong conclusion that initiative announcement impacts firm performance negatively. In contrast, differentiating between B2C and B2B companies reveals that this negative performance effect is driven mainly by the reduction in B2B performance following a sustainability initiative announcement, whereas B2C companies are associated with improved performance as a consequence of announcing an initiative. We extend the literature by introducing a more granular analysis of several types of sustainability initiatives: individual company green efforts, initiatives with supply chain partners, and broad cross-industrial alliances. We find that broader initiatives including supply chain partners or inter-industry collaborators are associated with higher performance gains than individual initiatives.
|
|
|
:
|
We also find that there is a significant interaction between initiative type and supply chain role, such that supply chain initiatives and broad horizontal alliances are associated with a higher marginal improvement of performance for customer-facing companies. On the question of drivers of sustainability initiative likelihood, our findings suggest that companies with larger top management teams (TMT) and lower average TMT age are more likely to announce a sustainability initiative. While the effect of average age is relatively stable for B2C firms, in B2B corporate settings, the effect reverses direction, as older teams become more likely to engage in a sustainability initiative. This is an important finding because it suggests that B2B TMTs with a lower average age are more likely to disregard stakeholder pressure for sustainability commitment and thus to expose the supply chain as a whole to the liability risks outlined by Villena and Gioia (2018). Lastly, Chapter 4 extends the application of network analysis in OSCM sustainability research to the analysis of an emerging phenomenon in humanitarian operations, the rise of digital volunteerism. Digital volunteerism is defined as the use of information and communications technology (ICT) for aiding disaster response and relief by gathering and exchanging data to help decision-making, deploying tools and methods such as online platforms and mapping, crowdsourcing data, microblogging, or maintaining and updating relevant wikis and social media (Twigg and Mosel 2017). Following the extended use of online resources in the aftermath of Hurricane Sandy, many humanitarian organizations (HOs) have initiated a process of restructuring their existing social engagement strategies to better communicate and collaborate with this new type of virtual helpers. Conventional HO outreach plans are largely limited to developing relationships with highly influential online users, i.e. those with a large social media following. Social network analysis, on the other hand, shows that these communities resemble scale-free networks, consisting of a disproportionately small number of influential actors compared to a vast majority of sparsely connected individuals (Barabási and Bonabeau 2003). Thus, finding ways to engage a more diverse base of online users may be beneficial. Since the Internet has made possible for virtual volunteers to contribute to humanitarian efforts regardless of their geographical location, two additional assumptions have been made by HOs regarding the nature of digital volunteerism. First, it is assumed that the disembodied volunteer opportunities enabled by ICTs would turn digital volunteerism into a routine role assumed by volunteers proactively rather than reactively and on an ad-hoc basis. Second, it is expected that there be a large number of remote digital volunteers not directly impacted by a disaster but willing to contribute through online engagement efforts. To test the assumptions about the influence status, geographic distribution and spontaneous vs. non-spontaneous behavior of digital volunteers, we analyze textual and user-information Twitter data from 18 hurricanes that have made landfall on the territory of the United States since the launch of Twitter in 2006. Using a topic modeling technique to uncover latent topics in the textual corpus, we focus on the topics most closely related to humanitarian messaging. Next, we use these humanitarian topics to detect Twitter users fitting the description of digital volunteers. The results of the analysis show that contrary to prevailing opinion, 68 percent of all identified digital volunteers have a follower/followed ratio between 0 and 2, suggesting that they do not have an influencer status since the number of Twitter followers they have is approximately the same as the number of accounts they follow. We further discover that although there are occasional Twitter users engaged in humanitarian messaging located in regions far away from the disaster epicenter, the vast majority of individuals exhibiting digital volunteer characteristics (91 percent) in the context of hurricanes impacting the United States are domestic. We do find however, that the states accounting for the highest number of digital volunteers are not necessarily the ones experiencing the most hurricanes. Despite Florida, North Carolina and the unincorporated territory of Puerto Rico experiencing the highest frequency of hurricane activity, we detect the highest percent of digital volunteers in New York (15.18 percent), Texas (12.28 percent), Florida (10.11 percent), and California (9.97 percent), the four most populous states. It can thus be concluded that digital volunteerism is still relatively local as opposed to global, and driven more by state population size rather than disaster occurrence frequency. We also find that contrary to expectation, the majority of digital volunteers (88 percent) are active only during a single hurricane event as opposed to several, suggesting that digital volunteerism is still a spontaneous, reactive act of prosocial behavior rather than a routine role. Taken together, the three essays establish a framework for the application of network analysis to assess pertinent sustainability problems in a dynamic and statistically rigorous way. This research has practical implications both for researchers, through the novel literature survey framework proposed in Chapter 2, and practitioners—corporate managers debating the performance consequences of announcing a sustainability initiative, and HO social media strategists developing ways to connect with prospective digital volunteers.
|
| Subject
|
:
|
Information science
|
|
|
:
|
Management
|
|
|
:
|
Operations research
|
| |