Article Preview
TopIntroduction
In recent years, the COVID-19 pandemic has forced organizations and institutions to develop newer work practices heavily involving the use of information systems and the internet. While organizations were constantly changing to adapt to new processes, technology, and even public-health regulations, many of these changes are complex, slow, and mostly ineffective (Howe et al., 2021; Jacobs et al., 2013). Prior to the COVID-19 pandemic, organizations engaged in traditional business practices while skirting around issues/challenges that ultimately created a negative workforce experience. These negative workforce experiences are typically the result of inefficient technologies, archaic ways of working, and poor organizational culture (Bordeaux & Lewis, 2021). As the pandemic progressed, many organizations began to implode from the weight and rate of mass attrition—both forced and voluntary. De Smet et al. (2021) referred to this phenomenon as “the Great Resignation/Reflection,” where workers were leaving their jobs at unprecedented rates and moving to new ones. Similar to a “renter’s market” in the housing industry, the United States (U.S.) was suddenly faced with an “employee’s market,” where employers were now at the mercy of employees who were reflecting on their senses of purpose and meaning at the workplace and opted to look outside their current employment situations to fulfill them (Dhingra et al., 2021).
Deloitte Consulting, a global management consulting firm, described the workforce experience as “the sum of a human’s lived experience at work and how they feel about their organization” (Bordeux & Lewis, 2021). For organizations to retain workers and safeguard human life during the pandemic, many employees received flexible work arrangements. Business meetings shifted to using teleconferencing applications such as Zoom and Microsoft Teams, and employees were able to work from remote locations without needing to physically be at the office. Spreitzer et al. (2017) classified these flexible work arrangements into three dimensions: (1) on-site workers with flexibility in their scheduling, (2) remote workers with a fixed schedule, and (3) flexibility in employment relationships. These changes to the organizational workforce have had lingering effects even after the critical stages of the pandemic. The COVID-19 pandemic was most threatening before the development of the vaccines and boosters, but after the proliferation of vaccines and access to them, organizations—and society in general—began to engage in several pre-pandemic practices, such as reducing the limits of social distancing and mask mandates. The flexible work arrangements offered during the critical stages of the pandemic created a paradigm shift in the lives of many employees, leading them to question the effectiveness, efficiency, and necessity of fixed schedules and permanent work locations.
The flexible working arrangement is a multi-faceted phenomenon with several considerations. As indicated by Spreitzer et al. (2017), one of these considerations pertains to the location at which employees work, which shall become the focus of this study. Specifically, this paper focuses on workers’ preferences of where they would like to work. We apply a data-driven approach to this study by using social-media data (specifically Twitter1 data) to better understand worker preferences as influenced by the COVID-19 pandemic. Our research is focused on the U.S. population—rather than a global audience—given the possibility of socio-economic and cultural differences that may influence worker preferences (which is outside the scope of this study). Our study is divided in two phases: Phase I pertains to understanding workers’ preferences and perceptions about the work locations, and Phase II aims to develop a suitable predictive model that could be used to determine what category a worker may fall into regarding work preferences. The study is outlined as follows: the next section discusses the research methodology of this study. This is then followed by data analysis and results for Phase I. Consequently, Phase II predictive models and results are then reported. Finally, the paper provides a discussion and conclusion of the study’s key findings.