Mapping the Transition of Work in Labor Markets and Entrepreneurial Organizations
Dewitt, Theodore
2019
Abstract
Jobs and technology have a complicated history. Advances in technology can enable people to do existing jobs more efficiently, or technology can completely replace the human element of a job. While fear of the impact of job automation is not new, the scope of this concern has expanded due to novel technologies such as robotics, complex neural network architectures, and artificial intelligence. Though wholesale replacement of all human labor may be distant, technology is reconfiguring the jobs landscape, as it has in times past. Understanding how requires a more complete understanding of the structure of jobs and how that structure is impacted by technology than currently exists in the organizations literature, a subject this dissertation explores. In this work, I advance a theory of jobs and technology that relies on conceptual analogy. This analogy allows me to use what we know about the structure of technology along its combination, recursive, and performative dimensions to infer a similar structure in jobs. I propose that just as a technology is a stack of technological components, a job is a stack of task components. Through this analogy I also propose that jobs and technology can coevolve. To examine the theoretical proposition of a job as a stack of tasks, I take task descriptions of jobs from the U.S. Bureau of Labor Statistics and decompose them into underlying task variables using natural language processing methods. This analysis allows me to represent the underlying theoretical concept of jobs as a stack of tasks that can be represented quantitatively, with jobs consisting of a group of latent task variables. Using paired T-tests and exploratory visualizations of these variables, I demonstrate that the underlying task structure of jobs varies over time, that jobs are becoming more diverse in terms of their tasks, and that jobs cluster together in sometimes unexpected groups. Next, to examine the theoretical position that jobs and technology may coevolve, and to understand how technology intersects the creation of new jobs, I performed a case study of a commercial cleaning services company whose workers use computer tablet-enabled, guided workflow software. I supplemented this case with interview data from recent entrepreneurs. Using thematic analysis, I uncovered three key findings across the data set. The first is that in contrast to existing literature on the pursuit of entrepreneurial opportunities, all entrepreneurs seem to experience both risk and uncertainty in the forms of market uncertainty, market risk, execution uncertainty, and execution risk. The second is that these aspects of risk and uncertainty make it very challenging to create fixed and formalized jobs until a company’s product and strategy have crystallized. The third is that the technology stack that jobholders use, and the job task stack that jobholders perform, coevolve through a process of on-the-job experimentation. This work makes three contributions to the literature. The first two are methodological: I demonstrate how natural language processing can help researchers engage in quantitative analysis to address questions past research has only addressed through qualitative methods. This research also provides a new method of examining the underlying task structure of jobs. Finally, this work shows that as new organizations create jobs, that initial task structure is provisional; jobs go through iterative processes of experimentation before they can be formalized.Subjects
job design entrepreneurship case study natural language processing job task structure technology
Types
Thesis
Metadata
Show full item recordCollections
Remediation of Harmful Language
The University of Michigan Library aims to describe library materials in a way that respects the people and communities who create, use, and are represented in our collections. Report harmful or offensive language in catalog records, finding aids, or elsewhere in our collections anonymously through our metadata feedback form. More information at Remediation of Harmful Language.
Accessibility
If you are unable to use this file in its current format, please select the Contact Us link and we can modify it to make it more accessible to you.