Publication Date

8-2023

Date of Final Oral Examination (Defense)

4-7-2023

Type of Culminating Activity

Dissertation

Degree Title

Doctor of Philosophy in Public Policy and Administration

Department

Public Administration

Supervisory Committee Chair

Jen Schneider, Ph.D.

Supervisory Committee Member

Stephen Crowley, Ph.D.

Supervisory Committee Member

Eric Lindquist, Ph.D.

Supervisory Committee Member

Michael Ekstrand, Ph.D.

Abstract

In policy sciences, data have traditionally been a tool used by scientists and technocrats to guide state policy. Boundaries around what counts as data generally fall along traditional understandings that data are neutral, objective, and abstracted from individual bodies and experiences. Unfortunately, this understanding of data has a history of perpetuating harmful social hierarchies and, especially in the era of “big data”, mirroring our racial and gendered prejudices (Kitchin, 2014). More recently, however, data have been claimed as a tool by a different kind of actor operating in a unique environment. These new actors, such as some police officers and citizen activists, are negotiating and redefining who is considered a data expert and what we understand data to be.

These conversations between traditional and novel understandings of data can be seen within the data for good movement, where actors from a broad range of backgrounds and training come together for the purpose of advancing some notion of social good. Given the history of data perpetuating social harms such as racial discrimination, how can these relatively new understandings of data promote the social good while avoiding data harms? Or, how can data be used to promote the social good? Using the theoretical framework of Data Feminism, the data from participant interviews suggests that shifting understandings of data rely on the emergence of the concept of embodiment. This research highlights the differences in how embodiment manifests in two dissimilar sites: Measure Austin, a non-profit advocacy organization for people of color, and the Big Data Hubs program within the National Science Foundation. The findings suggest that data for social good presents as a space where data advocates negotiate between embodied and disembodied meanings of data and where embodiment is more significant for street level bureaucracy and citizen activists.

The dissertation suggests that “embodied data” offers an alternative to the predominance of market-driven data approaches. This research ends with a discussion for how policy studies could benefit from incorporating the concept of embodiment in research related to data systems, including artificial intelligence and machine learning.

DOI

https://doi.org/10.18122/td.2103.boisestate

Included in

Public Policy Commons

Share

COinS