Contradicting Motivations in Civic Tech Software Development: Analysis of a Grassroots Project

Grassroots civic tech, where people create and use software to create positive change in their community, is based on volunteer and free/libre open source software participants. Earlier survey-based studies indicate that participant motivations are generally in alignment due to participant self-selection. In our case study, we examine one civic tech grassroots software group and their two projects to see how individual motives evolve and contradict as the project progresses. We found that even though participants agree on the big picture, there are significant contradictions in details. Two most significant contradictions involved planning, and converging on design and technical approaches. To address these contradictions and to sustain similar projects, we recommend involving skilled and committed facilitators that can keep volunteers pointed in the same direction and sustain participant engagement.

Grassroots civic tech, or software for social change, is an emerging practice where people create and then use software to create positive change in their community, such as using citizen sensing to record evidence of air pollution to support environmental activism [1]. It is related to Free/Libre and Open Source Software (FLOSS) movement and participants are diverse, calling themselves for example civic innovation, tech for good, civic crowdsourcing, free software, and community technology [2]. Like their names for the field, the participants’ and groups’ motivations and working practices are diverse.

Civic tech itself is a wider field [3, 4], with studies for example in: examining the role of information technologies in governance [5]; supporting participatory democracy practices of technology in society [6]; understanding the role of software in society [7]; developing agile methods and techniques to support software development in society [8, 9] and, more recently architecting smart cities, open data with and for people [10, 11]. However, civic tech software development processes from a systems perspective have been less often studied, especially when it comes to grounds-up development processes. Similarly, there have been studies on what drives individuals to participate and keep contributing in FLOSS [12, 13] or tech for good [14], but less analysis on how volunteer motivations interact once the development process is ongoing. We propose that this is an open question that is worth addressing, since development processes in action [15, 16] evolve based on team needs. A study on motivations of a project in action can reveal how different volunteer motivations can be negotiated towards a compromise.

In this paper, we apply Engeström’s expanded activity theory [17] as a theoretical lens to analyze motivations, how they relate to for example group goals or development tool supported processes, and what contradictions emerge when a civic tech group’s development process is in progress. The research case is based on the activities of Code from Ireland civic tech group from 2018 to 2019, with the group structure described in more detail in [18]. During the case study period, Code for Ireland was engaged in two projects: Transparent Water for publishing data on water quality and Finding Vacant Homes for identifying residences that could alleviate the Irish housing crisis.

More specifically, our research question is: What motivation, motive, or goal contradictions emerge in a grassroots civic tech project in action?

In the following Section 2, we detail related research on grassroots civic tech and motivations. We continue then in Section 3, about case study and analysis. Findings are presented in Section 4 and conclusions in Section 5.

Individual motivations for contributors have been studied in free/libre and open source software projects recently by Gerosa et al. [13]. They found that key motivational factors were social aspects (helping others, teamwork, reputation) and intrinsic motivations, such as fun, learning and altruism. Similarly, Barcomb et al. [12] found that social norms, satisfaction, and community commitment were supporting factors in episodic volunteers. Coelho et al. [19] who studied motivations of core developers mainly had similar results in friendly community and opportunity to engage in volunteer work. By contrast, core developers were also motivated by technical quality, such as improving projects they use, lack of project complexity, and lack of complex or buggy code.

Huang et al. [14] studied specifically FLOSS for social good project participant motivations and found that they are more motivated by project impact, such as the societal issue being addressed, and who are the project owners. Community and social characteristics were also important.

Pern and Kitchin [20] analyze frictions from a broader social and cultural systems perspective in civic tech development, using an earlier Irish case study as a data source. They found that complex urban issues can have only partial or imperfect solutions, that can in turn lead to or reveal further complications. The main frictions Pern and Kitchin discuss were stakeholder engagement, shifting goals, and varying commitment from participating developers.

The case we studied is focused on a grassroots group creating civic tech, Code for Ireland, which describes its mission as “developing innovative and sustainable solutions to real-world problems faced by communities across Ireland, by fostering collaboration with civic-minded individuals, businesses and public sector organizations.” The research and data collection occurred during 2018 and 2019. The process is summarized in this Section and is reported in further detail in an earlier study on the group [18].

Case study data includes interviews, development activity logs from GitHub, and an ethnographic study with a researcher embedded with the group for six months in the first half of 2018. Of these, the main data source was semi-structured interviews with six key group contributors. The interviewees included the community leader, a project manager, an open data advocate, and several programmers. The interviews were recorded, transcribed, and analyzed with qualitative data analysis software. The interview guide is archived in the Zenodo repository ¹¹1https://doi.org/10.5281/zenodo.2565434. Other collected data was used to contextualize the analysis findings.

Other related research on the group has been performed by Perng and Kitchin [20], who discussed Code for Ireland from a civic and cultural perspective and defined Code for Ireland as a civic hacking group, where “civic hacking binds together elements of civic innovation and computer hacking, with citizens quickly and collaboratively developing technological solutions” [20, pp. 2].

In this Section, we present our qualitative data analysis outcomes step by step as we progress towards increasing level of abstraction, following lightweight qualitative data analysis practises by Saldana [23] and Miles et al. [24].

In the first step of analysis, open coding, we analyze the interviews and note any motivations and motives. In the second round of analysis, we extract activity descriptions from interviews and online materials, and relate the motives to activities using our theoretical lens. In the final step of analysis, where we generate a situated explanation of the case, we present levels of contradictions inside and between the activities.

In this paper, we analyzed what motivation, motive, or goal contradictions emerge in a grassroots civic tech project in action. The main contradictions occurred inside activity systems on details of implementation or between system motives, instead of big picture motivations. Two most significant contradictions involved planning, and converging on design and technical approaches.

Our recommendations for practice are what perhaps professionals in established volunteering and non-government organisations already know: Have skilled and committed facilitators that can keep volunteers pointed in the same direction and sustain participant engagement. Based on our case study, it appears that it is essential to have: efficient communicators, shared task management tools, and a compromise on technologies that most participants can commit to. Finding an acceptable lowest common technology denominator is important, since participant skill levels in civic tech can vary [35, 36].

We extend the body of knowledge in civic tech software engineering through two main contributions. First, we demonstrate activity theory can be used as a theoretical lens to evaluate software engineering methods in action [15, 16] as evolving, in-action activity systems. Second, we demonstrate that examining participant motives together as an activity system can shed light on contradictions that arise during civic tech or tech for good development.

The trade-off between contextual richness of a single case study and generalizability of a multiple case study points towards a potential limitation in our work. However, seminal literature [37, 38] acknowledge single case study as a method of inquiry, especially when undertaking novel, exploratory research to capture the complexity of the context in which the case is situated (i.e. motives and motivations within activity systems in complex civic tech environments and processes). Looking forward, we advocate for more comparisons between different civic tech software engineering activity systems for broader findings and more generalizability.

We thank Dr. Victoria Palacin for her conceptual contributions and research contributions to the authors’ earlier research collaboration. This research is based on the line of thinking that started in collaboration with Dr. Palacin.