So this is the last roundup where I try to collect and present *everything* new in the world of civic tech research. Plus it’s 3 ⅕ months worth, grabbing back through February. As a result there’s a lot here and I apologize for the info overload. Hopefully it’s fun to read.
You can also jump to
- Findings
- State of the art (lit & evidence reviews)
- Useful research
- Case studies
- Concepts & frameworks
- Community & curation
- In the methodological weeds
- Miscellanea & Absurdum
Findings
Citizen engagement:
More evidence (this time from 41 Romanian cities over 4 yrs) that “increased decisional transparency does not automatically lead to greater citizens’ involvement in the decision making process.” But tweeting about sports is a great way for local governments to boost civic engagement, according to a study of Spanish social media use. Analysis of an Austrian city planning platform shows that external rewards motivate citizens to evaluate other citizens’ ideas, but not to generate their own.
Transformation and Trust:
Why do small municipalities adopt e-gov? It depends. A survey of small municipalities in Nebraska (population <5,000, n=68) suggests that the drivers are different than larger municipalities, and can are different for digital transactional and informational services.The most important driver for information services is management support. The most important drivers for transaction services are administrative capacity, commitment to open access to government, and educational attainment of civil servants. In other news, you might think e-govt is bad for snail mail, but this article examines how “postal operators are playing an active role in the success of e-government policies.”
A Bahraini survey (n=313) finds that ICT-enabled transformation of government “positively influences citizen trust mediated by transparency and government performance,” but admits that it’s hard to generalize from Bahrain.
Accountability and responsiveness:
Research in Albania suggests that community scorecards don’t help efforts to contest government officials and improve their responsiveness, but only “bolster officials’ agenda.” Analysis of Ghana’s extractive sector transparency measures (survey, n=3,500) suggests that to result in government accountability, government transparency should target NGOs, not citizens.
Things go wrong:
Case study of the Estonian e-participation portal suggests that “due to complex context, e-participation systems are prone to fail”, and emphasizes the “complexity of context and flaws in system management”.
Public sphere:
Analysis of tweets during a 2014 U.S. gubernatorial election (n=20,580) suggests that that middle‐level gatekeepers (w b/w 1,800-26,000 followers) are most influential on political twitter.
Design matters in participatory budgeting:
Hollie Russon Gilman and Brian Wampler review differences in the design of participatory budgeting in Brazil and the US (PB in the US is city-based and emphasizes representation, while in Brazil is municipal and emphasizes representation). A separate Brazilian study (survey data from 94 municipalities) finds that municipalities tend to use digital in PB, but that the mechanism seems flexible enough to meet multiple needs online and off, and allows leaders to adapt their approach on the go. Relatedly, CitizenLab presents three PB initiatives in French and Belgium cities, to show how PB can be designed to meet different policy objectives, there’s mixed results from PB in London and Paris, and case study research in the Philippines finds that a centralized institutional design of PB obstructs real empowerment.
The duh desk
Social networks matter for how tech facilitates online collective action. Case studies and cross-sectional survey data from Japan and South Korea suggest that the size and heterogeneity of social networks are particularly determinate.
Telling citizens that govt is performing well boosts trust in govt, telling them the opposite has the opposite effect, according to an “online randomized survey experiment conducted in Argentina” by the World Bank.
Best title: The Unbearable Lightness of Digital Activism: the effect of gender on social activist participation level of university students.
State of the art (lit & evidence reviews)
Most research on social media and participatory decision-making is produced in the US. This according to this lit review and bibliometric analysis (n=1,159 b/w 2008-2019), which also suggests that health and disaster management are prominent policy areas, and that ethics are under-addressed.
There’s also a lit review of Citizen Science and Crowdsourcing in Applications of Pluvial Flooding, a lit review on the public value of e-government (53 articles, finds that “there is a need for research on the public value of e-government in Least Developed Countries.”), and a review of 775 research papers on open data visualization, suggests that the most common approach to visualizing open data is mapping, and the most common problem is bad data.
3ie has mapped out the evidence base for ICT4D interventions. The resulting evidence map is organized according to sectors, and has some interesting implications. Most notably, a whopping 146 of the 253 impact evaluations included in the mapping target mobile health. Notably, most studies also “estimate the effect of a pilot intervention instead of a program” and are RCTs. R&E gives a good high level overview and links to the mapping.
The state of SDG research. Researchers from a prominent publishing industry brain trust (how else to describe them?), have mapped out research on topics associated with the SDG goals. While the title of LSE Impact’s blog post is misleading (they didn’t map the effect of the SDGs on research), there are some fascinating tidbits. Most research is in Ag or Health, with some small pockets of collaboration. Latam and Africa don’t do much cross disciplinary collab. Most research on poverty and inequality comes from Nigeria. Who knew?
Lastly, this systematic analysis of the corpus of e-governance research literature using tools of network theory is too confusing for me to summarize.
Useful Research
NRGI’s launched a new open National Oil Company (NOC) Database
This smart city stakeholder classification model distinguishes between 4 stakeholder roles and 9 aspects of city governance as a foundation for designing engagement and design initiatives.
This article presents a systematic evaluation and comparison of map-based crowdsourcing platforms (Ushahidi, Maptionnaire, Survey 123, ODK, GIS Cloud).
Chinese open data researchers have developed “a new flow-based planning support system for examining neighborhood quality of life and health for the City of Atlanta as a prototype”
“This chapter aims to present a prototype integrated popular report designed to promote citizen participation in financial sustainability decisions”
“To fill this gap, we present a prototype of an automated bias assessment tool for geographic data. This new tool will allow city officials, concerned residents, and other stakeholders to quickly assess the bias and representativeness of their data. The tool allows users to upload a file with latitude and longitude coordinates and receive simple metrics of spatial and demographic bias across their city.”
Case studies
From TAI weekly: “If you are heading to the Open Government Partnership Summit in a couple of weeks, you might want to bone up on nearly 100 innovative open government case studies from over 30 countries.” Wow. Otherwise:
Global leaders?
- Open Data in Estonia (poster child isn’t easy)
- Social media in the US justice system (not that interactive)
- Algorithmic risk assessment in US criminal justice (yikes)
- Open data, closed government in Singapore (yup)
- Collaborative e-gov in a Finnish municipality (play-by-play)
- Arabic activism on Twitter (enables lots of things previously prohibited, perils persist)
- Participatory service development in Belgium (there’s a gap between literature and practice)
- Participatory Budgeting in London and Paris (mixed results)
- Transparency in Denmark’s open aid portal (meh, not so much)
- Infrastructure feedback in Japan (digital is better)
- Barriers to OD engagement in Hong Kong (many, policy most imp)
- Impact of open data in Cape Town (only for a few)
- Internet Democracy and Social Change in Israel (it’s complicated)
Developing and authoritarian country contexts
- Public accountability in Thailand (automatic disclosures are weak)
- Government accountability in Thailand (Auditor-General is most imp stakeholder, !the public is perceived as the second”)
- Participatory Budgeting in the Philippines (centralization blocked real impact)
- Sub-national OGP in Indonesia (Bojonegoro) (everything is awesome!)
- Open Data in Kazakhstan (poster child, maybe)
- Budget monitoring in Nigeria (citizen engagement is good)
- Gender-Responsive Budgeting in Indonesia (still centralized)
- Open Government Data in Tanzania (there’s obstacles and inertia)
- E-gov and trust in Saudi Arabia (it’s super complicated)
- National Open Government Data (OGD) Portal of Saudi Arabia (data’s bad)
- E-gov in Kazakhstan (it’s a paradox)
Activism
- #ShoutYourAbortion Twitter Campaign (fighting stigma reinforces stigma)
- The #YesAllWomen student movement on Twitter (classroom convos matter)
- Youth Media Activism on YouTube (it’s complicated)
- Drones, activism and mountaintop removal mining (check your stereotypes)
- Activist leadership in 2011 Egyptian revolution (decentred, emergent and collectively performed)
Global
- Tech in the refugee system (lot’s of promise, less accountability)
Concepts and Frameworks
Governments and participation:
Dutch researchers suggest a “ladder of government participation” to assess how government roles change during the course of collaborative and participatory initiatives, also, here’s a messy e-government maturity model.
DEWEM is a conceptual framework for the design and evaluation of a government websites for democratic e-governance.
This “five-dimensional scale of public servants’ public values preferences” is based on Chinese survey data and validated with two surveys and explanatory factor analysis.
Activism and power:
This “integrative model of activism that explains why and how individuals in the networked society are engaged in contentious issues” is validated on individual engagement with gun ownership, immigration, and police use of power, via an MTurk survey (n=800),
Kersti Ruth Wissenbach suggests that the “de-westernisation discourse of communication scholarship” reveal power asymmetries within global civic tech activist communities.
Jonathan Gray sketches the concept of data witnessing “ to characterise how data is involved in attending to situations of injustice”, through an analysis of Amnesty International’s Decoders initiative.
Data is the new everything:
Alex Ingrams explores “the democratic governance impacts of big data in three policy areas using Robert Dahl’s dimensions of control and autonomy.”
Using data walks as a method for engaging non-tech-savvy citizens in co-design work.
Meta:
This article in the Journal of Applied Philosophy suggests five reasons why researchers should not avoid political activism (but still be mindful of their own bias).
Community & Curation
Responsible data in development: Join @USAID_Digital, @mSTAR_Project, @DIAL_community, @GlobalDevLab, @ICT_Works and the responsible data community are organizing a #5DaysofData Twitter Chat “to learn how you can use data responsibly and ensure that data risk is regularly considered and addressed.”
Luminate, the new Omidyar spin-off, reflects on the last 10 yrs of civic tech in the US, and released slides (partially redacted) from last year, which present contracted research from Dahlberg on the same. The core message is that civic tech has moved through phases of early exuberance and demonstration/dissemination, and is now entering a period of civic tech realism, which will have to contend with challenges related to funding shortfalls, impediments posed by formal procurement processes, and persistently low levels of literacy and capacity. There are also lessons, which will not surprise anyone in the field (civic tech is a means not an end, is embedded in the political, and could be more diverse).
Relatedly, there’s a fascinating google doc collaborative effort to develop a process-oriented definition of civic tech (defined by ppl, tech & impact) on this google doc. I’m not sure who’s behind it or how I found it. That’s the trouble with all these tabs and links you see.
Mapping AI governance. Whoa: Nesta is building a “comprehensive online resource about global governance activities directed at artificial intelligence and machine learning”
Mary Joyce takes issue with Matt Stempeck’s “10 Problems With Impact Measurement in Civic Tech”. Each of these 10 deserves a rigorous debate.
Goodbye to an almost amazing data source. After Seven Years of Helping People Govern Better Together, Article1 is shutting don Madison (Article 1 on Medium)
Noel Isama from Sunlight makes the case for user research for cities’ open data.
Book review: The Rise of Nerd Politics: Digital Activism and Political Change
Global Integrity continues their laudable efforts to ensure that the data they produce is useful and used in promoting better governance. They’re looking for advocates in civil society to give them feedback on the utility and usefulness of the Africa Integrity Indicators.
Information Communication and Society has a special issue on Data Justice, based on the Cardiff conference from last year, and including articles on counter-mapping and data witnessing (in Amnesty Int).
In the Methodological Weeds
Nathan Begbie digs into algorithmic bias in the South African context, illustrating the three types of bias proposed by by Danks and London (2017), outlining 3 strategies to mitigate that bias.
CDG has a wonderful post exploring the paradoxical tyranny of regressions in development research: you need them to get published, but the most important development problems don’t have the data to support meaningful regressions.
R4D is mapping and designing “fiscal governance indicators focused on measuring performance of organizational activities and programs.” TAI has a high level write up of the deep weeds.
Bloomberg Cities offers 5 secrets of successful community surveys
Miscellanea and Absurdum
- Researchers analyzed 25 years of papers (n=16,625) to understand broad trends in AI research. Two takeaways: machine learning is just a blip, and “the key thing to realize is that nobody knows how to solve this problem.”
There’s a new Map of Best Breweries in America, 2018 (Flowing Data), and in best city practice, Code for Anchorage has updated its bike to breweries map.
