Exploring the Fabrics of Civic Tech on Digital Media

Team Members

Kersti Wissenbach (project host), Nicole Pauli, Ingrid Woudwijk, Ángeles Briones, Marcelo Santos, Mette Storup, Cæcilie Laursen, Alejandro Martin del Campo

Contents

Introduction

Civic Technology has become a popular term over the past years. Whilst there is no clear definition of the term, the wider civic tech scene spans from business-oriented tech start-ups towards (digital) social and political activist groups. With specific attention to the utilization of technologies in the realms of opening of civil society space, increase transparency and hold governments to account, we witness the growth of new civil society groups and organisations alongside with increasing attention from traditional development cooperation (NGOs), and the startup scene.

Within the transparency and accountability field, civic technology gained attention among various actor groups. At its core stands the potential of new information and communication technologies and data at civil society’s disposal. However, attention is often disproportionately directed towards creation of tools and technologies at the expense of the development of other capacities needed to put them to work in the service of social and democratic goals.

First investigations into the structure of civic tech coders on Github exist (Baack 2015) but no research has been done so far aiming to understand in all bandwidth how civic tech as a new phenomenon is constructed beyond the technology and on a global scale. In order to shed light on one aspect, namely the ‘digital fabrics’ of civic tech, this summer school project will map the civic tech scene and its relations as apparent on the web and social media.

We will apply various digital methods in order to explore the fabrics of civic tech in the digital.

Research Questions

This explorative process set out to investigate the fabrics of civic technology on digital media, specifically twitter and the web. It has been driven by the question if a disjunction between promises and practises of “civic tech” exists. In order to inform this overall research question four main areas of focus have been investigated, namely:

1. What are the definitions and understandings of civic technology according to ‘the web’?

2. Who are the actors and what are their connections?

3. What is the geography of civic technology?

4. What is the temporality of civic technology?

Methodology

The four areas of focus informing the overall research interest of this process have been operationalized into a set of questions which allowed the team to consult the web as well as twitter as the platform most heavily used for professional purposes of outreach and engagement among actors involved in the wider civic tech environment. The decision to take distance from Github for this explorative exercise is grounded on the intention to avoid a tech-centristic lens by foregrounding the developer side. This exercise intended to cover the overall wider actor network and discourse. However, we recognize the specific digital discours bias to this exploration, particularly considering project implementation and funding structures on a more global scale. Further recognition is required towards the fact that a multitude of projects, ideologically and practically following same intentions but being ‘labelled’ differently, according to its overall intentions (e.g. transparency and accountability, Tech4TA, etc.) or its ‘predecessor’ (e.g. open government, open government data, etc.).

Google search engine querying in combination with the analysis of twitter account bios and respective user tweets as well as general twitter capture have been used.

The following datasets have served as point of departures for this work:

1. List of all twitter accounts entailing the terms ‘civictech’, ‘civic tech’, or ‘civic technology’ including bios and locations of all results > 362 accounts

2. Capture of all tweets from twitter account capture (1) > 544902 tweets

3. Tweet capture of ‘civictech’, ‘civic tech’, ‘civic technology’, ‘civictechnology’ (time frame 23.05.2016 - 05.07.2016)

4. Follower & followee network lists of twitter account capture (1) entailing ‘civictech’, ‘civic tech’, ‘civic technology’ (account names and bio descriptions, number of tweets per user)

5. List of Google results of ‘civic tech’ / ‘civic technology’ for complete .com and .br query (incl. all google results headlines, URLs, shown abstracts)

Departing from these preliminary collected datasets the following additional data has been collected during the project days:

6. Categorization of twitter account user dataset and respective tweets

7. List of Google search results of the search term “civic tech is” for the domain “.com” (incl. all google results headlines, URLs, shown abstracts) + categorization

8. List of Google search results of the search term “civic tech” / “civic technology” for the domains google.com.br (incl. all google results headlines, URLs, shown abstracts) + categorization

9. List of google search results for those countries in the third phase of the Open Government Partnership Action plan cycle (nine countries in total)

10. Filtered versions of tweet capture (only those tweets entailing ‘civictech’ or ‘civictechnology’)

• Categorized tweet account dataset

• extracting all tweets (from dataset 2) entailing the hashtags ‘civictech’ or ‘civictechnology’.

• Extracting ‘top hashtags’

• Extracting links from tweet set

• Extracting mentions from ...

• Co-hashtag and user hashtag networks (monthly and overall)

Findings

In the following chapter we will describe the main findings relevant to each of our main research pillars.

1. Definitions and understandings

Civic Tech is…

Qualitatively analysing google query results of ‘civic tech is’ provided us with a first overview of tendencies as in how civic tech is interpreted in various contexts. Three main tendency clusters could be identified, namely civil-society driven and technology driven and one on the intersection of technology and people. A smaller cluster relating to investment / business has been identified.

‘Civictech’ / ’civictechnology’ tweet capture > most contemporary ‘situation’

All tweets containing “civic tech“ or “civic technology“ for the time frame June 23, 2016 till July 05, 2016 have been analysed as a one week sample case to demonstrate the most contemporary state of civic tech discourse on twitter. The top one thousand hashtags have been extracted for a Gephi (Force Atlas 2) network analysis. For better visualisation the “civic tech“ and “civic technology“ tweets have been removed from the map since as central note they ‘pulled together’ the network, reducing clear network visibility.

Leaving nodes related to contemporary happenings aside (e.g. the big Brexit node), Open data, open government, and gov tech form the biggest, thus most linked nodes in the network with the strongest affiliation to civic tech. The wider network shows how the field assembles a number of main focus areas. The next biggest nodes assemble one community with aforementioned central nodes are tech, big data, startups, ICT4D, and civic engagement. Considering the limited granularity the data as provided in the graph could suggest the strongest civic tech affiliations with the older open data and open government movements. Furthermore, the wider clustering of interests around civic tech could suggest the dominating interests pushing into the scene. Thus, the startup / entrepreneur scene on the one side and international development cooperation, specifically the very popular ICT4D (information and communication technologies for development) show similar closeness to the civic tech field.

Hashtag frequency

Now we queried all tweets entailing the hashtags ‘civictech’, ‘civic tech’, and ‘civic technology’ from those users identifying via the term in their twitter bios. This exercise resulted in 11.878 tweets from 224 distinct users. Despite certain spikes mainly related to civic tech events (e.g. TICTeC, civictechforum) we find the same top affiliations as we can see in the network graph, with opengov, opendata, govtec high in the ranking and a bit further down also techforgood, ict4d.

Comparison of Co-hashtag analysis between unfiltered dataset and filtered dataset (excl. civictech)

For the co-hashtag analysis we used all the tweets of one year (July 5, 2015 - July 5, 2016) of the 271 twitter accounts entailing “civic tech“ or “civic technology“ in their bios. We worked with two comparable data sets a) the full data set of all user tweets, and b) only those tweets entailing the hashtag #civictech (a dataset of 10.351 from 209 distinct users).

For the very large unfiltered dataset we filtered by degree 50, meaning that all nodes with less than fifty links have been filtered out. The network shows similar affiliation as did the contemporary tweet graph. Open data and open government here being the closest nodes. Open data also being one of the biggest nodes alongside with tech and innovation. Interestingly enough government as such is to be found in relative distance to open government and civic tech but in close proximity to the tech and innovation environment which are also two of the biggest nodes. Here civic technology builds a community with the government and innovation clusters, meaning that those findings suggests a stronger community between the government, civic tech, and innovation scene than between civic tech and the open government, open data movement.

The filtered dataset in comparison provides information in regards to the strongest affiliations within the tweets also entailing the hashtags #civictech or #civictechnology. This filter has been applied under the assumption that the complete tweet capture can be very broad. Analysing only those tweets who entailed one of the above mentioned hashtags would zoom in on the actual discourse around civic tech. The results show an even stronger domination of open data and open government, including the government data cluster as already found in the most contemporary civic tech tweets capture. In significant proximity here we also find the reappearance of the ICT4D node.

2. Actors (and issues)

Actor dominance per country

> Google search results issue discovery and triangulation [.com & phase 3 OGP partners > 9 countries > triangulation > issue crawler results common to 8/9 countries]

In an attempt to create an understanding of the actor range around civic tech we applied google web query analysis in addition to the twitter analysis.

The top one hundred results of the google search query .com and .br have been selected and categorized [public, private, civil society, NGO, media, research]. The .com results have been explored with the DMI Issue Discovery Tool after which all single category results have been triangulated. The results provide an overview of which of the ‘issues’ discovered were unique to one category and which have been utilized by multiple categories.

Further, Google search query lists have been created based on the top 100 ‘civic tech’ Google results for those countries in the 3rd and thus last phase of implementing the Open Government Partnership action plan (http://www.opengovpartnership.org/). By using the Google Scraper we extracted the top 100 results from the different domains (when searching for civic tech). The URL lists were compared using the Triangulation tool. The overview below shows the appearance of sites in different domains. Red indicating that the site appears in all nine domains and blue indicating appearance in only one domain.

An additional comparison has been made using the host URLs extracted through the Harvester Tool. This provides a picture of more sites appearing in all nine domains, whilst very few sites occur in only one domain. The scheme represents that the strongest related actors are at the same time to be found in the top of most google queries per domain.

Assuming that top ranking search results will be those most consulted and thus potentially setting agendas, or the organisations behind those who will be leading the discourse such findings are significant. Specifically concerning is such concentration when it occurs across country domains.

Hashtags analysis

Mention frequency

We continued to explore the main actors in the civic tech environment according to twitter.

In line with earlier mentioned hashtag frequency, exploring the top hashtags, we then ran the same exercise running a mention frequency. Doing so we aimed to identify the top mentioned actors (@username).

The behind the scene story

Additionally, a case has been extracted deriving from one of the most central and recurring actors, namely Code for America. A brief deep-dive into the relational network based on the top mentions provides us a very first indication for cross-categorizational relations. Such exercise show the relations, often related to funding as well as associate or management related strong links between a key set of actors across the private, philanthropic and civil society sectors.

Hashtag co-occurrence

We then analysed the top co-occurring hashtags per categories as have been manually assigned to all twitter accounts referring to civic tech or civic technology in their user bios. The purpose here was to understand which kinds of actors are more prominent around each issue, represented in Twitter by the hashtag as a proxy. This procedure was performed for the 5 top co-occurring hashtags in the dataset and a few others that have shown unexpected results. The graphs show actor nodes coloured based on their category (see color code below) and we selected two of them to illustrate the analysis.

Whilst in the first case (#opendata) an overall diversity of actors can be identified, with a slight dominance of the NGO sector, the second network reveals an unexpected dominance of the private sector (green nodes) as well as an unexpected absence of government-related users (pink nodes). That reveals somewhat surprising, since the hashtag “gov20” refers to government 2.0 and could be interpreted as messages in the realm of a participatory, ICT mediated public administration. One possible explanation in this case is the interest of private actors in providing services or influencing the industry around the idea of government 2.0. Nevertheless, it would be interesting to dig deeper into the hashtagged messages that show this kind of discrepancy between expectation and results to better explain them.

Hashtag x type of user

When working with such great amounts of data, it is tricky to decide in which part of the dataset we should spend time and focus on. That happens with hashtags as proxies for issues as well: some hashtags may appear very important in the dataset from a purely quantitative descriptive analysis point of view, just because of its frequency. But, that doesn’t mean that such hashtag is a transversal subject in the community, for it could be an issue advocated firmly and frequently by one or a small group of very active users. As the graph below shows with #iot (Internet of Things), which is very attached in our dataset, to user @republiciot, Republic of Things, a private group that advocates and executes projects related to IoT (see www.republicofthings.com). The thick edge represents the weight of the connection between user and hashtag, in other words, the amount of times the user has employed the hashtag. We can see that other users are much less central and much less strongly connected to that hashtag.

Additionally, we took the sum of this hashtag co-occurrence frequency analysis and merged the nodes according to type of actor (public, private, NGO etc.) and filtered their position by the Network Analysis property degree in order to map which kinds of actors are more central or more peripheral to the overall discussion around civic tech in Twitter, based on our data.

The result, shown below, demonstrates that, interestingly, private actors and civil society seem to be more connected to the group of issues represented by our dataset collection of hashtags related to civic tech, while NGOs seem a bit off the center. Even though this is very preliminary, it could point to a more niche-specific strategy of NGOs, but this should be further revised in order to make final statements.

Top actor comparison

Finally, we also matched the top followed actors against the top mentioned actors based on our user’s tweetset for the time period 01.01.2007 till 04.07.2016.

Identifying those users mentioned in most tweets allows for an indication of the main influencers within the civic tech environment. Doing the same with the most followed actors and matching both lists against each aimed to strengthen a potential argumentation towards most influencing actors. The graph shows the most followed accounts on the left side and the most mentioned on the right side. We can see that whilst we have some overall dominance, such as e.g. of Code for America, other actors have a lot of followers, yet are not in the top ranks when it comes to their recognition through mentions.

3. Geography

Civic tech twitter accounts per category

We continued to attempt some conclusions regarding the geography of civic technology. Given the limited time available for this project and the limited location information at our disposal in those regards our findings are to be seen as high level and indicative.

We used our first data sets of all twitter accounts including ‘civic tech’, ‘civictech’, or ‘civic technology’ in their bios and identified their localities (city and country) by the location provided and via additional research where required. We then mapped the users with CartoDB in total as well as per category. No strong significances can be identified between the different actor categories. Across the sectors a strong centrality in the US and Europe can be found. Unsurprisingly, we can find stronger geographic diversity among the NGO actors, however with remaining dominance in the ‘Global North’. The strongest US and EU centrism is to be found in research and the public sector!

Also here, suspicious based on our findings, we adventured into some additional exploration.

We conducted a top level domain analysis for six countries. However, the results are found to be inconclusive due to different patterns of domain adoption per country.

Also, if we look, at the amount of results per country in the table below, we can conclude that civic tech is an expression centred in the English language area, and there could even be no equivalent in other languages such as Portuguese, French or Spanish. The mere translation of the words wasn’t enough to translate the concepts as used in each of those countries: the search in Google.es for “tecnología cívica” in websites in the Spanish domain (“site:.es”) resulted in just 28 results. The same pattern repeated for a few other countries such as France, Argentina and Brazil, leading us to abandon this strategy.

4. Temporality

Only limited findings could be generated regarding the temporality of civic tech. This is mainly due to the only very recently started tweet capture dataset we had at our disposal. Findings based on this ongoing capture will be provided in the future.

Consequently, for this part of our exploration we exclusively worked with the co-hashtag analysis for the time frame of June 2015 - June 2016, extracting the top one thousand hashtags of the extracted co-hashtag network per month and visualised it in a bump chart using RAW. The bump chart shows how the most dominant hashtags have been moving throughout the year.

We see a strong frequency and consistency of a few dominating hashtags (#iot, #opendata, #civicdata) and following more general ones (#tech, #data, #technology). Other dominant hashtags tend to relate to recent events or trends (see #cop21 or the annual and much advocated #opendataday).

Discussion

The findings we could extract in the limited time of this project should be considered as very first indications towards the attempt to define the civic technology field.

We set out wondering if this new field is really new or if it is rather a new label? What we were mostly interested in was to explore if there is a disjunction between the promises and practises of civic tech. First of all it has to be said that such mapping of the digital has clear built-in limitations to provide an exhaustive answer to such question.

To be considered here is not only the limited exhaustiveness of data we had the chance to work with. To be strongly considered are two limitations, that is

1. The aspect of different language use as well as different terminology for the civic tech phenomena across countries, language areas, and domains. For the least it would be necessary to identify the most used terminology per language area and to explore across all different country domains.

One recommendation for further analysis would be to find the “orbit” of related issues around the key concept civic tech and work with the group of keywords and hashtags derived from it. Some countries apparently don’t employ an “umbrella” concept like civic tech, but that definitely doesn’t mean the concept is not understood or is not existent. This methodological analysis also rises the problematic issue of the lack of worldwide standards for those domains. These standards could possibly be part of a basic URL/Internet grammar, but these variations go against a globalized comprehension of the URLs and domain system for the Internet.

2. The limitations of interpreting such top-level results from the web. Such a picture does not provide us with any nuanced information in regards to different ways of engagement, different motivations etc, and thus risks to lead to oversimplified conclusions.

3. Top-level domains are not a good way to approach data when analyzing different countries that use different systems. It may still be valid among countries that use the same schemes or within a single country, as long as the researcher understands clearly how it is used in such country.

Given those recognition we strongly recommend to take findings resulting from digital methods research with a grain of salt and to consider the combination with more qualitative research methods.

Keeping this in mind, we continue to discuss some of our findings against the background of our question if civic tech bears potential to be a new hope in civic-driven change and suggest possible paths to move forward. By the end of our project week, having explored our data from so many different ankles we kept getting back to wonder if we actually detected some sort of ‘race for the new power’ to be harnessed through the increased availability of data and means (tools) to use such data to drive change. A race of different actors (activists, NGOs, startup companies) with likely different interests.

What we generally found in the results of various methods is the thematic strong affiliation of the civic tech field with the open government and open data movement. This can indicate the strong roots of the civic tech field in the already older open government and open data field. It can also indicate the framing of its core business of opening up governance space towards stronger public participation or ownership. Such findings have been strengthened through our exploration using a pre-filtered data set which only entailed tweets including #civictech or #civictechnology.

Our twitter analysis, as to be seen in the network graphs supports such central relation. Looking at the wider affiliations then shows two fields, the ICT4D, and as such international development cooperation as well as the startup scene, and as such more business driven field. Such results suggest two fields dominantly driven by very different underlying ideologies engaging with the civic tech environment. Such picture suggests further investigation allowing to strengthen a potential argument towards the ‘race for power’. It can be assumed that the field will evolve differently if it will be dominated by profit interests or e.g. a rather civic driven participatory democracy agenda.

A further interesting finding deriving from our data (here the co-hashtag analysis) in those regards, is the differentiation or distance between government and open government. What we find here is a significant closeness of the government hashtag with hashtags such as IoT (Internet of Things), tech or innovation whilst the open government hashtag is, as discussed earlier, closer to the civic tech and open data movement. This could suggest indications towards a potentially rather movement / civil society related open government field and associated civic tech actors; and the actual happening within government agencies. It could be interesting to invest further research into main government alliances when it comes to increasing their ‘openness’. Stronger affiliations might be found here with the actual private sector. Considering our initial research interest this could suggest a reason for concern considering potential underlying ideologies impacting decisions to be made (in relation to main interest, context-considerations, civic engagement, etc.)

Exploring the actors we could identify a rather centricity of core actors which, following our little extra exploration into top actors relations does also suggest strong relations in regards to funding, ownership and association links between such rather small actor group. Analysing the main hashtags around the core actors could suggest to indicate which actors have been more successful in driving issues around the civic tech field on twitter. The private sector seems slightly closer to the core civic tech activity on the intersection of civil society and public sector actors, which are to be found closer to the center of the network. Overall, this could be some sort of indication for our ‘race to power’ hypothesis.

Deeper research into such actor engagements as much as a general more qualitative analysis into the extracted URL lists as presented in the former chapter would be interesting in order to investigate which sites are unique per domain. This analysis would also contribute from a more extensive analysis, including a larger N of country domains.

We did find a certain centrality of actors and discourse-dominance by those actors. A brief deep-dive into the relational network provides us a very first indication for cross-categorizational relations, suggesting a rather close and centralist network, particularly when it comes to private investment and philanthropic funding.

Our top-level domain analysis did suggest a strong ‘eurocentrism’ around the topic. However, we wish to emphasize the need for further research including the civic tech term dominating other language areas. Furthermore, it is to suggest to open up research towards the issue around the ‘civic tech’ umbrella term or the underlying ideology.

Drawing any bold conclusions form the density of hashtag co-occurrence should be treated with a caution. Just a few considerations coming to mind in those regards are:

• Are issues (hashtag proxy) egocentric? Or are there lots of actors talking about it?

• Which of those patterns could serve which research questions?

• Are there issues monopolized by a couple of actors in the Twittersphere? Be careful with bias due to these active actors.

In order to take the research to a more qualitative level, we developed an approach to the selection of issues to be focused amongst the myriad of hashtags used on Twitter. That is justified because the simple monodimensional criteria (such as most used hashtag) simply doesn’t answer to every research question on the field. In our case, this development was propelled by the necessity to understand the diversity of issues around the “civic tech” orbit in order to define the field. This diversity will most probably be different according to the criteria we design to select the hashtags. We would suggest then to concentrate on hashtags that: (i) are spoken of by a reasonable amount of different users; (ii) with a relatively high frequency regarding the dataset standards and (iii) that is not dominated by a few users who drive the conversation, inflating the total number of occurrences of such hashtag. To see more on the elaboration of this refer to the Appendix of this document.

The limited data available to inform our question addressing the geography of civic tech did provide some very rough indications regarding an overall actor and discourse dominance in the US and Europe. Most significant finding were the slightly stronger geographic dispersion of the NGO community as well as the exclusive ‘Global North’ clustering of research and private sector actors.

Obviously, any kind of conclusions based on such limited data we worked with is to be seen as highly indicative.

A few considerations here to be played out and providing potential interesting hints for further research are the following.

As for the public sector, taken into consideration earlier findings, such geo-centrism could likely be due to the apparent ‘relabeling’ underway. Where civic tech startups might be closely affiliated to public sector innovation in the US and Europe, as we saw earlier it could be that the open government, thus stronger civil society and non-profit associated work is strongly executed by NGOs in the ‘Global South’. Here many activities and thus discourse might therefore fall under the more traditional open government discourse or under the rather umbrella term of ICT4D.

What is rather concerning but not surprising given the general prevailing criticism towards research domination by Western researchers when it comes to e.g. the role of ICTs in development contexts (Diaz et al 2012), is the apparent centrism of research in the US and Europe.

Conclusions

We can list the following core findings:

• Strong relation to other ‘meta topics’ (open data, open government, IoT)

• Rather close network of actors

• Main bridge between civil society and public > but private sector closer than NGO sector > Such tendency could be perceived as concerning from a democratic-ideological point of few, where profit-oriented decisions might head the field.

• Need for more thick data for more meaningful interpretation of web and social media data

• Need to complement online analysis with other methods / qualitative research

• Different issues are related to different ecologies of actors (some more diverse, some more oriented to a type of actor, some are “monopolized” by one or a few actors etc.). That could be an interesting approach to understand the fabrics of public policy around some of the issues regarding civic tech and to map the battle of interests in the field.

We can furthermore conclude that the exclusive dependence on digital methods for above elaborated research bears various limitations rooted in the methods as much as the specific objects of analysis. Main risks for exclusions and consequently inaccurate results are:

• Overall limited focus on the digital sphere

• Language, different terminology in different languages and different countries

• URLs, domains (Lack of standardization in URL adoption through different countries limits its use as a port of entry to comparative international analysis)

• Representation

• Difference between funding (and thus potentially visible) versus executing actors in the field

Bibliography

Baack, Stefan. 2015. “Scraping the Global Civic Tech Community on GitHub, Part 2.” Civic Hacking and Journalism. http://sbaack.com/2015/11/19/scraping-the-global-civic-tech-community-on-github-part-2.html.

Gilman, Hollie Russon. 2016. Participatory Budgeting and Civic Tech: The Revival of Citizen Engagement. Georgetown University Press. https://books.google.nl/books?id=PR5ADAAAQBAJ&pg=PA2&lpg=PA2&dq=civic+tech&source=bl&ots=8w9_VhG2w3&sig=nl9bg3SLkeK-7fbbXk7ogitSL_Q&hl=en&sa=X&ved=0ahUKEwju95SksafNAhXBCBoKHb4qA4s4HhDoAQhLMAg#v=onepage&q=civic%20tech&f=false.

McCann, Laurenellen. 2015. “5 Modes of Civic Engagement in Civic Tech – Smart Chicago.” Smart Chicago. http://www.smartchicagocollaborative.org/5-modes-of-civic-engagement-in-civic-tech/.

Omidyar Network, and Purpose. 2016. “ENGINES OF CHANGE: WHAT CIVIC TECH CAN LEARN FROM SOCIAL MOVEMENTS.” Washington DC. https://www.omidyar.com/sites/default/files/file_archive/Pdfs/Engines%2520of%2520Change%2520-%2520Final.pdf.

Rumbul, Rebecca. 2015. “Who Benefits from Civic Technology? Demographic and Public Attitudes Research into the Users of Civic Technologies.” London: mySociety. https://www.mysociety.org/files/2015/10/demographics-report.pdf.

Stempeck, Matt. 2016. “Towards a Taxonomy of Civic Technology.” Microsoft on the Issues. April 27. http://blogs.microsoft.com/on-the-issues/2016/04/27/towards-taxonomy-civic-technology/.

Some initial background reading on civic tech labeled discourse

Bimber, Bruce. 2000. “The Study of Information Technology and Civic Engagement - 10584600050178924.” Political Communication 17 (4): 329–33. doi: http://dx.doi.org/10.1080/10584600050178924.

Cavanaugh, John W. 2000. “E-Democracy: Thinking About the Impact of Technology on Civic Life.” National Civic Review 89 (3): 229–34. doi:10.1002/ncr.89305.

Díaz Andrade, Antonio, and Cathy Urquhart. 2012. “Unveiling the Modernity Bias: A Critical Examination of the Politics of ICT4D.” Information Technology for Development 18 (4): 281–92. doi:10.1080/02681102.2011.643204.

Edwards, Duncan, and Rosie McGee. 2016. “Opening Governance.” IDS Bulletin - Transforming Development Knowledge 47 (1). http://opendocs.ids.ac.uk/opendocs/bitstream/handle/123456789/7686/IDSB_47_1_10.190881968-2016.103.pdf?sequence=1.

Hansson, Karin, Kheira Belkacem, and Love Ekenberg. 2014. “Open Government and Democracy: A Research Review.” Government Information Quarterly 14 (4): 397–406. doi:10.1016/S0740-624X(97)90035-6.

Longo, Justin. 2013. “Open Government - What’s in a Name?” The Governance Lab @ NYU. August 5. http://thegovlab.org/open-government-whats-in-a-name/.

Menéndez, L. S. (2003). Análisis de redes sociales: o cómo representar las estructuras sociales subyacentes. Apuntes de Ciencia y Tecnología, No 7, junio de 2003.

Wissenbach, Kersti. 2016. “Data Revolution or Data Divide? Can Social Movements Bring the Human Back into Civic Tech?” presented at the TICTeC The Impacts of Civic Technology Conference, Barcelona. http://lanyrd.com/2016/tictec/sfbccb/.

Wissenbach, Kersti. 2015. “Civic Tech and NGOs … Wait, and Donors – Can We Be Better Collaborators in Global Transparency and Accountability Work?” Open Knowledge International Blog. July 10. http://blog.okfn.org/2015/07/10/civic-tech-and-ngos-wait-and-donors-can-we-be-better-collaborators-in-global-transparency-and-accountability-work/.