- This piece was written over a year ago. It may no longer accurately reflect my views now, or may be factually outdated.
- After writing up most of the first day, I was distracted by other things and didn't return to this until November, at which point there wasn't must point in continuing the writeups. There was some good material, like the Facebook reps with a manic cultist's gleam in their eyes and a conversation with a reporter at L'assemblée nationale at the end of the second day, but alas.
As an early, but as yet not particularly active, Code for Lancaster member, I spent much of the last week in Paris for The Impacts of Civic Technology Conference (TICTeC), organised by mySociety—they of They Work for You and What Do They Know fame. As a recent convert I was keen to catch up with where the rest of the civic technology world was at, get to know the community and hopefully bring back some lessons and ideas to help with growing Code for Lancaster at its current, infant stage.
As with previous conference/event reports, I will be mixing in my own commentary and tangents with summaries of presentations and events. Click here to hide all commentary and leave only reportage. I will also be added links to presentation recordings as and when they are released.
Welcome to TICTeC 2019 and Welcome from the OECD
mySociety CEO Mark Cridge kicked things off by welcoming us all to TICTeC and to the OECD. Civic technology is at a vitally important point in its history, he said, finding itself in the midst of a
slow-burn environmental catastrophe, at a time when
trust in institutions is at an all-time low and whilst
local services are struggling under austerity—this was my first indication that I probably wasn’t going to run into any Tories whilst I was here (or any who would admit to it, at least)—and had to evolve with the times, from the
early days of volunteer-led hacking efforts to something more international, and something that could move beyond
tech. for tech’s sake. Summing up what would become the unofficial theme of the next two days he posed a question that civic tech. had to ask itself:
if tech. is the answer, what was the question again? He closed with a quick overview of some of the speakers to come, highlighting that the conference had a majority of women speakers for the first time.
Next up was the OECD’s Director of Public Affairs and Communications Anthony Gooch to introduce our hosts. Recognising that being an international organisation was currently
not in vogue in certain places, Gooch stressed that he and his colleagues
don’t believe [themselves] to be the centre of the world After asking the audience how many felt they knew what the OECD actually did and receiving only a few tentative hand raises, Gooch explained that the organisation—which emerged out of the ’99 Seattle WTO protests—was not just a talking-shop for governments. Gooch countered UK MP Michael Gove’s much-criticised claim that people were
tired of experts by suggesting that experts were making a comeback, but behaving differently. Their focus has shifted
beyond proving cold hard facts to
improving lives, and Gooch cited as evidence of this shifting approach the OECD’s invitation of Occupy leaders to visit following it being cited as a potential solution to the problem of income inequality in an op-ed by the movement (possibly this one).
Gooch explained that as part of this shift, the OECD’s Open Government Team was working to
push the needle from Echoing Cridge, Gooch reiterated that civic technology had, too, seen a shift in decades past from
Open Governments to
a fringe movement of hackers and coders to being presented as the saviour of governments in the pages of Le Monde. When
Parisians living within 20km of each other may have subjective wellbeing as varied as a Ukrainian and a Swiss, Gooch finished by plugging the OECD’s May Forum and warning that
the digital divide is not eradicated by civic tech., and more must be done to ensure that the rising tide raises all ships equally.
Personally, I had very little in the way of pre-conceived notions about the OECD going into the conference—I was one of the many whose hands remained lowered when asked if we knew what it even did—but I did notice one thing during my time there. On the walk up to the building, there were an awful lot of very nice cars bearing special license plates (I assume of a diplomatic or similar nature). Inside the building, however, the toilet paper was as cheap and scratchy as you can get. That some of the greatest cheerleaders of the global world order talk about improving lives whilst driving flash cars, yet are also walking around with sore bums, is an observation that you may take however your politics allow.
The Third Age of Civic Tech.
After first apologising for her
gimmicky title, mySociety Head of Research Dr Rebecca Rumbul—what a waste, not becoming a pro wrestler with a name like that—again detailed what she perceived as a shift in how civic tech.
talks about itself and how it is talked about over the past 12–18 months. Rumbul divided the evolution of civic tech. into three phases: first, the
ideas and building stage (what a later presenter would describe as the
throwing spaghetti at the wall to see what sticks stage); second, the
scaling and broadening stage; and, thirdly, the
maturing and re-evaluating stage. It was time for civic tech. to ask itself
if the tech. we’re putting out there is doing what we’re saying its doing, or whether
we [have] built our civic tech. castle on quicksand.
Rumbul was optimistic, however, describing this
third age as
like the third instalment of a really good trilogy (with the examples of Indiana Jones and the Last Crusade and Return of the Jedi to illustrate her point). In The Last Crusade, for example, it
turns out that Indiana Jones wasn’t actually called . Similarly, it may be that some of the fundamental views of civic tech. aren’t actually true, and that now is the time to discover which.
Rumbul concluded by posing a few such self-reflective questions.
Are we being ambitious or culturally sensitive enough?
Are we taking our eye off the ball in some places? She concluded by admitting to
not know[ing] if there’s a fourth age of civic tech., or whether what comes next
might be the first stage of something else. Whatever it would be, she was certain it would still be
cool and exciting.
Fostering democractic societies to fight platform populists: the Brazilian experience
The first keynote of the conference was delivered by Alessandra Orofino, Executive Director and co-founder of Brazilian activist organisation Nossas. She opened bluntly:
I’m here to talk about very depressing things, but hopefully leave you with reasons to be optimistic. Borrowing the term from New Power, Orofino described a rash of new leaders around the world—your Trumps and Bolsanaros—as
platform strongmen, combining old power politics with new, connected means of gathering it. As an example of such leaders, Orofino pointed to Rio’s recent election of someone
who hadn’t even been showing in the polls two weeks prior, but who was heavily bouyed by the support of evangelical Christians. Citing such evangelical sects as an example of
a successful distributed model, where anyone can become a pastor and set up a church, Orofino laid out what must be done:
we need to beat them at their own game.
This must be done, however, without compromising one’s ethics. It must be a long-term plan, the aims to
foster democratic societies, not just democratic governments. Orofino lamented that
at some point, democracy has lost the elections and the challenge now was to build a society that is
practicing democracy between elections. She divided this into two approaches: resistance, which manifests as activism; and regeneration, which manifests as regeneration. Chiding civil society NGOs for what she considered to be a reticence to
declare opposition through fear of falling out of favour with the ruling (or could-be-ruling) powers, Orofino detailed some of Nossas’ previous work, such as it’s 2012 Escola Não se Destrói campaign, where a webcam’s gaze was trained on a school slated for demolition and residents could sign up to receive alerts whenever demolition activity was beginning.
To demonstrate the promise of such activism/solidarity, Orofino introduced
Brazil’s first elected : João Doria, elected mayor of São Paulo in 2017 on a platform of reversing his predecessor’s car-related regulations and bolstered by an intensive social media campaign that saw him livestreaming regularly from 5am until 2am. In the midst of this sat Nossas’ Paulista Aberta campaign, which had successfully campaigned the previous mayor to institute car-free Sundays in a number of São Paulo locations and to turn the streets over to pedestrians. Doria was quick to respond to worries that the policy would end up another casualty in his undoing spree, declaring that as the policy was not his predecessor’s, but the people’s, it would stand. Orofino concluded that
the best home for the continuation of a policy is to ensure that is is backed by the people, this being
the counter to platform strongmen. Sounds like the social contract is back with a vengeance.
Fast-forward, and Doria is now governor of São Paulo. In his first week, he vetoed a Parliamentary proposal to set up a number of 24-hour female-staffed special police stations for victims of sexual crimes. Leveraging the pre-existing activist base of their Mapa do Acolhimento campaign, which connected volunteer lawyers and therapists with sexual assault victims, Nossas countered Doria. As the Mapo do Acolhimento activists had
been flexing their civic muscles for some time, Nossas were able to ask for more than just an email campaign or for people to sign a petitions. Shortly thereafter, Doria relented (and, of course, bigged up the new police stations on his social media accounts).
platform strongmen would not be defeated overnight. Orofino’s admitted to be
playing the long game, and that we
won’t see a lot of victories over the next few years. Despite this, however, she has cause for optimism:
I am certain that the next generation will be better than us—more connected, more open, more queer, more tolerant, less cynical, more honest. Personally, I’m not sure I’m as optimistic about Gen. Z as Orofino—as a friend of mine put it,
this is the generation that ate Tide Pods. Time will tell.
Orofino wrapped up with some advice for how to successfully strengthen civic society. Ultimately,
governments will come and go, but citizens will still be here, and
tech. is not an end, but a means. Organising is based on connecting, having a common cause and engendering trust—Orofino cited the Bolsanaro campaign’s co-option of pre-existing WhatsApp groups as showing the
need for connection and the advantage of having a pre-existing base that can be retasked (a la the Mapa do Acolhimento story).
Moving into the Q&A, one questioner likened Orofino to US Representative Alexandria Ocasio-Cortez and asked if she had any plans of
getting into politics, and whether she ever worried about her safety. She replied that
I’m already in politics, but assumed that he had meant specifically running for office. She didn’t rule it out as something she might like to do, but warned that
we assign way too much power to the person holding the pen, when they’re only responding to different groups’ pressures. Regarding her safety, Orofino said that it was a concern, particularly for those on her team from less privileged backgrounds, but that she often reminded her team that
we will not die until they kill us. As the
platform strongmen like to drape themselves in the trappings of
democratic normalcy, she warned that to be silenced by one’s fear would not been seen as being a result of the strongmen, and would only serve to strengthen them. It was better to force them to reveal their true face if they wanted to suppress critics with violence.
Finally, she was asked how one can expect people to have solidarity with those who dehumanise them, and how to do this sort of solidarity
at scale. Orofino replied that
one cannot extend solidarity to someone they do not see as an equal, but that part of the platform strongmen’s strength came from their early realisation that traditional demographics—16–25, white, etc.—don’t work. Instead, the strongmen target psychological types, such as
convinced by numbers. This lesson can be repurposed by others, and Orofino pointed out that many Nossas supporters were also evangelicals and Doria supporters—
it is important to never say that the strongman equals his supporters. In closing, Orofino reiterated the value in long-term campaigning and trust-building, suggesting that these supporters supported campaigns opposing their own candidates because Nossas
have been around for seven years, and people know we’re on their side regardless of who’s in power.
Evaluating the impacts of voter information campaigns
The last presentation before lunch brought David Alzate and Eliza Keller of MIT’s Abdul Latif Jameel Poverty Action Lab (J-PAL) together with Profs Jonathan Weigel (of the London School of Economics) and Yusuf Neggers (of the University of Michigan) for a discussion of data-driven impact assessment case studies.
Keller went first, summarising J-PAL’s approach and explaining how
low expectations lead to low accountability in government and that, as a result, J-PAL were currently interested in randomised control trials (RCTs) that could answer the question:
how to citizens form expectations in the first place? Their research had found that voter information dissemination efforts had significant impacts on voter choice, turnout and more. For example,
being presented with information about candidates impacts who citizens voted for, which led to more qualified and accountable candidates being elected, with a subsequent improvement in outcomes.
The key lessons of J-PAL’s research were that:
policy preferences can transcend in-group preferences;
information should be disseminated widely from a credible source (good luck finding one the everyone can agree on, it seems lately);
voter coordination effects lead to voters behaving differently if they believes others in the community are viewing the same sources;
voters’ prior beliefs matter; and
information that confirms rather than surprises voters (e.g. evidence of corruption in a political system already considered corrupt) has a discouraging effect on engagement, whilst surprises lead to an improvement in turnout.
Neggers went second, detailing a study carried out in India—the
world’s largest democracy, and one with a
strong, independent electoral commission. Incredibly, 21% of current Indian MPs have serious criminal charges pending against them. Neggers was interested in whether
voters are generally indifferent to, or even value, criminality in elected officials, or whether an information gap exists. If the latter,
can a light-touch approach address the gap and influence voting?
His study took place in Uttar Pradesh, which is the most populous and one of the poorest regions of India whilst simulteneously having the highest level of mobile penetration, during the 2017 state assembly elections. 25% of candidates were found to have serious pending charges, with 26% of races having at least one candidate with a murder-related charge pending. Having divided 3,800 rural villages (each containing ~1,200 voters) randomly into control and treatment groups, Neggers and his team sent voice and SMS messages two days before the election to all treatment village citizens detailing the number and type of charges facing each of their candidates. A subset of the treatment group also received co-ordination content (e.g.
many other citizens in your area are receiving the same message). Ultimately, almost half a million individuals were reached.
The researchers found that the messages induced a significant change in voter behaviour. Candidates with murder-related charges saw a 7.5% reduction in votes, whilst those with no charges saw an improvement. Those with less-serious charges saw little to no change. Overall, turnout increased by 1.6%, and impacts were almost doubled when the co-ordinating message was included. Though this boded well for future elections, Neggers identified some issues that may hold back wider deployment, not least of which was opposition from government—
at larger scales, parties may notice and react. In addition, other countries may not have the same requirements to file criminal and financial information with an electoral commission in order to run, and telecom. companies may be less willing to send potentially-sensitive political messages elsewhere.
Weigel was third, presenting his own work in the Congo looking into
where accountability demands come from in places without democratic institutions. He proposed that the origins of political institutions lie in taxation, which leads to participation, which leads to demands to accountability—
no taxation without representation, and all that. To test this tax-participation hypothesis he went to Congo, where he measured civic participation in the wake of the introduction of the country’s first-ever property tax campaign.
Kasaï-Central province, with a population of around 1 million, previously had a property tax on paper, but as it was self-declarative compliance was minimal—many did not even know they were supposed to pay that. The Democratic Republic of the Congo as a whole is by no means a democratic state, being as it is a
competitive authoritarian state that has cancelled elections for years and which has no institutionalised parties. Civic engagement takes place via the submission of written complaints and attendance at local meetings, rather than through voting.
Kasaï-Central, long-neglected by the capital, was hoping to increase tax payouts, the money from which would all go toward the province itself. To this end, they dispatched local bureacrats to create neighbourhood property registers and to appeal for tax payments. These neighbourhoods were Weigel’s treatment group, whilst his control group was those still on the old, self-declarative system, where compliance was near zero. Weigel reported a 64pp increase in tax collector visits, along with a 11pp increase in payment rates. Still, 89% of people in treatment neighbourhoods managed to avoid the tax.
With this modest increase in tax compliance, he returned half a year later to record attendance at provincial government town-hall-style meetings and suggestion card submission rates, finding that both forms of costly participation had risen slightly since the tax campaign. Treated individuals demanded better services, believing the provincial government should handle their provision rater than the national government or NGOs. However, Weigel noted that the participation increase did not appear to be payment-based—tax evaders were just as likely to participate as the taxed—which suggests that even just running the campaign
sends a signal of state capacity, raising the expected benefits of participation. In other words, government support depends on government being seen to do things, rather than necessarily doing them.
Welcome from Google Civics
Originally slated to be an hour-long discussion amongst four Googlers entitled Scaling Impact through Civic Product Partnerships, the non-attendance of one of the presenters had apparently scuppered the entire thing. In its place, Claire Foulquier-Gazagnes appeared to give a brief address to conference attendees on behalf of Google, a long-time supporter of TICTeC. She told those interested in civic tech. to
work in government as a means of developing
empathy for those who have to use outdated systems and to
see how law is made. She also talked about wearing multiple hats, delivering a welcome to attendees
…as a French person, as a person who has moved from government into the private sector and as a Googler, before introducing the handful of her colleagues who were also present and advising those present to take advantage of the location of this year’s conference and ask a French person for their perspective on civic tech., being as we were some way in to the gilets jaunes protests and Macron’s
Grand Debate in response. The premise of the originally-scheduled talk hardly filled me with passion, but its replacement did come off as very back-of-the-napkin,
we have to fill minutes, which seemed a shame. It seems they didn’t even bother to upload a recording to the YouTube channel.
Social, political and public engagement
I headed off for my first session outside of the OECD conference centre, in the very impressive Château de la Muette. I’m not quite sure what happened to the first scheduled presentation—Sabrina Wilkinson’s Reddit, surveys and online interventions: Evaluating public engagement efforts in communications policy development in Canada, but instead we began with Benjamin Snow of Civocracy, who presented Learning from setbacks: Civocracy and citizen consultations.
Snow began by admonishing us for not attending the workshop on Challenges and opportunities for women in Civic Tech going on next door, suggesting that
…none of you care about women gaining funding or well-being, which is…interesting—Snow’s rather dry sense of humour continued throughout which I found entertaining, though I wonder how well it may have translated for attendees from less sarcasm-heavy cultures than my own. The thrust of Snow’s presentation was an analysis of
when stuff doesn’t go well, and why that might happen, focusing on the different levels of success Civocracy have seen in their work in Potsdam, where they have only engaged a few dozen people, and Lyon, where they have reached thousands.
First, Snow suggested that asking for feedback on civic tech. projects was important. Whilst in Potsdam
no-one knew about [Civocracy’s] projects, in Lyon they were having great success with campaigns on primary schooling, which
affect everyone. Secondly, he identified the need to have multiple
champions in partner organisations. Civocracy had had only a single champion within the local Potsdam government and when that person had moved on, they had limited support for their projects. Snow here offered some advice, saying that
if you are the figurehead in your organisation, make friends. Finally, he highlighted the importance of design in successful civic tech. projects, suggesting that the tech. itself was
not very important in doing civic participation online.
Following on from Snow was Prof. Marko Skoric of the City University of Hong Kong with Being unsocial on social media: the implications for civic and political engagement. Skoric first proposed that
conversation is the soul of democracy and that
lots of mono- or dialogues are bad. For the first time, courtesy of the rise of social media,
daily conversations are being recorded and monetised. Framing his presentation as a timeline of the rise of social media, Skoric began with 2007—the start of the conversation about social media and democracy. At this point, the discourse focused on new technologies’ ability to
change everything, in a democratising way. Things like the then-novel Facebook were predicted to increase social capital (reversing the trends observed by Putnam 7 years earlier in Bowling Alone) and to promote context collapse, through exposure to diverse views and perspectives.
2011 brought with it the Arab Spring, in which
an army of Davids used social media tools to challenge authoritarianism in their own countries. Despite this, Skoric observed, democracy is currently in a decline. This tied in with the 2013 global surveillance disclosures, which came at a time where the
everpresent liberation narratives still focused on tech.—as a response,
people retreated into small groups and private settings for their conversations. This was further reinforced in 2016 with the political polarisation around the US presidential election and in 2018 with the Cambridge Analytica scandal and the introduction of the GDPR in the EU.
Skoric believed that
the social media sphere is being reconfigured and redesigned, citing the example of
Turks with 12 Facebook accounts—one for the wife, one for the mistress, one for the mistress from abroad, one for the kids, one for his friends. With users shifting to E2E-encrypted messaging apps and other forms of ephemeral social media (e.g., Snapchat), Skoric lamented the impact on his data collection abilities, complaining that he now had to rely on the
19th-century technique of interviews.
He then moved onto the recent phenomenon of political unfriending, which peaked during the 2017 post-presidental election period in the US (nominally a usual event), but which also saw jumps during the Hong Kong Umbrella protests and the 2014 Israel–Gaza conflict (both unusual events). Skoric identified that teenagers are increasingly vetting the online presence of potential connections prior to connecting with them, utilising the joint strategies of selective avoidance and post-hoc filtering. Rather than context collapse, this represented
context relapse as people shifted to
smaller, more intimate platforms that support homophilous interactions and content.
Unsurprisingly, confrontational discussion emerged as the most robust predictor of unfriending/unfollowing and content hiding—the
weakness of weak ties. Citing other research, Skoric discussed sender effects—i.e., someone who announces their intention to vote (especially in front of friends) will be more likely to vote—in the context of vocally political people online being the most likely to participate in political activity, whilst others withdraw. However, when people feel safe within small groups, they are more likely to express their own views, even when those views differ with other members of the group.
As people carve out their
digital safe spaces, research apparently shows the overall impact to be depolarising. This effect is especially important for disenfranchised and minority groups, who can build confidence and skills in a space space before going out into the big wide world to make a difference. Skoric closed but suggesting that in contemporary conversations about the breaking up of tech. giants/
data extraction companies, the balance between transparency and privacy must be maintained. We finished with a Q&A (including one from yours truly).
Note: At this point, life intervened and I was distracted from completing my writeup. I didn’t return to it until November, at which point there was little point. I’ll stick a few highlights below:
- Going to the separate OECD anti-corruption forum event on Integrity Idol, which highlights civil servants in corrupt nations performing their jobs with integrity and is massively popular in South Africa, Pakistan and Mali. One winner was called immediately after winning by the governor of Punjab to ask what job he would like.
- Faebook came to talk about their efforts in countering election fake news and bots. This included offering all candidates free video slots to be shown to their voters, as well as (are you really surprised) the retention of ever more data on users. I asked how this attempt to insert Facebook into the electoral infrastructure and this gathering of data on potential rivals might square with rumours that Mark Zuckerberg was considering a run for the presidency, to which the Facebook reps (with a manic, cultist-like gleam in their eye) said they had never heard of such rumours. Quite how you miss claims that your boss (who they were evidently close enough to to refer to as
Mark) might run for the presidency is beyond me. I also asked about how the creation of profiles on non-users orchestrating misinformation campaigns squared with Facebook’s insistence that they do not keep shadow profiles of non-users. The rep said it was a shame no-one from that division was available to answer me which, considering the talk was entirely about that division’s work, I can only assume was intentional. Someone also accused Facebook of profiting off of the genocide of the Rohingya, which was spicy.
- At the end of day 2 we had a drinks reception at l’Assemblée nationale. I spoke to a reporter who criticised how corporate the conference was, with keynotes and sponsorship from Google and Facebook. I concurred. I also had a chat with a French civil servant about Macron’s