The politics of participation

The Economist Intelligence Unit’s Democracy Index 2017report has a particularly revealing quote from Edward Luce, who in his 2017 book ‘The retreat of Western liberalism’ observes that “our societies are split between the will of the people and the rule of the experts—the tyranny of the majority versus the club of self-serving insiders”. Luce is focused on the West – Europe, the UK and US in particular – but his central thesis of democratic decline, because of a rise in populism and an authoritarian resurgence, finds resonance in helping explain the situation in Sri Lanka. The threat to democracy in this reading comes not because of manipulative foreign actors, but the disenchantment with and distrust of democratic dialogue and institutions by constituencies increasing taken in by a toxic recipe. Misinformation, a general decline in trust around media, rising intolerance of difference, increased social and religious clustering around identity markers that are exclusive, a proclivity to the violent resolution of conflict and an increasingly divided electorate on partisan lines are some of the ingredients in this recipe, which in fact, we are co-creators of. As the Economist notes, “The popular reaction to an economic and political system which many voters feel has left them behind is presented as the cause of democracy’s ailments rather than a consequence of them”.

The report is helpful to understand risk vectors in Sri Lanka that aren’t adequately discussed. The democracy recession can be seen in, according to the Economist, through declining popular participation in elections and politics, weaknesses in the functioning of government, declining trust in institutions, dwindling appeal of mainstream representative parties, growing influence of unelected, unaccountable institutions and expert bodies, widening gap between political elites and electorates, decline in media freedoms and the erosion of civil liberties, including curbs on free speech. Many of these one finds not just in Western liberal democracies but in Sri Lanka post-2015. We have a country with very high adult literacy losing faith in democratic government. The communication of the government’s failures, coupled with the failure of government to communicate, are two sides of a problem that is leading to the erosion of trust. The electoral implications are not theoretical. The 10thof February demonstrated the degree to which the government has lost popular appeal. This is not the same as saying that the Rajapaksas, JO or Pohottuwa have gained any greater appeal. The electorate is faced with a conundrum – on the one hand, a largely liberal and democratic government unable to fulfill its lofty promises and is insensitive and technocratic to boot. On the other, representatives of a more authoritarian form of government who seek a return to power and though essentially corrupt, brutal and violent, gets things done, puts everyone in their place and are masters at generating populist charisma by posing frequently with children with plats or pottu, infants, the disabled, soldiers, the poor, priests and old people. Embedded in this reading is an asymmetry of generating self-serving spin and positive optics for parochial gain. The current government is horrible at it. The former government wrote the rulebook on it.

This all feeds into what is a systemic problem of politics in the way it is negotiated, conducted and perceived. In 2014, the Economistgave Sri Lanka a score of 4.44 for political participation, a metric that measures the degree to which the population engages in electoral processes and more generally, is involved with governance mechanisms between elections. By 2016 this had increased to 5.00. It remains the same in 2017. There is also a metric for political culture. The Economistflags this as “crucial for the legitimacy, smooth functioning and ultimately the sustainability of democracy. A culture of passivity and apathy, an obedient and docile citizenry, are not consistent with democracy. The electoral process periodically divides the population into winners and losers. A successful democratic political culture implies that the losing parties and their supporters accept the judgment of the voters and allow for the peaceful transfer of power”. In 2014 and 2016, Sri Lanka scores 6.88. Intuitively, especially if one supports the current government, you would expect this score to be stable or rise. Instead, in 2017, the score goes down, to 6.25. What we see in these figures is a risk vector that ironically is pertinent precisely because of the numbers that turned out to vote in 2015’s Presidential and General elections. In both instances, a youth bulge in the electorate – 1sttime voters as well as 2ndto 4thtime voters, all between 18 to 34 – supported the elections of those currently in power. The social engineering to get this demographic go out and vote was conducted over social media almost exclusively for the Presidential Election. By August 2015, the apathy and disappointment with the new government had already taken seed, which is why another concerted effort to get young people engaged in political communications was needed. Most if not all of this content generation and strategizing was done by civil society – some admittedly with partisan bias and intent, others more involved and interested in generating interest amongst the youth in our electoral processes and the value of democratic institutions. Either way, what is evident today is that the heightened interest and participation in political conversations, just three years ago, has now led to deep disappointment and disgruntled disengagement. This fits very well with those who want to regain power, mirroring how in the US, Republicans in 2016 used against Democrats technologies and strategies first imagined, seeded and set in motion as part of Obama’s 2008 presidential campaign. The revolt of authoritarians or as the Economistcalled it in 2016, the revenge of the deplorables ironically happens on the very social, media and technological foundations put in place by more democratic forces to gain power. This is playing out in Sri Lanka.

To understand this is to grasp the increasing appeal of the JO. Sadly, it is a political reality that was given life to by those in power. What I’ve flagged in recent weeks – the weaponization of social media, the gamification of elections – all stem from the inability to capture the spirit of participation in January and August 2015 and animate it over the longer term. This is a failure of political vision, just as much as it is a failure in political communication. The danger is now reflected in the data – electoral contests ahead of us are going to be perceived as much more divisive, with losers unable to countenance those who gain power, and those unable to regain power unwilling to countenance those in government. 2015 was as moment to rewrite the grammar of our mainstream politics, where the conjugation of divergent political opinion was normalized so that violence wasn’t the intended or immediate result of partisan difference.

Through true, it is easy to say the government has failed us. Truth is, we have failed ourselves, no matter which party we vote for, and who we want to see in power come 2020.


First published in The Sunday Island, 22 April 2018.


Redefining our world: A review of ‘Factfulness’

Factfulness’ by Hans Rosling is the first book in a long time I could not put down once I picked it up. Published after the passing of Rosling, the book was just released and comes with glowing recommendations from Melinda and Bill Gates on its cover. Bill Gates is particularly effusive, calling it one of the most important books he’s ever read, and going on to review it on his blog earlier this month. It is easy to see why, for several reasons. Statistics and complexity are immediate turn-offs for most, even if as students in school, or University, or staff in an office, one is forced to deal with both. Few understand the subject well enough to explain it to others. In the world of media, the domain of data driven journalism which donors love to support is anchored to facts and figures, and the communication of both to audiences who aren’t conversant in computational modelling, time regression, pivot tables, charts, graphs, databases, tables or economic theories. The domain is heavily anchored to visual design. As rows and columns, numbers overwhelm. When presented through infographics, they intellectually or emotionally appeal to a broader audience. Rosling made a habit of provocative talks, at TED, Davos, on TV, in person and in print, exploring the world of statistics related in particular to health and development. To see and hear him speak is to understand the power of communication in shaping hearts and minds. This voice, even after his passing, shines through the book.

Rosling is interested in thirteen questions, the answers to which are counter-intuitive. The book starts with asking the reader to answer these questions, and it helps to not cheat because the rest of the book is based around why the answers readers most often think are the right ones are in fact, wrong.  The author stresses the point that most humans score worse than the statistical probability of giving chimpanzees in a zoo the questions and asking them to guess the answers. It is an interesting observation, anchored to Rosling’s life as a teacher, that is also fleshed out. The reason all of us get the answers so wrong is because we are basing our understanding of the world on outmoded and outdated definitions. These ways of seeing, to borrow a phrase from critic John Berger, are frames founded on economic theories that categorise and capture countries, populations and people in ways that haven’t accommodated how far and fast how the world has in fact progressed over the past seven to eight decades. Rosling calls us what we are – ignorant. The book is a voyage in discovering what the true state and nature of the world is through statistics, quite incredibly, open for all and in the public domain.

What interests me is not so much the data presented, but the voice of Rosling – how he writes, how he teaches, the use of self-deprecating humour, the wit and above all, the love of teaching. Rosling is never preachy, and it is precisely here that the book and indeed, his other output in the public domain hold value for countering violent extremism, a subject the author does not deal with directly. In a foundational chapter early on the book, Rosling notes,

“What do you need to hunt, capture, and replace misconceptions? Data. You have to show the data and describe the reality behind it… But you also need something more. Misconceptions disappear only if there is some equally simple but more relevant way of thinking to replace them.”

Rosling is writing against the labels ‘developing world’ and ‘developed world’, which Bill Gates himself admits having used often before he read this book. In its place, the author places a model based on four levels, in turn based on income. Most people in the world – roughly three billion – live in Level 2, earning around two to eight dollars a day. The author encourages us to discard a language that no longer captures accurately the world as it is, and instead offers a new framework, with an associated language, that more humanely categorises people in ways a bifurcated worldview just cannot accommodate.

Rosling teaches the reader early on the danger of statistics and the pitfalls of interpreting graphs. A simple or small change to the way numbers or trends are represented, and a completely different understanding or worldview emerges, from the same dataset. This is why he encourages in the main a healthy scepticism with opinion, and instead implores us to seek out, for ourselves, the numbers that make a story.

The lesson for countering violent extremism is that when debunking myth or misinformation, one always risks strengthening it. This is referred to as confirmation bias. Telling someone they are wrong, often ends up confirming in their mind they are right. Rosling also hints at motivated reasoning, which is the tendency to read into or project onto whatever we consume our bias. So, a speech, image, graph or policy may mean completely different things to different people, even if they are presented with precisely the same original material. Rosling’s book, if nothing else, is a lesson in communicating inconvenient facts. To the author, we are all idiots, and our answers to his thirteen pivotal questions places us, as he notes, in a position worse than chimpanzees. Rather than take offense, the reader cannot help but be taken in Rosling’s world, and through his incredible gift for writing, teaching and humour, understand at the end of this book a world different to what was seen before. As if not more importantly, the reader is left with tools to critically question the world as it is presented through media and others. Rosling asks us to measure our own knowledge, instead of calling out the ignorance of others.

There is another aspect to this book I have often talked about with Anushka Wijesinha, a friend who is also an economist by training. Rosling is a champion of open data, suggesting that the basis for financial stability and peace is international collaboration, in turn anchored to a shared and fact-based understanding of the world. At the very end of the book, Rosling notes how in 1999 he angered the World Bank by putting on the web and in a way that made it easier to access and understand, statistics distributed first and only on CD-ROM. Rosling’s argument was that the data was already paid for by taxpayers. By 2010, the World Bank had decided to release all its data for free, largely because of Rosling’s insistence. In Sri Lankan context, Wijesinha has also shared my frustration with the inability of institutions, some of which he has been a part of, to release data in the public domain for free, and communicate vital research in a way that kindles the imagination. Both are needed for the construction of empathy, something that in Rosling shines through in this book. Noting that many who read and purchase his book would be on Level 4, spending more than 32 US dollars a day – Rosling states that like looking down at buildings from the top of a skyscraper, it is possible to see many others who less fortunate, but not really easy to understand what their lives are like, or the conditions they live in and work around.

‘Factfulness’, contrary to what the title suggests, isn’t just about facts and statistics. It is about a shared humanity – and the communication of it through evidence already in the public domain. If it there is just one book you buy this year, let it be this.


First published in The Sunday Island, 15 April 2018.

Gaming politics

Asanga Welikala, writing last week on the high-drama around the No Confidence Motion and its fallout, avers that “With hopes crushed and expectations laid low by exactly the substandard culture of parochial politics that the majority of Sri Lankans uniting across ethnic and religious divides hoped to fundamentally change in that nation-building moment of 2015, we now have a Government that is in office but not in power. As was seen in the local government elections, the electorate will punish the self-indulgence of division and in-fighting within government.”

Two points stand out in that excerpt. The fact that we now have a PM who clearly no longer controls, directly or through coalition congruence, all of government. There are many visible manifestations of this just last week, but the problem is longer in the making – the inevitable consequence of an insecure, myopic, self-centred President seeking the security of office even to the extent of reneging promises, and as Welikala flags, the unwillingness and inability of the PM and UNP to go about what was expected of them after January 2015.

The other point about electoral pushback and verdict is an interesting one, pegged to the result of the local government election in early February, projected into the future. Humour me. In last week’s column, I flagged the possibility, going by what happened in Digana alone, that Sri Lanka’s intelligence services may well have personal allegiances to those outside of government that impacts their professional assessment, analysis and reporting. A government that isn’t pre-warned, or is blind-sided by violence clearly engineered by a few, is one that cannot then take measures to contain, control, mitigate or prevent. Worse, as CCTV evidence indicates, elements of the Army and Police are openly in favour of and participating in the violence. We then have strikes, hitting Colombo’s traffic choke points, close to or during rush hour. This is all amplified online. During Digana, accounts on Twitter took to the production of content with no basis in fact, which was then spread wider by other accounts – a network effect that results in the spread of misinformation. Closer observers of the violence in Aluthgama did not see anything remotely akin to this. In 2014, Twitter and social media in general was the means through which the violence on the ground was reported to the world, at a time when mainstream media was largely barred from capturing the extent of it. These vectors are now weaponised to fulfil the aims of those who inflict physical harm. This is done in two ways. One, by the production, spread and engagement with content that is geared to inflame and incite. Two, by latching on to individuals and institutions providing timely updates, eye witness accounts, factual reporting and insight – attacking them as agents of violence or its promotion. This is achieved by individuals, using pseudonyms and in vast numbers, as well as by automated accounts, engineered to spread and amplify the output of those who want to clamp down on inconvenient truths. The net effect is that the information most vital to be heard, seen or acted upon, becomes lost or hard to find – this is sometimes called the signal to noise ratio. With just too much of noise, it becomes hard to tune into to the signals that provide trusted information or updates. It has, also over the longer term, a more disturbing impact.

By making the news and information landscape toxic, it results in distrust and scepticism of all news and media. This works in favour of those promoting misinformation, because in the absence of media or information literacy, those who push out the most amount of content are often the architects of what is consumed. Think of it like a flood, controlled by those who want to overwhelm rivulets of vital reporting. The smaller flows of information are overwhelmed and subsumed by a much larger, faster moving body of content. This makes it possible to construct new and often false realities, by misguiding or misdirecting the attention and energies of a young demographic whose window to the world, and politics, is primarily social media.

This is why something that went unreported by any mainstream media – the unprecedented growth of fake accounts on Twitter over just the past week – is important to flag. The fake accounts mostly featured Muslim, Tamil and Sinhala names. Some had names that were more Western. Many of the profiles had images stolen from others, including notably, one that was taken from Indian cricketing star Virat Kohli’s Facebook page. Others have images of young girls, often Caucasian. Most of the accounts were created in March, and haven’t tweeted even once, though they often follow hundreds of accounts. Academic Raymond Serrato from the outfit Democracy Reporting International recorded around 4,800 accounts created during the time the violence in Digana was raging on the ground. Some of those accounts tweeted (which included the retweeting of content first published by others) over 5,000 times. Two factors are noteworthy here. The age of the accounts and the sheer volume of content production. Think of it like a crèche, with thousands of babies all crying loudly and at the same time for attention. In the same space we have a few more mature voices trying to be heard above the din, with something important to say. It is not a scenario with odds stacked in their favour.

This is precisely what the weaponisation of social media does to public discourse, especially around ethno-political emergencies and within fragile democracies. It is also what can over the longer term be engineered to influence, by the production of content aiming to disrupt, deny or decry through great volume and repetition, the thinking and perception of an electorate over certain issues, individuals, institutions or processes. Buying this sort of influence through fake accounts on social media isn’t hard. Companies sell it for less than fifty US dollars. Combining it with older, well-proven strategies like strikes, demonstrations that can be engineered to result in violence, real world disruption and mainstream media propaganda isn’t that much harder either. The visible physical manifestations of breakdowns in governance leads to venting online, and engaging with angry fellow citizens over social media leads to greater impatience with and anger towards the incumbent government. It is a vicious cycle, with legitimate grievances exploited by a few for the parochial, partisan pursuit of power.

This is, in a sense, a new dynamic within an old problem. Welikala’s submission around the electoral response to division and in-fighting in government is not a new political phenomenon, and arguably is an enduring feature of any coalition government in its sunset phase. What’s new here – what we will see more of but today understand little around, what will be used against us but we aren’t prepared to critically discern or respond to – is the subtle, sustained manipulation of perceptions. There is no quick fix. Studies show that misinformation and falsehood travels, by order of magnitude, faster, farther and wider than fact, or content that seeks to correct misinformation in the public domain. Picking on and exacerbating existing socio-political, communal, ethnic, religious and economic faultiness in society, these sophisticated disruption blueprints will undermine trust and faith in government. All roads lead to authoritarianism as a better form of government. As the New York Timesrecently flagged around Putin in Russia, the exercise of power and control from one central authority is deemed preferable by many to the paralysis and visible lack of progress when democratic institutions or governance is entrusted with reform.

The spectrum of disruption noted here and much more besides, from the physical to the digital, will have lasting effects on our society and politics. My fear is that those who cheer on and support pioneering architects of all this today are starting down a path that can only end up in a form of government that is deeply insecure about the very means through which it came to power. This insecurity has political and institutional manifestations we are all acutely familiar with from 2005 to 2015. My fear is thus not just the return of who was once in power. It is about what kind of politics, public discourse, electorate and government we invest in, inherit and inhabit moving forward, no matter which party or politician we vote for.


First published in The Sunday Island, 8 April 2018.

Icebergs of intelligence

As a mandatory visa requirement in order to undertake doctoral studies, I was asked to get a Police Clearance Report, which I had never done before. Online as well as in person, once my application was lodged, I was told to expect a response in around a fortnight. Six weeks passed. In desperation and as a last resort, I approached someone in government to inquire what the delay was about. In the interim, I had emailed the Clearance Branch four times to an official address given by them, receiving just a single response to my first email noting that my application was still being processed. The actual phrase used was that my application was “still under investigation”. This, ironically, wasn’t far from the truth.

When on the day of the local government election in February, I went into the Police HQ in Fort, armed with an introduction to a senior Police officer to look into my case, I didn’t know what to expect. The HQ is a maze of corridors and partitioned office spaces, and a visitor, like a Pac-Man, is guided by regular sentries in the general direction of the office one has to end up in. After much theatricality in good spirit – references to how such delays always happen to the best of citizens, inquires as to what I did professionally, where my village was, whether I was related to others with my surname and in case I forgot, how busy they were on that day but yet felt it imperative to help me and my case – I finally discovered the reason for why I hadn’t received my clearance. Technically, I was till then under investigation by the State Intelligence Service (SIS).

A file opened on me during the Commonwealth Heads of Government Meeting (CHOGM), five years ago, had never been closed. I was told that the reason for the file being opened in 2013 was threefold – one, the fact the I worked at the Centre for Policy Alternatives. Two, the fact that I ran the Groundviews civic media website. Three, the fact that I had appeared in an BBC programme, aired just before CHOGM, which painted the then government in a negative light as one unfit to host the event. None of this I knew. What particularly struck me was that the file was opened at all based on the stated reasons. Here, in a very corporeal form, was evidence of the deep or dark state. My file I never saw – and it could have been physical, digital, or some combination of both. I didn’t ask and do not know what being under active investigation entails – whether I was followed, whether all open communications were logged and monitored with the complicity of my ISP and telco, or whether family, colleagues, visitors and friends also came under the official surveillance dragnet. The senior Police official’s dismissal of the investigation was an interesting one, noting that it was opened by the previous government and thus had no relevance or merit at present. Interested only hastening the issue of my Clearance Report, I smiled and nodded. It did cross my mind however as to why he said this, unless it was known internally that investigations launched by the SIS under the Rajapaksa regime were politically motivated and had no real value after January 2015. Once the problem was flagged and cleared, my Clearance Report was issued in a matter of two days. My enduring thought is around how many more – dozens, perhaps hundreds – may be technically under active investigation, and worse, unknown today even to the agencies that opened their files in the first place, years ago.

Though there are obvious dangers to any generalisation based on personal experience, this interaction was a reminder of dangers in waiting under two types of government. One, the easiest to pin, is a reversion to the Rajapaksa regime. Dormant files will be reactivated, re-assessed and in most cases, escalated, because soon after regaining power will be the all-encompassing task of retaining it, taking out as early as possible those who are a threat to authoritarian entrenchment. This is why I fear and oppose Gotabaya Rajapaksa. His political campaign, pegged to what is clearly a vision for what government under him can and will do, already espouses violence, murder and even torture without any qualm. When in power and supported not just by a sliver of society, but a far broader coalition of even elites in society, the lessons learnt in how power was lost in 2015 will be leveraged to design comprehensive and pervasive architectures of power, control, censorship and containment. To recall what was done from 2005-2015, both within and outside the theatres of active war, is to project the blueprint of what will be in store if and when power is regained – a process already well underway and succeeding to boot. The glitter and glamour of economic development will be to the rapid and sustained detriment of human rights, with a pendulum shift away from the incumbents in essence supporting the establishment of government that only if conformist, supine and silent, delivers la dolce vita.

Speak out, and you are killed, or worse. There is a second danger. Or more accurately, two combined into one. The enduring role and relevance of the dark or deep state – a euphemism really for the intelligence apparatus – and its anchors within the present government. Threats against a state, arguably, outlive the tenure of governments, so there’s logic and merit in institutional memory that retains surveillance records over the long term. But given we live in a country without any constitutionally or legally guaranteed right to privacy, where surveillance as an exercise of coercive power and control is largely invisible, intelligence services and those in it have unbridled authority to do as they see fit. It is clear that oversight and accountability are weak, if they exist at all. Political allegiances of those in the intelligence community may well be with those outside of the incumbent government, which then calls into question the information senior decision makers get. The violence in Digana alone makes it obvious that there are serious issues around the accuracy, relevance and timeliness of actionable intelligence given to senior figures in government. The hidden hand of the deep and dark state is also evident in the substance and text of the Emergency Regulations in operation during the violence in Digana, the more recent draft Bills to govern civil society, as well as imprecise, overbroad and entirely unnecessary legislation being proposed to combat the generation and spread of hate speech. What is projected and presented as benevolent or beneficial to society, are in fact dangerous instruments of censorship and control.

Who benefits? Who are the architects of these draft Bills and legal instruments? In what guises do they appear in government, civil society and academia? What role do intelligence agencies play in all this, greatly animated by the degree to which they can control information flows as a consequence of the myopic decisions of this government? The dangers around a return to the Rajapaksas are well articulated. The greater danger of the intelligence operatives, never fully visible, never accountable, their tentacles extending to socio-political terrain we don’t know, embracing those who are close to us, corrupting trust and influencing policymaking is not known, studied or resisted.

Personally, the generation of my Police Clearance Report was extremely illuminating, and in hindsight, tragi-comic. More broadly, it reinforced my belief that true enemies of the people are not those most visible under Yahapalanaya or Viyathmaga, but others, always in and close to power, who like icebergs, wait for democracy to sail past, to wreak havoc from beneath.


First published in The Sunday Island, 1 April 2018.

Politics is data

The video and contracts are already forgotten. In 2011, the now disgraced and defunct Bell Pottinger PR firm based in the UK was caught on tape, as part of a sting operation run by British media, boasting that they had in fact written a key speech of then President Mahinda Rajapaksa to the United Nations in 2010, a year after the brutal end of the war. The speech was an interesting one eighteen months after the end of the war, calling for the re-evaluation of the Geneva Conventions, a tacit acknowledgement that the government’s victory came at the cost of compliance around established international humanitarian law as well as rules of engagement. Issues concerning accountability persist, and will endure for many years, if not decades to come. The speech by Bell Pottinger, the company’s Chairperson at the time David Wilson notes, was chosen over a version Sri Lanka’s own Foreign Ministry had drafted. A BBC article published at the time also noted that in 2010 alone, the Sri Lankan government paid the company 4.7 million dollars for what is euphemistically called ‘reputation management’.

The Rajapaksa regime needed blood washed off its hands. Bell Pottinger was, at the time, was good at doing just this and for many other brutal, violent regimes around the world, through a range of methods that included the strategic injection of content online in such a manner, over time, that Googling for information around a sensitive keyword, topic, person, place or issue would result only in content favourable to the government being presented first. In this way, inconvenient truths published online (at a time when domestic mainstream media was brutally suppressed) were buried from easy discovery, and over time, eventually lost under a mountain of content which was built around, or organically grew from what was initially seeded. How the company eventually faced bankruptcy and public shame last year for its insidious work in South Africa which the Public Relations and Communications Association (PRCA) of Britain said likely contributed to the inflammation of racial discord is, thankfully, not a story it was able to bury.

And yet, though the company is gone, the tactics have evolved, and how. What’s happening in the world of political communication today makes what Bell Pottinger did and stood for look positively benign or quaint. Bell Pottinger’s foray into the construction of alternative realities and manipulation of facts in South Africa dabbled with digital propaganda, more commonly known as fake news. The phenomenon involves the production, at great scale and speed, material that drowns out other narratives, or is directed to attack those who promote, produce and disseminate inconvenient truths. Either way, the aim is to create a tsunami of content partial to a specific narrative that in turn is favourable to the party that has invested in its production. The net gain, which can range from greater political clout or profit to the demise of competitors and opponents, outweighs the large expense to produce and sustain this sort of media operation.

Enter Cambridge Analytica. What Bell Pottinger did in South Africa, Cambridge Analytica did better for its client, Uhuru Kenyatta, in Kenya’s 2017 general elections, marred by widespread partisan and communal violence leading to many deaths. Kenyatta is the head of the Jubilee Party, which now rules the country. Last week’s revelations on Cambridge Analytica based on a sting operation conducted by the UK’s Channel 4 TV station captures Mark Turnbull, Managing Director of Cambridge Analytica, noting that his company had rebranded the entire Jubilee Party twice, written their manifesto, done two rounds surveys involving tens of thousands of people, written all the speeches and staged just about every element of Kenyatta’s campaign. The puppet master is revealed, and the sight is not pretty.

Kenyan writer Nanjala Nyabola writing to Al Jazeera this week made an important distinction in dealing with the fallout of Cambridge Analytica in Kenya, which has broader resonance in countries like Sri Lanka, where successive governments have tried their hand at reputation management and significant investments are already visible around the weaponisation of social media for political gain. Nyabola avers that precision is necessary in responding to the significant, dangerous and growing challenges to electoral processes and the timbre of democracy posed by entities like Cambridge Analytica, using information gleaned from Facebook and other companies. One part of the problem lies with what’s called data analytics – how companies like Facebook or Google are able to monetise information gathered by them, generated by us, in the aggregate. Companies like Cambridge Analytica are then able to use this information to base their content production and targeting on, which is at a level of sophistication that is mind-boggling, and as the senior management in the Channel 4 videos boast, clearly deliver and shape intended outcomes. The question will remain, despite assurances given by Facebook, whether greater regulatory oversight is needed around how large social media companies govern data generated by hundreds of millions of users, which in effect is to scrutinise in much greater detail how these global influence engines work. The focus Nyabola notes and I agree, should not be on the regulation of end user generation of content or use of social media but instead on what companies do with the data downstream, with third parties who are then free to retain and reuse this data as they see fit.

Nyabola flags that the other side of the problem is around political consulting by firms like Cambridge Analytica, which on the face of it isn’t illegal though extremely expensive. What usually happens though is that through progressive capture or enticement, clients are blinded to and campaigns get mired in tactics which justify anything as a means to a desired end. The problem in both Kenya and Sri Lanka is one of political culture, where elections are always zero-sum exercises, with a winner takes all approach fuelling partisan violence.  Late last year, I met with the Elections Commissioner to warn him of threats to Sri Lanka’s electoral architecture outside known risks and mitigation strategies. These range from highly technical cyber-attacks on critical infrastructure that disrupts and discredits results, to what I call the hacking of minds – content that’s appealing to and stokes the fears of swing voters, first time voters and in particular a demographic between 18-34, which in Sri Lanka today is around 15% of the electorate. There is not an insignificant amount of money spent on analysing the psychological, geographic, econometric and demographic makeup of our electorate. Underlying communal tensions, unaddressed by successive governments and indeed, exploited for expedient ends, feed into all this, since over and on social media, existing fears and anxieties of both majority and minority communities can be exploited or channelled in creative ways, often anchored to a political and partisan power dynamic.

To understand any of this, is power. Today’s politics go well beyond physical rallies, posters, the usual negative campaigning on TV and radio, spot ads, false covers, mugs, stickers or branded bric-a-brac. Social media mediates political opinion. Companies that own and operate social media platforms aren’t governed by domestic legislation. They are themselves struggling to accommodate the imperatives of making profit, the safety of their users, and maintaining their privacy. Lax governance and technical loopholes are exploited by companies like Cambridge Analytica – one of many others out there – which weaponise information voluntarily produced by individual users, often against them, without their knowledge. There are many implications around all this. In and for politics, we have entered a new era where data is more valuable than votes, an analyst able to crunch numbers is as important as a political mastermind able to broker coalitions of convenience, and any political communications specialist with expertise in social media is now in high demand because a younger vote base is both extremely important in the final count and extremely vulnerable to manipulation through viral, compelling, divisive content online.

How we interrogate this matters a great deal, because ultimately, it is not about Cambridge Analytica or Facebook we are talking about. It is about the quality, strength and vitality of our democracy and the dialogues that sustain it. This goes beyond party political lines. It is a question for all of polity and society to embrace, because the tools of political manipulation and their suave, mercenary agents know no loyalty or patriotism beyond the colour of money.


First published The Sunday Island, 25 March 2018.

The real evil

The women, they said, were alone as the men would be at prayers. Some commented on the youth and beauty of the women, noting how attacking them would be somehow pleasing to the eye. In other groups, an assorted array of knives was an invitation to bring whatever weapon possible to join a mob in cleansing an area of Muslims. Defiling Islam was common, and in some cases it seemed like a competition amongst those participating in the discussion as to how best to come up with insults that were the most heinous or violent. In a widely shared message reported to Facebook, the invitation was to kill all Muslim infants without sparing even one, as they were no different to dogs. And therein lies the rub. After six days, Facebook got back to the person who reported the post noting that it did not violate their community guidelines and policies against hate speech.

Social media platforms as we know them today are broken, badly. Facebook last week was called out by the United Nations in Myanmar as having a central role in fanning the flames of Islamophobia and hate. The situation is more complex. Governments, including those in Mynamar, the Philippines, India and most notably, Russia, are weaponising social media in various ways, using platforms, apps and services used the most daily as a means of communication and engagement against the very sections of society that trust in large part information received or shared on these platforms. This is akin to poisoning on an unprecedented scale, with the target not the body, but mind. And there are many actors competing to wrest control of public perception, with a view to informing their reactions and responses. In the middle of this melee are the social media companies themselves, notably, Facebook, which owns and controls in addition to the eponymous and almost ubiquitous platform, Facebook Messenger, WhatsApp and Instagram. Twitter, YouTube and a comparably smaller family of instant messaging apps like Viber or Telegram, which still have users in the hundreds of millions and are growing apace, and you have what the father of the world wide web, Sir Tim Berners-Lee called ‘platform power’, which is crushing the nature of the web as he conceived it, and wanted it to grow as. His concern is twofold – that Silicon Valley companies today are invisible yet ever-present interlocutors through their technology in how we all interact with the rest of society, and the world. Secondly, that given the sheer numbers online and on these privately controlled platforms, designed to maximise profit over all else, it is a challenge – to say the least – for companies to control their technology against being used for terrorism, the spread of hate or incitement to violence. Money, ethics, jurisdiction, monopoly, privacy, surveillance, oversight, control – there are many wars being fought simultaneously on and around these platforms, around the world.

Sri Lanka is now in the spotlight for just this reason. During the violence in Digana two weeks ago, the government decided to block social media they claimed was fanning the violence, including Facebook, Facebook Messenger, WhatsApp, Viber and Instagram. An immediate consequence of the block was the use of VPNs – apps or sites that easily by-passed the blocks. On Google, searches for popular VPNs over just the course of a week saw an astronomic rise, and VPNs entered popular, social discourse as a way to access what for many was their primary means of communication and engaging with their friends, peers and indeed, customers. Small businesses that rely on social media were badly hit. The telecoms companies that blocked social media sites, and later on, the most popular VPN services, also blocked popular discussion sites (in no way connected to the violence in Digana) and other random websites including blogs. All this was by no means unpopular. Somewhat simplistic surveys done by mainstream media suggested a high degree of public support for the social media blocks, in light of the violence and going by the explanations provided by government at the time. The blocks played into an older generation’s enduring fear – piggybacked on by government – that all social media is evil, all the time, and only serves to corrupt the minds of the young. In their incomprehension of how social media works, what it does and how it helps stem, prevent and combat violence, a blanket ban or block over social media remains one that many are partial to – just like the opposition to the Internet, the VHS recorder, TV and telephone in years past. The significant, multifaceted and growing ways through which social media help combat rumour, misinformation and shape public perceptions around democracy, governance, political participation and social mobilisation for good remain unexplored or often undervalued.

There is a far more sinister element also at play, increasingly evident as the blocks on social media continued long after the violence on the ground in Digana died down. The government, especially under Emergency Regulations, seemed overtly interested in engaging Facebook to stop the transmission of hate on its platforms, but covertly interested in negotiating ways through which dissent could also be controlled. Using Digana as a powerful, emotive reminder of what we as society should never again see, the government seeks to monitor and control the use of social media in particular, in ways yet to be publicly disclosed, and may never be fully revealed. True, governments have a legitimate purpose in ascertaining public mood and sentiment over social media, for obvious reasons related to governance, law and order and policymaking. But the constant refrain in Sri Lanka, including from Ministers in Government, is that it is the Ministry of Defence or the Army best positioned and capable of social media monitoring. This is dangerous for two reasons, at least. One, obviously, the Army isn’t really interested in human rights, the freedom of expression or privacy and the deep or dark State, dormant to date, now has the perfect vector through which to stamp its authority. Secondly, less evident, is that this government for the best of reasons is setting up the worst of templates – one ripe for abuse by more authoritarian, illiberal regimes in the future. Monitoring architectures, by their very nature, are turnkey solutions – which can be used for the benefit of society, or to severely restrict their rights. Who controls the architecture matters in a country without any constitutional or legal provisions to safeguard the right to privacy or oversight around surveillance.

Facebook, as a company, isn’t known for its ethics. For many years, data-driven and evidence-based arguments were presented to the company around the violence and hate produced, promoted, projected over its platform in Sri Lanka. It took Digana for the company to take serious notice. For years, Facebook was asked to strengthen its Sinhala language moderation in order to deal with the reporting of content. This wasn’t done. What is even more disturbing is that content explicitly, through direct translation and without any need for contextual awareness was in violation of community standards, passed muster – suggesting that those in charge of responding to user generated reports at Facebook who understood, or were Sinhalese, allowed their own personal bias and prejudice to take precedence over the company’s moderation guidelines. Facebook’s impunity regarding all this is indicative of Silicon Valley’s approach to the problem of hate writ large – profit first, public relations second, government satisfaction third, user capture fourth, ethics and rights – well, that’s for another day. Coupled with governments like ours who will use any excuse to control, contain and censor inconvenient narratives, and you have a perfect storm.

All this, of course, does nothing whatsoever to really address the root causes of the violence in Digana. Why did the STF attack innocent Muslim civilians? Why were the Police so impotent? Why hasn’t the government addressed racial hatred over three years? Why is a Buddhist monk, a central figure in the promotion of hate, openly seen with and beside the President in official tours abroad, even as he is wanted by the courts in Sri Lanka? Why isn’t the catastrophic failure of our intelligence services a matter of concern? Why did it take Digana for the government to wake up to the nature, volume, vector and sources of hate online, well-known, reported on, and flagged for years by civil society? What has any programme, policy or project on reconciliation done to address underlying communal and social grievances that give rise to this kind of violence? Why is there so much of hatred in those who are so young? Why is it that in Sri Lanka today, an accident, a drunken argument, a brush, nudge, poke or prod, a word or glance, a random action in sober or inebriated state, can immediately or days after, become a flashpoint for the worst communal violence, ignited by architects seeking political gain through chronic instability?

These are questions Facebook cannot answer, because they are our own creations. Social media wasn’t the cause of Digana, Gintota, Ampara or Aluthgama. These are two conversations here. One, around popular technologies and media, is emotive and helps conveniently gloss over the other more urgent, enduring and serious one, around the grievous failure of government. Our interest is perennially around the public optics of cosmetic solutions. Social media and Facebook today grab headlines because it’s easier than tackling the real issues that plague society and polity, post-war. The future if clear. We either fix and address them or face many Digana’s in the future.

Each one, worse than the other.


First published in The Sunday Island, 18 March 2018.

Black March

Unlike the anti-Muslim riots in Aluthgama four years ago, the horrible violence in Kandy, Digana and surrounding areas was covered in great detail by the mainstream media. As a consequence of dealing with the fallout of the violence over social media, some insights are worth sharing. This is particularly pertinent in light of the censorship of social media carried out under Emergency Regulations by the government, ostensibly with the intent of controlling, curtailing and containing the spread of content that incited violence or fomented hate. The public were also divided – with some noting that social media was the cause of, or certainly added to the violence, and other – like myself – noting that the myopic blocking of key platforms were extremely harmful on a number of fronts and set a terrible precedent for governments in the future to do as they saw fit to curtail information flows. But to this we shall arrive after some observations which mainstream media cannot report on, because their model of journalism isn’t linked to a deep-dive into, or the sifting of social media content.

The sheer volume of social media content generation during the violence was significant. Close to ten thousand tweets in Sinhala and English alone with the hashtag #digana, marking out the content as somehow anchored to what was going on in the area. There were dozens of pages on Facebook that were anchored to extremist Buddhist groups promoting falsehoods, vicious diatribes against the Muslim community and Islam, replete with memes and photography on top of which were often calls to protect Buddhism, congregate at a certain place or Temple to discuss and take action against threats to Buddhism, the accelerated birth rate of Muslims, continuing rumours and purported evidence of sterilisation pills sold or somehow smuggled by Muslims to be used against Sinhalese women, and language that suggested there were violent, invaders, alien, untrustworthy, hostile, ungrateful, ungracious and a community that needed to be taught a lesson or two. Importantly for a discussion on the merits of this content on social media and its influence on the violence over the week is the fact that a lot of this content was openly published, for years on end, by accounts, individuals and institutions who in many cases were openly named, and with contact details given. In other cases, the content was pegged to anonymous or pseudonymous accounts. There are dozens of videos still up on YouTube pegged to individuals and organisations promoting this line of thinking. Repeated reports to Facebook in particular have yielded no relief, since the company does not have the necessary resources to monitor the spread of hate on its platform in Sinhala. Another key development this week was in the form of around seven WhatsApp group invitations I got, with names indicative of the content and discussions they would have featured. It was not an option to enter these groups with my mobile number since I would either have been immediately targeted, kicked out or both. Screenshots sent by those who did infiltrate these groups reminded one of the stories now documented of pre-genocide Rwanda – photos of an assorted array of knives calling for all good men to rally around and deal with a problem, the targeting of women specifically, the highlighting of brick and mortar structures including mosques for destruction, the planning and plotting to destroy community symbols, the coordination of mobs, the fuelling of group think against community and religion, the sharing of videos, photos and other material that was trophy footage from individuals and groups who celebrated the wanton destruction and violence. There was also a link to a group on Telegram I received, and since the government also blocked Viber, it’s clear that instant messaging in general has become a primary vector in the communication of violent content leading up to and especially during riots.

The data on Google is also quite revealing. Over the period covering the height of rioting, the 6th to 8th of March, searches for ‘Molotov cocktails’ on Google indicate a sharp, significant increase in the Central Province. Searches for ‘How to make a petrol bomb’ over the same period shows a similar dramatic increase in Sri Lanka.

Finally, what we saw over the week was the weaponisation of social in an entirely new and unprecedented way with the advent of trolls and bots, adding fuel to rumour, misinformation and disinformation over Twitter in particular. With most leading social media platforms blocked, Twitter became a platform for the dissemination of information as well as a melting pot of ideas, updates, contestation, fact and fiction. Something interest, nay disturbing took place over the course of the week. As revealed by colleague, friend and researcher Raymond Serrato from Democracy Reporting International, who analysed well over fifteen thousand tweets over the course of the week, a few thousand accounts were created that went on to generate very high volumes of content on Digana. This initial production of information, fed into the Twitter ecosystem, was retweeted many times over, serving to mislead, misinform and often grossly simplify events to the benefit of a Sinhala-Buddhist perspective or narrative. These accounts featured fake names and fake photos designed to look like Sinhalese, Muslim and Tamil names. They targeted individuals like me and other senior journalists, as well as anyone on Twitter providing perspective to the rioting that held accountable Sri Lanka’s systemic racism, the silence or complicity of the Buddhist clergy, and the criminal nature of what was going on in and around Kandy.

In the middle of all this, and far better known and reported was catastrophic, shameful government inaction and impotence in the face of violence. Instead of attention and action, we got censorship. As noted by me in public over the course of the week, the blocking of social media directly increased distrust of and pushback to Government, contributed to international media headlines which have painted Sri Lanka in a very negative light, fuelled significant and growing concerns by foreign investors, severely impacted the operations of the civic media teams and professional journalists in Sri Lanka to respond to and know about ground conditions, severely and tragically curtailed the ability of victims of violence to make their voice and concerns heard and finally, emphatically did not contributed to a reduction in mob violence on the ground – despite what the Commander of the Army noted late into the week. This is because the use of VPNs was openly and first promoted by the very group whipping up the violence and hate, content continued to be exchanged over WhatsApp by a number of the groups, and pre-planned attacks, vectors of violence and targets, once mobs had assembled in an area, didn’t require further coordination or collaboration using social media – they knew what to do, how to do it, and where to go.

Early and sustained reports from the worst affected areas indicated that the Police did little or nothing and that they were grossly outnumbered by the mobs. There are dozens of videos around how the mobs roamed freely during curfew, and importantly, imposed under a State of Emergency. It is evident the government has no capacity whatsoever to monitor social media, and that its only approach to it is to shut it down completely no matter what the consequences. As reported by so many, until the weekend, neither the President nor Prime Minister had gone to Digana. Throughout the week, the Chief Prelates were silent. Episodic, inspiring stories of communities, Buddhist priests and ordinary citizens standing up to the violence, giving refuge, security and showing up in solidarity to mosques and shelters gave a glimmer of hope that even in the darkest of times, not all were consumed by hate. But fear and anxiety persist and grow, across the country, spanning the usual social, economic and political divisions – around what this violence truly highlights, and how it is controlled by forces that turn it on and dial it up with almost total impunity, at their whim.

The past week was a grim reminder that Sri Lanka is post-war, but yet not a country with or at peace. What lies ahead, for us all?


First published in The Sunday Island, 11 March 2018.