Thursday , August 16 2018
Home / Society / Big data BS

Big data BS

We are currently living through an even bigger tech bubble than the DotCom boom in the late 1990s.  Now as then, investors have quite literally gone insane – throwing cash at companies that have never made a cent and look for all the world like they will never make a cent.  Some tech – solar roadways and hyperloops for example – defies the laws of physics.  Other tech – like carbon capture and storage, flying cars and rockets to mars – is so prohibitively expensive that it will never be more than vanity projects for a handful of Godzillionaires who never got around to growing up.  Nevertheless, in a world awash with central bank stimulus even the impossible looks attractive to investors with nowhere else to go.

It is in this febrile environment that the worlds of tech and politics have collided in the shape of the Cambridge Analytica scandal.  In brief, a British tech company unlawfully accessed perhaps 50 million personal profiles from Facebook, and used these to inform and target political messages on behalf of the US Trump and UK Vote Leave campaigns in 2016.

Politically, this is highly embarrassing, since, by their own admission, Cambridge Analytica is guilty of exactly the crime for which Special Counsel Robert Mueller recently indicted the Russian Internet Research Agency… except that Cambridge Analytica did it more effectively and on a scale that the Russian troll farm could only dream about.

To the affluent classes, still traumatised from the discovery that the unwashed masses might have stopped listening to them, Cambridge Analytica is a gift insofar as it allows the now tarnished “Russia stole my vote” fantasy to continue.   In both the USA and the UK, neoliberal politicians, journalists and campaigners have jumped on the scandal as reason enough to nullify the Brexit referendum result and the Trump presidency.

Mass media have their axe to grind too.  It is no accident that traditional media across the political spectrum have gone after Facebook and Google rather than Cambridge Analytica.  After all, social media platforms – of which Facebook and Google’s YouTube are among the largest – are responsible for the plunging advertising revenue and consequent decline in journalistic standards of the mainstream media.  No question that might undermine the “social media is evil” narrative will be asked; still less answered.  It is not as if Facebook and YouTube users didn’t realise that the deal was that we get a free (in money terms) social media platform in exchange for their harvesting our data and selling it to advertisers (including political campaigns).  Mainstream media’s real gripe is that their platforms lack the necessary sophistication to harvest personal data in the same way.  That is, they would like to follow Cambridge Analytica’s lead, but they lack the means.

To the tech crowd, the Cambridge Analytica scandal is anything but sinister.  Rather, it is held up as a vindication of the power of big data… big data that they, too, can harvest… for a fee.  In the same way as an arms dealer views the aftermath of a drone strike as a marketing opportunity, the tech companies are busy incorporating the supposed successful evil genius of Cambridge Analytica into their 2018 sales brochures.

Missing from the media coverage, however, is what should be a key question: does it work?  After all, the tech sector is full of examples of things that don’t work; at least, not without massive state subsidies.  The question, though, is more or less dismissed.  Carole Cadwalladr, whose Observer report broke the story, raises it rhetorically:

“But does it actually work? One of the criticisms that has been levelled at my and others’ articles is that Cambridge Analytica’s ‘special sauce’ has been oversold. Is what it is doing any different from any other political consultancy?

In answer to this, Cadwalladr does not turn to tech experts in the university sector, but to an ex-employee of the company who claims that the $15m invested in the company in the wake of a series of pilots in some way proves that it works:

“I don’t think Mercer even cares if it ever makes any money. It’s the product of a billionaire spending huge amounts of money to build his own experimental science lab, to test what works, to find tiny slivers of influence that can tip an election. Robert Mercer did not invest in this firm until it ran a bunch of pilots – controlled trials. This is one of the smartest computer scientists in the world. He is not going to splash $15m on bullshit.”

This is to assume that tech billionaires like, say, Elon Musk, would never invest in something that didn’t work like, say, a hyperloop.  Mercer may not be primarily concerned with the money, but that hasn’t stopped him making it – earlier in the same article, Cadwalladr informs us that:

“It was with AggregateIQ that Vote Leave (the official Leave campaign) chose to spend £3.9m, more than half its official £7m campaign budget. As did three other affiliated Leave campaigns: BeLeave, Veterans for Britain and the Democratic Unionist party, spending a further £757,750.”

Just one referendum campaign returned more than a third of Mercer’s $15m investment.  We might reasonably assume that the Trump campaign reimbursed the rest, and that the company will have also enjoyed an income from the British 2017 general election and will currently be negotiating contracts for this year’s US mid-term elections.  That is to say, far from being unconcerned with the money, this looks like a highly profitable scam.

In the end, Cadwalladr’s report tells us only that Cambridge Analytica is a particularly nasty company that has some pretty unsavoury friends.  Nowhere do we find a shred of evidence to demonstrate that its algorithm actually worked; still less that it was effective enough to sway the outcome of any of the votes it was commissioned to influence.

An actual study of social media activity in the run up to both the Brexit and Trump votes, by Dr Julio Amador from Imperial College London points to a far more mundane reason for the defeat of both the Remain and Clinton campaigns – complacency:

“The most active users in the Leave group were shown to both engage in dialogue and provide information whereas people supporting Remain were more likely to just provide information to fellow users…

“We found that in the 12 hours before the polls closed there was a surge of 3,000 tweets using the hashtag #MakeAmericaGreatAgain (used by Mr Trump supporters) compared with only 1,400 for #I’mwithher (used by supporters of Hillary Clinton). We concluded that Trump supporters made greater efforts to win over voters on social media than Clinton supporters, right up until the polls closed.

“We also found a higher level of Twitter conversation about the election in swing states such as Florida and Pennsylvania.

“We can draw parallels between the level of engagement on Twitter in terms of volume and sentiment of tweets coming from Democrat supporters, to the level of Twitter engagement put out by the Remain campaign on the day of the EU Referendum vote.  In my view, these results show that supporters of Remain and Hillary Clinton could have done more to reach out to voters online.”

As was noted by commentators outside the mainstream media bubble – myself included – prior to both votes, the Remain and Clinton campaigns (both of which deployed the discredited “Project Fear” approach) were shambolic.  Like their supporters, they spent their energy and resources talking to fellow travellers while failing to notice that the political mood in their respective countries had changed.  Unlike their opponents, they made little attempt to reach out to those who did not support them; preferring instead to berate anyone who disagreed with them as a racist, a fascist, a moron or a deplorable.  Predictably, this had a far greater impact than Cambridge Analytica or any number of Russian troll farms in pushing people solidly into their opponents’ camps.

There is of course an unspoken class snobbery behind the faux concern about the use of data mining in the run up to key votes.  “We” educated people, after all, were not affected by the crude influence techniques deployed.  Only the little people – who refused to listen to the advice of their social highers and betters – succumbed.  It is all too reminiscent of the now infamous question posed by prosecutor Mervyn Griffith-Jones to the jury in the obscene publications trial of D.H. Lawrence’s Lady Chatterley’s Lover:

“Is it a book that you would even wish your wife or your servants to read?”

Unlike “us,” wives and servants, it would seem, are far more susceptible to the subliminal messages published by such unsavoury companies as Penguin Books, Cambridge Analytica and the Internet Research Agency.  The idea that wives and servants might be just as capable of distinguishing fact from fantasy simply does not get a look in.

The deceit is compounded by the failure to examine the data mining activities of “our” side.  It is highly implausible that Cambridge Analytica is the only data mining company to discover the “secret sauce” that allows its clients to win friends and influence people.  Dr Andrew Mullen, Senior Lecturer in International Relations and Politics at Northumbria University, points out that all sides were engaged in much the same activity:

“Utilizing big data mining – drawing upon canvassing returns, social media traffic, voter records and other sources (e.g. consumer databases about newspaper readership, shopping habits, etc.) – the Leave and Remain campaigns also used the internet and social media for intelligence gathering purposes to construct detailed and personalised voter profiles. Using analytics software – the Voter Identification and Contact System, developed in-house, in the case of the Leave campaign and NationBuilder in the case of the Remain campaign – with their in-built algorithms, the respective campaigns were able to assign each voter with scores (on a scale of one-to-five) based on how likely they were to vote and how likely they were to vote to Leave or Remain. This data was then used to compile target lists for digital advertising, door knocking (e.g. Get Out the Vote operations) and telephone contacts.”

Nor was this anything new – the 2012 Obama campaign is widely held to have brought big data into the mainstream of political campaigning.  Writing in the Washington Post, John Wagner reminds us that the Clinton campaign was doing the same:

“Ada is a complex computer algorithm that the campaign was prepared to publicly unveil after the election as its invisible guiding hand. Named for a female 19th-century mathematician — Ada, Countess of Lovelace — the algorithm was said to play a role in virtually every strategic decision Clinton aides made, including where and when to deploy the candidate and her battalion of surrogates and where to air television ads — as well as when it was safe to stay dark…

“The use of analytics by campaigns was hardly unprecedented. But Clinton aides were convinced their work, which was far more sophisticated than anything employed by President Obama or GOP nominee Mitt Romney in 2012, gave them a big strategic advantage over Trump.”

Nobody is expressing concerns about NationBuilder or Ada.  Nor, crucially, would they be doing if Remain and Clinton had been successful.  But there is no reason to believe that what Cambridge Analytica was doing was so far removed from what its opponents were up to that it can reasonably be said to have stolen the result.  The reality, is summed up by behavioural economist Dan Ariely, who quipped that:

“Big data is like teenage sex: everyone talks about it, nobody really knows how to do it, everyone thinks everyone else is doing it, so everyone claims they are doing it…”

It is more likely that Cambridge Analytica are the beneficiaries of a correlation similar to that enjoyed by advertising agency Saatchi and Saatchi in 1979.  Then, as now, the agency was accused of using underhand and morally dubious psychological profiling to unfairly sway voters in favour of Margaret Thatcher.  The reality, then as now is that the social, economic and political tectonic plates had shifted away from what had previously been viewed as the “centre ground.”  The unexpected result was that someone thought to be an extremist by her own party when she became leader in 1975 won a huge majority with which she shaped the political landscape for a generation.  As Ed Howker and Shiv Malik note:

“In 1979 Margaret Thatcher received a 16 percent swing in support from young people aged 18-35 – the baby boomers – significantly more than any other group… Without their support she wouldn’t have gained power.”

Without that generational shift, the efforts of Saatchi and Saatchi would have been as fruitless as those of the companies behind Ada and NationBuilder.  Instead, Saatchi and Saatchi went on to become one of the world’s leading advertising agencies.  No doubt Cambridge Analytica is hoping for a similar publicity boost – one that could prove highly lucrative at the very top of the “everything bubble.”

The bottom line, however, is that ordinary Americans voted for trump and ordinary Britons voted for Brexit.  We might not like their reasons for voting in this way, and we might not like the impact that their decisions will have on our way of life.  Nevertheless, their reasons were valid; and we ignore them at our peril.

As you made it to the end…

… you might consider supporting The Consciousness of Sheep.  There are four ways in which you could help me continue my work.  First – and easiest by far – please share and like this article on social media.  Second follow my page on Facebook.  Third, sign up for my monthly e-mail digest to ensure you do not miss my posts, and to stay up to date with news about Energy, Environment and Economy more broadly.  Fourth, if you enjoy reading my work and feel able, please leave a tip.

Check Also

Brexit is the backstop

Brexit is the backstop

As an exercise in creating unnecessary constitutional crises, Brexit is turning out to be a …