From Big Data to Big Brother

Big data analytics and the internet, originally seen as tech miracles that would lift all our psychic and economic boats, have suddenly become dark holes into which civilization could plunge. The enthusiasm in a 2013 book — Big Data: A Revolution that Will Transform How We Live, Work, and Think — has been replaced by gloomy warnings that the Facebook/google era is creating crippled societies controlled and manipulated by powerful forces armed with algorithms and trillions of data points.

Thanks in part to the sensational claims surrounding the Cambridge Analytica scandal and the allegations levelled by Canadian Christopher Wylie and others, it is now received wisdom that this new digital world is filled with grave and unprecedented threats that require major government action to bring it under control.

How much of the alarm is based on real risks, hard evidence and solid science? Less than we have been led to believe. To a significant degree, we are now engulfed in another round of junk science. This marks the 20th anniversary of FP Comment’s annual Junk Science Week. Junk science occurs when scientific facts are distorted, risk is exaggerated and the science adapted and warped by politics and ideology to serve an underlying agenda.

It should now be clear that the great Cambridge Analytica episode looks more and more like an overblown balloon of bunkum, based on hyped claims that secretive power brokers somehow collected and manipulated Facebook data to swing the popular vote for Donald Trump in the U.S. and Brexit in the U.K.

A good starting point for a review of the myths behind the Cambridge/facebook threat is a new book, Outnumbered: From Facebook and Google to Fake News and Filter-bubbles — the Algorithms That Control Our Lives, by David Sumpter, a British mathematician based in Sweden. In one chapter, aptly titled Cambridge Hyperbolytica (excerpted today in FP Comment), Sumpter concludes that the great Facebook manipulation of the U.S. election could not have happened for one very good reason: the science behind the claim was faulty. It could not — and did not — work as claimed.

As Sumpter told me in an email, Cambridge Analytica CEO Alexander Nix and his purported whistleblower, the outspoken media star Christopher Wylie, are victims “of their own hyperbole and incompetence. They exaggerated what they could do, appear to have fairly poor data practices and Nix was opportunistic.”

It would also appear that Nix and others knew their claims to have manipulated the U.S. electorate were untrue. Another Cambridge official said Facebook data was never used because of its “effective uselessness.”

Wylie, now an international celebrity, continues to make headlines, including via hyperbolic and disputed claims last month before a House of Commons committee in Ottawa. He claimed the Facebook/cambridge data scandal is the “canary in the coal mine” when it comes to shadowy use of citizens’ data to try and sway elections.

Despite many reasons to question Wylie’s credibility as a source of intelligence, and the likelihood that Cambridge Analytica is a fake big-data scandal, the company’s name has now become a launch pad for calls for governments everywhere to intervene.

Parliamentary and Congressional committees are holding hearings, media are fulminating and assorted academic interventionists are cranking out papers calling for national data strategies across the economy and the political system. Big data needs some nudging from Big Brother.

A recent example of this trend in the academic world is a new collection of essays from the Canadian Centre for International Governance Innovation (CIGI) called Data Governance in the Digital Age. In it, the Cambridge/facebook episode is portrayed as an illustration of the problem: “surveillance capitalism” which is part of the bigger problem known as the “surveillance economy.”

Through more than a dozen papers, CIGI’S big-data thinkers come to a consensus: We need national data strategies, with the government in control. “In a surveillance economy and society … effective democratic oversight of both the state and economic actors is essential to resolving the tension between the threats posed by such surveillance and the necessary role of surveillance in enabling the datadriven economy.”

One essayist writes: “Data has been called the new oil. Why? Because it creates class warfare. Those with it (‘the haves’) benefit from its enormous value and, ultimately, control the agenda.” The government should therefore “work to ensure that the value of data in our cities profits citizens and not private interests.” Canada “must develop a national policy that mandates two things: data ownership and open architecture.”

All this is a reminder of an old debate in economics: whether it is possible for government to manipulate economic activity and behaviour using planning and data. The dominant conclusion of the 20th century after the massive failures of planned economies has been that it is not.

But in this era of big data there is talk of new possibilities. Back in 1993, the two economists who brought us the book Towards a New Socialism, argued in a paper that with modern computer power and “a careful choice of efficient algorithms” there is a growing opportunity for “reassertion of the classic Marxian argument for economic calculation.”

As in the past, the doubtful science of big data is being used to promote another agenda.