With the collapse of cryptocurrency exchange FTX and troubles at Twitter, discussions on day two of this year’s Bristol Festival of Economics were particularly timely. Big data, social media and behemoth technology firms are here to stay – but calls for their regulation grow louder.
It was certainly auspicious timing to be holding a festival of economics in the wake of the UK's biggest political and economic crisis for a generation. Indeed, during the second day of the event, the debate was as sharp and direct as the budgetary cleaver that was soon to be wielded by the Chancellor of the Exchequer. The topics hewed around regulation of Big Tech, bubble spotting and the future of big data.
The day began with a lively discussion chaired by the inimitable Rory Cellan-Jones, a former BBC technology correspondent, who set the tone by remarking that we are ‘a slave to Big Tech’ for better or worse.
The giant technology companies – collectively known by the acronym GAFAM: Google (Alphabet), Apple, Facebook (Meta), Amazon and Microsoft – have a market capitalisation of $6.7 trillion, 7% of the global total.
Figure 1: 'Big tech' companies by market capitalisation
Source: Refinitiv Datastream
The panel explored how best to rein in these tech giants, to restore competition in the market and to ensure that their content complies with national broadcast standards.
Kate Bevan (a former journalist, now at Infosys Knowledge Institute) set out her three principles of regulation. First, it must be forward-thinking and try to anticipate technological changes, rather than using existing regulation and ‘retrofitting’ it. Second, it must be scientific and evidence-based rather than knee-jerk. And finally, it should consider how to regulate firms whose bosses, like Elon Musk of Tesla and, more recently, Twitter, appear to be opposed to regulation.
That is easier said than done. Tommaso Valletti (Imperial College London) pointed out that during his tenure as chief competition economist at the European Commission between 2016 and 2019, he failed in nearly all his antitrust cases against Big Tech.
Of the 1,000 or so acquisitions by GAFAM, fewer than 97% have been investigated, he remarked. It is a David and Goliath battle: the UK’s Competition and Markets Authority (CMA) has just 100 staff and the European Union (EU) just 35 PhD competition economists, against a battalion of lawyers and economists defending the tech firms.
Kimberley Long (Asia editor of The Banker magazine) pointed out that regulation of Big Tech has been more robust in Asia, particularly because it threatens political institutions. Ant Group, a sprawling financial conglomerate, had its proposed dual-listing in Shanghai and Hong Kong cancelled in November 2021 and its chief executive, Jack Ma, mysteriously disappeared for two months.
The message was clear: tech firms that thrive in China must do so with the consent of the state. Yet that has conversely given rise to super-apps like WeChat, which provide a universal portal for almost every daily digital task – and of course the accompanying censorship and surveillance by China's communist party.
The broad consensus in the panel was surprisingly echoed by the venture capitalist on the panel. Harry Destecroix (Science Creates Ventures) whose firm invests in early-stage deep tech – artificial intelligence (AI), blockchain, quantum computing – admitted that tech firms and their networks must be regulated.
He worried that nascent deep tech firms would be gobbled up by the insatiable tech behemoths, and suggested widening the National Security and Investment Act to protect them. Tommaso went further by proposing that GAFAM be banned from all acquisitions regardless of scope or size.
Although all panellists agreed on tech's dominance, their views on how much consumers care differed widely. Harry pointed out that history has a long record of knee-jerk political reactions that held back technological innovation: genetically modified organisms; nuclear power and now AI. GDPR, the EU data regulation, imposes costs that are substantial for small firms, he argued.
Rules to protect people from harmful content are no less tricky. Kate said that defining harm legally is difficult – particularly in the broad domain of social media – and the evidence for it is mixed. Regulating content and language more so: the comment ‘that's a lovely pussy’ could be related to something that is either indecent or entirely innocent, she remarked.
The festival's no PowerPoint rule was broken this year by the hosting of the British Academy's Keynes lecture in economics. In a session entitled ’Wage controversies’ Stephen Machin (director of the Centre for Economic Performance at the London School of Economics) produced a chart-packed tour-de-force of income and earnings dynamics in the UK.
Stephen remarked on the slowdown in real wage growth over the past 15 years and rising income inequality. He highlighted that the monopsony power of firms to set prevailing wage rates in small labour markets – such as the coal mining towns of yesteryear – still endures today with the NHS's dominance of employment in some areas of the UK.
The rise of zero-hours contracts and self-employment, led in part by Big Tech's tenuous relationship with labour, along with a decrease in union membership is another important factor that Stephen explored.
A timely history lesson was delivered for the third session of the day with Diane Coyle (University of Cambridge) interviewing William Quinn and John Turner (both of Queen's University Belfast), the co-authors of Boom and Bust: A global history of financial bubbles.
The timing was perfect, with the previous week's implosion of FTX, a cryptocurrency intermediary, whose founder Sam Bankman-Fried was earlier this year being compared to the imperious John Pierpont Morgan, the great financier of America's gilded age. The embers of FTX's collapse are still smouldering, with Bitcoin having depreciated by 60% this year.
Figure 2: Price of Bitcoin ($)
Source: Refinitiv Datastream
Will and John set out their necessary conditions for a bubble to form, a triangle of: intense speculation, marketability of assets, and cheap and easy credit. A technological or political spark is required to create the bubble, which are themselves difficult to predict but can easily be observed ex post.
Cryptocurrencies, they said, exhibit all the symptoms of a bubble: low interest rates, frantic buying and selling, and price manipulation by bad actors. As Diane remarked, ‘whenever you see adverts on the Tube, you know you're in trouble’.
The authors gave examples of their favourite bubbles. John talked the audience through the railway mania of 1843-48 and the lesser-known British bicycle bubble of the 1890s. Both cases provided examples of the benefits of bubbles: an extensive, albeit sub-optimally planned, train network still relied on today; and the engineering know-how that helped to propel Birmingham to the forefront of automotive manufacturing decades later.
Will’s favourite – the 1990s stock market and housing bubble in Japan – was a symbiosis of media speculation and murky back-room dealing where massive stock market bets were made based on prognostications from séances.
The discussion moved to what policy-makers can do to prevent bubbles inflating and to mollify the effects of their bursting. John reckoned that the plumbing of the financial system encourages too much speculation, which inherently results in bubbles forming.
Although central bankers; remit is famously to ‘take away the punch bowl just as the party gets going’, they so often preside over cleaning up the mess alone after everyone has trashed the house.
The final session of the day took on the future of big data and the growing role that uses and misuses of data play in our day-to-day lives. ‘Data is infinite, but our attention is not’, pointed out Dénes Csala (Economics Observatory), and so there are two ways to navigate the data world, via algorithms or visualisation.
Arthur Turrell (Office for National Statistics, ONS) pointed out that nearly every institution, be it firms and their algorithmic recommendation systems or governments using NHS health information rely on data to inform their decision-making. People want data and their insights ‘cheaper, faster, better’ but they cannot solve every problem, he said.
Anna Powell-Smith (Centre for Public Data) whose organisation works with government departments, said that data-based decision-making needs to be better seeded in government departments. To that end, the ONS is giving civil servants a crash-course in the principles of data science and embedding them across the civil service.
Increasing amounts of private data are being made available for public good, from anonymised mobile phone data to aid transport planning, or shipping data to monitor supply chains, to the processing of text data from websites. Yet Arthur cautioned that a lot of pre-processing of the data has to happen to make it comparable to national statistics and suitable as a policy-making tool.
Tiziana Alocci (Market Cafe Magazine) said that ‘having data is not enough’. She implored data practitioners to be more thorough, and ensure that people understand how data are collected, measured and presented. Users should be sceptical of the data, explore their potential biases and be aware of the data that are not collected and ask why.
To that end, the panellists' wish-list of data included a social-media-based study to understand social mobility in the UK (mimicking the research of Raj Chetty in the United States); data on our tricky-to-measure online behaviour and personas; and a time-use survey for the UK.
With that, our time on day two of the festival of economics was up. But the lessons will be ruminated over for some time to come.