Tag Archives: algorithm

Moscow’s pandemic in the not-so-smart city. Part 2: Social monitoring and prostheses of control

In the second part of our book chapter, on which these posts are based, we turn to a description of the steps the Moscow government made at the beginning of the pandemic. The Russian pandemic ‘began’ in Moscow. Perhaps out of admission of the Russian state’s low capacity, perhaps out of cynicism, the federal centre ‘delegated’ to the regions responsibility for measures against Covid. Certainly part of the logic was to preserve the austerity politics the centre has pursued for some time; from the perspective of today, Russian fiscal expansion to cope with Covid has been among the smallest proportions of GDP in the developed world.  For example, one of the early ‘responses’ was a capped 8% loan to a limited groups of SMEs – hardly comparable to the significant support in some other complex and service-orientated economies. Now problems are emerging with a promised 3% loan cap for small businesses. The GDP deficit never even reached 5% during the pandemic and is now back in surplus (compare this to the immediate fiscal response in Anglo-Saxon countries that was nearly 10% of GDP). The refusal to make use of fiscal ‘space’ by the government in 2020-21 is not only criminal, but economically illiterate. Poverty rates have risen by around 10%. For the first time I’ve encountered beggars in even the smallest towns.

The implementation of QR codes made Moscow’s response in 2020 famous, but the first use of these was actually in Kazan. From 15 April 2020, QR codes in Moscow were required for internal movement. In reality though, the limits on mobility without the use of codes was hazy – taking out rubbish, walking pets 100m from residence all allowed Muscovites to test the practical limits. Of note is the enrolment as ‘police’ of Moscow’s taxi drivers – now required to check QR codes of their passengers, as were turnstile controllers employed by Moscow Transit Authority. By the end of April, a ‘social monitoring’ app was imposed on the infected to enforce ‘home quarantine’.  The ‘mask-glove regime’ was introduced in May 2020 and is absurdly still in force as of November 2021, although since summer 2021 I have rarely seen anyone in transit wearing gloves. Interestingly, the full regime seems to be only enforced in one of the higher-end shopping chains.

In the chapter version of our research, we reflect on the Moscow authorities’ attempt to emulate China’s fangkong system of public health surveillance. We also contrast the Russian ‘Social Monitoring’ system with Singapore’s horizontal TraceTogether system, and Seoul’s use of mobile and banking app data to track individuals. Arguably, Moscow’s system was most similar to China’s Alipay Health Code, although the latter was both more sophisticated and less transparent in operation. The Moscow Social Monitoring app was plagued by bugs and ‘dirty’ code, seemingly slapped together in just a few weeks and was a far cry from the initial promises by City Hall that a system like Seoul’s personal data aggregation was planned.

What can we learn from re-reading Deleuze’s 1990 ‘Control Society’ essay? This is a post-institutional look at control. Deleuze pessimistically sees Foucault’s governmentalizing (the ‘sovereign’ person learns to love the policing of herself) as transient. Using the metaphor of a corporation, Deleuze foresees control as continual adjustment via the codifed ‘dividualization’ of information about people (did he read Marilyn Strathern on the quiet?). Deleuze anticipates how technology can create a kind of double of an individual based on her data trail – and that this trail can enable control mechanisms via real-time exploitation of data – what Deleuze and Guattari call a ‘universal modulation’. Presciently, Deleuze also sees this logic as destroying the rationale for traditional state institutions (why do we need a hospital or public health system based on evaluation of evidence and research, when an algorithm can be pre-programmed to optimise health outcomes on the basis of a simple risk calculation? Do you take a particular diabetes or asthma drug, asks the algo? Then your mobility card is automatically blocked when R-reproduction reaches a certain point). Judgement is suspended on the basis of simple big-data calculations of relative risk.

But what does the imperfect implementation of this logic look like in the case of Moscow in 2020? On 15th April, QR codes for essential journeys were supposed to be available for download from the City Hall website, but immediately the site was shown to crash repeatedly under the demand. The embedding of QR codes into the existing digital infrastructure of ride-hailing apps also failed ‘due to the providers not yet having found a way of doing this’. What we observe is an interesting example of improvisation based on using old tech for new purposes – a 2015 app from the Transit Authority was repurposed for use in hand-held tablets by law-enforcement personnel. Enormous queues at Metro stations formed, giving a fundamental insight into technology-led surveillance policy, one that is not so different from elsewhere: almost no ‘real-world’ contingencies were really considered. For example, what was the lone Rosguard officer supposed to do with someone whose code didn’t work? It became clear that only the first link in the chain was considered. The sight of law enforcement telephoning for advice and their superiors having no response was repeated over and over.

In late April 2020 the QR was finally linked to individualized travel cards (noting that these are a small minority of cards in use – the majority being anonymous cards that you can load with credit). Despite this advance, police did not change their protocol – they still required a barcode or QR and did not have the capacity to read the travel card’s Covid validity. The much admired private sector was no better. The Yandex taxi app is a sophisticated piece of flexible software allowing you to tell the aggregator whether your driver is wearing a seatbelt, has good taste in music, etc, and allows you to leave a tip or not. In our chapter, we discuss the potential for this IT giant to have assisted City Hall’s control society. It did aggregate its own data about Covid risk using geolocation data and published it publicly. But this was not integrated into the City’s response and clearly City Hall feared the dilution of central control.  In the next part of our chapter and in the next blog post tomorrow, we discuss the relevance of theories of ‘surveillant assemblage’.