Digital does not mean you cannot touch it
Understanding the analogue footprint of digital technologies
Often, when we hear the word digital, we immediately think about something immaterial. This is paradoxical for at least two reasons. First, the origin of the word digital is the Latin digitus (finger), which later acquired the numerical and computerized sense because numerals under 10 were counted on fingers. Second, the digital world depends on an analogue stack. There is no internet without computers, cables, satellites, data centers, communication towers or energy grids. This is obvious, but too often the debates around digital government ignore this material dimension.
This was the case in a very interesting paper by Luciana Cingolani recently published in the academic journal Governance. The article Infrastructural state capacity in the digital age: what drives the performance of COVID-19 tracing apps? explores how the notion of infrastructural state power – a central concept in political science and development studies - needs to be reconsidered in the digital age.
According to Mann (1986; 1993) infrastructural power refers to “the capacity of the state, despotic or not, to actually penetrate civil society and implement logistically political decisions throughout the realm.” Soifer and Vom Hau (2008) identify three elements of such power: endowments (financial and bureaucratic resources), state spatial reach, and state legitimacy or acceptance. Digital technologies can disrupt traditional notions of state infrastructural power in at least three ways:
1. Digital technologies can bypass physical organizations and structures, reaching directly to citizens through a different set of instruments such as facial recognition systems, digital registries, or mobile services.
2. State-society interaction is transformed by the use of digital technologies to carry out much more targeted and granular interactions with citizens.
3. Digital technologies simultaneously enable reduced spatial reach with more targeted interactions at lower costs, suggesting that governments will substitute traditional presence with digital means.
By studying the emergence and adoption of COVID-19 tracing apps in 150 countries, Cingolani tests the variables that may determine state’s digital infrastructural power. She finds that financial endowments and spatial reach (measured by the presence of territorial bureaucracy), do not seem to affect the emergence of tracing apps. By contrast, bureaucratic endowments are associated with the emergence of these apps and legitimacy (measured by an indicator of trust in government) is positively correlated with their adoption. In other words, the key determinants of governments’ digital capacity seem to lie on the professionalism of their bureaucracies and the trust of their citizens.
These findings are interesting, although their explanatory power is probably compromised by the coarse nature of the variables and a comparison of very different country contexts. What is surprising, however, is that Cingolani does not discuss what I thought was an interesting result. Table 3 shows that the control variable Mobile phone infrastructure is positively correlated (at a high degree of statistical significance) with user adoption rates of tracing apps. This is, I think, the only IT infrastructure variable included by the author in her model, and yet highlights that physical infrastructure is a necessary condition for governments’ digital capacity. States’ tech infrastructural reach – and therefore their ability to invest in and deploy digital infrastructures – is a key determinant in the development and take-up of digital governance tools.
Understanding the extent to which digital technologies may transform our lives requires a much deeper and more nuanced exploration of their infrastructural and physical conditions. For example, very little of the discussion around algorithmic governance devotes attention to how algorithms are shaped by their underlying technologies. A notable exception is the paper by Donald MacKenzie Material Signals: A Historical Sociology of High-Frequency Trading. This paper studies the algorithms involved in high-frequency trading and finds that they are the product of the present and past interactions of people, organizations, and machines with the algorithms themselves. The location of data centers and fiber-optic cables in North America affects the decisions of trading firms, since these infrastructural elements can affect the fraction of seconds in which information is received and processed by selling or buying algorithms, giving a commercial edge to those firms that get price information (signals) microseconds earlier than competitors. This may seem a trivial detail, but shapes how the algorithms operate and affects multi-million-dollar decisions around investments, infrastructure, and trading.
Most importantly, it shows that one cannot really think about digital governance and how it can affect governments’ capacity to make decisions and deliver services, nor how these will be received by citizens, if their analogue dimension is ignored. What are the physical investments required by digital governance tools? How do these infrastructures operate and how do they affect digital governance tools? What are the implications of their physical presence for the communities in which they are located? And for those served by the technologies? All these are thorny questions in search of brave researchers willing to tackle them.