In this issue of Other Worlds, Ana Isabel Carvalho and Ricardo Lafuente, founders of the F/LOSS-oriented design studio Manufactura Independente, rigorously dissect Stayaway, the contact tracing app launched in Portugal during the Covid pandemic. By analysing both the technical limitations of the app and its misleading visual metaphors, they prove that a grounded criticism of public health monitoring systems is possible. Reconsidering the app after its demise, they warn us against a novelty bias in the public discourse, which is too focused on promotion and expectations rather than consequences and effects.
The persistence of the pandemic has motivated many proposals for digital solutions to fight Covid. Contact tracing apps are one of such endeavours: by measuring how close smartphones are and how long they stay so, their purpose is to warn people of potentially dangerous contacts with infected people. As part of our work in Portugal’s digital rights association D3, we’ve dedicated a good amount of time to the analysis of the evolution and failure of Portugal’s contact tracing app, Stayaway Covid. As a result, we published a lengthy public report highlighting the shortcomings of this experiment. In this brief text, we offer a design perspective into the app’s interface and communication strategy, in order to outline the less evident pitfalls in rapidly deploying untested technologies as public health measures.
Stayaway embodied the hope that we could develop an effective digital weapon to ward off the virus. The promise was enticing: an app to keep you safe without being visually or physically obtrusive. No barriers needed, no special equipment required. Expanding on the power of a device “we all have” (a common statement that conveniently overlooks social inequalities and minoritarian choices), we could be just an install away from safety.
However, when applied, this solution exhibited multiple shortcomings. Its dependence on Bluetooth for measuring distances – a purpose that the Bluetooth protocol was not designed for – eventually revealed significant inaccuracies. Many people are already familiar with these, as broken connections or frequent pairing problems with audio speakers are to this day still common. Even with perfect usage, detection might not happen: one “positive” phone centimeters away from a “negative” one might not trigger the activation of the latter, due to wildly varying environmental factors such as reflective wall materials or signal interference from other devices.
The use of the app was especially encouraged in crowded spaces and public transport, where transmission risk is high. Practical studies confirmed that many false positives and negatives were to be expected, particularly in the context of large crowds where detection accuracy drops to 10%. It was also discovered that the app became virtually ineffective inside large vehicles with metal frames such as trams and buses. Destined to fail in its most crucial usage scenarios, contact tracing apps were promoted as our only hope, reminding us of the Silicon Valley mantra Fake it till you make it.
At no point did the app’s proponents ever acknowledge that the detection method was, to be generous, not fully reliable. Public statements were made blaming doctors, users or the health ministry for the app’s failure to have any significant effect. The app itself, its creators insisted, had no issue and worked properly. At the same time, glossy TV ads and public transport posters appealed to the general public to install Stayaway, showcasing relieved citizens with floating shields around them.
Official full-page ad for Stayaway Covid. It reads “The app which alerts if you were at risk of contagion. Take care of yourself. Take care of everyone.” Source: stayawaycovid.pt.
The visual metaphor of the floating shields has clear implications for the app. It promises protection, when it could only provide detection. It proposes peace of mind, even though it can’t reassure you that you’re not infected – only that it didn’t detect prolonged contact with an infected person. Both of these aspects of public communication point toward a projected optimism around the app’s capabilities and potential to fight Covid. Besides its practical purpose, Stayaway also was presented (and thus appeared) as a concrete source of hope to end the pandemic; public communication followed: it emphasised this aspect of faith in a digital device that could present a true solution if only widely adopted. The reality is that most of these implicit promises were inflated to keep up the sentiment of hope in the app’s effectiveness, which indirectly reinforced the perception of the government’s preparedness.
Stayaway’s user interface takes this encouraging view further, adopting the visual metaphor of traffic lights used to represent either a positive or negative status. There are only two possible states for the app (contact detected or not) with no context given to the user, no probability score, no warnings about signal instability, and very little useful advice on how to act on the reported results. When no contact has occurred, a reassuring message saying no contacts were detected is presented in the color green; in the case of detection, the user is brought to a yellow screen and urged to enact preventive measures.
Application screenshots for both “no contact detected” and “contact detected” states. Source: stayawaycovid.pt.
The use of yellow is appropriate: while it reassures the user that it’s not reason for immediate alarm – hence the avoidance of the color red –, it conveys the feeling of a situation that warrants caution. The use of green, however, contributes to a false sense of security since it establishes a notion of safety, when the reality is that contact might have occurred without the app detecting it. By not mentioning this possibility and presenting a green light the app ended up misleading the Prime Minister himself, who publicly remarked “I use [Stayaway], and it is with great satisfaction that, every morning, I have verified that until today I have not been close to anyone that could be a contact risk”. Of course, Stayaway can only report whether or not a contact was detected and cannot be read as a general assessment of your status, but the green light clearly encourages this legitimate (but dangerously wrong) interpretation.
It is significant that the app was mainly presented as a source of hope for society and not just as a public health device. This positioning was made concrete in its advertising, public messaging and UI design choices: Stayaway appears as something that can protect you and reassure you that things are okay, despite being able to guarantee neither. The privacy side-effects of employing a large-scale networked system as a response to the pandemic are also left out of the picture. It is then appropriate to ask what kind of social purpose does the misleadingly positive view of the app’s capabilities bring. And, more generally, we should ask whether exaggerated optimism is a good strategy to ensure public trust, particularly when reality cannot deliver.
We can find examples of apps that are, by design, clear about their limitations. For example, menstrual cycle tracking apps provide the users an estimate of fertile and non-fertile periods. Since cycles aren’t regular, such an app can never precisely state a yes/no result, but rather an approximation based on statistical data, an average value that the user can take into consideration but that isn’t a definitive result. And even though contact tracing apps operate on the same nebulous area of averages and approximations, furthermore based on unrealistic premises – everyone should have the app installed, everyone should carry their phone at all times, perfect physical conditions are always ensured for Bluetooth signal to propagate –, they present the user with a clear-cut binary status.
Still from the official promo video of Stayaway Covid. Source: stayawaycovid.pt.
Staying with the women’s health example, we are reminded of calendar-based methods of contraception, which propose a simple week-based calculation to determine fertile periods; they are now widely considered unreliable and reflecting an excessively simplistic view of people’s bodies. Likewise, contact tracing apps attempt to reduce a complex and nuanced problem to the premises of a simple framework that is weakly related to the actual problem. In broad strokes, this can be connected to critic Evgeny Morozov’s notion of technological solutionism: a fascination with technical possibilities leading to the application of digital technology to complex problems using excessively simple conceptual frameworks that do not address the problem’s true complexity.
The solutionist approach is also eager to restrict analysis to cold numbers. The app’s supposed efficacy was showcased through download counts and the global number of contacts detected. Results and consequences that cannot be easily quantified, such as the stress of receiving a false positive or the risk of going on with life after a false negative, were not considered at all.
[M]any new technologies and their products have entered the public sphere in a cloud of hope, imagination, and anticipation. In many cases these hopes were to begin with fictional, rather than real; even in the best of circumstances they were vastly exaggerated. Discussion focused largely on individuals, whether users or workers, and promised an easier life with liberation from toil and drudgery. Discourse never seemed to focus on the effects of the use of the same device by a large number of people, nor was there any focus on the organizational and industrial implications of the new technologies, other than in the vaguest of terms.
– Ursula M. Franklin, The Real World of Technology
Contrary to the media attention given to the promotion of the app, there was no public “post-mortem”, no follow-up debate, no analysis of what happened after it became clear that Stayaway didn’t meet its goals. This was a lost chance to further public discourse about the advantages and shortcomings of public health measures mediated by apps. The later introduction of digital vaccination certificates worldwide makes it clear that the solutionist doctrine has firmly made its way into our public health infrastructure.
The interface and communication details outlined above hint towards a design approach that is focused on projecting a sense of safety and control that these apps cannot meet in reality. The feeling of hope is treated as something that can be managed and furthered by the use of projected positive attitudes that disregard any minimum expectations for the app (“if it saved a life, it was worth it”), while dismissing any possible side-effects of the app’s workings and limitations, such as the wasteful stress inflicted to someone who self-isolates in response to a false positive. In the end, the general goal wasn’t so much to make an app that works, but to spread the belief that it does, even when it doesn’t. The realisation that this misleading discourse (even if well-intentioned) is the official line of public health policy warrants a debate on the role of novel shiny tech solutions in regards to the challenges society faces.
Still from the official promo video of Stayaway Covid. Source: stayawaycovid.pt.
Ultimately, these are design choices about how public health practice and policy are formulated. And even if designers haven’t had a seat at the table yet, it’s not late to join the discussion about the design of digital health solutions. Design discourse ought to play a role in the critique of these solutionist approaches, applying its own specialised lens to the understanding of the fine workings and the nuanced effects of digital pandemic response solutions. Calls still abound for “design thinking” hackathons focusing on novel concepts to fight Covid. We hear much less engagement with the actual design concepts that are being employed at scale, conceived (and designed) solely by technocratic task-forces and technical specialists – the concern voiced by the Institute for Technology in the Public Interest was a notable exception to this near silence. The solutionist wave has cast aside any field of study that is not mainly technical, like social sciences or concept design. If we don’t press for their seat at the table, it’s likely that they won’t earn it back in the face of an all-encompassing, technoaccelerationist app-based agenda.
Other Worlds is a shapeshifting journal for design research, criticism and transformation. Other Worlds (OW) aims at making the social, political, cultural and technical complexities surrounding design practices legible and, thus, mutable.
OW hosts articles, interviews, short essays and all the cultural production that doesn’t fit neither the fast-paced, volatile design media promotional machine nor the necessarily slow and lengthy process of scholarly publishing. In this way, we hope to address urgent issues, without sacrificing rigour and depth.
OW is maintained by the Center for Other Worlds (COW), at Lusófona University, Portugal. COW focuses on the development of perspectives that aren’t dominant nor imposed by the design discipline, through criticism, speculation and collaboration with various disciplines such as curating, architecture, visual arts, ecology and political theory, having in design an unifying element but rejecting hierarchies between them.
Editorial Board: Silvio Lorusso (editor), Luiza Prado, Francisco Laranjo, Luís Alegre, Rita Carvalho, Patrícia Cativo, Hugo Barata
More information can be found here.