Aarogya Setu won’t be effective without access equity

Why govt must address the question of access inequity before making mobile apps mandatory during COVID-19

With the app being made mandatory for workplaces and reports suggesting it will be used to access public/private spaces, the risks of exclusion are significant

Representative image. (Credit: DH Photo)

On May Day, the Ministry of Home Affairs issued guidelines pertaining to the extension of the lockdown. As part of these guidelines, the section which included instructions for workplaces stated that the Aarogya Setu app would be mandatory for all employees irrespective of whether it was a public-sector or private-sector operation.

Moreover, the responsibility of ensuring compliance was placed on the “head” of the organisation. It remains unclear how the penal provisions referenced in the document are applicable in this case, whether the liability clause falls on the head of the organisation or on an individual employee. 

We have since also seen further expansion of this mandate, with the Noida Police announcing penalties for anyone without the app installed on their smartphones. The Bengaluru Police too published a procedure to return to Karnataka from other states, which states that 'citizens have to mandatorily install Aarogya Setu on their phones henceforth.' The government has also announced an interactive voice response system-based solution under the 'Aarogya Setu’ branding. However, technologically it remains a different solution mainly providing a mechanism to self-report. It does not address contact tracing.

The issues

Several concerns have been raised about the implications of such an approach on multiple fronts. Privacy, and the risk of its evolution into a vehicle for mass surveillance, has been the most discussed one. Then there is the security aspect, and the potential information security risks to individual users as well as a large centralised database of citizen data. There is also the legal aspect, that is, whether the National Disaster Management Act confers the necessary powers on the government to make it mandatory, and the lack of a data protection legislation. Technological aspects have to do with the efficacy of contact tracing apps/algorithmic risk detection and the associated issues around false positives and false negatives. Finally, there is the transparency question: Opaque processes and the fact that the code has not been made open-source yet is worrisome. 

The Ada Lovelace Institute has published a rapid review of technological responses to the pandemic titled, Exit through the App Store, which warns of the risks of “rushed deployment of technological solutions without credible supporting evidence and independent oversight.”

On equity

An aspect which is under explored is equity, or the lack of it. In designing public policy, equity is a crucial part of policy design. It deals with the social allocation of benefits and deals with the questions of 'who pays' and 'who benefits'. In the book Policy Paradox, Deborah Stone lists three dimensions and eight issues and associated dilemmas with each distribution method. Ultimately, this is a complex undertaking and no matter what criteria is for distribution, some group or the other will feel that they have been left behind by the policy.

In this context, it is important to look at a mandatory national level-app from the lens of who might benefit and who may pay a cost by facing exclusion. It follows that internet access, smartphone ownership with the required feature-set become important criteria for the distribution of this 'benefit'. Note that they are certainly not the only criteria, but considering them helps establish a baseline. One must also consider that internet sharing is unlikely to work well for this use-case, and that happens more than we realise. And that even a significant number of people who have internet access do not have the luxury of staying connected all the time.

Estimates suggest that there are around 450 million smartphones in India, but there is limited data by State/Union Territory or divided by rural and urban settlements. A useful proxy under these circumstances is data on the number of internet connections per 100 people, which TRAI publishes periodically. There have also been statements from analysts at CCS Insight and Counterpoint Research who estimate that between 50 to 70 percent of devices/people will be ruled out of such a solution. Such data can be combined with census data divided by urban and rural areas to estimate what pan-India distribution will look like. The maps estimate what percentage of people by State/Union Territory, may not have access to an Internet + Bluetooth Low Energy-based solution, under three conditions. 

A) 50% of Internet Subscribers do not have access to a BLE-enabled phone

B) 40% of Internet Subscribers do not have access to a BLE-enabled phone

C) 30% of Internet Subscribers do not have access to a BLE-enabled phone

Note that the scales vary in each image. The numbers represent the estimate of people who may not have access to internet + Bluetooth Low Energy-based solution needed for most contact tracing apps

It is worth pointing out that while TRAI's data suggests that in urban areas there are more than 100 internet subscriptions per 100 people, it cannot and should not be assumed that every individual residing in an urban area owns a compatible smartphone. This is more likely to be the case in low-income neighbourhoods. The end result will be an increase in data blind-spots and a reduction in any perceived efficacy of the solution. Additionally, since the disease continues to be stigmatised, there are incentives for people to report incorrect information too.

And it is precisely here that the crux of the issue lies. With the app being made mandatory for every formally employed individual, and reports suggesting that it will be used as a 'pass' to access public/private spaces the risks of exclusion for those who are already vulnerable are significantly increased. A situation we can ill-afford as we try to rebuild from the economic costs the pandemic has already imposed. Also, with reverse-migration of migrants across the country underway, it is extremely important to index towards more equitable solutions, not less.

It is, therefore, necessary that before we impose such mandates on people, we first take steps to address the underlying imbalance in access in any upcoming stimulus or recovery package.

(Prateek Waghre is a research analyst at The Takshashila Institution)

Disclaimer: The views expressed above are the author’s own. They do not necessarily reflect the views of DH.