Blog

The 30 Per Cent Left Out of COVID-19 Big Data Analytics

On March 29th 2020, the Bangladesh Telecommunication Regulatory Commission announced that Bangladesh will use big data analysis to contain COVID-19 pandemic.

An article in the Daily Star reported that the pandemic will now be tracked through mobile phones. The idea is that all 16.62 crore mobile phone users of Bangladesh will receive an SMS asking them to call a hotline free of charge. The users will respond to this SMS and self-report basic information such as their age, their health condition, whether they have come in contact with anyone travelling from abroad or infected with coronavirus. This information will then be used by a2i under ICT Division to develop a map that will help take stock of the overall situation. 

This is surely a commendable step towards containing the pandemic and an opportune moment to utilize technology for public welfare. However, the question is how many people will be able to participate in this initiative? The success of effective big data analysis depends on how comprehensive and representative the data is. 

Fortunately for us, mobile usage is very high in Bangladesh. You and I will certainly support this initiative by responding to this SMS. There is a group of mobile phone users, however, who cannot read SMS on their phones. In a recent national rural survey on Digital Literacy, BIGD found that almost 30% of rural mobile phone users cannot read SMS. 46% cannot respond to SMS. This section of the population, at least 30%, will be left out of the mapping data. 

For responding to this government initiative, one must be able to at least read the SMS. A quick look at the data give us a profile of this 30% population. First of all, as one would expect, these households are poorer than others. 55% of these households have per capita income under BDT 4000, compared to 50% in the rest of rural households. The average per capita income is significantly lower in these households. The most digitally literate persons in these households are also less educated and have lower rates of literacy. The 30% have a higher proportion engaged in agriculture and self-employment. 

The following table shows the percentage of rural population who cannot read SMS. For the self-reporting and mapping purpose, Mymensingh, Rangpur and Sylhet are most likely to have a higher proportion of people not responding to the request for information.  

In this context, we need to think about how to get data from this population. One possibility is to call a select group of mobile users directly instead of sending them SMS. NGOs working at the grassroots level can also help in gathering information for this government portal. The 1,00,000 frontline BRAC workers who are going door to door to raise awareness can contribute. With data from all mobile users of Bangladesh, this map can make a huge difference in how quickly we tackle this crisis as a nation.


Mehnaz Rabbani is the Head of Operations & Strategic Engagement and Partnership at BRAC Institute of Governance and Development, BRAC University. 

Photo: Man takes a photo with his mobile phone in Dhaka, Bangladesh, by Abir Abdullah for ADB, licensed under CC BY-NC-ND 2.0. 

Up