Big Data ignoring the plight of the poor

My recent quest for answers to the current problems led me to Thomas Piketty's economic magnum opus, Capital in the 21st Century. The book is an extensive yet incisive economic rationale of the growing inequality in the world.
Some 600 pages later I realised that the book was entirely focused on the top 10 per cent of wealth earners who own more than half the wealth in US and Western Europe. He had little to say about the bottom half of the population on the wealth distribution ladder that barely owns 5 per cent of the wealth. This was ironic because the book's central thesis - inequality - mattered most to this bottom half. In the concluding chapter, I understood why his book appeared to not dwell enough on the poor. He did not have good data about the income and wealth of the economically disadvantaged - the bottom half. The dystopic economic view aside, he was largely forced to skim over the situation of the poor.
World Bank estimates that the pandemic will push 70-100 million people into extreme poverty in 2020 alone. These are people earning less than $2 a day. When in good times, the poor live from pay cheque to pay cheque, during tough times they are in dire need of direct government help. But despite the best intentions, government agencies in developing countries can never be sure if the benefits reach their target beneficiaries.
Governments of developing countries or the Global South need a rigorous system of data collection. No government, no matter how efficient it is, can run a successful programme to reduce inequality in the absence of credible data.
Nobel Laureate Abhijit Banerjee has been outspoken about the lack of reliable data on the informal sector in India, which makes estimates of poverty and even GDP untenable. So how would the government reliably provide any kind of direct monetary support during a crisis? International agencies like the World Bank has taken a dim view of the dated data from India, which is from the 2011 Census. Even the World Bank's estimates of global poverty levels are suspect because Nigeria, India, and the Democratic Republic of Congo are home to a third of the world's extreme poor. In all these countries data on the poor is questionable even though the use of mobile phones and tablets and GPS has improved surveys.
In Africa, data produced by national statistical offices is bad. There is no incentive to improve the process of data collection. The data is either incomplete or wrong and is rarely audited. Donors have no certainty that their funding for social programmes has any material impact.
The Indian government relies on Household Consumer Expenditure Survey conducted by the National Sample Survey Office to measure shifts in poverty. The 2017-18 survey results were withheld by the government for fear of losing political capital because the results reportedly showed a slowdown in consumption. A survey can tell you the 'what' if the sample is statistically significant, but it cannot tell you the 'who'. The distress of the migrant workers was compounded during the lockdown because they could not benefit from central or state government programmes. For any portability of welfare schemes across states, there needs to be a robust tracking of such workers. They have essentially been nobody's problem and have fallen through the cracks.
Good governance is data driven. It is backed by a full machinery for data collection on household consumption right from the village level. This does not have to be the government's problem alone. Tech can pitch in to solve this problem in a new paradigm. Big Tech and investment banks are great at applying analytics and AI to data on people who spend well. What about those who don't? Technology companies' efforts to change the world will remain on the fringes unless they can provide revolutionary tech to make this happen.
So, while we are caught up in a debate about a V-shaped or a U-shaped economic recovery, we will never be able to realise a real recovery, unless we strike at the global issue of inequality. For that, collecting good data on the poor is the first step. We need to learn to walk before we can run. 

Shalini Verma is CEO of PIVOT technologies