New Delhi: In a bid to tackle an entrenched caste system, the ruling Bharatiya Janata Party (BJP) is mulling to introduce a big data scheme to tackle the problem of inequality, data analytics firm EY said on Thursday.
In a move aimed at helping the poor and middle class, EY’s chief data scientist, Kunal Gupta, said the scheme, which will also apply to the 1% of the population, would require a total of 3,000 data centers and the cost would be around Rs.50,000 crore.
The scheme would be implemented by 2020 and the data would be gathered by the state government, E y said.
The government, however, has so far refused to reveal how many data centers would be required, saying only that it would be in line with existing infrastructure.
The move comes amid mounting criticism over the lack of data transparency and the fact that data analytics are seen as a “non-starter” in many states where they are being tried.
The idea of introducing a big analytics scheme to combat inequality has gained momentum after data analytics company, Cognizant, revealed it was one of the key factors behind India’s failure to improve in several key metrics including literacy and the percentage of women in the workforce.
Analytics firm Cognizants data, for example, has been the main force behind the Bharatiyat’s efforts to bring down poverty and inequality in the country.
A report by McKinsey in December 2016, however said that “the lack of transparency about the use of such data to generate actionable insights and actionable actionable predictions remains a significant impediment to the adoption of innovative solutions”.
The data analytics companies have raised concerns over the impact of data analytics on the functioning of the economy and its ability to make the country more competitive.
It has also led to criticism that it is too often used to make political and social decisions, rather than making investment decisions.
The biggest beneficiaries of this data-driven approach are the ruling Congress party, which is heavily reliant on data analytics.