Coronavirus, which emerged from Wuhan in China, has now gripped the entire world. India, which imposed an early lockdown, could only postpone the spread. In recent weeks, however, it has been reeling with the virus infecting ever more numbers. India has already surpassed the UK to become the fourth worst-hit nation. According to data compiled by John Hopkins University, India by June 22 has reported 4,25,282 confirmed cases against the UK’s 3,05,803. More worryingly, in the last few days, India has recorded over 14,500 positive cases and more than 350 deaths every day.

India is a densely populated country with a large base of migrant workers. According to the 2011 Census, 37 per cent of the population comprises internal migrants. The exodus of migrant workers following the imposition of the 21-day lockdown on March 25 elevated the risk of spreading the coronavirus. According to some recent media reports, several lakh migrants have already returned to their home-States since the lockdown started. Millions of others are still stranded in cities and waiting to return to their native States.

Bihar, Uttar Pradesh and West Bengal have seen the maximum influx of migrant workers. Odisha had received 5,13,103 domestic migrants by June 21, according to its official Covid-19 dashboard. The deadly virus has moved from more economically-active cities to less active town and villages, spreading the disease within States quickly.

Bihar, West Bengal, Odisha and Assam have witnessed huge spikes in positive cases since the return of migrants. The share of these States in the national tally has grown over the past few weeks except for Uttar Pradesh. More importantly, their combined share (excluding UP) in the country’s total positive cases has risen from 3.5 per cent in March to 8.1 per cent in June. Nationwide, new infections and deaths have also surged.


Correlation in States

Heatmaps and spread analysis clearly show the coronavirus becoming contagious in States. The States now have become more correlated. The spread analysis calculates the difference between short-term (five-day) and long-term (20-day) averages of confirmed cases and deaths. The spreads in both confirmed cases and deaths have widened for all most all States over the past few weeks. Haryana, Telangana, West Bengal, Karnataka and Jammu & Kashmir are the new States to report a surge in positive cases and deaths.

The heatmap of cross-correlations in confirmed cases signals a grim scenario of the spread of the virus within States. The States are found to be highly correlated in confirmed cases in the past one month in comparison to their respective correlations in March.

States like Assam, Arunachal Pradesh, Jharkhand, and Tripura, which did not report a single positive case till March-end, have registered a large number of infections over the past two weeks. Their correlation with other States has risen.

With high correlation among States, the risk of infections rising has become high. Detection of positive cases will rise as States undertake more tests. More deaths may follow, as fatalities generally lag by two-three weeks.

Fatality rates

Daily deaths at a fatality level (such as deaths between 1-25, 26-50 and 51-100) are calculated as the ratio of total deaths to the total number of days of that fatality level. According to the analysis, States have witnessed higher per day deaths at higher levels of fatality. For example, Maharashtra has recorded per day deaths as high as 132 at its current level of 1,001-4,000 deaths when compared with previous fatality levels. Delhi, Tamil Nadu and other States follow a similar pattern of critical per day deaths at higher levels of fatality.

The heatmap of per day deaths suggests steeper death curves for several States. Delhi has the highest per day deaths of 140 people currently. Its per day deaths have increased steadily through all levels of its fatality. Maharashtra and Gujarat have seen similar steep death graphs. Tamil Nadu recently crossed the grim milestone of 500 deaths, and it has reported its highest per day deaths of 28.

West Bengal, Uttar Pradesh, Madhya Pradesh and Haryana have also witnessed high per day deaths at their current fatality levels. Telangana and Rajasthan have seen huge spikes in deaths over the past few weeks. These two States have reported eight deaths per day in their respective current fatality levels, and may soon join other critical States in recording very high deaths per day.

A steep increase in the number of confirmed cases and deaths has been seen across States over the past few weeks. With a continuous surge in reverse migrations and highly associated States, what lies ahead is a slippery path.

Panda is Senior Data Scientist, IBM India and Biswal is Professor, MDI Gurgaon. Views are personal