Last week, India crossed the somber milestone of one crore confirmed cases of Covid-19 infections. The lockdown announced by Prime Minister Narendra Modi in March resulted in many hardships for Indians, especially for the millions of migrant workers who were stranded in urban centers without any work and with little savings. But the lockdown it failed to contain the spread of the coronavirus.

The government decided to effect Unlock phase on June 1, even when Covid-19 cases were on a steep rise in most parts of the country. The peak of the disease curve came several months later, in September.

Despite experiencing a lockdown that was arguably the most draconian anywhere in the world, Covid-19 spread widely across India. It is widely assumed that the Shramik Special trains, which were run from May 1 to allow stranded migrant workers to travel back to their home states, were responsible for this spread.

Any casual observer would agree that it is the conventional wisdom that the blame for the spread of the virus to rural India, even if it is inadvertently, lies with the workers who had traveled on these trains. A recent New York Times article makes a similar argument that the virus, hitherto concentrated in the cities, was taken to rural areas in “almost every corner” of the country by the migrant workers traveling in these trains, where it proliferated rapidly and unleashed devastation.

The article even goes to the extent of dangerously normalising class prejudices by calling these “virus trains”. Similar arguments were widely made during lockdown and subsequently, for instance, here, here, and here.

Were the Shramik trains really responsible for failure of lockdown to contain the virus spread?

Sampling bias

Most of these reports follow a similar pattern: based on ground reporting from one or two districts, they generalise the situation in these districts to the rest of the country by citing other media reports from the month of May. These media reports, in turn, note that a significant number of confirmed infections in many districts were of migrant workers who had just returned to these districts.

Importantly, none of these reports state the number of the tests that were being conducted on returning migrant workers as a proportion of the total tests being conducted in those districts – an important metric to assess the spread of the virus.

To provide some context to the situation prevailing in May, with apprehensions about the possibility of returning migrant workers carrying the virus along with them, the Indian Council of Medical Research had just widened the testing regime to include testing of returning migrant workers. Many states adopted strategies to test large number of returning migrant workers, even when their overall testing capacity remained relatively low.

This testing strategy, unsurprisingly, resulted in a significant proportion of total confirmed infections being found in the returning migrant workers, which was regularly amplified by a sensationalist media. This is similar to the sampling bias that furthered the vilification campaign during the initial phase of the lockdown against the people who attended a Tablighi Jamaat event in New Delhi.

A family in Greater Noida waits to board a Shramik special train. Credit: Adnan Adibi/Reuters

Now, consider the three states that received the highest number of Shramik Special trains in the month of May – Uttar Pradesh, Bihar, and Jharkhand. These three states together accounted for about 75% of the total trains that were run until May 27. Data available at the Covid19India website shows that by May 1, confirmed cases of Covid-19 had already been reported in 65 of the total 75 districts of Uttar Pradesh, 30 of the 38 districts of Bihar and 12 of the 24 districts of Jharkhand.

Taken together, 107 of 137 districts – or nearly four out of every five districts in these three states–
had detected infections even before the first Shramik train reached the destination. This indicates that reality is far from narrative about the trains being responsible for “taking the virus to every corner of the country”.

With regards to number of tests on migrant workers, taking into consideration the little publicly reported data, in Uttar Pradesh, data until June 3 indicates 27.5% of total positive cases are of migrant workers, 24% of the total samples tested were also of migrant workers.

In Bihar, considering official testing data of migrant workers and total tests data until May 17 shows that while 49.3% of total positive cases were of migrant workers, 18.2% of total samples tested were also of migrant workers.

In Bihar, by July 22, only two of the top 10 highest migrant worker receiving districts figure in the top 10 districts with highest number of confirmed cases, indicating that the return of migrant workers had little role in the surge of cases in the state.

In Jharkhand, calculations based on data until May 22 show that 48.5% of total confirmed cases and 46.4% of total tests were of migrant workers. It is also to be noted that many of the returning migrant workers were made to undergo institutional quarantine, limiting the possibility of virus transmission.

All of this suggests that while a portion of the returning migrant workers did carry the virus along with them, there isn’t a case to suggest that their journey back home is primarily responsible for the spread of Covid-19 across the country.

Serological survey data

Based on this author’s experience in help coordinating a nation-wide relief effort for migrant workers during lockdown through Migrant Workers Solidarity Network, even anecdotal evidence of a Covid-19 surge after the return of migrant workers was limited to very few places, one of them indeed being Ganjam district of Odisha on which the New York Times article was primarily based.

Most significantly, evidence thrown up by the results of the first nation-wide serological survey conducted by the Indian Council for Medical Research, suggest that by early May, there were as many as 8.56 lakh undetected infections in the 233 districts across India – most of them predominantly rural – which were deemed “zero caseload” districts at that time.

This makes it clear thateven before the return of the migrant workers, the virus, being highly contagious and with a significant proportion of people that it infects being asymptomatic, had spread across the length and breadth of the country.

Despite strong evidence to the contrary, what explains the narrative about the returning migrant workers being primarily responsible for spreading the disease becoming the conventional wisdom? This could, in great part, be attributed to the class prejudices against these workers, most of whom live precariously working in the informal sector in towns and cities where they have few social connections.

It is the entrenched prejudices that resulted in bleach being sprayed on them, police dehumanising them or brutally beating them up. Even before the pandemic struck, migrant workers were being threatened and called infiltrators, were fleeing targeted violence and lynching by Hindutva mobs. In this atmosphere, describing the Shramik trains as “virus trains”, as the New York Times article does, only serves to perpetuate the vilification of migrant workers.

To be sure, there were many things wrong in the operation of the Shramik trains. There is a strong case that they should have been run much earlier in the lockdown. It was only after widespread and spontaneous protests by workers for their legitimate right to return home that the Modi government was forced to start them.

When they did begin, they were shrouded in secrecy and complex procedures resulting in extortion, high fares were charged from already distressed workers, not being provided with food and water, huge delays, among many other problems. But being responsible for Covid-19 running amok isn’t among them.

Sunil Tamminaina is pursuing an M Phil at Jawaharlal Nehru University. During the lockdown, he helped coordinate a nation-wide relief effort for migrant workers under the banner of the Migrant Workers Solidarity Network. He can be contacted at