Hunger Grows as India’s Lockdown Kills Jobs

Results of a Survey from 12 States
Rahul Lahoti, Amit Basole, Rosa Abraham, Surbhi Kesar, Paaritosh Nath
Issue: June 5, 2020

Amit Basole is Associate Professor of Economics; Rahul Lahoti is Visiting Faculty, and Rosa Abraham, Paaritosh Nath, and Surbhi Kesar are Research Fellows; all at Azim Premji University.

Heading home at last: On a Shramik Special Train | Sunil Singh (Wikimedia)

The lockdown has devastated livelihoods at unprecedented levels, causing widespread job losses and hunger. The measures so far will not solve the crisis. India should universalise the PDS, expand cash transfers, and create a national job guarantee scheme.

The Covid-19 global pandemic and its associated containment measures have taken a heavy toll on economies and societies worldwide. In India, the national lockdown, imposed on 24 March and subsequently extended three times to 31 May, has had a profound effect on employment and earnings.

Livelihoods have been devastated at unprecedented levels. Food insecurity and economic vulnerability have increased to staggering proportions. Hunger deaths and suicides linked to economic stress are being reported from various parts of the country.

Any recovery from this crisis will be slow and very painful. The immediate relief measures do not appear to be in proportion to the severity of the situation on the ground.

[U]nequal development manifests as migration flows of millions of workers, creating dense populations of the urban poor [...] when put together with India’s persistent underinvestment in health and housing, our vulnerabilities to the present crisis become painfully clear.

In a survey of nearly 5,000 self-employed, casual, and regular wage workers across 12 states of India, conducted between 13 April and 23 May in collaboration with civil society organisations, we have found a massive increase in unemployment and an equally dramatic fall in earnings.1 

Two-thirds of our respondents have lost work. The few informal workers who were still employed during the lockdown saw their earnings drop by more than half. Almost 8 in 10 are eating less food than before. The impact of job losses and food insecurity has been higher for certain groups of people: Muslims, Dalits, women, and those with lower levels of education.

As late as May, even the grossly inadequate government relief measures, in the form of cash transfers or increased rations, had not reached large sections of the economically vulnerable population. Half of our respondents reported not receiving any cash transfers.

Three structural characteristics of the Indian economy and decades of underinvestment in public goods have combined with the sudden and severe lockdown to generate widespread misery.

First, for India’s predominantly informal labour force, which earns income day-to-day, any stoppage of economic activity instantly destroys employment and earnings. Second, the lopsided nature of economic growth has created a large divide between rich and poor states. Third, the same uneven growth process has created employment opportunities in larger cities much faster than in smaller towns and villages.

This unequal development manifests as migration flows of millions of workers, creating dense populations of the urban poor. When these long-run factors are put together with India’s persistent underinvestment in health and housing, our vulnerabilities to the present crisis become painfully clear.

The [relief] measures recently announced by the government [...] fall well short of what is required, since the bulk of the stimulus package focuses on increasing liquidity rather than direct spending.

The vulnerable population in urban areas has been severely impacted, but relief measures have not reached them. The union government should follow the lead of states such as Odisha and Himachal Pradesh, which are trying out an urban employment generation programme, or of Kerala, which already has such a programme running. State finances may not allow effective urban employment guarantees, but the central government can certainly do this.

The measures recently announced by the government go some way in addressing the crisis. But they fall well short of what is required, since the bulk of the stimulus package focuses on increasing liquidity rather than direct spending.

Survey Coverage

We partnered with 10 civil society organisations for phone surveys in 12 major states of India. They provided us a list of telephone numbers from their databases, from which we randomly selected households for the organisations to interview.

Our results are a conservative estimate of what the most vulnerable population might be experiencing.

The sample is purposive and non-representative of the states and regions it covers. However, we ensured that the sample is diverse by selecting respondents across rural administrative blocks and by talking to respondents across diverse occupations. In total, we completed 4,879 interviews across rural and urban areas in 12 states.2 Our survey covered 161 districts and seven large cities (Ahmedabad, Ajmer, Bengaluru, Bhubaneshwar, Delhi, Jaipur, and Pune). Respondents interviewed in urban areas spanned over 30 different occupations. Our findings are broadly consistent with what other similar surveys reveal.

Since our respondents are in the network of civil society organisations, they are more aware and have more access to government schemes. They are not among the most economically vulnerable sections in the country. Hence our results are a conservative estimate of what the most vulnerable population might be experiencing. The results presented are not weighted, and the conclusions below are not generalisable. But with the paucity of representative data, they present a useful broad picture to guide policy.

In some of the states, our sample is predominantly rural or largely urban, and is indicated in the labels in Figure 1.

Overall, 58% of our respondents were in rural areas, 52% were women, 79% were married, and the average age was 38 years.3 About half of the sample was illiterate or had only primary education. Of those interviewed, 85% were either the main income earners of the household or spouses of the main income earners. More than 8 in 10 households had a monthly income of less than Rs 10,000 in February.

In rural areas, more than half of our respondents were either farmers or casual workers. About 7 in 10 households in rural areas owned some cultivable land. In urban areas, more than 4 in 10 respondents had in February been non-agricultural casual workers (Figure 2). A majority of workers in urban areas were service providers or construction workers. They worked as rag pickers, domestic helps, tailors, sanitation workers, street vendors, plumbers, painters, carpenters, taxi or auto drivers, balloon sellers, cleaners, beedi makers, construction labourers, or security guards.


Impact on Livelihoods

The lockdown has resulted in unprecedented losses in livelihoods and earnings.4

Amongst those who had been working in February, 66% reported they lost employment during the lockdown. They could not find even a single day’s work during the lockdown or had to completely stop their self-employment activities. The loss of employment was more severe in urban areas, with almost 8 in 10 respondents reporting losing work, compared with just over half in rural areas (Figure 3).

About half of regular wage workers contacted by us received a reduced salary or no salary during the lockdown period. This was higher for those in urban areas (54%) than those in rural areas (39%).

“I have worked for so many years in the factory, I should get paid at least a little amount for now. The government should also help. Otherwise how will we bear daily expenses?” asked a 39-year-old male regular-salary worker from Uttarakhand.5

The non-agricultural sector had more job losses than the agriculture sector, with casual workers and the self-employed being the worst-off. (Figure 4).

Individuals who had some degree or diploma seem to have experienced relatively lower loss in employment. Members of the Other Backward Castes experienced relatively lower loss in employment compared with those from the Scheduled Castes or the Scheduled Tribes. Muslims in our sample have experienced substantially larger losses in employment than Hindus, at 81% versus 65%.

Respondents living in vulnerable urban households—those with monthly income of less than Rs 10,000 in February—have seen greater levels of loss of employment (79%) than those in higher-income households (76%). In rural areas, fewer respondents from land-owning households reported loss of work, relative to those from landless households: 67% versus 51%.

For those who managed to retain work during the lockdown, incomes were more than halved. For casual workers, median monthly earnings dropped to Rs 1,727 from Rs 3,750, and for the self-employed in the non-agriculture sectors, to Rs 2,165 from Rs 5,000. The median earnings drop was higher in rural areas than in urban areas. Among farmers, as many as 85 per cent faced severe difficulties in harvesting and selling crops. They were unable to harvest due to lack of machinery or labour. They could not sell their harvest or had to sell at reduced prices due to lack of transportation.

Impact on Households

The lockdown has led to severe levels of food insecurity, depleted savings, and increased household borrowing. In urban areas, 83% reported consuming less food than before. In rural areas, the corresponding figure was 73%. Food insecurity was higher among women, Muslims, Scheduled Castes, people with lower educational qualifications, and in vulnerable households.

Almost half of our respondents reported not having enough money to buy even a week's basic essentials.

“Living has become hard. We have two children and are unable to provide them three meals a day. During many instances, we go hungry or eat one meal to give food to them,” said a 31-year-old female casual worker from Karnataka.

“There is no help from the government.”

Almost half of our respondents reported not having enough money to buy even a week's basic essentials. The contrast here between rural and urban areas is stark: the lack was reported by 65% of urban respondents but just 35% of rural respondents. This probably reflects the greater dependence of urban respondents on market-provided goods and services.

More than 8 in 10 respondents did not have money to pay the following month's rent. More than a third were taking loans to cover daily personal expenses during the lockdown. The median size of such loans was Rs 5,000. Urban respondents were more likely to take loans (41%) than rural (33%) (Figure 5). Less than a tenth of these loans were taken from formal sources, indicating the limits of banks to reach out at times of distress. Half of the respondents borrowed from friends and family. About 2 in 10 borrowed from moneylenders.


Reach of Relief Measures

Large sections of the economically vulnerable population had not received any government relief. (Figure 6). Almost 2 in 10 vulnerable households were unable to get any rations. Muslim respondents were also more likely to not be able to procure rations than Hindus: 27% versus 17%. The median quantity of grains obtained by those able to access rations was 5 kg per person, far less than the promised quantity.

“The villages are in bad shape. We can’t buy essentials even. And we can’t go to the town for work or for buying items for daily needs,” said a 37-year-old male self-employed worker from Rajasthan.

“Savings are also getting depleted. It will be difficult to keep going for months like this.”

Since our respondents are connected to civil society organisations, they are more likely to be able to access the ration system than other socio-economically similar households. The inability to get rations is likely to be worse for other groups.

The saga is far worse when we examine the reach and extent of the promised cash transfers. Half of our respondents reported not receiving any cash transfer under central or state-specific schemes.6 The reach of cash transfers was particularly weak in urban areas, where 63% did not get even a single cash transfer.

[The Jan Dhan cash transfer] translates to Rs 4 per person per day [...] Compare this to the poverty line of Rs 50 per person per day in rural areas and Rs 73 in urban areas.

In absolute terms, on average, our respondents had lost Rs 2,000–3,000 of income per month. The announced cash transfer that is supposed to reach most households is grossly inadequate to compensate for this loss.

For instance, the central government announced cash transfers of Rs 500 per month for three months to every woman with a Jan Dhan bank account. This translates to Rs 4 per person per day, assuming the eligible woman has a family of four. Compare this to the poverty line of Rs 50 per person per day in rural areas and Rs 73 in urban areas (based on the Rangarajan committee estimates of 2011–12, converted to current prices).

Even the reach of Jan Dhan accounts is very limited. Only 1 in 3 vulnerable households reported receiving the promised cash transfer. This was mostly because a large majority did not have Jan Dhan accounts. Overall, 6 in 10 vulnerable households did not have Jan Dhan accounts.

The suspension of MGNREGA work has deprived the vulnerable of a safety net. “We want to work in MGNREGA but no work is available [...] The contractor has not started work this year,” said a 40 year old female casual worker from Jharkhand.

“My payment from last year's work is also pending, it was wrongly deposited in someone else's account.”


The disruption from Covid-19 and the lockdown in the economy and the labour markets is enormous. Any relief measures for this crisis should at minimum consist of three prongs: a universalisation of the Public Distribution System (PDS), the expansion of direct cash transfers, and an expansion of the rural employment guarantee scheme and its introduction in urban areas. We expand on these recommendations here.

Governments should [...] universalise PDS for the next six months. India has enough food reserves for this.

The PDS has the widest reach in providing relief. But there still are a large number of households, especially amongst the most vulnerable groups, that do not receive rations due to lack of ration cards. Governments should issue temporary ration cards or through other means universalise PDS for the next six months. India has enough food reserves for this. Governments should also open cooked meal centres for the hungry. The mid-day meal scheme in schools, anganwadi centres, and public feeding programmes existing in several states can be utilised for this effort.

To cover the income losses, governments should provide a monthly cash transfer of at least Rs 7,000 for the next three months to households by combining administrative lists of welfare schemes such as MGNREGA, PDS, and Ujjwala. To relieve pressure on the rural banking system, panchayats could be used for cash transfers, as is being done in Odisha.

A nation-wide large-scale employment guarantee programme should be launched, expanding on the experiments in some states.

There is large demand for work, both from returning migrants and from those who have lost jobs. Large-scale worksites under MGNREGA should be started immediately. The allocation for MGNREGA work should be increased by at least Rs 1 lakh crore. The number of days of guaranteed work per household should be raised to 150 days. MGNREGA workers should be paid without delays, preferably with cash in hand, and their past dues should be cleared. To provide relief to the urban poor, a nation-wide large-scale employment guarantee programme should be launched, expanding on the experiments in some states with urban wage employment programs.7

The pandemic and the resultant economic distress is likely to stay with us for some time. This unfortunate situation has already caused a lot of suffering for many. We as a society need to realise the scale of the problem, be empathetic to the plight of the underprivileged, plan for it, and provide much needed direct and immediate relief.


The survey was conducted by a team of researchers at the Center for Sustainable Employment at Azim Premji University. The team includes Mridhula Mohan, Sanjana Rajashekar, Zaeen D’Souza, Sravan Pallapothu, Neha Joshi, and Adithi Rao. We greatly acknowledge their efforts in the project and several others who advised us at numerous stages of the project. We also acknowledge support and funding for the project from the research center at Azim Premji University and the Azim Premji Philanthropic Initiative. We thank the staff at the various civil society organizations that did the interviews and the numerous students volunteers from Azim Premji University. For any further questions get in touch with us at rahul.lahoti AT


The India Forum welcomes your comments on this article for the Forum/Letters section.
Write to

  1. The data and details of the survey and of the organisations are available at
  2. We made 7,000 telephone calls and our response rate was close to 70%. This response rate is far higher than response rates obtained in random number telephone dialling. This was partly because respondents were called by organisations with whom they had a prior relationship. We have 12 states where more than 180 respondents each were interviewed. Overall we have observations from 15 states.
  3. The distribution of sample by sex is not uniform across states. In a few states like Gujarat, our sample is mostly women, while for some others it is dominated by men.
  4. The findings reported here differ slightly from the ones presented in a webinar on 12 May conducted by the Center for Sustainable Employment at Azim Premji University, as they have been updated with results from about 1,000 households that have been surveyed since then. The broad patterns remain the same.
  5. Respondents were interviewed on conditions of anonymity, hence they are not named here.
  6. Detailed results for each cash transfer scheme are available at
  7. A detailed proposal has been made by the Center for Sustainable Employment.

Latest Articles

Idea of Social Stock Exchanges for India

Social enterprises have the potential to make a significant impact on India's society and economy, but they struggle to find sustained funding. Social stock exchanges can help bridge the funding gap and create an investment ecosystem for such enterprises.

School Teachers in NEP 2020

The New Education Policy pays obeisance to school teachers but while the policy makes some interesting recommendations, there is no radical change in the offing that will improve the quality of teaching or the working conditions of the all-important teacher.

Desi or Not?

The Indian-American community’s mixed feelings about Kamala Harris’ candidacy for US vice president spotlight its problem of racism and social conservatism.

Can Conventional Metrics of Employment be Accurate During a Lockdown?

Labour force surveys that yield estimates on job numbers are designed to capture work inputs in 'normal' situations. The concepts and methodology used can lead to errors in estimates during a lockdown when production has been completely or partially halted.

The National Education Policy on School Education

The National Education Policy lacks a coherent strategy to strengthen public education. It does not own the narrative of education as a public good, instead the policy pushes for private practices antithetical to public goals.
Back to Top