August 31, 2020

Assessing Socio Economic Vulnerability of Occupations In India

Ashwin MB +, Rajeswari ParasaSneha P




The strict national lockdown from March 24th disrupted the economy and worsened the pre-existing socio-economic vulnerabilities of the informal workforce. India saw a precipitous rise in unemployment rates - from 8.8% before the lockdown to around 23.5% by the end of April with approximately 122 million workers losing their jobs. Moreover, household surveys found that over 45% of households reported a reduction in income compared to a year ago.  


The pandemic has exposed the precarious nature of the livelihood of workers. Without a constant stream of income, many workers face the risk of hunger and eviction. This prompted a mass reverse migration - people travelling from their work states to home states - which increased the risk of transmission of COVID-19 from urban to rural areas.


One of the reasons for increased unemployment and reverse migration is the lack of social security at workplaces. Benefits such as formal contracts, eligibility for paid leave, healthcare insurance and maternity benefits provide a safety net for workers in a time of crisis. Without these safeguards, workers are more socially and economically vulnerable, and at higher risk of losing their incomes and livelihoods.


Other workers may experience health vulnerabilities, as they might be forced to continue working even when unwell. This could put their co-workers at risk as well. To understand this vulnerability better, we analyze the Periodic Labour Force Survey (PLFS) 2017-18 and identify the vulnerability of different sections of the Indian workforce. First, we describe the data used for the analysis followed by some limitations of the study. Then, we discuss the findings and our recommendations from the analysis before concluding with some future research ideas.


Data and Methodology


The PLFS 2017-18 was conducted by the National Sample Survey Office (NSSO) from July 2017 to June 2018. It covers about 4,30,000 individuals and 1,02,000 households across different states in India. PLFS classifies workers based on 113 occupation groups (referred to as occupations from now on) using the National Classification of Occupation (NCO) 2004. These occupations are divided into nine broad categories called occupation divisions based on the skill1 demanded and type of work. For example, administrative/ managerial roles like General Managers, Directors and Chief Executives are categorized under the occupation division 'Legislators, Senior Officials and Managers'. Similarly, clerical occupations such as Secretaries, Keyboard operators come under 'Clerks'.  


We use four indicators: social security benefits, contract types, paid leaves, and median income to gauge the socio-economic vulnerability of workers in all 113 occupations of the survey. Social security benefits identifies whether an employee is eligible for Provident Fund/pension, gratuity, healthcare benefits, maternity benefits. Contract type helps us identify whether the employee has a formal contract with the employer or not while paid leaves indicates whether the worker is allowed to take paid leaves. 


Using the data from the first three indicators, we identified the proportion of workers engaged in a particular occupation without access to any of these three provisions from their employers. In other words, we calculate the percentage of workers within each occupation that are not eligible for: (a) any social security benefits, (b) any paid leave, and (c) have no contract. It must be noted that we used binary outputs (eligible/not eligible) for ordinal variables like social security benefits and contract type2 (paid leave is recorded as eligible/not eligible).


The last indicator that we considered, to understand the vulnerability of workers in an occupation, is their income. The PLFS records monthly income for regular and self-employed workers but the daily income of the previous seven days for casual workers. To maintain consistency, we converted daily income to monthly income3 and calculate the median monthly income across different occupations. 


In the PLFS data, regions are classified into rural and urban based on the location of the household at the time of the survey. Since the survey was conducted quarterly in urban households, our analysis only uses data from the first visit of the PLFS to avoid duplication.


Before presenting the results, here are a few caveats. First, PLFS data only includes whether a worker is eligible for paid leave or not but not how much paid leave. This limits our ability to capture the variation within this benefit. Second, the data is silent on the population of migrant workers - who could be at a higher risk compared to local workers and an important indicator for socio-economic vulnerability. Finally, the PLFS data cannot comment on virus exposure risk, which is also a form of social vulnerability.




Split by region, we see that 72% of workers reside in rural areas while 28% reside in urban areas. Skilled Agricultural, Forestry and Fishery occupation division employs the largest share of the rural workforce at 42.8% while Craft and Related Trades occupation division employs the largest share of the urban workforce. (Figure 1)



Figure 1: Occupation divisions and their share of the total workforce by region


Figure 2 shows the top 15 occupations in terms of the share of the total rural workforce. The top 15 occupations make up almost 84.7% of all the rural workforce. Unsurprisingly, the majority of the employees in rural areas are engaged in agriculture-related occupations with more than 37.1% of the rural population working as Market Gardeners and Crop Growers while Agricultural and Fisheries, and Mining and Construction labourers make up 15.3% and 7.7% of the total rural workforce respectively. 


Figure 2: Top 15 occupations by share of the total rural workforce


Similarly, Figure 3 describes the urban workforce and their distribution across occupations. The workers in the top 15 occupations constitute around 57.3% of the total urban workforce, which is much lower than their rural counterparts and indicates a greater diversity of occupations in urban areas.


A closer look at the individual occupational types reveals some interesting findings - ‘Directors and Chief Executive’ occupies the first rank with 11.3% of the urban workforce followed by ‘Shop Salespersons’ and ‘Motor Vehicle Drivers’. We find that almost 97.9% of the Directors and Chief Executives are either self-employed or unpaid family members5, which indicates that these workers are largely MSMEs.


Figure 3: Top 15 occupations by share of the total urban workforce


In Figure 4, on the y-axis, we show the median income of different occupations while the x-axis shows the proportion of the workers in the occupation with no social security provisions. The size of the bubble represents the proportion of the workforce in the rural region engaged in a particular occupation. Finally, the colour of the bubble shows the occupation division of the occupations as per the NCO - 2004 classification. Figure 5 captures similar indicators for the urban region. While Figure 4 and 5 show the top 50 occupations (by the proportion of workers) in each sector, the numbers on indicators for all the 113 occupations can be found in Table 1.



Table 1: Full list of 113 occupations with information on the share of the workforce by region, median income, and % of workers in the occupation without any of the social security provisions.


Hover over bubbles in Figure 4 and 5 below to explore each occupation, its median income, percentage of the workforce engaged in the occupation as a share of the total workforce and their vulnerability status. Occupations near the lower right corner are the most vulnerable to economic shocks.



Figure 4: Income and social security status of top 50 occupations in the rural workforce.


Overall, at least 62.7% of the rural workforce and 49.7% of the urban workforce are not eligible for any social security provisions. Specifically for the rural regions, occupations which fall under the division of (1) Elementary Occupations6, and (2) Skilled Agricultural Fishery Workers are the most vulnerable. Under the former, at least 86.7% of the labourers employed in Agriculture, Fisheries and Other Related occupations, which constitute 15.3% of the rural workforce, do not have access to social security provision. Under the latter, at least 72.2% of the workforce employed as Market Gardeners and Crop Growers - which constitute more than a third of the rural workforce - do not have access to basic social security provisions.


Figure 5: Income and social security status of top 50 occupations in the urban workforce.


In urban regions, we see a similar trend with occupations under the divisions (1) Elementary Occupations, and (2) Skilled Agricultural Fishery Workers being the most vulnerable. Additionally, some occupations under services’ division such as Shop Salespersons and Demonstrators, and Housekeeping and Restaurant Services - which combined form almost 11.8% of the total urban workforce, are highly vulnerable with at least 60% of the workers lacking any social security provisions. Further, the majority of construction labourers and motor vehicle drivers, who form almost 10.1% of the urban workforce, do not have access to any social security provisions.




Identifying vulnerable occupations and the corresponding share of workers they employ is a critical step towards mitigating the effects of the economic impact of COVID-19. This analysis aims to provide a basic measure of the types and scale of vulnerability that the workforce is exposed to.


While certain sections of the workforce particularly health workers, other frontline workers, construction workers had targeted economic relief programs, several other occupations remained without any social security. Only a few states like Tamil Nadu, Karnataka and Rajasthan made social provisions for street vendors and other daily-wage workers. However, these too were often provisions for registered workers. The national MSME relief package has also been subject to several exclusions. Given the extent of informality of the Indian labour market, relief packages need to be better targeted (when it is not financially feasible to be universal) to include all vulnerable sections of the workforce in its ambit. However, this also requires other forms of information and data such as migration status. There has been some discussion around capturing migration data for the next PLFS survey. This would be a welcome step as assessing the vulnerability of migrant workers post-COVID pandemic and lockdown is vital for designing relief packages. 


Further, it would be policy-informative to collect more granular data on occupation titles. Currently, several different professions are clubbed together under one occupation. For example, ‘Other Health Professionals’ covers a wide ambit of dissimilar professions including dentists, pharmacists, physiotherapists among others. Some titles - like ‘Directors and Chief Executives’ are misleading and indicate formal, secure occupations which in fact are often poorly/un-paid household-based productions. 


The pandemic has demonstrated the need for good data to inform policy decisions under time constraints and labour market data in India has several limitations. To this end, we recommend that data collecting agencies (NSSO, MoSPI, Labour Bureau) collect additional information such as the distance of the workplace from home (domicile and temporary), access to means of transport, access to toilets, regular breaks, meals, terms of employment, ability to work from home, and physical working conditions (high temperatures, radiations, hazardous equipment) & exposure (average number of individuals interacted per day, duration of interaction)7. These data points will provide much greater insight into the vulnerability of particular occupations to COVID-19 or any other potential health crises and thus, allow the government (centre and state) to respond appropriately.


Future Scope 


Our analysis of the socio-economic vulnerability of occupations is at the national level. It would be useful to conduct this analysis at the state or district level and compare the findings. However, one caveat could be the lack of sufficient data points.


Further, we conducted the analysis by urban and rural regions. Analyzing the social security provisions across gender, caste and religion would provide a better picture on the intersectionality between occupational vulnerability and other forms of social vulnerabilities.


Finally, although our analysis here uses PLFS 17-18, the Ministry of Statistics and Programme Implementation released the latest PLFS numbers in June 20208. Thus, we recommend replicating this study for the latest data. Resources to replicate this analysis can be found at the GitHub repos here and here.


+Ashwin MB, former Teach For India fellow, currently interning with us, is pursuing Masters in Public Policy at UC Berkeley.

** The unknown values in ‘social security benefits’ indicator have been considered as eligible for social security provisions. This means x-axis shows the minimum percentage of workers within an occupation without any social security provisions.

1According to NCO 2015, the skill level of each newly identified Occupation was decided on the basis of information about academic and technical qualifications and experience requirement as also the average job description of the occupation to see whether the job requirement was of administrative, managerial, supervisory nature or of a subordinate/repetitive nature in the Indian context.

2We categorized workers with no contract as “not eligible” while the other three contract types (1-2 years, 1-3 years, 3-4 years) were bundled as “eligible.”

3Monthly income for casual workers is given by: av_incomei x employed_daysi x 30/7, where av_incomei is the average daily income for the worker i; and employed_daysi is the number of days employed in the last seven days.

4We use median income as it is less sensitive to outliers.

5Unpaid household members are defined (according to PLFS 17-18 annual report) as those who assisted in the operation of an economic activity in the household farm or non-farm activities.

6Elementary occupations (as described by NCO 2015) consist of simple and routine tasks which mainly require the use of hand-held tools and often some physical effort.

7Similar to data points collected by O*NET program.

8This survey too was conducted before the pandemic.


Acknowledgements: We are grateful to Vaidehi Tandel, Sridhar Ganapathy, Isalyne Gennaro, Harsh Vardhan Pachisia, Sofia Imad, and Vikram Sinha for their methodology, data, and editorial inputs.

x Close Window

Please verify your email address to access this content