Why India needs unemployment insurance

  • Blog Post Date 11 March, 2015
  • Articles
  • Print Page
Author Image

Priya Ranjan

University of California, Irvine

The central government is contemplating reform of labour laws, including those that restrict the ability of firms to fire workers. This column contends that social protection needs to be kept in mind while designing these reforms. It makes a case for publicly-provided unemployment insurance in India and argues that labour laws should protect workers, not jobs.

Human beings are generally risk averse and seek insurance to protect themselves against various kinds of risk. In this column, I make a case for launching a broad-based social insurance programme in India, such as publicly-provided unemployment insurance, that allows individuals to mitigate labour-market risk.

Why do we need publicly-provided insurance programmes?

The market for private insurance against labour-market risk is missing due to the well-known information problems (moral hazard1, adverse selection2 ), and the issue of covariant risk3 . In this situation social insurance increases the welfare of workers by providing them protection against income risk. Moreover, Acemoglu and Shimer (1999) show that the production efficiency of the economy (or the aggregate output produced) reduces when risk-averse workers are unable to insure against labour-market risk. This happens because workers avoid taking up more productive but riskier jobs.

In a recent paper, I show that when unemployment arises because firms want to fire some workers who turn out to be bad matches for the job they are hired for, risk aversion of workers gives rise to inefficiency (Ranjan 2015). If a firm is very selective and keeps only high-productivity matches while firing other workers, then risk-averse workers, knowing that they have a high chance of being fired, will demand a high wage before accepting a job at such a firm. In this case, a firm is better off firing fewer workers by keeping low-productivity matches, which allows it to offer a lower wage.. In such a scenario, globalisation, which is generally welfare enhancing, can not only reduce the welfare of workers, but can also bring down social welfare overall4. This inefficiency can be addressed using social insurance devices such as unemployment insurance or severance payments5 made by firms to fired workers.

One may argue that in a developing country like India informal risk-sharing mechanisms exist that allow individuals to smooth consumption, obviating the need for social insurance. However, Chetty and Looney (2006, 2007) find that households may resort to very costly coping mechanisms, such as withdrawing children from school and sending them to work, to survive a negative shock to income in the form of unemployment.

Besides, India is undergoing rapid structural transformation with labour force moving out of agriculture and into the industry and services. According to the Economic Survey of India (2014), between 2004-05 and 2011-12, the workforce in agriculture and allied activities declined by 36 million, while the workforce in industry and services increased by 31 million and 20 million respectively. This trend is also manifested in a decline in the share of workforce in agriculture from 60% in 2000 to 47% in 2012 (World Bank). A consequence of the structural transformation is that the percentage of people living in urban areas has increased from 27% in 1997 to 32% in 2014 (World Bank). Given that informal insurance networks are weaker in urban settings as compared to rural, the argument for providing social insurance to protect workers against labour market risk becomes even stronger.

Moreover, India is becoming increasingly globalised6 , and hence, more vulnerable to external shocks7. Social protection in the face of external shocks acquires greater salience due to political economy reasons. Without social protection, an external shock that causes a decline in the incomes of a large number of voters (such as the East Asian crisis in the late 1990s) can cause a backlash against globalisation.

Existing social insurance programmes in India

India has various organisations to provide social insurance programmes of different types. The Employees State Insurance Corporation of India (ESIC) provides medical care, maternity, and work-related disability benefits for employees earning less than Rs. 15,000 a month in factories with 10 or more workers (20 or more in other establishments). It is funded by contributions from both employees (1.75% of wages) and employers (4.75% of wages). In 2005, the Rajiv Gandhi Shramik Kalyan Yojana (RGSKY) was added to the ESIC programme. RGSKY is the only form of traditional unemployment insurance in the country. Under this scheme, an insured person who loses his job after contributing premium for at least three years is eligible for unemployment allowance equivalent to 50% of wages per month for a maximum period of one year. While the benefits are moderate, the eligibility requirement of three years of contribution is quite stringent. Also, the ESIC requirement that the monthly wage must be less than Rs. 15,000 makes a lot of workers ineligible. The information on how many of the unemployed are actually getting unemployment allowance is not readily available8.

In 2006, the Government of India launched an ambitious programme of social protection for the rural poor under the MNREGA (Mahatma Gandhi National Rural Employment Guarantee Act). MNREGA guarantees 100 days of work to each rural household at the prevailing minimum wage. The design of this programme is different from the standard unemployment insurance programmes that provide temporary income to the unemployed while they are looking for a job. Rather, a person must work on a public project to obtain income. Therefore, it is more suited to the rural poor who cannot find employment all year round.

One may think that the traditional unemployment insurance schemes prevalent in developed countries are not suitable for a developing country like India. However, studies show that these programmes can be successful and enhance welfare significantly in developing countries (Gerard and Gonzaga 2012). As a result, several developing countries have launched unemployment insurance programmes (Holzmann et al. 2011). Table 1 below provides a comparison of the unemployment insurance programme in India with other BRICS (Brazil, Russia, India, China, and South Africa) members as well as Vietnam, a developing country slightly poorer than India that has launched an ambitious unemployment insurance programme. It is clear from the table that all these countries have more developed unemployment insurance programmes than India, in terms of coverage and generosity of benefits.

Table 1. Unemployment benefits in India and comparator countries

Country Coverage Qualifying condition Level of benefits Funding
India Factories with 10 or more workers; other organisations with 20 or more workers; employees earning less than Rs. 15,000 a month. Three years of contribution 50% of wage per month; for up to one year Employee: 1.75%; employer: 4.75% of payroll*
China All employees of urban enterprises and institutions At least one year of covered employment Determined by local governments: less than local minimum wage but more than the level of local public assistance benefit; benefit paid for up to one year with five years of coverage, up to 1.5 years with 5-10 years of coverage, up to two years with more than 10 years of coverage Employee: 1 % of gross earning; employer: 2% of payroll; provincial and local government subsidies
Russia All citizens of Russia At least 26 weeks of employment in last 12 months 75% of the previous average monthly wage for the first three months, 60% for the next four
months, 45% for the five months, and thereafter 30% of the local minimum subsistence level for a further 12 months
Federal and local government budget, using revenue from general taxation
South Africa All employees working more than 24 hours a month Must have sufficient credit before becoming unemployed.
For every six completed days of employment, eligible for one additional day of benefits, up to 238 days in the prior four-year period
38% to 68% of average earnings in the last six months, depending on the insured´s period of
Service; paid for up to eight months
Employee: 1% of earnings; employer: 1% of payroll
Brazil Persons employed in the formal private sector and other categories of workers such as household workers and fishermen Must have worked at least six of the last 36 months 80% of average earnings for average earnings up to 1,090.43 reais; 50% of average earnings plus a lump sum of 872.35 reais for average earnings of 1,090.44 reais to 1,817.56 reais; lump sum of 1,235.91 reais with average earnings greater than 1,817.56 reais. Benefit paid for three to five months, depending on the insured´s duration of coverage Earmarked taxes
Vietnam Vietnamese citizens with employment contracts of one to three years or permanent contracts who are employed by private and
public sector organisations with 10 or more workers
Must have at least 12 months of contributions during the last 24 months, must be registered as unemployed,
and must not have found a job within 15 days of registration
60% of the average monthly earnings in the prior six months paid for three months with 12 to 35 months of contributions, for six months with 36 to 71 months of contributions, for nine months with 72 to 143 months of contributions, or for 12 months with 144 months of contributions or more Employee: 1% of gross earnings; employer: 1% of payroll; government: 1% of insured’s gross monthly earnings and administrative cost

Note: *In India, the contributions are to the ESIC for medical, maternity, work-related disability benefits, in addition to the recently added unemployment allowance under RGSKY.
Source: International Social Security Association: Country Profiles.

In addition to traditional unemployment insurance, severance payments also allow workers to smooth consumption during unemployment. Mandated severance payment programmes require employers to make compensatory payments to employees upon separation. In India, workers in the organised sector working in firms employing more than 50 workers have access to a mandated severance payments programme. Minimum tenure required for eligibility is one year9. Upon separation, an employee earns the right to get 15 days’ salary for every year of service. Comparing the severance payments regime in India with other members of BRICS according to the generosity index computed by Holzmann et al. (2011), we find that India’s severance payments regime at 2.1 is more generous than that of South Africa (1) and Brazil (1.7), but less generous than China (4.3) and Russia (3.8)10.

Protecting workers, not jobs

As the new government in India is contemplating major labour market reforms, the issue of social protection of workers has to be kept in mind. It is true that several labour market policies are anachronistic and need to go. Among the issues of major concern is the firing restrictions imposed on firms. For example, according to the Industrial Disputes Act (IDA) of 1947 a permanent worker can be dismissed only for proven misconduct or habitual absence. Also, companies employing more than 100 workers must get government approval before they can fire workers or shut down. In practice these approvals are rarely granted. The IDA also stipulates that severance payments must be given to the fired worker, set at 15 days´ average pay for each complete year of continuous service. Usually all these regulations are lumped together under the broad term “employment protection”.

My recent research shows that not all elements of “employment protection” measures are the same. In particular, a distinction needs to be made between measures that ‘protect jobs’ and measures that ‘protect workers’. Government approval for firing would be a measure protecting jobs while severance payments would protect workers by providing them insurance. The former tends to reduce the efficiency of an economy while the latter, at appropriate levels, can increase efficiency. Therefore, as a part of labour reforms, while the government should remove ‘job protection’ it should strengthen ‘worker protection’.

In the absence of any formal unemployment insurance, it would be a good idea to extend severance payments to all firms. In fact, these payments should be viewed as an unemployment insurance programme, which is funded by contributions from firms whereby the contributions depend on the number of workers that a firm lays off. In certain situations they can be more efficient than the traditional unemployment insurance programmes funded by contributions from workers or a payroll tax on firms. If there is a concern about non-compliance by firms, an alternative is to collect a layoff tax which goes into a common pool that can be used to pay the fired workers. In fact, Blanchard and Tirole (2008) advocate an unemployment insurance programme funded by layoff taxes on firms. This would basically replicate a mandated severance payment programme.

While MNREGA provides insurance to rural workers, aside from a small unemployment insurance component in ESIC, no unemployment insurance programme exists for urban workers in India. Going forward, the government must set up a comprehensive unemployment insurance programme for urban workers, which can be funded by contributions from both firms and workers and possibly supplemented by revenues from general taxation. Instead of a payroll tax, the contribution from firms could be in the form of a layoff tax (similar to the ‘experience rating’ in the US) which would induce them to internalise the cost of firing workers. This would be an alternative to extending severance payments requirement to all firms and will also deal with the issue of possible non-compliance by firms with regard to their severance payments obligations More generally, the programme must be adapted to the circumstances prevailing in India11. For example, to alleviate moral hazard the contribution by employees can be made mandatory as in the case of China, South Africa and Vietnam, and ESIC in India. The current eligibility condition of three years of prior contribution in ESIC to get unemployment allowance is too stringent, and it needs to be changed to something like six months of work prior to becoming eligible, as in Brazil. Monitoring of job search may be difficult due to limited administrative capacity in a developing country context, and therefore, it may be a good idea to keep the job search rules for continuing eligibility flexible (Vodopivec 2013). While India currently lacks a comprehensive social security database to monitor the work history and benefit receipt of workers - an essential requirement for running a successful unemployment insurance programme, the completion of the Unique Identification (UID) scheme would be a big step in this direction.


  1. Moral hazard in the context of unemployment insurance arises from the fact that a worker receiving unemployment insurance will have less incentive to search for a job.
  2. Adverse selection refers to a situation where one party in a transaction has private information that can potentially affect the transaction adversely. Since workers that are more likely to lose their jobs have a stronger incentive to buy unemployment insurance, a private company offering such insurance will not survive.
  3. Covariant risk affects all parties involved in a transaction and arises out of factors that prevail on all parties equally. In a recession that leads to large-scale unemployment, a private company offering unemployment insurance will go bankrupt.
  4. In general, globalisation enhances welfare via lowered prices of imported inputs. However, imported inputs substitute for domestic workers thereby, lowering their wages and increasing unemployment. This increases labour-market risk and reduces welfare of risk-averse workers that do not have unemployment insurance. This reduction in welfare of workers can more than offset the gains from globalisation and hence, reduce social welfare overall.
  5. Severance payment refers to compensation that an employer provides to an employee when his/ her employment is terminated. It may include pay as well as other benefits.
  6. For example, trade to GDP (Gross Domestic Product) ratio has more than doubled from 25% in 2001 to 53% in 2013.
  7. Milani and Park (forthcoming) find that while in 1990 most of the volatility (about 70%) in output, inflation and interest rate in South Korea was driven by domestic shocks to preferences, technology, and monetary policy, by 2010-11 about 70% of the volatility was driven by external shocks such as terms of trade and exchange rate changes. Over the same period, South Korea’s trade to GDP ratio doubled from about 60% to 120%. One can expect a similar pattern in India as it becomes more globalised.
  8. Other prominent social insurance programmes are: (i) Employees’ Provident Fund Organisation (EPFO) that provides old-age disability and survivor pensions. (ii) Rashtriya Swasthya Bima Yojana (RSBY), which is a health insurance programme for Below Poverty Line (BPL) workers. So far, the programme has only covered about 10% of the target population.
  9. The eligible contingency for receiving severance payments is redundancy - laying off workers due to retrenchment, reorganisation, restructuring, structural changes, technological change etc., and disability.
  10. Generosity index is defined as average benefits (in weeks of salary) per year of service.
  11. See Vodopivec (2013) for a comprehensive discussion of designing unemployment insurance schemes in developing countries.

Further Reading

  • Acemoglu, Daron and Robert Shimer (1999), "Efficient Unemployment Insurance", Journal of Political Economy, Vol. 107, No. 5 (October), pp. 893-928.
  • Blanchard, Olivier and Jean Tirole (2008), "The joint design of unemployment insurance and employment protection: a first pass", Journal of the European Economic Association, 6:45-77.
  • Chetty, Raj and Adam Looney (2006), “Consumption smoothing and the welfare consequences of social insurance in developing economies”, Journal of Public Economics, 90: 2351-2356.
  • Chetty, R and A Looney (2007), ‘Income risk and the benefits of social insurance: evidence from Indonesia and the United States’, in Ito, T and A Rose (eds.), Fiscal Policy and Management, East Asia Seminar on Economics, University of Chicago Press, Chicago.
  • Gerard, F and Gustavo Gonzaga (2012), ‘Social Insurance under Imperfect Monitoring Labor market and welfare impacts of the Brazilian UI program’, Working Paper.
  • Holzmann, R, Y Pouget, M Vodopivec and M Weber (2011), ‘Severance Pay Programs around the World: History, Rationale, Status, and Reforms’, Social Protection Discussion Paper 1111.
  • Economic Modelling, Special issue: Current Challenges on Macroeconomic Analysis and International Finance Modelling.
  • Ranjan, P (2015), ‘Globalization, Jobs, and Welfare: The Roles of Social Protection and Redistribution’, CESifo Working Paper No. 5191, January.
  • Vodopivec, Milan (2013), “Introducing unemployment insurance to developing countries”, IZA Journal of Labor Policy, 2(1).
  • World Bank:
No comments yet
Join the conversation
Captcha Captcha Reload

Comments will be held for moderation. Your contact information will not be made public.

Related content

Sign up to our newsletter