EPFO, ESI overhaul required to unlock India's formal jobs potential

This column has been updated to reflect a rejoinder and the authors’ clarifications 

---

  In 1958, Mao Zedong’s Great Leap Forward included a quirky “Four Pests” campaign to eliminate flies, mosquitoes, rats, and sparrows. Research suggests that this culling of 2 billion sparrows, which ate locusts and rice borers that attack crops, resulted in the unintended deaths of 2 million people. This is an extreme case, but it highlights the second-order effects of noble policy intentions under the licence raj, which have led to the unintended absence of at least 10 million formal employers and 100 million formal employees. 

Two licence-raj legacy institutions — they are not government-funded — that have contributed to this sabotage of formalisation are the Employees Provident Fund Organisation (EPFO) and the Employees State Insurance Corporation (ESIC). Indian employers quietly honouring the ₹40,000 crore cost to comply with the labour Codes should earn them radical reforms to the EPFO and ESI, such as  creating an Aadhaar-linked, fully portable, lifetime universal social security account that gives employees greater choice. 

The ESI and EPFO deserve radical reform for similar reasons: Poor policy outcomes on employer and employee coverage; high dissatisfaction of employers and employees; sabotaging formal employment with a flawed design out of touch with a cost-to-company compensation world;  technology systems inadequate for a world where employment has shifted from a lifetime contract to a taxicab relationship; excessive administrative charges relative to benchmarks; weak governance with dozens of people on their unrepresentative boards where “anybody can say no and nobody can say yes” to proposals for improvement. Reforming social security in any country is complex, but the 10-year delay in implementing a Union Budget announcement for radical reform of both organisations makes reform even more urgent. Let's dive into each organisation.  

The EPFO has failed to deliver work-based social security; it covers only 2 per cent of our employers (63 million) and 13 per cent of employees (560 million). Employer surveys cite the EPFO as one of the top three pain points due to corruption, poor employee service, administrative costs collected separately from employers (₹8,000 crore), and high charges that defy benchmarks (4 per cent vs National Pension System, or NPS 0.03 per cent), bloated by an irrational calculation formula based on wages rather than contributions.  

Employee dissatisfaction is high; 78 per cent of the 325.6 million subscribers are non-contributing due to the employer-linked design and poor technology adoption, lower returns than NPS, and an involuntary diversion of a substantial portion of the contribution to the sub-account of the Employees’ Pension Scheme (EPS), whose birth defect of defined benefits and defined contributions (one must be variable) suggests a deficit of ₹3 trillion that can only be covered by a fiscal contribution or benefit reduction (maybe why 75 per cent withdraw this lifetime benefit within four years).  

The ESI has failed to deliver work-based healthcare financing; less than 1 per cent of employers and 6 per cent of employees are covered. Employees vote with their feet; only 50 per cent of ESI’s 32 million contributors use their services. Most painfully, ESI overcharged employees and employers to generate a ₹8,900 crore surplus last year via a low 63 per cent payout ratio (this would be illegal in most countries, where payouts below 85 per cent trigger a refund to contributors by law) and ₹3,000 crore in high administrative charges (calculated at 15 per cent of contributions versus 2 per cent global and local benchmarks) separately collected from employers. ESI’s undercoverage and low benefit payout have huge human capital costs.  

The pathologies suggest clear reforms for both programmes. The EPFO should offer employees a choice: Pay their 12 per cent employer contribution to either the EPFO or the NPS, and decide among three levels of employee contribution (0/6/12 per cent). It should cut employer administrative charges to NPS levels (4 per cent to 0.03 per cent). The EPFO should explicitly clarify that no coverage applies to employees above the Act’s salary threshold (₹15,000 per month) by allowing voluntary discontinuation or withdrawal above that threshold. It should restructure the EPS by filling the funding hole, re-evaluate promised benefits, give employees the option to opt out by directing their monthly contribution to their defined contribution account, and converge with Atal Pension/other schemes. Employees must be given the option to link their accounts to Aadhaar rather than Universal Account Number (UAN). The lowest-hanging reforms are operational: Strengthening investment operations by linking returns to earnings, mandating service-level benchmarks (SLBs), integrating accounts with citizen DigiLocker and Entity DigiLocker, reducing the board size from 42 to 12 members, and moving away from digitisation road maps involving PDFs to application programming interface (APIs by aligning with the paperless and presenceless design of digital public infrastructure.  

The ESIC should give employees a choice to divert their monthly health insurance contributions to insurance regulator-licensed insurers. It should reduce administrative costs charged to contributions from 15 per cent to 2 per cent by benchmarking them against regulator-licensed insurers. It should immediately reduce premiums by 30 per cent to reduce excess collection, and mandate future payouts of 85 per cent of the contribution; failing which, the excess should be refunded.  

The previously collected ₹1.5 trillion surplus should be used to provide a two-year contribution holiday for micro, small and medium enterprises (MSMEs), women, backward states, and Scheduled Castes/Scheduled Tribes. All employee accounts should be linked to Aadhaar for portability (backpack benefits), and all employer compliance must move from PDFs to APIs to enable straight-through processing. The most difficult but impactful reform would be cutting the board size from 59 to 12 to ensure governance targets stakeholder - employer and employee - welfare rather than self-interest.  

The EPFO and ESI have resisted radical reinvention since 1991, illustrating how self-interested bureaucracies preserve their own interests by creating improbable scenarios of chaos. Sir Humphrey Appleby — a civil servant in the BBC Sitcom Yes Minister — explained to his deputy that the best way to deter a politician from taking a radical course of action is to tell him that this decision is courageous. He said, “Calling it controversial only means that the policy will lose you votes. Courageous means that it will lose you the election.” The ESI and EPFO have made significant contributions to India’s development, but it's time to update their mandates, incentives, and technology. The political upside of courage is the duas (blessings) of 1 million current employers and 100 million employees who are hostages, not clients. But the economic upside is millions of new formal employers and employees. 

---

The authors are, respectively, board member and research head at Teamlease Services 

---

 

POINTS OF VIEW 

In response to this opinion piece, Dhanraj Balakrishna of Safe in India Foundation provided additional context to its arguments. In order to facilitate greater understanding, Business Standard shared Mr Balakrishna's views with Mr Sabharwal and Ms Patil, who explained the background in which they made their arguments. The said rejoinder and the authors’ clarification are as follows: 

---

 

Rejoinder received from Mr Balakrishna

 

The opinion column in Business Standard, dated 27 April 2026, makes several claims on ESI that need additional context.

 

Coverage: The claim that ESI reaches less than 1% of employers and 6% of workers uses the entire workforce as its denominator, a figure that includes the self-employed, informal workers, government employees, those above the wage ceiling, and establishments below the statutory threshold, none of whom ESI is designed to cover. 

 

ESI currently covers 3.84 crore insured persons and an estimated 14.91 crore beneficiaries, including dependents. PLFS 2025 estimates regular wage/salaried workers at 23.6% of all workers. Against even this broader relevant denominator, ESI insured persons are around 26 per cent of regular wage/salaried workers. Once higher-wage workers, workers in establishments below the ESI employee-count threshold of 10 employees, excluded geographies, and other excluded groups are removed, the real share would be even higher. 

 

Payout Ratio: The cited 63 per cent ratio is not current. Following the reduction in contribution rates from 6.5 per cent to 4 per cent in 2019-20, the benefit-to-contribution ratio ranges from 83 per cent to 104 per cent. For comparison, in private health insurance, the 2024-25 incurred claims ratio was 88 per cent for private-sector general insurers and 69 per cent for standalone health insurers (Irdai Annual Report)

 

Proposals for a 30 per cent contribution cut or a two-year holiday must be assessed against current outlays and future liabilities. Critically, the surplus in the ESI corpus belongs to workers, not to employers. If deployed at all, it must be deployed for worker benefit, not to subsidise employer costs as recommended in the opinion piece. 

 

Administration Expenses and Employer Contributions: Comparisons of ESIC’s administration expenses with a 2 per cent insurance benchmark (private insurance or global standards) are misleading. For standalone health insurers in 2024-25, Irdai shows operating expenses of about 16 per cent, excluding commissions. 

 

ESI's statutory contribution is 4 per cent of wages, 3.25 per cent from the employer and 0.75 per cent from the worker. There is no separate 15 per cent administrative charge on employers as indicated.

 

Privatisation: ESI is not hospitalisation insurance alone as the article implies. It covers OPD, medicines, sickness benefit, maternity benefit, disablement benefit, dependants' benefit and occupational-injury protection. Any private substitute must replicate this full statutory package. 

 

It must also account for workers' socioeconomic vulnerability, private insurers' well-documented tendency to deny claims, and the reality that employer compliance depends on regulatory compulsion, and not choice architecture. 65% of 15,000+ workers assisted by Safe in India did not even get their ESIC cards before their injuries. In such a setting, “choice” will become the employer or contractor's choice to deny insurance altogether. 

 

Compliance and Employer Responsibility: Employer responsibility is largely missing from the article, and if this is addressed, ESI and workers can do better. Safe In India Foundation has found that 16 per cent of injured workers were registered only after their injuries. Compliance is already failing workers. Proposals that weaken the mandatory structure and hand over poor people’s health and benefits wholesale to private sector will worsen, not resolve, this, though there are indeed several processes that can be partially privatised to improve ESIC’s effectiveness and challenge the current norms of service. 

---

 

The authors’ clarification 

Coverage: Our article used the broader denominator of total workers because our argument relates to the overall reach of formal social security within India’s labour market, which is the master of ESI. Its ingenuity to execute on its coverage mandate cannot lead to a reduction of the original rationale for the organisation of employment-based financing. particularly in the context of increasing informalisation, contractualisation and fragmented employment relationships. ESI workers still represent a relatively small share of India’s total workforce, despite that being the mandate when it was formed. The article was referring specifically to direct contributing workers (insured persons) as a share of India’s total labour force, not total beneficiaries including dependents. 

 

Payout ratio: Aggregate expenditure-to-contribution ratios have increased materially following the reduction in ESI contribution rates from 6.5 per cent to 4 per cent in 2019-20. However, the article’s analysis was intended to focus on the core beneficiary payout rather than the total institutional expenditure booked in ESIC accounts. 

 

Our approach distinguished between: 

 

• direct healthcare and cash benefits received by insured workers and 

 

• broader institutional and operational expenditure embedded within the medical benefits category in ESIC accounts. 

 

According to the ESIC Annual Accounts 2024-25, total medical benefits expenditure was reported at about ₹12,467 crore. However, this category also includes expenditure related to repairs and maintenance, equipment and infrastructure, establishment and operational overheads, training/education, depreciation and other institutional costs. For analytical comparability with health-insurance payout ratios, we therefore focused on estimating core medical consumption by including hospitalisation/treatment, medicines, diagnostics, direct patient-care expenditure and cash benefits/transfers while excluding non-core expenditure components. 

 

Since ESIC accounts do not separately disclose an adjusted core medical expenditure figure, a conservative analytical adjustment for institutional and non-core expenditure components suggests that core medical benefits are likely lower than the headline medical-benefits figure reported in the accounts. On this basis, the estimated core medical benefits are about ₹10,000 crore. After adding cash benefits of around ₹2,541 crore, estimated core beneficiary benefits remain significantly below total contribution income of about ₹20,255 crore, reinforcing the article’s broader concern regarding the proportion of contribution income that is ultimately translated into direct healthcare and income-support benefits for insured workers. This comparison is intended as a flow-based assessment of annual utilisation, not an actuarial assessment of long-term liabilities. 

 

The article’s broader concern was not about total institutional expenditure alone but about the proportion of contribution income ultimately translating into direct healthcare and income-support benefits for insured workers. The larger reform argument therefore relates to utilisation efficiency, contribution calibration, governance, portability and administrative cost structures within a worker-funded social security framework. 

 

We also agree that the ESI corpus fundamentally belongs to workers and should be deployed for worker welfare. The suggestion regarding calibrated contribution relief for MSMEs, women, backward states and SC/ST workers was intended as a labour-market formalisation measure within a worker-welfare framework, since contribution costs directly affect formal hiring incentives and worker formalisation outcomes.

 

Administrative expenses and employer contributions: The statutory contribution structure under the ESI framework currently remains 4 per cent of wages, comprising 3.25 per cent employer contribution and 0.75 per cent employee contribution. However, this clarification does not materially alter the underlying analytical concern of a separate statutory levy, but the effective utilisation efficiency of the contribution pool collected under the ESI framework. The analysis specifically evaluates the proportion of aggregate collections and institutional expenditure that is absorbed by administration and overheads, as opposed to direct medical care and insured-worker benefits. 

 

The observation compares ESIC against private standalone health insurers and cites Irdai operating expense ratios of about 16 per cent excluding commissions. However, such a comparison requires contextualisation for three structural reasons: 

 

• First, private commercial insurers operate under a fundamentally different model involving distribution costs, acquisition expenses, underwriting, marketing expenditure, intermediary commissions, profit margins and product-level risk pricing. ESIC, by contrast, is a compulsory social-insurance institution with statutory enrolment, payroll-linked collections, sovereign backing and a captive contribution base. Therefore, the administrative efficiency expectations applicable to ESIC should reasonably be higher than in fragmented commercial insurance markets.

 

• Second, the article’s benchmark was intended to reflect efficient public/social-insurance administration rather than private commercial insurance economics. International social-insurance systems — particularly those operating at scale with integrated digital administration — often function within a lower administrative overhead range, generally around 2%, notwithstanding differences in institutional structure. 

 

• Third, the issue is not whether ESIC performs better or worse than private insurers in absolute terms, but whether there exists scope for significant rationalisation of administrative expenditure through greater digitisation, consolidation of administrative layers, streamlined compliance architecture, improved procurement systems, technology-led claims and records management, integration with broader public-health infrastructure and reduction in institutional duplication.

 

Privatisation and compliance: The observation correctly points out that ESI is a far broader social-security arrangement than conventional hospitalisation insurance. ESI provides not only medical treatment and medicines, but also sickness benefit, maternity benefit, disablement benefit, dependants’ benefit, occupational-injury protection and wage-linked income support. We fully recognise the importance of preserving these worker protections, particularly for vulnerable and lower-income workers. The proposal sought to introduce greater employee choice, portability, accountability and service competition within a regulated framework, while preserving mandatory employer contributions and core worker entitlements. The argument is therefore not for replacement of ESIC, but for calibrated competition and choice within a mandatory social-security floor. Any alternative arrangement would necessarily need to replicate the statutory benefit package and remain subject to regulatory oversight and minimum standards.

 

Compliance and employer responsibility: We agree that weak employer compliance is a serious concern. Evidence that workers are often registered only after injuries highlights implementation failures that harm workers and undermine the objectives of social security. However, we would argue that these compliance failures also demonstrate the limitations of the current architecture and strengthen the case for portable, Aadhaar-linked, technology-enabled systems with real-time registration, straight-through processing and greater worker visibility over contributions and entitlements. Mandatory contributions, enforcement and worker protections remain essential. The proposed reforms were aimed at improving outcomes for workers through better governance, portability, competition in service delivery and stronger accountability mechanisms. We agree that there is significant scope for selectively expanding private participation in areas such as healthcare delivery, claims processing, diagnostics, technology infrastructure and service standards without undermining the core social-security objectives of ESI.