akistan is growing old before it grows rich. The Population Census 2023 shows that around 6 percent of the population (more than 13 million) are 60 and above. Yet the one national scheme meant to protect private sector workers in old age, the Employees’ Old-Age Benefits Institution, still covers only a small fraction of the labour force and pays pensions that cannot keep pace with the cost of living. In the current situation, most Pakistani workers will reach old age with only their prayers to live on.

For decades, policymakers have responded with cosmetic fixes: ad hoc pension hikes, sporadic “drives” to register employers and endless seminars on strengthening social protection. The result is a structurally weak scheme that neither meets international standards nor offers credible security to the workers who finance it. The choice now is stark. Pakistan can continue to muddle through with an under-funded and under-covering EOBI or rebuild the scheme aligned with the International Labour Organisation and the International Social Security Association standards, with clear legal guarantees, sustainable financing and coverage that actually matches today’s labour market.

On paper, the EOBI is simple enough: a contributory, defined-benefit pension scheme for private-sector workers, created by the Employees’ Old-Age Benefits Act, 1976. Employers and workers pay contributions, and in return, insured workers receive old-age, invalidity and survivors’ pensions. But the legal design was crafted for a world where “the worker” meant a “male employee” in a formally registered factory or large establishment. The law applies only to firms with ten or more workers, with contributions set as a percentage of the statutory minimum wage, and coverage has never extended meaningfully to the informal, agricultural, domestic, gig or platform workers, street vendors, home-based workers, construction workers, self-employed workers and other informal sector workers. The result is dismal: while the total number of workers registered with the EOBI since its creation in 1976 exceeds 9 million, the actively covered workers (who are contributing or receiving pensions) are fewer than 4 million. Given Pakistan’s labour force of 83 million, this translates to less than 5 percent of the total workforce.

Coverage is not the only problem. Are EOBI pensions adequate, predictable and protected against inflation for those who qualify?

Again, the answer is troubling. The EOBI’s benefit formula is 2 percent of average monthly earnings (well, actually the minimum wage for unskilled workers) per year of service, with a statutory minimum pension. In principle, a worker with 30 years of contributions could receive 60 percent of the reference earnings, above the 40 percent minimum replacement benchmark embedded in ILO Convention No 102 for a standard case. The minimum EOBI pension stands at Rs 11,500 per month. When compared with the minimum wage for unskilled workers, it provides a replacement rate of less than 30 percent.

In practice, two things go badly wrong. First, many workers never accumulate enough continuous, registered service to reach that theoretical replacement rate. Short-term contracting, casualisation and non-registration of workers mean that a vast number of workers either fail to vest at all or qualify with very low credited service. Their problem is not a 30 percent vs 40 percent replacement rate; it is the absence of any contributory pension.

Second, for those who do qualify, EOBI pensions are eroded by inflation because adjustments are discretionary rather than rules-based. Pension revisions have historically been made on an ad hoc basis, driven by political decisions rather than a statutory formula linked to consumer prices or wages. There is no obligation in primary legislation for annual or periodic indexation, nor any transparent criterion for the size of increases.

From the viewpoint of ILO Convention No 102 and Recommendation No 202, this is a serious flaw. Both these instruments emphasise not only adequacy at the time of award, but also the regular adjustment of benefits to maintain purchasing power and a decent standard of living over time. A scheme that allows pensions to wither in real terms between irregular, politically expedient adjustments does not meet that test.

Before 2010, the federal government made several amendments to the EOBI Act 1976 through Finance Acts. However, in later years, the High Courts and the Supreme Court held that these amendments were ultra vires of the constitution, effectively restoring the legal position to its pre-2005 state. But that means that the total contribution payable to the EOBI is only Rs 170 (Rs 150 from the employer and Rs 20 from the worker), instead of Rs 2,400 per month. This makes EOBI severely underfunded and unsustainable.

Given this background, the Centre for Labour Research has been working on a draft to amend and re-enact the EOBI law to align it with international standards and the labour market’s needs.

Scope of “insured person”: ILO Convention No 102 mandates coverage of at least 50 percent of all employees or 20 percent of economically active residents. The Social Protection Floors Recommendation (No 202) calls for universal coverage, ensuring that all residents have access to at least basic social security guarantees. Under the current law, an insured person is essentially an employee in insurable employment (i.e., a contract of service/ apprenticeship). Under the amendment proposal, the “insured person” covers employees and self-employed persons. The definition of “self-employed” explicitly includes platform workers, online freelancers, street vendors, domestic workers, home-based workers and other informal sector workers. It also allows overseas Pakistani workers to be registered with the EOBI. Both the self-employed and overseas Pakistani workers can voluntarily register themselves with EOBI and self-contribute.

Under current law, there is no individual pathway for workers outside standard employment, as coverage depends on being an “employee” in the formal sector. The amendment proposal creates an explicit pathway for self-employed, overseas Pakistanis and informal workers to opt in. Moreover, it allows young workers aged 16 and above to register with the EOBI and start contributing.

Coverage Trigger for Establishments: The current law applies where ten (10) or more persons are employed (and can apply to smaller units if notified. It allows voluntary registration for establishments with fewer than 10 workers. The amendment proposal drops the application to five or more workers. Enterprises may apply for voluntary registration even if they have fewer than five workers. According to the 2023 Economic Census, 95 percent of the 7 million economic establishments in Pakistan have fewer than 10 workers. Therefore, it is essential to expand access to the formal scheme by lowering the applicability threshold.

Adequate Contributions and Benefits: Under the current law, “wages” are the basis for contributions. The system has historically suffered from ambiguity and erosion. The amendment proposal introduces/ clarifies “contribution base,” annually recommended by the board and approved by the federal government and ties the floor to the minimum wage for unskilled workers (in Islamabad Capital Territory), rather than allowing contributions to collapse to very low wages, as is the case now. The reform targets benefit adequacy and reduces contribution evasion/ under-declaration. Moreover, while the government’s contribution is discretionary under the present law, the amendment proposal requires that the federal government make a matching contribution (equal to the employer’s and insured person’s contributions), as was the case before 1995. In line with international standards, contribution levels should be sufficient to ensure the system’s financial sustainability. The current contribution levels are too low, even compared to regional levels. The amendment proposal suggests a phased increase in contributions from 6 percent to 10 percent over the next five years.

Currently, if there are gaps in contribution history (e.g., unemployment/ spells out of covered employment), the beneficiary might lose eligibility for benefits. There is no provision in the current law regarding entitlement loss due to intermittent work. The amendment proposal allows an insured person who leaves insurable employment to continue contributing (treated as insurable employment for scheme purposes) for a defined period.

Contracting and Subcontracting chains are where the informality thrives. The current legislation does not impose any responsibility on companies in this regard. The amendment proposal, on the other hand, makes the principal/owner liable and allows withholding the contribution amount from the contract sum and remitting it directly to EOBI.

Under the current law, the minimum pension is set by the federal government. Given that the nominal pension floor erodes rapidly with inflation, it should be pegged to a wage-linked base to prevent poverty in old age. In line with ILO Convention 102, the minimum pension is proposed to be re-anchored: it cannot be less than 40 percent of the contribution base (in today’s rates, it would not be less than Rs 16,000), and the Schedule clarifies that the pension cannot fall below the “minimum pension.” Though the amendment proposal does not include this, to save retired workers from impoverishment, there is a need to add a statutory indexation rule requiring annual adjustments to pensions by at least the increase in the consumer price index, or a combination of price and wage growth.

Women Face Multiple Disadvantages in Pakistan’s pension system: a lower retirement age (55 vs 60 for men) reduces contribution periods. Besides the lack of recognition for unpaid care work, lower labour force participation rates mean fewer contributors, resulting in significantly lower pension coverage and benefit levels for women. Under the current law, the gender-differentiated pensionable age exists (60 years for men; 55 years for women), reflecting older design choices. While this might seem protective of women, it actually results in lower pensions for women over the long run. The amendment proposal moves to equal pensionable age (60 years) with a transitional equalisation schedule for women from 55 to 60 over time (raising the retirement age for women by 6 months each year). The amendment proposal introduces the concept of care credits with 2 months of contribution credits for each child (6 months maximum for childcare), a dependent elder, or a person with disability, recognising women’s disproportionate unpaid care responsibilities. The maximum contribution credits for childcare, elderly care or care of PWDs cannot exceed 12 months. These care credits count toward both vesting requirements and benefit calculations.

Portability of benefits is limited under current law, as benefits can be stopped if an insured person (eligible for a benefit) is outside Pakistan. The amendment proposal adds the provision on payment of benefits outside Pakistan through banking channels.

A pension system that covers only formal-sector workers in large firms, pays inadequate benefits and is constantly at the mercy of politics (whether it should be devolved or not) is not just unfair; it is also economically inefficient. It encourages informality, undermines trust in public institutions and shifts the burden of old-age income security onto families and ad hoc charity. The population aged 60 and above is projected to double (to more than 25 million) by 2030. Rather than leaving the older people to their prayers, rebuilding EOBI on international best practices is a moral and constitutional imperative. It is the bare minimum of a just economy.

The authors work at the Centre for Labour Research, Pakistan. They were members of the drafting team for the Labour Codes in the Punjab and Sindh.