NPS Tier 1 vs Tier 2: Which One is for you?

Editor’s Analysis

I should say plainly that I do not hold NPS myself. That is a decision about my own finances at this stage rather than a verdict on the product — I value the flexibility more than I currently value the deduction, and I would rather be honest about that than pretend otherwise. But my view of the scheme is genuinely positive, and I think it deserves to be far more popular with people entering the workforce now than it is. Read the table again and notice that the lock-in and the annuity, which are listed as costs, are also the mechanism. They are the reason the money is still there in thirty years.

The design supports that. Under Auto Choice the equity share steps down on a fixed schedule without you having to do anything — LC75 holds up to 75 per cent equity until age 35 and tapers towards roughly 15 per cent by 55, LC50 is the default at half that, and the Balanced Life Cycle option introduced in 2024 keeps 50 per cent equity until 45 before it starts stepping down. Fund management charges are capped at 0.09 per cent of assets, an order of magnitude below what most mutual funds cost. So you are getting professional management, an equity taper that happens whether or not you remember to do it, and a tax break, at close to no cost.

That combination matters more now than it did a decade ago, not less. Geopolitics, currencies and job security have all become less predictable, and the generation starting out today has more ways than any before it to take risk and fewer default mechanisms forcing it to hold a safe base layer underneath. A regulated, professionally managed retirement account that you cannot raid in a bad month is one of the few such mechanisms left. I would like more young salaried people to have one, even if I have not yet opened mine.

Editor’s Analysis

My own employer does offer corporate NPS, and it sits inside the usual flexible-benefit menu alongside meal cards, fuel and telecom components and a driver-salary head. That menu got materially more valuable this year. The Income-tax Rules, 2026, notified in March and in force from 1 April 2026, lifted the per-meal exemption from Rs 50 — a figure last set in 2001 — to Rs 200, and, more importantly, did not carry forward the proviso that had blocked the exemption for anyone on the New Regime. Fully used, that is up to about Rs 1.05 lakh a year of salary taken tax-free under either regime. The chauffeur component in the perquisite valuation moved from Rs 900 a month to Rs 3,000 in the same exercise. These are not trivial sums once you add them to the 14 per cent employer NPS route.

The problem I see is not what gets offered. It is what gets used. Employers increasingly route the whole menu through a single multi-wallet card — Pluxee, the employee-benefits business formerly branded Sodexo, is one such provider — where the employer loads separate meal, fuel, telecom, driver and reward wallets onto one instrument and the employee spends from whichever applies. In principle that is a clean design. In practice, whether an employee gets any of it depends almost entirely on how much work the employer is willing to do: whether the declaration window is explained rather than emailed, whether the wallets are pre-configured sensibly, whether anyone tells the employee that a component left undeclared simply reverts to taxable salary. Where HR treats the flexi-benefit declaration as a form to be collected, most people leave the default and quietly lose the benefit.

Corporate NPS is the sharpest case of this because it is the one item on the list the employee usually has to initiate. It rarely appears as a pre-ticked box; someone has to ask payroll to restructure a slice of CTC into an employer contribution. That single administrative step — not the lock-in, not the annuity — is why the take-up in most companies is lower than the tax case would predict.

Editor’s Analysis

In my experience the reason NPS uptake stays low splits cleanly by income, and the two halves have completely different causes — which is why a single explanation never works. Higher up the salary scale, the refusal is informed. People in the 30 per cent bracket understand the tax case perfectly well and decline anyway, because for them the lock-in to 60 and the compulsory annuity are the binding constraints, not the deduction. A 32-year-old is being asked to accept 28 years of illiquidity, and at the end of it to convert at least 40 per cent of the corpus into an annuity at whatever rates prevail then. At today’s rates that is roughly 7.5 to 8.1 per cent for a plain life annuity and around 5.7 to 6.4 per cent if you want the purchase price returned to your family — and that income is taxable at slab in the year it is received. Someone who already has an EPF corpus, a home loan and equity exposure looks at that and reasonably concludes they can build a better-shaped retirement themselves.

Lower down the scale it is not a considered rejection at all. It is that nobody has explained it. The employee has never had the difference between Tier 1 and Tier 2 laid out, has confused the annuity with an insurance product, and has no way to weigh a deduction against a lock-in because nobody in their working life has done that arithmetic in front of them. You can see the size of that gap in the base: there are only around 90 lakh non-government NPS subscribers, in a salaried population many times larger. PFRDA’s chairman has spoken of taking that to 35 to 40 crore within five years — treat that as a statement of ambition rather than a projection, but an ambition on that scale only makes sense if the regulator also believes the product has never properly been put in front of most people. That matches what I see.

The practical implication is that these two groups need opposite interventions. The higher earner needs an honest account of the exit terms so they can decide against it if they want to. The younger, lower-bracket employee needs someone to sit down and explain what the account actually is — and that is a job employers are better placed to do than anyone, and mostly do not.