The real cost of poor employee mental health in India

The cost of poor employee mental health in India is not a line item you will find in any budget review, which is exactly why it keeps getting paid. It does not arrive as an invoice. It arrives as a developer who is at her desk but cannot concentrate, a team lead who quietly checks out months before he resigns, a high performer whose replacement takes a quarter to ramp. None of that shows up as a discrete charge. All of it is real money, and most organisations are already spending it without ever deciding to.

The useful question for a People leader or CFO is not whether to start paying this cost. You are paying it now. The question is whether it stays invisible, spread thinly across payroll and attrition and missed deadlines, or whether you measure it and decide what to do about it. This piece tries to make that invisible cost legible, using figures you can check, and grounded in Indian workforce realities rather than imported benchmarks.

When people picture the cost of poor mental health at work, they picture absenteeism: someone is unwell and does not come in. That cost is visible. It shows up in leave records, in shift gaps on the factory floor, in the scramble to cover a service desk. For shift-based and frontline roles, absenteeism is often the first signal HR sees, because attendance is tracked tightly and a missing person is immediately felt.

The larger cost is presenteeism: people who are physically present and on the payroll, but working at a fraction of their capacity because of anxiety, depression, burnout or sustained stress. Workplace-health research has repeatedly found presenteeism to be a bigger drain than absenteeism, often several times larger, precisely because it is continuous and unmeasured. Deloitte's UK analysis, for example, attributed substantially more of employers' mental-health cost to presenteeism than to outright absence. Treat the exact split as country-specific rather than universal, but the direction is well established: the part of the cost you cannot see tends to be bigger than the part you can.

In India this matters even more, because a salary on the books carries loaded costs around it. Provident Fund and ESIC contributions, gratuity accrual, variable pay and the overhead of managing a role all mean that a disengaged employee is not just costing you their take-home pay. You are funding the full loaded cost of a seat while getting a fraction of its output.

The World Health Organization estimates that depression and anxiety cost the global economy roughly 12 billion working days and around US$1 trillion in lost productivity every year. Those are population-level figures, and no single employer should pretend to map them onto a spreadsheet line. But they establish scale: this is not a soft, peripheral concern, it is one of the largest sources of lost output in the modern economy.

On the other side of the ledger, a WHO-led study published in The Lancet Psychiatry in 2016 estimated that scaling up treatment of depression and anxiety returns on the order of US$4 for every US$1 invested, mostly through improved productivity and recovered ability to work. That ratio is an estimate at the level of economies over a 15-year horizon, not a guarantee for any one programme, so it should be read as a signal of order of magnitude rather than a promise. Deloitte's recurring work on the employer cost of poor mental health points the same way: when firms measure their own absenteeism, presenteeism and attrition costs against the cost of support, the support has tended to more than pay for itself.

The honest caveat is that these are ranges and estimates, sensitive to assumptions. The right response to that uncertainty is not to dismiss the cost, it is to measure your own.

Of all the ways poor mental health shows up on the books, replacing people is the most concrete. Gallup and SHRM have long estimated the cost of replacing an employee at somewhere between roughly one-half and two times their annual salary, depending on seniority and how hard the role is to fill. For a senior engineer or a hard-to-source specialist, the upper end is realistic once you add recruitment fees, the manager hours lost to hiring, lost institutional knowledge and the months a replacement needs to reach full productivity.

Burnout sits directly upstream of this. The WHO's ICD-11 classifies burnout as an occupational phenomenon resulting from chronic workplace stress that has not been successfully managed, defined along three dimensions: exhaustion, increased mental distance or cynicism toward one's job, and reduced professional efficacy. Read that definition closely and it describes the exact arc of a regretted resignation. The efficiency drop is presenteeism. The mental distance is disengagement. The exit is your replacement cost crystallising. People rarely leave in a single dramatic moment, they leave gradually and then suddenly, and the bill lands long after the warning signs were visible.

This is why mental health and retention cannot be treated as separate programmes. The same underlying condition that quietly lowers output is the one that eventually walks out of the door.

If this cost is so large, why does it stay off the radar? Partly because the signals are quiet by nature. Edmondson's research on psychological safety, echoed by Google's Project Aristotle study of what makes teams effective, found that the strongest teams are ones where people feel safe enough to admit they are struggling. The corollary is uncomfortable: in workplaces without that safety, distress goes underground. Nobody is filing a ticket that says they are burning out. They are simply producing less, and you are paying full price for it.

There is also a measurement problem specific to mental health, and it is partly a trust problem. Asking employees to self-report wellbeing only works if they believe the data will not be used against them. In India, the Digital Personal Data Protection Act of 2023 sets out the ground rules that make this trust possible: consent, purpose limitation, and clear rights for the person the data is about. Handled well, that framework is not a constraint on measuring wellbeing, it is the precondition for measuring it honestly. People share truthfully when they can see that responses are anonymised, aggregated and used to change the workplace rather than to rank individuals.

Get the trust right and the hidden cost starts to become visible. Get it wrong and you are left guessing, while the spending continues.

None of this argues that mental health support is charity, or that it should be justified on sentiment. The argument is narrower and more durable: the cost is already real, already paid, and currently undecided. It sits inside your payroll as reduced output, inside your attrition as replacement spend, and inside your missed timelines as work that quietly did not get done. The only choice in front of a People leader or CFO is whether that cost remains an accident or becomes a decision.

Tools like Aha exist to help organisations make that cost visible and act on it within a framework employees can trust, with anonymised, aggregated insight rather than individual surveillance. But the first move does not require any vendor. It requires putting an honest number on what poor employee mental health is costing your organisation in India today, and then choosing, with eyes open, what to do with that number.