Search

Beyond the Iron Rice Bowl: What GovTech's 93 Job Cuts Really Tell Us

2 days ago
7 min read

When GovTech announced it was letting go of 93 staff in the first phase of a shift to a new operating model, the story travelled fast — and the reactions travelled faster. Within days, the conversation had moved well past GovTech itself: comment sections were relitigating whether the civil service is still an "iron rice bowl," whether the retrenched staff had been treated fairly, whether age had quietly played a role given that 42 per cent of those affected were in their forties, and whether this was a sign that Singapore's economy — and AI's march through white-collar jobs — is in worse shape than we'd like to admit.

Every one of those reactions is understandable. None of them, on its own, quite captures what's actually going on. Worth saying upfront: GovTech's own account is that this was not an AI-driven or cost-cutting move, but a shift from project-based delivery to in-house product ownership — the agency "changing shape, not shrinking," in its chairman's words. Whether that framing fully satisfies the 93 people affected is a separate, fair question. But it's worth holding both readings — the official rationale and the public's discomfort — rather than assuming one cancels the other out.

The iron rice bowl left the building years ago

The idea that a civil service job comes with a job-for-life guarantee has been fading quietly for close to two decades. Term contracts — often for two years, with extension at the ministry's discretion — have long been a standard entry point into public service, not an exception. Public Service Division's own figures from the late 2000s show that even then, well under half of new hires went straight into permanent posts; the rest were on contracts designed explicitly as a "mutual evaluation period" before either side committed further. None of that is new, and none of it is secret. What's different this time is scale and visibility: 93 people let go in one announcement, covered by name-brand media, is simply a much louder version of something that has been happening for years. The surprise, in other words, may say more about how invisible the shift has been than about how sudden it is.

A real break — just not the one people think

Where GovTech's announcement is genuinely unusual is here: the civil service has, by long-standing practice, avoided retrenchments even when belts tighten elsewhere. Civil service attrition — voluntary departures — has typically run below the private sector; the most recent figure we could verify with confidence is a 9.9 per cent resignation rate for the Management Executive Scheme in 2021, itself a ten-year high. In leaner years, the public service has at times been a source of additional hiring rather than cuts, partly to cushion the broader labour market. Retrenchment, as a tool, sits outside that pattern — which is probably the real reason this story landed with so much force. It's less "the civil service broke a promise of lifetime employment" (that promise was already gone) and more "the civil service broke its own habit of not doing what the private sector does in a downturn."

Worth pausing briefly on why, though, before reaching for the obvious answer — because it's tempting to read GovTech's cuts as a proxy for a broader economic downturn, and I don't think the data supports leaning on that. Take GovTech at its word: its chairman was explicit that this was not an AI-driven or cost-cutting move. Singapore's aggregate labour market doesn't look like one under real strain either — employment has grown for nineteen consecutive quarters, resident unemployment sits at 2.9 per cent (historically low), and job vacancies still outnumber jobseekers. Retrenchments did tick up to a five-year high in Q2 2026, but stayed well below Global Financial Crisis or pandemic levels. That's not to say individual anxiety isn't real — job security surveys do show many Singaporeans feeling less secure than the headline numbers suggest, and that gap between sentiment and statistics is a genuine story in its own right — but it's a different story from this one, and I'd rather not blur the two. The more durable explanation for what's happening at GovTech, I think, sits somewhere else entirely: in the demographics.

The quiet arithmetic behind the scenes

A year-by-year table drawn from Department of Statistics civil service headcount data and Ministry of Manpower resident labour force figures, running from 1991 to 2025 (with a handful of gaps in census years), and it's precise enough to build the real argument on rather than an approximate one. Civil service headcount as a share of the resident labour force fell sharply through the 1990s and into the mid-2000s — from 4.52 per cent in 1991 to a low of 3.43 per cent in 2006 — and has since settled into a much narrower band, roughly 3.5 to 3.8 per cent, that has held for close to twenty years. In 2025 the ratio stood at 3.61 per cent — near the middle of that band, and in fact slightly higher than 2022 and 2024, both 3.55 per cent. Worth being clear about what this does and doesn't show: civil service headcount hasn't been shrinking. It grew from 62,062 in 1991 to 88,664 in 2025, a 43 per cent increase. It has simply grown more slowly than the resident labour force it's benchmarked against, which nearly doubled over the same period, from 1.37 million to 2.46 million. That's consistent with a long-standing and fairly disciplined policy: don't let civil service headcount grow faster than the resident labour force — a labour force made up almost entirely of citizens and permanent residents, since civil service hiring doesn't draw on foreign labour the way much of the private sector does.

Here's the part that matters more looking forward than looking back. Resident labour force growth has decelerated sharply in recent years — from an average of roughly 2 to 2.4 per cent a year across the 1990s and 2000s, down to around 0.6 per cent a year between 2021 and 2025. That's the demographic story most of us already sense in daily life: an ageing population and a low birth rate translating into a resident labour force that is still growing, but only just. It lines up with the wider picture — resident labour force participation has declined for four straight years, the share of residents aged 55 and above has climbed from about 30 per cent in 2015 to nearly 38 per cent in 2025, and the old-age support ratio (working-age residents per retiree) has fallen from 5.7 to 3.3 over the past decade, projected to fall further to 2.7 by 2030. If that deceleration continues, holding civil service headcount growth below the resident labour force's growth rate — the same policy that has kept the ratio in a stable band for two decades — becomes a materially tighter constraint than it has been for most of the last thirty years. To be fair to the data, that tightening hasn't visibly happened yet: 2025's uptick to 3.61 per cent was driven by civil service headcount growing faster than the resident labour force that particular year, which cuts the other way. What I'm describing is a reasonable read of where the underlying trend is headed, not something the historical ratio already confirms. But if resident labour force growth keeps sliding toward flat, the same fiscal discipline that has held this ratio steady for two decades would eventually require civil service headcount to grow more slowly still, or shrink outright, simply to stay within it. GovTech's cuts may be an early, visible instance of an agency starting to plan around that eventuality — that reading is my own inference, not something a source confirms — and it's a more durable explanation, I think, than reaching for AI or a weak economy.

None of this erases the human cost to the 93 people affected, or the discomfort of watching a familiar institution behave differently than expected. Those concerns deserve to be taken seriously on their own terms, not waved away by population charts. But it's worth naming what this isn't: a betrayal of some old guarantee. The civil service hasn't been an iron rice bowl for at least two decades, given how long term contracts have been standard practice. What it has offered, and by most of the evidence above still does, is closer to a hard plastic rice bowl — meaningfully more secure than most of the private sector, given how rarely the public service has resorted to retrenchment, but never actually unbreakable. GovTech's cuts are the clearest illustration yet of where that boundary sits, not proof that the boundary just moved.

And if this is indeed the leading edge of a broader, quieter recalibration across the civil service, a smaller headcount needn't only be read as decline. Fewer people can mean fewer layers between a decision and the person who has to live with it: fewer approvals to chase before something ships, faster calls made, and a clearer line back to whoever made them when things go right — or just as usefully, when they don't. Speed, flexibility, and accountability aren't consolation prizes for a shrinking organisation; they're plausible upsides of one, if the redesign is done with intention rather than left to happen by attrition alone. Whether GovTech and the wider civil service manage that well is, again, a genuinely open question — but it's one worth asking honestly, rather than assuming a smaller public service is automatically a worse one.

The more useful question for the rest of us, mid-career professionals in particular, may not be "is my job as safe as I assumed" — for civil servants and private-sector employees alike, the honest answer has for some time been "safer than it looks, but not guaranteed." It's whether we're building the kind of judgment, adaptability, and comfort working alongside AI tools that makes us valuable regardless of who signs the paycheque, or how many layers of approval sit above us. That's a shift in mindset the civil service is now navigating as publicly as anyone else — and one every organisation, ours included, is trying to get right.