Four predictions for tonight's National Day Rally

At 8pm tonight, Prime Minister Lawrence Wong will deliver the National Day Rally in English at ITE College Central in Ang Mo Kio. It is his third Rally. I think it will be the family Rally.

I am going to make four specific predictions about what he will announce, and a fifth about how he will end. A vague prediction cannot be marked, so each one is specific, and each comes with its reasoning. These are opinions built on inference from published numbers and from what ministers have already said in Parliament, and I may well be wrong by 10pm.

The numbers first

Singapore recorded about 27,500 resident births in 2025. That is a fall of roughly 11 per cent from the 30,808 births in 2024, and the lowest number in our recorded history. The resident total fertility rate fell from 0.97 to 0.87 in a single year. A decade ago, in 2015, it was 1.24. Replacement level is 2.1.

The Government has not been quiet about what this means. In February, Deputy Prime Minister Gan Kim Yong told Parliament that without new interventions, the citizen population may start shrinking by the early 2040s, and that Singapore plans to take in 25,000 to 30,000 new citizens a year over the next five years. In April, an inter-ministry workgroup on marriage and parenthood was set up. In his National Day Message this month, PM Wong said a major review of how Singapore supports families is underway.

Governments do not announce major reviews in August and then say nothing at the Rally two weeks later.

If you are my student, you have seen this essay before

Every J2 student I teach can recite the consequences of an ageing population, because it appears in the A-Level syllabus and it appears in past papers. Tonight the essay happens in real time on national television.

Walk through the chain slowly, one link at a time. The 27,500 babies born in 2025 are the workforce of the 2040s and 2050s. A smaller cohort of workers will support a larger group of retirees, so each worker carries more of the cost of supporting the old. A smaller labour force also lowers potential growth, because the quantity of labour is one determinant of long-run aggregate supply. At the same time, more elderly residents raise government spending on healthcare and eldercare, while a smaller workforce narrows the income tax base that pays for it. Singapore also became a super-aged society this year, which means more than one in five residents is 65 or older.

So the state faces rising expenditure and a shrinking base to fund it, decades ahead of time, with near certainty. Exam answers list three standard responses: raise productivity, import labour, and raise the birth rate. The Government is already acting on the first two. The 25,000 to 30,000 new citizens a year is the immigration response, already operating at scale. Tonight, I think, is about the third.

Lee Kuan Yew was famously sceptical of the third response. In Hard Truths he said he would offer a baby bonus worth two years of salary just to prove that money would only have a marginal effect on fertility, because the causes were changed lifestyles and mindsets. He may still be right about cash. But cash incentives and structural conditions are different things. A S$16,000 grant does not change the fact that your family of six is deciding whether it fits into 110 square metres. All four of my predictions are structural for that reason.

The conditions a larger family actually faces

Judge tonight's speech by whether it deals with what larger families are actually up against.

Start with space. A typical 5-room HDB flat is around 1,000 square feet. A family of six plus a helper is 7 people, which works out to about 142 square feet per person. A two-person household in a newer 900 square foot flat near town has about 450 square feet per person, three times as much, in a newer flat, in a better location.

Space affects daily stress, privacy, the study environment and long-term liveability. If spatial conditions are tight from the outset, the marginal decision to have a third or fourth child becomes materially harder. Incentives shape behaviour. Among the young couples around me, including former students who are now married adults, a recurring choice is to have no children and become paw-rents instead. When I ask why, the answer is consistent: housing is expensive, and raising children feels expensive.

Costs jump again at Primary 1. A teacher friend of mine paid about S$300 a month for childcare after subsidies. When his child entered Primary 1, school itself became close to free, and student care rose to about S$1,300 a month, because the subsidy structure changes at that boundary. His reaction was direct: how does one justify more than one child under those conditions? For a family with three or four children, the same costs compound across the household.

CDC vouchers are issued per household. A ten-person household receives the same amount as a two-person household. The policy intent is clear and the administration is simple. The arithmetic is simply amusing.

Prediction 1: larger flats for larger families

I think the Government will announce a new class of BTO flats sized between today's 5-room flat and the jumbo flats of the 1990s, which puts the range at roughly 125 to 145 square metres. I expect them to be launched in non-mature estates, where land costs make the subsidy bill defensible, and I expect the accompanying grants to rise with the number of children rather than being flat per household.

The reasoning is a supply gap that is now more than twenty years old. HDB created jumbo flats in the 1990s by merging unsold smaller units, at 147 to 199 square metres. Executive flats reached up to 179 square metres before they were phased out in the mid-2000s. Since then the largest flat in regular BTO supply has been the 5-room at about 110 square metres. The 3Gen flat introduced in 2013 is only 115 square metres, and its eligibility rules are written around generational structure rather than household size, so a couple with four children may not qualify while a five-person multi-generation household does. In a private market, scarcity of large homes would raise their price, and the higher price would draw in developers. Public housing does not work that way. HDB decides what gets built, so a missing flat type stays missing until HDB decides to build it again.

Why non-mature estates? Because the two objections that ended this flat segment were the subsidy cost per unit and the resale windfall, and both shrink where land is cheaper. A 140 square metre flat in Tengah needs a smaller subsidy and creates a smaller windfall than the same flat in Queenstown. If windfalls remain a concern, resale gains can carry a clawback. The starting point is that the units exist at all.

Why subsidies scaled to family size? Because Budget 2025 already accepted the principle: the Large Families Scheme gives up to S$16,000 of additional support for each third and subsequent child. Support that rises with the number of children exists in grants. Housing is the larger cost, and the same principle applies to it.

Prediction 2: preschool moved towards primary school

I think the Government will announce a significant expansion of preschool subsidies and capacity, with a stated destination: preschool that is close to free for those who choose it, the way primary school is close to free today. I expect more government-supported places, lower fee caps, and possibly a timeline for universal access at primary-school-like cost.

The direction is already set. From January this year, full-day childcare fee caps at government-supported operators came down by S$30 to S$610 a month at anchor operators. A working mother receives a S$300 basic subsidy, and lower-income families receive more. From 2027 the income ceiling for additional subsidies rises from S$12,000 to S$15,000 a month. These are real reductions, and they are small ones. The 2025 birth numbers give the Government a reason to move faster than S$30 a year.

When a student calls primary school free, I give the same reply every year: when you say free, it is not really free. The government is paying for it. It pays because education gives society benefits beyond those the child and the family capture, so at a market price, less schooling would be bought than society wants. Every part of that argument applies to a four-year-old. The returns to early childhood education accrue over an entire working life. If the argument justifies free schooling at seven, it justifies it at four. Preschool also does something primary school does not: it allows both parents to work. If care costs are what stop couples at one child, care costs are the thing to subsidise.

The detail I will listen for most closely is student care. The jump from S$300 to S$1,300 arrives at Primary 1, precisely when parents expect the expensive years to end. If tonight's announcement covers student care for primary school children as well as preschool, the review has looked at what parents actually pay.

Prediction 3: parental leave, redesigned

Three parts to this prediction. Paternity leave doubles from 4 weeks to 8. Maternity leave extends from 16 weeks towards 26 weeks, which is six months. And the Government takes over the full cost of maternity leave, replacing today's design where, for a family's first two children, the employer pays the first 8 weeks and the state only reimburses the last 8.

Start with the recent record. Paternity leave went from 2 weeks to 4 in April 2025. Shared parental leave reached 10 weeks for babies born from this April. The Government has moved on leave twice in two years, in the same direction, and the fertility rate has since fallen further. A second doubling of paternity leave repeats a move it has already made once and already defended once.

The funding redesign is the part I feel most strongly about. Under the current structure, an employer who hires a young woman expects to pay up to 8 weeks of her salary for each of her first two children, on top of covering her role while she is away. Her expected cost to the firm is higher than that of an otherwise identical candidate, and some employers respond to that difference at the hiring stage. If the state reimburses all 16 weeks, or all 26, that cost difference shrinks, so the incentive to pass over young women at hiring weakens. There is a fairness logic underneath the accounting. Society gains the future worker and taxpayer, so society, through the state, should carry the leave that produces them, rather than whichever single employer happened to make the hire.

The counterargument: six months of fully paid leave is expensive, and a small employer still has to cover the role even when the state pays the salary. Both points are true. Not all spending is short-term consumption, though. Some spending is long-horizon capacity building, and a shrinking citizen base carries its own bill. Still, I hold this prediction with less confidence than the first two, and the six-month figure is the part most likely to arrive as a destination with a timetable rather than an overnight change.

Prediction 4: cars, the long shot

I do not think this one will happen. I include it because the mechanism already exists for a different policy goal, and because if it does happen, it is the strongest signal of the night.

Since 2021, Singapore has offered EV buyers a rebate of 45 per cent off the Additional Registration Fee, capped at S$20,000 in its first years and S$7,500 this year. The state wanted more EVs on the road, so it cut the ownership taxes for buyers who chose one. Cat B COE closed at S$131,001 this month. Cat A closed at S$128,501, a few hundred dollars off its all-time record. A family with three or four children cannot fit into a small hatchback, and ferrying four children on public transport is possible, but it costs the parent doing it hours of every day. If the state can discount ownership taxes to put more EVs on the road, it can discount them for a family car carrying four children. A partial or full ARF rebate for families with three or more children would use the same mechanism for a different objective.

Why do I still doubt it? The COE exists to ration a fixed supply of road space, and every exemption becomes the argument for the next exemption. A rebate for large families would concede that for some households a car is a necessity, and COE policy since 1990 has been built on refusing that concession. I put the probability low. If it happens anyway, it tells you the Government is treating the birth numbers as the emergency they are.

And a fifth prediction: the work continues

I think PM Wong will end the family section by placing tonight's measures inside a longer effort: the schemes are the start, and the review continues.

He has good reason to say so. Ask a parent of young children what wears them out, and the answer is usually the week itself. The daily logistics fall mostly to mothers. A playful child at a restaurant gets glares as often as smiles. Leaving the office at 6pm for the school run can read as a lack of commitment. None of this appears in a subsidy schedule, and all of it enters the decision on a second or third child.

So I expect announcements of further reviews: into daily life for mothers, into workplace norms around caregiving, into how children are treated in public spaces. These will draw fewer headlines than a new flat type. I think they matter as much. The full cost of a child is money plus time plus energy, and a Budget can move only the first. The other two are set by employers, restaurants, MRT carriages and the rest of us, which is why a Prime Minister raises them at a Rally. It is the one speech the whole country watches.

The scorecard

Evaluating my predictions

I make my students write an evaluation for every essay, so here is mine.

The strongest argument against all four predictions is fiscal. Larger flats carry larger subsidies and larger resale windfalls. Near-free preschool commits the Budget to a permanent recurring cost. Fully funded six-month maternity leave costs more with every birth it encourages, so its own success raises its price. A finance-trained Prime Minister may prefer to announce the review's direction tonight and deliver the details at Budget 2027, where the funding can be shown. If tonight is framing and the substance arrives in February, my predictions are deferred rather than wrong, and I will accept half credit.

The second argument is Lee Kuan Yew's: mindsets may have changed in ways no policy can reverse. Perhaps no flat size and no subsidy schedule brings back a fertility rate of 1.4. I cannot disprove that. I would point out one thing. Singapore has run the cash experiment for two decades, while flats stayed small, care stayed expensive and working hours stayed long. The structural experiment has not been run. The absence of certainty does not justify resignation. If demographic decline constrains everything else Singapore wants to do, then housing supply, care costs and leave design have to be aligned with that objective, and we may need to accept deliberate opportunity costs today to secure capacity tomorrow.

You do not need to agree with any of this. You can build your own predictions from the same public numbers before 8pm. Watch the Rally tonight, and see how many of these hold up.

Want this on paper? Grab the free 112 page Summary and Diagrams pack.