New flats near a school, and what happens to P1 demand
A question worth an answer with numbers behind it: when thousands of new flats go up near a primary school, does that school become harder to get into, and when? We measured it across every primary school, over sixteen years of registration.
Data covers: HDB completion years and P1 outcomes 2010 to 2025, compiled 1 August 2026
after the keys are handed over is when Phase 2C is most contested.
applicants per place, from the quietest point to the peak. A real shift, not a transformation.
schools ballot either every single year or never, so for half of them the yes or no cannot move at all. The length of the queue still can.
More flats, then what
Every bar below compares a school against its own normal, not against other schools. That matters: a school everyone wants and a school nobody fights over would otherwise drown each other out. Read left to right as the amount of new housing going up within 1km in the years before, and the two panels answer two different questions about the same schools. Both panels are drawn to the same scale, each measured as a share of its own normal, so a taller bar on one really does mean a bigger effect than a shorter bar on the other.
applicants per 2C place, against the school's own average
Each row is a school-year grouped by how many new flats were completed within 1km in the three to ten years before.
percentage points, against the school's own rate
Each row is a school-year grouped by how many new flats were completed within 1km in the three to ten years before.
The left panel is the answer to "do new flats make a school harder to get into". Yes, and you can see the dose: a school in a year following 7,000 or more new flats runs 0.17 applicants per place above its own normal, while the same school in its quiet years runs slightly below. That is measured on 2,594 school-years.
The right panel is the one worth slowing down for, because it is the question everybody actually asks. Balloting does become more likely, by 4.4 percentage points at the top of the range, which is 9% above what those schools normally do. The queue over the same jump moves 16%, so roughly twice as far. That gap is the whole point, and the reason for it is not subtle: 50 schools balloted in every single year we have, and 40 balloted in none of them. For 90 of 173 schools the yes-or-no is already settled before a single flat is built, so new housing has nothing to flip. It changes how many families are standing in the queue behind that answer.
Put plainly: if your school already ballots, new flats will not save you and they were never going to. What they change is how far down the list you are standing when it does.
When it arrives
We took every case where 1,500 or more flats were completed within 1km of a school in a single year, 207 of them between 2000 and 2018, and tracked that school against its own long-run average. The bars below are not raw numbers of applicants. They are how far that school sat above or below its own normal, so a school that is always competitive and a school that never is can be read on the same axis.
Scroll sideways to see the whole chart.
The same chart as a table
| Years from handover | After a delivery | No delivery | School-years |
|---|---|---|---|
| -3 | -0.14 | -0.02 | 95 |
| -2 | -0.15 | -0.04 | 105 |
| -1 | -0.15 | -0.03 | 117 |
| 0 | -0.12 | -0.03 | 124 |
| +1 | -0.08 | -0.04 | 127 |
| +2 | +0.03 | -0.06 | 130 |
| +3 | +0.06 | +0.00 | 133 |
| +4 | +0.15 | -0.01 | 135 |
| +5 | +0.18 | -0.04 | 137 |
| +6 | +0.14 | -0.02 | 139 |
| +7 | +0.14 | -0.02 | 145 |
| +8 | +0.03 | +0.01 | 139 |
| +9 | +0.07 | +0.00 | 130 |
| +10 | +0.09 | +0.00 | 137 |
| +11 | +0.05 | -0.01 | 122 |
Why five years, and not one
Families move into a new flat and some already have a three or four year old, so a few extra applicants turn up almost immediately. MOE also tends to open a few more Phase 2C places early on, which cushions it.
The squeeze comes later, and from the other direction. Five to nine years after handover those same families' younger children register in Phase 1 as siblings. The school's total intake goes up, but a larger share of it is taken before Phase 2C opens. Applicants rising and places falling at the same time is what produces the peak. It fades as the estate matures and the first wave of children moves on to secondary school.
How much of the story this is
Not all of it, and it is worth being exact about the size. Across 2,712 school-years at 179 schools, 49% had more Phase 2C applicants than places. New housing moves that, but weakly: the correlation with balloting is 0.06, against 0.23 for applicants per place. Part of that gap is real and part of it is arithmetic, because a yes-or-no outcome cannot track a number as closely as another number can. The bar charts above are the better comparison, and they show the same ranking: the queue responds more than the verdict.
On the queue itself the relationship explains roughly 5% of the variation between schools and years. It is one real force among several, not the explanation. Where a school sits in parents' minds, whether it has a popular affiliation, how many siblings are already inside, and how many places MOE opens all matter, and none of them is housing.
Three schools, close up
Bars are flats completed within 1km that year. The line is Phase 2C applicants per place. All three panels share the same scale, so they can be compared with each other.
10,469 flats completed within 1km since 2013, 30% of all the HDB flats in its 1km band.
Scroll sideways to see the whole chart.
14,616 flats completed within 1km since 2013, 42% of all the HDB flats in its 1km band.
Scroll sideways to see the whole chart.
15,085 flats completed within 1km since 2013, 64% of all the HDB flats in its 1km band.
Scroll sideways to see the whole chart.
Read this one with care. Angsana moved to its current campus on 6 June 2022 and was formed in 2015 by a merger, so the years before 2022 were registered somewhere else. The new flats and the move arrived together and cannot be separated here.
Where the next wave is already built
Flats completed between 2019 and 2025 are homes families are living in now. On the timing above, the demand they carry lands roughly between 2024 and 2030. 73 schools have 1,000 or more such flats within 1km. 61 have none at all.
| School | Flats 2019-2025 | Share of its 1km stock |
|---|---|---|
| Bukit View Primary School | 16,463 | 48% |
| Pioneer Primary School | 15,567 | 50% |
| Angsana Primary School | 10,689 | 46% |
| Northshore Primary School | 10,409 | 77% |
| Cedar Primary School | 8,872 | 68% |
| Maris Stella High School (Primary Section) | 8,325 | 59% |
| Poi Ching School | 7,439 | 25% |
| Huamin Primary School | 7,107 | 18% |
| North View Primary School | 7,107 | 25% |
| Sengkang Green Primary School | 5,979 | 27% |
| Jurong Primary School | 5,954 | 23% |
| Princess Elizabeth Primary School | 5,852 | 17% |
A large share is not a prediction that the school will ballot. It says the pool of families near it grew quickly, which on the evidence above shows up as more applicants per place, most strongly around five years after the flats were handed over.
Common questions
Do new HDB flats near a school make it harder to get a place? +
They make Phase 2C more contested, and the effect is real but modest. Comparing each school against its own long-run average, flats completed three to ten years earlier correlate with applicants per place at r = 0.23, which explains roughly 5% of the variation. Plenty of other things matter more.
How long after the flats are built does demand arrive? +
About five years. Around the 207 cases where 1,500 or more flats were completed near a school in one year, Phase 2C was easier than that school's own average while the flats were being built, crossed over about two years after handover, peaked at year 5, and faded from about year eight.
Does that mean the school will go to a ballot? +
Not reliably. Whether a school balloted at all barely moves with new flats (r = 0.06). Across 2,712 school-years, 49% had more Phase 2C applicants than places, and that share is about the same near new estates as far from them. What changes is the size of the queue, not whether there is one.
Can I use this to predict my own school's ballot? +
No, and it is worth being blunt about it. This is a pattern across 179 schools over 16 years, not a forecast for one school in one year. Your own year turns on the number of siblings, alumni and parent volunteers ahead of you in the earlier phases, which no amount of housing data reveals. Use MOE's published figures for your school and year.
Which flats count as near a school: the 1km band is SLA's OneMap School Query answer for each individual block, the service MOE tells parents to use, not a distance we measured ourselves.
Flats and their completion years: HDB's own Property Information records on data.gov.sg. Private and landed homes are not in this analysis, so a school ringed by condominiums will show a smaller figure here than the number of homes actually near it.
Registration outcomes: per-school Phase 2C vacancies and applicants, 2010 to 2025. "Went to a ballot" on this page means applicants exceeded places, which is the standard reading but is not the same as MOE's own published balloting flag. For your school and your year, use MOE's vacancies and balloting data.
How the correlations were calculated: each school is compared against its own average before anything else, so a school simply being popular cannot produce the result. Seven schools whose Phase 2C applicants doubled in one step and stayed doubled are excluded from the headline figures; the finding holds with or without them. Analysis compiled 1 August 2026.
This is a pattern across schools and years. It is not a forecast for any one school, and it should not be used as one.
Find your child's group
Every cohort has its own WhatsApp group of parents going through the exact same year, real-time registration updates, school reviews, balloting news and answers from people one step ahead of you. Pick the year your child starts Primary 1.
The 2026 registration exercise runs from 30 June to 30 October 2026, with the citizen and PR phases finishing in late August. This is the live cohort, phase dates, balloting and school choice are happening right now.
Registration in mid-2027. Get the lay of the land early, school shortlists, the 30-month address rule, and what to prepare.
Plenty of runway. Useful if you're weighing a home move for distance priority or choosing a kindergarten with an eye on P1.
Free to join. A friendly, parent-run space, no spam, no selling.