Joining in the TFR talk

Recent news of TFR hitting a new low made me reach for the spreadsheets again!

The overall direction of the TFR trend (downwards) is indisputable but I already talked about how the actual TFR number needs to be taken with a pinch of salt - here. Quick recap, TFR is affected by numerators on number of babies and denominators on number of women; some years with an increase of resident babies do not have a corresponding increase in resident TFR.

Today, I want to look at births vs. immigration coz we have taken to importing people when we don't produce enough.

Overall, the number of Singapore citizen (SC) births had been declining for the past six years after a slight uptake between 2018 and 2019 (+1.3%). The latest decline between 2024 and 2025 was the worst at -10.4% since 2001 (-12.5% from 2000).

So 20-30y later, we won't have enough local workers. At this rate, native SC population will start to shrink at some point.

But fear not, we can import new SCs! Ready made workers! SC pop won't shrink (as much?)!

The SC population as at end June 2025 was of an older profile compared to the new SCs granted in 2024. In particular, the two largest groups of new SCs are at the prime working ages of 31-40yo and up to 20yo, probably children of the former. This should be the case since immigration is meant to augment the young and the working population. 

Note: The age group categories are different for the SC pop data and for the new SC data, but no choice, those were the only data I could find. But close enough to compare.

But what is the impact of immigration on TFR? Does it really help?

Firstly, I was curious about the TFR among the new SCs. If I assume the new SC children were all born in 2024 and that 50% of those aged 21y and above were women of childbearing age, their "TFR" would be a ghastly 0.85 in 2024, lower than local TFR of 0.97 in the same year. But the first assumption is not sound and the second also dubious. So maybe look at the number of added births rather than TFR. But before that, I want to look at correlation between TFR and SC population.

I plotted TFR against SC pop, new SCs granted and a derived variable assuming no new SCs from 2007, created by subtracting new SCs granted from 2007 onwards (that's the first available number) from the SC pop each year. 

Now, I'm departing from facts to inference. 

Visually, there seems to be some correlation between the increase in all SC numbers and the decrease in TFR. It's probably maths as highlighted in my previous post. 

But - visually again - TFR seems to decrease more in recent years, when the delta between SC pop and SEC pop less new SCs becomes greater. Dare I propose that besides maths, this also suggests that new SCs have a lower propensity for children? But then again, many may come here already parents so they don't add to the TFR. If they conform to local norms, they are unlikely to have more children than the average SC. So I guess that brings me back to number of added births.

By added births, I'm referring to how the new SC kids add to local birth cohorts of previous years. 

Since it takes time for a foreigner to become a PR and then an SC, I assume the new SC children are at least 5yo and they are evenly spread across the ages. So the annual average of 6,600 kids since 2012 (when age data were available) covers 15 years of births or 440 per year. Let's call these births "newfound births".

As we get new SCs every year, each cohort of new SCs bring 440 newfound births in each of the first 15 of the past 20y. This means:

• In 2021-2025, there are no newfound births from the new SCs granted in 2025.
• In 2020, there were only newfound births from the 2025 cohort of new SCs, i.e. 440.
• In 2019, there were only newfound births from the 2024-2025 cohorts of new SCs, i.e. 440 x2. 
• Thus, in 2006, there were 440 x 15 = 6,600 newfound births from the 2011-2025 cohorts of new SCs combined.

6,600 represents a 20.7% increase in the number of SC births in 2006! I thought that was significant. 

So to answer my earlier question, I think immigration can bolster our low TFR but only if we're selective about it. 

A similar calculation can show that new SCs of working age can be considered newfound births decades ago and are useful to augment our current working pop. But as we have a greater dearth of births recently, it is more worthwhile getting new SCs who already have children to augment our current younger pop and future working pop. Bonus: These children will grow up like native SCs. 

Conversely, it's not worthwhile getting new SCs who do not have children at the point of being granted new SCs because they will likely conform to local norms, i.e. not likely to have many children, if at all.

Of course, don't add to our ageing pop!

The other part of the TFR discussion is on increasing local births. I feel like this is another discussion altogether. Maybe next time..

References:

• New SCs before 2012: Population in Brief, NPTD 
• All other data generated in early March 2026: DOS search by theme