A data trend is not a certain future
Originally written in Chinese. Translated with AI, reviewed by Gary. Read the Chinese original →
I just came across a piece of AEC news with an eye-catching headline: by 2060, Malaysia's Chinese population will fall to around 15%. The comments were full of sarcasm and jokes. You could tell people had complicated feelings.
But I don't want to talk politics. I'd rather use this news to talk about data, statistics and prediction.
Why is Malaysia's population still growing?
It's true that Malaysia's birth rate keeps falling. On average, a couple has fewer than two children. That's a fact. Yet strangely, the total population is still increasing.
Why? Because medicine has improved and people live longer. Health awareness has grown, lifespans have naturally lengthened, and the share of older people keeps rising.
So does the idea that "Malays have lots of children" still hold today?
Not necessarily.
Look at the Malay friends around us, especially families with stable jobs and good incomes. Many choose to have only one or two children, hardly different from Chinese families. Because economic conditions affect the whole of society, not one particular race.
A point from Factfulness
What decides how many children a family has isn't religion, culture or race. It's economic status.
The poorer the family, the more children it tends to have.
But once the economy develops, people tend to have fewer children, however much you try to explain it with religion or culture.
The blind spot in data and statistics
This is a blind spot for many people when they make decisions.
They see a data trend and assume it will go on forever.
For example, China's one-child policy was introduced because of fears there wouldn't be enough food, so strict controls were put in place in advance. The starting point of the policy was to "prevent famine".
But the policymakers at the time didn't foresee that agricultural technology would leap ahead over the following decades. The result is that China today faces new problems: an ageing population, a shortage of productive workers, and young people who don't want children.
What does this have to do with starting a business?
Does data matter?
Of course it does.
But many founders, me included, are sometimes too easily led by "the data".
A data trend is not a certain future.
Often, business decisions can't just be based on "past data continuing". Because putting a decision into action can trigger new changes, and those changes can make the original trend no longer hold.
Keep testing, adjust in real time
The essence of starting a business is really:
- Find a problem
- Set a goal for solving it
- Keep measuring, checking and validating
- Keep improving based on feedback
Decisions shouldn't be made on a whim because of one news story or one set of numbers, then followed blindly all the way.
Back to the opening example: if there are fewer and fewer Chinese Malaysians, should business decisions switch entirely to the Malay market?
The conclusion sounds reasonable, but I don't think it will necessarily happen.
Because changes in population structure aren't a straight line. There's a statistical principle here: "regression to the mean".
As the economy improves, fertility rates across all groups tend to become similar.
That means the gap in birth rates between Malays and Chinese may not keep widening in the future, but slowly narrow.
Add complicated variables like better healthcare and migration, and the ethnic ratio in 2060 may look much like it does now.
Of course, this isn't "I'm definitely right". It's just a guess made with some scepticism.
Starting a business is a continuous process of solving problems.
The environment we face changes every day. Data is a reference, not the answer.
It's fine to make decisions with data, but what matters more is to keep testing, getting feedback and adjusting after the decision. Don't jump to conclusions too early because of a "forecast of the future".
Can the future be predicted? Yes. But the future is even more worth verifying and rewriting, bit by bit.