How scientists forecast future populations

How do scientists figure out how populations will change in the future?
02 March 2021

Interview with 

Brienna Perelli-Harris, Southampton University


this is a picture of people standing together


Quantifying population growth now and forecasting into the future is an important task - for instance, to know how big a nation’s future workforce will be, how many school places you need etc. So how do scientists actually accomplish this? Brienna Perelli-Harris is an expert on human populations from Southampton University, and spoke to Katie Haylor...

Brienna - Well, demographers tend to use three different types of forecasting methods. The UN uses Bayesian estimation techniques, which takes into account many different previous estimation trends, and when they calculate their forecast they tend to have a lot of uncertainty. So they might have a median-variant, but they do include lots of uncertainty based on what has happened previously. The IHME, which produced a study that was very publicised - it was in the Lancet - over the summer... they tend to make assumptions about contraceptive use and education, but it was rather controversial because many of the assumptions, especially in Africa, didn't really line up with what other demographers were seeing had happened in Africa, because of the stall in fertility in certain regions of Africa.

And then the other way that demographers tend to forecast populations is by estimating how education is going to change over time, because women's education has been one of the most important factors for declining fertility. And this technique also takes into account expert opinion. So these different approaches have been used to forecast global population in general.

Katie - I guess it's important to distinguish between biological fertility, and reproductive choices and agency, isn't it? How does biological fertility stack up as a potential impactor on population change?

Brienna - In general we've seen that the most important factor is age. So as you know, women's reproductive age tends to end around 45-49, and although this has been pushed forward a little bit with assisted reproductive technologies and IVF, it really hasn't pushed that boundary very far; and we see very few births occurring after the age of 45. That is what we take into account most in terms of the biological capacity. And in fact, there have been studies... for example, I did some research in Ukraine, which as you remember was impacted by Chernobyl, and there was a lot of pollution in the environment. And in the early 2000s, I found that only 5% of the population remained childless. This is really due to social norms that have really pushed couples to have children really early, and if you start early, or if a population starts early - around the age of 20 - then childlessness can be very rare, or relatively rare. Although we take into account fecundability on the population level, it's not the main consideration that we demographers use to think what populations are going to do in the future.

Katie - Okay. So what you're saying, is the main biological factor is maternal age. What is the overall biggest factor? Because I would have assumed it's probably finances. Is that true?

Brienna - Yes, and preferences, and... but you're right that economic uncertainty has been one of the most important reasons why certain countries on average have had lower fertility rates. Also, family policies are really important: to what extent couples can balance work, and family, and childcare, and maternity leave. That has also been really important. Also - we talk about gender equality, and to what extent men actually contribute to household labour, as being also really important for these decisions about how many children to have.


Add a comment