#demography

Methodology (Summary)

This is where me not being a demographer is an important thing to know. I neither know nor normally care about the statistical methods used to determine demography, just that the demographers aren't retracting papers over it. However, I do know that in general the CCF50 (total cohort fertility before the age of 50) is a neater and more accurate measurement to build projections on than the TFR (total fertility rate by year) and that's the methodology the paper's authors went with. This is good and promising. TFR for known years and CCF50 projections sounds like a solid method. 👍

We additionally produced forecasts for multiple alternative scenarios in each location: the UN Sustainable Development Goal (SDG) for education is achieved by 2030; the contraceptive met need SDG is achieved by 2030; pro-natal policies are enacted to create supportive environments for those who give birth; and the previous three scenarios combined.

I'm very hopeful about these forecasts! They'll show a few different hopeful scenarios.

To evaluate the forecasting performance of our model and others, we computed skill values—a metric assessing gain in forecasting accuracy—by comparing predicted versus observed ASFRs from the past 15 years (2007–21). A positive skill metric indicates that the model being evaluated performs better than the baseline model (here, a simplified model holding 2007 values constant in the future), and a negative metric indicates that the evaluated model performs worse than baseline.

This is a very responsible thing for the authors to have done, and I am interested to see how this is reflected in the models.

.

Findings (Summary)

During the period from 1950 to 2021, global TFR more than halved, from 4·84 (95% UI 4·63–5·06) to 2·23 (2·09–2·38). Global annual livebirths peaked in 2016 at 142 million (95% UI 137–147), declining to 129 million (121–138) in 2021. Fertility rates declined in all countries and territories since 1950,

(Emphasis mine. The numbers in parentheses are the confidence interval.) I think this is the most important takeaway from the whole damn paper. Makes sense, since it's the first line of the findings. If you read nothing else, read these three sentences. Global birthrates are barely above replacement (which, if you recall from my other essay, is generally considered to be ~2.1). To me, this implies lot of problems that traditionally have been considered solvable with population redistribution (meaning, mostly, immigration) may not be solvable that way even if fertility were to stop declining today and hold constant for the rest of the century.

Future fertility rates were projected to continue to decline worldwide, reaching a global TFR of 1·83 (1·59–2·08) in 2050 and 1·59 (1·25–1·96) in 2100 under the reference scenario. The number of countries and territories with fertility rates remaining above replacement was forecast to be 49 (24·0%) in 2050 and only six (2·9%) in 2100, with three of these six countries included in the 2021 World Bank-defined low-income group, all located in the GBD super-region of sub-Saharan Africa.

Holy shit. I cannot emphasize enough how low a TFR of 1.59 is. This is approximately the current TFR of the United Kingdom, and they're beginning to freak out even though they have relatively easy sources of additional replacement recruitment through the Commonwealth. Imagine that for the whole Earth. With only six countries as a potential source of surplus population to be redistributed.

Funding from the Bill & Melinda Gates Foundation. Makes sense. This is the kind of thing that foundation funds.

.

Introduction

Low levels of fertility have the potential over time to result in inverted population pyramids with growing numbers of older people and declining working-age populations. These changes are likely to place increasing burdens on health care and social systems, transform labour and consumer markets, and alter patterns of resource use.

Oh man, I wish I'd gone through this paper earlier, I could have just quoted this bit and been done instead of trying to explain it from scratch! 😂

The UN Population Division estimates of past fertility are not compliant with the Guidelines on Accurate and Transparent Health Estimates Reporting (GATHER) statement in important respects; notably, they do not provide all code for statistical models or explicit details on criteria for exclusion or adjustment of primary data sources. Furthermore, the validity of UN Population Division projections has been questioned due to the assumptions applied in countries experiencing low post-transition fertility dropping below replacement level.

YES GO OFF 👏 The UN Population Division is so much more cagey about their data than the World Bank, it's so annoying, and they keep predicting increases that don't happen. I thought it was so weird as an undergrad but figured it was because of ~bureaucracy~ or privacy laws or whatever. It's nice to be vindicated [redacted] years later.

Our forecasts also suggest that, by 2100, the largest concentrations of livebirths will shift to low-income settings, particularly a subset of countries and territories in sub-Saharan Africa, which are among the most vulnerable to economic and environmental challenges. Extreme shifts in the global distribution of livebirths can be partially ameliorated by improved female education and met need for modern contraception. Outside of this subset of low-income areas, most of the world's countries will experience the repercussions of low fertility, with ageing populations, declining workforces, and inverted population pyramids, which are likely to lead to profound fiscal, economic, and social consequences. National policy makers and the global health community must plan to address these divided sets of demographic challenges emerging worldwide.

This is such an important point for them to make. Demography isn't a vacuum; it has significant real-world effects. By 2100, most babies born will be born in Africa, and we need to plan for that now. By 2100, most countries will not have enough workers, and we need to plan for that now. 2100 is not that far into the future. I, personally, will live to see the beginnings of the effects of this demographic shift, and I'm an adult who pays taxes and has a college degree and shit.

.

The Data Sources and Processing section is pretty standard and unremarkable. Good job.

.

Fertility Forecasting

We produced forecasts of fertility using an updated modelling framework (appendix 1 section 3) that improved on the methods in the 2020 study by Vollset and colleagues. In our updated methods, we used not only estimates of female educational attainment and contraceptive met need as covariates, but also estimates of under-5 mortality and population density in habitable areas to account for a larger variation in CCF50 across all countries in the sub-models (appendix 1 section 3.1, appendix 2 figure S2). Similar to Vollset and colleagues, we continued to forecast fertility with CCF50 rather than TFR, because modelling in cohort space is more stable than in period space.

Niiiiice. Covariates are things that, well, vary, alongside the thing you're trying to measure. For fertility, the most obvious one might be age of the mother at first birth; if someone is 16 at first birth, she probably will have more kids than someone who is 30 at first birth, for example. This model also includes how much schooling the mother gets, whether she has contraception, the mortality rate (that is, how many of them die) of children under five, and population density! That's a lot of statistical crunching and their model will be more precise for it. Precise isn't the same as accurate, but I think that with the variables they selected, they will travel in the same direction.

What a pretty equation. I don't understand it, but it's got a certain je ne sais quois.

For the education SDG scenario, the forecasts assume that by 2030, all people will have 12 years or more of education by the age of 25 years and then maintains the same rate of change as the reference scenario up to 2100. For the contraceptive met need scenario, to reflect the SDG scenario of universal access, the forecasts assumed a linear increase in contraceptive coverage to reach 100% by 2030 and then stay constant up to 2100.

I love how optimistic these scenarios are 😂 This truly is the best-case scenario for both the education forecast and the contraceptive forecast! I do hope everybody has 12+ years of education and 100% contraceptive coverage by 2030. Make it happen, António!!!!

(Joke explained: António Guterres is the current Secretary-General of the United Nations, and these goals are absolutely not going to be met by 2030.)

In the pro-natal scenario, we assumed a country will introduce pro-natal policies, such as childcare subsidies, extended parental leave, insurance coverage expansion for infertility treatment, 33 and other forms of support for parents to afford high-quality child-care services, once TFR decreases to less than 1·75. We then made three assumptions on the effects of such policies. First, we assumed the full effect of pro-natal policies will be to increase TFR by 0·2. Second, it will take 5 years after the policy is introduced for the full increase in TFR to occur, and TFR will rise linearly over that time span. Last, we assumed that both the policies and the increase in TFR by 0·2 will endure for the remainder of the century.

The pro-natal scenario is also incredibly optimistic. This kind of response simply hasn't happened in any country that's tried pro-natal policies as envisioned by the authors (my reference cases, just off the top of my head: Japan and France).

The optimism makes sense. They represent extreme cases, in order to contrast possible outcomes versus the reference case. This is good practice! It's just also funny.

.

Results

The Reference Case

I hate the embedded tables. They have the confidence interval in the same cell as the estimate. How very dare they, that's incredibly inconvenient for me personally.

The chart in Figure 1, however, I think speaks volumes:

It speaks so many volumes that I'm gonna go up and put it above the cut, brb. This chart shows the reference case; that is, it shows the fertility rate if the fertility trends of 1950-2021 continue into the future.

At the national level, estimates of TFR in 2021 ranged from 0·82 (95% UI 0·75–0·89) in South Korea to 6·99 (6·75–7·24) in Chad, with below-replacement levels of fertility (TFR <2·1) in 110 of 204 countries and territories (table 1, figures 2A, 3).

I think this range is neat and goes to show that while the trend is world-wide, it's still not even. Chadian women still give birth to about 7ish kids on average. That's more than 3x replacement, and more than 8.5x the average fertility of South Korea. South Korea is going to have different problems than Chad; Chad probably doesn't have to worry as much about their workforce being unable to sustain a large elderly population. (Don't look so cheerful about it. They've got lots of other stuff to worry about. 😬)

These charts are fascinating to look at to me. I think this really showcases just how dramatic the projected decline is. It's not just the Europe, it's not just wealthy post-industrialized countries, but everywhere. It's in Eswanti, it's in Indonesia, it's in Burkina Faso, it's in China. It really shows just how much Chad is an outlier (adn should still be counted, btw, just because it's an outlier doesn't mean we should discard it; it's dependent on study structure and you can't just throw out entire countries because they have high birth rates on a study of birth rates).

Our estimates indicate that there is approximately a 30-year gap between the time when TFR falls below 2·1 and when the natural rate of population increase turns negative. We forecast that 155 (76·0%) countries and territories will have fertility rates below replacement level in 2050; by 2100, we project this number will increase to 198 (97·1%), with 178 (87·3%) having a negative natural rate of increase (figure 3).

A 30-year gap sounds reasonable. That's about how long it takes for people to have/not have kids, and for their own parents to potentially die, in about equalish numbers (on a global scale, anyway). I do think this gap number is likely to increase as healthcare improves in places that are worse today and as fertility technology increases the age at which people can become pregnant, but 30 is a perfectly respectable number with actual statistical backing.

Alternative scenario fertility forecasts

This is the part I'm really excited about!!!

The first scenario, which assumes meeting the SDG education target by 2030, is estimated to result in global TFRs of 1·65 (95% UI 1·40–1·92) in 2050 and 1·56 (1·26–1·92) in 2100 (table 2). The second scenario, which assumes meeting the SDG contraceptive met need target by 2030, will produce global TFRs of 1·64 (1·39–1·89) in 2050 and 1·52 (1·21–1·87) in 2100. The third scenario, which incorporates pro-natal policy implementation, is forecast to yield global TFRs of 1·93 (1·69–2·19) in 2050 and 1·68 (1·36–2·04) in 2100. The combined scenario, in which all three other alternative scenarios are applied, is projected to result in a global TFR of 1·65 (1·40–1·92) in 2050 and 1·62 (1·35–1·95) in 2100.

So recall the reference scenario projections: 1·83 (1·59–2·08) in 2050 and 1·59 (1·25–1·96) in 2100.

I find it interesting that all cases are so incredibly close to reference, with overlapping confidence intervals. Functionally, there's not a lot of difference between a TFR of 1.68 and 1.52. They're both still well below replacement. It's about the difference between Sweden (1.67) and Russia (1.51). Russia, you may have noticed, is waging war about it.*

*This is not a stated goal of the Russian Federation in the Ukraine War. This is me personally making an assertion that the shifting demographics of the Russian population, including the below-replacement birthrate beginning to put pressure on their lacking social safety networks, has contributed to the many complicated and interconnected reasons why the Russian Federation invaded Ukraine, but please do not take me to be the final authority on the matter or interpret this statement as implying that demography of all things is the sole or primary reason for the war.

Discussion

The aforementioned changes in fertility over the coming century will have profound effects on populations, economies, geopolitics, food security, health, and the environment, with a clear demographic divide between the impacts on many middle-to-high-income locations versus many low-income locations. For nearly all countries and territories outside of sub-Saharan Africa, sustained low fertility will produce a contracting population with fewer young people relative to older people before the end of the 21st century. These changes in age structure are likely to present considerable economic challenges caused by a growing dependency ratio of older to working-age population and a shrinking labour force. 42 Unless governments identify unforeseen innovations or funding sources that address the challenges of population ageing, this demographic shift will put increasing pressure on national health insurance, social security programmes, and health-care infrastructure. These same programmes will receive less funding as working-age, tax-paying populations decline, further exacerbating the problem.

This is why the Economist article talks about birthrates the way it does. It's not about white babies or whatever people in the notes are sarcastically ascribing to an article they haven't read. It's about the whole world. There are 150 countries outside of the Sub-Saharan Africa region, and 44 of the 46 countries within Sub-Saharan Africa are projected to feel the many or all of the same effects as well.

It's about the way social security nets are structured and how they're going to fail. It's about the way that elderly people are going to be treated by our societies. It's about me, and it's about you, and it's about making sure that there are enough humans to take care of the other humans that need taking care of.

If we don't increase global fertility rates above replacement, which it increasingly looks like we won't, we need other solutions. The fertility one is easy fuckin' pickings compared to a complete overhaul of society, and you saw how little difference it actually makes. So did the authors:

To date, one strategy to reverse declining fertility in low-fertility settings has been to implement pro-natal policies, such as child-related cash transfers and tax incentives, childcare subsidies, extended parental leave, re-employment rights, and other forms of support for parents to care and pay for their children.49, 50 Yet there are few data to show that such policies have led to strong, sustained rebounds in fertility, with empirical evidence suggesting an effect size of no more than 0·2 additional livebirths per female. [...] Moreover, although pro-natal policies primarily aim to increase births, they also offer additional benefits to society, including better quality of life, greater household gender equality (ie, more equal division of household labour),53 higher rates of female labour force participation,54 lower child-care costs,55 and better maternal health outcomes,56 depending on policy design and contextual factors. In the future, it will be beneficial to perform an in-depth analysis on varying impacts of pro-natal policies in selected countries that have a meaningful impact on population. [...] Importantly, low fertility rates and the modest effects that pro-natal policies might have on them should not be used to justify more draconian measures that limit reproductive rights, such as restricting access to modern contraceptives or abortions.

I just want to highlight that the study authors explicitly argue for certain pro-natal policies that increase quality of life and caution against pro-natal policies that limit rights. These people aren't heartless.

They also discuss at some length the implications of the changing distribution of live births, where by the end of the century most live births will take place in the poorest nations, which are also the ones that will be hardest-hit by climate change. These nations already face famines, military rule, civil wars, terrorism, and climate changed-caused severe heatwaves, droughts, and floods. They advise politicians to take this into account when making policy decisions but don't go into what policy decisions should be made, which is wise since they're demographers and not political scientists, but disappointed me, the political scientist reading the demography paper and hoping to find something to criticize.

.

.

My takeaway: