By John McKeown
Empty Planet moves confidently from an optimistic premise to unwarranted conclusions. It has been reviewed favourably by Steven Pinker, The Wall Street Journal and the New Statesman, among others. Its premise is that world population will peak far lower and sooner than the UN forecasts, and that because of a faster-than-expected decline in fertility rates rather than any disastrous rise in mortality. Given this, the authors claim that “gloom” about species going extinct, crowded out by human impact, is misguided (page 2). The second part of this review will argue that even if the premise were accurate that conclusion would still be unjustified. But first I examine the credibility of their premise.
World population growth is slowing but demographers are unsure what peak number will be reached, and when that will happen. The UN medium projection is 9.77 billion by 2050, rising to 10.9 billion by 2100.1 The book claims that an “increasing number of demographers from around the world believe the UN estimates are far too high. More likely, they say, the planet’s population will peak at around nine billion sometime between 2040 and 2060 … By the end of this century we could be back to where we are right now” (p. 2), which would be 7.6 or 7.7 billion. The first claim, that some demographers think the UN forecast is too high, is true and the authors interviewed a leading critic, Wolfgang Lutz of the IIASA (International Institute for Applied Systems Analysis). However, Empty Planet misrepresents the IIASA. The other alternative forecasts cited on pages 45-47 are outliers: Jørgen Randers, and a Deutsche Bank report by Sanjeev Sanyal. The book contains no graphs, so I compared its claims about peak timing, and population by 2100, with its cited sources. Figure 1 shows that the number for 2100 in Empty Planet is far below the Lutz/IIASA Medium forecast, but close to Sanyal’s forecast.
A report for Deutsche Bank in 2013 by economist Sanjeev Sanyal, titled “Predictions of a Rogue Demographer” argued that: “Urbanisation is the strongest contraceptive” and so “the world’s overall fertility rate will fall to replacement rate by 2025” causing population to peak around 2055 just above 8.7 billion.2 His forecast for 2100, at 8 billion, is slightly higher than Empty Planet anticipates. Sanyal’s methodology is not open, but in an earlier report he claims that global convergence to a TFR of 1.6 is “no less likely” than convergence to replacement, and “the likeliest scenario is somewhere in between”.3 Tiny differences in the pace of fertility change lead to widely varying outcomes in long-range forecasts, as seen in Figure 1. Since these forecasts by Sanyal and Randers are a few years old they can be checked against population in recent years, as estimated by the Global Burden of Disease project (GBD). Setting aside the issue of differing estimates for my 2010 starting year in Figure 2, the subsequent changes in the UN, IIASA, and GBD estimates are similar (lines are roughly parallel), whereas the older forecasts by Sanyal and Randers are diverging over time further from (estimated) reality.
Jørgen Randers is Professor of Climate Strategy, and co-author of The Limits to Growth which offered scenarios, dependent on variables including policy choices, and in some scenarios global population peaked before 2050 but the cause was a sharp rise in death rates.4 However in the 2014 TED talk which Empty Planet cites, Randers is optimistic that birth rates will decline much faster than the UN expects, leading to a peak at 8 billion around 2040. That is based on Randers’ book 2052 – a global forecast for the next forty years.5 It cites as a baseline the UN WPP 2010 Low variant, which is a sensitivity analysis (at minus 0.5 children per woman), and for World aggregates these typically run far below the lower bound of the UN probabilistic projection’s 95% prediction interval.6 Randers’ book deals with many systems (including ecology, energy, climate, and economy): only a small part of his book considers population. Empty Planet fails to mention that Jørgen Randers in 2016 revised his forecast upward, to peak at 8.3 billion.7
The third source from whom Bricker and Ibbitson claim support is an eminent demographer: Wolfgang Lutz directs the IIASA World Population program at the Wittgenstein Centre and co-leads the Centre of Expertise on Population and Migration (CEPAM). The IIASA uses a different methodology from the UN, and its 2014 Medium forecast peaked at 9.4 billion around 2070, then declined to 9 billion by 2100.8 When this was updated in 2018 the peak moved slightly higher so the IIASA 2018 Medium forecast now peaks around 9.7 billion in 2070-75, then declines to 9.3 billion by 2100.9 This does not support the Empty Planet claim that peak population “will occur in three decades, give or take” (p.2). Also the population size for 2100 is higher than Empty Planet suggests.
So how can Bricker and Ibbitson claim Lutz and the IIASA as supporters of their view? The reason is that the IIASA developed a number of scenarios based on “shared socioeconomic pathways” (SSP): five in 2014, and three of those were kept in 2018. There is a “Medium” pathway (SSP2); a worse future is “Inequality” SSP4; a better future is SSP5; the worst is “Stalled Development” SSP3; and the best is SSP1 “Rapid Development”.10 Lutz and his colleagues advised in 2018 that SSP2 is “a medium pathway that can be seen as most likely”; whereas SSP1 is “the most optimistic scenario … Since these goals of universal secondary education by 2030 are extremely ambitious for some of the least developed countries, the results may not necessarily be considered likely”.11 Under SSP1 global population peaks at 8.7 billion around 2055 and then declines to 7.2 billion by 2100; at the other extreme under SSP3 it rises to 13.6 billion by 2100. The authors do not make clear that they chose among paths, and chose a future regarded by the IIASA as less likely (p.46).
Beyond numbers and dates, the most striking contrast between IIASA’s work and Empty Planet is that IIASA’s hopeful scenarios are presented as possibilities if policy choices are implemented. Lutz and colleagues warn that which SSP happens “will depend largely on policy choices made over the coming years” and so “we as humanity are at the crossroads, yet to choose between very different futures.”12 By contrast Empty Planet, though it occasionally admits uncertainty, portrays a future near SSP1 as likely since countries “continue their march toward urbanization, modernization, and the inevitable 2.1 [TFR] and below” (p.238).
The authors’ confidence arises from two sources: personal conversations, and a simplistic theory. They travelled “to Brussels and Seoul, Nairobi and São Paulo, Mumbai and Beijing, Palm Springs and Canberra and Vienna” and they “talked to young people: on university campuses and at research institutes and in favelas and slums” (p.3). This gives them, they think, better insight than UN demographers because: “Some of the indications of an accelerating decline in fertility can be found in scholarly research and government reports; others can only be found by talking to people on the street” (p.3). I find this unconvincing: demographers use large and carefully sampled surveys. For example since 1984 USAID’s Demographic and Health Survey (DHS) in 90+ developing countries has assisted 400+ surveys each with a sample size between 5,000 and 30,000 households.13
Empty Planet follows a simplistic theory of the origin of fertility decline: “Indisputably, the most important factor is urbanization” (p.18). They also emphasise women’s empowerment and education but in a chain of causation these are made secondary because “urbanization leads to the empowerment of women, which leads to a decline in fertility rate” (p.88). Elsewhere other factors are mentioned but urbanization is the primary driver: “As a society urbanizes, and women gain more power, the ties of kin, the power of organized religion, and the dominance of men declines” (p.59). Since the UN forecasts global urbanization rising to 68% by 2050, they think fertility decline will be faster than expected. The theory is flawed. One clue is where fertility decline began, in France not in Britain which lagged by 80 years despite being more urban. To their credit,14 Bricker and Ibbitson admit: “In France, oddly, fertility declines were already underway by the late 1700s” due to changing ideas (p.61). Tim Dyson observes “when the French began to employ birth control in the 1790s, only about a quarter of France’s population lived in towns” but “when the English turned to birth control in the 1870s around three-quarters of the population lived in towns”.15
There is no universally applicable formula. The factors mentioned by the authors are all important, but from one country to another they have interacted in varying sequences and patterns. One factor Empty Planet neglects is initiatives to disseminate new ideas favouring small family size (for example Population Media Center have produced radio and TV shows spreading this message successfully). In some countries, for example in Algeria, Egypt, and Indonesia, a period of active promotion of smaller family size coincided with a faster rate of fertility decline, and a subsequent weakening of those family planning initiatives in the mid 1990s began a period in which fertility decline slowed or stalled temporarily, despite a simultaneous increase in urbanization and education.16
Most of the numerical differences between forecasts from the UN, IIASA and others, depends on what happens in mid-Africa (South Africa, neighbouring Botswana, and northern Africa have already substantially reduced fertility rates). The IIASA cautions that “the future of world population growth will largely be decided in Africa with the future education of women as a main determinant … There has been recent moderate progress in education expansion but continued progress is not guaranteed”.17 Bricker and Ibbitson visited Nairobi and found trends accelerating fertility decline, but as they admit Kenya is doing better than many African countries (p.118). In the 1980s Kenya was one of the few African countries with a family planning outreach policy personally backed by then President Daniel arap Moi. The countries with the highest and slowest changing birth rates are elsewhere, and it only takes a few persistently high outlier countries to perpetuate growth in global population. John Bongaarts in an illuminating 2017 article showed that the difference in fertility between mid-Africa and developing countries elsewhere is due not only to a later start, but also often a higher TFR than would be expected at a given level of development.18
The UN 2017 forecast accurately reflects recent trends.19 I hope that above-replacement birth rates do decline faster than expected, so forecasts can be revised lower. An earlier peak is possible, but the pace of change needed is unlikely to result from urbanization alone. It would also require accelerated action to increase the proportion completing secondary education, spreading the small family ideal and wider access to family planning technologies enabling women to achieve a smaller family size. Part 2 of this review will consider Empty Planet’s forecast of long-term population decline, the consequences they expect, and their prescriptions for how governments should respond.
John McKeown (PhD Liverpool, M.Sc. UWE Bristol) formerly worked at an environmental non-profit, and as a university lecturer. He now lives in Cornwall.
- United Nations. World Population Prospects 2019. https://population.un.org/wpp/
- Sanyal, S. Predictions of a Rogue Demographer. Deutsche Bank; 2013.
- Sanyal, S. The End of Population Growth. The Wide Angle. Deutsche Bank; 2011.
- Meadows, D., Randers, J., Meadows, D. Limits to Growth: The 30-Year Update. Earthscan; 2005, page 169.
- Randers J. 2052: A Global Forecast for the next Forty Years. Chelsea Green; 2012.
- With thanks to Patrick Gerland for explaining the effect of aggregating deterministic variants, and pointing to https://population.un.org/wpp/Graphs/Probabilistic/POP/TOT/
- Lutz W., Butz, W., Samir KC. (eds.) Executive Summary : Demographic and Human Capital in the 21st Century. IIASA; 2014.
- Wittgenstein Centre Data Explorer http://dataexplorer.wittgensteincentre.org/
- Lutz et al. 2014, http://pure.iiasa.ac.at/id/eprint/11189/
- Lutz W., Goujon, A., Samir KC, Stonawski, M., Stilianakis, N. Demographic and Human Capital Scenarios for the 21st Century: 2018 assessment for 201 countries. IIASA; 2018. p.124
- Lutz et al. 2018, pp. 121, 128.
- USAID. The DHS Program – Team and Partners. https://dhsprogram.com/Who-We-Are/About-Us.cfm.
- A popular video falsely portrays (playtime 3:20) fertility decline originating in Britain as its origin in France did not fit a similar theory. https://ed.ted.com/featured/t1pefMVY
- Dyson, T. Population and Development: the Demographic Transition. Zed Books; 2013.
- O’Sullivan, J. Synergy between Population Policy, Climate Adaptation and Mitigation, in: M.Hossain et al (eds.) Pathways to a Sustainable Economy. Springer; 2018.
- Lutz et al. 2018, p. 9
- Bongaarts J. Africa’s Unique Fertility Transition. Population and Development Review, 43.S1 (2017): pp. 39-58.
- Gerland, P. et al. World population stabilization unlikely this century. Science 346 (2014) pp. 234-7.