Plausible reductions in future population growth and implications for the environment

Bradshaw and Brook (1) conclude that a substantial slowing of global population growth is infeasible over this century short of drastic reductions in fertility, and therefore could not contribute much to lessening pressure on the environment. However, their methodology is too crude to produce credible results.

Bradshaw and Brook’s (1) principal results are produced with a population model that treats all people in the world as identical. Rather than treating Nigeria, which has a total fertility rate of 6.0 births per woman, differently from Italy, with a fertility rate of 1.4, Bradshaw and Brook use a single global region, essentially treating everyone in the world as if they were from Mexico (which is close to the global average). It is not possible to specify meaningful scenarios about future paths of fertility or mortality in a highly diverse world in such a model. For example, Bradshaw and Brook’s “Scenario 1” assumes constant global fertility and mortality rates and projects a population in 2100 of 10 billion. However, a scenario that truly held rates constant in countries around the world would actually produce 20 billion people by the end of the century (2).

Furthermore, in their “realistic” Scenario 2a, Bradshaw and Brook (1) make the unrealistic assumption that mortality rates for every age group decline by half by the end of century. It is well known that as countries transition from lower to higher life expectancy, mortality rates decline first in very young age groups and only later at older ages, a dynamic with substantial effects on population growth. Similarly, Bradshaw and Brook’s (1) simulation of increasing age at childbirth by allocating half of the fertility of women aged 15–24 evenly to those aged 25–49 is arbitrary and not based on evidence.

Bradshaw and Brook (1) also make selective use of scenarios. Their low-fertility scenarios include either the assumption that global fertility declines only to two births per woman, similar to central or best-guess scenarios from demographic institutions (2, 3), or that it declines all of the way to one birth per woman, simulating the imposition of a global one-child policy. It is no surprise then that the authors conclude that there are only two possibilities: little change in the expected global population growth path or draconian limits on child-bearing. Bradshaw and Brook (1) ignore the possibility of plausible low-fertility outcomes in between these scenarios (2, 3), which are consistent with experience in many countries that have already achieved low fertility and show that there is scope for slowing future population growth modestly by midcentury and substantially toward the end of the century. Although such scenarios are not considered the most-likely outcomes (3, 4), they would be facilitated by accelerated progress in female education, delayed first unions and first births, increased time interval between births, and expanded access to family-planning options.

We have used such scenarios to conclude that although slowing population growth would not be the largest nor necessarily most important factor in reducing greenhouse gas emissions, it could make a significant contribution, particularly in the long run (5). We stand by that conclusion.