No, having children won’t destroy the environment - Part 1
Why the carbon emissions of having a child are way smaller than you think
Note: Just before publishing, I realised that Boom has already done a very similar piece and comes to the same conclusion as me.
When you google “climate impact of having a child”, you see statements like “adding a child to the planet can add up to 9,441 metric tons of carbon dioxide to the environment” and “Having one fewer child will save 58.6 tonnes of CO2-equivalent per year”. These numbers come from two papers, Murtaugh and Schlax 2009 and Wynes and Nicholas 2017. While these numbers are widely cited and make some people think that they really shouldn’t have children, it turns out that they are based on outdated data and pessimistic assumptions that don’t match reality.
The papers claiming that the carbon footprint of having an extra child is massive
Murtaugh and Schlax 2009 focus directly on calculating the carbon emissions from an additional child. Wynes and Nicholas 2017 compare different actions that individuals can take to reduce their greenhouse gas emissions. The second paper surveys the literature for estimates of climate impacts of different actions so they actually end up relying on the numbers from the first paper.
The results in Wynes and Nicholas 2017 look like this:
Source: Wynes and Nicholas 2017, visualisation by Claude Opus 4
Every single action that you could take is totally dwarfed by the impact of having one fewer child!
The carbon emissions paths that they assume don’t match reality
Okay, but how do they get to the total impact and should we actually trust their methodology?
Murtaugh and Schlax 2009 calculate the carbon emissions not only of the child but also of all future expected descendants in order to quantify the impact of having an additional child. They attribute 50% of the child’s life-time emissions to each parent, 25% of any grandchildren’s emissions, and so on. This means they have to model fertility rates and carbon emissions over time in order to get to a final number of extra emissions added by having an additional child. I won’t look at their fertility assumptions in detail and just show how the carbon emission modelling goes quite wrong.
They use three different scenarios for carbon emissions: One optimistic one, a “constant” one, and a pessimistic one. Their baseline values are per capita emissions in 2005. In their constant scenario, emissions stay at the 2005 level indefinitely. In the optimistic scenario, emissions linearly fall from the 2005 level to 0.5t CO2 per person per year in 2100 and stay there indefinitely. In the pessimistic scenario, emissions linearly rise to 1.5 times the 2005 level by 2100 and stay there indefinitely. It’s probably immediately obvious that none of these scenarios is particularly realistic and I will show you in a second that in many developed countries, per capita emissions have actually been falling faster than even the optimistic scenario from the paper!
Wynes and Nicholas 2017 want to give recommendations for individual actions to take in developed countries and choose the US, Russia, and Japan from the list of countries that Murtaugh and Schlax 2009 do calculations for. It seems weird that their 58.6t number is supposed to give you a sense of the impact of having one fewer child in a developed country and is based on an unweighted average of the US, Russia, and Japan. That is hardly a good proxy for the average developed country.
Let’s see how the data actually compares to the different scenarios based on the countries in the paper1:
Source: Our world in data
The dashed lines indicate the three different scenarios and the solid lines are actual per capita emissions. We can see that while Russia is currently on track for the pessimistic scenario, Japan is most closely following the optimistic scenario and in the US, per capita emissions are clearly falling even faster than in the optimistic scenario. If we assume the pessimistic scenario for Russia2 and the optimistic scenario for the US and Japan, the impact should actually be 48.0t per year per person in Russia, 2.7t in Japan, and 7.0t in the US3.
Now we immediately have a result where the impact of one fewer child becomes comparable to some of the other high impact actions! For example, the action with the next highest impact is living car free at 2.4t per person per year, almost the same as the 2.7t for having one fewer child in Japan. Keep in mind that this is all calculated from the point of view of being in 2005: For the US and Japan, the numbers for thinking about having a child in 2025 would be even lower and I also haven’t even taken into account yet that the US is doing better than the optimistic scenario.
Other countries are doing even better
And the US and Japan aren't even the poster children when it comes to per capita emissions. Let’s look at Sweden which is one of the richest countries with the lowest per capita emissions. I’m also adding Germany and the UK, the country where I grew up and where I live now:
All these countries are doing better than the optimistic scenario! Per capita emissions are also significantly lower than in the US and even lower than in Japan in all those countries (though Germany is pretty close to Japan). That means that for these countries, the impact of having a child will again be even lower than the 2.7t per person per year that we found for Japan.
Conclusion
After taking a closer look at the methodology of these two papers, I think it’s clear that their numbers just don’t hold up. The climate impact of having an additional child is nowhere near as bad as has widely been claimed.
An obvious objection to my analysis is that the impact of having a child might still be significant, even if I have shown that it is much lower than often assumed. I will tackle this objection in my next post where I will look at a new paper studying the aggregate impact of population changes on global warming. This aggregate analysis shows clearly that under realistic assumptions of the timing of population changes and decarbonisation paths, having fewer children is not an effective strategy to combat climate change.
Many people don’t realise how much per capita emissions are already falling in developed countries. We should focus on the technological changes necessary to bring per capita emissions down further instead of making people feel bad for wanting to have children.
Thanks to Ben Snodin for helpful discussions while working on this post.
Note that this is per person per year so they take the total impact of having one fewer child and divide that by the average lifespan.
Of course I don’t believe that things will actually be so bad in Russia in the future but for the sake of the argument I’m just assuming that we’re on the path the data is currently suggesting.
The total emissions in the optimistic scenario are 562t for the US and 233t for Japan. The pessimistic scenario in Russia has 3497t. Wynes and Nicholas 2017 seem to use expected lifespans of 80.2 years for the US, 85.5 years for Japan, and 72.8 years for Russia to convert the totals to the impact per year. I follow this approach to get to the corrected numbers per year.