Hide table of contents


What socially beneficial causes should philanthropists prioritize if they give equal ethical weight to the welfare of current and future generations? Many have argued that, because human extinction would result in a permanent loss of all future generations, extinction risk mitigation should be the top priority given this impartial stance. Using standard models of population dynamics, we challenge this conclusion. We first introduce a theoretical framework for quantifying undiscounted cost-effectiveness over the long term. We then show that standard population models imply that there are interventions other than extinction risk mitigation that can produce persistent social benefits. In fact, these social benefits are large enough to render the associated interventions at least as cost-effective as extinction risk mitigation.


In the coming century, humanity may face global catastrophic risks stemming from climate change, nuclear war, pandemics, and emerging technologies such as artificial intelligence (H¨aggstr¨om 2016; Ord 2020). Many interventions for reducing these risks are likely to be cost-effective by the light of standard cost-benefit analysis (Posner 2004; Shulman and Thornley, forthcoming). However, it has often been argued that, under a zero rate of pure time preference[1], special priority should be given to the subset of these interventions that most effectively reduce the risk of human extinction. In a widely cited passage, Parfit (1984: 453) introduces this line of argument by comparing three possible outcomes: 

(i) No catastrophe occurs. 

(ii) A catastrophe kills 99% of the existing world population. 

(iii) A catastrophe kills 100%. 

Insofar as human life is valuable, (i) is clearly socially better than (ii), which in turn is better than (iii).[2] But which of these two differences is greater in terms of welfare loss? Counterintuitively, Parfit and many others have argued that, although the welfare difference between (i) and (ii) is greater if only the current generation is considered, the welfare difference between (ii) and (iii) is greater if all generations are considered equally. The motivation for this is that, while any global catastrophe would lead to an immense welfare loss for the current generation, human extinction would additionally lead to an even greater welfare loss by irreversibly preventing all subsequent generations from coming into existence.[3] Therefore, in Parfit’s view, “[w]hat matters most is how we respond to various risks to the survival of humanity” (Parfit 2017: 436, emphasis added). This line of thought has been invoked in cost-effectiveness analyses of interventions that reduce the risk of human extinction posed by asteroids (Matheny 2007), climate change (Ng 2016), and pandemics (Millett and Snyder-Beattie 2017). We refer to it as the longtermist argument for prioritizing extinction risk mitigation (or simply, ‘the longtermist argument’). 

An important assumption underlying the longtermist argument for prioritizing extinction risk mitigation over other types of risk mitigation is that the welfare effects of human extinction would be permanent, whereas the welfare effects of a nonextinction catastrophes would not. More precisely, the argument assumes that, if a non-extinction catastrophe were to occur, humanity would have a good chance of eventually recovering. However, the likelihood of such recovery depends on people’s fertility decisions, which in turn depend on economic and social factors. Understanding these factors is necessary for determining whether extinction would indeed be uniquely consequential in the long run, or whether some non-extinction catastrophes would have comparably persistent effects on long-run population and welfare levels (cf. Ord ms-a). 

In this paper, we explore how shocks to the size of the current population might affect long-run population levels, and what this implies for philanthropic priority setting. We start by introducing a theoretical framework for quantifying the undiscounted cost-effectiveness of risk reduction efforts. A heuristic implied by this frame work is that the undiscounted cost-effectiveness of reducing the risk of a negative population shock is proportional to the ratio of lives lost in the long run (in percentage terms) to lives lost in the short run (in percentage terms). 

In the remainder of the paper, we assess the implications of various population models for the relationship between decreases in current population levels and longrun population levels. First, we discuss shocks that reduce the current population level, but that leave all other factors of production unaltered. We show that, for such shocks, the assumption that population levels eventually recover after any nonextinction shock is implied by the Malthusian model of fertility (Malthus 1798). Importantly, however, this assumption is not implied by models that take fertility choices to be primarily determined by social norms. Nor is it implied by the Barro-Becker model (Becker and Barro 1988; Barro and Becker 1989), which is the workhorse model for studying the economic determinants of modern fertility dynamics. Indeed, in our calibration of the Barro-Becker model, non-extinction shocks to current population levels can result in permanent drops in long-run population levels that are disproportionately larger than the size of the initial shock. We then proceed by analyzing events that reduce both population size and other factors of production proportionally by the same amount. Given constant returns to scale technology, such events leave economic determinants of fertility choices unaffected and therefore result in a permanent, proportional reduction in the size of the global population. Interventions that save lives and increase the capital stock in equal proportion therefore have permanent effects in standard economic fertility models. Our undiscounted cost-effectiveness framework suggests such interventions could be as cost-effective as extinction risk mitigation. Moreover, a back-of-the-envelope calculation suggests that these interventions may be even more cost-effective than extinction risk mitigation provided that the determinants of population levels remain suciently stable far enough into the future. While these cost-effectiveness estimates should be interpreted with considerable caution, they nonetheless suggest that interventions other than extinction risk mitigation could have significant impact on long-run social welfare.

Read the rest of the paper

  1. ^

    Adopting a zero rate of pure time preference amounts to not discounting future welfare. Note that this is fully consistent with discounting future consumption based on the expected rate of economic growth and the diminishing marginal utility of consumption. Surveys of the arguments for and against adopting a zero rate of pure time preference are found in Dasgupta (2008), Greaves (2017), and Groom et al. (2022).

  2. ^

    For views to the contrary, see e.g., Benatar (2013) and Pettigrew (2022).

  3. ^

    Parfit writes, “Earth will remain habitable for at least another billion years. Civilization began only a few thousand years ago. If we do not destroy mankind, these few thousand years may be only a tiny fraction of the whole of civilized human history. The difference between (2) and (3) may thus be the diffference between this tiny fraction and all of the rest of this history. If we compare this possible history to a day, what has occurred so far is only a fraction of a second” (Parfit 1984: 453-454).





More posts like this

Sorted by Click to highlight new comments since: Today at 12:43 AM

I only skimmed this paper, but as far as I can tell it never responded to one of the most central reasons for why a Malthusian model makes sense, which is selection effects and evolutionary pressure. The paper brings this up themselves, but then doesn't seem to do anything with that: 

The case for nonetheless considering the Malthusian model is that Malthusian population dynamics may reemerge in the long run. First, evolutionary pressures for higher fertility might increase long-run population growth to the extent that natural resource constraints become binding once more (cf. Bostrom 2004; Collins and Page 2019).14

Over the course of hundreds of generations, we should expect huge memetic and genetic selection towards higher fertility rates, so it seems pretty implausible to me you end up with a population permanently substantially below carrying capacity, unless you also posit the development of some enforcement mechanism that prevents people from having children.

I don't really know why we should assign much validity to the alternative population models you outline, on the timescales that we are talking about (100,000+ years). The basic selection effect argument seems much stronger than the support for these other models on their long-run fit, so it seems pretty confused to me to consider them seriously.

Thanks – this is a very important point and I am glad that you raised it! Overall, I think we should be very uncertain about what the long-run population dynamics might be after a catastrophe. I am not sure how much we disagree, but I tried to add some thoughts below (Note: Maya might not necessarily agree with my thoughts on this issue).

As you point out, we write in the paper that Malthusian population dynamics may reemerge in the long run and that evolutionary pressures are one of the main reasons to expect that this might happen. We do not directly argue that such reemergence won’t happen, but in the footnote that follows immediately after that passage you quoted we write:

"In contrast, Arenberg et al. (forthcoming) argue that “empirical facts and models of heritability do not provide reason to conclude that positive population growth is bound to continue via the dynamics of a higher-fertility type making up an ever-increasing share of the global population”.

Here's an attempt to quickly explain Arenberg et al.’s argument: Arenberg et al. point out that we “should not conflate higher fertility within a heterogeneous population with high or above-replacement fertility: it is an empirical question whether future higher-fertility sub-populations will have above replacement fertility”. They then argue that “there is strong historical and global evidence that even higher-fertility groups will eventually trend to near or below replacement fertility”. Drawing on these insights, they introduce a model indicating that “long-term population growth can be negative even with both strong heritability and an above-replacement-fertility sub-population”.

Now, it could reasonably be argued that evolutionary pressures will nonetheless determine fertility rates on 100,000+ year timeframes. However, even if this is correct, such timeframes are only relevant if there is at least a decent probability that humanity will still be around in 100,000 years (conditional on surviving this century). This is not obvious; for instance, if the background probability of human extinction is 1% in each century, then the probability that humanity is still around after 100,000 years is only 0.004%. It is therefore not clear to me that the Malthusian model is correct, so it seems sensible to take other models of fertility seriously.

Thanks again for engaging with the paper!