PRELIMINARIES

Hooker's preferred brand of rule-consequentialism holds that an act is wrong if and only if it is forbidden by the ideal code. This is the set of rules whose internalization by the overwhelming majority of everyone everywhere in each new generation has maximum expected value in terms of well-being (with some priority assigned to the worse-off).Footnote ¹

I will focus on the prevent disaster requirement in this article. It instructs us to sacrifice more than usual – and break the other rules if necessary – in order to promote the good when faced with calamity.Footnote ² The merit of everyone having internalized such a rule is clear. In exceptional circumstances where a sufficiently large amount of expected value is at stake, it would be far better on the whole were people capable of enormous sustained sacrifice. As Hooker puts it, ‘[when] necessary to save the world, or even just some significant proportion of humanity, one may be required to make an extreme – even the ultimate – self-sacrifice’.Footnote ³

Here's the rub. The prevent disaster requirement threatens to make rule-consequentialism problematically overlydemanding. Take global hunger, for example. The starvation of innocent people surely counts as a disaster. However, an unlimited requirement to prevent disaster would entail, in so far as there is a staggering number of persons affected by food insecurity, that the relatively affluent should give up nearly everything, even forgoing personal projects or deep personal relationships – at least to the point where they are themselves impoverished.Footnote ⁴

The reply from Hooker is that this criticism overlooks an important variable in our moral evaluations. The expected value of a rule is only partly determined by the benefits to the poor of getting aid and the costs to the affluent from donating. Missing from this equation are the costs of getting the rule internalized by everyone – including those who are least able to promote the good.Footnote ⁵ Even the poorest, most disadvantaged persons, after all, may find themselves in a position to save another's life.

Now, the safe assumption is that inculcating and sustaining in everyone the desire to abide by a rule will be more difficult the more demanding the rule is. Such costs militate in favour of less demanding rules (in general) for two reasons.Footnote ⁶ First, some logically possible codes are so overbearing, alienating or intricate that too few if any persons are capable of adhering to the rules, no matter how incredible the benefits promised by such a code.Footnote ⁷

Second, the time, energy, attention and mental duress of internalizing a very demanding rule about preventing disaster could be ruinous.Footnote ⁸ And note that the benefits received by the global poor or hungry would themselves be (to some extent) counterbalanced by the costs of internalizing in them a completely impartial altruism towards others. Furthermore, internalizing the code is not a one-off cost, but rather something incurred with each new person added to our population – repeating indefinitely. ‘It is not as if infants begin with as much altruism as their parents eventually internalized. . .[each] new generation would need to be transformed from beings concerned mainly with immediate gratification, personal comfort, and self-assertion to impartial beings willing to make virtually endless sacrifices for others.’Footnote ⁹

At first blush, the internalization costs of a very demanding altruism towards (among other things) abating global hunger would outweigh the benefits of improving the lives of the poorest in our population. If so, some point short of this level of sacrifice would instead produce the most well-being, given the costs.

I accept that the costs of internalizing a very demanding rule about preventing disaster could, given the state of a world,Footnote ¹⁰ produce less overall good than a moderately demanding rule. However, I will argue below that existential risk – that is, a risk that ‘threatens the premature extinction of Earth-originating intelligent life or the permanent and drastic destruction of its potential for desirable future development’Footnote ¹¹ – does not support this being true of our world.

PROLONGING HUMAN HISTORY

My case rests on the astronomical waste produced by failing to prolong mankind's place in the world.Footnote ¹² I will start by rehearsing a classic thought experiment from Derek Parfit. He writes in Reasons and Persons,

Certainly, nuclear war would bring about tremendous losses for all of those alive at the time. This alone provides good reason to take steps towards avoiding a nuclear disaster.Footnote ¹⁴ But perhaps the greatest loss occurs when the last of humanity is snuffed out by nuclear winter, given the whole of humanity's future follows them to the grave. Indeed, the vast reduction of the possible sum of happiness due to our extinction is literally astronomical, given the long-term potential of humanity. The destruction of mankind would be, as is often said by the likes of Parfit, Sidgwick, and others, the greatest of all conceivable crimes.Footnote ¹⁵ It constitutes what Bostrom calls an existential catastrophe.Footnote ¹⁶

This brings us to the point I now wish to press. Yes, Hooker is correct to alert us to the cost of internalizing in everyone the prevent disaster requirement. But we should likewise recognize that – on the assumption that additional lives worth living are of contributive value for our populationFootnote ¹⁷ – there are goods, not just in improving the well-being of the poor or worse-off, but in prolonging humanity. My contention is that once we take stock of these goods in our moral evaluations, then a more demanding rule about preventing disaster (that is, our extinction) will produce higher expected value.

This can be demonstrated by first imagining a population of n persons. n stands for the complete population of mankind across time in this possible world. Let's say that their average welfare is A and that this already includes the cost of internalizing a moderately demanding code. Straightforwardly, the total sum of well-being in this population corresponds to n × A. As seems likely, let's also say that there are lots of catastrophic things that have some chance of happening to this population in the future. The total risk of doomsday occurring is not important for our purposes. All that is required here is that we accept that doing more to lower the chance of mankind's extinction thereby (a) increases the expected size of this population,Footnote ¹⁸ and (b) increases the cost of internalizing the ideal code by everyone in this new population. Further, m will stand for the members added to this now larger population.

Let g be a finite positive number (1 ≥ g ≥ 0) that represents how much the good life for this population is reduced, given higher internalization costs. We can then say that the total sum of well-being in the new population is equal to (n + m) (A × g). Notice that if g = 1, then their average welfare remains the same as before (and m = 0). Conversely, if g = 0, then they are scraping by on muzak and potatoes, a life barely worth living (and m > 0). In other words, the lower g is, the worse the quality of life of the population. A more demanding prevent disaster requirement is of greater expected value whenever n × A < (n + m) (A × g). That is to say, there is some value of m such that a more demanding rule is better than a moderately demanding rule, namely $m > \frac{n-n \times g}{g}$ .

This counter-intuitive result can be neatly summarized as follows: If what's supposed to be holding back the demandingness of the ideal code are the costs of internalization, then the goods promised by a long future for humanity will swamp those costs in our moral evaluations.

Notice, the problem can be cast two different ways. First, the loss associated with humanity's premature extinction is so great that even if the probability of a catastrophic event is very low, an expected value calculation suggests that we should strive to prevent its possible occurrence. And yet, there is something deeply puzzling about ruining the lives of all actual persons for the sake of humanity eking out a longer stay in the universe.

Second, you may have realized that the above implication bears close resemblance to the dreaded Repugnant Conclusion. The Repugnant Conclusion states that for any population, all with a very high quality of life, there must be some larger imaginable population whose existence, all else being equal, would be better despite their lives being barely worth living.Footnote ¹⁹ The mistake, as countless critics have noted, is that quantity (that is, size of population) should not be able to compensate for a stark reduction to their average quality of life.

I'm inclined to agree that this looks worrisome. For some, if this were the end of the story, it would surely act as a reductio ad absurdum of the view. But this is not the full story.

AN INDIRECT APPROACH TO LOWERING THE THREAT OF EXTINCTION

In setting out our earlier comparison of the two populations it was assumed that only costs go up, never benefits. That is to say, A was fixed and the total sum of goods went up merely because the size of the population grew, despite internalization costs reducing average quality of life. Colouring in the picture, this corresponds to the scenario where, all else being equal, existential threats are directly targeted. To illustrate, this could amount to putting a lot of resources towards asteroid deflection programmes.Footnote ²⁰

I now wish to argue that we could instead reduce existential risk by indirect means, and in so doing make the world in two ways go better. As noted earlier, we would prolong humanity's place in the cosmos. Furthermore, an indirect approach improves the average welfare of persons, particularly the worse-off in our population.

Certainly, it would be a mistake to concentrate exclusively on indirectly lowering the probability of doomsday. Returning to our earlier example, reducing global poverty cannot prevent an Earth-bound asteroid the size of Texas from making impact. Nevertheless, if we were also to adopt an indirect approach, then this would contribute to existential risk reduction by curbing the negative ripple effects of readily preventable illnesses, global hunger, and so forth.

Ripple effects are a class of phenomena that affect the far future in significant ways, shaping how our history unfolds over time.Footnote ²¹ A ripple effect is initiated by a particular event that has some causal influence on the course of events that follow it. These events, in turn, may have their own impact on how further events play out. And so on it goes, reaching wider and wider as time passes.

Consider the following example. A doctor is in a position to cure some infant's blindness. Sure, the infant will probably have a better life after the operation. Most of us are quick to hone-in on this feature of the situation. And many other goods go unacknowledged by us as a result. Just a few of the proximate advantages we might reasonably expect to find after curing the infant's blindness include: her parents will be less worried about her, subsequently finding more free time to develop their own personal projects; the government will spend fewer resources on providing her education; this child will grow up with more opportunities, as well as perhaps being inspired to start a grassroots initiative or develop an anti-malarial drug. All of these consequences will have some role in shaping our future due to their own ripple effects. This network of ripple effects might go so far as causing ‘[her] country's economy to develop very slightly more quickly, or make certain technological or cultural innovations arrive more quickly’.Footnote ²²

My claim is that intervening in persistent global harms will contribute to existential risk reduction by removing harmful ripple effects in two important respects. I will focus on global hunger in making my case. My comments, however, are meant to apply generally – at least, to all cases of persistent global harms that might affect our leverage opportunities with respect to averting threats to humanity's future.

To begin, global hunger tends to make things worse around the world by its ripple effects, culminating in a range of independent doomsday scenarios. Indeed, not doing as much as is now possible to alleviate world hunger will mean that many more die from hunger and malnourishment-related illness. All of these deaths, consequently, produce their own growing burden in their communities. For example, a loss of working-hands in the fields means that their community's agricultural output suffers. An economy that depends on its agriculture sector could fall into disrepair and itself churn out further burdens as jobs and investments are lost. Consequently, the entire region could become socio-politically destabilized, fraught with civil war and marred by monstrous human rights violations. And this may introduce a hundred-year period of international hostility which invites the potential for nuclear engagement between warring parties. Or perhaps this region instead refuses to abide by treaties, such as by pumping out excessive greenhouse gases by mining coal in a fragile, protected rainforest.

Second, by failing to do more now to alleviate global harms our civilization is not just at growing risk of a catastrophe occurring, but also at growing risk of being unable to respond to an eventual existential threat. In other words, our leverage with respect to preventing a catastrophe could be severely weakened. It is not beyond the realm of possibility that in the aforementioned region, institutions equivalent to FEMA (that is, the Federal Emergency Management Agency) could become badly under-resourced, lacking the necessary funds and technical expertise to handle certain situations such as massive flooding – due to cancellation of training programmes, for example – or even non-existent if the government is corrupted by its military leaders, for example, and refuses to continue funding social programmes altogether.

I will now comment on some further benefits made possible by this approach.Footnote ²³ Just as global hunger has ripple effects, so does the absence of starvation in this region. The primary benefit that springs to mind here is that global coordination is far less likely to degrade between well-fed, happy populations than it is with military dictators of oppressed states. And when it comes to existential risk reduction, our success in coordinating globally is of high importance.Footnote ²⁴ Moreover, well-fed persons will be free to become educated, grow their economy, perhaps even getting involved in our civilization's scientific advancements, and so forth. They will make their children's future better. Lastly, if we solve global hunger, we will have, in effect, recruited more moral agents to aid us in eradicating malaria, poverty, or whatever happens to be next on our moral agenda. The altruistic burden will be spread out, and, perhaps after some extended period of incredible sustained sacrifice by our population, A will be higher than it has ever been in the past.

Tyrants and so on do not come from nowhere. They have a history. It seems to me clear that one plausible explanation for their presence in our world is that things tend to get worse around the world if (among other things) global hunger persists. In the hypothetical region we have been considering, if no one went hungry, then it is far more likely their crops would be tended, their cause for civil war and monstrous human rights violations jettisoned, and so forth. Therefore, the risk of humanity's extinction goes down by removing this fodder for catastrophe.Footnote ²⁵ And, of course, abating global hunger means that A (in expectation) goes up. Taken in tandem, the conclusion we reach is that rule-consequentialism ought to prescribe a very demanding rule about existential risk reduction, and, by splitting our efforts between direct and indirect approaches, humanity's future could be very bright indeed.

SUMMARY

I have argued in this article that the cost of inculcating in everyone a very demanding rule about minimizing existential risk (or curbing the astronomical waste produced by a broken world) does maximize expected value. My main thesis, however, was that this only looks problematic at first blush. As we have seen, taking an indirect approach to existential risk reduction, and so removing persistent global harms along the way, can improve the well-being of our population in significant ways. In the long run, we can expect things to get better, not worse, for our population.

So how bothered should we be about the severe costs of internalizing a very demanding impersonal concern for humanity's future? Does the conclusion we have reached count against rule-consequentialism? I do not see why it should. Hooker himself notes that this ‘possibility is a consequence of the idea that moral rules are to be selected from an impartial point of view. And this impartiality is a deeply attractive feature of rule consequentialism.’Footnote ²⁶ I wholeheartedly agree. And for those who still believe something has surely gone wrong here, the article closes by reminding them of the broken world that might otherwise lie ahead. As Mulgan puts it, ‘[can] we coherently picture the grim reality of their moral lives, and still refuse to make similar sacrifices for ourselves?’.Footnote ²⁷ Buck up.Footnote ²⁸