A place for new ideas to settle.

16 July 2014

Why culling Earth's human population isn't rational


In a discussion on Facebook tangentially related to my last blog post (insofar as it was about how political views aren't rational for a lot of people), my friend Briana made the following comment:
I've never seen it [politics] as rational... there is an ideal that it can be, but I've never seen it in practice. Otherwise, we'd have started culling our own populations by now.
I objected pithily that we don't need to cull humans because humans are capable of using birth control (including the decision not to reproduce), to which she elaborated:
No, we cull other animals because we don't mind killing them from a moral standpoint (i.e. we are less repulsed emotionally). If we did mind as much as we do for ourselves, we'd find other ways to manage their populations like we do for ourselves. Culling is the rational choice (most economical, quickest and easiest). Finding other ways to appeal to our personal security that are less effective, take longer, and are more costly is the irrational but morally acceptable choice.

It reminds me of that study posted a while back (in Freethinkers?) about the difference in choices for moral dilemmas depending on the language used. Essentially, that showed the less "personal" the dilemma was, the easier it was to make the rational choice about it. Most political issues are very personal in order to get people interested (this is where your boring thing comes in) and so positions are irrationally conceived.

I'm not saying we should cull humans, btw, I mean, I think "rationality" is put on too much of pedestal by western society. [Link added.]
My reaction to that last statement is... complex. On the one hand, there's a strain of Western culture that does exalt "rationality," or at least it exalts the appearance of rationality. But there are broad currents of general anti-intellectualism in Western (particularly American) thought (cf. Richard Hofstadter's essay) so it's not so easy to say one way or another.

But that gets us too far afield of the primary question: is culling Earth's human population rational?



But let me elaborate why.


First, let's consider why we cull any populations at all, from deer to kudzu vines. I can think of several primary reasons:
  1. The species is overfeeding;
  2. The species is overbreeding;
  3. The species is detrimental to human life or civilization.
Of course, overzealous invocation of reason #3 can lead to, say, too-sparse population density of natural predators, which leads to reasons #1 and #2.

For example, the deer population has exploded in the wake of overhunting wolves. The deer population, freed from natural predators, will fluctuate wildly as they overeat the local foliage and then starve to death. This is essentially reason #1.

As another example, nitrate fertilizer runoff into the Mississippi delta has resulted in vast algal blooms. These blooms rapidly deplete the oxygen in the local volume of water, killing all the fish. It's not that the algae has no natural predators, but an outside influence is greatly increasing its reproductive rate. This is essentially reason #2.

As for reason #3, consider smallpox or polio. As these are infectious diseases that cause widespread human misery, we culled their population to near-zero in the Western world (to extinction in the wild, in the case of smallpox).

Note, though, that really all these reasons involve human civilization. And fair enough; as far as we know, we're the only species with a meta-level concern for other species, even one rooted in self-interest.

Now let's examine each of these reasons in turn, vis-a-vis the human population.


Question #1: Is the human population overeating?

This and the next question are really two sides to a more general question: What is the earth's carrying capacity for humans (HCC), and is the current population above or below that capacity? And this is a really hard question to answer. Other species only have limited ability to change the environment or adjust their own behavior patterns, compared to humans. So algae can cause blooms and deplete oxygen, but there's no Algal EPA or Algal Greenpeace warning against the dangers of oxygen depletion to the local waters. Deer can defoliate a local grove, but there's no Deerpartment of the Interior to re-plant those ferns and bushes.

In short, the HCC is only contingent on current human activities and behavior patterns, which are themselves changeable as technology is invented, science advanced, and culture shifted. A quick look at recent history shows, contra pessimists on all fronts, that we have made stunning progress in a few areas.

Incidentally, HCC has been estimated as high as 10 billion people, so we're not quite there yet. Let's put this question in the Maybe pile.

Question #2: Is the human population overbreeding?

Okay, so we may not be facing a Malthusian crisis right now, but maybe it's only a generation or two away. Except that human societies tend not to reproduce too much if they're sufficiently well off. Hans Rosling, in particular, has made much of the fact that rich countries (the West, but this is also evident in the middle economies) are at or even below replacement rate; he projects that the human population will hit a steady state (about as many births as deaths, year over year) around... 10 billion people.

Of course, it's all contingent on technology, culture, and so on. But notable in Rosling's analysis was that his model depended only on economic choices made by women, that is, whether or not to have children.

I'm a fan of Rosling, so I'll commit to No on this one.

Question #3: Is the human population detrimental to human civilization?

Well, this is a fascinating question. Personal preferences and moral sentiments (more on those later) point me towards No but I don't really know how to nail this one down, so I'll answer Mu.

But I also don't want to be glib. Various ideologies really do have various and definite answers to this question. Most broadly, Western liberalism would answer No: humans are on the whole virtuous and productive and we like civilization, liberalism tends to say, so therefore the more humans living together and doing things, the better we all are.

Western conservatism would probably answer Yes: humans can't in general be trusted to get along, so it's better to live in small close-knit groups and make sure nobody else touches your stuff. Large populations tend to be chaotic and therefore dangerous to stability. Neo-reactionaries (N.B. Scott Alexander is not a reactionary) seem to be less inclined to beat around the bush here.

Western leftism would probably answer No, but. Humans are useful and human society can be coordinated in a virtuous way, but you have to make sure your humans are doing it the right way first. So in general, a random newcomer human should be regarded with suspicion because they're Schroedinger's shitlord until proven otherwise. (Leftists also don't understand quantum physics, but then again nobody does.)

American libertarianism would probably answer either way, depending on how much any given American libertarian is convinced that non-libertarians are statist low-information sheeple. Heh.


There's another thing; culling isn't the only form of population control. This is especially important if your reasons for concern are more like #2 or a soft version of #3: if you can just keep the population from actually getting to a crisis point, that'd be much easier for everyone.

So we try to spay and neuter our domesticated animals, for example, rather than throw bags of kittens into the river or whatever. I mean, people still do that, but in places where veterinary clinics are easy to come by it's kinda frowned upon.

With humans, we have several programs of population control:
  1. Wealth. Yes, the more well-off a woman is, the less likely she is to have more than a couple kids. See this excellent presentation by Hans Rosling, detailing the negative correlation between per capita income and children per woman. So strong is this correlation, in fact, that it kinda trumps the following, more proximal reasons.
  2. Abstinence/celibacy. We can choose not to reproduce. Not every species is so lucky.
  3. Contraception and abortion. We have the ability to prevent or terminate pregnancies at will. Again, not every species is so lucky.
  4. Sterilization. A more permanent solution than contraception and a less ephemeral than a vow of abstinence. Historically our society has even done it to people without their consent.
  5. Infanticide. Also a permanent solution. Also grisly as hell, and unfortunately not a thing of the past. At least it's not institutionalized.
  6. Genocide. Just remember that as soon as you start killing people in other countries, the world won't stand for that after a few years. Stupid man...

Now let's move on to practical considerations. Somewhat paradoxically we'll ignore stuff like legal restrictions (systematized mass murder is, after all, a crime against humanity) and step into hypothetical territory. The first thing to note is that humans have moral sentiments. In particular we have a sense of fairness, and a compassion for those close to us. A program of culling, absent any otherizing propaganda, is bound to generate lots of fear and vengefulness.

Moreover it would be hard to convince a typical human to just start killing other humans, even with mildly good reasons. You'd need inevitability on the order of a trolley problem, except that really any good reason (a true Malthusian catastrophe) has a relatively long time horizon, so Joe Schmo will probably opt to do nothing (proximal agency in a mass murder is way worse than distal generalized suffering).

One might say that moral sentiments are irrational. I don't think that's right; I think they're arational, that is, decoupled from rationality. Rationality is a pattern of thoughts, while moral sentiments are feelings, or at least they originate as such. Yet they do have a place within rationality, because they provide evidence for shaping our beliefs. If I strongly feel that something is unfair, for example, that's a signal to investigate further. Yes, it may be that our moral sentiments are at times too coarse or misdirected. And they can be trained or retrained, like any other perception. But, properly considered, they're just another tool in the rationalist's toolbox.


And with that, let's consider one possible moral argument against culling the human population. Again, that the suggestion seems ludicrous and horrible to so many people, is evidence that maybe it's a bad suggestion. (To be fair, though, there are enough things that seem horrible at first glance, which are actually good in some way. So it's worth checking out.)

Suppose that we decide to build Ultron, a machine intelligence with a mission to save the human race who decides that the most efficient way to do that is to just start killing the least useful humans until we reach parity. Ultron has enough processing power to calculate all this stuff out to arbitrary precision, so we know it can't screw up too badly.

Well, again, unless Ultron decides that it also needs to propagandize everyone, there would be a massive freakout backlash that would probably cause even more death and destruction than intended by the culling program.

But let's imagine further that, since Ultron is a very powerful machine intelligence, it can cull instantly with micromissiles: whirr-bang, you're dead, no announcement necessary.

What sort of world would this be? It's one in which a hard lower limit has been set on the value of human life, enforced by death. Does this promote human virtue, to know that you're only alive because a distant power deemed you useful? That if you slip even slightly over that line, you're damned to die by what amounts to a cold, utilitarian god?

This is not a world I want to see. It's a mockery of civilization. It's worthless to humanity, and in it, humanity is worthless.


Note that we gave our hypothetical killing Mind zero compassion and arbitrary computing power. Humans have compassion (among other moral sentiments) and finite computing power, indeed very limited computing power. There's no guarantee that we can create a perfectly utilitarian machine intelligence. And this program failed even with zero compassion and arbitrary computing power!

But don't think that I'm going to fall prey to #humanormativity here. Already we're loath to cull species that are close to us in sentience (it helps that many of these are endangered). Better if we could limit our killing even further, while still managing ecosystems and our place in them.

Except pathogens. Fuck those guys.