摘要：全球生育率持续暴跌，人口收缩已成定局。是文明危机还是转型契机？《经济学人》说：少子化并非末日，人工智能与寿命延长将重塑社会结构。民族主义的生育鼓励政策为何注定失败？老龄化社会真正需要怎样的变革？本文理性拆解人口焦虑，带你看清未来世界的真实图景——减少的人口未必

有趣灵魂说

来自《经济学人》杂志编辑的推荐语：

译文为原创，仅供个人学习使用

The Economist ｜Leaders

经济学人｜社论

Peak human人类峰值

Don’t panic about the global fertility crash

不必为全球生育率暴跌惊慌

A world with fewer people would not be all bad

一个人口更少的世界并非全是坏事

1968年出版的《人口爆炸》中，生物学家保罗•埃利希（Paul Ehrlich）写道，人类繁衍速度如此之快，粮食将不可避免地耗尽，“数亿人”很快将饿死。在戏谑地提出“用星际运输解决过剩人口”的想法后，他主张实行严格的生育控制，“如果自愿方法失败，就采取强制手段”。

许多人仍在担忧人口过剩。但越来越多的人，尤其是富裕国家的人，却在担心相反的问题：人口暴跌。“低生育率将终结文明，”育有多子的埃隆•马斯克（Elon Musk）预测道。

尽管人口总数仍在增长，但生育率——即一名女性一生中预期生育的婴儿数量——一直在急剧下降。而且不仅仅是富裕世界如此： 目前全球三分之二人口所在国家的生育率低于2.1的“更替水平”——这是维持人口稳定所需的标准估计值 。哥伦比亚首都波哥大目前的生育率（0.91）甚至低于东京（0.99）。

根据联合国的主流预测，全球人口将在2084年达到104亿的峰值。但正如我们本周报道所指出的，其假设前提值得商榷。它假设从此刻开始，势头将发生突然转变：许多低生育率国家的生育率将停止下降或反弹，而高生育率国家急剧下降的生育率将放缓下降速度。如果这个假设错误，人类峰值来临的时间要近得多。如果当前趋势再持续仅仅十年，之后才出现联合国更乐观假设所预期的情况，那么全球人口将在2065年达到96亿的峰值，随后到2100年骤降至89亿。即便如此可能也过于乐观。

无论峰值何时到来，低于更替水平的生育率意味着全球人口将先缓慢收缩——然后急剧减少，这恰似此前指数级增长的镜像，正是那种增长使人口从1800年的10亿猛增至如今的80亿。这样的前景令许多人警觉。

一种担忧是广泛的经济层面的。 更少的人口意味着更少的头脑，因此创新速度会放缓。这意味着专业化和劳动分工的范围缩小。（如果你所在的城市只有1000人，想找到埃塞俄比亚餐馆或为你小众爱好服务的俱乐部，就只能祝你好运了。）快速收缩可能极具破坏性。沉重的公共债务将突然由更少的肩膀承担，其中许多还是日益老化的肩膀。特大城市或许没问题，但随着最后一座学校关闭，小城镇可能空心化。

另一种担忧则更为狭隘，带有民族主义色彩。 不同国家和群体的生育率差异很大。因此，有些人担心未来像自己这样的人太少，而被他们视为文化异类或威胁的人太多。这就是整个西方世界的民粹主义者倾向于贿赂家庭多生孩子的原因之一，也是唐纳德•特朗普承诺要成为“生育总统”的原因之一。

人口预测是确定性（所有到2070年将年满50岁的人都已经出生）与不可知性（如今的20岁年轻人会选择生育多少孩子？）的奇特混合。在漫长的时间尺度上，指数级收缩看起来快得惊人。然而，在社会必须应对该问题的初始阶段，变化的速度应该是可控的。

有若干理由可以质疑这些末日预言者。人工智能（AI）或许被夸大了，但其发展速度明显快于人口可能收缩的速度。因此，它或另一项尚未知的技术，必将缓解人类学者数量减少对创新造成的拖累。

另一个乐观的原因是，人类健康寿命不断延长，使人们能够保持更长时间的生产力。在41个国家的样本中，2022年70岁老人的认知能力相当于2000年53岁的人。也许这种进步会终止。但只要它持续下去，就会减缓劳动力的萎缩，为社会提供至关重要的额外几十年来适应。浪费人力资本的国家可能会找到减少浪费的方法，例如更好地喂养和教育年轻人，并消除女性工作的障碍。总之，人口减少未必意味着更贫穷。日本人口近二十年来持续减少，但生活水平显著提高。

民族主义者认为世界人口构成将发生变化，这一点是对的。即便是联合国的预测也显示， 到2100年中国人口将减少一半以上。印度人口保持稳定的时间会更长 。欧洲和美国可能通过移民推迟人口减少——或者他们选择不这样做。未来的世界将比现在更具非洲色彩，但非洲的生育率也在暴跌。巨大而渐进的地缘政治和文化转变是正常的。世界过去曾成功应对这些转变，未来也必定能再次应对。

生育主义者希望动用公共资金提高本国生育率，以对抗这些结构性趋势。他们必将失败。政府在为家庭生活提供便利方面可以发挥作用，但试图花钱让人们生育比原本意愿更多的孩子，要么成本高得惊人，要么根本无效。就连将GDP的6%巨款用于亲生育政策的匈牙利，其生育率仍低于更替水平，并且一些研究表明，其膨胀的婴儿补贴主要影响的是生育时间，而非生育总数。

前进与分化

人口的萎缩以及随之而来的人口老龄化，最终需要重大的经济和社会调整 。高龄老人需要照料（即使他们的花费不比年轻人高，后者通常需要长达二十年的抚养）。老年人更可能，因此他们的观点将塑造政治格局。这可能使得根据预期寿命提高退休年龄变得更加困难，但政府迟早必须这样做。

适应一个更空旷的星球并非易事，但这是可以做到的。所有关于人口灾难的预测在本世纪似乎都难以成立，而2100年如此遥远，对其后的预测似乎毫无意义。谁知道呢？到那时父母们或许拥有让育儿不再那么耗费精力的技术，家庭规模或许会再次扩大。但这仅仅是猜测。就目前而言，我们有理由关注，但没有理由恐慌。■

In “The Population Bomb”,published in 1968, Paul Ehrlich, a biologist, wrote that humans were breeding so fast that food would inevitably run out and “hundreds of millions” would soon starve to death. Having toyed with the idea of “interstellar transport for surplus people”, he advocated strict birth control, “by compulsion if voluntary methods fail”.

Many people still worry about overpopulation. But an increasing number, especially in rich countries, fret about the opposite: a population implosion. “Low birth rates will end civilisation,” predicts Elon Musk, a father of many.

Though the number of people is still rising, the fertility rate—the number of babies a woman can expect to have in her lifetime—has been plummeting. And not just in the rich world: two-thirds of people now live in countries where it is below the “replacement rate” of 2.1—the standard estimate of what is needed to maintain a stable population. Bogotá, Colombia, now has a lower fertility rate (0.91) than Tokyo (0.99).

The global population will peak at 10.3bn in 2084, says the UN’s central estimate. But as we report this week, its assumptions are questionable·. It assumes a sudden change in momentum, starting now: that fertility rates in many low-fertility countries will stop falling or rebound, and that plunging rates in high-fertility countries will fall more slowly. If it is wrong, peak human is much closer. If current trends continue for just ten more years before the UN’s more optimistic assumptions kick in,the global population peaks at 9.6bn in 2065, then tumbles to 8.9bn by 2100. Even that may be too optimistic.

Regardless of when the peak arrives, sub-replacement fertility implies that the global population will shrink slowly at first—and then dramatically, in a mirror image of the exponential growth that made it soar from 1bn in 1800 to 8bn today. Such a prospect alarms many.

One type of fear is broad and economic. Fewer people means fewer brains, so the pace of innovation would slow. It means less scope for specialisation and division of labour. (If only 1,000 people live in your city, good luck finding Ethiopian food or a club for your niche hobby.) Rapid shrinkage could be hugely disruptive. Heavy public debts would suddenly fall on fewer shoulders, many of them ageing. Megacities might be fine, but small towns could hollow out as the last school closes.

Another kind of worry is narrower and nationalistic. Fertility rates vary a lot between countries and groups. So some people fear a future with too few people like themselves and too many they see as culturally alien or threatening. That is one reason why populists all over the West favour bribing families to have more children, and Donald Trump has promised to be the “fertilisation president”.

Demographic forecasts are an odd mix of the certain (all the people who will be 50 in 2070 have already been born) and the unknowable (how many nippers will today’s 20-year-olds choose to have?). On a long time-scale, exponential shrinkage looks astonishingly fast. However, during the initial phase, which is when societies must grapple with the problem, the speed of change ought to be manageable.

There are several reasons to doubt the doomsayers. Artificial intelligence may be hyped·, but it is plainly advancing faster than populations are likely to shrink. So it, or another as-yet-unknown technology, will surely ease the drag on innovation from dwindling numbers of human boffins.

Another cause for optimism is that healthy human lifespans keep stretching, allowing people to stay productive for longer. In a 41-country sample, a 70-year-old in 2022 had the same cognitive abilities as a 53-year-old had in 2000. Perhaps such progress will end. But as long as it continues, it will slow the shrinkage of labour forces, giving societies crucial extra decades to adapt. Countries that waste human capital may find ways to waste less of it, by feeding and educating young minds better, and removing barriers to women working. In sum, a declining population need not mean a poorer one·Japan has been shrinking for nearly two decades, yet living standards have risen markedly.

The nationalists are right that the world’s make-up will change. Even the UN’s projection has China’s population collapsing by more than half by 2100. India will hold steady longer. Europe and America may postpone shrinkage via immigration—or they may choose not to. The future will be more African than the present, but there, too, fertility is plunging. Big, gradual geopolitical and cultural shifts are normal. The world has coped with them in the past, and can surely cope again.

Pro-natalists hope to counter these tectonic trends by using public money to boost birth rates at home. They will fail. Governments have a role in making life easier for families, but trying to pay people to have more children than they otherwise would is either staggeringly expensive or does not work. Even Hungary, which spends a colossal 6% of GDPon pro-natal policies, still has sub-replacement fertility, and some studies suggest that its bloated baby bonuses have mostly affected the timing of births, not the total.

Go forth and divide.

Shrinking, and thus ageing, populations will eventually require big economic and social adjustments. The very old will need caring for (even if they are no costlier than the young, who often spend two decades needing support). The old are more likely to vote, so their views will shape politics. That could make it harder to raise pension ages in line with life expectancy, but sooner or later governments will have to.

Adapting to an emptier planet will not be easy, but it will be doable. None of the predictions of demographic disaster seems plausible this century, and 2100 is so far away that forecasts beyond it seem pointless. Who knows? By then parents may have technology that makes child-rearing less exhausting, and families may expand again. But that is mere speculation. For now, there is reason to pay attention but not to panic.

来源：左右图史