Autocracy and Stagnation: How Imperial Exams Shaped China's Destiny | Hacker News Reader

Autocracy and Stagnation: How Imperial Exams Shaped China's Destiny

chinatalk.media

88 pointspseudolus84 comments

Contrary to popular belief, the original Keju (imperial exam) wasn't designed to select talented folks from commoners, but rather a royal ranking system to reduce the voilent political fights between noble houses during the Great DEI era of Wei-Jin-Northern-Southern states, when lots of deadly racial conflicts happenning in east asia for centuries.

The's a saying "寒门出贵子", big names rise from small houses, yet it requires a house to rise from.

Which Huang also points out in his book "Rise of the E.A.S.T", which this article is directly referencing.

The Keju system helped minimize the risk of military intervention in politics along with the internal fights amongst power clans and families, but at the expense of innovation due to ossification.

Highly recommend reading anything by Yasheng Huang or Yuhua Wang about the development of Chinese administrative capacity.

Yeah, the examination system may have been an elegant way for emperors to prevent consolidation of power among the gentry.

But it really seems a stretch in the interview say the exam system provided political mobility to farmers. The exams required an absurd amount of dedicated study.

That someone like Fan Zhongyan could overcome his privileged family falling into difficult circumstances and have the opportunity to rise to heights as the result of success in exams shows a certain amount of meritocratic mobility within elites, but that's not all the same thing as the exams being about recruiting talent from commoners.

  > The exams required an absurd amount of dedicated study.

This is a problem in general with trying to achieve meritocracy. These merit based tests only work when all things are equal. Like you're suggesting, a noble has time to dedicate to studying while a farmer must work much harder since they cannot abandon their means of putting food on the table.

I don't think this means we shouldn't try to make our societies more meritocratic, but I think a lot of people miss key points like this. The assumptions underlying the idea. Especially with survivor bias and that there will always be examples of those who overcame their disadvantages. But it's critical to understand these if we're actually to progress towards are more meritocratic society than we are now.

[Side note] I do often wonder why we equate recognizing an advantage as diminishing the hard work one needs even when advantaged. Especially since our stories are the hero's journey and a critical retirement from that is meeting a wizard to get a magical power. Be it in The Matrix Neo's natural talent, Luke Skywalker being born a powerful Jedi, or King Author's sword and help from Merlin, we don't see these as tainting the struggles they overcome or diminishing their efforts. But we do when it's stories about ourselves.

> You can’t expect to wield supreme executive power just because some watery tart threw a sword at you. —D

> This is a problem in general with trying to achieve meritocracy. These merit based tests only work when all things are equal.

This assumes that the goal is egalitarianism, and filtering by "merit" is merely a means to achieve that goal.

If instead the goal is to get your tasks done as effectively as possible, it doesn't really matter whether someone's effective because of inherent ability vs education quality. People not reaching their full potential due to lack of resources does reduce how effectively your tasks get done, but is a separate issue from how to best select someone from the people who are actually available.

If instead the goal is to displace problematic things like factionalism or nepotism or corruption as the mechanism for filling official positions, it's only a problem to the extent that one group (along a dimension you actually care about) can keep resources away from all other competing groups and to the extent that any such monopolized resources are actually necessary.

  > This assumes that the goal is egalitarianism, and filtering by "merit" is merely a means to achieve that goal.

This is a deceptively false assumption. I think it makes sense that anyone would make this assumption, because it is correct via low order approximation.

The issue here is that there's an implicit assumption that this is a zero sum game that is also locked in time. The assumption being made is that by spending resources to seek out these others (let's say "farmers") that can contribute that you have to give up places for "nobles" or that this effort comes at a cost of time to progress. But neither of these are true! There is no need to have a finite and fixed set of positions (unless you have a strong belief that an increased number of workers decreases efficiency, due to some other process coupled with this one or through some self-feedback) nor does this mean we can't perform merit testing by current standards while simultaneously seeking to improve those standards. Those can be done in parallel! In fact, if you are increasing the number of workers you have, then you clearly have no problem of diminishing efforts. The question is then more about how these workers' efforts are best spent and we can't just look at the first order approximation because if we're defining a task to look at increasing our workers then we have a rate of change factor and so need to analyze at what point the investment pays off.

For a more concrete example, let's use scientific progress. We'll take the reasonable assumptions that the difficulty of scientific progress increases over time (we have strong evidence of this as topics become significantly more nuanced as you dig deeper into them), as well as the reasonable assumption that one's educational level directly correlates to the likelihood of them contributing to progress. We'll also take the naive assumption that progress is random, in a Monte Carlo process; i.e. scientists stumble into progress (presumably in reality scientists are better than random, but there is clearly some randomness to the process).

With this system, scientific progress advances faster by the number of scientists we have (very much like we've observed over history and even in specific fields and niches!). In such a system, it warrants allocating some "resources" to increasing the efficiency of the training and selection process.

In other words: do you think the world of science would advance faster if there was 1 Einstein or 100? If you believe the latter, how many Einsteins would you be willing to allocate to achieve a total of 1000? Inherently, these are coupled, not separate issues; which is what my post was about.

I think it's important that there is a big difference between a farmer as in someone who is out in the field doing laborer, and a farmer as in a large landowner who has a lot of other people out in the fields doing labor which historically in China, and also Europe and the US, was considered to be a kind of ideal person in many time periods. So I doubt anyone who was out doing labor in the fields could do great in the exam, but if you were a small local elite, this provided a path for your children to potentially rise up.

There was also an official exam for becoming a military officer. One part of it was manipulating an extremely heavy guan dao (which is likely where the stories of Guan Yu wielding a pole arm between 40 and 90 lbs come from).

Training for the test is still around in the cultural practices of some people groups in China:

https://www.youtube.com/watch?v=XqnulYO5890

I have seen some pictures of PLA soldiers aiming their rifles with weights hanging off the end of the barrel. I wonder if this is a distance relative of that old practice.

> aiming their rifles with weights hanging off the end of the barrel

That sounds like it has a practical explanation: Strengthening certain muscle-groups so that the soldiers won't easily tire of handling their weapon.

It might be even better if the rifle had a laser on it, so that the exercise involves not just lifting the barrel but also centering it onto a target. (However shooting live-rounds could metaphorically backfire, since users would be getting trained to expect the wrong recoil.)

And the manchurian qing dynasty had a test where you had to be able to draw a 300 plus pound composite recurve bow! Which folks were still doing in the 20th century!

Edit I was mistaken: https://www.mandarinmansion.com/glossary/wuke-gong-wukegong

These military exam bows went up to 160 lbs.

In at least one instance a 240 lbs bow was manufactured which a single elite archer was able to pull

Interestingly enough, standardized testing was unknown to Europe until they made contact with China.

One company that was inspired by this was the British East India company which introduced an exam for employees. This contributed to the company becoming the largest most powerful corporation in the world! Down to running private militaries and overthrowing kingdoms. At one point the company's private armies were 3x larger than the British military.

As for China itself, it kept doing these exams, but the content never changed. Straight into the 20th century the scholar exam involved memorizing Confucian classics. There was even a military exam that wasn't killed until the first decade of the 20th century. It involved being able to hit multiple targets with a bow and arrow whilst riding a horse, and being able to draw a 300 lb bow without straining your shoulders.

So for all ya'll software companies... one day it might be prudent to update these leetcode problems...

With quizzes about tensors, Einstein summation, and recent scientific papers

I think there must be a conversion issue there. An English Longbow topped out at like, 110 lb draw weight I don't think you could make a 300 lb bow, and if you did I don't think anyone could draw it without straining their shoulders.

Here are the historical draw weights for Qing dynasty bows: https://www.manchuarchery.org/historical-draw-weights-qing-b.... 80 lbs was considered minimum for military usage and 150 lbs was not unheard of

Imperial examination bows were giant versions of these military bows that were intended to assess the strength of the examinee. I couldn't find reliable numbers on the draw weights for these, but needless to say they had to have significantly higher draw weights than standard military bows.

Even the Qing Emperor was expected to be able to draw a 120 lb+ bow or at least had such bows in his personal collection. The Manchurians came from a culture that valued hunting and horseback archery even more than the Han Chinese.

120 lb draw weight I could see, even 150 for a test or a little more. That still leaves us 100 lb short of whats being said here so I still think the 300 lb figure is either an exaggeration or a translation/conversion issue. I did some googling and indeed I cannot find any example of a 300 lb bow being made ever, let alone before modern materials science and I am skeptical that anyone short of an Olympic athlete would be able to draw it.

From the same source above professor Mark Elliot writes:

"In regularly held contests, expertness might be demonstrated by accuracy of aim, length of shot, or strength in drawing. It was the last of these that occasioned the most detailed comments by supervising officers. Bows were graded by "strength" from three [40 pounds] to as high as eighteen [240 pounds], according to their stiffness. Ability at strength of six (probably a pull of about 80 pounds) was considered minimal for a grown man, and strength of ten (about 133 pounds) was required for participation in hunts. A 1736 report found that of 3,200 troops at the Hangzhou garrison about 2,200 were able to draw bows of strengths six to ten [80-133], and 80 could handle bow strengths of eleven to thirteen [147-173 pounds]… …In comparison, the 500 troops at the small Dezhou garrison acquitted themselves with honor, all of them being able to take a five-strength bow [67 pounds], 203 a six-strength [80 pounds], 137 a seven strength [93 pounds], and 85 a ten-strength bow [133 pounds].

And: "The champion in a 1728 contest between the one hundred top bowmen in the empire won one hundred taels when he hit the bull's-eye using an eighteen-strength bow an estimated drawing weight of almost 240 pounds!"

So in 1728, the draw weight topped out at 240 lbs which only one guy in the entire empire could draw.

I've attempted to pull an 80 lb longbow once with no training, and I'm pretty sure I risked tearing a shoulder. It takes a lifetime of training to get into military bow pull strength

It also seems that the 80-110 numbers for the English longbow may be a low estimates. Bows and skeletons were recovered from Henry VIII's flagship, which sank in 1545. Mary Rose longbows had draw weights estimated between 110 to 180 lb. Similarly, it has been estimated that to penetrate an armour from the first half of the sixteenth century, a draw weight of about 110 to 130 lb was needed.

This makes anecdotal sense to me. 70-80 lbs is a common draw weighed for modern compound hunting bows routinely used by people with comparatively little training. I shoot about 75 lbs with no more training than going to a range once a month and am a 6ft male with no weight training.

Given the historical accounts of the required training and dedication required for the English longbow, I expect 110 could be on the lower end.

Guinness world records has contemporary archers shooting longbows up to 200lbs, although without precision.

80lbs is fine for deer. Gotta go higher for armored deer

> One company that was inspired by this was the British East India company which introduced an exam for employees

I would assume they were based on the exams in the English navy introduced in 1677 (which I think was almost 200 years before EIC)?

They weren't fully standardized by modern standards but I'm not sure that they were influenced by the Chinese exams as much as common practices in European universities, guilds etc. It's not that obvious that standardized in the 1800s didn't just gradually develop from previous practices with limited external influence.

AFAIK was the first country in Europe, they had standardized for judges in the 1740s (so 100 years before the EIC) and the "Abitur" exam for school graduates was introduced in 1788. I doubt had any links with the Chinese exams. Pretty much all Western Educational systems were significantly influenced by Prussia.

Also the East India Company was founded in 1600, predating those English Navy exams by 1677. I'm not sure when they started adding standardized exams.

Edit: It's possible these EIC exams were only introduced in the 19th century after the company was taken over by the British government. Could not find any sources stating that they were introduced earlier.

They reformed the civil service system in 1853. I'm not sure it was something that the EIC came up with on its own, rather it was the British government/parliament, possibly trying to fix the arbitrariness/mismanagement/inefficiency that was so rampant in the EIC.

There might be some Chinese influence coming through the Jesuits as early as the 1600s. However the early European "exams" in the 1700s seem to have been very different e.g. basically you'd join the navy, or the civil service (in Prussia) in an entry level position, learn on the job for a few years and then take some semi-formal oral exam before by a panel of senior officials. So probably much closer to the guild style system than actual standardized exams which didn't really came until the 1800s. And of course you are already hade to come from a fairly affluent/well-connected background just to get in.

> Surprisingly, candidates from wealthy families performed worse during this round. This is counterintuitive, especially when compared to contemporary educational systems where socioeconomic background often correlates positively with educational access and performance. Our research revealed a negative correlation in imperial China.

> We attribute this to the emperor’s reluctance to appoint officials from independent power bases to central government positions. The final round selected the most important officials for the imperial court - those who would design future keju examinations, advise the government, and have direct access to the emperor. This system effectively marginalized the nobility and successful merchants.

This is a pretty brilliant mechanism: an emperor normally draws a certain amount of jealousy and hate from the nobility and successful merchants, but by promoting rustic wonks over them via this system, might the emperor have been playing "let's you and him fight": diverting their animosity onto a bunch of "academic, gate-keeping, elites"?

IIRC, there was discussion comparing these exams to the leetcode hazing rituals that comprise tech interviews today.

Spain has civil service exams and people can spend years after university preparing for them, full time, the official syallbi from the public service are tens of thousands of pages long. I don't think leetcode is anywhere near that arduous.

Is the knowledge practical in any way? Or largely irrelevant in the day to day?

Is there a modern justification for such an onerous process? I am all for well trained professionals, but at some point people have to recognize gate keeping. Years of full time study is an impossible ask for many without external support.

My friends who studied for the Pakistan Civil Service exam told me you can get questions like:

When did Gen Musharraf visit India on so-and-so trip?

A) Oct 10th 2004 B) Oct 8th 2004 ...

(I made the answers up, but I distinctly remember the days being 2 days apart)

Is cheating rampant? I do not understand how a “test” like this can work where any irrelevant piece of trivia is fair game.

The Indian Railway exam has questions like this too. Here are some examples from previous years:

https://prepp.in/rrb-recruitment-exam/practice-papers

> 1. In Computers, what does ALU stand for?

> 5. BJP had won how many seats in the Lok Sabha election 2019?

> 20. In which of the following states of India is the Moatsu festival celebrated?

> 51. Who among the following was the first cricket player to win Arjun Award?

> 62. Which is the shortest part of small intestine?

I have heard that the only way to pass is to know someone who can give you the questions in advance.

> only way to pass is to know someone who can give you the questions in advance

RRB is remarkably transparent. You can download past papers going back two decades & study them. There are multiple youtube channels that solve all of these problems & tell you the strats explicitly. Current events & GK culture is a thing in India. Lots of common folk pride themselves on knowing the answers to these questions. I wouldn't consider this gatekeeping by any means.

This is being discussed in the context of the Chinese test which required memorizing enormous numbers of facts. An example question for a train test requires knowing human anatomy to the degree of classifying individual organs.

I am failing to see how that is anything but blocking those who cannot dedicate themselves to the test.

There are more sailors than ships. In this context, something like 10000 applicants for a single rrb vacancy. You don’t want the railways to be nepotistic. Nor do you want legacy admits. Nor money power to dominate. What’s left ? Well, how about - leetcode has 6720 permutations. Find the 100th one by hand, when ranked alphabetically. One of my friends got this question last week at a bog standard Java IT job, in Bangalore, with over 1000 applicants. Its a fair question, just a little bit of math and arranging alphabets. Doesn’t mean he has to devote itself to handcomputing permutations for the entirety of his life. Its one day of interviews and you get to know if you are selected by 4pm same day. Perfectly fair imo. The standard gs interviews during my time took 12 separate rounds with 12 teams, were completely non-transparent, and you spent a whole month not knowing whether you are in or out. I would rather permute leetcode. Similarly, let the railway board applicant sweat a little. Its a government job for life with handsome pension to boot.

> Similarly, let the railway board applicant sweat a little. Its a government job for life with handsome pension to boot.

So just have a general aptitude, “IQ style”(of course bot directly) test? Or better yet base your questions on things that are actually relevant to the job?

The purpose of tests like that is to select to people who have nothing better to do but waste their time learning pointless trivia. Seems extremely suboptimal..

Or you could just set the test for the actual prerequisites, then run a lottery on those who pass to decide who to interview.

Statistically the individual's chances are the same, but you free up a not-insignificant amount of work and stress that could be spent on more productive areas.

Presumably they don't repeat the same questions so I don't really see how the past question bank is so helpful when the possible subject matter is incredibly broad. That said apparently what I've heard is wrong and people do manage to pass these just by studying.

On the larger point, trivia is popular here in the US too but I would be amazed to see sports trivia questions on a government exam. It's hard to understand what the legitimate purpose of these questions could be for someone applying to be a train conductor.

Aside: In case anyone else wondered GK is apparently General Knowledge, it took me a while to find that initialism since I wasn't familiar with it.

> trivia is popular here in the US too

When you interview at Blackrock in SF, they walk you thru a bunch of rooms. Each room has a door. On each door is the name of a famous economist. When the interview begins, you have to recollect all of these economists and then drop some tidbits about each of them. You know, to signal that you are part of the in-group. Like, you can say hey wasn’t Krugman at Princeton when he got the Nobel but doesn’t he teach at cuny now because that’s where your phd advisor saw him last week hint hint. Now, would you rather do such obsequious ass-kissing, or remember that ALU stands for arithmetic logic unit ? Compared to Blackrock, Railway Board is god, son and holy ghost rolled into one.

That's an interesting fact but a weird take away. Of course you can always find an arguably worse candidate selection process.

That seems pretty weird but at least you are expected to demonstrate that you have a genuine interest in the field outside some narrow specialization (or would talking about their actual work not count as “tidbits”?

The listed Indian train service exam questions just seem beyond absurd in comparison. If the questions were about random niche train related trivia it would make a whole lot more sense (of course still stupid..)

> I wouldn't consider this gatekeeping by any means.

Seems to be pretty absurd, though? Basing your hiring decisions on entirely pointless trivia? Looks like a good way to filter out candidates unwilling to demean themselves to such nonsense

Some of it is, a lot is pointless trivia and ephemeral legalese. But there are many candidates and they need some way to filter through them. There's also scholarshipa for studyydor the exams for some of the positions, for example judges.

I agree it's not as arduous. But I feel the concept is still similar.

In the case of leetcode, it's not limited to people out of university either. Mid career professionals are just as exposed to them as fresh grads.

I know many senior and staff level SWEs who grinded leetcode for years to land their FAANG level job (or gave up trying midway).

Great podcast and great episode.

Keiju system was a tool for the emperor, not for the ruling classes. The first two rounds were anonymized and in the highest level exam the gentry performed worse than those coming from lower ranks. The incentive for the emperor was not to allow potential competition into high offices. It was meritocratic, but it also homogenized the ideology and stifled innovation.

...yes, there is that incredible mobility. You could be a farmer, and then you took the exam, you succeeded, and then you became an official. So in that sense, it’s not social mobility, it is political mobility.

https://www.gutenberg.org/files/48612/48612-h/48612-h.htm#Pa...

> In the language of Vilfredo Pareto, this would probably be termed "capture of the rising elite"; in the language of present-day Marxists, this would be described as "utilization of potential leadership cadres from historically superseded classes"; in the language of practical politics, it means "cut in the smart boys from the opposition, so that they can't set up a racket of their own." –PMAL

A previous discussion about the exams:

https://news.ycombinator.com/item?id=33888398

I'm skeptical of some of the claims about technological innovation being higher during periods of fragmentation. Correlation doesn't equal causation - there could be other factors at play. Still, an intriguing hypothesis worth exploring further.

War is a well known factor for innovation, and it tends to go together with periods of fragmentation. Probably less innovation than a war against a rival external power.

This is one of those claims that people tend to repeat without giving any evidence. Seems like there's some sort of attempt using Needham's data.

There was a lot more innovation in Europe, particularly in military technology, in the 17th-19th centuries, due to many smaller states competing against each-other, compared to the Ottoman Empire and China. That's why Europe was able to so easily subjugate China.

> to many smaller states competing against each-other, compared to the Ottoman Empire and China

Did wars really have a positive impact on that though? I don't think the 30 years war etc. had such a positive impact, rather it probably slowed down progress by quite a bit.

It could easily be the other way around; new technology leading to new wars as nations feel empowered by their new toys to act on their ambitions. The first world war occurred shortly after the introduction of smokeless powder and machine guns. These inventions were certainly funded by some military spending and perhaps even prompted by earlier wars, but in large part they were technologies who's time had come; precision manufacturing and chemical technology had only made these things possible shortly before they were invented.

The development of "smart bombs" like the Walleye television bomb happened about as soon as they were technologically feasible; the idea for a fire-and-forget television bomb came from an engineer who was playing around with a new model of commercial TV camera. Emboldened by these new weapons, American politicians started wars which previously they might have considered too politically costly if they had to be fought using older methods. Desert Storm particularly.

The Russo-Ukrainian war saw a huge explosion of drone warfare, but I am mostly certain that it didn't happen due to military drones coming of age.

Those drones weren't developed because of the war. But the existence of those drones (among other newish technologies) might plausibly have something to do with Ukraine believing they can successfully resist.

If you follow the war closely, the development of the drones during the war is absolutely breathtaking.

Originally, Ukraine had a few Bayraktars (nowadays hopelessly obsolete) and some observation drones. They couldn't even use drones against armor. The original losses of Russian armor in 2022 were mostly due to Javelins, NLAWs and similar specialized weapons.

Nowadays, the Ukrainians deploy jet-powered Palyanitsya drones to the deep rear of the enemy and regularly destroy heavy equipment with explosive drones.

> factor for innovation

That's not so obvious, at least in fields not directly related to warfarce e.g. Byzantine Empire was probably the most innovative, progressive and developed state in Europe (if not the world by ~1000 AD) constant endless warfare turned it into an empty shells by the 1400s.

Extreme instability, violence etc. rarely has a positive effect. Relatively peaceful primary economic competition (perhaps with some limited warfare) between multiple states might lead to very positive outcomes, though.

I suspect that the motivating factor is lack of deployable human capital. In a large centralized society, there may always people to deploy at the problem. The more people you deploy at the problem, the more time the people at the top spend thinking about people issues and the bureaucracy that comes with that. The bottleneck usually happens when the people at the top become so focused on bureaucracy/politics/corruption that nothing really gets done.

Fragmented societies are less prone to bureaucracy, and more prone to not having enough people available to solve a problem. These factors absolutely help innovation.

> Fragmented societies are less prone to bureaucracy, and more prone to not having enough people available to solve a problem.

Competition probably also played a part. If the emperor in China or Ancient Rome decided to ban or restrict certain fields/groups/ideas you were pretty screwed, innovation likely stopped, most progress was lost and the next generations had to start from scratch. In Europe you could just move to the next state or city. Same applies to major invasions and societal collapses, e.g. the Byzantine scholars fleeing to Italy during/after the Turkish invasions.

Maybe too much of meta tangent: mccarthyism 2.0 = take scholarship from Chinese Americans with grain of salt. Or at least be aware of motivations. Yasheng Huang slowly transiting to "house chink" - suspect Huang's trying to avoid becoming a Gang Chen in the next wave of China Initiative purges, so not surprising some of his analysis has gotten more "motivated" over the years. The one's who don't play the game, i.e. Cheng Li from Brookings, don't get the podium anymore. There's still "libtard" audiences in PRC who eats Huang's work up, the main propaganda pitch of this book is:

> I really want to push back against this view that automatically you have a Western ideological view because you emphasize diversity, freedom, and competition, and that somehow is ideology rather than fact. I’m telling that story from China’s own history. I do hope that I win some hearts from people who have an automatic immunity to Western ideas.

...

> I aim to convey this message through Chinese history, hoping to persuade people to reconsider their aversion to diversity and competition.

...

>"Too much autocratic stability is detrimental"

Gets challenged on Song innovation. No, no, I've rescaled importance of inventions, downplaying "four great inventions" that Chinese find significant, which conveniently illustrates Song is less innovative than Han-Sui interregnum, I hope the take away for PRC is that Chinese innovation the strongest when China most fragmented and mired in chaos. Get's challenged on commerce. No, no, historic/current Chinese commerce too restricted to launch industrial revolutions, see Jack Ma. NVM PRC explicitly did not want industrial revolution in unchecked financialization. Or PRC biasing scale in lieu of scope is pushing involution tier industrial revolution in many strategic sectors, i.e. PRC in a place to pick what revolution to scale. "In contrast, Western countries resolved this tension by embracing scope", except this came at the expense of scale, and one of PRC's biggest advantage is acknolwedging and pursuing quantity having quality of it's own. Funny how blob incentives map to imperial bureaucracy, think tanks another incarnation of posturing eunuchs vying for position (or not losing it) at the end of the day.

Interesting parallel with our own modern day society - ostensibly meritocratic but preparation for the imperial exams began at the age of 3-4, the practical effect being that the well off were already advantaged. Similar to our present age where the well off can afford to allocate resources for their preschoolers that the less well off simply don't have. In both instances a class perpetuates itself.

What I've read is: Income and (nuclear) family size are inversely correlated, and one explanation for this is that as income increases, the total cost of an additional child increases in relative terms. Put another way: phenomena like the preschool rat race mean each child takes a larger portion of your income than it does at lower levels of income.

But children education is not a nuclear family issue, it’s a societal infrastructure matter.

If you have more children, you can’t give the same amount of individual love and care to each. That said they can give love and care to each others, though not with the same level of experience and self-control as (some) adults can do.

Yes love and care are tremendously important to foster mentally healthy people.

But this alone won’t teach them literacy, swimming, mastering of self attention, math, music, biology, active listening, critical thinking, humility, sane cooperation at scale, and so on. For all that you need a safe generous educative environment that could be logistically scaled to all humans but that we will fail to put in place as we let the most capricious wealth-drainers deteriorate harmony in societies.

That’s hard for me to wrap my head around. I have friends who want a second kid but worry about whether they could still pay their mortgage and buy food. No wealthy family is worrying about that. The cost of raising a kid might go up with income, but probably not at the same slope as income, which would mean they get cheaper in relative terms.

My attempt to explain would be a high income family aim to provide their kids the same opportunity, and they know with seven children they won’t be able to provide the personal attention to provide that. For example, consider how parental involvement in a child’s schooling is the number one predictor of educational outcome. Well, you can’t spend thirty minutes each night helping each of seven kids with their homework.

> That’s hard for me to wrap my head around.

You seem to be trying to model social phenomena as linear functions. That's going to fail on almost every time.

The GP is also claiming monotonic behavior. Please understand it as constrained into a "reasonable" interval that he didn't disclaim. Otherwise it certainly won't hold either.

In my observation kids with older siblings tend to develop faster. As in they walk at an earlier age. I guess they get fed up with watching the older kids running around and being left behind. Older kids also often act as deputy parents. So I don't think kids from large families surfer from attention deficit. I am a single dad and often struggle to keep my child attention right through the day.

> No wealthy family is worrying about that.

They are, but in a different way. They worry if they can still afford to pay the mortgage while also paying for a second daycare/preschool tuition and then later a second private school tuition.

> The cost of raising a kid might go up with income, but probably not at the same slope as income, which would mean they get cheaper in relative terms.

FAFSA allows colleges to perform perfect price discrimination, so the cost of raising a kid through college absolutely goes up with the same slope as income. When colleges know you have significant retirement savings and a high income, your kids will get zero financial aid.

"For if every instrument could accomplish its own work, obeying or anticipating the will of others, like the statues of Daedalus, or the tripods of Hephaestus, which, says the poet, 'of their own accord entered the assembly of the gods'; if, in like manner, the shuttle would weave and the plectrum touch the lyre without a hand to guide them, chief workmen would not want servants, nor masters slaves." (Aristotle, Politics, Book I, Part IV)

And here we are, with all these automatons, but still with a societies that prepare people for more slavery systems anyway.

we're still very far away from complete automation. one person now can do the work of 10 or 100 workmen in the past but people are still needed to transport and set up the feed stock, monitor, configure and service the machines, remove, inspect and assemble the finished products, on and on there are so many other tasks that only a general purpose agent like a human can do. you could construct an assembly line that is so completely integrated that it can almost run untended except for maintenance and recovery from failures but it will only ever produce exactly one model of one product and if you even need to change the weight of one of the parts it produces and then assembles you'll need to manually reconfigure a good deal of it creating significant down time.

it doesn't help that we impede the progress of automation by outsourcing labour overseas where the cost of labour makes manual processes still viable or by importing temporarily cheap labour until they realize they're getting a raw deal and move up the ladder with everyone else.

if we want to develop the technology needed to alleviate the burden of manual labour then we have to disallow these temporary quick fixes. if we do nothing they'll run their course in a few decades anyways, it'll just be a slow walk to the same destination giving the people who are acting in exploitative ways ample time to stuff their coffers with the fruits of their schemes while letting the rest of society rot on the vine mere steps away from the solution that would benefit everyone.

> In principle, membership in these three groups [proles, outer party, inner party] is not hereditary. The child of Inner Party parents is in theory not born into the Inner Party. Admission to either branch of the Party is by examination, taken at the age of sixteen. —EG

Sounds like Winston Smith wound up in the Outer Party due to poor A level results?

> "Similar to our present age where the well off can afford to allocate resources for their preschoolers that the less well off simply don't have."

A talking point endlessly repeated by those trying to push an agenda but those of us who climbed the ranks from poverty because the children of the well off chose not take advantage of the privilege they had and instead chose to be indolent would beg to differ. Meritocracy is alive and well, thank you very much.

This effect would be much smaller if there were a high-quality, free-at-point-of-service state education system. When you have to privately school or tutor or homeschool children to get them to learn things like reading [1], having extra resources at your disposal is a much bigger lever.

[1] https://features.apmreports.org/sold-a-story/

Its a bit tiring to read similar utopistic comments so often here. Completely unhinged in reality, utterly ignoring human nature, disregarding whole human history and so on.

No solution to anything in sight, just complains about unfairness. I know its an easy position in life, but its a lazy one.

But what about some concrete, achievable realistic steps to improve current situation even further? Or even better, what about bringing less fortunate parts of the world at least to this, apparently still miserable level? Now that would be an interesting discussion.

> Completely unhinged in reality, utterly ignoring human nature, disregarding whole human history and so on.

Unhinged from any concrete counter-critique, latching on to nebulous nonsense like “human nature”, hand-waves towards “whole human history” as if that means anything. That’s your comment.

> preparation for the imperial exams began at the age of 3-4, the practical effect being that the well off were already advantaged.

This reminds me of the recent experiments with dropping standardized tests (or making them optional) for college admissions.

The critics bemoaned the advantage that rich kids had when studying for the exams... but when colleges temporarily switched to essay writing, grades and extracurriculars as the bar for admission, the incoming class was even more heavily skewed to rich kids.

It turns out that rich kids have much less of an advantage on standardized exams than any other form of evaluation.

The standardized test, scored by an anonymous grader, still requires you to know some things and to engage your brain. Being born to rich parents (or members of the imperial court) doesn't automatically give you passing scores... You still have to work for them.

Most of the people who wanted standardized testing done away with knew exactly what was going to happen. They knew that asian immigrant families in particular did well on standardized testing and hoped to keep them down by shifting the focus of admissions to BS like sports and after school club participation. It wasn't an accident, it was the intent from the start.

> The critics bemoaned the advantage that rich kids had when studying for the exams... but when colleges temporarily switched to essay writing, grades and extracurriculars as the bar for admission, the incoming class was even more heavily skewed to rich kids.

Was there some mechanism to ensure/confirm that the scoring for those remaining factors didn't get changed during the process?

I imagine someone judging an essay or extracurricular activities would behave differently if they thought it mattered a lot versus whether they thought it was just a tiebreaker to apply after a standardized test.

How much does "allocating resources for preschoolers" actually matter? My sense is not a whole lot.

The West is fairly meritocratic. The attempt to deny this is catastrophizing e.g. "everything is awful and unfair." And in practice, critics of Western meritocracy have mainly had a negative effect by erroding the good things we do have.

Certainly primary schooling has much much much larger effects than university. I'm not sure if preschools matter though.

In a gold rush, those with the diamond shovels will go further. The only way to no perpetuate class is to not have access to the consumer class depend on a certain selection of jobs. The problem is the size of the middle class, not the selectivity of participating in the middle class.

The fundamental problem imho is that standardized tests, which start out as well-intended, start to replace actual skills as a focus of interest. This is especially problematic when the tests are only correlated 0.1-0.4 with whatever is actually of interest. The tail starts to wag the dog.

One common example of this is in policy discussions about the tests, where figures tend to only present means, with no variances or scatterplots. I remember this, for instance, from a report at Dartmouth not too long ago arguing for the use of standardized tests, where the authors only presented means, and not the huge variances around them which is really the heart of the issue.

It starts to become especially problematic when people get canalized into lifetime paths based on essentially invalid measures, especially when the targets themselves are changeable. "Meritocracy" starts to become sort of a misnomer at that point.

Of course, relying on anything too heavily starts to be an issue, and tests are a source of information. But when you start to overrely on anything, or rely on it to an extent that is empirically unwarranted, you start to have problems institutionally and societally.

> "Meritocracy" starts to become sort of a misnomer at that point.

Even if the variance is so high that it becomes a lottery, it still looks better than any alternative people have come up with.

I see it as a sort of leaky inheritance.

The Duke of Westminster is going to pass Mayfair down to his kids forever, but the CEO of Google can only do so much for his kids. In all likelihood, they will be well educated but never get the top seat at this kind of corporation.

If you couldn't pass down anything, eg if you had to put your kids in the same schools as everyone else and not give them any advantage, it would seem to be against our instincts as parents. Yet it also isn't good for society if we gave guarantees.

Duke of Westminster is worth like $12B, and Sundar Pichal is worth almost $1B which is much less but still generational wealth. The CEO of Google is kept in line by the founders who keep control with dual class stock options though.

you say:

> the practical effect being that the well off were already advantaged. ... In both instances a class perpetuates itself.

by contrast, the article says:

> You could be a farmer, and then you took the exam, you succeeded, and then you became an official. So in that sense, it’s not social mobility, it is political mobility.

> Ilari Mäkelä: But there was genuine political mobility.

> Yasheng Huang: Oh, absolutely.

> Ilari Mäkelä: How would you convince a skeptic of that?

> Yasheng Huang: Very easily. Again, going back to our earlier conversation about data, this is not something you can just use stories to illustrate. You kind of have to use statistics and data. An earlier historian, Ping-ti Ho 何炳棣 who really pioneered research on the keju showed that it was meritocratic. There are later historians who study keju, and they say, “Oh, so-and-so came from a prominent family, and he also succeeded. He got promoted.” But the problem with that is, how do you know it is the prominent family part that contributed to his mobility? Maybe he was extremely smart.

> I’m not saying that the second view is necessarily wrong, but all I’m saying is that just telling that story itself is not convincing enough to make your point. In my work with Clair Yang from the University of Washington, we ran statistical tests to determine the effect of family backgrounds on exam performance. Whether or not you came from a rich family, whether or not you came from a politically prominent family, your father was a government official, your grandfather was a government official. Statistically, if that background influenced your exam score, we should be able to see it. We should be able to detect that effect in the statistical results we generate.

> Indeed, because you had three rounds, in the first two rounds that were anonymized, we didn’t see any effect of family backgrounds on exam scores. That’s quite impressive — going back to the 6th century, 7th century, 9th century, and 13th centuries — that your backgrounds didn’t influence your exam scores.

it sounds like you disagree with dr. huang's research conclusions, but you don't seem to be offering any argument against them, just bare contradiction

Reply on news.ycombinator.com