China's Vanishing Jobs Target, Flexible Work, and Gen Z's Move to Smaller Cities

Alice Han

Welcome to China Decode. I'm Alice Han.

James Kynge

And I'm James Kynge.

Alice Han

In today's episode of China Decode, we're discussing economic indicators, from unemployment to inflation, raising flags; Tencent's move to buy Manus back from Meta; and why Gen Zed is moving to smaller cities across China looking for a different lifestyle. That's all coming up. But first, let's do a quick check-in with how the Chinese markets are starting the week.

Markets were down, with the Shanghai Composite falling around 2.06% and the Shenzhen Component dropping 3.48%, partially attributed to risk-off sentiment surrounding escalating US-Iran tensions. The Hang Seng Index was up slightly, at 0.16%.

All right, let's get right into it.

Alice Han

We'll start with the broader economic situation in China right now, as it's facing a bit of an uphill battle. China's Ministry of Human Resources and Social Security released its 15th five-year plan just a few days ago, and for the first time in decades it didn't set a numerical target for urban job creation over the next five years. This marks the first time since at least the 1990s that a headline number for new urban jobs has been left out of the country's medium-term economic plan. It's being blamed in part on the uncertainty created by the AI boom in China.

The official target for new urban jobs over the past five years was just over 55 million — slightly higher than the target set in the late 1990s, at about 40 million. That's fitting, given this is a country really driven by targets: economic, social, and political. So the fact that we've had this change to the way they think about the urban employment market — basically getting rid of this target — suggests a lot of what we've discussed previously: uncertainty about the job market, uncertainty about demographics, uncertainty about how AI could displace a lot of jobs, especially gig economy jobs but increasingly white-collar jobs too, as we've seen elsewhere in the West.

I think in general we're starting to see public consternation globally about the implications of AI for the labor force, and the way it manifests in China is quite interesting to me. On the one hand, the government is trying to proactively set up guardrails, whether through court rulings or through policy statements about how companies should not be displacing jobs on the grounds of AI disruption. But on the flip side, you still see, in general, a lot more enthusiasm for AI among the general public compared to, say, America and Europe. How do you see this decision to take away the target? Do you agree with my intuition that it has a lot to do with uncertainty about how AI is going to pan out for the labor market?

James Kynge

Absolutely agree, Alice, and I think it's a really important topic — I'm glad we're focusing on this one today. I'd say that if there's one area in which the wheels might come off China's extraordinary economic advances over the past few decades, it has to be employment. What you've already explained — the spreading adoption of artificial intelligence over the past year, and the older adoption of factory automation over probably the last decade or fifteen years — makes China a guinea pig for the rest of the world to watch. I don't say that in any pejorative way about China. What I mean is that China is moving ahead so quickly on all of these fronts that it's now become an experiment for what happens to employment and social stability when you have hyper-rapid AI adoption and automation throughout the economy.

I witnessed some of this myself when I was last in China a few months ago. I took a train journey from Shenzhen, in the south, to Beijing — 24 hours on the train. I got to chat to a lot of fellow passengers, and I have to say I think every single one of them voiced concern about AI taking away people's jobs. It came over to me in a really concrete way.

Reading around the subject, a couple of examples stood out. The first is embodied AI — physical AI, AI used to power machines and robots. I'd point to the Xiaomi car factory in Beijing, which is 100% automated on its production lines. Not a single person is on those production lines — there are people in the factory doing things like quality control, checking the machines aren't breaking down, but no human being is actually making the cars. Those lines produce a new car every 76 seconds, with 700 robots employed. This has to be one of the most advanced factories in the world.

Another example is a robot expo I went to in southern Beijing. When I got there I found robots able to do just about everything — scores of them. Robot surgeons, robot shop assistants, robot waiters, robot basketball players, robot bartenders, robot encyclopedias you could ask any question under the sun, robot vacuum cleaners, and on and on.

So the Chinese job market is being pinched from both sides: AI taking away white-collar jobs, as you mentioned, and embodied AI — robots — taking away blue-collar jobs, on the back of about a decade of the most intensive factory automation drive we've seen anywhere in the world. All of this leads to one of the most stunning statistics I've read about China in a long time: at the moment, the number of people in flexible employment — another word for the gig economy — is about 320 million workers this year, up from about 280 million last year. That accounts for about 44% of China's workforce, which to me is an incredible number.

This does seem to be a really big issue. How do you perceive all of this? How do you put it in China's context?

Alice Han

I agree with everything you've said on the labor market side. Another dimension we haven't touched yet is demography — the uncertainty about how quickly not just the labor force but the population itself will shrink. If you think about that level of uncertainty, it becomes clear the government is also unsure how many new jobs it can target in the next five-year plan, because it remains to be seen how quickly the population is shrinking. China now has one of the lowest replacement rates in the world, at around one — only South Korea is slightly lower, at around 0.75 to 0.8. That could move quite quickly and throw a lot of uncertainty into the projection models for job growth.

So if I think of it as these twin problems — the demographic challenge and the AI displacement challenge — it makes sense the government has decided not to issue targets, and to see how this actually plays out over time. But it's very clear, from all the anecdotes and the stats you've just raised, James, that China is in a way more exposed than some of the countries we're familiar with in the West, because it has a huge portion of its labor market in this flexible, gig employment situation. You said 44% — that's much larger than Europe, the UK, and the US.

On the flip side, the argument I've heard is that household debt levels are pretty low in China, and the cost of living is much lower — even rental prices have been cut substantially because of the property bust that started in 2021. So there's an argument that if the government can cushion the blow to the labor market through more support for the social security system, or some kind of universal basic income to support everyday citizens, then maybe it can weather the storm. But I think the biggest takeaway for me, in this decision to cut targets — remember, this is a political system that loves targets — is that Beijing simply doesn't know what the future pathway is for jobs, for population growth, for fertility.

James Kynge

Right, absolutely, Alice. I think Beijing is hurtling into an unknown and untested future here. Far be it from me to be overly pessimistic about the abilities of China's economic planners — we've got at least four decades of the Chinese economic miracle burnishing their reputation, and they've been proved right again and again, even when Western commentary has pointed out various problems. But on this topic, honestly, I am concerned that China may be hurtling ahead too fast into an uncertain future. 320 million people in the gig economy sounds to me like a very lopsided, very imbalanced economy that could engender social stability issues.

Alice Han

That number you just cited is just a little bit under the total US population, so people get a sense of the scale — 320 million compared to 350 million Americans. That's a huge amount of people.

But James, one question I have for you, going back to China in the 1990s: when China was going through its own kind of technological and economic cycle, trying to get rid of overcapacity and overbuild in the state-owned enterprise sector — you'll remember the era of "grasping the large and letting go of the small." Hundreds of millions of SOE workers in the industrial sectors lost their jobs, and the government introduced the "iron rice bowl" — subsidies for employment, housing, and other benefits. Don't you think that if something in the same vein were to happen in this AI revolution, the government in China would be a lot more responsive than the US or Europe, just as it has been historically, in the 1990s?

James Kynge

Well, actually, I think it's a really good analogy you've drawn. At the time of smashing the iron rice bowl and letting go of tens of millions of state-owned enterprise workers, around the end of the 1990s and early 2000s, that was a huge social experiment as well as an economic reform program. A lot of people in China were unsure whether it would work, and of course it worked tremendously well in terms of building up the efficiency of Chinese industry. But it was very bruising for the tens of millions of SOE workers who were let go, or whose factories were sold off — in some cases, sold for a single token payment in renminbi, put on the chopping block.

So what we may be going into now is, in some senses, an equivalent of that — potentially a bruising period of economic reform in which China pushes automation and AI at the expense of workers' interests. From that perspective I'd draw that analogy. We just don't know how it's going to pan out. As I said, around the turn of the millennium a lot of commentators were saying this wasn't going to work, that China's social stability would be impinged upon. With hindsight, it obviously did work, certainly from an economic efficiency perspective. So I don't want to say this is going to bring the Chinese economy down. But I certainly think a lot of gig economy workers in China now are leading very tough lives.

Alice Han

Yeah, and I think you're right that the scale is slightly different. The 1990s layoffs were around 35 million SOE workers, who had their benefits, pensions, and housing subsidies — the iron rice bowl — smashed, as you say. But then, fast forward a couple of years, China joins the WTO in 2001, and you see a big boost in China as a manufacturer to the rest of the world. You also see a big boost in the property sector, which becomes a problematic bubble for the next two decades.

The big question is whether AI will be enough of a productivity boost, the way the property sector and manufacturing were in the early 2000s, to help transition this workforce into a new growth trajectory, into new jobs. That's the trillion-dollar question we don't yet know. But it's helpful to think about the '90s in the sense that we've been through a similar period of labor disruption and dislocation before. In that instance, there was strong leadership in Beijing that said the SOEs and their employees had to carry the burden of the cost. But they were also lucky in terms of timing, because then there was the WTO accession and the property sector boom.

James Kynge

At the time I was a reporter in China. I used to go to a lot of those factories where all the workers had been laid off and the buildings sold for a few renminbi, and there were demonstrations in a lot of places in northern China — these demonstrations would go on for weeks, sometimes months. It was a period of social instability. Some of the stories you heard from these state-owned enterprise workers, who'd spent their whole lives in the iron rice bowl, working for a pittance because the tradeoff was that their healthcare and pension would be taken care of, and then suddenly their iron rice bowl was smashed and they had no way to support their family — some of those stories were heart-rending in the extreme.

I think the idea the outside world sometimes has of China as a sort of managed communist state has always been vastly wrong, because a lot of people in China have had to sacrifice a great deal and live very difficult lives to build what we see now. China has, in many cases, not been communist in the sense of taking care of its workers at all — it's been much less communist than the societies we've built up in the West. And right now, with 320 million gig workers, I think that's just the latest example of the real nature of the Chinese economy: how hard-scrabble it is, how difficult it is for a lot of people to get by.

Alice Han

We'll be back with more after a quick break. Stay with us.

Alice Han

Welcome back. Looking more closely at one particular AI company in China now: Tencent Holdings is in discussions to become the largest shareholder in Manus, a Chinese-founded agentic AI startup that builds general-purpose AI agents. You'll remember that in December of last year, Meta announced plans to acquire Manus for roughly $2 billion to bolster its agentic AI efforts. But Beijing rejected this on regulatory and national security grounds and ordered the deal unwound. Tencent was already among Manus's early-round investors.

James, this is another story that shows how tech companies are increasingly victim to geopolitical conflict and the nationalization of technology issues. In this case, Beijing got ahead of Washington in trying to block this — remember, Washington was also uneasy with the idea of a US company spending $2 billion on effectively a Chinese AI company. But Beijing went ahead and said this isn't in line with its regulations on purchasing Chinese technology IP, and has effectively strong-armed the Manus founders — the two founders are now stuck in mainland China. This is supposed to be a shot across the bow for any companies looking to relocate to Singapore and be treated as non-Chinese.

But the bigger story that's just emerged is the Tencent story. Tencent seems really keen to get into agentic AI — thus far it hasn't been a leader in the space, hence the desire to increase its shareholding in Manus. Manus reached a recurring annual revenue of $500 million earlier this year; for a sense of scale, Tencent's revenue last year was over a hundred billion. So Manus is still very early-stage, but there are interesting use cases, with Tencent pushing to put the embedded AI engine inside the WeChat app, which serves about 1.4 billion people. That could be a huge upgrade to WeChat, which hasn't really changed that much in my use of it over the last couple of years.

Do you think that, moving forward, we're just going to see what we've discussed previously — this iron curtain when it comes to foreign stakes in Chinese companies, where Beijing and Washington both draw the line and don't allow foreign ownership or stakeholding in their homegrown AI talent?

James Kynge

Yeah, I'm glad you asked that, Alice — that's exactly the conclusion I was coming to as well. I think this example really shows that the idea the US is the leading party in the tech war with China, putting restrictions on entity lists and semiconductor sales, is not quite right — it might not be right at all. Here we've got a clear example of the Chinese regulator jumping in before the US regulator, to prevent one of America's biggest companies getting hold of this very smart tech, in this case Manus's agentic AI.

I think that's absolutely the right lens. On the iron curtain idea — I wouldn't be in the least bit surprised if this is just the first of many examples of China trying to hold tight to its crucial AI technology and prevent foreign actors from getting control of it. And the other thing worth highlighting is that Tencent is China's biggest company by market capitalization — about $533 billion, which by my research makes it the biggest Chinese listed company, listed in Hong Kong. And WeChat, as you said, is used by 1.4 billion people — basically you can live your whole life through WeChat if you live in China. So the idea that Manus will be used to develop an agentic AI tool for use within WeChat makes this doubly important from a tech perspective. If Manus's agentic AI is put to the service of China's biggest app, that has obvious implications for WeChat going forward.

One other thing: the co-founders of Manus were summoned to Beijing for a meeting with the National Development and Reform Commission — China's big ministry that controls most major decisions relating to the Chinese economy. They were questioned about potential violations of the foreign direct investment rules tied to Manus's onshore Chinese entity. So we get a version here of how China can exercise its regulatory authority through the people involved — in this case, the co-founders are summoned back to Beijing, questioned, and a lot of pressure is put on them. There seems to be an investigation into the foreign direct investment disclosure requirements of Chinese law related to this Manus deal with Meta. It seems to me the Chinese authorities are using Chinese law, applied to individuals, to frustrate the flow of Chinese technology into foreign hands.

Alice Han

Yeah — and maybe people won't realize this, but I think it's pretty significant that it's the NDRC doing the disciplining. I'd have expected, as happened with Didi a couple of years ago and some of the tech companies trying to list overseas without Beijing's approval, that it would be the Cyberspace Administration of China doing a lot of the disciplining and talking. But the fact that it's the NDRC, which sits under the State Council — a really important body when it comes to the administration of Chinese industrial policy and economic policymaking at a high level — says to me that this is the central government trying to get a bigger say over technology issues, over these AI issues. Am I reading that correctly? I was quite surprised it was the NDRC doing it.

James Kynge

Yes, absolutely. I also noted the language was quite colorful — Chinese officials reportedly described the transaction, Meta's acquisition of Manus, as a "conspiratorial attempt to hollow out China's technology base." That's very revealing to me — it shows the level of emotion involved. China is just as determined to block the US's acquisition of crucial Chinese technology as the US is to block China's acquisition of US technology.

Then I guess the question becomes: because Manus was trying very hard to do the ByteDance model, moving to Singapore, not being seen as a Chinese company — will this be a blowback to Manus's foreign expansion goals? It's clear they were trying to pitch themselves as a global company by moving to Singapore, seeing themselves as a potential leader in agentic AI. And this is a powerful space in itself — if you have more and more consumers and corporates using AI agents for different parts of their daily lives and jobs, that's a very compelling use case that can scale quickly. Do you think this sets them back from globalizing the way ByteDance has successfully, but Tencent hasn't?

Alice Han

Maybe the subtext here is that Manus is being called home, to be put into the service of Tencent, WeChat, and other domestic applications, and then it's going to go global from China rather than from Singapore. I just don't know, but I think it's a very interesting point. If we look at the divergence between Tencent and ByteDance — you just said Tencent's the biggest Chinese tech company listed by market cap, but ByteDance, which hasn't IPO'd yet, would probably exceed that if it ever did, just looking at its revenue streams, its growth, its user scale. And yet if you compare the two, Tencent has really just stayed domestic, or grown with the Chinese diaspora of users using WeChat outside China, whereas ByteDance has done exceptionally well with TikTok, and also with some of its enterprise AI software, used throughout Southeast Asia outside China.

I raise this because I think it's quite important in terms of setting a precedent, but also setting the tone for future Manuses, future Chinese AI companies — whether or not they can truly be global, the way ByteDance has succeeded in doing. China's pivoted very much from the model of traditional incumbents — Baidu, Alibaba, Tencent, JD — to these new AI startups we've talked about, like Moonshot, Zhipu, Manus. It remains to be seen — I think this is a big question for China's AI revolution — whether they can become global and shake off the China moniker or association, the way ByteDance has done successfully.

James Kynge

Those are really good points — I think absolutely important for the future of this whole Chinese tech sector.

Alice Han

Let's take one last quick break. Stay with us.

Alice Han

Welcome back. Finally today, a look at what some of this economic news might be doing to Gen Zed in China. James, we've heard of "lying flat," and we've heard the youth unemployment numbers. We're now seeing more movement from Gen Zed people in China who are shying away from the hustle and bustle of major cities — tier-three and tier-four cities in the country are becoming increasingly popular among Chinese youth.

I was first alerted to this by a Financial Times piece by your colleagues, or former colleagues, that I thought was really interesting — it framed it as ghost cities being revived by youth in China, Gen Zed opting to shun tier-one and tier-two cities and go to these lower-tier cities, where rent and the cost of living are noticeably cheaper. In a way, it solves some of the problem a lot of people associated with China — all these ghost cities, all this overbuilding in the property sector. One statistic from Goldman was that the shadow supply — the overbuild of apartments — is apparently around 100 million units, which is a considerable amount.

The fact that you have graduate students saying the cost of living and job market are too high and competitive in Shanghai and Beijing, and that they're going to move out further afield, into cities like Nanjing or Suzhou, is a trend that's documented and happening, at least from what I see, including on Chinese social media. It also changes, I think, the shape of cities in China, and the lifestyle people expect. If we look at recent history in China, people were all struggling to get into the best schools, to get into tier-one cities, fighting to get household registration. Now there seems to be a reverse trend, where people are saying it's not too bad to go back to their hometown, or to a lower-tier city, where the cost of living — rent, food — is three, four, five times cheaper. Maybe that's the future of China, and it weaves rather nicely into the first segment, which was really about people's fears of job displacement and rising youth unemployment — maybe the flip side of that is that this will allow for the revival of the overbuilt ghost cities.

James Kynge

Yes, absolutely — I think that's the positive impact of this. But it comes from a very difficult place for China. I think this gives us a glimpse into the flip side, the dark side, of the Chinese economic miracle. It very much reminds me of a previous period in Japan — Japan in the 1970s. I lived there in the late 1980s for a while, and there was a famous book called "Japan in the Passing Lane," all about the relentless, grueling reality of Japanese salarymen's lives in the 1970s — the forced overtime, the physical exhaustion, even death from overwork. I think China's got a version of that going on right now. This is China in the passing lane.

What we're seeing is the collateral damage of all the effort individual people are putting in to spur China's emergence as an economic superpower being so considerable, and Gen Zed just deciding, you know what, I'm going to escape from this, I'm jumping off the treadmill, I'm going to swap my big-city existence for life in a tier-three or tier-four city. We have similar trends in the US and the UK — in the US they call it opting for a "soft life," to minimize stress and avoid burnout; in the UK we have "downshifting," leaving a high-pressure job for a less demanding, more balanced life. What happens in China, I think, is much more extreme. First, the stress of the 996 existence — working from nine in the morning to nine in the evening, six days a week — is much more extreme than what we see in the US or Europe. And second, the escape is more extreme — it's more of a social statement.

Moms and dads in China are not happy, generally speaking, if their kids say they're going to leave their tier-one city existence for a tier-three or tier-four city. Xi Jinping, we know, is on record as not supporting this trend — he's been exhorting Chinese people to "eat bitterness," as the Chinese phrase has it, meaning to undergo hardship in order to get along and build the country. So from those perspectives, I think this is a considerable trend. The issue, as you've described, is that once somebody's moved from a tier-one or tier-two city to a tier-three or tier-four city, they can probably afford somewhere better to live, but they're downshifting their earning potential, so they have to tighten their belt, live a life of what's sometimes called minimalist consumption. They can't afford to buy luxury goods, but they feel their life is freer, more carefree, less pressured — probably a lot saner than some of their former colleagues who remain in the tier-one cities.

This is obviously a social issue, but it's become a political issue in China too. In recent weeks, the Ministry of State Security has accused overseas groups of funding influencers to constrain China's development by "brainwashing" people into lying flat. I don't think there's any evidence for this, but that's what the intelligence agency is saying, and there's been a flood of social media comments in China along the lines of, "oh, so my exhaustion from overwork — that was the CIA all along." So it's become a bit of a political issue, tinged with politics, though obviously it's a big social issue really. Do you know people in China who are following this so-called low-desire life, moving to a third- or fourth-tier city?

Alice Han

I do sense, from what you've talked about with soft living and lying flat in different contexts, that Gen Zed is a bit more indexed for quality of life than for working non-stop — that's been a generational shift I've noticed. It's also viral on social media, and we haven't discussed it yet, but there's a trend in China called Dreamcore, reflecting back on the nostalgia of the 2000s, when a lot of things were better, more innocent — internet cafes, KFC coupons, QQ Messenger, the 2008 Beijing Olympics, a sense there was still hope and optimism, innocence in China's economy during that period.

But what you just said about the government being annoyed with laziness, seeing it in political terms, seems to be at cross purposes with another government initiative — trying to get young people, through loans and tax incentives, to start businesses in rural parts of China, in some of these lower-tier cities, to boost their growth. So I feel like we're getting conflicting messages from the government. On one side, I do sense the government likes the fact that more people are moving into lower-tier cities, because it increases productivity in other parts of China, relieves burdens on the higher-tier cities, and enables a more even distribution of productivity and benefits across the country, instead of it all being pulled into tier one. But at the same time, the government still wants people to be high achievers, pushing at the frontier of technology, business, AI. So I sense we're getting conflicting messages, which I think are part of China's transition.

My biggest takeaway from being in China over the last couple of years is that, in general, people — not just young people — are shifting away from the 996 mentality. This obviously doesn't include people working in AI companies, at the frontier of technology. But in general, your everyday person is more interested in traditional Chinese medicine, in quality of life, in going out into nature, in wellness. That's a shift that's been fast-occurring in the last couple of years.

James Kynge

Absolutely — it's remarkable how many graduates are opting for the lower-city type of life. According to statistics, about 20% of graduates relocate to lower-tier cities for employment after they graduate, primarily due to the high living costs we've discussed in tier-one cities. This is not a small trend at all — it's becoming part of the mainstream.

I've known Chinese people who did this with their lives more than ten years ago, and I think they've really enjoyed the lives they've had in lower-tier cities. Or perhaps they're people who've kept a kind of minimalist consumer existence in the tier-one cities, but spent a lot of their time in lower-tier places pursuing a more spiritual life — including tier-four or even tier-five cities, in beautiful parts of China, running a coffee shop or something like that. So I think this is a fairly well-established social trend in China — the idea that you don't have to be in the rat race to live your life, you don't have to dedicate your life to the accumulation of monetary capital, you can go for intellectual capital, or social capital, or spiritual capital, different types of capital, and seek fulfillment through that. Some of these places in China — if you go to Dali, or some of the other places in Sichuan and Yunnan provinces, near the border with Tibet — you can find a lot of people living their best life down there.

Alice Han

Yeah, I particularly like Dali, if anyone's been — it's a hippie kind of capital in China. They love their cafes, they have their own type of coffee, they love wearing tie-dye — there's a particular blue tie-dye that's very popular down there. If anyone hasn't been, I highly recommend it. But that's a great example of people opting out of the rat race, to your point, James, and deciding to start up coffee shops and live in places like that.

I was running some numbers on the Yunnan province area — home prices there are apparently now about three times cheaper than Shanghai, and in some parts of China you can get away with rent as low as $200 a month. At that point, how much work do you actually need to do? There are stories of people who do some part-time work online and live in a cheaper province or city in China, enjoying their lifestyle as a result. It's a shift away from what I'd call an Americanized "live to work" mindset, to a more work-to-live European mindset in some parts of China.

All right, James, you know what time it is — it's prediction time. As you peer into the future this week, what do you see?

James Kynge

I'm going to predict that about half of China's workers next year will be gig economy workers — that will be about 362 million workers, about 50% of all China's workers. I'm picking this extraordinary number to dramatize the point that Chinese society is really undergoing massive change at the moment, and I think this is a rather undercovered story. This could have big impacts on what China is, and where China goes, for many years to come.

Alice Han

Fascinating. Mine is more in the realm of AI. We just saw some data suggesting that, since the beginning of the year, Chinese models — the Chinese open-source models — have really outpaced American models in terms of token usage, by some metrics now six times more used. My prediction is that as we continue to scale AI use through tokens, Chinese models will continue to outpace the American ones, because they can leverage what we've talked about previously: cheaper costs, being quite easy to use, host, and fine-tune. On the back end, people are increasingly using them to minimize costs and do more AI integration. I think it's really interesting that Zhipu is now, in some respects, almost as competitive as Claude for coding. So I'm starting to get more bullish on Chinese open-source models, and I think the use cases for corporates and consumers are becoming more and more compelling.

That's all for this episode. Thank you for listening to China Decode. This is a production of ProfG Media. Make sure to follow us wherever you get your podcasts, so you don't miss an episode. Talk to you again next week.