Employment Down by 9.35 Million, Yet GDP Still Grows 5%! Output per Capita Would Have to Rise 7% — Four Times the US; Excluding Computing, Electricity Use Grows Under 3%; All Five Lie-Detection Triggers Fired; FX Settlement Rate Down to 40% Means Legal Capital Flight; Households Saved 7.58 Trillion While Loans Actually Fell! Real Growth at Most 2.5%, Possibly Zero

The host cross-referenced multidimensional physical indicators (electricity consumption, freight volume, employed population, working hours, foreign exchange settlement rate, etc.) with the official GDP data of the Chinese Communist Party (CCP), pointing out that the officially claimed GDP growth is severely falsified. Through Kalman filtering and lie-detector model analysis, the real growth rate of China's economy is between 0 and 2.5 percent, revealing the true economic backdrop of legal capital flight by enterprises, a surge in household precautionary savings, and a precipitous demographic cliff.

Video Link:

Full Text:

Observe things as they are and reject fragmented understanding. Recently, the National Bureau of Statistics of the CCP released the GDP data for the first half of the year, claiming that the constant-price GDP grew by 4.7%, and then grew by 4.3% in the second quarter, with the nominal growth rate exceeding the real growth rate for the first time in 13 quarters. Then, the rhetoric came back again: oh, economic performance is once again stable and improving. Well, let's take a look at how China's so-called nominal growth rate actually bounced back. It mainly relies on the deflator turning positive, and what does the positive deflator rely on? It relies on oil prices and price hikes in the AI industry chain, rather than demand. We just talked about this point previously. In addition, the real growth rate of exports in 2026 is only around 3 percentage points, while out of that 20% increase in paper exports, 80% consists of chip price hikes, exchange rate conversions, transit inventories, and export data padding. If exports are like this, what about the entire Chinese economy? Today, we won't obsess over the single digit of the CCP's GDP; instead, we will switch to a new way of thinking to see the truth. Let's see what the dashboard group looks like. It's a very simple question: that CCP GDP meter of yours has basically become the same thing as the CCP's current water and electricity meters. Even if you don't use the thing, it can still spin like crazy. It can run on its own. Although the CCP's data in other areas may not necessarily be very accurate, using multidimensional data at least shows us one thing: even if your other data are already very fake, they still cannot prop up the GDP data the CCP is bragging about. What do we use? We use factory electricity consumption data, highway and port freight data, the unemployment rate plus per capita working hours, and even data on marriage and childbirth, all treated as an independent dashboard. We bundle these data into a dashboard cluster, and then lay all these scales out on the table and calculate them with weights. What is the true color of the economy really like? It becomes instantly clear. Today, we are putting a whole table of these lie-detector instruments on display, piece by piece, and we will take them apart to look at them one by one.

Let's tackle the hardest part first: people. As early as last year's statistical bulletin, the Ministry of Human Resources and Social Security of the Chinese Communist Party (CCP) released a very interesting piece of data. What data was that? It stated that the total number of employed people nationwide was 725 million, which was 9.35 million fewer than in 2024, a drop of 1.27 percent. Meanwhile, the number of people collecting unemployment benefits surged by more than 20 percent in a single year. Yet, what was the growth rate of real GDP for that same year? It grew by a full 5 percent. Here, I want to introduce an indicator called labor output elasticity. In plain human terms, for every 1 percentage point GDP rises, employment is forced to change by a certain percentage point. Dividing the rate of change in employment by the growth rate of GDP gives you this elasticity. Now, if you take the total urban employment in all of China to calculate this elasticity, it is ridiculously fucked up: the elasticity is a negative 0.25. What does that mean? In other words, judging by the CCP's GDP statistical methods, employment, and workers' working hours, for every 5 points GDP grew, the number of working people had to drop by 1 point. Make sense of that. Broken down by city, these numbers are even more fucking absurd. Last year in Beijing, the number of people participating in urban employee pension insurance only increased by 0.74 percent, but GDP grew by 5.4 percent, resulting in an elasticity of 0.41. This means Beijing's GDP grew by over 5 points, while employment basically didn't move at all. As for Shanghai, the number of insured persons jumped by 1.15 million in a single year. At first glance, it might look like some kind of employment explosion, but of course, it's not an employment explosion. Wasn't it because last year in September, the CCP rolled out a new supreme judicial interpretation regarding social insurance? That is, any private agreement between a company and its employees not to pay social insurance is completely void, and an extra constraint was added: if employees sue the enterprise, the enterprise still has to make retroactive social insurance payments. Most of Shanghai's 1.15 million added employments are of this nature; it's completely not an increase in jobs, so the elasticity there is also zero. Look, aren't these tier-one cities under the CCP pretty awesome? They can conjure GDP out of thin air. What does an elasticity close to zero mean? It means growth through useless effort. And when elasticity is negative, it gets even more absurd: the more people work, the fewer people there actually are. If it were a normal country in a normal year, with GDP growing or the economy doing relatively well, what conditions should appear? Job postings on recruitment websites should also go up, right? Construction sites should resume work, and office building occupancy rates should go up as well. But what has the situation been since 2025? Fucking GDP is rising, but jobs are being cut, and office vacancy rates are also climbing. Cushman & Wakefield has a report with detailed statistics: what was the vacancy rate of Grade A office buildings in Shanghai at the end of 2025? It was 23 percent, while Shenzhen's once exceeded 30 percent, and the total office space absorbed by the four major tier-one cities in a year was only one-third of 2021 levels. An empty building means people are leaving. That's our standalone dashboard for employment. Now let's talk about the working hours dashboard. Regarding working hours, as we mentioned before, the average weekly working hours in 2026 decreased by nearly 1 percent compared to 2025, bringing an end to nine consecutive years of growth. In other words, combining headcount and working hours, the labor input alone was cut by around 2 percent. If you calculate based on the CCP's claimed GDP growth of 4.7 percent, then per capita output must have grown by a fucking 7 percent. How dreamy is this productivity boost? It is four times the labor productivity growth of the United States, and twice that of Taiwan, an economy centered around high-value-added industries like semiconductors. Moreover, ordinary growth doesn't even cut it; it has to be compounding growth year after year. If the CCP's GDP growth figures are true, then this isn't data—it's fucking science fiction. The existence of Trisolarans in this world would be officially confirmed. With Trisolarans providing technology over there, maybe they could achieve a month-by-month boost in labor productivity like this. Furthermore, we also know a fundamental problem: the CCP's employment data is definitely significantly overestimated. Because things like social insurance retroactive payments and statistical caliber adjustments are all ways to push the numbers to where they look bigger, right? And what gets left out are flexible jobs. But as for flexible-employment workers, we already discussed in the previous segment: 20 million food delivery riders fighting over 4 million people's worth of work. That isn't employment; it's just another form of unemployment.

Let's look back at this physical dashboard. From January to July 2026, total social electricity consumption rose by 4.7%, which is roughly in line with the CCP's GDP data—and remember, these are the CCP's own figures. In July alone, it only rose by 1.7%, while commercial freight volume rose by 2.8%, and port cargo throughput only rose by 7.1%; in fact, for the single month of July, it actually fell by 2.9%. Let's look at the railway sector: China's railway freight volume rose by 0.7%. What is mainly transported by China's railways? Isn't it coal, iron, and steel? If railway freight isn't moving, what does that mean? It means that goods related to heavy industry and infrastructure aren't moving either. However, during the same period, the value-added of industries above designated size rose by 5.3%—an increase of over five points. Now we can sort this data. You see, industrial value-added rose by over five points, but freight only rose by over two points, and port cargo only rose by over one point. The freight might be real, but what is the three-point gap calculated from the value-added? It's just price and statistical methodology. If these aren't goods in transit, where are they? They must be in warehouses, right? Look, the sales rate of products from industrial enterprises above designated size was 96.9% in July, which is 0.6 percentage points lower than the same period last year. What does that mean? It means more of the goods produced weren't sold and are piling up in warehouses. And what does value-added calculate? It only counts what is produced, not what is sold. Therefore, the portion that is produced but not sold is also counted as growth. This is another source of why value-added can be higher than freight volume. So why did electricity consumption still rise by 4.7%? I looked at the composition of electricity consumption, and what is it mainly? It's data centers and charging piles. Look at the data from the CCP's National Energy Administration by industry. Everyone look: from January to July 2026, electricity consumption for internet data services rose by 43.3%, while electricity consumption for charging services rose by 55.8%. The growth in these two areas doesn't correspond to productivity, but to what? Graphics cards and electric vehicles. It's not people working, nor is it goods being shipped. Once you strip away the growth from these two areas, the real growth in total social electricity consumption is less than 3%. So after breaking down the 2026 electricity data, what is rising? It's electricity being converted into computing power. And who is buying this computing power? Whether anyone is paying for it is another matter. This ledger has nothing to do with factory operating rates or household electricity consumption. These gauges aren't just for casual viewing.

I have established five lie detectors for this method, each corresponding one-to-one with a piece of CCP data and a physical gauge. Once the gap exceeds one standard deviation of the normal distribution calculated over the past decade, the lie trigger is activated. Then you just have to see how much urine is in the CCP's performance-enhancing drugs. Let's look at the first anomaly: the gap between the growth rate of industrial value-added above designated size and the growth rate of commercial freight volume has exceeded three percentage points for three consecutive months. In July 2026, this gap was four percentage points, which is a direct trigger for a lie. What is the second lie trigger? It's the export growth rate denominated in RMB divided by the growth rate of container throughput; only when it's within 2.2 is it a standard normal distribution, but from January to July 2026, this ratio reached 2.7. Hey, this lie item is also triggered. Let's look at the third lie trigger: the gap between GDP growth and the growth in the total number of employed persons also exceeds five percentage points. This part is also in a triggered state. The fourth is that the monthly fluctuation in the unemployment rate is less than 0.2 percentage points, but the number of people receiving unemployment benefits has risen by more than 15% in a year. The annual average unemployment rate of 5.2% corresponds to a 20.3% increase in the number of people receiving unemployment benefits. This part of the lie is also triggered. Even the last and most solid item is triggered. What is the last item? Looking at the electricity growth rate of the secondary industry lagging behind the industrial value-added growth rate by more than one percentage point, and this has persisted for more than two quarters, this part is also in a triggered state. All five are triggered. If this were a real lie detector, it would probably catch fire and explode on the spot, just like the battery in that shitty Huawei car. No need to think about it, there's no chance to think. The lies are just so fucking outrageous that the CCP's output data has absolutely no credibility to me. You can see that these five ratings are not emotional at all; they are just five physical dashboards that don't match a single bit of the bullshit the CCP is bragging about.

Let's continue to present evidence. What kind of evidence? The evidence of money. Has the money earned from exports actually entered the Chinese economy? Previously, we mentioned that half of the export growth in 2026 came from chips. And who are the main manufacturers shipping these chips? It is foreign-owned wafer fabs doing the shipping, while the payment for the goods is only settled within the territory. Therefore, the gap between the trade surplus according to customs statistics and the balance of payments surplus according to the State Administration of Foreign Exchange narrowed from $260 billion in 2023 to $47.6 billion in the first half of 2026, making the accounts match better than before. But matching accounts does not mean the money has actually stayed. From January to July 2026, the bank surplus in foreign exchange settlement and sales was $289.4 billion, while the customs surplus for the same period was $686.4 billion. Only 42% of the customs surplus was converted into RMB; the remaining 60% either sat in offshore accounts as foreign currency deposits or went straight out in the form of foreign investment. So, none of these channels will turn into what? They won't turn into orders for Chinese factories, payrolls, or goods. Why don't companies convert the US dollars they receive into RMB? Because the prerequisite for converting to RMB is that there must be a place to spend money within the territory. Companies spend money on only three things: expanding production, hiring people, or buying land. But in 2026, companies are not doing any of these three things. The dollars just sit in their accounts or go straight out. What does the settlement rate represent? It represents a gauge of corporate willingness to invest domestically. The reading on this gauge is only 40%, and it is not just the investment willingness that is skewed; the profit dashboard is also severely skewed. Look, from January to July 2026, the national industrial profit for enterprises above designated size grew by 17.6%, but the integrated circuit industry alone contributed 9.3 percentage points. If you exclude chips, the growth rate of the remaining industrial profits is only in the single digits. Moreover, aren't most of the chip profits recorded on the books of foreign-owned factories like Samsung in Xi'an or Hynix in Wuxi? Since dividends have to be remitted back to South Korea, they won't turn into orders for China.

Furthermore, there is a mirror image in Vietnam. Let's look at this: in the first eight months of 2026, Vietnam's trade surplus with the US was $106.6 billion, while Vietnam's trade deficit with China was $107.7 billion. These two figures are almost identical, meaning that China's intermediate goods are shipped to Vietnam, get a new country-of-origin label, and then enter the US. This means that only processes like packaging, testing, and logistics remain within China. The labor density of this part is much lower than that of full-unit assembly. The fact that the foreign exchange settlement and sales rate has been cut in half to about 40% essentially means that private foreign trade enterprises are institutionally refusing to inject liquidity into the domestic monetary system. Companies not converting foreign exchange is not because they have nowhere to spend it for the time being, but a systematic hedging behavior against the garbage-like asset yields and investment returns within China, as well as policy expectations. Therefore, the huge amount of US dollars remaining overseas or in offshore dollar deposits is, in fact, a legal capital flight completed through formal trade channels.

Let's look further at what? Let's look at Chinese household data. According to the CCP, the per capita disposable income of national residents in the first half of 2026 grew by about 5.2%. However, even according to CCP data, the nominal growth rate of per capita consumption expenditure was only 3.7%. By this logic, earnings are growing faster than spending. What does the CCP use this part to explain? They use it to explain the increase in savings. They say that household deposits increased by 7.58 trillion in the first half of the year, while household loans decreased by 366.8 billion, with a cumulative decrease of 827.1 billion from January to July. That means in less than a year, nearly 1 trillion in household loans has been wiped out. That's interesting! Even by the CCP's own logic, when would a family save money while paying off loans at the same time? Isn't that when income expectations are falling? The current situation is not at all the 5.2% wage growth narrative pushed by the CCP. This figure is fucking impossible when the real unemployment rate is as high as 10.8%. The situation is very clear: wages are stagnant—no, let's not even talk about stagnation; if a company can still pay wages on time and in full, it is already considered a very, very top-tier, very, very good company. At the very least, they should be considered quite decent companies.

So what should the real job market look like right now? Job positions—well, we can't even talk about stability anymore, because they are still dropping, and housing prices are still falling. At a time like this, cash in hand is the only certain thing. Every extra yuan saved counts, and anyone who can pay off their loans early should do so. This is called precautionary savings. In northeastern dialect, what is this called? It's money for the common people to get through the winter, or what some call money to get over a hurdle; it's money saved out of a sense of insecurity. So when everyone looks back at the data the Chinese Communist Party (CCP) starts putting out—that total retail sales of consumer goods grew by 1.2 percent from January to July, and a mere 0.6 percent for the month of July alone—well, this is also data spoken in a dream. Fixed asset investment dropped by 0.7 percent, private investment dropped by 9.4 percent, and real estate development investment dropped by a fucking 19.2 percent. Money is being spent on the cheapest possible things. Under these circumstances, I have no idea what the CCP is using to drive up retail sales. In the first half of 2026, the country saw a net decrease of 410,000 catering establishments. Under such circumstances, how can the CCP tell me what on earth is growing? Therefore, what do I judge the real situation to be? Wage incomes are declining, social retail sales of consumer goods are also dropping, and the decline in retail sales is actually faster than the decline in actual wages.

A surge of over 7 trillion yuan in savings is, in my view, a much more logical interpretation of the current situation. Then the demographic ledger is also trailing right behind. Look at this: the nationwide birth population in 2025 was 7.92 million; in the first half of 2026, marriage registrations—even under the nearly hysterical propaganda campaigns of the CCP—dropped by 264,000 couples compared to the first half of 2025; and in 2025, the country had a staggering 21,400 fewer kindergartens than in 2024. Furthermore, this trend is accelerating this year, and the areas where closures are most numerous and dense are none other than the suburban fringes of Tier-1 cities where migrant populations cluster. As we mentioned before, the migrant population is contracting, which completely overlaps on the map, further corroborating the fact that people are truly retreating back to their hometowns. Moreover, after retreating, they are no longer having children. The kindergarten sector is the most lagging indicator on the entire dashboard, because the kindergartens closed in 2025 correspond to what? They correspond to the children born three years prior. And since the marriage figures in the first half of 2026 continue to trend downward, the readings on this dashboard for the next few years are already sealed.

Now let's ponder another question: even if these people retreat back to the county towns, can the county towns absorb them? They still can't. Because all a county town can offer is a cheap cost of living—well, perhaps you have a house in your hometown, so at least you don't have to spend money on that. But the hometown cannot provide any job positions. Some financial media outlets nowadays are also fucking out of their minds, starting to crazily hype up the concept of resolving local government debt, claiming that this year over 80 percent of financing platforms have exited, and that there is also 800 billion yuan in special bond funds for debt resolution. The media of the CCP right now is truly fucking blind and talking absolute bullshit, claiming that this portion of money will generate some investment. What are you actually resolving with debt resolution? You are resolving money that was fucking borrowed in the past, and the essence of the CCP's debt resolution isn't even to make the debt smaller, but to swap local government credit with central government credit. This fucking money cannot provide new job positions either. Then, around August 10th, the Ministry of Human Resources and Social Security of the CCP issued another extremely, extremely heavyweight piece of news—except this news failed to ferment. What did it say? It stated that during the 15th Five-Year Plan period, it is necessary to strengthen the governance of wage arrears for primary and secondary school teachers as well as medical personnel. The CCP's original wording was that teacher salaries in some regions have experienced delayed issuance, and the performance-based pay of medical personnel cannot be fully paid out on time. Look at that—even the salaries of teachers and doctors fucking require a separate official document to govern unpaid wages, and you're actually expecting county towns to open up job positions for the returnees?

Why single out childbearing? Because this part is also a major chunk of Chinese family expenditures. Daily expenses for the Chinese public mainly consist of three major things: housing, cars, and children. As for cars, as already mentioned, they directly plummeted by 17% in July. In other words, not a single one of these three major areas is seeing spending. So, can someone please tell me how on earth this total retail sales of consumer goods managed to rise? Alright, now that the dashboard has been laid out, we begin weighting. How do we weight? This method is called the multi-sensor Kalman filter, which in plain human words is actually very simple: it means taking a row of dashboards to simultaneously measure something we cannot see, and then whoever has the larger noise gets assigned a smaller weight. If someone's reading is chronically too high, we first deduct those excessively high parts and then apply the weighting. The dashboard for the the Chinese Communist Party (CCP) GDP has actually been a large chunk higher than the combined readings of electricity consumption and freight transport since 2021, so I simply deduct this chunk as a fixed over-reporting margin, and then set the sequence noise of the CCP's data to three to five times that of physical dashboards, with the rest being weighted.

Let's take a calculation example: suppose at the starting point, looking at the whole year of 2025, even assuming the data provided by other economic sectors in China are all true, the true China economic growth rate is only around 2 percent. Then in the first half of 2026, all five dashboards are reporting this anomaly. The GDP growth at constant prices by the National Bureau of Statistics of China reports 4.7%, but society-wide electricity consumption, after excluding data centers and charging piles, only grew by 3%, commercial freight volume only grew by 2.8%, labor input calculated from headcount and work hours decreased by 2%, and adding the 3.5% per capita output gives 1.5%, plus the CCP's GDP itself must first have a deduction taken off. From 2023 to 2025, whether independent economists or economic data groups like Rhodium Group, their estimates of China's true growth rate averaged over two points lower than the numbers reported by the CCP. I take a relatively conservative 1.5 points as a fixed over-reporting on the CCP's dashboard. Under this circumstance, what does the CCP's GDP reading become? 3.2%.

Let's look at the weights again. The CCP's GDP data is chronically high year after year and has extremely high noise, so the weight can only be given as 1—not giving it 0 is already giving them face. Meanwhile, electricity, freight, and personnel have relatively small noise, so each is given a weight of 3, and the starting point of 25 is also given a weight of 1. The weighting is 3.2 times 1, plus electricity's 3 times 3, and so on, plus 2.8 times 3, plus 1.5 times 3, plus 3 times 1, and then divided by 11, which equals what? It equals 2.7%. This is what the Kalman filter does. There is no need to argue with those fucking idiot data points from the CCP; first, carve away that historically high-reporting section on the CCP's dashboard, then assign weights based on whose noise is smaller, combining these dashboards into a single reading. I still use that 4.7% from the CCP, hey, but only after it has been deducted and discounted can it be used. The final calculated, relatively optimistic growth rate is merely 2.7%.

Let's calculate one step deeper: the CCP's urban surveyed unemployment rate is 5.2%, but if calculated based on an actual unemployment rate of 10.8%, that hides 43 million people without work in the middle. Moreover, based on the long-term relationship between the CCP's urban unemployment rate and the corresponding youth unemployment rate, which has hovered around 300 over the years, the youth unemployment rate should be at least 30%, and this part yields an extra 4 million people. If these 47 million people are calculated as accumulated over three to six years, the true employment figures are dropping between 2.2% and 3.2% a year. Coupled with a 1% drop in weekly work hours, if labor input decreases by around 3.6% a year, and per capita output is still given full face to the CCP and calculated under the 5% condition, even with such a calculation, China's true GDP growth rate in 2026 would only be between 0% and 1.5%, with the midpoint around 1%. If we switch to the rulers used by the United States and Taiwan—that is, the labor productivity ruler—zero growth up and down is already the greatest respect shown to the CCP.

Finally, let's combine the previous export data with today's GDP. What does a bad macro-environment mean? There are phenomena to support it now: people are retreating back to their provinces, working hours are plummeting, and the growth of real exports is hovering around three percent. The readings for electricity consumption and freight transport are also around three percent. The household sector is clearly saving money rather than borrowing. And the last areas of spending are marriage and childbirth. The three curves for marriage, births, and kindergarten enrollment are all dropping together, and they are dropping in perfect alignment. In the first half of 2026, the optimistic real economic growth rate for China will be between two and three percent, with the midpoint being 2.5 percent. This so-called optimism assumes we still trust a portion of the CCP's data; if we calculate based on real unemployment rates and labor productivity, zero growth would be the most optimistic scenario. So, what is the greatest function of these dashboards? It is to tear the CCP's GDP facade to pieces, leaving not a fucking shred behind. As for the CCP's current propaganda paradigm, could you please innovate a little? You just keep harping on about this GDP and exports. Before, you were firmly blowing the bullshit of five percent growth; now, even your fucking bullshit is being discounted. Oh, and there is also that trillion-dollar trade surplus narrative. These are the only two lonely brothers left to prove that the Chinese economy is still doing well.

But faking isolated data points is useless because every cent in the real economy is fucking voting with its feet. The foreign exchange settlement rate has been slashed in half to forty percent; this is the market slapping the CCP's GDP in the face with the loudest possible sound. That is, if the economy were really galloping forward at a five percent rate, why isn't money pouring in, but instead scrambling to escape? Isn't this just the reality that even foreign trade bosses have recognized? That they would rather treat the US dollar as a safe haven than convert it into RMB to serve as fuel for this internal circulation. In the end, isn't this GDP, which is detached from the monetary cycle and detached from actual wage income, just a complete and utter act of statistical self-defense? Or, in other words, the CCP's cyber-crisis? Even if you, the CCP, fucking inflate these two data points to the sky, what good does it do to make them look so fucking pretty? Brother, your data looks so fucking good, yet the real economy is hemorrhaging, and the common people haven't even found a sip of soup to drink in this digital self-defense of the CCP. After seeing these garbage numbers, won't the common people just get even more fucking angry? Therefore, this garbage data from China is not just self-defense, not just a cyber-crisis, but what else? It forces you to look at it. Tell me, what normal person could fucking stand that?