In China, the Summer Camp Is Subsidised in a Way You Wouldn’t Imagine

When a Chinese cost structure seems to defy the rules of the market, look for state assets and institutional relationships.

---

Introduction: Six days, 6,580 yuan, an impossible programme

I signed my daughter up for a six-day tech summer camp in Shanghai, run by Shanghai Jiao Tong University, one of China’s top four universities. Price: 6,580 yuan, full board, or about 822 euros. Accommodation, meals, supervision, classes, visits, equipment.

The programme:

• Day 1: arrival on campus, accommodation at the university hotel, Unitree robots (宇树), introductory AI class
• Day 2: visit to the automated Yangshan port, SJTU’s Sino-German Industry 4.0 innovation centre, collaborative robotics
• Day 3: National Integrated Circuit Innovation Centre, a class on “from grain of sand to chip”, circuit design, principles of large language models
• Day 4: Baidu showroom, calibration workshop, a ride in a Luobo Kuaipao Robotaxi in live service, drone class, disassembly and reassembly
• Day 5: Shanghai Tower, smart building, scale model of a future AI city
• Day 6: presentation of the three mini-reports written during the week

Looking at this programme, one question arises. At 1,097 yuan a day, full board included, how is this economically possible? Pointing to Chinese costs would be a convenient shortcut. The real answer is more interesting.

---

I. The benchmark: what this programme would cost at market price

In China, tech camps in Shanghai range from 5,380 to 9,900 yuan. Two references: Discovery Education Science Camp, 5 days, 6,380 yuan. AI 脑动力 camp, 5 days, 7,588 yuan. These programmes offer classroom-based science workshops. No visits to industrial or state infrastructure. No full board.

In France, a comparable tech camp costs between 870 and 2,500 euros a week depending on the operator and the level of supervision. Telligo, GoColo, IT School and KidnKod offer robotics and programming stays with Lego Mindstorms kits, Scratch and Python workshops, and one instructor for every 6 children. No visits to industrial infrastructure. No full board.

The decisive point of comparison is not the headline price. It is what that price buys.

A non-residential coding course without accommodation, at a private school in Shanghai or Paris, already costs between 3,000 and 6,000 yuan a week, as much as or more than the entire SJTU stay. The classes alone exceed the price of the full package. Without board. Without round-the-clock supervision. And above all, without the visits that are the real point of the programme.

Because this is the point parents sometimes forget. Children aged 10 to 14 do not spend their days coding. No one turns out an engineer in six days. What matters at that age is what they see: a fully automated port without a single human on the docks, a driverless vehicle taking them somewhere, a national research centre where chips are made from sand. Open-mindedness, discovery, a sense of the scale of things. Those hours are worth far more than one more Scratch workshop.

And it is precisely those hours to which private organisers cannot get free access.

Rebuilding the SJTU programme at market price, for a private organiser with no university ties:

So is SJTU selling at a loss? Of course not. The programme makes a margin, because it rests on a double invisible subsidy: state assets mobilised at marginal cost, and institutional relationships that money cannot buy. The cost exists. It is simply recorded on the state’s balance sheet, not in the price parents pay.

---

II. The invisible subsidy: six mechanisms

The subsidy operates on two levels. The first is tangible and quantifiable: public assets mobilised at marginal cost or through internal transfers. The second is intangible and unquantifiable: access, through the university’s institutional relationships, to places money cannot buy. The first lowers the price. The second determines the value.

---

Level 1: Public assets at marginal cost

Campus infrastructure, during its idle period. The children stay at the SJTU campus hotel, eat in the university canteens and take their classes in the university’s teaching rooms. In July and August, these facilities are empty. The campus runs idle, but fixed costs keep running: maintenance staff, energy, depreciation. Any use during the off-peak period is marginal revenue, never a cost. The cost charged to the organiser is an internal cost, bearing no relation to the market price of a hotel, a caterer and equipped rooms in Shanghai.

Available university staff. Classes are taught by Jiao Tong staff. A university of this calibre employs thousands of administrative, technical and teaching staff, a significant share of whom are structurally underused during the summer holidays. Their pay is already covered by the public budget. Teaching at a camp represents a marginal top-up, not a full cost.

Students on summer internships as supervisors. The group leaders are university students doing their summer internship, often for academic credit. The cost to the organiser is marginal, sometimes zero. In France, a BAFA-certified youth leader costs between 60 and 90 euros a day, including social charges, or roughly 500 to 750 yuan per child over six days at a 1:6 ratio.

Equipment from the labs. The Unitree robots, circuit design kits and drones very likely come from SJTU’s in-house laboratories. A Unitree Go2 quadruped robot costs between 16,000 and 100,000 yuan depending on configuration. A private organiser would have to buy or rent it at market price. For the university, it is an internal transfer between departments, of equipment already depreciated and available during the off-peak period.

This first level explains the price gap. It probably represents 2,000 to 3,000 yuan per child. It is significant, but it remains quantifiable, and theoretically replicable by an operator owning its own assets.

---

Level 2: Institutional access, the asset money cannot buy

This is where the heart of the matter lies, and where the subsidy becomes both the most intangible and the most decisive.

The automated Yangshan port. The National Integrated Circuit Innovation Centre. The Baidu showroom and its calibration workshop. A ride in a Luobo Kuaipao Robotaxi in real commercial service.

This access stems from an architecture of relationships that Shanghai Jiao Tong University has built over decades: joint laboratories with Baidu, research agreements with the National Chip Centre, recruitment pipelines into Shanghai’s port SOEs, industrial partnerships with Unitree. These companies and institutions open their doors to SJTU because they recruit their engineers there, fund projects there, and find their partner laboratories there. The summer camp is merely a marginal use of a relationship built for entirely different reasons.

For a private organiser, this access is either closed or very expensive. There is no price at which a private travel company could buy a demonstration Robotaxi ride, a visit to the National Integrated Circuit Centre, or entry into an automated strategic port. The service is not in the catalogue. There is no ticket office.

And that is precisely what the programme sells.

The asymmetry, then, is not an asymmetry of price. It is an asymmetry of access. A private organiser willing to sell at a loss still could not match the programme, because the core of its value cannot be bought.

---

III. Takeaways for foreign companies in China

1. Never analyse a Chinese price without identifying the operator’s ownership structure. Before benchmarking a competitor, the question is: who owns its assets, at what price did it acquire them, what institutional relationships does it draw on? The question is not “how do they manage to be cheaper” but “what assets do they mobilise that I cannot?”

2. Determine early whether the asymmetry is prohibitive or can be worked around. If the advantage rests on legacy land holdings and deep institutional relationships, the asymmetry is prohibitive: do not compete head-on. If the advantage rests on negotiable access, it can be worked around by building your own relationships, over five to ten years. If the advantage rests on genuine operational efficiency, it can be challenged through innovation.

3. Seek partnership rather than competition. Faced with a structurally advantaged competitor, the effective strategy is rarely direct competition. A foreign company that becomes a supplier, service provider or co-organiser for a state-backed player gains indirect access to the same cost structure.

4. Map institutional relationships, not just competitors. Conventional competitive mapping (market share, pricing, range) is not enough in China. It must be complemented by institutional mapping: which universities, which SOEs, which local administrations are active in my sector, and where can I fit in?

---

Conclusion

The invisible subsidy is not a Chinese anomaly. It involves neither fraud, nor concealment, nor any breach of international trade rules. It stems from an ownership structure and a way of relating between institutions that have no equivalent in the West.

It is invisible precisely because it is nowhere. No budget line, no accounting trail. A Western analyst looking for the subsidy in the documents will never find it.

What a 6,580-yuan summer camp demonstrates on a small scale, the Chinese economy practises at every scale. A foreign company that fails to build this mechanism into its competitive analysis, partnership strategy and financial models is making decisions based on an economic reality that does not exist.