My body and I have drawn two conclusions from a couple of weeks of traveling through southwest China and eating copious amounts of tongue-numbingly spicy food: China’s public bathrooms have (very marginally) improved, but the urban planning has not. (I had a lot of time to think about urban planning while utilizing those public bathrooms, due to the aforementioned spicy food). Provincial capitals like Kunming and Changsha were expansive last time I visited a few years ago; the cities have only grown larger since, sprawling further and further from what used to be the city center. Thirty-story apartment buildings sit stacked in neat rows, stretching for miles in every direction, as though they were stamped there by a bureaucrat over-using the copy-paste function on his computer. Huge new tracts of former farmland on the outskirts of the city have been razed and flattened in preparation for further expansion. It takes an hour to drive from one side of the city to the other not because of bad traffic but because there are tens of millions of housing units, spaced far apart, that take up an incredible amount of physical space. It should be obvious to any observer that these cities have way more housing units than people. People may be living in houses in the city center, or they may be living in houses on the outskirts of town, but they can’t be living in both.

For a number of years, fears of a Chinese housing bubble rested on “ghost cities”: brand-new cities, built on local government debt and backroom deals, rising from nothing in the middle of nowhere and devoid of people. Developers were building with no regard for demand, so apartment blocks would sit unbought, unwanted, and slowly crumbling into oblivion. Excess housing stock is a problem in some places, but it pales in comparison to a related, and possibly more sinister issue. Vacant homes sit empty, but they have already been purchased as a second, third, or fourth home. Unwanted and unused are not synonyms in a country with no property taxes, an economy driven by real estate, negative real returns on regular bank deposits, and a volatile stock market.

Housing demand is insatiable not because people want to live in houses, but because they want to own houses. As I wrote about last year in Hebei (the New Jersey of China, if you will), even working class families purchase multiple houses. (If you trust the advice of a real estate company’s blog, a single family should buy six homes to feel financially stable: one to live in, one for each set of parents to live in, two for their children’s future use, and one to rent out.) Up to 25 percent of all housing in China is owned but not occupied as of 2015, a rate far higher than in other countries around the world. In China, housing is simply money that you can sometimes live in. There is no tax on holding property, so unused housing can sit there and increase in value. When you earn more money, you want to store it away in a place that is relatively safe and will earn high rates of return, which it seems, in China, means buying houses that nobody lives in but might come in handy later when your children can’t afford to buy one or you need to get a lump sum of cash to send your child to school in the United States.

We tend to think of bank deposits as safe and real estate investment as risky. Such a view, however, is built on the premise that the government will protect our money and that in the long term, interest rates will gradually create a small but stable return. Neither of those is obvious in China. Without much political trust in the banking system (controlled and operated by the state), why would a citizen choose to put their money in an invisible, liquid asset rather than an actual tower of concrete and steel? A fixed asset is more reliable in the minds of many Chinese citizens because it literally cannot be moved or disappeared with one stroke of a pen. After decades of real estate investment, from officials with hundreds of off-the-books apartments to single families saving up to buy a second or third home so their children can be socially eligible to marry, enough people’s assets are in the form of housing that housing has become a de facto banking system.

With so many people’s wealth tied up in the housing market, the ultimate fear is that housing prices will decline. If housing is a bank, allowing housing prices to drop is equivalent to banks losing money and all of their customers taking a hit. It would presage deflationary pressure and financial instability, if not a crisis, for housing prices to take a hit, so the government has to implicitly guarantee that they will not decline. The policy options for corralling an overheated market are somewhat constrained by the simple fact that allowing housing prices to decline, even if they are wildly overvalued, would presage large-scale financial instability. And if you haven’t heard by now, the Chinese government is not a big fan of instability, nor are they looking for massive deflation.

In this way, it mirrors China’s urban land market. All urban land in China is state-owned in the form of leases ranging up to 70 years. Nobody has determined what happens when leases expire, however, and when a set of leases expired in Wenzhou in 2016, the government decided to not decide and offer temporary renewals without actually solving the issue. According to the original plan, the state should be able to charge for renewal (or reclaim the land when the lease expires). Yet, as Peter Ho argues, the system has evolved endogenously to create a credible system of property holdings even without official property rights. The property market has been operating with the belief that property rights will be renewed indefinitely, and so the actual policy itself is now constrained so it must reflect that or risk massive instability. The possibility of charging for renewal or refusing renewal is so anathema to homeowners (and developers) that such an outcome is politically impossible. The same applies for housing prices, or for implementing a property tax on holdings: since everyone has been operating for so many years on the assumption that housing will appreciate over time, actions that would cause significant depreciation would create instability and chaos.

Thinking about housing as money means that regulators will be very careful to avoid allowing home prices to drop, and will use all of the tools in their arsenal to prevent such an outcome. Yet, for the same reason, even a bubble of such bubblicious proportions does not have any reason to crash any time soon. The purported value of vacant apartments in China far exceeds the actual value that they could possibly have, but this is a representation of implicit government backing rather than actual value. A house is like a stack of bills, rather than an actual asset. Financial stability now depends on keeping the housing market stable, because housing is money. And as long as people believe that the government will step in to protect housing from losing value—and has the capacity to do so—there’s no reason to think it will decline.

But Chinese housing is not like money, because money is not going to crumble and fall apart soon after it comes into existence. The normal estimate is that Chinese buildings, constructed as quickly and cheaply as possible (hence the cookie cutting), will last thirty years; after talking to some architects, I am led to believe that even thirty years is an optimistic estimate. Buildings are built as quickly and cheaply as possible, which means they will slowly crumble from the inside. If housing cannot depreciate because it is money, but the walls are caving in, the value question becomes contradictory. It is hard to believe that people will think that a pile of rubble is worth the same thing as an apartment unless there is an implicit guarantee or suggestion that the housing will be rebuilt every thirty years, free of charge. China’s housing bubble seems unlikely to pop, but it may well crumble.

For the sake of the environment, I’m glad that most of the world chose small pieces of paper rather than ungainly, poorly built, and resource intensive skyscrapers full of empty space to represent government-backed value. But at the end of the day, nothing is real–at least until it collapses on itself due to shoddy building materials.