The dysfunctional tiger
How to upzone Hong Kong
Events are one of the most important things we do at Works in Progress. We find them immensely valuable for meeting new authors, finding new ideas for articles, and getting suggestions on how to improve. If you’d like us to invite you to events in your city, please give us your details here.
By many metrics Hong Kong has the world’s most expensive housing. It is difficult to get firm and comparable numbers, but property in the territory costs about $24,000 per square meter. In London it is $10,000 per square meter. Like many cities, Hong Kong’s affordability crisis is due to a shortage of housing. But on the face of it, the territory has a good excuse – its terrain. Our image of Hong Kong is a city where skyscrapers nestle in between mountains. It is tempting to infer that its housing shortage is because the territory has simply run out of space where homes can be realistically built.
But it would be a mistake to infer that the territory’s housing shortages are because there is no more space in between the mountains. Like in nearly every country, Hong Kong’s problems ultimately derive from uncertain property rights regimes that do not let those who would benefit from development trade with those who have the power for it to be blocked.
It is untrue that Hong Kong has run out of flat land to build on. For one, 49 of Hong Kong’s 1,117 square kilometers of land is agricultural, but remarkably, in 2016 the government estimated that only seven square kilometers of this is actually farmed: the rest lies fallow. For another, there is 17 square kilometers of brownfield land, which includes open storage, car parks, vehicle repair yards, and port facilities. Much of the brownfield and agricultural land is owned by private developers, in the hope that one day it will be rezoned for building.
Another 34 square kilometers is zoned as ‘village-type development land’ reserved for the ‘Small House Policy’. This policy allows male indigenous Hong Kongers to apply for permission to build a small house. Applicants must be descended in the male line from a villager who lived in 1898 in the New Territories, the area leased in that year by China to Britain, which expanded the size of the colony sevenfold. Small House Policy houses are limited to three stories and must be built in the applicant’s own village. The policy was adopted in 1972, but since the Handover, it has been protected by Article 40 of Hong Kong’s Basic Law, the territory’s constitution.
Not all of this 34 square kilometers of ‘village-type development land’ is actually built on. Just over a quarter of this land is vacant, because it has been reserved for future claims by indigenous villagers. In other words, almost 1,000 hectares of the most expensive city in the world has been left vacant just in case an indigenous Hong Konger might want to build a three story house.
The politics of housebuilding in Hong Kong
In summary, Hong Kong has about 90 square kilometers of land which is being inefficiently used, either because it lies fallow, hosts low-value brownfield uses, or is reserved for three-story houses for indigenous Hong Kongers. By comparison, this is twice the total area used for housing, excluding areas set aside for the Small House Policy.
The land hasn’t been developed because the politics of housing in Hong Kong are toxic. Immediately after Handover, the Chief Executive proposed the ‘85,000 policy’ to build 85,000 homes a year. This was scrapped following the Asian financial crisis in the same year, which led to a collapse in housing prices that was widely (and falsely) blamed on the 85,000 policy.
Hong Kong also has many anti-development groups. Activists complain about the dispossession of farmers whose ancestors settled in the New Territories in 1900, just missing out on the Small House Policy. They lament ‘special treatment’ given to developers and worry about the ‘disappearance of Hong Kong agriculture’ when the government tries to build on rural land. Even though country parks occupy 40 percent of Hong Kong’s land area, when the Hong Kong government attempted to extend a landfill site into a country park by just five hectares, it was forced to back down following pressure from activists.
It is tempting to assume (in the democratic world) that we could ‘crush the NIMBYs’, if only we had a sufficiently determined leader who could seize power in a landslide – a Margaret Thatcher, a Lyndon Johnson or a Lee Kuan Yew. But even though Hong Kong is not really a democracy, and its government has the power to change its planning laws, in practice the territory is unable to build on its ample available land.
Land law in Hong Kong
On top of all its powers, Hong Kong’s government actually owns 100 percent of the land in the country. Since the colonial period, all land in Hong Kong has been owned by the government, which sells leases to developers, who in turn sell sub-leases or rent apartments to tenants. In the colonial period, these base leases were granted for 99-year terms, but since handover this has been 50 years. The leases function quite a lot like freeholds: they can be bought and sold, and allow the lessees most of the rights of freeholders.
The important difference from normal freehold is that the Hong Kong government has the power to bring the lease to an early end, in exchange for a payment to the lessee of just compensation for the remaining term of the lease. This is known as resumption. It means that the Hong Kong government can in principle take control of land at any time, for example to densify it.
In theory, this offers a straightforward way for the Hong Kong government to enable more house building. The government could buy out the lessees, upzone land currently reserved for agriculture or indigenous houses, and then re-lease the land to a developer. The only cost would be the compensation payable to the lessees, and here the deal for the government is sweetened by the way this compensation is assessed. It is not determined by the value of the land after upzoning, but by reference to the land uses permitted in the lessee’s deed. Land zoned for agriculture would be compensated at its agricultural value, not at its value if it were used for housing.1
Even though this tool is available, the Hong Kong government rarely uses lease resumption. The 50- or 99-year leases are effectively treated by the government as a freehold. The government (rightly) regards protecting private property as important, and is also aware that Hong Kong’s free and open economy depends on maintaining property rights. Even though, as a matter of black-letter law, the Hong Kong government owns all land in the territory, it rarely acts like a conventional landlord. Indeed, the government deliberately binds its own hands: the statute authorizing the Hong Kong government to order land resumption only allows its use for a ‘public purpose’. This has been interpreted as excluding private housing development.
When the government does try lease resumption, it frequently gets bogged down by legal battles over the value of compensation payable. Assessing the value of the land at its current use is not always straightforward, and judicial reviews of land resumption decisions are common.
Instead of lease resumption, the Hong Kong government uses the ‘conventional’ method of permitting development. Lessees maintain control of their land, but need to apply to the government to have their land upzoned. As well as running into the same problems as in Britain and America, with local objectors trying to resist changes to the status quo, Hong Kong adds a further barrier. Any lessee whose land is upzoned needs to apply to the government for a modification to their lease, because the terms of the lease dictate what they are allowed to use the land for.
In theory this should not pose any substantive barrier: the modification of the lease could mirror the terms of the upzoning. But the Hong Kong government has a duty to make money out of lease modification, so it charges lessees a premium in exchange for modifying their lease, equal to the gain in value of the land following the upzoning.
This stage of the process is fraught with legal uncertainty, in a manner similar to Britain’s attempt to introduce a land value tax in the early 1900s. There is a right of appeal from the initial valuation, and after that there is the possibility of seeking judicial review.
Most importantly, the Hong Kong government has no power to compel a lessee to accept its terms for lease modification; to do so would be tantamount to expropriation. This, in turn, leads to protracted negotiations between the government and the lessee over what they are allowed to build. These negotiations, not the changes to zoning, are the primary cause of delay to getting things built.
Bilateral monopoly
This gridlock is what economists call a ‘bilateral monopoly’ problem. The transaction involves, in effect, the lessee buying a lease modification from the government, in exchange for the premium. But neither side is in a competitive marketplace. On one side of the negotiating table, if the government tries to impose an extortionate premium, then the lessee cannot really do anything about it. There are no other sellers of lease modifications. On the other side, the government is unable to go elsewhere if the lessee does not like the government’s terms, because there are no other buyers of lease modifications in respect of that plot of land.
This encourages the parties to engage in drawn-out, bad-faith negotiations: the lessee is incentivized to downplay how much they value the land, in the hopes of securing a lower premium, while the government has an incentive to over-egg the value of the land to make more money out of the lease modification.2 To make matters worse there is an information asymmetry: it is likely the lessee will have a better idea of the true value of the land than the government.
This is a problem familiar to economists, and the theoretical solution is also familiar. The goal is to ensure that the asset (here, a lease modification) ends up in the hands of the party who values it the most. A neutral arbiter values the asset, and the government is forced to sell the lease modification to the lessee at this objectively-determined price. If lessee does not accept that price, then the government gains information about how much they truly value it. Economists have come up with all manner of exotic variations on this theme, involving auctions, shared ownership or self-assessment, to ensure that the initial valuation of the asset is as accurate as possible.
But all of these ideas have remained theoretical toys, both in Hong Kong and in other places with similar problems, because they ignore the politics. Property owners form a political bloc – a constituency – of people who share an interest in avoiding any downgrading of their rights. They have a sense of entitlement to those rights, and they are keen to block any change.
In Hong Kong, two constituencies have an interest in maintaining the status quo. Property tycoons are enormously influential. Representatives of indigenous villagers can also claim to have justice on their side – preserving their right to build a three-story house was a promise made to them by the British in 1972. Both of these groups are supporters of the pro-Beijing government of Hong Kong.
In short, legal reform to change the lease system is unlikely to happen. The fact that Hong Kong’s housing shortages are caused by land law rather than land-use policy is unusual, but the barriers to change are familiar to any Anglosphere YIMBY. Existing laws and regulations benefit a group of people, who can get themselves politically organized to block any reform to the status quo. Theoretically neat reforms will founder unless they can find a way to break the deadlock. Hong Kong needs to be smart, not clever.
Breaking the deadlock
An interesting paper proposes creating a new kind of property right to sidestep the gridlock: Land Options for Housing.
Under the proposal, the Hong Kong government would create a number of land options which would be sufficient to meet the territory’s housing needs, and would distribute them among the population. The Hong Kong government would conduct a comprehensive planning exercise to define the land suitable for high-density residential housing, such as fallow farmland and low-value industrial land. On this land development would be permitted by default if a developer owned enough land options: holding a land option would serve as a kind of certificate of pre-approval for construction. Developers that wished to build would be required to bid against each other to buy land options, either in kind (in exchange for a unit in the proposed development) or in its cash equivalent.
The sale of the land option would be tied to a particular project, and on completion the seller would be paid or given their new property. This means the land option holders would have to bear some of the risk of the project’s failure: if their option were sold at too high a price, making the development ultimately unprofitable, then they would not be paid. Likewise, a project that went bust due to poor planning cost overruns would not pay out to option holders (although they would not be on the hook for costs). This, in turn, means that the cost of options for a particular development would price in the riskiness and quality of the development.
The land options scheme sidesteps the bilateral monopoly problem by creating competition on both sides. Option holders who try to hold out for a high price are likely to be underbid, and they are unable to try to extract concessions such as public infrastructure from developers, as the monopolistic government is able to do. (Land option development would be subject to a uniform levy to fund infrastructure instead.) On the other side, developers who try to lowball their offer run the risk of being outbid by their rivals. Furthermore, by designating a large amount of land in the territory as buildable, the government no longer has to haggle with developers over a project bespoke to a particular plot of land: the right to build becomes standardized.
Land options also avoid removing any rights from Hong Kong’s powerful constituencies. Indigenous villagers currently have the right to build a three story house, and the property developers that speculatively own open storage yards have no right to build houses at all: nobody has the right to build residential high rises. This means that nobody can complain that their rights are being expropriated.
Most importantly, the scheme would allow those who benefit from development (developers and the new residents they build apartments for) to effectively trade with other residents who would otherwise oppose it. The benefits of development would automatically be shared with the residents of Hong Kong who hold options, creating a structured and dependable bargaining process out of the endless haggling of the current system. These residents would become a constituency in favor of development, because they would directly benefit financially from new housing.
In spirit, land options are similar to New York City’s reforms to its air rights. The space above buildings in Manhattan is some of the most valuable real estate in the world, but they often go underused. Broadway theaters, for instance, do not want to allow themselves to be demolished to build a skyscraper. The City allowed theaters to sell their right to build upwards to the owner of another site, who would then be allowed to build something taller than they otherwise would be. This meant that historic theaters could raise funds and, crucially, that their audiences were enthusiastically in favor of the reform.
They are also similar to how Dublin reformed its taxi medallion system. The city gave every existing licensed taxi driver another license, which they could sell onto another driver. By creating new property rights and giving them to the opponents of change, Dublin was able to double the number of taxis, without forcing a stand-off with a politically powerful taxi lobby who were opposed to ending their monopoly.
What else might work
There is no shortage of proposals to enable more housebuilding in countries that have shortages. Some of these reforms get passed. But bitter experience has taught housing reformers that passing a reform in the first place is only half the battle. Policies that enable more houses to be built are vulnerable to repeal: witness the experience of New Zealand’s upzoning or the London Borough of Croydon’s attempt to densify its suburbs. Land options could be the solution to Hong Kong’s housing troubles, by broadening the constituency with a vested, durable interest in development.
One idea could be to allow a group of lessees of neighbouring plots – say, a city block – to vote by a supermajority for their leases to be resumed and re-leased to a particular developer, with the land being upzoned at the same time. The lessees would be able to choose which developer the land would be re-leased to: the only direct involvement of the Hong Kong government would be in facilitating the transaction. The compensation that the lessees would receive would be funded by the proceeds from the new development – but rather than being based off the land’s value at its current use, which under-prices the compensation and is easily subject to litigation, the compensation would be agreed between the lessees and the developer.
This would enable Hong Kong to densify areas of the city that are in need of regeneration. Large swathes of Hong Kong consist of housing built cheaply in the 1960s and 1970s. By allowing lessees to vote in favor of their own redevelopment, Hong Kong could build more homes with the consent of the people who already live there.
Benedict Springbett is a writer and Bar student based in London.
This is similar to how the British government is planning to reform compulsory purchase so as to enable the compensation for expropriation to exclude ‘hope value’, the value of the land that comes from the possibility of granting planning permission.
This has similarities to the haggling over section 106 contributions in Britain, which are contributions made by property developers to local governments as a condition of being granted planning permission, negotiated on a case-by-case basis for each development. See Housing, Communities and Local Government Committee, Land Value Capture (HC 2017–19 766), para 50.