There are many voices in Hong Kong proclaiming unanimously that the city is in the midst of a housing affordability crisis. Another chorus of voices appears to proclaim that Hong Kong has a poverty problem. The latter is sometimes described as a gaping inequality between the haves and the have-nots. Poverty of course makes housing unaffordable, but this is a result of poverty, not its cause. The best remedies for poverty are very often either a direct income supplement or an indirect subsidy to provide education and training – not a solution focused on housing.
But to come back to the question of housing affordability, we have to ask whether there is indeed such a crisis in Hong Kong. And if so, what are its causes and what can be done about them?
Housing prices per square foot increased by 1022% from 1979 to 2011 (see Figure). This is equal to an annualized growth rate of 7.6%. Over the same period the tender prices per square foot for building works conducted in the public and private sectors respectively increased by 316% (based on the Architectural Services Department Index) and 250% (based on the Rider Levett Bucknall Index). This is equivalent to annualized growth rates of 4.5% and 3.9%. The tender price indexes are approximations for the expected construction prices and include labor, materials and management expenses as well as anticipated profits at the time of tendering. The average gap between housing and construction prices during this 32-year period was 147%.
Demand Outstrips Supply
Housing prices are of course determined by both demand and supply. High housing prices reflect high consumer demand for housing. However, they also reflect some sort of restriction on the supply of housing. The very large gap between housing and construction prices leads one to naturally conclude that this must be due to restrictions on the supply of land. The traditional answer for why housing prices are so high in Hong Kong is that land in Hong Kong is intrinsically expensive, especially in some areas. According to that view, there is a great deal of demand and land, by its very nature, is limited in supply; therefore, housing prices must rise. The issue here is whether this large observed gap between housing and construction prices can be interpreted as reflecting land values.
The traditional line of thinking is that the high cost of land is a result of the so-called high land price policy of government. But there is another view, namely the cost has been driven by government regulation in the form of planning, zoning, and other restrictions on building. According to this view, properties are expensive because of artificial limits on construction created by the regulation of new properties.
There is no doubt that property values are relatively high on Hong Kong Island compared with those in the New Territories, and this reflects at least partially a strong demand to live in high-amenity areas. However, a casual walk through some parts of the island suggests that there is still available land in high-cost areas and new construction might be able to reduce the gap between housing and construction prices.
I can think of many plots of land occupied by the police force as quarters and training facilities or of reservoirs whose meager capacities are of little use as protection against water shortages. These facilities have long outlived their uses. No doubt the members of our respected police force and our many environmental groups have plenty of ready answers to show how mistaken I must be to fail to appreciate the intrinsic value of preserving these things. Nevertheless, it is still worth pointing out that these barriers to building create a potentially massive wedge between housing and construction prices.
That gap is a combination of two things: the land value and what can be called the regulation tax. The regulation tax includes at the very minimum the cost of assembling land, acquiring planning and building permission, and overcoming political resistance and bureaucratic inertia.
Housing prices (P) and construction prices (C) can be found by observing the data. However, it is not possible to directly observe the land price (L) and the regulation tax (R) in the data. For this reason L and R cannot be easily and separately identified. Indeed it has been common to view the sum of L + R, which is the difference in P – C, as the land value.
So which is the more significant driver of the wedge between housing and construction prices: land value or the regulation tax?
If the driving force is land value, then houses with bigger lots should be much more expensive than similar houses on smaller lots. If you double the lot size, you should double the gap between the property price and construction price since L measures a land’s use value as a physical lot. L is therefore land value measured at the “extensive margin”.
Land Value and Regulation Tax
But if the regulation tax is driving the wedge, then the gap should be quite constant for houses on various-sized lots. That is, the lot’s ability to accommodate a house in accordance with land-use regulations produces the lot’s value. The gap will therefore be wider if the regulation tax is higher. The value of L + R is a measure of land’s use value at the “intensive margin”.
Professor Edward Glaeser of Harvard University used single family homes on an individual lot in the US to test the significance of these two factors (the “intensive margin” and the “extensive margin”) as drivers of the wedge. He devised two different measures of land values.
First, Glaeser compared two properties on lots of different sizes in the same neighborhood, but with similarly built houses on them. By taking the difference in the property values of these two properties he would get a measure of land value (L) on the “extensive margin” on the assumption that the regulation tax on these two properties should be quite similar. The land value at the “extensive margin” is therefore the value of physical land. The difference between the two property values is:
P2 – P1 = (C2 + L2 + R2) – (C1 + L1 + R1)
Assuming C2 = C1 and R2 = R1 allows Glaeser to impute the land value difference on the “extensive margin” as:
L2 – L1 = P2 – P1
Second, Glaeser estimated land values in the traditional way as the gap between housing and construction prices or as L + R = P – C. This is the estimate of land value at the “intensive margin”, which includes the unobserved regulation tax. The estimated difference between land values estimated on the “intensive margin” and “extensive margin” provides a good gauge of the relative importance of land values versus regulation tax as the driver of the gap between housing and construction prices.
Glaeser found that estimated land values at the “intensive margin” were of the order of 20 times higher than land values at the “extensive margin” in major cities like Boston, Chicago, New York and Los Angeles. In San Francisco they were around 10 times higher. His results clearly showed that in the US, regulation tax was by far the most important driver of higher housing prices. He concluded that the US was facing a nationwide affordable housing crisis not because of either high construction prices or high land values (at the “extensive margin”), but rather, because of the high cost imposed by zoning and other land use controls.
Failure to Convert Industrial Land Use
It is useful to draw on Glaeser’s insight to examine why Hong Kong housing prices are so high. In an urban high-rise setting, if high land values were the key driver of the wedge then it would make little sense to build 30 floors on a one million square foot lot and 60 floors on a half million square foot lot to develop 60 million square feet of usable space efficiently. Indeed it would only make sense to build 40 floors on both lots to yield the same total completed floor space. The implication is that in a free market, housing structures completed in the same year or when the same market conditions prevailed would all have roughly similar plot ratios in order to maximize total land values.
As the urban economy prospered over time, higher demand for property would cause plot ratios to be raised to meet the rising demand. As a consequence, the gap between housing and construction prices would not widen significantly if the regulation tax were small and stayed small.
But if the regulation tax were high, then in an urban high rise setting it is entirely possible for different plot ratios to appear on different lots since the role of land value becomes much less important. The final plot ratio on each development would depend on what level of regulation tax happened to apply in each case.
From the Figure we find that the gap between housing and construction prices in Hong Kong was very narrow until the late 1980s. After China’s economy began to open up from 1979, manufacturers from Hong Kong gradually relocated their operations across the border until there was a considerable scale of operations there by the late 1980s. This coincided with a rapid rise in re-exports. Many manufacturers and their employees were making fortunes and purchasing properties in Hong Kong. Housing prices rose dramatically because supply could not respond in the immediate short term. As a result the gap between housing and construction prices started to widen.
It is useful to divide the post-1979 period into two: one from 1979.Q4 to 1988.Q4, the other from the end of 1988.Q4 to 2011.Q3. The average gap between housing and construction prices in the private sector during the earlier period was only 98% higher, but the corresponding figure in the later period was 165%.
The intriguing question is why didn’t the average gap in the later period narrow over time and return to its earlier level? Even during the deflationary period from 1997 to 2003, immediately after the onset of the Asian Financial Crisis, the average gap between housing and construction prices was 156%, representing a very modest reduction in the gap compared to the levels of the earlier period. Supply had failed to adjust.
I believe two factors altered the housing supply environment. The first was the failure of government to efficiently rezone industrial land after manufacturers had left. More bold and sweeping measures to facilitate the conversion from industrial land use to commercial and residential uses would have supported more rapid redevelopment. Instead there was case by case negotiation of conversion premiums, a cumbersome process that failed to address the massive need to redeploy land resources into better alternative uses in the aftermath of rapid structural changes that were taking place in Hong Kong and on the Mainland.
It would have been better for government to announce rule-based premiums on all properties based on updated rating values. The land conversion premiums for many properties would then have become public information. Moreover, an additional step to allow these premiums to be paid not as a lump sum at the outset but as a new component of land rent to be collected in perpetuity, could have been explored. The recent development strategy for East Kowloon is a limited step in this direction.
The second factor was the approval of the Town Planning (Amendment) Bill on 7 July 2004 to replace the old pro-development Town Planning Ordinance, which had been in force since 1939 and fostered an environment that lowered the transaction costs for property development. According to the government, the main objectives of the amendment bill were to streamline the plan-making process and planning approval procedures, enhance the openness and transparency of the planning system and strengthen planning enforcement control in the rural New Territories. What it achieved was to alter the rules of the development game.
The much hailed improvements encouraged more open consultation with all stakeholders on a much wider range of planning and development concerns. This was seen as opening up the system to greater accountability, but in fact it reduced the supply of housing and other types of property by delaying development and creating a permanent increase in the regulation tax, which raised land values at the “intensive margin”. The enormous gap between housing and construction prices resulting from China’s opening therefore could not be closed, and housing prices, as well as other property prices, continued to escalate.
New Planning Ordinance Benefits Existing Owners
The old Town Planning Ordinance was pro-development and this is why the gap between housing and construction prices remained narrow until the late 1980s. It fostered an environment that lowered the transactions costs for property development. The new ordinance raised transactions costs.
It has often been claimed that the new ordinance is an improvement over the old one because the rights of property owners are better protected. For example, under the old ordinance if I owned an apartment with a sea view, I had very little protection against a new development that would completely block my sea view. The new ordinance allows me to raise objections, ensures a hearing, and offers a higher probability to stop such a development. But this is a false argument.
If I knew that under the old ordinance I had hardly any chance to stop an offensive development that would block my apartment’s sea view, then I would not have paid so much for that apartment in the first place. It is reasonable to assume that I have therefore already been appropriately compensated for the probable loss of the sea view. The new ordinance gives me additional rights that I had not paid for under the old rules of the game and I have experienced a windfall gain as a result. But society now bears the cost of development delays. What the new ordinance accomplished was to benefit existing property owners at the expense of prospective owners, and society is made worse off. It is a form of protectionism for the haves.
Almost everyone in Hong Kong believes the government deliberately reduced the supply of land in order to raise its price so as to maximize government revenues from land sales. This is the so-called high land price policy. To an economist, it is not entirely obvious why total government revenues would necessarily be maximized by such a policy. Government revenues depend not only on land sales, but also on the overall prosperity of the economy. How could government revenues increase if businesses and employees were driven out by higher land values? At the very least one should investigate the tradeoff between revenues gained from land sales and revenues lost from a less prosperous economy. But perhaps the real culprit is not a high land price policy, but a rigid land planning and control system.
High Property Prices Divides Society
Another view that has become popular recently is the alleged rise of “property hegemony”. The idea is based on the observation that the property development industry has become more highly concentrated. It is alleged that the industry is now dominated by a few firms that wield enormous influence over government and that Hong Kong now has a classic case of government-business cronyism that conspires against the interests of the public. Whether this is indeed the case has to be settled by careful research and not popular assertion.
But two points are obvious. First, a rigid, high-transactions-cost planning and control system surely works in the interests of all existing property owners and not just property developers. The rise in land values at the “intensive margin” helps property owners, who are over one-third of households in Hong Kong, but at the expense of prospective property owners. Changing the present planning and control system will not be easy because there are many stakeholders in the system. Second, property developers who have a competitive advantage in a high transactions cost environment would thrive better under the new rules. Many former property developers have therefore ceased to be active in Hong Kong in the last two decades.
Housing has become increasingly expensive in Hong Kong in the past 20 years. The initial reason was the surge in local demand fuelled by Hong Kong’s growing prosperity following China’s opening. Subsequently the approval of a new Town Planning Ordinance increased the regulation tax permanently. Land value at the “intensive margin” continued to rise as supply failed to catch up because of various planning and control rigidities. The concentration of the property development market is also the result of these rigidities rather than its cause.
Addressing these issues entail tackling bureaucratic inertia, divided interests, and opposing views in society over development. Changing the present planning and control system will not be easy because there are so many stakeholders. If Hong Kong has to continue to live with high property prices then the demand for more provision of public housing will not cease. And the division of society into haves and have-nots will also become permanent. For this reason, the provision of public housing has to confer bona fide property ownership to qualified occupants to avoid a divided society.
Edward L. Glaeser, Joseph Gyourko, and Raven Saks, “Why Have Housing Prices Gone Up?” NBER Working Paper No. 11129, February 2005