The goals and tasks of a long-term housing strategy at a time when the number of housing units is greater than the number of households is different from the situation when the opposite is true. In the immediate post-war period most immigrants arriving in Hong Kong had relatively undifferentiated housing needs. Their immediate goal was to find shelter – any shelter. When the only objective was a simple one of tackling a severe housing shortage, a plan and command approach fared reasonably well.
But by the mid-1980s the housing situation had completely changed. An increasingly prosperous community had diverse housing and property investment needs. When shortages occur today they are no longer general shortages, but shortages in a particular segment, for example, a defined type of housing unit, a particular type of housing tenure, a choice between investment and shelter purposes, or demand from a specific group of households. Under such a scenario, the plan and command approach became unsuitable and a market approach was needed. But combining the two approaches in a long-term housing strategy was never going to be easy.
Matching demand and supply today is no longer simply about matching the number of units to the number of households. It is also about meeting the preferences of different households for different types of units, recognizing why these preferences will change over people’s lifecycle, and understanding why they are influenced by market conditions such as business cycles. The demand of these households is far more dynamic than that of previous generations and requires a multi-dimensional dynamic market approach, not a plan and command approach.
An early but flawed attempt at merging the two approaches happened in the MacLehose era when the Housing Authority devised the Homeownership Scheme (HOS). It encountered a problem in that initial batches of the new HOS units in the early 1970s were sold to eligible buyers at a subsidized price set at 70% of the prevailing market level. The pricing policy embodied three features: (1) the price was set to be affordable to eligible households, (2) the subsidy rate was fixed, and (3) the unit was sold to the household with complete ownership (including the right to dispose of it on the open market 10 years after purchase).
This pricing model had to be subsequently abandoned when property prices in the market increased rapidly a few years later. The HOS units became unaffordable to eligible households. The HOS pricing policy had failed to appreciate the full implications of this possibility and an important lesson was learned. When markets demand change, property prices often change even faster leaving many households behind in the dust. More importantly demand changes cannot be foreseen by the planner. Even if the planner turns out to be a clairvoyant it is unlikely he will be able to persuade other stakeholders to change an approved plan.
It was apparent that the level of subsidy had to be varied to make the units affordable, but the authorities chose not to do this and sidestepped the issue. Instead they chose to deny the purchaser title to the full value of the land under the pretext that only the user’s right to the property was sold, not the right of free disposal. The latter right became conditional on repaying the remaining portion of the unpaid land premium.
The purchaser therefore had only a half title under his name. Unfortunately for him the value of the unpaid land premium appreciated over time at a rapidly rising rate as land values surged ahead of income growth. The government could easily have extended him a mortgage loan at the time of purchase to cap the value of the unpaid land premium, but for reasons I am unable to fathom they failed to do so.
This attempt to blend the housing strategy with the market had an unfortunate outcome. The required interface between the two was very limited and involved merely the proper pricing of publicly built homeownership units. And purchasers ended up with an awkward compromise of half title to the property; they became pseudo homeowners of housing units that have had no turnover and have been put to an inferior use that destroys value.
As a consequence, the HOS has become locked into the public housing sector rather than gradually integrating into the market, which would have enabled land values to appreciate to the benefit of the homeowner. And in future when the HOS buildings are run-down with age, the issue of redevelopment will become impossible to resolve without addressing the question of titles. The owners will have no right to redevelop the buildings without the government’s approval nor could they afford to do so on their own.
Complicating the housing scenario is the fact that, because of allocation decisions made in the past, many households in the public housing sector are well-off, but many households in the private sector are not. As a result the mix of large and small units between the public and private sectors is wrong, and certainly misallocated. Forecasting changes in demand and supply in such a market is theoretically and practically difficult, as it is in any highly regulated market, but errors in prediction can have profound economic and political consequences.
Supply Factors
Supply is affected by the straightforward production costs of supplying land, labor and materials for building and construction. The lack of productivity gain in the local construction industry and the shortage of skilled construction workers will drive up costs.
But more significant are the numerous regulatory-related transaction costs involved in property development, including navigating planning rules, complying with building codes, negotiating costs associated with land acquisition due to holdouts, and holding public consultations that often result in long delays. The uncertainty over how long these things will take makes development ever more costly. Large teams of advisors and professionals are now retained by developers and government to facilitate development. The expertise required to successfully manage these transaction costs forms a formidable barrier to entry for many less experienced prospective developers. Many have left the market.
I have estimated that since 1989 such regulatory uncertainties and delays have increased the gap between property prices and construction cost by about 67% on average. In my article (Why is Housing so Expensive? in the HKEJ 14 March 2012) I called these the regulatory costs of development.
The case of the old Kai Tak airport site is a vivid illustration of this. It has taken an inordinate amount of time to develop the site yet it is vacant land. Imagine how long it takes to develop sites that are not vacant. All the time delays add to the final cost of development. Reducing these transaction costs will make a huge difference to the success of any long-term housing strategy – and, indeed, have an impact on the high prices of non-residential properties, which are also affected by transaction costs. It may not be customary for a housing strategy to address the cost of property development, but in Hong Kong today this factor is a major determinant in the success of any housing proposals.
Many have complained that the scarcity of land in Hong Kong is why property prices are high. This is blatantly false. Consider Singapore where over 90% of the land has been developed; by contrast, less than 25% of the land in Hong Kong is developed. The real problem is that our planning rules and regulations have made most land unavailable for development or subject to long and uncertain delays that deter development.
There are many ways to lower the regulatory cost of development. The most radical approach is to remove zoning rules that differentiate land into agricultural, residential, industrial, and office and commercial categories. Town planning approval could then focus only on the intensity of development and narrowly specified incompatible uses of land. The volume of development each year or every 3 to 5 years could be controlled. This approach may sound iconoclastic but should be studied seriously. Quite a lot of property in Hong Kong is indeed zoned for mixed uses and it does not appear to be a problem in the service economy of today.
Another approach is to set and pre-announce each year the rates of compensation for land use conversion on every plot of land. These pre-announced rates would be totally transparent to all and be valid for a period of say 12 months. This would significantly lower the negotiation costs in settling land conversion premiums. It would also remove suspicion and concern over the opportunity for corruption in these negotiations.
Would government revenues suffer as a consequence? This is not obvious because the transaction process would be speeded up and this could lead to more negotiated successful outcomes. Moreover, even if government revenues suffered, society would benefit from less costly transactions, speedier development, and perhaps fewer opportunities for corruption. It is useful to appreciate that when government revenues suffer, society benefits from the transfer. In recent years government transfers of benefits have come under much criticism, but this is targeted at selective transfers that benefit the few and not the many. My approach benefits all equally and is therefore a non-discriminate transfer to the public.
Demand Factors
Demand is affected in the long-term by economic and population growth. The critical policy variables relevant for economic growth are investments in human capital (education and health care) and lowering the transaction costs of development. The relevant variables for population growth are immigration policy and drivers of household formation. Most of these variables can be estimated and some of them depend on what other policies will also be implemented and when they will take place.
A difficult issue in forecasting the demand for new housing units is that it depends on the pace of formation of new households. For example, children will leave the household as they grow up to form new households. But the pace at which this occurs depends on the available supply of new housing on the market and in the public housing sector. When supply is tight, this process will be delayed. In other words, estimating the growth in the number of new households is itself dependent on the demand and supply of housing units. This means one cannot use past trends (or even recent past trends) to make accurate forecasts of the future pace of new household formation. Untangling these interactions is not going to be trivial.
The formation of new households is worth special attention. It usually occurs when adult children leave the home. There are many considerations influencing whether and when this occurs. A particularly intriguing question is what happens to the original housing unit. The answer depends very much on whether it is a private or public housing unit.
If it is a public housing unit then the adult children move out and the parents stay behind in order to retain the unit. When the original unit is private then one can think of three other possible outcomes – the parents leave and the children stay, the children and the parents stay and both leave to find a new arrangement that may include them living together or living apart. Household formation is part of the lifecycle of households but while it leads to demand for different housing types and tenures in the private market, in the public housing sector it mandates only one solution.
Many young new purchasers of private housing units rely on their parents to make the initial down payment and parents may refinance their existing home to do so. This phenomenon is common in other societies as well. But those living in public housing do not have such a choice. As a consequence, a large portion of the population – the tenants and pseudo homeowners in the public housing sector – see limited prospects for homeownership for future generations. For many in urban societies, homeownership is an important channel to enhance the value of one’s savings and upward mobility.
Faced with fewer choices, young adults who live in public housing have chosen to become early applicants for their own public housing units. Table 1 shows that the number of non-elderly single person households on the Public Rental Housing Applicants’ Waiting List has grown from 35% in 2007 to 47% in 2012. Those aged below 30 have risen from 13% to 24% of applicants. As the table shows, most of them live with their parents and the majority has post-secondary or tertiary education.
Table 1: Characteristics of Non-Elderly One-Person Household Applicants on the Public Rental Housing Waiting List
|
2007 |
2008 |
2009 |
2010 |
2011 |
2012 |
|
| Non-Elderly Households on Waiting List |
35% |
35% |
38% |
40% |
42% |
47% |
| Below 30 |
13% |
13% |
14% |
16% |
19% |
24% |
| 30 or above |
22% |
22% |
24% |
24% |
23% |
23% |
| Not Living Alone |
67% |
61% |
71% |
77% |
86% |
85% |
| Living with Parents |
|
|
|
|
79% |
81% |
| Below 30 |
|
|
|
|
93% |
92% |
| 30 or above |
|
|
|
|
58% |
65% |
| Post-Secondary and Tertiary Education |
|
|
21% |
23% |
33% |
34% |
| Below 30 |
|
|
37% |
40% |
53% |
55% |
| 30 or above |
|
|
7% |
7% |
11% |
12% |
| Private Permanent Housing |
53% |
54% |
53% |
51% |
48% |
45% |
| Public Rental Housing |
28% |
28% |
31% |
27% |
28% |
29% |
| Subsidized Sale flats |
14% |
11% |
12% |
18% |
21% |
23% |
Source: Housing Authority, Memorandum for the Subsidized Housing Committee of the Hong Kong Housing Authority, Survey on Waiting List Applicants for Public Rental Housing 2012, 8 October 2012.
Meanwhile, the percentage of the non-elderly one person households on the Waiting List who live in private permanent housing has decreased from 53% in 2007 to 45% in 2012. A growing number of applicants are now residents in subsidized sale flats, increasing from 14% in 2007 to 23% in 2012.
A society with a large public housing sector reduces upward mobility of the next generation because the occupants are denied the opportunity to benefit from appreciating housing values available to those who are homeowners in the private sector. The long term consequence is to create a more polarized society. The next generation is therefore looking to the government for assistance. Their parents’ could have helped them if the sale of HOS units had not ended up in the awkward compromise of selling only half a title.
Demand and Market Equilibrium
The most difficult factors to forecast are the demand for housing from non-domestic sources and the business cycle. For an open city like Hong Kong, external factors inevitably play an important role in influencing housing demand.
For example, when property prices are high in Hong Kong, some of the domestic demand may spillover across the border to Shenzhen. Another source of additional housing demand in Hong Kong is cross border family reunion. For some people, regular commutes across the border have become a way of life and will be facilitated if regulatory barriers are lowered (for example, by lifting local residency requirements for receiving welfare payments or privatizing public housing).
This of course raises the question of whether a long-term housing strategy can be developed if Hong Kong is viewed as an isolated entity. There is a need for a broader vision of the future of the Special Administrative Region so that all related policies can be coordinated to form a coherent whole.
Similarly the availability of cheap credit in recent years has fueled property prices in Hong Kong, including demand coming from buyers on the Mainland. Forecasting overseas demand is difficult because it is driven primarily by business cycle factors. By its very nature foreign demand is likely to be primarily for investment rather than consumption. Forecasting such demand separately is simply not feasible and most probably cannot be differentiated from business cycle effects, such as the Asian Financial Crisis and the Financial Tsunami.
The most difficult thing with forecasting long-term changes in housing demand is that demand is subject to considerable short-term fluctuations, some of which may last a few years. Given that the supply of new housing units takes several years to come on stream, the housing planner faces a dilemma: if he heeds political calls to speed up supply and demand then reverses its course a few years later, society will be faced with a huge property market glut. What was supposed to be a counter-cyclical intention has turned into a pro-cyclical mistake! And society belatedly screams that the property market bubble has burst and goes after the skins of the politicians and bankers.
The fluctuations in demand and long lags in supply mean property prices in the market inevitably make large upward and downward swings. These periodic swings are necessary for demand and supply to equilibrate in the market; no long-term housing strategy can avoid them because this is the nature of the property market.
Many recent commentators have proposed that a large land bank should be established so that the supply of housing can be readily increased when demand rises and vice versa. This view fails to appreciate that it is difficult to know when demand will increase until after it has happened. Such calls are always made only after prices have started to increase. Even then most policy makers are still not certain whether the increase is temporary or the start of something permanent. They may consult economists and businessmen, but most likely these experts will advise the government to wait for more signs as they think it is too early to tell.
By the time everyone from Tuen Mun to Ap Lei Chau knows what has happened, so does the policy maker, but it is already too late to commit. The politician, however, has no choice and commits to path in order to curry favor with his constituents. If the market collapses in a few years’ time then he just re-enters the scene as a new white knight to save the fair maiden from another dark night.
Just how difficult it can be to forecast demand is evident in the failed housing plans proposed by various administrations. The plans of Youde, Akers-Jones, and Wilson met their targets but failed to build enough units to meet the surging demand. The Tung housing plans also met their targets but built too many units because demand collapsed. Tsang had no housing plans and no targets to meet so he is now blamed for doing nothing in the face of surging demand. This short history is not to apportion blame or praise. It merely observes that getting it right is a difficult job.
Political Reality
The government’s central role in providing land and housing means any long-term housing strategy requires not only good forecasts but also continuity in policies across administrations. This was more possible under British Colonial rule because we did not have political elections. Today the new political dynamics of electioneering are changing public perceptions of the long-term continuity of any policy.
The continuity of any long-term housing strategy that has to straddle administrations is less secure today. Policy uncertainty does not depend on the credibility of any administration or its determination to act; it reflects the underlying uncertainty of a political system that is still in transition. This adds an additional dimension to the market – policy uncertainty. Such uncertainty magnifies the existing market uncertainties.
The question is what kind of long-term housing strategy would be prudent given market fluctuations and the need for political approval to meet private housing demands? (To be continued.)
