When Hong Kong’s resettlement housing program was introduced in the 1950s, it initially targeted a limited number of squatters but quickly mushroomed into a massive public rental housing program. At its peak it accommodated two-thirds of the population. Like all public sector programs it provided a uniform standardized product administered by one set of rules and regulations with limited flexibility, an approach intended to avoid criticisms of unfairness and corruption.
One important standard was the size of the accommodation. The initial standard was set with reference to the cramped conditions of the old private tenements. But once it was laid down, it became formally institutionalized. Changing the standard became always politically divisive, subject to criticisms and reservations from all quarters. Numerous protracted bureaucratic meetings were required to achieve consensus, leading to prolonged indecision. As a result, the standard of accommodation in the public rental housing sector changed very slowly and lagged far behind developments in the private market. The gap between the public and private sectors grew wider over time.
I found in the period 1982-1996 that the average public rental housing unit was about 60% the size of its equivalent in the private sector (see Wong (1997)). The updated figures for the median size of the various housing units are given in Table 1. There is evidence that the size of public rental housing units have grown over time relative to private housing units with the average size around 60% for the period 1980-2010. In my 1997 article, I proposed a new measure of the housing stock by counting each public housing units to be equivalent to 60% of a private housing unit based on the average size of the units called the standardized housing stock. Based on this measure, I demonstrated that the “housing shortage” situation between 1982 and 1996 was substantially worse if we compared the number of households with the standardized housing stock. The shortage would create pressure for housing prices in the private sector to rise because there was unsatisfied demand for housing accommodation.
Table 1: Median Size of Existing Stock and New Supply of Housing Units by Type 1980-2010
| Year | Existing Stock of Public Rental Housing Units |
Existing Stock of HOS Units | Existing Stock of Private Housing Units |
| 1980 |
23.1 |
51.3 |
53.9 |
| 1985 |
24.5 |
51.8 |
46.1 |
| 1990 |
28.3 |
52.4 |
47.1 |
| 1995 |
29.6 |
53.7 |
48.0 |
| 2000 |
31.9 |
54.0 |
49.5 |
| 2005 |
34.0 |
– |
50.4 |
| 2010 |
33.8 |
– |
51.1 |
Sources: Wong
(1997), Housing Authority, Rating and Valuation Department.
It is important to note that the nature of the “housing shortage” in this case is not a shortage of housing units compared with the number of households. Rather, it comes from the small size of the housing units in the public rental housing sector. If tenants in that sector are all very poor, then this housing shortage would not be present because they would not be demanding larger private sector units. They would be willing to settle for the very cheap and small units in the public sector.
But if some significant fraction of these households is not so poor and can afford private housing units, then there will be pressure for private property prices to rise. This was the situation we observed in the 1980s and 1990s before the Asian Financial Crisis.
The Alarming Consequences
Therefore, a crucial question is whether most households in the public housing sector are on the whole poorer than those in the private sector. In other words, do the distributions of household income among public housing occupants and private housing occupants overlap significantly or minimally?
A well known result from the economics of housing demand is that households with higher incomes prefer or demand bigger homes and the size of accommodation is expected to be positively related with income. Given the 40% difference in the median size of most of the housing units between these two sectors, an efficient or optimal housing arrangement would require very different income levels between the members of these two sectors.
Research undertaken by myself with Chinese University’s Professor Liu Pak-Wai and by Dr. Yan Wai-Hin (see Wong and Liu 1988 and Yan 2000) had found that the distributions of household income between public and private housing tenants and between public and private homeowners overlapped significantly. Table 2 presents updated figures on these distributions in the period 1976-2006. In Table 2 all the households in the population are divided into 10 deciles, from the top 10% to the bottom 10%. We then assign households in the 4 categories of public tenants, private tenants, public homeowners, and private homeowners to each of these 10 deciles and determine what percentage of their respective categories are in each of these household income deciles.
Table 2: Percentage of Public and Private Tenants and Homeowners by Income Deciles 1976-2006
| Public Tenants |
Private Tenants |
Public Homeowners |
Private Homeowners |
|||||
|
1976 |
2006 |
1976 |
2006 |
1981 |
2006 |
1981 |
2006 |
|
| Bottom Decile |
8.4 |
13.1 |
11.1 |
5.7 |
0.8 |
7.9 |
7.7 |
10.0 |
| 2nd Decile |
9.0 |
17.4 |
11.8 |
9.5 |
1.9 |
5.7 |
7.1 |
5.5 |
| 3rd Decile |
10.9 |
13.5 |
10.3 |
8.2 |
1.9 |
6.3 |
7.5 |
5.1 |
| 4th Decile |
12.2 |
16.0 |
10.9 |
11.6 |
3.0 |
10.6 |
8.7 |
8.4 |
| 5th Decile |
11.2 |
11.7 |
9.8 |
9.9 |
7.2 |
10.8 |
8.5 |
7.1 |
| 6th Decile |
9.6 |
9.4 |
8.1 |
8.7 |
9.6 |
13.1 |
8.8 |
5.7 |
| 7th Decile |
14.1 |
8.9 |
10.3 |
9.6 |
16.5 |
16.2 |
10.3 |
14.2 |
| 8th Decile |
11.4 |
5.6 |
8.8 |
9.3 |
17.3 |
12.9 |
11.0 |
11.7 |
| 9th Decile |
9.4 |
3.4 |
8.4 |
11.4 |
20.0 |
11.5 |
12.8 |
15.2 |
| Top Decile |
3.8 |
1.0 |
10.5 |
16.1 |
21.8 |
5.0 |
17.6 |
17.1 |
| Bottom 5 Deciles |
51.7 |
71.7 |
53.9 |
44.9 |
14.8 |
41.3 |
39.5 |
36.1 |
| Top 3 Deciles |
24.6 |
10.0 |
27.7 |
36.8 |
59.1 |
29.4 |
41.4 |
44.0 |
Source:
Estimated from population census data (various years)
Note: The income deciles are different from year to year, but they are respectively applied to all four types of housing tenures – public renters, private renters, public homeowners and private homeowners – in any given year. The income deciles are constructed from the sample of all households by distributing an equal number of households, i.e., 10% of the households into each decile.
The results are alarming indeed. In 1976, 51.3% of the public tenants are in the bottom 5 deciles. For private tenants the proportion is 53.9%. Since 50% of all households must be in the bottom 5 deciles, our findings clearly show that the bottom 50% of the household income distribution is filled by both public tenants and private tenants almost equally. The distribution of household income between public and private tenants in 1976 was essentially identical. There is a complete overlap of the two distributions. The overlap is also considerable even if we compare the percentage of households decile by decile from bottom to top.
The public rental housing program clearly fails totally as a program for achieving equity in housing consumption. In 1976, the incomes of the wealthy half of the public tenants were equal to the wealth half of the private tenants. The bulk of the wealthy half of the public tenants was living in housing units that were about 60% of the median size of private housing units. Obviously these public tenants were living in housing units that were too small relative to what they could afford in the private market. Those that stayed accepted their housing condition only because of the exceptionally cheap rent. It is also likely that many had private housing units but did not return their public unit.
In 2006, 72.5% of the public rental tenants are in the bottom 5 deciles. For private rental tenants the proportion is 45.4%. Although the inequities of the program were reduced, nevertheless the overlap of household income distribution of public and private tenants remains enormous. Table 2 shows that 27.5% of the public tenants are found in the highest 5 deciles of the income distribution, however, 45.4% of the private tenants are still found in the lowest 5 deciles of the income distribution.
Compared with 1976, the proportion of lower income households has clearly grown considerably among public rental tenants. This is mainly a result of the increase of low income households admitted into the public rental housing program some of which became available through the recovery of public rental housing units from well-off tenants in the program. Ever since the 1980s the Housing Authority has attempted to recover public housing units from the well-off tenants through a “double rent policy” and preferential treatment for Green Form applications for Homeownership Scheme (HOS) units.
The Housing Leviathan
But progress on this front has been painfully slow and I should add this is to be expected. Administrative measures have limited effects unless they are draconian ones, but these are unlikely to be adopted when the objective is not to drive out the tenants, but to entice them to leave with an HOS unit as the carrot. Most of the public rental units that were recovered were surrendered after a successful application for HOS units as a Green Former. Since 1978, a total of 184,000 HOS units have been awarded to existing public housing tenants. In the same period successful White Form applicants for HOS units were only 146,000.
Expanding the Housing Authority’s reach to build new HOS units became the modus operandi for recovering public rental units from the well-off tenants. To recover one public rental unit a new HOS unit had to be built. The Housing Authority’s land portfolio grew to house more public rental and HOS units. Since these units had very low circulation, therefore, large tracts of valuable land were put into inefficient and inequitable use.
In 1986, 44.7% of the public homeowners were in the top 3 deciles. For private homeowners the proportion was 41.5%. Again one finds an enormous overlap in the distribution of household income between public and private sector homeowners. The proportion of public homeowners that were in the top 3 deciles of the household income distribution was even slightly greater than private homeowners. In 2006, the proportion of public homeowners in the top 3 deciles declined to 29.9%. For private homeowners the proportion was 44.4%. There was less overlap but it remained significant.
The public rental housing program benefits many households, but not all of them are the less well-off in society and certainly many of the poorest do not get the benefits, including low income households who are private homeowners who do not qualify to apply for public sector housing. The failure to target housing benefits to the poorest in society is not surprising. The resettlement program was aimed at re-housing squatters and was not means tested in the beginning. Early squatters were unlikely to be the poorest members of society since they had paid market rents in squatter areas, where housing units were more spacious than the old private tenements. Later, squatters were primarily those evicted when old private tenements were torn down or those who took advantage of the resettlement policy and turned themselves into squatters by exiting from old private tenements. There was no presumption that they would be the least well-off in society.
The public housing program began to acquire a reputation of helping the poor only after over a million squatters had been resettled. But by then the main bulk of the public housing population had been admitted without a means-test. The government’s reputation as a provider of housing for less well-off households should be viewed with some very serious reservations.
A means-test was eventually introduced for those who were not squatters, but this was administered only at the time of entry. A well-known trick to lower measured household income was to have one’s working spouse temporarily dropout of the labor force at the time of assessment. Tenants did not have to be continually means assessed as a condition for remaining in the program after their admission. Over time many tenants became well-off.
The Hidden Truth
It was only in the 1980s when land availability became tight again that there was an attempt to reintroduce means-tests for existing tenants. A double-rent policy was imposed on well-off tenants 10 years after they were admitted into the program. This has had only a limited impact on recovering units from existing tenants, although it has made tenants who did not wish to be means tested pay higher rents.
The inequities of the program are well known to all those who have been involved with it. That the program has gained a widely and firmly held reputation for being Hong Kong’s most successful benefits redistribution policy is a truly inexplicable myth. This mistaken view is even popularly held among broad segments of academia is of course another puzzle that probably deserves a separate essay someday.
What is obvious now is that many registered public housing tenants are not the less well-off. In the 1980s and 1990s they were in fact an important source of rising property prices in the private sector. Many were buying another unit in the private sector to expand their living space or as an investment. The Housing Authority estimated in the early 1990s about 13 percent of public rental housing tenants, or about 74,000 households, owned private domestic properties. Their flats accounted for as much as 11 percent of all private housing units; 84 per cent of these units were owner occupied, with the remaining 16 percent being rented out.
What was happening was that some members of the household registered as living in the public rental housing unit, were in fact residing in the household’s private flat. They were the sons, daughters, or parents of the heads of these households. In extreme cases, all members of the household resided in their own private flats, leaving the public rental units vacant.
This trend of private home ownership by public rental housing tenants was on the rise. A random sampling of 2,000 transactions between October 1992 and March 1993 provided by the Rating and Valuations Department showed as many as 24% of private housing units were purchased by public rental housing tenants, suggesting considerable absenteeism in public rental housing. A survey taken by the Housing Department in August 1992 showed as many as 18% of public rental tenant households had some registered members not currently living in the flat.
Even when the data show a surplus of housing units over the number of households, there can still be a housing shortage because the size of the public rental units is too small for a considerable fraction of the households living there. As long as this situation continues, we will have a two-sided problem. The well-off public housing tenant will buy property in the private market, but they will not relinquish their public rental unit, thereby withholding these units from less well-off households and individuals in society. Property prices in the private sector will continue to rise but old public rental units will not be released because there is no market for them. This state of affairs is one of the terrible legacies of MacLehose’s Ten-Year Housing Program (October 1972) and Long Term Housing Strategy (April 1987).
As mentioned, many households living in public housing estates are well-enough off to have alternative accommodation in the private sector. The longer those households have remained in the public housing sector, the greater the chance they are well-off. And they are probably more represented in the older estates, which are closer to the urban areas.
A simple solution is to revive the halted Tenant Purchase Scheme (TPS). A non-trivial and substantial share of these units would become available for rent in the open market provided this was allowed by the Housing Authority. It would meet the needs of those who are currently occupying sub-divided rooms and are on the Waiting List for public rental housing. Moreover, competition from the TPS would help to make these sub-divided rooms more affordable.
References
YCR Wong and PW Liu, “The Distribution of Benefits Among Public Housing Tenants in Hong Kong and Related Policy Issues”, Journal of Urban Economics, Vol. 23, No. 1, January 1988, pp. 1-20.
YCR Wong, “How Severe is the Housing Shortage in Hong Kong?” HKCER Letters, Vol. 42, January, 1997
Wai-Hin Yan, “Efficiency in the Distribution of Hong Kong Public Housing Resources (70s’-90s’), Ph D Dissertation, School of Economics and Finance, The University of Hong Kong, 2000.
