Social housing construction vs. housing allowances:
from antagonism to integration

In their article "Sociale woningbouw versus huursubisidies" Prof. E. Buyst and A. Soete show a strong preference for public investment in new social housing over housing allowances.

As I differ somewhat from this position, I shall comment briefly on their evaluation of both housing policy instruments and then present some suggestions of my own.

ad A) Sociale woningbouw

1) Housing construction produces tax revenue and social security contributions and part of public investment, therefore, is returned to the government.

However, the revenue benefits mostly the federal government whereas investments in housing are a regional expenditure.

2) In the construction sector idle capacity and increasing unemployment is expected for the near future.


No doubt, under the present tax systems a certain part of the production costs represent taxes and levies on materials and labour. However, this is true for all kinds of housing construction, both public and private, both subsidised and non-subsidised.

The gap between tax revenues and public expenditures at the different levels of government is due to the given constitutional structure. Similar situations occur in other countries and are attended to by tax revenue sharing agreements.

Employment in a particular sector of the economy depends on demand and productivity. If demand slackens, output will be reduced over time and excess labour will be shed, or construction companies will close down. Short-time variations of demand may, though, be counteracted by public investment trying to stabilise the construction sector. Alas, these counter-cyclical measures often come too late.

In the long run, all production sectors have to adjust structurally to arrive at an adequate size. No sector should be kept artificially inflated by public funds which are anyway too limited for essential tasks.

Housing output should be adequate to cover demand (as manifest in

the housing market) and to satisfy needs (as defined by public housing policy).

3) (Large scale) public housing projects offer the possibility of lower unit cost.

This is basically true. However, experience with large housing estates has shown that (too) large projects exploiting these economies of scale may have negative social consequences which by far outweigh this advantage. Even in public housing the international tendency is towards smaller projects, infill construction etc.

4) An active housing construction policy is able to make use of the latest technology.

Yes, but the implementation of technological innovations deemed beneficial would be more effective if they became part of the Building Code and would hence be mandatory for all new construction, and would not just be applied de facto in new public housing.

5) The aims of structural planning could become reality by an active policy of new construction.

Yes, but the same is true for all new construction if the authorities enforce established planning regulations strictly.

ad B) Huursubsidies

1) Housing allowance is a static, housing stock oriented instrument.

Private landlords would have little incentive to improve the quality of their rental dwellings.

The necessary controls for determining the individually correct amounts would be cumbersome and expensive.

The character of housing allowances varies according to its regulation. Households entitled to housing allowance may be tenants of public or para-public dwellings only, or of private rental dwellings as well. In some countries even owner-occupiers benefit from housing allowances.

Even as a demand-side subsidy, housing allowances will affect new housing production if explicitly combined. E.g. the Austrian housing allowance system established in 1972 is integrated into the financing of new subsidised housing in order to reduce the otherwise unaffordably high rent levels.

Private landlords do tend to absorb a (large) portion of housing allowances. This is an aspect of all subsidies. Even supply-side subsidies are frequently diverted in the land, finance and construction markets. For a system which would reduce this negative effect see the proposals below.

The undeniable cost of administering a housing allowance scheme is part of the system and has not prevented its introduction in a large number of countries.

As a matter of fact, Flanders already provides some sort of housing allowance by adjusting housing company rents to tenant incomes, which in part results in below-cost rents and requires similar controls.

2) There would hardly be any economies of scale.

See comment on A 3).

3) Housing allowances would not generate any tax revenue nor employment.

See comment on A 1).

4) Excess capacity in the construction sector would become manifest.

See comment on A 2).

5) The effects on physical planning are doubtful.

It need not be the aim of housing allowances to assure the implementation of particular planning goals.

General observations

Strictly speaking, there is no "social" housing.

A house consists of walls, a roof, doors, plumbing etc. All these elements are physical objects which - in an adequate combination - provide "housing" as a bundle of services to be consumed by the occupant household. None of this has any "social" character.

"Social" usually means societal or something benefiting society or - commonly in housing policy - something in favour of population strata on a low socio-economic level. Therefore, the "social" aspect of some forms of housing consists of public subsidies covering part of capital, finance or operating costs through grants, low-interest loans, operating subsidies etc.

The principal aim of social housing policy is to provide low-income households with adequate dwellings at affordable expenses. Theoretically, this can be achieved both by subsidising supply and by providing income transfers.

As an example we might consider the production cost of a typical dwelling at BEF 3,000,000 and, therefore, the monthly cost rent, somewhere near BEF 15,000. If, now, a first-quartile household cannot pay more than e.g. BEF 9,000 per month, the resulting gap could be bridged either by a capital grant of BEF 1.200,000 (supply-side subsidy) or by a housing allowance of BEF 6,000 (demand-side subsidy).

However, and this is the main drawback of supply-side subsidies, it is extremely difficult and often - for a variety of reasons - not even intended to assure that precisely those households benefit from subsidised dwellings which actually need them. Quite regularly, subsidised housing is occupied by better-off households which could well pay the full cost. As a consequence, considerable amounts of public funds (actually taxpayers' funds) are drained away by production sub-markets (see above) and by households able to cover their housing needs on the market.

This situation drives up tax rates. As there is a limit to the tax burden citizens are willing to carry, many low-income households are not attended to.

The social aspect of housing policy is best assured by individual housing allowances if regular income leaves adequate housing out of reach. However, if governments were to rely on this transfer instrument exclusively, private rents would go up and the positive effect of increased purchasing power would largely evaporate. Also, any capital or finance subsidies awarded to non-public investors in rental dwellings with the aim of assuring below-market rents will be gladly accepted, but their expenditure-saving effects tend to fade out.

Therefore, the decision for the optimum system is not a question of choosing either subsidised "social" housing construction or a general housing allowance scheme, but rather a double approach of assuring sufficient (public) housing and of providing necessary individual income transfers.

Proposals (1)

  1. Subsidised rental housing should only be promoted by public bodies (usually local authorities).

  2. The supply subsidy provided should not entail any loss of public capital. It should be restricted to indexed capital loans. This means, that no real interest is charged but the capital balance is regularly adjusted for inflation. For public rental dwellings these loans could have a lifetime of 50 years. Every year 2.0 % of provided capital would be repaid in real terms.

  3. The public rental stock should be operated strictly at long-term cost rents. This is important in order to avoid arbitrary political decision and also to let tenants be aware of the real cost of their dwellings.

  4. Actual housing production and estate operation should be outsourced as much as possible and at competitive market terms. As the essential public attribute only ownership - and therewith the right to determine rent conditions - should be retained.

  5. Local authorities should continually observe their housing markets and produce and operate a sufficient rental housing stock in order to prevent scarcities which would provoke rent increases on the private market.

  6. Housing allowances should be provided to tenants of local authority housing stock. The formula for determining its amounts would have to assure that sufficient income is left available for non-housing expenditure. Also, the formula should be designed in a way which leaves households an appreciable share of any additional income (from work) instead of confining them in a "poverty trap".

  7. Initially, housing allowances should be awarded to all households in the public rental sector according to size and income. Eventually, when the housing market reaches sufficient equilibrium, housing allowances could be extended to the private rental sector as well, but only at amounts corresponding to equivalent economic rents in the public rental stock.

  8. Additionally and following a philosophy of tenure neutrality, owner-occupation should be provided with finance support in a similar form. Every household should have a once-in-a-lifetime right to real-interest-free public finance to cover its basic (!) housing needs through the acquisition of owner-occupied dwellings. In this case annuities should be 4.0 % of capital so that the public loans would be fully amortised over 25 years.

  1. For details, see: Donner, Ch. 1995, Das Ende der Wohnbauförderung ...back to reference