Supply-side subsidies

Supply-side subsidies create more distortions in housing markets as they can affect where housing is built, who it is provided by, and when it is produced. However, they can be an important boost to developers, particularly where local construction industries are not quite developed enough to supply a particular market segment profitably, and subsidies are integrated into a policy package to grow the local construction industry over time. Subsidised housing construction can be implemented in a variety of ways including direct government provision, capital grants, government-subsidised loans, or tax breaks for developers. However, research in this area highlights two key considerations:

  1. Subsidies alone do not solve underlying cost problems. Bringing housing costs down requires reducing the cost of housing’s three key inputs: land, infrastructure and property construction. One way to do this is through government subsidies. However, these are generally insufficient to plug affordability gaps alone; a more sustainable solution is to tackle the factors that make the costs of these inputs so prohibitively high in the first place.

Furthermore, when implemented in the context of restrictions (e.g. on land supply) that create housing supply inelasticity, supply-side subsidies cannot increase the total housing stock, but rather serve to crowd out unsubsidised housing with housing built by subsidised developers (see technical annex).

  1. Transparent calculation of subsidy levels enables effective policy. Such calculations need to include hidden costs, such as the opportunity cost of lost tax revenues, and be expressed in terms of present values, as in the case of subsidies on future interest rates. Under-attention to the often complex calculations required for this has led to significant subsidy cost underestimates 1.

Where governments do subsidise developers, often in return for the construction of a certain level of housing at affordable prices, effective provision is aided by:

  1. Incentivising developers to build in well-located areas. This can be aided by ensuring that restrictions on infrastructure and property quality that accompany subsidies do not force developers to cut costs on land. South Africa’s housing subsidy programme for example, by fixing ceiling costs, minimum floor areas and land-use standards, establishes land as the only variable which developers have to reduce costs to make housing more affordable. Developers can be further incentivised to build in well-located areas by removing the constraints to often well-located and under-utilised public land being used for housing development.
  2. Reducing monopoly power of developers. This can be aided by an open and competitive bidding process for construction subsidies. However, open and competitive initial bidding processes are insufficient to tackle monopoly power. Monopoly power is reestablished once the government contracts with a particular developer and incurs initial sunk costs related to negotiation or to construction. This power can then be used to renegotiate the contract at government expense. In the transport sector in Latin America, 78% of public-private partnership contracts have been negotiated within 3 years, and typically cause cost increases of up to 3-15% of the initial investment 2. Knowing this, many firms deliberately produce ‘low-ball’ bids for government contracts in order to later renegotiate.

Internationally, methods for mitigating the effects of costly renegotiations include the establishment of independent panels of experts to adjudicate proceedings; pre-emptively identifying and allocating risks in the initial contract; instituting a ‘freeze period’ of 3-5 years on contract renegotiations; and the preparation of alternative provision mechanisms to ensure the government can credibly allow projects to go bankrupt if they do not meet cost/revenue estimates 3.

  1. Effective risk sharing systems between public and private actors. When the public sector contracts out subsidised housing delivery, some of the uncertainties associated with costs and with revenues are transferred to the private sector. It is important not to shift all of these risks to the private sector, since they will demand a premium for taking on this risk, resulting in higher subsidies. Instead, it makes sense for contracts to allocate risk according to the ability of each actor to manage it. The private sector is often best placed to handle the risk of construction and operating cost overruns, except in the case of changes in construction regulations. The public sector is often better placed to cover some of the revenue risk the private sector faces in terms of consumer demand for housing.
  2. Targeting subsidies at alternative housing providers, particularly where formal-sector large-scale housing developers have little experience in providing housing at costs affordable to lower-income residents. Alternative providers could include semi-formal micro-developers who currently provide low-income housing, but have limited access to large-scale financial markets to scale up operations. In Porto Alegre, Brazil, the Social Urbanizer project has legalised and encouraged the activities of previously illegal developers and land subdividers by reforming land-use policies. This has enabled the delivery of fully-serviced land at US$25-28/m2 in contrast to US$42-57/m2 in the previously formal market 4.

Alternative housing providers could also include community-groups that finance home improvements. In Thailand’s 2003 Baan Mankong Housing Programme, for example, community collectives were encouraged to pool together resources to negotiate formal land purchases or land-leases from official owners. The government supported this by issuing grants, and subsidised loans to collectives at 4% interest rates not only for land acquisition, but also for housing improvements and infrastructure costs. This enabled the effective upgrading of existing stock and the addition of second and third floors. The share of housing made from durable materials increased from 66% in 2000 to 84% in 2010 5.