Prepared for:
Market Housing Branch,
Ontario Ministry of Municipal Affairs and Housing
Prepared by:
Steve Pomeroy
Focus Consulting Inc.
July 2005
This paper was commissioned by the Ontario Ministry of Municipal Affairs and Housing and examines the area of operating assistance to narrow/fill the gap between tenant incomes and rents that is either attached to a rental unit (rent supplements); or, moves with the tenant (shelter allowances) – essentially demand-side approaches to address non-market housing need.
The experience in three countries has been examined:
- Australia , which has a national rental assistance program – Commonwealth Rental Assistance (CRA). This is a stand alone rental assistance program, separate from mainstream housing programs delivered at the State level, and is targeted specifically to private sector tenants;
- US implemented demand-side assistance in 1974 under a Section 8 (Certificates and Vouchers) – now consolidated in the Housing Choice Voucher Program. This is also a stand alone rental assistance program funded nationally through the Department of Housing and Urban Development (HUD), but administered locally by Public Housing Authorities; and
- UK has a national entitlement program, Housing Benefit, that is embedded in the social security system but designed specifically to provide housing assistance separate from (but complementary to) general income assistance.
In all three countries these rental assistance programs are targeted to deep need – all are means-tested programs with the benefit being withdrawn gradually (tapered) as income rises. In Canadian terms, all are targeted to deep core need, but do not necessarily remove households from core need.
Eligibility varies but is closely linked to eligibility for general income assistance:
- In Australia and the UK , the programs are a means-tested universal entitlement and thus broadly subscribed by low-income households;
- The US voucher program is rationed. Many more households are eligible than the number receiving assistance. Vouchers are allocated more or less on a chronological first come first served basis to households on public housing waiting lists, although there are a series of special allocations relating to expiring contracts, and allocations for tenants in public housing being redeveloped.
The level of assistance provided is determined by formula that take into account income and actual rent paid:
- In the UK the post assistance net rents range from zero to well below 30%, and involve deep subsidy;
- In Australia, there is no explicit affordability objective and roughly one-third of recipients still pay more than 30% even after CRA assistance (9% still exceed 50% of income), so in many cases the level of assistance is quite shallow,
- Former versions of the US voucher program sought to provide assistance to reduce shelter costs to 30%; however recent revisions permit actual costs to be as high as 40%, so households are not fully removed from need.
In Australia the CRA is exclusive from other social housing – public housing tenants are not eligible so assistance cannot be stacked. In the US and especially the UK there is stacking of rental assistance and supply assistance as a way to both reduce rents to more affordable levels and, in the UK , to provide the cash flows necessary to secure private financing of capital loans for new social housing development. Accordingly there is a symbiotic relationship between rental assistance and separate supply programs.
Expenditure levels for these programs are steadily increasing in all three countries. Efforts are ongoing to reform and modify the benefit structures or eligibility to help manage the expenditure trajectory. In both Australia and the UK , annual expenditures on rental assistance programs have overtaken and exceed annual expenditures for subsidies on existing or new supply.
Critical issues exist in all three countries in terms of horizontal equity (especially between households that reside in market versus assisted stock) and this is likely to increase as eligibility criteria continue to be tightened to contain expenditures.
There is also a very significant overlap with general income assistance systems, especially in Australia and the UK . This is an inevitable consequence of means-tested targeting – the client populations overlap.
There is extensive literature in all countries discussing the effects of income assistance on work incentive and disincentive. There is a general sense that rental assistance programs do little to alleviate these issues and in most cases reinforce disincentive structures (especially in the UK since Housing Benefit often covers 100% of rent and leaves tenant with no share of cost burden and related housing consumption decision).
In the Australian and UK approaches, individuals out of work (equivalent of EI in Canada ) become eligible for rental assistance and in the UK continue to receive Housing Benefit for a defined period after returning to work. Such approaches may be useful design features to explore more as a way to reform disincentives to seek work into incentives.
There is mixed evidence on the geographic distributional impact. US approaches in particular have explicitly sought to use voucher assistance in a strategy to disperse and reduce high poverty concentrations. However program criteria (e.g. on rent levels of eligible units, as well as the requirement for landlords to enter into contracts) tend to limit choice resulting in migration to fewer market areas.
All three countries (but to a more limited degree in Australia) suffer from low levels of new rental production with excess demand for lower rent units putting pressure on the lower rent stock, while redevelopment and conversion options squeeze out part of the existing supply. Inevitably demand pressure in the eligible rent stock pushes up program costs.
Where constrained supply exists (as it does in low-rent ranges in all countries) demand-side programs could be complemented by supply initiatives (potentially including stimulus to private development as well as assistance for social housing). Without sufficient new supply there will be an inevitable upward pressure on rents and program costs.
It is important to clearly define the overall objectives of housing policy and assistance. Is policy narrowly focused on specific housing outcomes, such as post assistance shelter to income ratio, or are there broader objectives, such as labour market attachment, deconcentration of poverty or neighbourhood regeneration effects?
Depending on objectives and target populations, rental assistance may be effective as part of an overall strategy, but may not fit for some specific target groups (e.g. long term dependents) or in tight markets. In particular, use of rental assistance in the context of transitional and temporary assistance, appropriately targeted to clients that have a higher probability of graduating on to self- reliance is an area to carefully explore.
Clear articulation of objectives is critical due to the symbiotic context in which rental assistance is, or can be, used. Inevitably it will be targeted to some degree to lower income housing and as such interfaces with more general income support/social security systems. It also interfaces with the labour market (incentive and disincentive structures) the housing market and health systems. Good policy needs to recognize and understand these interactions and to massage the respective sets of policy objectives and program designs (including eligibility criteria) to optimize synergies and minimize inadvertent negative consequences.