Scottish Government

Determined Differences: Rent Structures in Scottish Social Housing

Alison More, Jeanette Findlay, Kenneth Gibb, Diana Kasparova and Carl Mills,
Department of Urban Studies, University of Glasgow

This document is also available in pdf format (108k)

Rent Setting Systems

The local authority survey established that almost half used some form of comparability system to apportion rental income across their stock (see Figure 1). However, 27 per cent still based rents on Gross Annual Values (GAVs) and the remaining quarter operated 'ad-hoc' systems comprising a mix of rent schemes inherited at reorganisation in 1996. All RSLs, in contrast, had developed comparability systems, most of which were based on points.

Figure 1 Rent setting systems

Seventy per cent of council rent schemes had been in place for at least 20 years with few changes made since introduction. RSL schemes were newer (reflecting the introduction of assured tenancies in 1989) and were reviewed more frequently. Three in 4 RSLs also managed Fair Rent and/or stock transfer rent regimes.

Forty six per cent of landlords had a rent scheme review underway or under consideration, because of anomalies or inequities in rent charges. Some problems related to the amalgamation of previously separate rent schemes but GAV systems had become difficult to maintain and some other systems simply suffered from neglect, with differentials eroded through years of flat rate rent increases or other policy decisions. Over a quarter of councils had abandoned earlier attempts to restructure rents.

Rent Scheme Attributes

The specification of comparability rent schemes varied significantly in terms of the number and type of property attributes used to differentiate rents. Property size and type were incorporated in almost all schemes and usually had the greatest effect in determining rents. Over a third of council comparability and GAV schemes made no adjustment to rents to reflect changes in the physical standard of properties, creating some inequity between tenants.

While GAVs originally reflected area amenity, the values set in the 1980s were often out of step with current perceptions. In comparability schemes, RSLs were more likely (53%) than councils (35%) to adjust rents for properties in more or less desirable areas. Some landlords held strong views that such a policy could exclude poorer households from better areas. In some RSLs location adjustments were only to reflect lower council rents. Also among RSLs, some rent schemes had added extra points to rents in order to secure development funding, a practice which introduced anomalies.

Rent differentials

The values attached to different attributes also varied between landlords, so that there was a wide range in rent differentials between different types or sizes of properties. On average, council rent structures were 'flatter' than those in RSLs, but there was considerable variation. For example, a 3-bedroom property cost just one per cent more than a one-bedroom in North Ayrshire, while there was a 50 per cent difference in Shetland.

As indicated in Figure 2, GAV schemes typically generated wider differentials than other council rent schemes. There was little difference between the schemes used by RSLs but on average rent differentials were, like GAV schemes, wider, sharing similarities with patterns in the private rental sector.

Figure 2 Rent differential ratios by property size

Explaining differences in rents

Comparison between council and RSL rents is complicated by differences in service charging policy. A minority of councils (28%) charged tenants directly for the costs of services (in mainstream housing) compared to 80 per cent of RSLs. Councils typically charge all or a proportion of service costs to the Housing Revenue Account (HRA) thus spreading the costs across all rents.

Rents also differ because of differences in stock composition. In April 2001, average property size in councils ranged from 3.44 apartments in East Dunbartonshire to 2.8 in Angus. Adjusting average rents to reflect property size led to some changes in the rents 'league table': East Dunbartonshire, for example, moved from 8th to 14th while Angus moved up from 28th to 19th, though those at the top and bottom of the table remained in the same position. Since RSLs tend to own smaller properties, this also increases their average rents on a per apartment basis.

In the case studies, comparisons were made between council and RSL rents in selected localities, allowing differences in property size, type and quality to be taken into account. While differences in rent could often be attributed to quality variations, there were some instances in which properties of better quality and in better locations had lower rents. There were also marked variations across council areas in the difference between rents for modernised council property and RSL new build.

The case studies also found that while private sector rents were consistently above social rent levels, average council rents across the country did not mirror market patterns in any consistent fashion. Rents for 2 apartment council properties, for example, ranged between 44% (in Edinburgh) and 64% (in North Lanarkshire) of market rent levels for equivalently sized properties. However, policy has not required any specific relationship between social and market rents. Therefore, differences between landlords and areas that cannot be explained by variation in stock composition and quality will reflect differences in running costs, the primary driver of rents in social housing, rather than market factors.

The research examined the scope for creating indices that could compare differences in rents and rent differentials. With appropriate consultation on modifying existing data collection it was concluded that indices could be constructed and maintained relatively easily.

The impact of rent variations

Despite the variation in rents and rent differentials between landlords, tenants and applicants were not generally aware of these, and were much more likely to be concerned about the location and type of property on offer. However, the insulating effect of the current Housing Benefit regime has a major impact in blunting the effect of rent-based incentives for tenants to adjust their housing consumption. Those paying full rent were, however, more likely to expect high standards of service, though rent levels were rarely implicated in refusals of tenancy offers.

Conclusions and recommendations

The complex mosaic of rental patterns in the social rented sector begins to make sense once the nature of individual rent schemes is explored. Judged against either landlord objectives or market rent differentials, some rent schemes were more 'rational' than others. Some no longer generated 'fairness' in rents, having become distorted through the effects of flat rate rent increases, failure to review attribute values, or other policy decisions. This degradation, like the variation in rents and rent dispersion among more 'rational' rent schemes, is the inevitable outcome of permitting landlords discretion in determining the fairness of rents. Such discretion also complicates the requirement for rent polices to demonstrate 'comparabilty' with other social providers.

As well as identifying a need for several rent schemes to be reviewed and a greater role for regulators in monitoring rent policies, the research indicated scope for greater consistency in rent setting. The complexity of revising rent schemes tends to create inertia so that guidance may be necessary. However, a key question is whether local discretion should be retained or whether the key parameters of fairness need to be consistently defined.

Achieving equity (though not necessarily equality) in rents would not be achieved through rent scheme reform alone but would also require adjustment to landlords' average rent levels. A more detailed study of market rent differentials could be a useful input to policy development.

As illustrated by the prolonged process of rent structure revision in some councils, any policy change is likely to be slow and difficult and would be facilitated by a broad consensus on the key objectives of change. At the same time, the prospect of Housing Benefit reform, already initiated in the private rental sector, must be taken into account.

Research Methods

The research was carried out in two stages between November 2001 and September 2002. It involved: