Written evidence submitted by Haringey Defend Council Housing [RSH 101]

Regulation of Social Rent setting

for First Let properties

This contribution addresses the questions, ‘Is the current regime for regulating social housing fit for purpose?’ and, ‘Does the current regime allow tenants to effectively resolve issues?’

Based on original research, we can demonstrate that there is no effective regulation of the amount of First Let Social Rents (for new builds, newly acquired properties, and newly converted properties). 

Introduction and Context

Council tenants, housing association tenants and welfare campaigners have for many years valued and defended social rent as the Gold Standard of real affordability. As the Joseph Rowntree Foundation stated in its 2013 report, The links between housing and poverty: An evidence review, ‘Social housing is highly targeted on people with low incomes and has been shown to be the most ‘pro-poor’ and redistributive major aspect of the entire welfare state.’

Existing social rents are a hotchpotch of rents which were set over a long period of time by different housing providers, under many varying rent setting policies. There is a good deal that we know about the existing level of social rents, available in government and other statistics in the form of averages; but there is also a great deal that we do not know.

The question of the amount of social rent came to prominence after 2010 when government decided to move away from social rent to so-called affordable rents; and after 2016 when the Mayor of London introduced London Affordable Rent as a ‘social rent equivalent’ product (this was a claim which we strongly refuted); and then again after 2017 when partly in response to the Grenfell fire, plans for new Social Rent housebuilding came back onto the policy agenda.

The new housebuilding plans have not been supported by adequate grant rates for development, and this has led to sustained upward pressure on new build rent levels.

Some key elements of the policies of 2010-16 have now been pushed back.  The affordable housing programme has been opened up to bids for Social Rent; and the rental component of the affordable homes programme in London 2021-26 is based on Social Rent, rather than any form of affordable rent.  So our campaigning had an impact. However, the fallout from these battles has loosened the definition of just how high a Social Rent can be.  Councils sometimes set new Social Rents £50pw above the rents of their existing homes. We face a future of a two-tier system of council rents, with the danger that at some future point, under some future government, the existing really-affordable rents will be assimilated up to the new, higher levels. 

We now have the phenomenon of unaffordable social rent, because there can be a higher rent, plus high service charges (which have been increasingly disaggregated from rents and then charged in addition to rents, over the past 20 years).  So social rent no longer provides the safety net that it used to.  A high proportion of materially deprived children are in social renting households.  47% of food bank users are social rent tenants.  Welfare benefit restrictions also affect tenants who are being charged ‘higher’ Social Rents: the reach of the Household Benefit Cap is such that even two unemployed adults with two children in a two bed property paying council rent could be benefit-capped as a result of social landlords setting Social Rent at these higher levels.

It must be emphasised that the setting of social rents for existing properties has been increasingly controlled in detail by central government, through rent convergence, by the four years of rent reductions, and now by a formula based on annual increases of CPI plus one per cent.

It is the rents on new properties which concern us here. It is notable that there is NO substantive published evidence about the actual levels of what are called ‘First Let’ Social Rents: that is, the rents of new builds, newly acquired properties, and newly converted properties.

Social Rent setting

Social landlords may use intermediate or affordable rents, however their First Let Social Rents must be set according to the procedure set out in the government’s Policy statement on rents for social housing (MHCLG, February 2019)[1]. This uses a formula, originally dating from 2002, indexing all rents and costs to January 1999 values, with:

• 70% of the rent adjusted for relative county earnings in January 1999 (using amounts in the statement, p 20).
• 30% of the rent adjusted for the relative property value (compared to average housing association stock in January 1999). 
• A weighting for the number of bedrooms.
• The rent to be updated by annual increments, as specified by government since 2000/01; with annual rent adjustments at CPI plus 1%;
• The rent to be limited by the Social Rent Cap as published annually by government for each number of bedrooms.

Here is the highly prescriptive presentation of the Government’s policy statement on rent setting:

The weak point of the otherwise highly prescriptive formula is fact that the determination of the free market value of the property is open to question. Given that the house is not actually on the market, and no real transaction takes place, this is an assigned value, identified by the landlord alone. 

There is a fundamental inequity here. Because the valuation process is controlled by one party, the landlord, and is kept secret from the other interested party, the tenant (and also from board members, Councillors and the public), there is a tendency for rent outcomes to be too high. This was noted as early as 2004 in a government review of the rent convergence policy, when Housing Today reported, ‘Housing Corporation to scrutinise inflated rents: Some landlords will face regulatory checks on property valuations in wake of ODPM report’ (23 July 2004).[2]

As reported in the Social Housing website article in the article Rent and regulation: when registered providers get it wrong (06/08/2020), Richard Petty of real estate company JLL said that some registered providers were “a bit optimistic at the development appraisal stage in setting their target [social] rent term’s existing use value for the rent formula”.  He continued: “We have certainly seen one or two instances where the original development staff have been very optimistic in their calculations and something has been approved when it probably shouldn’t have been approved. They’ve ended up charging more than they should have done.” [3]

In real free market values based on a willing buyer and a willing seller, the professional valuations are disciplined by the price at which the parties actually agree to exchange. However in landlord valuations for social rent setting, this discipline is completely absent. The valuations must be carried out to Royal Institute of Chartered Surveyors (RICS) Red Book standards, but as yet we know little or nothing about the actual process by which they take place.

Researching Social Rent setting

Social Rent setting has been researched in two different ways:  by using government CORE data; and by using mathematics to identify the market values which are being used to set the rents.

CORE data on new lettings

Data on rents for new social sector tenancies is collected by front line housing workers at tenancy sign up under CORE (COntinuous REcording of the details of new tenancies). Specific data for First Let rents is collected by CORE but NOT published by government; however it is available using Freedom of Information requests.

CORE data shows that First let rents in London for both housing associations and councils closely followed relet rents until 2013, and then moved higher than them.[4] Please note that these higher rents are unrelated to faster house price increases in London, because the Policy Statement, paras 2.24 and 2.27, says that ‘the valuation must be made at January 1999 prices’.

A long series of FOI questions brought out the detail at borough level in London, showing many cases where average Social Rents were set above the Cap, which is supposed to be the maximum for all individual rents. For the four years 2016/17 to 2019/20, average First Let Social Rents were obtained for 352 separate  borough/year/tenure/bedroom-size categories. In 58 of these categories (relating in 48 cases to housing associations and in 10 cases to local authorities) the mean average rent was set above the Cap. These cases occurred in 17 out of the 33 local housing authorities in London.  Please see the example data sets for 2018/19, showing both London-wide figures and information on specific private providers in Wandsworth, with rents set above the Cap.[5]

There were very few instances where Social Rents seemed to have been Capped, as required by the government rent-setting guidance. Instead it seemed that many social landlords had disregarded the Cap completely.

The breaches of rent regulation are concentrated amongst private providers in a minority of boroughs. Most council and housing association lettings go through local authority choice based lettings systems. The substantial variation in housing association rent setting practices in different boroughs suggests that local authority officers play a significant role in influencing rent setting by housing associations.

It should be emphasised that there are a large number of cases where rent setting is carried out properly and fairly by social landlords; and a healthy number of cases where rents for new homes are no higher than those of the existing stock.

The use of mathematics to determine assigned market values

Because all of the elements which go into the rent formula (except one) are published by government, once we know the rent outcome we are presented with a simple algebra problem allowing us to determine the one unpublished element: the assigned market value. These calculations demonstrate that some of the rents being set using the Formula could only be achieved by assigning the market values at extraordinarily high and improbable levels.

For 2021/22 the Social Rent Cap (the maximum permissible Social Rent) is £157.62 pw for a two bed compared to the estimated average two bed rent of £109.10 pw for London’s existing Council Social Rent stock.  So the Cap is £48.52pw or 44% higher. Where Social Rent is set at the level of the Cap, as is increasingly the case on new developments, the assigned market value must be at least £156,988 at January 1999 prices. These figures apply anywhere in Greater London.  This equates to a value at April 2021 prices of £734,331 compared to the average market house price for all bedroom sizes of £494,402 (48.5% higher).

From Freedom of information replies, we know that mean average rents for First let Social Rent General Needs two beds in London in 2019/20 were £129.65 pw, compared to a median average two bed London Council relet rent of £106.35 pw.[6] The assigned market value for the first lets must have been equivalent to at least £515,664 in April 2019, an amount £46,170 (9.8%) higher than the average London house price in April 2019.  This is preposterous, especially because new Social Rent housing is not disproportionately located in high value areas. 

In April 2021 Haringey Council began to set Social Rent for its new build council homes at Cap.  These rents were justified as the valid outcomes of using the Social Rent formula. After representations, these rents were reduced in August 2021, allowing some useful insight into the detail of rent setting.

Looking at a one bed flat in Stroud Green Ward in Haringey, we see that the Social Rent which was set at £148.88 pw (the Cap) in April 2021 was reduced to £112.96 pw in August.  The assigned market values at December 2020 prices were £814,698 for the April valuation, and £468,970 for the August valuation. The April valuation was 39.4% higher than the average Haringey house price (of all bedroom sizes) in December 2020 of £584,144.

If we assume for a moment that the August valuation was accurate, then the assigned value in April was 73.7% too high.  We can have no confidence in valuations with such widely variant outcomes. Instead, we need transparent, comprehensive and accountable social rent setting policies.

Lack of effective regulation

The government says, ‘CORE provides a national standard dataset for examining high-priority issues, such as high and low-demand housing, allocations policy, and housing affordability… Policy makers and practitioners regard the system as an essential tool for monitoring housing costs, assessing affordability and developing policy.’[7]

In 2020, we asked the Regulator of Social Housing to investigate the overcharging revealed by the CORE data, but has refused to do so, on spurious grounds. The Regulator stated that it would not rely on CORE data to support regulatory interventions because ‘CORE data groups together formula rent properties with other tenures, including, for example, temporary housing accommodation, or specialist supported housing, which are exempted from the formula rent and Rent Cap requirements’.[8]  However, these arguments were perverse. In fact temporary accommodation is not covered by CORE, as is detailed in the annually revised CORE Manuals issued by the housing ministry; while our research as presented to the Regulator was exclusively concerned with general needs housing, and not any kind of supported housing.

Correspondence with the Regulator did not indicate any concern on their part about the possibility of social landlord breaches of Social Rent setting regulations. There was no concern about any reputational damage from such malpractices being allowed to continue.

A note on the scope of research and on data quality

Solely to reduce the data to manageable proportions, this research on Social Rent setting has been limited to General Needs, Social Rent properties in London.  It seems likely that the freedom which the Regulator allows to Social Rent setters may be very similar elsewhere in England.

The Regulator was asked to begin an investigation based on the CORE data showing evidence of overcharging. The CORE data is completed by front line workers, and sometimes different individuals answer the questions differently.  The Data is therefore indicative: but because the message from the CORE returns is so overwhelming, it should form the basis for further investigation.

Conclusion and recommendations

It is quite clear that rent rigging is taking place. The question is, what can we do about it.   

We need to bring Social Rent setting back under control, and back into line (or at least into a realistically close connection) with the rents of the existing properties. The Regulator of Social Housing must do its job and regulate.  The government should publish the CORE statistics on First Let Social Rents. We need to open up the process at local level, developing local Social Rent setting policies to drive out the current bad and unaccountable practices. Assurances about the adoption of general policies will be insufficient, and the publication of all rent setting data may be the only way to ensure good practice for all providers.

One reason for rents being set too high is to fill gaps in funding for new housebuilding. Hence, the national affordable housing programme must be adequately funded, so that rent increases are not used as a means to subsidise building costs.

We are willing to give verbal evidence in support of the case being made here, and to answer questions about it.

December 2021