Private Sector Rent Statistics, Scotland, 2010 to 2022

Annex C – Methodology (Source Data, Sampling Methodology, Sample Sizes)

This publication uses data from the Rent Service Scotland 'Market Evidence Database', which is a database used to meet the needs of determining annual Local Housing Allowance levels.

The market evidence data on private rents is sourced through a variety of means, including advertised rental information, private landlord and letting agent returns, and mailshot initiatives.

The database excludes any rents related to social housing, mid-market rents, halls of residence, and private tenancies known to be the subject of housing benefit and regulated tenancies.

In the latest year to end September 2022, an estimated 86% of records were based on advertised rents, with the remainder being based on sitting tenant rents from landlord returns.

The data collected includes a minimum level of address, property attributes and tenancy details. Rents relating to studio/bedsit properties, properties with 5 or more bedrooms, and bed and breakfast lodgings have been excluded from this publication due to small sample sizes. Rents for bedrooms in shared properties are presented as 'rent only' figures, i.e. do not include the additional cost of shared services where these are known.

If a particular property has more than one piece of market evidence available in a given year, then only the most recent item of evidence for that year has been used in the average rent calculations for this publication.

The private rented sector is de-regulated, meaning that landlords are free to charge an open market rent for their property. There is currently no legal obligation for landlords or agents to provide Government, or any other organisations, with details of the rents achieved on their lettings. Therefore rent officers have to actively seek, collect, validate and maintain a suitable dataset.

There is no requirement for rent officers to collect 100% of rents that are agreed between landlord and tenant. Neither is it realistic to assume that all landlords and letting agents would be able to co-operate with this requirement.

Rent Officers instead aim to capture a representative sample of around 10% of private rents based on the total number of records obtained (the amount of records used in average rent calculations may be slightly less than this due to removal of any multiple records for a single property/address in a given year). Landlord registration data and census data is used as a baseline for establishing and monitoring the total sample proportion that is aimed to be achieved.

As rent officers do not have access to every letting that takes place in the market the use of a random sample is not feasible, and given the variations in the size of the markets in each Broad Rental Market Area a simple quota based sample would be unlikely to produce representative results either. The sample should ideally reflect the profile of the market in terms of the type of property, its distribution, and the letting sources within each Broad Rental Market Area.

There are no definitive measures for these so rent officers monitor local market activity and take every opportunity to acquire feedback from landlords, agents and tenants. This market intelligence means that rent officers are able to continually evaluate the composition of the list of rents used for Local Housing Allowance, and where necessary divert resources from their regular program of data collection to address any perceived weakness in the data.

This combined approach of regular and targeted collection based on market intelligence aims to produce a representative sample for each property size for each Broad Rental Market Area. This approach in turn reflects the structure of the legislation which allows for rent officer judgment on a number of these factors.

The private rented sector is very complex and is continually changing as it reacts to market forces. The overall target of a 10% sample therefore only represents a guide figure at Broad Rental Market Area level. Local knowledge, confidence testing and interpretation of other available data may be applied to refine the guide level. This contributes towards achieving a representative sample for each property size category at a Broad Rental Market Area level.

It is important to note that the data collected on individual rents may encompass different property types and addresses for each data collection year. The Broad Rental Market Area Profiles in Section 9 show the sample sizes for each rental area. It can be seen that there have been some variations in the number of records by rental areas over time, and also the proportions by size of property. Some of this may be due to changes in the underlying rental stock over time, and some may be due to sampling variations over time.

Also note that some methodological improvements were made to the 2020 publication, which have been carried forward again to this publication, in relation to how Scotland level average rent figures have been estimated. These include the use of a weighted stock approach, which is set in further detail in Section 8.

Table C1 and Chart C1 below illustrate the different sample data profiles by rental market area.

It can be seen that the sample data profiles differ by rental area. For example for Dumfries and Galloway, 1 bedroom properties make up 21% of all sample records and 3 bedroom properties make up 25% of the total. This compares to Perth and Kinross for which 30% of records are 1 bedroom properties, and 13% of records are 3 bedroom properties. This emphasises that it is generally not appropriate to compare an overall "average" rent figure (averaged across all property sizes) between different areas of the country.

It is also important to note that there are some sample data profiles that have changed over time by property size, which may also introduce some bias into comparing overall Broad Rental Market Area averages over time.

Chart C2 available in the Supporting Documents Excel Workbook shows an example of this for Argyll and Bute, for which the sample data profile has changed over time both in the total number of records and the proportion in each property size category. The proportion of records relating to 1 bedroom shared properties increased from 3% in 2010 to 16% in 2013, after which this percentage has dropped to 8% in 2021. Whilst the proportion of records relating to 2 bedroom properties decreased from 43% to 34% from 2010 to 2014 but then increased to 45% in 2021. This would likely add bias to the trends if an overall "average" rent figure was calculated (averaged across all property sizes) each year.

Table C3 and Chart C3 available in the Supporting Documents Excel Workbook show the proportions of the total samples that are within each Broad Rental Market Area each year, by property size.

For most property sizes each rental area has a relatively consistent proportion each year, however there have been some changes over time. For example in the 2020, the proportion of all 2 bedroom records that are located in Lothian increased by 4 percentage points compared to the previous year, and the proportion of all 4 bedroom records that are located in Lothian increased by 7 percentage points compared to the previous year.

Given the variability in some years of the sampling numbers by area and property size as shown in Table C3 and Charts C3, separate weights by area and property size have been calculated based on Scottish Household Survey (SHS) data, which are also available in the Supporting Documents Excel Workbook. These SHS weights have been used to estimate Scotland level average rents in this publication.

The Scottish Household Survey (SHS) weights have been built up by using an approximate mapping of local authority areas to Broad Rental Market Areas.

The weights from the survey data have been constructed using 3 year rolling data periods to help with precision of results at Broad Rental Market Area and to smooth out annual variations seen in the survey results. Each 3 year data period being centred on the rental year to be estimated with a lag applied of 2 years, due to the time delay between 3 year survey results being available and the latest rental year in question.

For example in the 2020 publication, for the rental year 2020, a 3 year data period 2017 to 2019 was used to construct the stock based weights, which equates to the year 2020 lagged by two years (i.e. to be centred on 2018).

One aspect to note for this publication is that given that the Scottish Household Survey results for the year 2020 are not available, the previous set of weights based on the 3 year data period 2017 to 2019 have been rolled forward as the basis for the rental year 2021.

For each property size category, the stock based weights have been calculated based on the proportion of private rented households in each category that are located in each of the Broad Rental Market Areas.

The exception is for 1 bedroom shared properties, where the weightings are instead based on the estimated proportions of owner occupier or private rented households in each areas that contain two or more adults who are not related or who are not in a married / civil partnership or a co-habiting partnership, with the aim to identify households where there may be an individual room being rented in some form.

To note one limitation of the weighting approach applied is that there may be some level of mis-match between the SHS weights (covering all PRS stock) and the sample of records collected from Rent Service Scotland (i.e. excluding those with housing benefit and regulated tenancies)