Analysing the 2017 Ratings List

When the Valuation Office Agency (VOA) published the 2017 Ratings List, they made all of the underlying data from 1.8 million ‘bulk valuations’ of non-residential property available in a .csv file totalling over 740MB, with an accompanying Technical Guide.

Whilst the raw data is extremely rich and comprehensive, the recommendation that users import the file into a database package such as Microsoft Access in order to gain insights from the data fails to meet two of the UK Statistics Authority’s Code of Practice for Statistics principles:

  • V3 — Clarity and insight: Statistics and data should be published in forms that enable their reuse.
  • V5 — Efficiency and proportionality: Statistics and data should be presented clearly, explained meaningfully and provide authoritative insights that serve the public good.

I can see why the VOA might be reluctant to have their data used to inform policy and decision making; however, they would be hard pressed to deny that they often hold the ‘best available data’ for exactly that purpose, even after taking into account its limitations. In particular, whilst sectoral and productivity analysis is often carried out on an employment / financial basis, short term inelasticity of supply in the property market (potentially caused by a variety of factors) can represent a major barrier to economic growth or improved productivity.

This article aims to provide a flavour of the sorts of information and insights that the VOA could publish following each ratings epoch, as well as highlighting methodological limitations that they could address in future publications. I am not a ratings surveyor, so there may be aspects of the Ratings Manual that I have overlooked. Analysis may also be more revealing at finer grain than the national overview provided herein. Indeed, provision of data by electoral ward would also be desirable.

The article includes a sample of charts and maps, and I have published a Microsoft Excel spreadsheet allowing users to generate charts for (almost) every billing authority, and a suite of maps in .png format for the key sectors. The maps show the boundaries of 12 regions emerging from another analysis project that I’ve been working on in purple — more to follow!

Before we dig into the analysis, it’s important to note that the valuation data alone provides little insight into business rates revenue to billing authorities, due to reliefs and exemptions provided to businesses, and the redistribution carried out through the local government finance settlement. For example, around £5.6bn of Rateable Value (9.0% of England and Wales) lies within the City of Westminster, but they reportedly experience a net loss on business rates income after redistribution and appeals.

Our first chart shows the spread of the 1.8m ‘bulk valuations’ in England and Wales by sector. I was surprised at the roughly equal spread between the number of industrial properties, offices and shops — and how small the storage and distribution sector is (just 5.5% of properties). The ‘OTHER’ category (presented in uppercase to distinguish it from the VOA’s various ‘Other’ categories) makes up the remaining 16.7% of properties.

In terms of floorspace, storage and distribution and industrial properties are unsurprisingly larger on average (1,109 and 739 sq m, respectively) than shops and offices (219 and 203 sq m, respectively). Industrial properties have a broad range of floorspace, often several hundred thousand sq m. Together, industrial and storage and distribution properties make up 58.1% of rateable floorspace.

In terms of rateable value, shops and offices have higher average rateable values per sq m (£155 and £109 /sq m, respectively) than industrial and storage and distribution properties (£54 and £44 / sq m, respectively). Shops and offices have a broader range of rateable values, often running into several thousand pounds per sq m. Together, shops and offices make up 63.4% of rateable value.

Billing Authorities

Office floorspace is the most spatially concentrated property type, with the City of London providing a staggering 7.7 million square metres across 16,000 properties. The floorspace per resident is 1,067 sq m (almost double every other billing authority combined) due to the low proportion of residential floorspace in the area, coupled with high levels of in-commuting (daytime population of 423,000 compared to an overnight population of just 7,000). The office floorspace per worker is still the highest in England and Wales at 18.27 sq m, but this is much closer to the second-highest of Tower Hamlets at 14.39 sq m per worker. Tower Hamlets is also home to the largest office hereditament in England and Wales at 100,828 sq m.

Due to the geographic concentration described above, the colour-coding in the maps is derived using the Jenks optimisation method, rather than using equally-sized classes or intervals.

There are also notable concentrations of office floorspace elsewhere in Inner London, as well as along the M4 corridor, Warwickshire, and around the university cities (e.g. Bristol, Cardiff, Cambridge, Exeter, Leeds, Norwich, Oxford, Plymouth).

London also dominates the education property type, with three London Boroughs (Westminster, Kensington and Chelsea and Camden) providing more floorspace per resident than Cambridge. The City of London and Richmond upon Thames also feature in the top class (no pun intended), with over 0.20 sq m of education floorspace per resident. The HS1 / M20 and M4 corridors also feature consistently in the top 15% of billing authorities.

Storage and distribution floorspace is highly concentrated in the ‘Golden Triangle of Logistics’, which offers access to over 90% of the UK population within four hours’ drive.

The billing authority of North Warwickshire lies at the heart of the Golden Triangle, and provides the highest amount of storage and distribution floorspace per resident of any billing authority in England and Wales (34.47 sq m)— as well as the second-highest overall floorspace per resident after the City of London. Storage and distribution floorspace per worker is sixth-highest at 48.40 sq m.

The Golden Triangle is located at the corner of three of the regions, with areas of lower population density such as Harborough, Daventry and East Northamptonshire featuring prominently in the top 10 billing authorities.

The port locations also feature prominently in the top 10% of billing authorities — the Humber, Thames Estuary, Kent coast, Bristol Channel, and North West England. Swale is home to the largest storage and distribution hereditament in England and Wales at 344,322 sq m — a massive vehicle storage facility (the photo below is only a tiny part).

Imagery © 2018 Google, Map data © 2018 Google

In terms of industrial floorspace, the map below illustrates a ‘deficit’ of floorspace per resident in the highly populated areas of the south, and a corresponding ‘surplus’ in the highly populated areas of the north.

Derby is home to the largest single hereditament in England and Wales — an industrial property of 380,249 sq m.

Retail floorspace is highly concentrated in a relatively even geographic pattern, surrounded by catchment areas with significantly lower amounts of retail floorspace per resident. Billing authorities with boundaries constrained around settlements therefore feature more prominently on the map (e.g. Cheltenham, Gloucester, Ipswich, Lincoln, Norwich, York).

The City of London again presents an anomaly again with the highest retail floorspace per resident at 29.80 sq m — but the lowest retail floorspace per worker at just 0.51 sq m. The City of Westminster is the only other billing authority in the top class at 7.83 sq m per resident, but provides below average retail floorspace per worker at 3.43 sq m.

Analysis of the VOA data over several epochs could provide a temporal analysis of trends such as consolidation of comparison floorspace into a smaller number of regional cities.

I must stress that the analysis above has been carried out to generate discussion, and is not intended to be comprehensive. Obvious areas for further exploration would be comparison against productivity data (including sales densities for retail floorspace) and analysis of smaller geographies.

The views expressed are also my own, and not those of any organisation that I am associated with. Please do get in touch if you have any comments or questions, and hit the ‘Applause’ button if you found the article interesting!

Technical notes: Whilst this analysis focuses on the 1.8m ‘bulk valuations’ based on measured floorspace (referred to hereon as ‘properties’, the spreadsheet also provides the number and rateable value of all 2.0m hereditaments by sector — including everything from advertisement billboards and telecoms masts to power plants and car parks.

The rateable values per sq m do not distinguish between the different valuation methods for different types of property, or types of floorspace within properties. The VOA also has its own standards for measuring practice, which differ from those used by other organisations and/or for other purposes.

The maximum and average rateable values per sq m exclude properties with floorspace under 10 sq m, in order to provide more typical estimates by removing a small number of very high values. However, this also has the effect of excluding a large number of ‘hotel, guest and boarding, self catering, etc.’ and ‘retail — financial and professional’ properties from those calculations (63.7% and 51.6%, respectively).

Billing authorities may not align exactly with local authorities.

The data provided for Powys and Carmarthenshire billing authorities is split in the VOA dataset, and they are not included in the analysis. Part-time employees data is missing for Rutland and Oadby and Wigston. These gaps mean that the proportions quoted for England and Wales and floorspace per resident / job rankings may be skewed.

Planning and development surveyor / economic development specialist. Lifelong learner, writing about cities, society and sustainability.