Dear North Kesteven District Council,

In terms of the Freedom of Information Act of 2000, and subject to section 40(2) on personal data, could you please provide me with a complete and up-to-date list of all business (non-residential) property rates data for your local authority, and including the following fields:

- Billing Authority Property Reference Code (linking the property to the VOA database reference)
- Firm's Trading Name (i.e. property occupant)
- Full Property Address (Number, Street, Postal Code, Town)
- Occupied / Vacant
- Date of Occupation / Vacancy
- Actual annual rates charged (in Pounds)

If you are unable to provide an absolute “Occupation / Vacancy” status, please provide the Exemptions and / or Reliefs that a particular property may be receiving.

Please provide these data as machine-readable as either a CSV or Microsoft Excel file, capable of re-use, and under terms of the Open Government Licence (meaning reuse for any and all purposes, including commercial).

I last requested this data three months ago, and this is a request for an updated and current dataset.

I'm sure you get many requests for business rates and responding and managing these FOIs must be costly for you. Could I request that - as 71% of local authorities now do - you update and release this dataset via a dedicated page on your local authority website or on an open data service. I can recommend approaches for you to consider, and you should find that this reduces the time and cost of this request process.

Yours faithfully,

Gavin Chait

Dear NK - FOI,

Thank you for providing me your data. However, the occupancy status of each property (occupied/void) is not included. Please could you correct this and provide an updated version of your data.

Yours sincerely,

Gavin Chait

Dear North Kesteven District Council,

Please pass this on to the person who conducts Freedom of Information reviews.

I am writing to request an internal review of North Kesteven District Council's handling of my FOI request 'Complete Non-Residential / Business Property Rates Data (Q1-2019)'.

On 4 April 2019, I sent an FOI request for a complete and up-to-date list of all business (non-residential) property rates data, and including the following fields:

- Billing Authority Code
- Firm's Trading Name (i.e. property occupant)
- Full Property Address (Number, Street, Postal Code, Town)
- Occupied / Vacant
- Date of Occupation / Vacancy
- Actual annual rates charged (in Pounds)

My request has been refused in terms of Section 31(1)(a). According to the Information Commissioner's Office, "Section 31 is a prejudice based exemption and is subject to the public interest test. This means that not only does the information have to prejudice one of the purposes listed, but, before the information can be withheld, the public interest in preventing that prejudice must outweigh the public interest in disclosure."

Section 31(1)(a) deals specifically with "the prevention or detection of crime".

You don’t state what the nature of the risk of crime is, but no matter. I have prepared my own research into crime in vacant commercial properties by sending FOIs to every police authority in England and Wales. What follows is my submission to ICO and to the Information Tribunal in various cases over the past few years. Apologies if it seems pedantic, but it is critical that we have a common set of probabilistic definitions.

Vacancies form a small proportion of any commercial property base. Vacancy rates vary from as little as 2.6% in Cornwall and the Isles of Scilly, to 19.4% in the West Midlands.

Assuming that crime is evenly distributed (i.e. that an occupied commercial property is as likely to experience crime as an unoccupied property), for a vacancy rate of 10% you would expect a ratio of 9 crimes in occupied properties for each crime in an unoccupied property (ratio of 9:1).

In a street of 10 commercial properties, if 10 crimes were committed, and every property had an equal chance of being affected, then – all things being equal – each property, vacant or not, would experience one crime. That is the definition of “equal probability”, any property – irrespective of occupation state – is at the same risk.

Therefore, if we know the absolute number of property-related crimes, we can estimate a predicted number of crimes for both occupied and vacant properties. If the real data reveals more crime than predicted we have a positive correlation for increased risk; if the real data reveals the opposite, less crime than predicted, we have a negative correlation for increased risk.

According to UK Crime Stats (https://www.ukcrimestats.com/Police_Forc...) Lincolnshire Police Authority experiences the following level of crime March 2018 to Feb 2019 relating to Burglary + Criminal damage & arson + Robbery + Shoplifting = 15,313.

The ratio of 10 to 1 implies that, if there were 15,313 commercial property-related crimes, an average of 1,513 of these types of crimes would have been committed in vacant properties between March 2018 and February 2019.

Compare your data and offer some evidence as to why you believe you are experiencing a greater-than-expected level of crime in vacant commercial properties.

It is insufficient to simply produce evidence that crime occurs in vacant properties. It is necessary to produce evidence demonstrating that vacancy property crime occurs at a rate higher than expected. I do not believe that you have done so; or could.

I requested statistics on crimes by vacancy status from Gloucestershire Police in March 2016. These data are not aggregated and Gloucestershire Police were unable to comply unless I was willing to pay to have them go back through each written crime report and extract that data manually, if it existed at all (https://www.whatdotheyknow.com/request/a...). If Gloucestershire Police have no data, then Stroud has no data.

If you have data that supports crime levels in vacant properties significantly in excess of what is expected per year, then we can assess whether or not vacant properties are at greater risk to crime. Note, however, greater risk does not identify the cause of that risk. I imagine that this would be of interest?

When I started my research in early 2016, 70 out of 348 local authorities were already regularly publishing their vacancy data to their open data websites. However, 66 local authorities in England and Wales refused to publish citing the risk of such data being used to commit crime. Their justifications cited previous findings by the Information Tribunal (such as Mr. Yiannis Voyias v the Information Commissioner and Camden EA/2011/0007) to validate their decisions.

They did not conduct their own risk assessments, they simply used Voyias as justification.

During 2016, I sent FOI requests to all police services across England and Wales requesting total number of incidents of criminal activity in empty commercial properties. At the time, I had secured 66% of vacancy data from local authorities through FOI requests, and direct access.

The combination of crime incident data and the list of local authorities would permit easy comparison between areas that regularly disclose and those which choose not to in order to assess whether there is a greater risk as a result of disclosure.

Out of 44 police services, only two were able to provide data on incidents in empty commercial properties. The remaining police services do not specifically collect such data and have no way of knowing what the incident rates are.

The two who have are Thames Valley Police and North Wales Police. If a relevant police authority is advising you on a Section 31(1)(a) exemption, then - unless they are either of these - they do not have data to validate their opinion.

There are about 45,000 commercial properties in North Wales and vacancies are about 8%, leading to an expected crime ratio of 12:1. North Wales Police state that there are an average of 1,780 crimes a year in occupied properties, and 26 crimes a year in unoccupied properties that largely have to do with theft, vandalism or arson (note that squatting in commercial property is not a crime and so unrecorded).

The total of 1,806 implies the expected number of vacant property crime incidents to be 71; 270% higher than the actual 26 recorded.

In other words, there is a large negative correlation between crimes committed in vacant properties relative to what was expected. A vacant property is less likely to experience crime.

This tells us that crime in empty commercial premises is extremely rare, and explains why most police services do not bother routinely aggregating such data.

The data from the Thames Valley police gives us a better idea of the variation between authorities which publish, and those which don’t.

For example, in 2015 Oxford had 4,038 commercial properties and suffered 2 cases of empty commercial property crime at a cost of £1,259. In comparison, they had 3,133 cases of crime committed in occupied business premises, at a cost of £507,956.

By comparison, Reading, with 5,659 commercial properties suffered 2 empty commercial property crimes that caused no damage at all.

Oxford refuses to publish under Section 31(1)(a) while Reading publishes regularly.

In total, across the Thames Valley, for 74,027 properties - of which about 7,000 are empty (9.5%) - only 11 crimes related to empty properties were recorded. Compare that to occupied and actively-traded properties which experienced 29,946 reported crimes.

For an approximate 9:1 vacancy ratio, and 30,000 crimes, you would expect 3,000 vacant property crimes. 11 is a statistical rounding error. Crime in vacant properties is not a meaningful risk.

The Thames Valley data are here: https://www.whatdotheyknow.com/request/a...
And the North Wales Police data are here: https://www.whatdotheyknow.com/request/a...

In two distinct parts of the UK, with different approaches to recording crime, empty property crime is extremely rare, and – more importantly for publication considerations – publication of data listing empty properties does not have any impact on the number of incidents of crime.

Some local authorities who claim exemption under Section 31(1)(a) claim to have their own data to validate these claims. I sent FOIs requesting these data to 64 local authorities (including both those who do publish empty property data, and those who do not). Not one has any data - other than very sparse anecdotal data to support their claim.

All the data I collected indicates that there is no substantive basis for concern that publishing a list of empty properties will lead to prejudice under Section 31.

In November 2016, I appealed this class of refusal – using this evidence – to the Information Commissioner’s Office.

My central argument was that, in order to justify a Section 31(1)(a) exemption it is not sufficient simply to claim that crime occurs in empty commercial premises. Neither should anecdotal tales suffice. It is also necessary to provide evidence that publishing a list of empty properties in real, measurable terms increases the expected incidence rate for crime in such properties.

My submissions to ICO resulted in the following findings in my favour:

FS50628943 vs Cornwall Council; citing Section 31(1)(a);
FS50628978 vs London Borough of Kensington and Chelsea Council; citing Section 31(1)(a);
FS50681246 vs London Borough of Camden; citing Section 31(1)(a);
FS50681250 vs London Borough of Greenwich; citing Section 31(1)(a) and Section 38(1)(b);
FS50681266 vs Royal Borough of Kingston Upon Thames Borough; citing Section 31(1)(a);
FS50681283 vs Westminster City Council who cited Section 31(1)(a);
FS50681292 vs London Borough of Wandsworth; citing Section 31(1)(a);
FS50681297 vs Knowsley Metropolitan Borough; citing Section 31(1)(a), 40(2), and 41;
FS50681322 vs West Berkshire Council; citing Section 31(1)(d);
FS50681326 vs Vale of Glamorgan; citing Section 31(1)(a);
FS50681332 vs Doncaster Metropolitan Borough; citing Section 31(1)(a);
FS50681336 vs Sheffield City Council; citing Section 31(1)(a), 31(1)(d), and 41;
FS50685343 vs London Borough of Ealing; citing Section 31(1)(a), 31(1)(d), and 31(2);
FS50685350 vs Liverpool City Council; citing Section 31(1)(a);
FS50685375 vs West Lancashire Borough Council; citing 31(1)(a);

In each of these decisions, the Commissioner either dismissed the application of the exemptions cited, or declared that “the public interest in the information being disclosed outweighs that in the exemption being maintained”.

Any claim in support of Section 31(1)(a) must also consider the public interest in the data requested.

In support of public interest, I offer that 95% of 348 local authorities in England and Wales now publish commercial vacancy data. 71% publish these as regular updates to their authority websites or open data services.

Retail and commercial vacancy, and its impact on commercial rates continue to be of great public interest (cf, as a small example, “City figures warn high street landlords face ‘tsunami’ of pressures” https://www.ft.com/content/fa7c48d0-6993..., “After the retail apocalypse, what next for the high street?” https://www.theguardian.com/society/2018..., “An investigation of the impact of 2017 business rates revaluation on independent high street retailers in the north of England” https://www.emeraldinsight.com/eprint/wT...).

These claims of a ‘retail apocalypse’ are having a very real impact on the capacity for local authorities to collect commercial rates. Most recently, Debenhams – the department store chain – has demanded business rates cuts to support their viability (https://www.ft.com/content/1bcbb9fc-6b3e...). It is critical that researchers have the data to test such claims, both now, and over time, in order to ensure that it is possible to differentiate between poor individual business management, and general economic difficulties.

I ask that you reconsider your refusal, and publish. Failing that, I will refer this matter to ICO for further review.

A full history of my FOI request and all correspondence is available on the Internet at this address: https://www.whatdotheyknow.com/request/c...

Yours faithfully,

Gavin Chait