Business rates - British Property Federation

Business Rates: Who Pays and Why it Matters

A Report by Regeneris Consulting

Business rates: Who Pays and Why it Matters?


British Property Federation, British Council for Offices, British Council of Shopping Centres

Business Rates: Who Pays and Why it Matters

December 2015 Regeneris Consulting Ltd www.regeneris.co.uk


Contents Page Executive Summary

i

1.

Study Purpose and Objectives

3

2.

The Policy Landscape

5

History of Business Rates

5

Recent Policy Developments

5

Future Changes

7

The Investigation: Methodology and Expectations

8

The Issues

8

Expectations

9

Study Design

10

The Regeneris Study in Context

10

3.

4.

Results

11

5.

How Rates Affect Landlords

12

Development Foregone

12

How Rates Affect Occupiers

15

Impact Assessment

15

Issues for Policy Makers

18

6.

7.

Appendix A - Previous Literature Appendix B - Date and Sample Definition Appendix C - Detailed Results Appendix D - Correlation Co-efficients Appendix E - Case Studies and Wider Trends


Appendix F - Regional Datasets by Use Class: Prime Sample Appendix G - Regional Datasets by Use Class: Constant Sample


Executive Summary i.

Business rates are a £26 billion tax at a crossroads: concerns about the operation of the system led to a Treasury-led review announced earlier in 2015; whilst the Chancellor’s recent Autumn Statement confirmed that, by 2020, business rates would become a fully devolved local tax.

Who ultimately pays business rates? ii.

Economists talk about the “incidence” of a tax: this refers to who ultimately bears the cost of the tax. In the case of business rates, the key issue around incidence is whether the occupier really pays the tax, or whether it is reflected in higher or lower rents with the ultimate incidence falling on the landlord/land owner.

iii.

This research has examined existing studies and carried out a new analysis of a large sample of data on rates paid and rental values over the period 1990 to 2010. Its broad conclusions are: 

In the medium to longer term, changes in rates paid are reflected in corresponding adjustments in rental values. This relationship is clearest in the retail sector, but can also be seen with offices. The explanation for the different strength of relationship may lie in lease structures or in the relative balance of power between landlords and occupiers in these two use classes.



There is clearly a lagged relationship between changes in rates and the feed through to property rentals. This is only to be expected as most rents are paid at a fixed rate over a three-to-five-year typical lease period. Therefore the degree of flexibility for occupiers to change their rents payable is limited in the short term, but grows over time. The effect is largely passed on from occupiers to landlords after three to four years (with around 75% passed on after three years).



The relationship between business rates and rents is stronger in regional markets than in London. This may be because leases tend to be longer for prime London property, or it could again be down to the relative balance of power between landlords and occupiers in different markets.



The relationship between business rates and rents appears to break down after 2008. This is hardly surprising as this period: (1) contains historically unprecedented changes in rental values, rents paid, capital values across the country and especially in the retail sector; and (2) is unusual in that there has been no revaluation since 2010 and rateable values are still based on 2008, pre-recession, values.

iv.

Over the period 2011/12 to 2014/15, businesses in rented properties saw their rates bills rise by around £1.5 billion in England. Based on this research, this equates to £0.7bn in development capital foregone; an amount that would be increased by the gearing of borrowing to land values.

v.

More generally, our calculations suggest that, if the rates burden is 75% capitalised into rents over a three year period, then an increase of £100m per annum for three years would – over the course of those three years – lead to: 

A £150m reduction in development capital for landlords, affecting their ability to invest



A £150m loss in income for rental tenants, equating to approximately 1,000 jobs foregone.

i

Business rates: Who Pays and Why it Matters? vi.

As time passes, a greater proportion of the increased financial liability will be passed on to landlords.

vii.

Finally, it is important to emphasise that this research also highlights the complexity of the factors that determine rental levels, and that isolating the effect of rates is challenging.

What conclusions should policy makers draw? viii.

There are a number of important policy conclusions: 1)

The current structure of periodic business rate revaluation – every five years and the current seven year gap –creates uncertainty in future rental returns as it is very difficult to calculate future business rates costs. Such uncertainty is unique as far as the UK tax base is concerned and is likely to reduce investment in property below the level that is economically desirable. Implication: more frequent revaluations, which are technically feasible, would avoid much of this uncertainty and also obviate the need for complex transitional reliefs.

2)

To the extent that the business rates burden is capitalised into lower rental values this reduces the development capital available for re-investment, further multiplied by current gearing ratios. To the extent the burden falls on occupiers, this is an extra cost to business and reduces their ability to invest, profitability and ability to employ staff. Implication: whichever way the incidence falls there are significant economic effects from business rates. In the absence of rapid changes these will tend to fall on owners of property rather than occupiers.

3)

The use of rates reductions as a medium to long term incentive for occupiers to invest in a location (either an Enterprise Zone or a whole local authority area) is likely to be ineffective. The benefit of longer term predictable reliefs is likely to be capitalised into a one-off uplift in land and property assets values, which may spur additional investment by property owners. Implication: local authorities and Local Enterprise Partnerships need to be realistic about the limited incentive effects of business rates relief for occupiers.

4)

Temporary reliefs targeted at particular groups of occupiers are, if only lasting one to two years, likely to benefit the target occupier group rather than landlords/property owners. This suggests that the system of temporary reliefs for SMEs is likely to be effective. However, to the extent this becomes a permanent feature to the business rates system, it would become less effective. Implication: specific reliefs can create unintended consequences.

ii


1. Study Purpose and Objectives 1.1

In 2015 Regeneris Consulting were commissioned by the British Property Federation with the British Council for Offices and the British Council of Shopping Centres to explore the evidence on the economic effects of non-domestic rates (business rates). This was to help inform the national review into business rates announced by the then Coalition Government in the March 2015 Budget1. Our work has focused on the following: 

First, to provide a definitive view on the economic incidence of business rates. That is to address the question: who ultimately pays business rates?



Second, and connected to the above, to assess whether the extent to which economic incidence falls on occupiers varies according to use class and location.



Third, to quantify the unintended consequence of rising economic incidence of business rates on landlords/investors and occupiers.



Fourth, to reflect on the implications of our findings for policymakers.

Context 1.2

The UK has the highest level of property taxation in the OECD group of developed nations. This includes taxes on occupancy, taxes on transfers, taxes on capital gains and taxes on planning gain.2

1.3

This report is about business rates, how they operate, who they affect and who pays them. These are important questions, not least because in 2014/15, central government received £25 billion3 from business rates, over half the amount generated by corporation tax. It is also important because the Government has announced a radical policy to localise the operation of business rates after two and half decades of a fully centralised system.

1.4

1.5

1

For such an important tax, there is surprisingly little understanding of its effect on occupiers, landlords and the wider economy. This report examines the existing evidence base, provides some important new analysis, and models a number of scenarios to explore the interaction between rents, rates and capital values.

Figure 1.1 Business Rates in Context, £m £50,000 £43,005

£40,000

£30,000

£24,835

£20,000

£10,738 £10,000

£Corporation tax

Business Rates

Stamp Duty Land Tax

Source: HMRC: HM Revenue and Customs Receipts, ONS Publication: PUSF Public Sector Finances First Release 2014

The research is important because better data and improved statistical techniques mean that it provides a more comprehensive picture of the relationship between business rates and rents than has previously been thought possible. It will inform policy makers, investors and occupiers. Most importantly, it will show how the reform of the business rates regime and accompanying national

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/413070/business_rates_review_final.pdf

2

According to OECD Revenue Statistics 2014, the UK has the highest share of GDP accounted for property taxes and the highest share of taxes accounted for by property taxes of any OECD country (the data does not distinguish domestic and business property taxes 3

ONS publication, PUSF Public Sector Finances First Release 2014

3

Business rates: Who Pays and Why it Matters? revaluation exercise could have a series of unintended and potentially long lasting economic effects. 1.6

The report is structured as follows: 

Section Two outlines the current and future policy landscape, examining the current tax design, policy issues and moves for reform.



Section Three considers the scope of our current knowledge, the nature of our investigation and the data sources used.



Section Four describes the results of our econometric analysis, and what this means for our understanding of the effect of business rates.



Section Five considers the implications of this study from the landlord’s perspective.



Section Six considers the implications of this study from the occupier’s perspective.



Section Seven examines the policy issues raised by this study, focusing on those related to reform of the business rate system.



A series of technical appendices complete the analysis.

4


2. The Policy Landscape 2.1

The occupiers (or owners if empty) of approximately 1.8 million commercial properties in the UK are liable to pay business rates. The amount billed to individual properties each year depends on the ‘rateable value’ of the property and the national ‘business rates multiplier’. The individual rateable values have been until recently reviewed every five years by the Valuation Office Agency (VOA). At each revaluation date they are based on an estimated annual market rent value for individual properties as of April two years prior to the revaluation date.

2.2

The business rates multiplier (officially, the Uniform Business Rate or UBR) is set annually by central government and is the number of pence per pound of rateable value an occupier must pay before any business rates relief is applied. Properties that can qualify for business rates relief include some rurally located properties, properties occupied by charities, properties with a low rateable value and properties located in enterprise zones.

2.3

The business rates payable on any property is therefore the nationally set business rate (pence in the pound) times the rateable value of the property (less any reliefs that may be applicable).

History of Business Rates 2.4

Taxes on commercial property have a long history in the United Kingdom. Business rates were first introduced in their current form in 1990. Prior to 1990, the revenues derived from business rates accrued directly to the local government and the business rate multiplier was under local authority control. In April 1990, locally varying non domestic rates were replaced by national non-domestic rates (NNDR) and central government became responsible for setting the business rates multiplier.

2.5

The then government suggested that the introduction of a uniform business rates multiplier would:

2.6



remove ‘arbitrary’ variations in the amounts demanded of businesses by local authorities



allow for greater stability in rate payments over time (increases in UBR cannot be set higher than the prevailing annual change in the retail price index)



lead to ‘radical simplification’ of the local authority grant equalisation system.

Since 1990, the business rates multiplier has been set nationally and is normally increased annually in line with RPI inflation. When revaluations take place every five years, the multiplier is adjusted so that the change in the overall rates burden across the country is the same as the RPI change.

Recent Policy Developments 2.7

In October 2012, the Government took the decision to postpone the scheduled 2015 revaluation until 2017. This decision attracted much criticism at the time as rateable values now remain based on April 2008 valuations, prior to the significant changes in rental values that occurred in many locations as a result of the property crash and recession of 2008 to 2012. This means that 

First, business property as an economic entity saw a significant increase in the overall effective rate of tax (as rateable values have risen relative to rental values yet the business rates multiplier did not fall to offset this).



Second, the very large shifts in relative values since 2008 have not been taken into account so the effective rate of taxation on some classes of property in some locations has shifted quite dramatically.

5

Business rates: Who Pays and Why it Matters? 2.8

2.9

In a statement to Parliament, Communities and Local Government Minister Brandon Lewis supported the revaluation delay asserting that: "a revaluation at this point would be likely to result in sharp changes to business rate bills in many parts of the country and in many sectors." However, those opposed to the delay believe it places an unexpected and additional financial burden on businesses as they continue to pay higher businesses rates, unreflective of current economic conditions.

Figure 2.1 Net Business Rates Collected by Region 2014 Region

London South East North West East of England West Midlands Yorkshire and The Humber South West East Midlands North East

Business Rates collected 2014 (£bn) £6.6 £3.4 £2.5 £2.1 £1.9 £1.8 £1.8 £1.4 £0.8

% of Total 29.3 15.1 11.2 9.4 8.6 8.2 8.1 6.3 3.7

In 2013, local councils were given Source: DCLG, National Non Domestic Rates Return 2014/15 net amount permission to retain 50% of the receivable from business rate payers by local authority area proceeds of any future uplift in their business rates base, with the intention of giving councils an incentive to encourage enterprise and job creation. Figure 2.1 and Figure 2.2 illustrate the net business rates receivable per region and per local authority area for the eight core city locations and the four London boroughs we analyse in this study. It shows that the business rates base is far from evenly spread, with London as the dominant source of business rates reflecting the high value of commercial property there.

Figure 2.2 Net Business Rates Received by Local Authority Areas in 2014

Source: DCLG, National Non Domestic Rates Return 2014/15 net amount receivable from rate payers by local authority area

6


Future Changes 2.10

On the 5th October 2015, the Chancellor announced plans to allow councils to retain 100 per cent of all business rates collected (existing business rates base as well as any new business rates), whilst grant from central government will be phased out by 2020. There is still a live debate around the future of the current system of top up, tariffs and the safety net. Either way, some element of redistribution will remain in place. The government plans to abolish the uniform business rates multiplier and give local authorities the power to cut business rates to boost enterprise and economic activity in their areas. Local areas which successfully promote growth and attract businesses will now keep all of the benefit from increased business rate revenues not just 50%.

2.11

Whilst all local authorities will have the ability to cut business rates, only areas with directly elected mayors will be able to increase rates. These areas will be able to levy a premium of up to 2p on the rate in order to pay for infrastructure, with any premium needing the support and approval of an area’s Local Enterprise Partnership (LEP).

2.12

These changes undo one of Margaret Thatcher’s last legacies when her government introduced the national business rate with the current Chancellor George Osborne stating it will be the ‘biggest transfer of power to our local government in living memory.’

2.13

It is hoped that the incentive of a 100% rate retention will require local governments to become more engaged with the business community and develop a greater understanding of the impact business rate changes have on these businesses.

7


3. The Investigation: Methodology and Expectations 3.1

It is generally recognised that business rates are far from a perfect tax4. Business property is an input to the productive process of a company along with other factors of production such as labour and capital. It is an important principle of the economics of taxation that an efficient tax system should not distort choices firms make about inputs into the production process, and hence that intermediate goods—those used in the production process—should not be taxed. The principal effect of business rates is that economic activity in the UK is artificially skewed away from propertyintensive production. The same argument could of course also be made about employers national insurance contributions (a payroll tax).

3.2

In spite of this inherent economic problem with business rates they have remained a popular tax with successive Chancellors. Largely because they are relatively easy to collect and extremely difficult to evade. They are also non-cyclical in the UK as the tax base does not rise and fall in the economic cycle.

3.3

However, as with all taxes the nature of the tax and the way it is collected and any reliefs introduced all have economic consequences which are far from obvious.

The Issues 3.4

Our investigation looked at four principal research questions: 1)

On whom does the economic incidence of business rates fall (occupiers, land owners, landlords)?

2)

How has this changed over time?

3)

How does this vary between different locations?

4)

How does this vary between different use classes?

3.5

Underpinning these questions is an important distinction between the financial and the economic incidence of business rates. In general, the financial or statutory incidence simple indicates who the law says will pay the tax. In contrast, the economic incidence of a tax indicates the extent to which someone is made worse off by the tax. The economic and financial incidence are often very different, and will depend on demand and supply conditions in the relevant markets.

3.6

The financial incidence of business rates is borne by occupiers. This makes the tax easy and relatively efficient to collect, particularly when compared with the much more complicated regimes governing payroll and profit-based taxes. However, the economic incidence of business rates has been the subject of both policy and academic debate for some time. The debate is important because the share of economic incidence between occupiers and landlords can help

4

For instance the Institute for Fiscal Studies’ “Tax by Design: the Mirlees Review” (2011) explored the taxation of land and property, including Business Rates, and commented that the Business Rate is not a good tax. It taxes business property as an input to the productive process of an enterprise, and so distorts firms’ choices and so the IFS argue that the effect in the UK is to artificially skew economic activity away from property-intensive production. This combines with an additional distortion in that Business Rates are zero on unused or undeveloped land which provides perverse incentive to use land inefficiently. When commercial property is subject to tax but land is not, then there is an incentive to withhold land from development, but also an incentive to demolish empty or unused property.

8

Business rates: Who Pays and Why it Matters? determine investment decisions, the shape and scale of new developments, and the wider attractiveness of commercial property as an asset class. 3.7

The essential problem is whether or not differences in property tax rates are reflected in the values of the properties to which they apply. The dynamic nature of this relationship is captured in Figure 3.1 (below), which demonstrates how rate revaluations are assumed to have short to medium term effects on rental levels. Figure 3.1 Business Rates Adjustment and Methodology

Source: Regeneris Consulting

Expectations 3.8

Existing research into the relationship between business rates and rents is dominated by a 1996 paper by Bond, Denny, Hall and McCluskey. This paper used data on 2,964 investment properties from 1987-1992, disaggregating the model into different use classes and geographies. It produced poor results for office and industrial properties, but identified a significant relationship between rents and business rates in retail properties across three market areas – Retail London, Retail South East and Retail Elsewhere.5 The Bond et al paper is regularly cited in policy related studies, such as the 2014 IFS Green Budget.6

5

Bond. S, Denny K, Hall, J, McCluskey, W, (1996), Who pays business rates? Fiscal Studies, Vol. 17, No. 1, pp.19-35.

6

See IFS Green Budget 2014 see http://www.ifs.org.uk/publications/7072, Chapter 11.

9


The startling thing is that the original paper is very limited in scope (1987-1992), and only produced strong results for one use class. Appendix A places this study into context with two other recent investigations.7 It should be clear from this summary that the foundation for the view that changes in business rates feed into changes in rental and capital values is based on relatively restricted evidence and quite selective modelling results. Nonetheless, this view has been accepted by policymakers and commentators since 1996 without any real questioning.

Study Design 3.10

In order to test the underlying relationship between rents and business rates, we have designed a four part investigation: 1)

A descriptive analysis of the trends and patterns between business rates and rents from 1990 to 2014.

2)

An econometric analysis of the relationships, using established modelling techniques and specifications

3)

Further econometric analysis, using new model specifications and data transformations.

4)

Modelling incidence between regions and use classes, to determine the nature and effect of business rate changes on occupiers, owners and the wider economy.

3.11

Full technical details of the study design and methodology may be found in the appendices, however it is worth noting at this point that to a certain extent the study and its parameters were constrained by the data available.

3.12

This meant that it was unfortunately not possible to account for the impact of business rates transitional relief on business rates bills. Had the impact of the relief been included it would have dampened down the strength of the relationships between business rates and rents in some instances. However the direction of any relationships would not be affected.

The Regeneris Study in Context 3.13

The current investigation into business rates and rental levels is important because it is geographically comprehensive, and covers five revaluation episodes (1990, 1995, 2000, 2005, 2010). It also covers several phases of a full property market cycles, including the biggest dislocation in the market since 1945. Although this last episode was extreme, the cyclical nature of commercial property investment means that any insights that we can gain from data may be an invaluable guide to future events. It was on this basis that we proceeded with the four stage econometric investigation.

7

Tyler, P, Bond, S, Gardiner, B. (2012). The Impact of Enterprise Zone Tax Benefits on Local Property Markets in England: Who Actually Benefits. Journal of Property Research. Mehdi, M, (2003) The Capitalisation of Business Rates: an Empirical Study of Tax Incidence in Six London Boroughs, unpublished PhD, London School of Economics and Political Science.

10


4. Results 4.1

The evidence suggests that over period a period of two to three years approximately 75% of the value of business rate change is capitalised into rents. This estimate is based on previous studies and the application of established modelling techniques to our new data. We can therefore conclude that the economic incidence of any changes in this tax is borne initially by occupiers, but soon transfers to owners and landlords.

4.2

Taken as a whole, we find evidence that is similar in its robustness to that found in previous investigations8. The difference is that these findings apply to use classes across the UK and over a period of 24 years. The evidence is therefore more comprehensive. While the scope of the investigation has been determined by the nature of the available data, our new work is the longest macro based time series study of the relationship between rents and rates currently available.

4.3

The relationship between business rates and rents is clearest for the retail sector. The relationship exists to a lesser degree for office properties and our model suggested that larger revaluation events have a weaker influence in that context. This implies that office rents tend to primarily track changes in wider macroeconomic forces, with business rates exercising a secondary influence.

4.4

The association between business rates and rental levels becomes less noticeable from 2008/09 onwards. The apparently weaker relationship from 2008/09 onward is hardly surprising as: (1) this period contains historically unprecedented changes in rental values, rents paid, capital values across the country, but especially in the retail sector; and (2) an unusual period in that there has been no revaluation since 2010 and rateable values are still based on 2008 pre-recession values.

4.5

The relationship between business rates and rents appears to be stronger in regional markets, including Newcastle, Manchester, Birmingham and Liverpool. London rentals appear less responsive to changes in business rates.

4.6

Figure 4.1 provides our best approximation of the speed and scope of likely business rate capitalisation. This profile is used as central guide in our subsequent analysis of policy impacts.

Figure 4.1 Economic Incidence Model 100% 90% 80% 70%

60% 50% 40% 30% 20% 10% 0% Year 1

Year 2 Occupiers

Source; Regeneris Consulting

8

Detailed results are set out in Appendix C to E

11

Landlords

Year 3


5. How Business Rates Affect Landlords 5.1

The landlord’s or investor’s perspective on business rates is conditioned by two factors. The first is that financial incidence rests with occupiers. Because of this, tenants deal with tax collection and valuation. While valuations and the scale of occupiers’ taxes may affect the marketability of an asset, business rates are generally seen as an occupiers’ issue. The second point is that the differential economic incidence of business rates is only a concern where properties are rented. As the PIA’s 2015 Property Data Report9 outlined, some 43% of commercial property stock is owner occupied. For these properties, the question of economic incidence is straightforward as it simply reflects the financial incidence.

5.2

The usual presumption has been that if the economic incidence of business rates is passed onto landlords, then any tax increase will reduce landlords’ yields and profits. By contrast, a falling tax burden would feed through into higher margins and yield. We consider that this is rather a simplistic view of the marketplace. Typically real estate assets are held as part of balanced portfolio with different risk and reward profiles. The different classes of real estate perform different functions within these portfolios, and investors will have target rates of return for each class of asset. A reduction in rental value on one part of the portfolio as a result of rate changes does not necessarily reduce investment in that area as it may be offset by rises elsewhere

5.3

In such circumstances one of the best measurements of the impact of the economic incidence of on landlords is an assessment of potential development foregone. In other words, a rise in business rates will reduce the rents that a landlord is able to achieve and therefore reduce the potential level of new real estate investments by a sum equal to the value of the tax burden transferred from occupiers to landlords. This is a rule of thumb rather than an economic law, but it provides a useful impact measure which can be used in policy analysis.

5.4

A further potential role of business rates is the introduction of uncertainty into the analysis of value in property as an investment class (compared to other assets). If the incidence of business rates and changes in business rateable values every 5 years falls on property and land values then this will affect development returns. The less well rateable values follow actual rental values, the more noise will be introduced, and so uncertainty about development returns. This can readily be priced into investment decisions. But this risk factor, the large swings in rateable values relative to property values and so impact on property returns and rental growth, will tend to increase the risk profile of commercial property as an asset class.

Development Foregone 5.5

9

Our assumption is that the proportion of business rates increases that are capitalised into rental values may result in a net loss of reinvestment capital for the landlord or investor. Different proportions of business rate capitalisation (25%, 50%, 75% and 100% over a three year period) are modelled below, demonstrating potential reinvestment capital foregone. As approximately 57% of non-domestic properties in the UK are rented and would therefore be impacted by these capitalisation effects. Our approach is shown in Figure 5.1.

http://www.bpf.org.uk/sites/default/files/resources/PIA-Property-Data-Report-2015-are.pdf

12


Figure 5.1 Development Foregone Methodology


5.6

Table 5.1 illustrates the potential development capital foregone in the UK between 2011/12 to 2014/15 assuming 25%, 50%, 75% and 100% of the annual changes in business rates paid are capitalised over a three year period. If we assume over a three year period 75% of the value of business rate change is capitalised into rents, this equates to £0.7bn in development capital foregone between 2011/12 and 2014/15. The actual investment foregone may be considerably higher, as investors would have been able to procure additional debt capital secured against that £0.7bn. Table 5.1 Cumulative Development Foregone Nationally (2011/12 – 2014/15) 25% Capitalised Development Foregone

£220m

50% Capitalised £450m



Source: DCLG, National Non Domestic Rates Return 2014/15 net amount receivable from rate payers in England, Regeneris Consulting

5.7

The development capital foregone was also analysed for individual core cities and use classes. Table 5.2 illustrates the potential development capital foregone for the landlords of the top 50 most valuable properties in each core city and in each use class between 2011/12 and 2014/15. The model once again assumes that 25%, 50%, 75% and 100% of business rate changes are capitalised into rental values over a three year period.

5.8

Using Newcastle as an example, if we assume over a three year period that 75% of the value of business rate change (paid by the top 50 properties in each use class) is capitalised into rents, this equates to a cumulative loss of £1.4m in rental income and therefore development capital for landlords of those top 50 properties between 2011/12 and 2014/15. Again, the actual amount of investment foregone is likely to be higher once the ability to borrow is taken into account.

13


Figure 5.2 Development Capital Foregone by Core City and use class, top fifty properties only (2011/12 – 2014/15), £ms City Newcastle Leeds Sheffield Nottingham Manchester Liverpool Birmingham Bristol City Westminster Islington T. Hamlets

25%

50%

£0.48 £0.71 £0.57 £0.45 £1.10 £0.41 £0.71 £0.69 £1.58 £4.11 £0.45 £1.79

£0.96 £1.41 £1.13 £0.89 £2.21 £0.83 £1.41 £1.38 £3.17 £8.23 £0.90 £3.59

75% £1.44 £2.12 £1.70 £1.34 £3.31 £1.24 £2.12 £2.07 £4.75 £12.34 £1.35 £5.38

100% £1.92 £2.83 £2.27 £1.79 £4.41 £1.65 £2.83 £2.76 £6.34 £16.45 £1.80 £7.18

Source: BPF Members Business Rates Database, Regeneris Consulting

5.9

These are admittedly imprecise measures of potential impact. But they do underscore the message that business rate incidence is not a purely academic concern. If a proportion of business rate increases are capitalised into lower rents, then there is a likelihood that this will impact on the level of new funded commercial development. Given that our research suggests that the relationship between rents and business rates is stronger in the regions, it is the level of investment in regional areas that might suffer the most, should business rates increase sharply.

14


6. How Business Rates Affect Occupiers 6.1

From an occupier’s perspective, business rates are an important financial and administrative burden. This is true whatever the economic incidence of the tax. We know that business rates constitute around 40% to the cost of occupying retail and office property and have become a more significant cost in recent years. This is because business rates have increased at a much faster rate than rents, and, on average, in line with inflation. As the PIA’s Property Data Report suggests, the big divergence between retail sector rents and rates reflects the substantial uplifts in rateable values in the revaluation introduced in 2005 and 2010.10

6.2

As we have seen, the financial burden associated with the payment of business rates may be mitigated if changes are capitalised into rents. However, because the financial incidence never shifts, occupiers are left to navigate through the reliefs, policy changes and valuation appeals. Some of reliefs that have been added or amended since 2010 include:

6.3



Small business rates relief



Enterprise Zone relief



Reoccupation relief scheme



Retail and food and drink discount



New-build empty rates relief



Flooding relief.

Interestingly most of these reliefs are squarely aimed at occupiers either as incentives (Enterprise Zones) or to offset the perceived rates burden on smaller firms. Notwithstanding the administrative burden, the underlying policy assumption underpinning most of these changes is that business rates are not capitalised into rents and so land values but rather change occupiers’ total occupational costs and so business performance. We explore below the impact of various scenarios suggested by our technical analysis of incidence patterns.

Impact Assessment 6.4

Using national data on net business rates collected, the impact of changes in business rates on rental tenants can be measured in terms of jobs foregone. Figure 6.1 shows our methodology, which is a standard form of impact analysis used in policy studies.

6.5

This type of exercise can provide some useful metrics for policy and impact analysis, but is quite blunt in scope. In this case, the fundamental issue is that the increase in business rates will have been caused by an increase in rateable values and an expansion of the tax base (more properties). However, both will be subject to the capitalisation process as landlords and occupiers negotiate new lease terms and conditions.

10

http://www.bpf.org.uk/sites/default/files/resources/PIA-Property-Data-Report-2015-are.pdf

15


Figure 6.1 Rental Tenants Jobs Foregone Methodology


6.6

Between 2011/2012 and 2014/15 businesses renting their premises in England cumulatively paid an additional £1,520m in business rates. If 100% of these changes in business rates were capitalised into rents immediately, businesses renting their premises should face no real extra burden. However, previous research and our knowledge of institutional lease structures suggests that any relationship between rents and rates will tend to be subject to a timing lag.

6.7

Table 6.1 below illustrates the potential jobs foregone in the UK between 2011/12 to 2014/15 assuming 25%, 50%, 75% and 100% of the annual changes in business rates paid are capitalised over a three year period as opposed to being fully capitalised immediately.

6.8

If we assume over a three year period that 75% of the value of business rate change is capitalised into rents, this equates to a cumulative loss of £850m in business profit or income which is equivalent to 6,000 jobs foregone. Table 6.1 Changes in Business Rates and Rents Paid Nationally (2011/12 – 2014/15) 25% Capitalised Increase in Business Rates Paid (2011/12 – 2014/15) (1) Business Rates capitalised (2) Loss of business profit or income (a-b)

50% Capitalised

75% Capitalised

100% Capitalised

£1,520m

£1,520m

£1,520m

£1,520m

£220m

£450m

£670m

£890m

£1,300m

£1,070m

£850m

£630m

5,400 1,900 2,000

4,400 1,600 1,600

3,500 1,200 1,300

2,600 900 900

9,200

7,600

6,000

4,500

Potential Jobs Foregone (3) Retail Office Industrial Total Jobs Foregone (4)

Notes: (1) the increase each year on the previous year added together based on the 2014 proportion of 57% of properties being rented; (2) based on three year transition to the end point of capitalisation; (3) impact averaged over three years; (4) may not equal due to rounding

16


Using this analysis, we can conclude that a hypothetical £100m increase in business rates sustained over three years, capitalised at 75% over a three year period, would have the following impact on landlords and commercial property tenants over the course of those three years: 

£300m increase in total financial liability



£150m reduction in development capital for landlords, compared to if business rates were not capitalised into rental values



£150m loss in net profit for rental tenants equating to approximately 1,000 jobs foregone, compared to if business rates were fully capitalised into rental values immediately.

Our findings suggest that as time passes, a greater proportion of the increased financial liability will be passed on to landlords.

17


7. Issues for Policy Makers 7.1

From a policy perspective, the question of business rates incidence matters because it affects investment, employment and regeneration. From 1990 until now decisions on the level of the business rate multiplier and allowances have been made at a national level. From 2016 onwards, this will no longer be the case.

7.2

So what are the issues that policy makers, both national and now local, need to consider when designing their new approach?

7.3

First, policy makers need to recognise that despite the strict financial liability on occupiers to pay rates, business rates are not an occupation tax. The evidence suggests that over a period of three to four years, the majority of any change is likely to be capitalised into rents. This may have unexpected consequences on the appetite and ability of investors to fund new development. In contrast, because the tax burden is transferred from occupiers to property owners over a number of years, this means that occupiers pay a large proportion of any transition costs.

7.4

This has important implications for place-making and wider regeneration activity: 

Short term, time-limited reliefs will tend to benefits occupiers as they will not be capitalised into rents. To this end, the policy of announcing time limited reliefs and then extending them periodically will tend to focus the benefit of these reliefs on the occupier rather than landlord.



Conversely, reliefs that extend over time such as those in enterprise zones can be expected to be capitalised into rents and so ultimately land values. These reliefs do not therefore act as an incentive for occupiers. However, enterprise zones provide a number of incentives of which relief from business rates is only one, and arguably a fairly minor one at that.



The greater the divergence between rental values and rateable values over time the more this adds to the uncertainty and risk for investment in property as an asset class. Indeed arguably the effect of frozen rateable values during a period of very great turmoil in rental values (such as since 2008) will have added to the divergence of returns between locations and use classes.

7.5

Many of these unintended consequences of business rates would be removed or lessened by more regular revaluations that better align the rateable and rental values. The Netherlands’ model of annual revaluations has much to commend it in this regard.

7.6

Second, the evidence suggests that retail and retail warehouse properties may exhibit a closer relationship between rent and rates than other use classes. This means that any business rates rises that affect retail properties are likely to have less impact on occupiers’ economic activity, and a greater impact on landlords and investors.

7.7

Third, regional markets seem to show stronger relationships between rent and rates. This suggests that policies aimed at reducing business rates in order to foster regeneration will have a greater proportionate effect in these locations.

7.8

Fourth, our understanding of the London investment market remains incomplete. We believe that the statistical relationship between rent and rates in the capital is overpowered by wider macroeconomic forces and investor demand. This does not mean that business rate changes cannot be used in a regeneration context, but that more targeted research is needed

7.9

Finally, an investor’s appetite to bear an increased tax burden is likely to be stronger where increased charges will lead to improvements in the built environment or local transport

18

Business rates: Who Pays and Why it Matters? infrastructure. This is because investors may see a direct link between the increased tax burden, infrastructure improvements and future private returns.

19


Appendix A - Previous Literature Table A.1 Empirical Research into the Relationship Between Business Rates and Rents Study Who Pays Business Rates? Bond, Denny, Hall and McCluskey (1996)

Methodology and Sample Regressed estimated rental values (ERV) against business rates payable for institutionally-owned commercial properties between 1987 and 1992.

Findings and wider observations No statistically significant results linking business rates to rents for industrial or office properties. Large and statistically significant effects of increased rates on estimated rental values for retail properties.

Split data into three locations: London, South East (excl London) and Rest of England and Wales Total sample size: 2,964; retail sample: 1726, industrial sample: 334; Office sample: 904.

Increases in non-domestic rates put downward pressure on commercial property rents. However, takes several years to adjust and long run effects are uncertain.

Additional variables used: year dummies, county employment vacancy rates, district unemployment rates, proportion of long term employment in the district. The Impact of Tax Incentives on Local Real Estate Markets: the question of incidence Bond, Gardiner and Tyler

Examined effect of NNDR on rents at a time when properties in Enterprise Zones received full exception from paying NNDR. Data was derived from lease payments of properties on and off the EZs.

Initial sample size of 4,861 - produced no statistically significant results. Reduced to leases which commenced during operation of enterprise zones (2,214) - produced no statistically significant results. Final sample was for properties constructed in the 1980’s, (203) produced results significant at 10% level The Data collected for business properties in capitalisation six London boroughs. Cross-sectional data of business and longitudinal tests was carried out. rates: An empirical Matched pair data collected for properties study of tax in adjacent local authorities. Properties incidence in were of a comparable size and quality but six London had different rate burdens at the time of boroughs. study. Mehdi (2003) Time Frame: 1973-1988 Matched Pair sample size: Industrial (35), Retail (21), Office (21)

Occupiers of business properties likely to be main beneficiaries of temporary provisions such as transitional relief schemes. Main beneficiaries in the long run of any permeant reduction in business rates will be the property owners.

A large part of tax savings are captured in higher rents by landlords. Reductions in local property taxes as part of a local area regeneration package feed through into changes in rents. Average capitalization effect was not significantly different from 100% implying all of the local tax exemption benefits accrue to the owners of the property.

Results confirm property tax differentials in the six London boroughs were shifted to rental values and thus capitalized into property values. Results confirmed significance for all types of properties between 5-9% significance Over the long term is it likely that the full amount of the UK property tax is capitalised and the tax is borne by the owners of property rather than leasehold occupiers.

Source: Bond. S, Denny K, Hall, J, McCluskey, W, (1996), Who pays business rates? Fiscal Studies, Vol. 17, No. 1, pp.19-35. Tyler, P, Bond, S, Gardiner, B. (2012). The Impact of Enterprise Zone Tax Benefits on Local Property Markets in England: Who Actually Benefits. Journal of Property Research. Mehdi, M, (2003) The Capitalisation of Business Rates: an Empirical Study of Tax Incidence in Six London Boroughs, unpublished PhD, London School of Economics and Political Science.

A-1


Appendix B - Date and Sample Definition B.1

We used two sources of data in our work. Full details of the data and definitions are provided below.

B.2

Data on Business Rates was obtained from a database compiled from British Property Federation members. This dataset provides the rateable value for all non-domestic properties across the UK. The dataset includes the rateable value for each revaluation year from 1990 to 2010, including any alterations as a result of appeals.

B.3

Annual data on changes in rental values for institutionally-owned commercial properties was obtained from the Investment Property Databank (IPD). The dataset provides annual data from 1981 to 2014 disaggregated by local authority area, county and region. The dataset includes information on retail, retail warehouse, office and industrial premises.

B.4

Data was extracted for the following locations, use classes and years: 

Core Cities: Newcastle, Leeds, Manchester, Sheffield, Nottingham, Birmingham, Liverpool, and Bristol



London Markets: Tower Hamlets, Westminster, Islington and City



Use Classes: Retail, Office and Retail Warehouse



Time Period: 1990 to 2014, inclusive

B.5

The choice of locations was informed by two factors. The first was the nature of the current policy debates surrounding devolution, which center on the core cities as a possible first wave of new locally independent areas. The second was data availability. The Core Cities are reasonably mature markets for investment properties, and offered what we believed to be the best chance of securing a comprehensive rental database.

B.6

This strategy and the relevant sample sizes are summarised in Table B.1 below.

B-1


Table B.1 Number of properties sampled per year (1990-2014) by core city location

Source: MSCI IPD UK Key Centres 2014 Database, BPF Members Business Rates Database, Regeneris Consulting

B.7

The data was extracted on two bases. The first basis produced what is described as our Prime Sample. The Prime Sample is based on the fifty most valuable properties, in each year, in the business rates database. This is used to calculate average (median) levels of rates paid annually. The properties in the Prime Sample will therefore vary each year, but should always reflect the top 50 most valuable commercial properties for each use class and each location.

B.8

By contrast, a second data series was extracted which we describe as the Constant Sample. The Constant Sample tracks the rates paid for the same properties in each year, location and use class in the Business Rates database. The properties in the Constant Sample are the same in each year, and allow us to examine the relationship between business rates and rents on a more consistent basis.

B.9

Table B.2 below illustrates an example of the data extracted from the business rates database for an individual retail property in Newcastle which was included as part of the Constant Sample.

B-2


Table B.2 Newcastle Retail Property – Business Rates Payable 1990 - 2015 Property A: Newcastle Retail 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Rateable Value £133,500 £133,500 £133,500 £133,500 £133,500 £200,000 £200,000 £200,000 £200,000 £200,000 £276,000 £276,000 £276,000 £276,000 £276,000 £320,000 £320,000 £320,000 £320,000 £320,000 £265,000 £265,000 £265,000 £265,000 £265,000 £265,000

Business Rates Multiplier 0.348 0.386 0.402 0.416 0.423 0.432 0.449 0.458 0.474 0.489 0.416 0.43 0.437 0.444 0.456 0.422 0.433 0.444 0.462 0.485 0.414 0.433 0.458 0.471 0.482 0.493

Source: BPF Members Business Rates Database

Data Extracted B.10

From the IPD dataset, annual initial yield and capital value data was extracted from 1990 to 2014 for three use classes; retail, retail warehouse and office. This data was extracted for nine Core Cities; Birmingham, Bristol, Leeds, Liverpool, Manchester, Newcastle, Nottingham, Sheffield and London.

B.11

From the rateable value database, address, square footage and rateable value data for each revaluation year including any subsequent alterations, was extracted for each of the core cities and for each of the three use classes; retail, retail warehouse and office. As the database contains data for all non-domestic properties throughout the UK, only data for the top 50 properties with the highest rateable value for each use class, core city and rateable value year was extracted.

B-3


Data Transformation B.12

Using a combination of IPD and rateable value data, a dataset for each core city and use-class was developed. The data was only collected for the top 50 properties with the highest 1990 rateable values, the other properties were discarded. The properties with the highest rateable values were used to better reflect the property stock contained in the IPD database, which only contains institutionally-owned commercial properties. These institutionally owned properties are often high value, large properties and are therefore likely to have some of the highest rateable values within any one city. This method was repeated for the 1995, 2000, 2005 and 2010 revaluation years, resulting in five different datasets for each city and use-class.

B.13

This method was used to ensure that the 5 yearly datasets accurately represented the property stock in existence at that time. The isolated data more accurately represents the IPD rental data to which it will be compared, given the IPD property sample changes on a yearly basis to reflect the changing property stock.

B.14

Any subsequent revaluations as a result of appeals were also documented in the data. The rateable values were then multiplied by the annual business rates multiplier, to calculate the business rates paid on a year-by-year bases from 1990 to 2015.

B.15

Median annual business rates paid were then compared to rental data from the IPD database. This methodology was then repeated for each core city and for each use class.

B-4


Appendix C - Detailed Results Stage One: Descriptive Analysis C.1

The first stage of the analysis was to interrogate the raw datasets to identify problems which may have a disproportionate effect on later modelling. This exercise revealed a number of difficulties related to IPD data coverage. The major issues were: 

Office data for Tower Hamlets displays a unique pattern, which is probably related to the inclusion of certain properties and portfolios from Canary Wharf. This means that the suitability of this data for inclusion in the wider econometric investigation needed to be considered



For some core cities and use classes, a fully comprehensive data series covering all years was not available from IPD. In these cases, regional values were substituted for city based estimates.



The retail warehouse class had the poorest overall coverage in the IPD data. For some locations, city data was simply not available, especially towards the start of the time series. Again, regional values were used to fill any gaps.

C.2

Once collected, the data was used to generate estimates of year on year changes in business rates and rents. This data was then plotted for each location and for each use class from 1991 to 2014. This exercise was completed for the prime sample, as we were interested in the performance of the best performing and most valuable assets. This means that the sample of properties included in the analysis changes after each revaluation (every 5 years). Given the nature of the expected relationship between rents and rates, our a priori assumption would be that capitalisation of business rate changes would occur over a number of years. Accordingly we plotted one, two and three lags year for rental changes in each use class and location. The full set of graphs is provided in Appendix F.

C.3

We include below an example drawn from the Newcastle retail sector.

C-1


Figure C.1 Newcastle Retail (t+1)




C.4

Figures C.1, C.2 and C.3 show the relationship between the change in business rates and the change in rents in the Newcastle retail market, with a one, two and three year lag.

C.5

The expected inverse relationship between rents and rates becomes easier to identify in the later graphs, but the relationship is by no means straightforward. However, closer examination of these graphs and the other plots in Appendix F do highlight a number of general points.

C.6

First, the choice of lag can be critical. On the one Source: Regeneris Consulting hand, previous research and our knowledge of institutional lease structures suggests that any changes in business rates will take time to feed through to rent levels. Indeed, the strongest implied relationships appear to be in Years 2 and 3. There are some exceptions (see Bristol Office 1 Year Lag), and in other cases (Bristol Retail Warehouse, Year 2 to Year 3) an additional twelve month lag can imply a reversal of the expected pattern.

C.7

In order to examine the lag patterns in more detail we calculated simple correlation coefficients between business rates and rents, lagged from 1 to 2 years. This analysis showed stronger associations in the second year, consistent with our a priori assumptions and the existing literature.

C.8

Second, patterns in Retail and (to a lesser degree) Retail Warehouse appear to be stronger than in the Office sector. There is also a regional dimension to this pattern, with regional markets showing stronger relationships than London. Examples of strong relationships in regional markets include Manchester Retail and Retail Warehouse, Birmingham Retail & Liverpool Retail.

C.9

In general, data for the four London markets imply a weaker relationship between rent and rates. At this stage, it impossible to say whether this is a real effect, or the result of data inadequacies. It may be that in London properties are subject to stronger macro and wider pressures, which cannot be seen in a simple data plotting exercise. This could be controlled for in a more structured manner, using the econometric techniques discussed below. Possible other explanations for the disparity between London and the rest of the UK could be related to lease structures. The 2015 UK Lease Events Review notes that although average lease lengths continue to grow (20% since the bottom of the market), the larger tenants that tend to dominate the London market tend to prefer longer leases. The report also notes that as demand has increased, the London office space market has

C-2


Business rates: Who Pays and Why it Matters? turned from a tenants to a landlord’s market.11 This means that larger tenants have been increasingly pressurised into accepting less flexible lease terms. The disparity in lease structures could go some way in explaining the discrepancies between regional markets and London. C.10

For clarity, and to verify the findings of our analysis for the Prime Sample, we also produced a number of similar plots for the Constant Sample. This time we plotted the growth of estimated rental value, against business rates paid for selected geographies and use classes. Each bar represents an individual property. In order to better see the effect of revaluation episodes, we also excluded annual variations related to changes in the business rates multiplier. Figure C.4 City of London Office Rate and Rental Growth , Constant Sample 100

30

80

10

20

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

-20

0

1991

0

1990

Rateable Value £m

40

ERV Growth (%)

20

60

-10

-40 -60

-20

-80 -100

-30

Source: MSCI IPD UK Key Centres 2014 Database, BPF Members Business Rates Database, Regeneris Consulting

C.11

In figure C.4 the cumulative rateable value for the same sample of 50 properties is plotted against estimated rental value growth. The diagram is instructive in highlighting macroeconomic shocks and the related influence of property market cycles. The plot suggests that for this group of London office properties, rental values seem to have fluctuated primarily in tandem with movements in the wider economy. On a practical basis, this means that isolating the effect of business rate changes on rental changes will only be possible if we are able to control for wider macroeconomic effects. This is where wider econometric analysis may be useful.

Stage Two: Econometric Analysis Existing Models C.12

11

The purpose of econometric modelling is to examine the relationship between business rates and rents while controlling for other factors that might determine rent change (e.g. wider economic circumstances). The first stage of this work is to revisit the existing literature, and to apply the model specifications and techniques used by previous investigators to our new data sources. This will help us to assess whether current policy assumptions are based on a sound footing.

MSCI, Strutt & Parker (2015) UK Lease Events Review, November 2015, pp5-6.

C-3


Stage One Analysis C.13

Our starting point was the approach used by Bond et al (1996). Their model regressed the estimated rental value per square foot for a property on the estimated rental value per square foot for that property in the previous year, the rates bill per square foot for that property in the current year, and the rates bill per square foot for that property in the previous year. Additional variables included to control for other factors were county vacancy rates, district unemployment rates and the proportion of long-term unemployment in the district.12

C.14

After much experimentation, we were able to replicate a simpler version of this analysis using the data across the three use classes. Our expectations were that for each property type, the coefficients relating rents to rates would be negative (implying an inverse relationship) and statistically significant at the 10 per cent level. The specification and results for our retail sample are shown below: Rentit = α+β1Gdpgrt+ β2Unemt + β3Vacantt+ β4BRt+1 +β5BRt+2. . .β7BRt+4 + εit GDPgr Unem BRt+1 BRt+2 BRt+3 Brt+4 Number of obs. Wald Chi2

5.04* (2.88) 3.35 (4.36) 0.01 (0.09) -0.14 (0.08*) -0.14 (0.08*) -0.03 (0.08) 160 74

**significant at 5% *significant at 10% Standard errors are in parentheses C.15

Our analysis for retail warehouse (sample size of 200 observations) failed to discover anything of significance, and the coefficient signs did not conform to expectations (positive in the majority of cases). The same was true for offices which were tested with London properties included and excluded. By contrast the nature of the relationship between rents and rates in retail was as expected, and statistically significant for Year 2 and 3 lags. This means that we found evidence for a reasonably strong causal relationship, implying that business rate changes do exert a strong and observable influence on rents (in the retail sector at least). These findings were entirely in keeping with the previous work, which had identified similar relationships between rents and rates in the retail sector (but not in other sectors). The findings also matched our tentative observations derived from the earlier plotting exercise.

C.16

The validity of these results could be questioned for several reasons. The most important reservation is the technical difficulty involved in estimating a panel model with lagged (dependant) variables. The standard approach now is to use what is called an Arellano-Bond dynamic panel-

12Estimation

was by instrumental variables after taking first differences to control for unobserved property-quality attributes.

C-4

Business rates: Who Pays and Why it Matters? data estimation.13 We therefore estimated a new model on this basis, and added a further variable that interacted rates and labour markets. The principle was that if employment vacancies decline and rates decline there will be a greater level of upward pressure on rents. This seems to have had the expected effect and also drove a greater degree of significance in the other estimated coefficients. Growth in GDP, declines in vacancy rates and the interactive variable were all significant for retail properties. But although rates were of the expected sign, we could not find a credible specification that gave a statistically significant result beyond the existing 10% level.

New Models C.17

The second stage of our econometric work attempted to model the relationship between rent and rates using new specifications and approaches. Our first change was to use to a Generalised Least Squares (GLS) estimator. GLS is a technique for estimating the unknown parameters in a linear regression model. GLS can be used to perform linear regression when there is a certain degree of correlation between the explanatory variables (independent variables) of the regression. In these cases, ordinary least squares and weighted least squares can be statistically inefficient, or even give misleading inferences.

C.18

We next considered the composition of the independent variables. We decided to include a new variable to cover the annual change in the business rates multiplier. This created interpretative problems, and we therefore included another an additional variable (change in median rental value) but excluded the change in median rates. The model specification was therefore: Rentit = α+β1Gdpgrt+ β2Unemt + β3Vacantt+ β4Dubrt + β5Drvalit+ εit

Where: Rentit = % change in average rent paid in city i for year t Gdpgrt = % Growth in real GDP for year t Unemt = % Change in unemployment rate for year t Vacantt = % Change in employment vacancy rate for year t Dubrt = % Change in the uniform business rate for year t Drvalit = % Change in the median rateable value for city i for year t C.19

13

Random effects panel estimators were used in each case. Estimates were generated for offices (excluding London), offices (excluding Tower Hamlets), retail (all cities), retail (excluding London), and retail warehouse (all cities). The results are shown below

See Cameron, A. Colin; Trivedi, Pravin K. (2005). "Dynamic Models". Microeconometrics: Methods and Applications. New York: Cambridge University Press. pp. 763–768.

C-5


Figure C.5 Estimation Results (Dependent Variable: Rental Value Growth)

GDPgr Unem Vacant DUBR Drval _cons Number of obs. Wald Chi2

Office excl. London

Office excl. Tower Hamlets

Retail

Retail Warehouse

2.88** (1.32) 5.41* (2.92) -0.07 (0.09) -0.21 (0.33) 0.11 (0.07) 0.01 (0.07) 184

2.27** (1.05) 1.80 (2.31) -0.13* (0.08) -0.47* (0.28) 0.02 (0.05) 0.02 (0.03) 253

2.00** (-0.72) 0.64 (1.57) -0.07 (0.05) 0.29 (-0.22) 0.06 (0.05) 0.02 (0.02) 230

1.77** (0.87) 0.25 (1.93) 0.02 (0.06) 0.21 (0.24) -0.01 (0.05) 0.03 (0.02) 184

9.66

13.00

13.18

9.02

**significant at 5% *significant at 10%, Standard errors are in parentheses C.20

As one would expect, the macroeconomic variable GDPgr, is of the right sign and statistically significant in all cases. This shows how important macroeconomic conditions are in determining rents. The change in the median rateable value for each city location is not significant, and in the case of most specifications, produces the wrong sign. However, we do find statistically significant results for the change in the uniform business rates multiplier in one of the Office subsets. This suggests that annual variations in the multiplier do feed into rents for these properties. The model implies that the larger revaluation events have a weaker influence and that rents tend to track wider macroeconomic forces. These specifications show that there is some evidence for office space that changes in the UBR are capitalised. It also suggests that this model specification is a closer fit to office, rather than retail properties.

C-6


Appendix D - Correlation Co-efficients Table D.1 Correlation Between Business Rates and Rental Values Location Birmingham Birmingham Birmingham Bristol Bristol

Use Class

t 0.03 0.16 -0.18 -0.12 -0.21

t-1 0.01 0.28 0.10 0.22 0.08

t-2 -0.10 -0.00 0.37 -0.31 0.29

t-3 0.02 0.18 -0.04 -0.13 0.13

Office Retail Retail Warehouse Office Retail

Bristol Leeds Leeds Leeds Liverpool Liverpool

Retail Warehouse Office Retail Retail Warehouse Office Retail

-0.11 -0.11 -0.08 0.10 0.12 -0.00

-0.30 0.10 0.13 -0.16 0.25 -0.09

0.18 -0.03 -0.11 0.03 -0.07 -0.16

0.26 -0.16 0.18 0.28 0.04 -0.15

Liverpool Manchester Manchester Manchester Newcastle upon Tyne Newcastle upon Tyne


-0.06 0.20 -0.17 -0.07 0.38 0.20

0.12 0.21 0.25 0.17 -0.22 -0.25

-0.02 -0.13 -0.19 -0.12 -0.03 -0.01

0.61 -0.21 0.37 0.18 -0.11 0.31

Newcastle upon Tyne Nottingham Nottingham Nottingham Sheffield Sheffield


-0.04 -0.03 0.22 0.11 -0.30 0.21

0.03 -0.26 0.31 -0.40 0.33 -0.05

-0.14 0.10 0.09 0.51 -0.11 0.21

-0.26 -0.00 -0.01 0.05 0.23 -0.41

Sheffield City of London City of London Islington Tower Hamlets Westminster

Retail Warehouse Office Retail Office Office Office

0.22 -0.02 0.22 0.00 0.00 0.06

-0.01 0.17 -0.40 -0.41 -0.08 0.45

0.15 -0.06 0.12 -0.35 0.02 -0.21

0.19 -0.21 0.07 0.01 0.96 0.07

Westminster

Retail

0.37

0.04

0.20

-0.26

D-1


Appendix E - Case Studies and Wider Trends Case Studies E.1

The figures below illustrate the annual business rates paid from 1990 to 2015 for a retail property in Newcastle and a prime office property in the City of Westminster, London.

E.2

The retail property in Nottingham experienced significant increases in businesses payable during during four out of the five revaluation periods. The most significant increase was as a result of the 1995 revaluation which increased the property’s business rates payable by 53% between 1994 and 1995. The 2010 revaluation episode together with the fall in the UBR from 0.485 to 0.414 caused the overall business rates payable (before relief) for the property to decrease by almost 30% between 2009 and 2010. This trend was not untypical for Newcastle retail properties with almost three quarters of the 50 most valuable retail properties in Newcastle experiencing a fall in business rates payable between 2009 and 2010. Table E.1 Specimen Newcastle Retail Property – Business Rates Payable 1990 - 2015 Property A: Newcastle Retail 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Rateable Value £133,500 £133,500 £133,500 £133,500 £133,500 £200,000 £200,000 £200,000 £200,000 £200,000 £276,000 £276,000 £276,000 £276,000 £276,000 £320,000 £320,000 £320,000 £320,000 £320,000 £265,000 £265,000 £265,000 £265,000 £265,000 £265,000

Business Rates Multiplier 0.348 0.386 0.402 0.416 0.423 0.432 0.449 0.458 0.474 0.489 0.416 0.43 0.437 0.444 0.456 0.422 0.433 0.444 0.462 0.485 0.414 0.433 0.458 0.471 0.482 0.493

Source: BFP Members Business Rates Database

E.3

In contrast to the Newcastle retail property, the office property in Westminster experienced significant decreases in businesses payable (before relief) during four out of the five revaluation periods. The most significant decrease was as a result of the 1995 revaluation which decreased the

E-1

Business rates: Who Pays and Why it Matters? property’s business rates payable by 66% between 1994 and 1995. The properties business rates bill peaked in 1994 before the revaluation when the occupier had to pay £629,191. The most recent business rates bill for the property was £415,530, 34% lower than it was at its peak. This pattern is consistent across office properties in the other three London Boroughs we analysed, (Tower Hamlets, The City of London and Islington) which all experienced a significant drop in their business rates bills between 1994 and 1995 with those in the City of London also paying lower business rates today than they were in 1994. Table E.2 Westminster Office Property Property B: Westminster Office 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Rateable Value £1,487,450 £1,487,450 £1,487,450 £1,487,450 £1,487,450 £492,235 £492,235 £492,235 £492,235 £492,235 £921,400 £921,400 £921,400 £921,400 £921,400 £825,000 £825,000 £825,000 £825,000 £825,000 £810,000 £810,000 £810,000 £810,000 £810,000 £810,000

Business Rates Multiplier 0.348 0.386 0.402 0.416 0.423 0.432 0.449 0.458 0.474 0.489 0.416 0.43 0.437 0.444 0.456 0.422 0.433 0.444 0.462 0.485 0.434* 0.453* 0.478* 0.491* 0.502* 0.513*

Source: BFP Members Business Rates Database *Business Rate Supplement for Crossrail = 2p for every £1 of rateable value for properties with a rateable value > £55,000

Wider Trends E.4

When analysing how business rates payable have changed for prime commercial properties from 1990 – 2015 across different cities and different use classes, some clear trends emerge. 

The past trajectory of business rate changes in London are very different and distinct from patterns found in the regional markets. 

Between 1994 and 1995 business rates payable (before relief) substantially decreased in four key London Boroughs (City of London, City of Westminster,

E-2

Business rates: Who Pays and Why it Matters? Islington and Tower Hamlets). This is in contract to the other eight core cities which all experienced significant increases in rateable values during this time. 

Between 1999 and 2000, the market in London once again dislocated from the other core cities as business rates substantially increased during this revaluation episode versus elsewhere where the majority of business rate bills decreased.



In the 1995 revaluation episode dislocations between retail and office properties emerged with the majority of offices business rates bills decreasing and the majority of retail property’s business rates bills increasing.



Despite the 2010 rateable values being based on 2008 rental estimations at a time when the property market was considered to be at its peak, for the top 50 most valuable commerical properties across use classes and cities, the vast majority experienced a decrease in their business rates bills between 2009 and 2010. The exception to this trend was in Nottingham, Newcastle and Islington.

E-3


Appendix F - Regional Datasets by Use Class: Prime Sample F.1

The following diagrams show the annual changes between business rates and rents for our Prime Sample of properties across geographies and use classes. The annual changes in business rates are lagged over 1, 2 and 3 years.

F-1


Birmingham Figure F.1 Retail (Rent – 1 Year)

Figure F.2 Retail (Rent – 2 Year)


Figure F.4 Retail Warehouse (Rent – 1 Year)



Figure F.7 Office (Rent – 1 Year)



F-2


Bristol Figure F.10 Retail (Rent – 1 Year)









F-3


Leeds Figure F.19 Retail (Rent – 1 Year)









F-4


Liverpool Figure F.28 Retail (Rent – 1 Year)









F-5


Manchester Figure F.37 Retail (Rent – 1 Year)









F-6


Newcastle Figure F.46 Retail (Rent – 1 Year)









F-7


Nottingham Figure F.55 Retail (Rent – 1 Year)









F-8


Sheffield Figure F.64 Retail (Rent – 1 Year)









F-9


City of London Figure F.73 Retail (Rent – 1 Year)






F-10


Westminster Figure F.79 Retail (Rent – 1 Year)






F-11


Islington Figure F.85 Office (Rent – 1 Year)





Tower Hamlets Figure F.88 Office (Rent – 1 Year)

F-12


Appendix G - Regional Datasets by Use Class: Constant Sample G.1

The following diagrams show the effects of the five year revaluation episodes on our Constant Sample. Each colour in the bar chart represents an individual property’s rateable value which remains constant until a new revaluation occurs (1990, 1995, 2000, 2005, 2010). The blue line graph represents the estimated rental value for institutional properties for each city and use-class.

Appendix G - G-1


Nottingham Figure G.1 Office

Figure G.2 Retail

Figure G.3 Retail Warehouse

Birmingham Figure G.4 Retail


G-2


Newcastle Figure G.6 Office

Figure G.7 Retail


Figure G.10 Retail


Liverpool Figure G.9 Office

G-3


Leeds Figure G.12 Office

Figure Table G.12 G.1Office

FigureTable G.13G.2 Retail Warehouse

Sheffield Figure G.14 Office

Figure G.15 Retail


G-4


Manchester Figure G.17 Office

Figure G.18 Retail


Figure G.21 Retail


Bristol Figure G.20 Office

G-5


City of London Figure G.23 Office

Figure G.24 Retail

Islington Figure G.25 Office

G-6


Tower Hamlets Figure G.26 Office

G-7

Regeneris Consulting Ltd Manchester Office 4th Floor Faulkner House Faulkner Street, Manchester M1 4DY 0161 234 9910 [email protected] London Office 3rd Floor, 65 St. John’s Street. London EC1M 4AN 0207 336 6188 [email protected] www.regeneris.co.uk