Compared to authorities in other capital cities of developed economies around the world, the Mayor of London and the London boroughs have very limited powers over taxation.
Nearly 70 per cent of London’s revenue comes from central government, compared to 26 per cent in New York, 16.3 per cent in Paris, and 5.6 per cent in Tokyo. The city has very limited fiscal autonomy. National government sets the tax base and bands for Council Tax, and both sets and levies non-domestic business rates and property transaction taxes. There are no London-specific rates of income or sales taxes, or locally-assigned revenues. London’s fiscal powers operate at the margins: setting the level of Council Tax and the Metropolitan Police precept, within clearly defined limits; charging a supplementary non-domestic levy for Crossrail; and operating the Congestion Charge.
These constraints on London’s fiscal powers limit its ability to respond to changing economic and social needs. It cannot levy new taxes to fund the infrastructure it needs to grow, or properly capture the uplift in land values that is created by public investment in transport or local amenities. It is constrained from taxing the wealth and income generated in the city in order to reduce London’s high levels of inequality and poverty, or to provide more and better housing. It is not accountable for the majority of the taxes that are levied on its residents and businesses for the public services they receive, and by the same token, it cannot propose different or better ways of raising those taxes.
The political pressures facing national governments – and the political risks national politicians are prepared to take – effectively determine the scope for fiscal policy changes in London.
Reform of London’s property tax system is particularly pressing: the current system is inefficient and unfair. Council Tax is regressive in that more expensive properties benefit from a much lower tax rate, and provides no incentive for efficient use of land or property. The tax base is also outdated: bills are calculated using 1991 property values, meaning the proportion of the property value that is charged has halved on average across London, and even more in areas where prices have risen the fastest. This has weakened the incentives to use property more efficiently. Different approaches to tax policy across boroughs have also built up wide variations over time – a band D property in Barking and Dagenham is liable to higher Council Tax than the most expensive properties in Westminster.
Stamp Duty Land Tax is more progressive, but only taxes property wealth at the time of transaction, inhibiting transactions and liquidity. Additionally, our property tax system is very poor at capturing land value uplifts: it allows for nearly all the property wealth generated by public investment to flow to landowners untaxed. Transport for London estimated that its rail projects would yield twice their cost in (uncaptured) land value uplifts.
The case for increasing fiscal devolution to London has been made in the two reports of the London Finance Commission, a body of independent experts and interested parties convened by the Mayor of London. In its most recent report, Devolution: A Capital Idea, the Commission argued that the power to make long term commitments on infrastructure, housing, and employment policy would best enable London to meet the challenges of uncertainty caused by Brexit. It proposed the devolution of the full suite of property taxes to London, including Council Tax, business rates and Stamp Duty Land Taxes; the assignment of a proportion of income tax and VAT yields to London; the creation of new powers for London to levy business rate supplements, charges and minor taxes; and consideration of a new charge to tax the uplift in property value created by new transport or regeneration schemes, such as tube line extensions. In a recent report, Transport for London sketched out how such a charge could work. A levy on new buyers and renters of properties around new transport links would enable capture of some of the value uplift resulting from improvements in connectivity, and its reinvestment into further transport upgrades.
The case for greater fiscal devolution to London has been strengthened in recent years by the transfer of powers over property and income tax powers to the governments of Scotland and Wales. The Scottish Government has levied its own Land and Buildings Transaction Tax and Landfill Tax since 2015, and a Scottish rate of income tax has been in place since 2016. From this year onwards, Scotland will have full powers over tax rates and bands on non-savings and non-dividend income, and will receive a half share of VAT receipts in Scotland.
The Welsh Government now has power over Stamp Duty Land Tax, landfill tax and business rates, and will set a new Welsh Income Tax rate from 2019. Both the Scottish and Welsh governments have also introduced marginal reforms of Council Tax. There are neither principled nor practical reasons for not extending a similar measure of devolution to London.
In what follows, we model some of the options proposed by the London Finance Commission, and consider a series of incremental reforms to taxation in London. We analyse the effects of potential Council Tax reform; examine the revenue raised by Stamp Duty Land Tax under different scenarios for the London housing market; and sketch out what a tax on land value in London could look like.
Council Tax Reform
Council Tax is levied in nine bands, based on 1991 property values. Each London borough sets tax levels annually to meet its funding needs (after taking grants into account), as well as costs for the GLA, and Police and Fire Services. The ratio between the bands is fixed: the tax charge for the highest value properties (Band H) within any local authority is fixed at three times the tax rate for the cheapest properties (Band A).
Numerous experts have recommended that Council Tax be reformed. Because of its exceptional levels of house price growth, the issue is most pressing in London. In research to support the London Finance Commission, the Greater London Authority modelled in detail a dozen reform options that would be at least revenue-neutral if they followed devolution to London. The options are summarised in the box below.
Options for Council Tax reform
Properties that have gained most in value since 1991 pay more (they are mostly in inner London); properties that have gained least in value pay less tax. This does not change to the way Council Tax is set. The tax is still regressive.
1+2 Adding (up to 10) more bands
The more expensive properties are placed in new bands that are charged higher rates than in the current top band H, though the ratio between existing bands is retained. The tax becomes less regressive and Council Tax bills increase by larger amounts for more expensive properties. But the tax is still regressive, i.e. higher value properties still pay a lower proportion of their property value in Council Tax than the cheaper properties.
1+2+3 Changing the ratios between bands
This would be required to make Council Tax progressive. In the example provided by the GLA, the most valuable properties would pay three times more than band D properties – instead of a third more today. At the opposite end, the cheapest properties would pay three times less.
4 Introducing a rate-based tax
Instead of bands, the tax would be charged as a proportion of property value/rental value. The tax is neutral in distributional terms since all households pay the same proportion of their property value in Council Tax.
Centre for London has modelled a similar system for domestic property to that currently used for business rates – that is, charging Council Tax as a proportion of the annual rental yield of the property. We modelled revenue in the London Borough of Hackney – one of the boroughs where property values have risen most rapidly since the 1990s. In Hackney, taxing seven per cent of the annual rental yield (assumed to be four per cent of property value) sustains revenue at current levels, and equates to a 0.27 per cent tax on property values.
This level of tax would keep Council Tax bills at the same level for the median band D property. Lower value properties would see their tax bill fall; higher value homes would pay more – as set out in the table below. These findings are comparable with the GLA’s: they suggested property values be taxed at 0.2 per cent London-wide to raise the same revenue as current Council Tax.
Council Tax bills in the London Borough of Hackney, under the current and a rate-based system
|Band||1991 Banding||Number of properties in band (2015)||2015 Council Tax bill (£)||Bill based on taxing 7% of rental yield – for the median property in each band (£)|
|A||Up to £40,000||6,730||863||599|
|B||£40,001 to £52,000||31,520||1,007||827|
|C||£52,001 to £68,000||33,200||1,151||1,035|
|D||£68,001 to £88,000||21,210||1,294||1,294|
|E||£88,001 to £120,000||10,800||1,582||1,689|
|F||£120,001 to £160,000||4,100||1,870||2,344|
|G||£160,0001 to £320,000||1,110||2,157||3,541|
|H||More than £320,000||50||2,589||N/A|
Source: Centre for London estimates
Because Council Tax has not been reformed since its creation, changes to the way it is operated will inevitably shift the tax burden between households and across boroughs – with gains and losses greater as the tax is made more progressive. Carefully designed transitional arrangements would be required to make the reform more politically palatable, in particular for lower income people in higher value properties, who might be able to defer payments until properties were sold.
There would also be significant governance issues. For Council Tax to be operated more effectively and fairly, the GLA would need to play a greater role in setting London-wide bands and tax rates. A system redistributing revenue between boroughs will also be needed to ensure revenue matches council tax requirements.
Devolving Stamp Duty Land Tax to London
The devolution of stamp duty powers to Wales and Scotland in recent years provides clear precedents that London can follow. Reforms enacted in 2014 to remove the “slab” structure of Stamp Duty in England and Wales mirrored the model established in Scotland, demonstrating the policy innovation that can come from devolution.
Nearly half of all Stamp Duty paid on residential property transactions is levied in London. In recent years, some 98 per cent of properties sold in London were sold for more than £125,000; 85 per cent of properties were sold for more than £250,000; and about 4 per cent of properties were sold for more than £1.5 million. Property sales over £925,000 represent 11 per cent of total sales but half of Stamp Duty receipts. Stamp Duty receipts are therefore highly contingent on higher value property transactions.
Catherine Barnaby from the University of Bath modelled the volatility of Stamp Duty receipts for different housing market scenarios, in order to measure the future volatility of the tax receipts. These are set out in the box below.
Scenario testing Council Tax receipts
Scenario 1 – the market develops in line with trends over recent years
If total sales follow trends since 2010, and proportion of properties sold in different price bands follows the patterns of the last twenty years, then Stamp Duty revenues in London will rise from £3.8 billion in 2017 to £5.38 billion in 2021 (2016 prices).
Scenario 2 – the market is flat
If the total volume of sales and the prices of properties are stagnant, revenues fall from £3.35 billion in 2017 to £3.11 billion in 2021 (2016 prices).
Scenario 3 – there is a housing market crash, with volumes of sales falling by 44%, as they did in 2008-9
We modelled one year’s worth of revenue loss. If properties sold by price band broadly mirror recent trends, a hypothetical crash in 2017 would see revenues fall to £1.59 billion – about 47 per cent of 2016 revenues. If prices fall more sharply for all properties above £250,000, revenues drop to £1.19 billion, about 35 per cent of 2016 revenues.
These projections demonstrate the likely magnitude of the risks and rewards – on the upside and downside – to both central government and London government in devolving stamp duty receipts. Formula arrangements would be necessary for determining the fiscal relationship between London and the Treasury to apportion these risks and rewards, as with Scotland and Wales, and ensure that regions outside London do not lose revenue.
Land value based taxes for London
Even with reform, Council Tax, Stamp Duty and Business Rates remain taxes on property – as such they would only partially incentivise a more efficient use of land, a particularly scarce commodity in London.
The Institute of Fiscal Studies’ 2010 Mirrlees review highlighted that the economic case for replacing our property taxes with a land-based taxes is a very strong one:
“Taxing land ownership is equivalent to taxing an economic rent—to do so does not discourage any desirable activity. Land is not a produced input; its supply is fixed and cannot be affected by the introduction of a tax. With the same amount of land available, people would not be willing to pay any more for it than before, so (the present value of) a land value tax (LVT) would be reflected one-for-one in a lower price of land: the classic example of tax capitalization.
Owners of land on the day such a tax is announced would suffer a windfall loss as the value of their asset was reduced. But this windfall loss is the only effect of the tax: the incentive to buy, develop, or use land would not change. Economic activity that was previously worthwhile remains worthwhile.
Moreover, a tax on land value would also capture the benefits accruing to landowners from external developments rather than their own efforts.”
Despite this strong in-principle case, pure LVTs (or Land Rental Value Taxes) are relatively rare. Historically, the introduction of LVTs has accompanied the settlement and development of land – in Australia and the USA, for example – when the state creates land value by bestowing property rights and providing infrastructure. Developed urban economies tend instead to levy taxes on the occupation of land and buildings in combination, or simply on property values and transactions.
Introducing LVTs is politically risky, since it requires major reform, creating winners and losers across the whole population of a defined territory. It is also practically difficult, since land in urban areas is rarely sold without existing property on it, and therefore cannot be readily valued through market prices. An additional complication in the UK is that the ownership of around 12 per cent of the land in England and Wales is not registered with the Land Registry.
Nonetheless, there is considerable international experience of land valuation worldwide: the interim report of the Lyons Review into local government finance in 2005 noted that there were over 700 cities worldwide in which land and buildings were valued separately. In Denmark, which has had a land tax since 1926, land is valued on the basis of its best economic use every two years (with indexation applied in the alternate year) using a computerized hedonic model. The problem of “asset-rich, income-poor” pensioners is overcome by allowing the over 65s to roll up LVT due until the property is sold, when it is levied in full.
Examples of Land Value Tax rates
What would a LVT in London look like? Despite the limitations in the data on land ownership and values, the University of Bath used benchmark land values to estimate what tax rates on land values would be needed replace existing revenue from Council Tax, Stamp Duty and Business Rates. The graph below shows that a Land Value Tax of 2.3 per cent per annum could generate the same revenue as Council Tax and Stamp Duty in London, while a slightly higher rate on commercial land would generate the same revenue as business rates did in 2016/17 (a significant increase in business rate revenue is anticipated in 2017/18).
Current tax revenue in London and equivalent Land Value Tax rates, 2014/15
Source: Barnaby C. (2017), Estimation of Land Value Tax Revenues in London. Bath Institute for Mathematical Innovation. Annex Report. Using Valuation Office Agency land value data for March 2015.
Issues such as the inclusion or exclusion of private gardens have a significant effect on the rate charged (and on the distributional impact). A LVT of 0.7 per cent on both buildings and private gardens raises the same revenue as a LVT of 2.3 per cent on buildings only.
Land Rental Value Tax
Taxing the rental value of land is an alternative to land its capital value. Rental values are indeed a more accurate estimate of the economic utility of land than land values, which take into account speculative factors.
The table below illustrates how observed market rents can be used to derive land rental, removing the buildings component by taking indicative rebuild costs and spreading them over 50 years, i.e. at two per cent per annum and deducting this element from the total rent.
Illustrative Domestic Land Rental Value Calculation
|Parcel||Development Plot size||Floor space||Description||Households||Annual rental value per sqm floor space||Annual rental value per sqm land||Annual rental value||Build cost||Land Rental Value|
|27 Oak St
|50 sqm||250 sqm||5 rented flats||5||£200/sqm||£800/sqm||£50k
|29 Oak St
|50 sqm||250 sqm||Single family, owner-occupied||1||£50k
|2 Oak St
|100 sqm||500 sqm
|Planning permission for 10 residential units||10||£100k
Source: University of Bath Institute for Policy Research calculations
We cannot afford to leave our property tax system untouched given London’s chronic challenges. The current system is outdated, regressive and hampers mobility within the housing market. As the London Finance Commission has argued, the full suite of property taxes should be devolved to the Mayor of London and London boroughs to enable reform and respond to the city’s specific challenges.
Brexit opens the door to and strengthens the case for considering radical options for reform after devolution, to create a tax system that is fairer for Londoners, supports infrastructure and housing delivery, and send better signals to the market. This chapter has reviewed some possible routes for this – including the case for land-based taxes – and we recommend that further research is undertaken to explore their distributional impacts (in line with the recent report by the London Assembly Planning Committee).
After over two decades of status quo, any reform to property taxes would create a large number of winners and losers, making transitional arrangements essential. Reform would also require governance reform to enable effective administration and distribution of tax between the Mayor of London and London boroughs. But the reforms reviewed in this chapter should be considered as sequential. Revaluation and reform of Council Tax as a first step (accompanied by the creation of a London Valuation Office to take on these duties) would make marked progress towards fairer and more effective taxes on land
 Slack E. (2016). International Comparisons of Global City Financing: A Report to the London Finance Commission. Institute on Municipal Finance and Governance Munk School of Global Affairs. University of Toronto.
 Mirrlees, J. A., Adam, S. (2010). Dimensions of tax design: the Mirrlees review. Oxford University Press.
 Wingham M. (2017). Council tax in London. A publication for the London Finance Commission. Working Paper 80. GLA Economics.
 Transport for London. Infrastructure investment and land value uplift. Presentation to the London Finance Commission 20.10.2016.
 London Finance Commission 2017. WIPissuess receiving EU funding, allndon.and Inequality in London. Centre for London.
 Transport for London (2017). Land Value Capture.
 Marsden J. (2015). House prices in London – an economic analysis of London’s housing market. Working Paper 72. GLA Economics. and Portico. London Rental Yield Map. https://www.portico.com/yields (accessed on 27.04.2017).
 Wingham 2017.
 The distribution of sales across bands was estimated from Land Registry sales data (average for years 1995/1996/1997 (earliest available) and 2013/2014/ 2015). We assumed that the distribution of sales in both time periods was similar. The Land Registry sales data was then cross-referenced with ONS’s House Prices estimates.
 Average property in each band.
 There are too few properties in Band H in the London Borough of Hackney to estimate a reliable average property price – but a £12m property would face bills of £34,000 a year under this rate-based system.
 Full details of the methodology and results are given in a background paper.
 Mirrlees 2010 do acknowledge that planning regulations can marginally change the supply of land
 Mirrlees 2010.
 HM Land Registry (2016). Annual Report and Accounts 2015/16.
 Estimates made by Catherine Barnaby of the University of Bath Institute for Mathematical Innovation – the full report can be found at www.bath.ac.uk/ipr/publications
 Land value is impacted not only by current economic utility but also significantly by speculative sentiment on rental growth (and thus the future disposal value) and by cyclical swings in interest rates over time
 2% of rebuild cost @ £2k/sqm of floor space
 To simplify, we assume rental values per sqm floorspace are identical.
 London Assembly Planning Committee (2016). Tax Trial: A Land Value Tax for London.