Mark Stephens argues that Scotland’s new system of housing transaction tax (LBTT) should be allowed to bed down, but that as the Parliament’s Finance Committee considers “A Scottish Approach to Taxation” it should consider property taxation in the round.
In April 2015 the Scottish Parliament gained the right to set the rates of transaction tax on residential and commercial property transactions. The Parliament’s Finance Committee is currently reviewing its first year of operation.
The Scottish Government announced that it would change the structure of the former “Stamp Duty Land Tax” for residential properties, shifting from a “slab” to a “slice” structure – so removing the “cliff edges” that previously existed. Rather unhelpfully, the UK Government introduced its own reformed structure shortly before the Scottish system – known as Land and Buildings Transaction Tax (LBTT) – was due to be introduced. This prompted the Scottish Government to tweak its structure.
Table 1. Land and Buildings Transaction Tax (from April 2015)
| Purchase price | LBTT rate |
| Up to £145,000 | 0% |
| Above £145,000 to £250,000 | 2% |
| Above £250,000 to £325,000 | 5% |
| Above £325,000 to £750,000 | 10% |
| Over £750,000 | 12% |
Source: Revenue Scotland
Additionally, the Scottish Government followed the UK in introducing a 3% surcharge for second properties, ostensibly because not to do so risked attracting a sudden rise in investment in buy-to-let property from south of the border.
So how did the first year go?
Revenues were somewhat lower than forecast, with the focus being on the properties in the £325,000-£750,000 range. This has caused some disquiet in the housebuilding industry, and the Finance Committee is considering whether the bands should be changed.
Taking the long view
Whilst all of the submissions to the Finance Committee have focussed on trying to unpick the figures for a couple of years, it seems sensible first to step back and consider the new tax regime in perspective.
Although LBTT was introduced in its present form in April 2015, its antecedents date back to the 17th century when they were introduced seemingly because it was easy to do so.
Most economists argue that transaction taxes are inefficient because they create a wedge between buyers and sellers, suppress market activity and lead to an inefficient allocation of properties. In a practical sense they may deter mobility, which is counter to the objective of labour mobility. In contrast, recurrent taxes on land and property values are in principle non-distorting. If, as the Committee’s parallel enquiry into A Scottish Approach to Taxation seeks to do, one were to consider the taxation system in the round then the possibility of shifting from taxing property transactions to the values of land and property ought to be a central consideration.
In examining the impact of a tax, we need to distinguish between its formal (who legally is liable) and actual (who actually pays for it) incidence. The price of housing is determined by the anticipated value that we attach to living in it over time if one is an owner-occupier, or, if one is a landlord, from the anticipated rental income. Property taxes (whether transaction or recurrent) reduce the owner’s rate of return, which to be equalised results in the price of housing falling. Thus, although the purchaser formally pays the tax, over time we would expect it to be borne by the seller, i.e. reflected in the price. In contrast when domestic rates were abolished in favour of the poll tax, there was an upward pressure on house prices. Similarly, when mortgage interest attracted tax relief, it undoubtedly inflated house prices. Moreover, when changes occur the market may take some time to adjust, so this so-called “capitalisation” effect may take some time. Buyers and sellers are also likely to be influenced by anticipated changes to taxation.
Long-term experience of Transaction Taxes
Stamp Duty (as the tax was known until 2003) was, for many decades, a very straightforward tax. It was levied at 1% on property transactions over a certain threshold, which because it was rarely uprated, led to large-scale “fiscal drag” as rising house prices brought more properties over the threshold. In the 1980s, the generally low level of property transaction costs (including legal, surveying charges) were often cited as a reason for internationally high levels of home-ownership in the UK. It was claimed that they facilitated entry into the home-ownership market at a young age, and did not deter frequent moves necessitated as children were born and larger properties were needed.
The first attempt to use Stamp Duty to micro-manage the housing market occurred during the housing market recession in the 1990s. Stamp Duty was effectively suspended for a pre-determined fixed period in an attempt to entice buyers back into the market. The most notable effect of this concession was that there was an upswing in transactions only when the end of the “holiday” was imminent and buyers became convinced that the “holiday” would be extended.
After 1997, the tax became more elaborate as the number of bands (and rates) grew, and were altered frequently. In addition, a variety of concessions were introduced for first time buyers, purchasers in depressed areas and even (according to the Treasury) “to help kick-start the market in zero-carbon homes, encourage micro-generation technologies, and raise public awareness of the benefits of living in zero-carbon homes.” These concessions eventually began to collide with one another (for example, a concessionary threshold for first time buyers was greater than the general concession in depressed areas). However, perhaps the most notable side-effect of the system was the clustering of transactions below the thresholds – evidence of a deeply distorting effect which is attributable to the “slab” structure of the tax.
Stamp Duty and its successor Stamp Duty Land Tax became an increasingly lucrative source of revenue for government. UK receipts rose from £675 million in 1996/97 to a peak of £6.7 billion in 2007/08 (the eve of the Global Financial Crisis). However, transaction taxes are also very unstable, reflecting volatility of prices and transactions in the housing market.
Fig. 1. Revenues from SDLT in Scotland (£ million)
In Scotland, revenues from SDLT fell from £340 million in 2007/08 to £135 million in 2009/10. The percentage annual rates of change in (nominal) revenues from SDLT in Scotland and the rest of the UK are shown in Figure 2.
Fig. 2. Annual percentage change in (nominal) revenue from SDLT
The Finance Committee requested opinion on a number of aspects of the new tax.
Impact on revenue and whether this justifies altering bands/ rates in the draft budget for 2017/18
The first year of LBTT has been marked by the following:
- A reduction in transactions of higher value properties (esp. £325,000-750,000)
- A lower revenue than forecast (£208m vs £235 m – although it should be noted that the £235m figure relates to the anticipated revenue before forestalling was taken into account).
The examination by the Scottish Fiscal Commission attributes the difference to forestalling, but also suggests that it is possible that transactions in the £325-750,000 price range may be subject to “ongoing behavioural responses” to the tax.
Such “ongoing behavioural responses” may not be permanent. For example, the market might still be adjusting to the change. Would-be sellers may be holding out for prices that purchasers are unwilling to pay. The submission from Mr Alan Mickel illustrates this point. He attributes the difficulty he has experienced selling his property at the level suggested by valuers at least in part to LBTT. It may take some time for would-be sellers to lower their prices to a new level. (The “stickiness” of house prices in a downward direction is a feature of the housing market. When markets turn down, owners appear to be reluctant to accept that prices will fall. This in part explains the volatility of transactions.)
The submission by Aberdeen City Council also suggests that it is difficult to disentangle the effect of LBTT on the market from that of the downturn in the oil and gas industry.
Particularly given the disruptive short-term impacts of tax changes, and the additional disruption caused by anticipated changes, it would be premature to alter the bands or rates until the new system has had a chance fully to “bed down.”
If it is then established that the structure of the tax is causing a permanent reduction in revenue, then there would be a case for restructuring.
Consistency of bands and rates with the principles of “fairness, equity and the ability to pay”
The Scottish Government seeks to base the tax system on the following principles:
- Be proportionate to the ability to pay;
- Provide certainty to the taxpayer;
- Provide convenience / ease of payment, and;
- Be efficient.
Setting aside the band ranges and the rates of tax, it should be uncontroversial to assert that the “slice” structure of the tax is both fairer and more efficient than the SDLT’s slab structure that preceded it. The sudden jumps in liability and the consequent bunching of transactions around band thresholds were both unfair and inefficient.
A fundamental question is whether it is fair that property taxes should fall disproportionately on people who are mobile. Even if the actual burden of the tax falls on the seller, it is still people who move frequently who will pay the most LBTT through their lives. It might be regarded as being both fairer and more efficient (in terms of mobility) if property-related taxes were instead borne across the community as a whole through a recurrent property tax.
Views on the fairness of actual band structures and rates are largely a matter of opinion. Nonetheless, it is worth noting that liability for LBTT begins only when property values exceed £145,000, which is roughly the median house price (and about 80% of mean average house price) – implying that half of properties incur no tax.
Table 2 Entry-level Thresholds for Stamp Duty compared to Average House Prices
| Threshold for Stamp Duty liability | Ave. House Price (UK) | Threshold as % Ave. House Price | |
| 1984 | £30,001 | £29,106 | 103 |
| 1993 | £60,001 | £62,333 | 96 |
| 2000 | £60,001 | £101,550 | 59 |
| 2005 | £120,001 | £190,760 | 63 |
| 2006 | £125,001 | £204,813 | 61 |
| 2010 | £125,001 | £226,064 | 55 |
Source: M. Stephens (2011) Tackling Housing Market Volatility in the UK, Table 4
The table shows that the entry-level for Stamp Duty fell as a proportion of average house prices over time. In 1984 and 1993 it was set at around the average house price (though would have fallen in between). But when the threshold was raised in 2000, 2005 and 2006 it was set at around 60% of average prices. In 2011 it fell further to 55%. After 1997, the system was also becoming more “progressive” in the sense that higher rates were introduced (within the slab structure).
The idea of a tax allowance is well established in income tax, but even given the policy of raising the personal allowance, it is set well below median full-time earnings, and a long way below average earnings. By way of illustration: if mortgages are granted on (roughly) the basis of 3x gross income, with a minimum deposit of 10%, then a minimum income to purchase a house priced at around the entry threshold would likely to be earning in excess of £40,000. Clearly, this is a political decision, but is one that could be better informed by an analysis of the incomes of house purchasers and sellers. Such an analysis might also inform the setting thresholds and rates for higher band rates.
Conclusions
On the basis of the evidence available, LBTT should be allowed to bed down to establish whether the thresholds and rates are leading to a permanent diminution of revenues. This would also help the tax forecasters to improve their model.
Assessed against its base (i.e. the value of properties transacted), it is undeniably progressive. However, it is unclear how well LBTT relates to the “ability to pay” (which generally refers to current income). An analysis of purchaser and seller incomes would give a clearer idea whether the entry threshold is too high (it seems likely that many people on above-average earnings are liable to no tax), and how well higher bands and rates relate to incomes. Moreover, it is not clear why it is equitable that people who move more frequently should be liable to pay more tax than those who do not.
There is a strong case for looking at property taxation in the round, and to consider whether LBTT might be better absorbed into the system of recurrent property taxation. With the Council Tax subject to modest reform, this really ought to be a topic addressed by the Finance Committee as it considers A Scottish Approach to Taxation.
This blog is based on Mark Stephens’ written evidence to the Scottish Parliament Finance Committee. He gave oral evidence to the Committee on 28 September. The proceedings, which also included contributions from Homes for Scotland, the Scottish Property Federation and RICS, can be viewed here.