Abstract
The Selected Issues paper examines the dynamics of the inflation process in the United Kingdom, particularly the influence of external shocks. The paper provides a brief summary of existing empirical work. It presents the expectations-augmented Phillips curve model used in the analysis. The basic conclusion is that low inflation in recent years can be explained reasonably well by a combination of increased competitive pressures. The paper also presents some estimates of the degree of possible overvaluation in housing prices and outlines the links between housing prices and the macroeconomy.
IV. Property Taxes and the Housing Market 1
A. Introduction
1. Fluctuations in house prices, and their impact on activity, have been a central feature of the U.K. business cycle. Over the past 30 years, real house prices have not only grown stronger than in the euro area, but have also been more volatile and associated with overall macroeconomic fluctuations. Housing market peculiarities were underscored by HM Treasury’s EMU studies, which identified house price fluctuation as a key reason for cyclical divergences, as well as differences in the transmission of monetary policy impulses, setting the United Kingdom apart from the euro area economies.2 The macroeconomic implications of sharp movements in house prices are analyzed in Chapter II of this paper3 where the effectiveness of policy in mitigating adverse effects on output and inflation is discussed. Because of the importance of reducing volatility in the U.K. housing market, a number of researchers have studied the role of microeconomic and market structures, and the potential for fiscal instruments to improve market outcomes.
2. Flexible-rate mortgage financing and supply constraints have been identified as the major reasons for the volatility in house prices. Two recent reports, commissioned by HM Treasury, concluded that mortgage financing and inelastic housing supply explained much of the distinctiveness in the U.K. housing market, and explored ways to reduce volatility in house prices. The Miles Report looked into prospects of increasing the share of fixed-rate mortgages in housing finance, while the Barker Report studied the impact of supply constraints.4 The Barker Report also reflected on possible ways for fiscal policy to help reduce volatility in the housing market, including through reform of property taxation (the council tax).
3. In the United Kingdom, the council tax on housing is determined by local financing needs. The council tax5 is local authorities’ main discretionary revenue instrument, covering about 25 percent of local expenditures. It was introduced in 1993 as a replacement for the unpopular community charge (“poll tax”). It is a tax on housing, paid by owners and renters, and based on the assessed value of a property, currently as of April 1991. Properties are divided into eight tax-liability bands. As higher income households tend to own houses in the upper brackets, these households pay on average higher property taxes in absolute terms. But the implicit tax rate, the ratio of taxes paid to property values, tends to fall when values increase. This puts in effect a disproportionately larger burden on lower incomes (income regressiveness). Similarly, communities with low average house values are burdened with a higher effective council tax rate, because local authorities’ spending varies less than house prices across communities (regional regressiveness). This tax has drawn heavy criticism for a number of reasons, most prominently the two noted, its regressiveness across households and regions.6
4. A redesigned council tax could address these shortcomings, and, as the Barker report argued, strengthen fiscal automatic stabilizers as well. A redesigned tax could withdraw a higher proportion of household income when property values increase because it could take advantage of the fact that house prices typically fluctuate considerably more than before-tax income. With annual property assessments (or indexation of housing values between multi-annual assessments) the tax base would move in line with house prices and therefore in excess of household income. Combined with a fixed tax rate, the council tax could play a countercyclical role.7
Effective Council Tax Rates by English Regions
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
The effective council tax rate is defined as the ratio of average council taxes for 2 adults in band D, before transitional relief and benefits, to simple average house prices of all dwellings.Source: ODPM, Local Government Finance Statistics.Effective Council Tax Rates by English Regions
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
The effective council tax rate is defined as the ratio of average council taxes for 2 adults in band D, before transitional relief and benefits, to simple average house prices of all dwellings.Source: ODPM, Local Government Finance Statistics.Effective Council Tax Rates by English Regions
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
The effective council tax rate is defined as the ratio of average council taxes for 2 adults in band D, before transitional relief and benefits, to simple average house prices of all dwellings.Source: ODPM, Local Government Finance Statistics.5. Effective council tax rates are negatively correlated with house price inflation across regions in England. Since 1997, house prices have risen more sharply in the United Kingdom on average than local council tax revenues, mainly because property values are not assessed annually. As a result, the effective council tax rate has fallen over time. The strongly negative relation between the level of effective council tax rates (defined as the ratio of average council taxes per dwelling to the respective average house prices) and the average rate of house price increases by English regions, suggests that the council tax might have contributed to house price inflation (Figure IV.2).
Level in Implicit Council Tax rate Against Increase in Real House Prices (1998 to 2003)
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
Level in Implicit Council Tax rate Against Increase in Real House Prices (1998 to 2003)
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
Level in Implicit Council Tax rate Against Increase in Real House Prices (1998 to 2003)
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
6. The focus of this paper is on an empirical evaluation of the impact of the council tax on medium-term house price volatility through its impact on the user cost of housing. Although important dimensions of the broader issue of tax reform, neither the council tax’s short-term macroeconomic and distributional impact nor its role as a reliable source of local government income are explicitly examined. The empirical analysis finds evidence that declining effective council tax rates have had a positive impact on house price volatility over time and across regions. However, the magnitude of these effects was found to be small relative to other factors influencing house prices.
7. The remainder of this paper is organized as follows. In Section B, the theoretical model of house prices is presented highlighting the role that the council tax can play in house price volatility through its effect on the user cost of housing. The panel estimates of the model are presented in Section C along with some robustness checks. Some conclusion are offered in Section D.
B. Property Taxation Affects House Price Volatility through User Costs of Housing
8. A house price equation is obtained by inverting the demand function for housing services. Housing demand is assumed to increase with the number of households (or population) and the average household income, and to fall with rising real user costs of housing services. Other factors, such as changes in preferences for housing services relative to other consumption items, may also play a role.8
Equation (1) describes housing demand (H) as a function of demographic factors (pop), average real household income y, user costs of housing (uch), and (Z) representing any other factor that might shift the demand-for-housing curve. The slope of the housing demand curve is determined by the elasticity of demand for housing services with respect to the user costs of housing. With a high elasticity, a relatively small change in user cost is sufficient to minimize the impact on prices of an exogenous shift in demand.
The user costs of housing are proportional to house prices (P). They increase with the (tax adjusted) nominal mortgage rate (r), the rate of depreciation including maintenance costs (δ), and the property tax rate (τ), and fall with expected capital gains on the housing stock (πe), or expected house price inflation.
Using (2) the demand equation (1) can be inverted to obtain an equation for house prices.
9. In the short run, the supply of housing is either fixed or inelastic with respect to price and demand factors. Housing supply will increase as long as prices of existing houses remain above construction costs (C). Space limitations, zoning and other restrictions may however prevent equilibrium house prices from settling in the neighborhood of construction costs. Consequently, the supply elasticity (η) is very small (the slope of the supply curve very steep) in the short run.
The supply curve is:
10. In the long run, however, supply constraints, such as planning restrictions, are often eased. Planning is necessary to preserve the environment, avoid excessive congestion and limit other unwelcome by-products of building activities. However, the weights on political preferences might shift. When excess demand persists, economic considerations are likely to gain weight at the expense of other concerns. Therefore the elasticity of housing supply with respect to prices (η) is likely to be larger in the long run. But even with a time-invariant price elasticity, the supply curve becomes more positively sloped when the existing housing stock gets replaced or upgraded. Depending on the depreciation rate (δ), the long-run elasticity (η/δ) is larger even at an unchanged but non-zero price elasticity (η).
11. Short-run supply constraints lead to overshooting in house prices. The model is closed by assuming equality between demand and supply. An increase in demand will initially cause prices to rise above its new long-run equilibrium, and subsequently fall back (Box IV.1). The extent of price overshooting depends on a number of factors, which are typically identified to play a role in U.K. house price fluctuations, most prominently supply constraints and flexible-rate mortgage financing, but taxes also play a role. The simple model serves as a useful analytical tool to help understanding the possible influence of property taxes on house price fluctuations in the United Kingdom.
12. Mortgage rates tend to rise in cyclical upswings, and therefore mitigate price fluctuations. In the United Kingdom, largely due to the predominance of variable-rate mortgages, current and expected monetary policy moves are quickly transmitted to mortgage interest payments. Following a forward-looking inflation targeting rule, the Bank of England hikes policy rates when inflation would otherwise be expected to rise above target. As inflationary pressures build up when demand is growing faster than supply and the output gap is closing, policy interest rates will often move counter-cyclically with income. Changes in policy rates are quickly transmitted into mortgage rates, dampening housing demand relative to income, and therefore mitigating price fluctuations.
Housing Market Mechanics
In a linearized version of the model the slope of the demand curve in the H-P plane, holding all else constant, is simply given by -1/(r + τ + δ + πe) An increase in income,
population or other factors will shift the demand curve to the right, and lead to a stronger price increase when the mortgage, property tax or depreciation rates are low and expected house price increases are high (right picture). In the short run, when supply is predetermined, an increase in demand, shifting the demand curve out, will cause prices to increase from the original equilibrium in point A to a new temporary equilibrium point B. In the long run, supply reacts (the supply curve become positively sloped) and house prices fall back to point C. The degree of house price overshooting is inversely related to real user costs of housing, with the effective council tax rate as one component.
While the overshooting effect can be large in theory, an illustrative calculation based on average data values shows that the council tax’s contribution is more likely to be small in practice. Over the 1970 to 2003 period, the average value of (uch-τ) has been 2.7 percent, while the spread between the highest and lowest effective council tax rates was 1.5 percentage points. Given these numbers, and assuming a unitary income elasticity of housing demand H, a two standard deviation shock to real income (equivalent to 10 percent) causes house prices to overshoot by 0.1 percent at the high effective tax rate and by 0.2 percent at the low rate. However, the exact magnitudes of these effects in practice is an empirical question that is addressed in the next section.
13. Taxes can both exacerbate or dampen house price volatility. For example, the deductibility of mortgage payments from income taxes reduces the effectiveness of a given change in interest rates. Likewise, stamp duties and the U.K. type of property taxes can undermine endogenous stabilization of house prices through their impact on the user cost of housing variable,9 if constructed as they currently are in the United Kingdom, or they can reinforce automatic stabilizers if redesigned to serve this purpose.
C. Estimating the Effect of Council Taxes on House Prices
14. The price equation is estimated using regional data from 1969 to 2003. Efficient use was made of the available information by pooling data over time and across sections. The cross sectional dimension is given by the nine regions in England:10 North East, North West, Yorkshire and the Humber, East Midlands, West Midlands, East of England, London, South East and South West. Changes in real house prices (in each region) are linked to changes over time in regional demographics, regional real disposable household income, and regional real user costs of housing.
Real house prices
where
User cost of housing
rt: mortgage interest rate (Nationwide);11
mit: maintenance cost, calibrated [0.025];12
ϑ: depreciation rate, calibrated [0.015];12
ctit: effective council tax rate, calculated as tax revenues divided by house values;13
πeit: expected capital gains (HP trend of nominal house price inflation)
Inflation expectations
15. The user costs are strongly influenced by expected house price inflation in times of high price increases, but expectations are not observable and need to be estimated. While rational (or model consistent) expectations are theoretically satisfactory, it is difficult to constitute them in a partial equilibrium context. Also, households often buy a property to live in it, and enjoy non-pecuniary advantages of ownership, which may not be entirely reflected in rational equilibrium models. Applying the Hodrick-Prescott filter-which is essentially a centered moving average-implies a combination forward- and backward-looking behavior. It is assumed here that this representation fits households expectation well. The HP filter however suffers from an ‘end-point problem’, which limits its usefulness at the sample period end. It has therefore been corrected in the following way: (i) as from 2005, annual real house price increases are gradually returning to their average pace over the past 35 years (0.5 adjustment speed); (ii) a two percent CPI inflation rate is added; (iii) the resulting series is HP filtered over the period 1970–2008.14 The following Figure IV.3 shows annual percentage house price increases in London and its adjusted and non-adjusted HP filtered trends.15 Despite recent valuation gains, there has been a remarkable decline in nominal house price volatility after 1990, which probably owes much to the switch in the monetary regime and the resulting greater overall price stability.16
House Price Inflation
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
Source: ODPM, Simple average house prices, all dwellings.House Price Inflation
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
Source: ODPM, Simple average house prices, all dwellings.House Price Inflation
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
Source: ODPM, Simple average house prices, all dwellings.Calibrating the effect of taxes on the real user cost of housing
16. Taxes modify the pattern of the user cost of housing. Figure IV.4 illustrates how alternative calibrations of housing taxes affect the user cost of housing variable (in this example for London17). The solid line shows the real user cost of housing without taxes. The second line shows the impact of unlimited mortgage interest tax relief (MITR) assuming a 35 percent marginal income tax rate and 100 percent mortgage financing over the entire 1970–2003 period. In reality, the income tax treatment of mortgage payments has been more complicated. Tax laws and other factors, including the loan-to-value ratio, have been changing over time. Before the introduction of a 25,000-pound ceiling in 1974, MITR was available on any mortgage loan size. The tax relief was gradually reduced by under-adjusting nominal loan ceilings and later also by introducing restrictions on the applicable income tax rate. In April 2000, after its real value was largely eroded, MITR was finally abolished. The third line shows real user costs of housing including the effective council tax rate since 1993. The effective council tax rate in London declined from 0.7 percent to 0.4 percent, a 30 basis points decline, compared with a 340 basis points fall in the mortgage rate during the same period. The drop in house prices in the early 1990s was preceded by a sharp increase in the user costs of housing,18 while the recovery in house price inflation was accompanied by a steady fall in user costs. About two thirds of the spike in user costs, as well as its subsequent easing was due to changes in mortgage rates, about one third may be explained by the measure of expected house price inflation.
Calibrating the Effect of Taxes on the User Costs of Housing
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
Source: Fund calculations.Calibrating the Effect of Taxes on the User Costs of Housing
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
Source: Fund calculations.Calibrating the Effect of Taxes on the User Costs of Housing
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
Source: Fund calculations.17. The impact of the council tax is small within regions, but larger across regions reflecting sizeable differences in regional effective rates. In general, the inclusion of tax variables changes the pattern of user costs over time only marginally. The effective council tax rate explains only about 3 percent of the fall in user costs of housing in London (about 5 percent in North East where the fall in user cost was smaller). Widespread mortgage credit constraints during the earlier period of very favorable tax treatment, and its erosion since the mid-1970s, have mitigated the effects of the MITR on house prices. Therefore the user cost variable used in the equations reported below excludes taxes.
Estimation results
18. The price equation performs well; all parameters have the expected signs, reasonable magnitudes, and are robust over time and across estimation methods. Estimation results over the full time range (1972–2003) confirm expectations; all coefficients have the right sign, as well as a sufficient significance level, and lie well within the range of results reported in the literature. The overall test statistics are also satisfactory. More importantly, demographic effects were the major difference between the first equation, covering also Wales and Scotland, and the other equations, covering only the regions of England. In England, the elasticity of house prices with respect to real income is around 1.6, and with respect to the user costs of housing around -0.5. Both are plausible magnitudes, given important supply restrictions. The equation seems also to be reasonably stable over time. The major difference again being demographics, which seem to have played a larger role in the earlier period 1972–93 than over the full sample. Despite first differencing of all the variables, heterogeneity across regions might still be present in the estimated equation. This is partly so because regional effective property tax rates—the stamp tax and council tax rates, as well as its predecessor rates—have been time varying. Regional heterogeneity is captured by fixed effects (FE) or random effects (RE) estimation. Overall, the coefficients changed little compared to simple OLS. With respect to the coefficient on uch, which is of particular interest here, this implies that heterogeneity, whether or not caused by declining effective council tax rates, has very little effect. The estimated income elasticity was also quite robust, but the influence of demographic variables was sensitive to the estimation method.
Estimation Results
Estimation Results
| Great Britain | England | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Sample (adjusted): | 1972 2003 | 1972 2003 | 1972 2003 | 1972 2003 | 1972 1992 | |||||
| Included observations: | 34 | 32 | 32 | 32 | 21 | |||||
| Cross-sections included: | 11 | 9 | 9 | 9 | 9 | |||||
| Total pool observations: | 352 | 288 | 288 | 288 | 189 | |||||
| (all balanced) | ||||||||||
| Estimation method | FE | OLS | FE | RE | Swamy-Arora | OLS | ||||
| Variable | Coefficient | t-Statistic | Coefficient | t-Statistic | Coefficient | t-Statistic | Coefficient | t-Statistic | Coefficient | t-Statistic |
| constant | -0.009 | -2.1 | -0.008 | -1.0 | -0.015 | -1.5 | -0.008 | -1.0 | -0.012 | -1.2 |
| D(LOG(Hit)) | 0.393 | 2.9 | 1.081 | 1.5 | 1.811 | 1.9 | 1.081 | 1.5 | 1.611 | 1.8 |
| D(LOG(Yit)) | 1.494 | 26.8 | 1.600 | 24.7 | 1.600 | 24.4 | 1.600 | 24.4 | 1.689 | 20.1 |
| D(UCHit) | -0.395 | -2.3 | -0.512 | -2.4 | -0.509 | -2.3 | -0.512 | -2.3 | -0.896 | -2.9 |
| Effects Specification | ||||||||||
| fixed | fixed | |||||||||
| Cross-section random S.D. / Rho | 0 | |||||||||
| Idiosyncratic random S.D. / Rho | 0.058257 | |||||||||
| R-squared | 0.717 | 0.692 | 0.692 | 0.694 | 0.724 | |||||
| Adjusted R-squared | 0.705 | 0.689 | 0.689 | 0.682 | 0.719 | |||||
| S.E. of regression | 0.054 | 0.058 | 0.058 | 0.058 | 0.062 | |||||
| Sum squared resid | 1.001 | 0.941 | 0.941 | 0.937 | 0.701 | |||||
| Mean dependent var | 0.033 | 0.035 | 0.035 | 0.035 | 0.022 | |||||
| S.D. dependent var | 0.100 | 0.103 | 0.103 | 0.103 | 0.116 | |||||
| Durbin-Watson stat | 1.771 | 1.625 | 1.625 | 1.612 | 1.688 | |||||
| F-statistic | 60.894 | 213.187 | 213.187 | 56.912 | 161.643 |
Estimation Results
| Great Britain | England | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Sample (adjusted): | 1972 2003 | 1972 2003 | 1972 2003 | 1972 2003 | 1972 1992 | |||||
| Included observations: | 34 | 32 | 32 | 32 | 21 | |||||
| Cross-sections included: | 11 | 9 | 9 | 9 | 9 | |||||
| Total pool observations: | 352 | 288 | 288 | 288 | 189 | |||||
| (all balanced) | ||||||||||
| Estimation method | FE | OLS | FE | RE | Swamy-Arora | OLS | ||||
| Variable | Coefficient | t-Statistic | Coefficient | t-Statistic | Coefficient | t-Statistic | Coefficient | t-Statistic | Coefficient | t-Statistic |
| constant | -0.009 | -2.1 | -0.008 | -1.0 | -0.015 | -1.5 | -0.008 | -1.0 | -0.012 | -1.2 |
| D(LOG(Hit)) | 0.393 | 2.9 | 1.081 | 1.5 | 1.811 | 1.9 | 1.081 | 1.5 | 1.611 | 1.8 |
| D(LOG(Yit)) | 1.494 | 26.8 | 1.600 | 24.7 | 1.600 | 24.4 | 1.600 | 24.4 | 1.689 | 20.1 |
| D(UCHit) | -0.395 | -2.3 | -0.512 | -2.4 | -0.509 | -2.3 | -0.512 | -2.3 | -0.896 | -2.9 |
| Effects Specification | ||||||||||
| fixed | fixed | |||||||||
| Cross-section random S.D. / Rho | 0 | |||||||||
| Idiosyncratic random S.D. / Rho | 0.058257 | |||||||||
| R-squared | 0.717 | 0.692 | 0.692 | 0.694 | 0.724 | |||||
| Adjusted R-squared | 0.705 | 0.689 | 0.689 | 0.682 | 0.719 | |||||
| S.E. of regression | 0.054 | 0.058 | 0.058 | 0.058 | 0.062 | |||||
| Sum squared resid | 1.001 | 0.941 | 0.941 | 0.937 | 0.701 | |||||
| Mean dependent var | 0.033 | 0.035 | 0.035 | 0.035 | 0.022 | |||||
| S.D. dependent var | 0.100 | 0.103 | 0.103 | 0.103 | 0.116 | |||||
| Durbin-Watson stat | 1.771 | 1.625 | 1.625 | 1.612 | 1.688 | |||||
| F-statistic | 60.894 | 213.187 | 213.187 | 56.912 | 161.643 |
19. However, house price dynamics might have changed since the mid-1990s. When the estimation range is significantly shortened to cover the era before the introduction of the council tax (in 1993) the estimated parameters changed significantly, though still remaining rightly signed and within a reasonably range. The impact of the number of households has sharply declined (however, its coefficient was only marginally significant). The income elasticity remained little changed at 1.69 (compared to 1.60 for the full sample), and the user cost variable became more important with a semi elasticity of -0.90 (against -0.51 for the full sample). The robustness of the estimation results over time with respect to income and user cost of housing is reassuring, and the direction of change is congruent with a more stable monetary environment since the mid-1990s and the decline in relative importance of credit restrictions in the 1970s. Though the reasons for the declining influence of demographic variables since 1993 are unclear.
20. The effect of council taxes on house price inflation is more important across regions than over time. To disentangle the two effects, calculations based on estimated parameters show that the contribution of the effective council tax rate to regional house price increases varies between -0.9 percentage points in the North East to +1.1 percentage points in London and the South West, or about half of the difference in average inflation rates across regions (Table IV.2). However, the impact over time of a declining effective tax rate when house prices were rising was minor, and at 0.2 a little stronger where the decline in percentage points was largest (North East). In the regions that are often seen as setting the pace of house price increases (South West and London), the level of the effective tax rate is simply too small to have a major impact, because even a sharp increase in house prices will reduce the effective tax rate only by a few basis points.
Back-of-the-Envelop Impact of Declining Council Tax Rates on House Prices
Back-of-the-Envelop Impact of Declining Council Tax Rates on House Prices
| (percentage points) | Dimension over time Effective tax rate | Dimension across Regions Real house price increase | Total | |||||
|---|---|---|---|---|---|---|---|---|
| Decline 97-03 | Impact on house price | (annual aver.) | deviation from mean | Impact on house price | Impact (row 2+5) | |||
| North East | 0.4 | 0.2 | 7.7 | -2.1 | -1.1 | -0.9 | ||
| North West | 0.4 | 0.2 | 7.9 | -1.9 | -1.0 | -0.8 | ||
| Yorkshire & the Humber | 0.3 | 0.2 | 8.4 | -1.4 | -0.7 | -0.6 | ||
| East Midlands | 0.4 | 0.2 | 9.6 | -0.3 | -0.1 | 0.1 | ||
| West Midlands | 0.3 | 0.1 | 8.9 | -0.9 | -0.5 | -0.3 | ||
| East of England | 0.2 | 0.1 | 11.3 | 1.5 | 0.8 | 0.9 | ||
| London | 0.2 | 0.1 | 11.8 | 2.0 | 1.0 | 1.1 | ||
| South East | 0.2 | 0.1 | 11.0 | 1.2 | 0.6 | 0.7 | ||
| South West | 0.3 | 0.2 | 11.8 | 1.9 | 1.0 | 1.1 | ||
| England | ||||||||
| average | 0.3 | 0.2 | 9.8 | 0.0 | 0.0 | 0.2 | ||
| max deviation | 0.2 | 0.1 | 4.1 | 2.1 | 2.2 | |||
Back-of-the-Envelop Impact of Declining Council Tax Rates on House Prices
| (percentage points) | Dimension over time Effective tax rate | Dimension across Regions Real house price increase | Total | |||||
|---|---|---|---|---|---|---|---|---|
| Decline 97-03 | Impact on house price | (annual aver.) | deviation from mean | Impact on house price | Impact (row 2+5) | |||
| North East | 0.4 | 0.2 | 7.7 | -2.1 | -1.1 | -0.9 | ||
| North West | 0.4 | 0.2 | 7.9 | -1.9 | -1.0 | -0.8 | ||
| Yorkshire & the Humber | 0.3 | 0.2 | 8.4 | -1.4 | -0.7 | -0.6 | ||
| East Midlands | 0.4 | 0.2 | 9.6 | -0.3 | -0.1 | 0.1 | ||
| West Midlands | 0.3 | 0.1 | 8.9 | -0.9 | -0.5 | -0.3 | ||
| East of England | 0.2 | 0.1 | 11.3 | 1.5 | 0.8 | 0.9 | ||
| London | 0.2 | 0.1 | 11.8 | 2.0 | 1.0 | 1.1 | ||
| South East | 0.2 | 0.1 | 11.0 | 1.2 | 0.6 | 0.7 | ||
| South West | 0.3 | 0.2 | 11.8 | 1.9 | 1.0 | 1.1 | ||
| England | ||||||||
| average | 0.3 | 0.2 | 9.8 | 0.0 | 0.0 | 0.2 | ||
| max deviation | 0.2 | 0.1 | 4.1 | 2.1 | 2.2 | |||
The impact of the council tax rate on house price inflation over the 1997 to 2003 period has been calculated by adding up the product of the decline in the effective tax rate by the estimated semi elasticity of user costs of housing (-0.51) and the product of the difference in regional house price increases from the English average by the same semi elasticity. The council tax has uniformly contributed to house price inflation over time, though the extent of its contribution was minor (0.1 to 0.2 percent per year). Its contribution across regions was more significant, ranging from -0.9 in the North East to +1.1 in London and the South West.
21. The last decade has seen important regime changes that might have modified house price behavior. The estimation results need to be interpreted cautiously. There are some econometric shortcomings, such as the short time period, the representation of expected capital gains, or the omission of potentially important variables.19 More importantly, the 1990s was a period of regime changes, in particular monetary policy was reoriented towards inflation targeting. As a result consumer price inflation converged towards a lower rate and became more stable, as did the economy as a whole. Also house prices, have been less volatile (see Figure IV.4 above) since the mid-1990s. Finally, the council tax was introduced in 1993. However, to what extent this represents a regime change rather than a modification of previous tax regimes remains an open question.
D. Conclusions
22. The council tax may explain a significant part of cross-regional differences in house price inflation, but its influence over time was minor. Although property taxes contribute to the real user cost of housing, they have only a minor impact compared to the effects of mortgage interest rates and expected capital gains. At about one percent of actual property values, the effective council tax rate is of a comparable magnitude with rates in other countries in the EU and the USA, but is too small in absolute size to have a major impact on house prices at high house price inflation rates. Consequently, the direct effect of a reformed council tax on house price volatility in the United Kingdom is likely to be small.
23. A reformed property tax could nevertheless make a contribution to reducing overall macroeconomic volatility through its effect on disposable income. The contribution that a reformed council tax could make would have to work through income withdrawal in response to strong house price increases. This could then help stabilize overall economic activity to the extent that some households face constraints that prevent them from borrowing to smooth consumption. Annual property reassessments would ensure that the council tax no longer suffers from the adverse decline in its effective rate when house prices rise. Also, a uniform national council tax rate would help to reduce regional disparities.
24. Even so, these stabilization effects would come at the expense of local discretion over revenues, and might reduce certainty with respect to local government income. Reforming property taxes will therefore have to balance potential, and probably small, macroeconomic benefits against local governments’ interests in greater fiscal autonomy and predictability of revenues. To better inform its decision making, the U.K. government has appointed Sir Michael Lyons to carry out an independent inquiry into the detailed case for changes to the present system of local government funding. The inquiry is also supposed to make recommendations by the end of 2005 on how best to reform the council tax, taking into account the forthcoming revaluation of domestic property.
APPENDIX IV.1
House Prices and Expectations by English Regions
(annual percentage change)
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
House Prices and Expectations by English Regions
(annual percentage change)
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
House Prices and Expectations by English Regions
(annual percentage change)
Citation: IMF Staff Country Reports 2005, 081; 10.5089/9781451814286.002.A004
References
Adam, S. and J. Shaw, 2003, A Survey of U.K. Tax System, IFS, Briefing Note 9.
Aoki, K., J. Proudman, G. Vlieghe, 2002, House Prices, Consumption, and Monetary Policy: A Financial Accelerator Approach, Bank of England, Working Paper 169.
Barham, G., 2004, The Effect of Taxation Policy on the Cost of Capital in Housing – A Historical Profile (1976 to 2003), in: Bank of Ireland, Financial Stability Report.
Barker, K., 2003, Review of Housing Supply – Securing our Future Needs, Interim Report – Analysis, Chapter 7.
Barker, K., 2004, Review of Housing Supply, Delivering Stability: Securing our Future Housing Needs, Final Report – Recommendations, Recommendation 2.
Choi, I., 2001, Unit Root Tests for Panel Data, Journal of International Money and Finance, 249–272.
Danmarks Nationalbank, 2003, MONA – A Quarterly Model of the Danish Economy.
Gallin, J., 2003, The Long-Run Relationship Between House Prices and Income: Evidence from Local Housing Markets, Federal Reserve Board DP.
HM Treasury, 2003, Fiscal Stabilisation and EMU: A Discussion Paper.
HM Treasury, 2003, Housing, Consumption and EMU: EMU study
Institute for Fiscal Studies, 2004, Green Budget, 5. The taxation of housing.
Kearl, J.A., 1979, Inflation, Mortgage, and Housing, The Journal of Political Economy, 1115–1138.
Maclennan, D., J. Muellbauer, M. Stephens, 1998, Asymmetries in Housing and Financial Market Institutions and EMU, Oxford Review of Economic Policy, 54–80.
Meen, G. and M. Andrew, Modeling Regional House Prices: a review of the literature, Department of the Environment, Transport and the Regions.
Meen, G. and M. Andrew, 1998, On the Aggregate Housing Market Implications of Labour Market Change, Scottish Journal of Political Economy, 393–419.
Miles, D., 2004, The U.K. Mortgage Market: Taking a Longer-Term View, HM Treasury.
Muellbauer, J and A. Murphy, 1997, Booms and Busts in the U.K. Housing Market, The Economic Journal, 1701–1727.
Muellbauer, J., 2003, The U.K. and the Euro – the Role of Asymmetries in Housing and Credit Markets, EMU Studies 19.
Muellbauer, J., 2004, Property and Land, Taxation and the Economy after the Barker Review, Nuffield College, Oxford.
Muellbauer, J. and G. Cameron, 2000, Five Key Council Tax Reforms and Twelve Reasons to Enact Them, New Economy, www.housingoutlook.co.U.K./Papers/tax.html.
OECD, 2004, OECD Economic Surveys: United Kingdom, Chapter II: Reducing the risk of instability from the housing market.
Poterba, J.M., 1984, Tax Subsidies to Owner-Occupied Housing: An Asset-Market Approach, The Quarterly Journal of Economics, 729–752.
Poterba, J.M., 1991, House Price Dynamics: The Role of Tax Policy and Demography, Brookings Papers on Economic Activity, 2, 143–183.
Smith, L.B., K.T. Rosen, G. Fallis, 1988, Recent Developments in Economic Models of Housing Markets, Journal of Economic Literature, 29–64.
Thwaites, G., and R. Wood, 2003, The Measurement of House Prices, Bank of England Quarterly Bulletin: Spring 2003, 38–46.
Van den Noord, P., 2003, Tax Incentives and House Price Volatility in the Euro Area: Theory and Evidence, OECD Economics Department Working Paper No. 356.
Weeken, O., 2004, Asset Pricing and the Housing Market, Bank of England Quarterly Bulletin, Spring 2004, 32–71.
Wood, R., 2003, The Information Content of Regional House Prices: Can They Be Used To Improve National House Price Forecasts?, Bank of England Quarterly Bulletin: Autumn 2003, 304–314.
Prepared by Werner Schule.
See HM Treasury (2003), Housing, consumption and EMU, EMU study.
How Should Policymakers Respond to a Decline in House Prices?
D. Miles (2004), The U.K. Mortgage Market: Taking a Longer-Term View, HM Treasury; and K. Barker (2004), Review of Housing Supply, Final Report, HM Treasury.
Of the five current property (housing) taxes in the United Kingdom - capital gains tax, inheritance tax, VAT on repairs, stamp duty and council tax - stamp duty and council tax share the property that the effective tax rate tends to fall with rising house prices. For an overview of housing taxation in the United Kingdom see Institute for Fiscal Studies (2004).
Cameron and Muellbauer (2000) list 12 reasons for reforming the council tax.
A similar model was developed by J.R. Kearl (1979), and modified in many ways by Poterba (1984,1991), Muellbauer and Murphy (1997), Meen and Andrew (1998) and many others. This presentation draws heavily on Van den Noord (2003).
Beyond the user cost of housing channel, housing taxation affects disposable income, the allocation of savings and labor mobility.
Council tax revenues by English regions are available online from 1993/4 to 2004/5 only. Regional data on prices, income, and demographics are available over a longer time range (at least 1970 to 2003), but other potentially relevant factors are available only for shorter time periods (housing stocks by regions), or difficult to find (rents, regional construction costs). The community tax regime in England differs from those in Scotland and Wales, while Northern Ireland continues to tax properties according to the old national tax regime. Therefore the discussion here focuses on the more homogenous English regions.
Alternatively the tax adjusted mortgage rate rt(1 - ti, j) could be applied, using time varying tax rules, and marginal income tax rates across regions and time.
Following standard practice in the literature.
More precisely the ratio of average council taxes 1993/94 to 2003/04 for 2 adults in band D, before transitional relief and benefits (for 1997/98 to 2004/05 ODPM, Table 2.2f in sterling, for 1993/04 to 1996/97 consolidated local authorities series), to simple average house prices of all dwellings (ODPM, Table 511, in sterling).
Annual regional house price inflation was HP filtered using the standard θ=100. Unadjusted for end-point problems for 1970 to 2003, adjusted for 1970 to 2008.
Muellbauer and Murphy (1997) found that house prices in the South East are predicting U.K. house prices. Wood (2003) confirmed a leading role of house price inflation in London and the South East for the late 1980s and early 1990s, but not during other periods. In line with Wood’s observation, the HP filter was applied to each region separately.
The decline in volatility is also visible in some regions outside London, see Figure IV.A1 in the appendix.
The picture looks similar for the other regions (available on request).
House prices fell considerably earlier in London than in the northern regions.
Efforts to include the housing stock by regions were unsuccessful, producing implausible signs; good rent data were unavailable, neither were data on regional wealth holdings.
