Kingdom of the Netherlands—Netherlands
Selected Issues
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar
Close

1. Developments in house prices are of considerable interest in the Netherlands. During the boom of the late 1990s, the rapid increase in house prices added to household wealth and helped boost consumer spending. 29 As the momentum of house price increases moderated considerably in the new millennium, some of the fire under consumer spending was extinguished. While the recovery has now gathered steam, falling house prices, in light of prospects for interest rate hikes, are an ever-present concern that could impact negatively on the economy as well as banks. In this context, it is important to take stock of recent developments in house prices, with a view to continuing to assess the risks that prices may be out of line with fundamentals or fall in the future. 30

Abstract

1. Developments in house prices are of considerable interest in the Netherlands. During the boom of the late 1990s, the rapid increase in house prices added to household wealth and helped boost consumer spending. 29 As the momentum of house price increases moderated considerably in the new millennium, some of the fire under consumer spending was extinguished. While the recovery has now gathered steam, falling house prices, in light of prospects for interest rate hikes, are an ever-present concern that could impact negatively on the economy as well as banks. In this context, it is important to take stock of recent developments in house prices, with a view to continuing to assess the risks that prices may be out of line with fundamentals or fall in the future. 30

II. House Prices in the Netherlands: Determinants, Concerns, and Considerations Related to Phasing out the Tax Deductibility of Mortgage Interest Payments28

A. Introduction

1. Developments in house prices are of considerable interest in the Netherlands. During the boom of the late 1990s, the rapid increase in house prices added to household wealth and helped boost consumer spending. 29 As the momentum of house price increases moderated considerably in the new millennium, some of the fire under consumer spending was extinguished. While the recovery has now gathered steam, falling house prices, in light of prospects for interest rate hikes, are an ever-present concern that could impact negatively on the economy as well as banks. In this context, it is important to take stock of recent developments in house prices, with a view to continuing to assess the risks that prices may be out of line with fundamentals or fall in the future. 30

2. Associated developments in household debt are also of economic importance. In this connection, residential mortgage indebtedness and household leverage in the Netherlands have been rising, reaching levels that are high compared with other industrialized countries. High debt can raise vulnerabilities, for banks and the economy more generally.

3. With various fiscal incentives making mortgage debt attractive, this paper pays special attention to the tax deductibility of mortgage interest payments. The Financial System Stability Assessment of 2004 recommended, over the medium term, to phase out the tax deductibility of mortgage interest, preferably in a gradual fashion to avoid disruptive effects. 31 Among the reasons for eliminating this incentive, (i) tax deductibility of mortgage interest introduces a distortion in the housing markets—making holding mortgage debt attractive and reducing the incentive to pay back the principal; (ii) this policy favors wealthy households, as the tax advantage increases with higher tax brackets and is larger the larger the real estate assets of a household; (iii) removing tax deductibility would help the budget in the context of rising costs related to population aging; and (iv) the revenue received from removing tax deductibility could be used for policies that more directly aim to make housing affordable for low-income households that currently cannot afford a house due to high house prices. Against this background, and based on the experience of other countries that removed tax deductibility of mortgage interest payments, this paper attempts to quantify and shed some light on the expected price impact of phasing out this incentive.

4. The paper is organized as follows: Section B summarizes recent developments in the housing market in the Netherlands, with Section C providing a cross-country comparison. Section D reflects on the unusual developments in housing markets in industrialized countries. Section E discusses policy issues, in particular the tax deductibility of mortgage interest payments. Section F draws lessons from the experience of other countries that phased out mortgage interest deductibility. Section G discusses how tax deductibility of mortgage interest payments could be phased out, drawing on the U.K. and Swedish experiences. This is followed by concluding remarks in Section H.

B. What is the Current Situation in the Housing Market in the Netherlands?

5. House prices accelerated again after a brief slowdown. Following two decades of strong house price inflation, house price growth decelerated sharply in the fourth quarter of 2000 and then remained subdued until recently. In 2005, prices accelerated again, with the quarter-over-quarter increase in the third quarter registering 3.1 percent. Whether this increase is just a temporary revival in prices or reflects a return in house price growth to its exceptional upward trend remains to be seen.

6. Reports suggest that credit quality is deteriorating because of the continued expansion of mortgages. This is not all that surprising, since most European banks have been struggling to find growth and profit opportunities in an increasingly competitive banking environment. In many European countries, banks have been faced with a drop in corporate demand for loans, and residential real estate loans have been one of the few growth opportunities. The Netherlands has been no exception in this trend: the debt-to-equity ratio of the corporate sector fell from 69 percent in 1998 to 32 percent in 2004, causing a shift in the loan portfolio of banks; corporate sector loans of the banking sector fell from 38 percent of total loans to 30 percent over the same period; and, in contrast, household sector loans as a percent of total loans remained constant at 50 percent over that period. At the same time, mortgages guaranteed by the National Mortgage Guarantee have been expanding rapidly and reached €60 billion in 2004. For these mortgages, banks have little incentive to scrutinize the loans, as they are essentially risk free.

7. Meanwhile, households have become more sensitive to interest rate changes. Reports by De Nederlandsche Bank (DNB) indicate that the proportion of mortgages subject to an interest rate adjustment within two years increased from 21 percent in 2003 to 31 percent in 2005. Moreover, the percentage of newly provided loans with loan-to-value ratios (LTV) exceeding 100 percent has been increasing significantly (Figure 1). Households are also more indebted, insofar as the ratio of mortgage debt-to-GDP is high at more than 100 percent and has been increasing.

Figure 1.
Figure 1.

The Netherlands: Percentage of Newly Provided Loans with LTV>100%

Citation: IMF Staff Country Reports 2006, 284; 10.5089/9781451829556.002.A002

C. How does the Situation in the Netherlands Compare to Other Industrialized Countries?

8. Price developments in the Netherlands have, in a number of respects, been unique. A feature that differentiates the Netherlands from most other European countries is that house prices have been increasing for two decades. One of the few countries that had a similarly long expansion is Belgium. However, while prices in Belgium have been growing at a moderate pace for this entire period, price increases in the Netherlands accelerated in the 1990s. Hence, if one were to look at the cumulative real house price increase from trough to peak, the increase of 103 percent in the Netherlands surpasses all its neighbors. In addition, countries such as the United Kingdom and Spain have been experiencing much more volatile housing markets, with real estate bust and booms over shorter cycles. Table 1 and Figure 2 provide an overview of the differences among various countries in their real estate price cycles.

Table 1.

Accumulated Inflation-Adjusted House Price Increases

(Percent cumulative change as of end-2004)

article image
Sources: Inflation-adjusted real estate prices are from BIS calculations using national data; and author's calculations. Note: Nominal price increases are deflated by CPI.
Figure 2.
Figure 2.

Inflation-Adjusted Residential Poperty Prices 1/

Citation: IMF Staff Country Reports 2006, 284; 10.5089/9781451829556.002.A002

Sources: Provided by the national authorities; and BIS.1/ BIS calculations based on national data, cumulative real growth rates (1970=0).

9. On a cross-country basis, mortgage debt in the Netherlands is among the highest in industrialized countries (Table 2). In 2004, the ratio of overall mortgage debt to GDP was 46 percent in EU-15 countries. At 111 percent in the Netherlands, this ratio was the highest in Europe, followed by Denmark at 90 percent and the United Kingdom at 73 percent. Even in the United States, mortgage debt-to-GDP at 65 percent is significantly lower than in the Netherlands. Per capita mortgage debt in the Netherlands was around €32,000, surpassed only by Denmark. Furthermore, despite the already high debt level, mortgage debt in the Netherlands increased again in 2004 by 14 percent, compared with increases of 11 and 6 percent, respectively, in the United Kingdom and Denmark.

Table 2.

Mortgage Debt in 2004

article image
Source: European Mortgage Federation.

10. The Netherlands is one of the most densely populated countries in Europe, and supply restrictions tend to contribute to the excess demand for housing that has been met by price increases. With 478 people per square kilometer, undeveloped land is in short supply. In addition, zoning laws are very strict. Consequently, the supply elasticity of housing is usually regarded as very low.

D. Is There a Reason to be Concerned?

11. Viewed from a cross-country perspective, the current real estate boom cycle in industrial countries is unusual by historical standards. The unprecedented low interest rate environment and the development of new mortgage products contributed to housing markets reaching unprecedented highs in many countries. Moreover, housing price cycles have generally tended to last only 3–4 years, with cumulative price increases averaging around 30 percent during the boom phase and, in those cases of a complete cycle, the entire increase was often reversed during the bust phase. 32 By comparison, the duration and the magnitude of the current expansion in prices in several industrial countries has surpassed this record by far.

12. No country is immune to real estate boom and bust cycles—the Netherlands is no exception. Expert and media accounts of residential property markets—historically and across countries—describe price developments in these markets in terms of the recurrence of booms and busts;33 the empirical literature confirms cyclicality and high autocorrelation in house prices. 34 Even though somewhat distant from the present, the Netherlands did experience a real estate boom in the 1970s followed by a painful and complete reversal of the price gains in the early 1980s. What is striking about the Netherlands it that the current price increases have been exceptional both on a cross-country basis and compared to its own historical experience.

13. Estimating whether the housing market can be characterized by overpricing can be tricky and varies from study to study. One reason for large discrepancies among studies is differences in the sample period chosen and the period over which the overvaluation is calculated. Further, estimation methodology differs from study to study, and estimates using annual vs. quarterly data can differ substantially. Finally, in all studies, coefficients are somewhat unstable and jump around depending on inclusion and exclusion of various variables. 35 Having listed these reservations, the estimation by Hofman (2005), using a quarterly sample during 1970–2004, suggested that house prices (as of the second quarter of 2004) had not moved beyond what could be explained by changes in fundamentals. Other studies, by comparison, found evidence of overpricing: a recent study by Verbruggen and others (2005), using an annual sample from 1980–2003, found Dutch house prices “somewhat” overvalued; and Van den End and Kakes (2002) found a positive long-run correlation between stock market and house prices, at a two to three year lag, suggesting that there may have been substantial scope for downward movement in current house price levels at the time of the study, in light of the severe losses in the Dutch stock market at the beginning of the new millennium.

14. Nonetheless, all studies suggest that housing prices are highly sensitive to interest movements, so housing prices could fall and negatively affect consumption if interest rates were to rise further. In this regard, Hofman (2005) finds that a 1 percent increase in the real interest rate reduces the long-run equilibrium house price by about 10 percent. Verbruggen and others (2005) confirm the significant impact of interest rates on housing prices. Further, various studies suggest that aggregate consumption is sensitive to house price developments. According to calculations using CPB’s quarterly model SAFE, for the period 1995–2002, 40 percent of consumption growth could be attributed to the increase of households’ stock of dwellings. Estimates by the IMF (Nadal de Simone, 2005) suggest that consumption is sensitive to changes in house prices insofar as the long-run elasticity of consumption with respect to per capita real net wealth is 0.21.

15. An interesting and important result from a study by the CPB is that price developments are asymmetric. In this connection, Verbruggen and others (2005) show that downward adjustment in prices are slower. Nevertheless, a prolonged adjustment process could still potentially keep consumption at a lower level for a longer time period.

16. Our estimates suggest that real house prices in the Netherlands are highly responsive to changes in disposable income and the real interest rates (Table 3). The model in this paper explains historic price developments quite well (Figure 3). Further, the coefficients are highly significant, stable, and do not change significantly if we change lag length or change the sample period. It is likely that this is due to the inclusion of the nominal interest rate in the cointegration relationship, after correctly identifying it as a stationary variable with a trend. 36 Our findings suggest that the 1990s episode, with its very large price increases, was similar to the 1970s episode when housing markets were overheated (in the sense that actual prices exceeded fitted values). However, when prices slowed down in the new millennium, this allowed fundamentals and house prices to come more in line. The current pickup in prices may lead to a renewed upward deviation from fundamentals—and it remains to be seen whether the upward trend in prices will continue in the future. Nonetheless, our model predicts that, should fundamentals change sufficiently—for example because of rising interest rates—house prices would experience a downward adjustment.

Table 3:

The Netherlands: Regression Results, 1970-2005

article image
Source: IMF staff estimates.

Critical value 0.17.

Denotes signifcance at the 5 percent critical value.

Figure 3:
Figure 3:

Netherlands: Real House Prices vs. Fitted Values from Estimation

Citation: IMF Staff Country Reports 2006, 284; 10.5089/9781451829556.002.A002

17. The risks in the system also seem to have moved into unchartered waters. Current indebtedness levels in industrial countries and particularly in the Netherlands are unprecedented. Further, various new mortgage products have allowed households to leverage themselves to the limit, not pay back principal, and take on interest rate risk and other risks associated with the increased utilization of these products. With price developments in various countries, including the Netherlands, also exceptional compared to other cycles, it would seem that countries’ abilities to deal with the associated risks have not been tested historically. In a related vein, at this juncture, it makes sense to review policies that encourage households to leverage themselves excessively or policies that affect banks’ lending practices.

E. Policy Issues

18. Various steps could help reduce the incentives for the risks that households and banks are taking. This paper concentrates on the tax deductibility of mortgage interest payments but briefly discusses other issues.

19. The NHG guarantee encourages banks to extend mortgage loans and shifts the risk of default to the government. For banks, mortgages guaranteed by the NHG carry no credit risk, and are inexpensive, since mortgage loans guaranteed by the NHG have a risk weighting of zero. In contrast, the current solvency requirements for Dutch banks regarding the treatment of mortgages provide for a distinction according to the (perceived) riskiness of the loan. Mortgage loans that do not exceed 75 percent of the liquidation value of the property have a risk weighting of 50 percent. When the mortgage loan exceeds 75 percent of the liquidation value of the property—so called top mortgages—the “loan surplus” on top of the liquidation value will have a risk weighting of 100 percent. While the reasons for NHG guarantees may be commendable to some on social grounds, the implicit cost of this guarantee is covered by the government. That this guarantee comes at no cost to banks and households introduces a price distortion.

20. Tax deductibility of mortgage interest payments is another fiscal advantage that strongly encourages households to leverage themselves. Various empirical studies found strong evidence that household leverage is highly sensitive to the tax advantage created by interest rate deductibility (Dunsky and Follain, 2000; Follain and Dunsky, 1997; Follain and Ling, 1991; Ling and McGill, 1998; and Hendershott, Pryce, and White 2003). These studies show that removing interest deductibility provides a strong incentive to homeowners with existing loans to pay off their loans and causes new homeowners to choose loans with less leverage. This is particularly true for households that are not income-constrained and are in higher tax brackets, since they tend to chose the higher leverage because of tax advantages.

21. In addition, the extent to which interest deductibility meets social goals is not entirely clear. Even though this policy is intended to make housing more affordable, it is not clear whether it does so, since housing prices adjust to tax incentives. Further, interest deductibility favors households in higher tax brackets—thus using foregone government resources in ways that are less advantageous to those of lower income. It also encourages households to leverage themselves while reducing the incentive to pay off the principal part of the loan.

22. The next section takes a selective look at the country experience in removing the tax deductibility of mortgage interest payments. In particular, the experience of the United Kingdom and Sweden suggests that the timing of the phase-out of the tax deductibility of mortgage interest payments in terms of prevailing economic conditions, as well as the speed with which the phase-out occurs, are important variables to keep in mind for policy design.

F. The Experience of Other Countries that Phased Out the Tax Deductibility of Mortgage Interest Payments

The United Kingdom

23. One country that has gradually but fully eliminated the tax deductibility of mortgage interest payments is the United Kingdom. The United Kingdom phased out the tax deductibility over the course of 25 years (1974–99). Before 1974, mortgage interest payments in the United Kingdom were fully deductible at the marginal tax rate. In 1974, a ceiling was introduced on the size of mortgages that were eligible for tax deductibility of interest payments. This ceiling was set at £25,000, and households with a mortgage loan that exceeded this limit could only deduct interest payments for the portion that was at or below this limit. In 1983, this limit was raised to £30,000. Subsequently, the limit was never raised again. Later, the maximum tax rate at which interest could be deducted was cut from a 40 percent maximum income tax rate to 25 percent in 1992, to 20 percent in 1994, to 10 percent in 1995, and finally down to zero in 1999 (Table 4).

Table 4:

United Kingdom: The Gradual Phase-out of Tax Deductibility of Mortgage Interest Payments

article image
Source: Hendershott, P.H., Pryce, G., White, M. "Household Leverage and the Deductibility of Home Mortgage Interest: Evidence from U.K. House Purchasers," Journal of Housing Research, Vol. 14, Issue 1.

24. The success of the U.K. experience reflected the introduction of a nominal ceiling—on the size of mortgage loans that qualified for tax deductibility of mortgage interest payments—that became binding over time (Figure 4). In 1974, when the ceiling was set at £25,000, the median price of a house in the United Kingdom was only £10,800, and just a few very rich households were affected by this restriction. However, since the cap was nominal and the limit was only raised once (to £30,000), the rapidly increasing house prices and the rapidly increasing loan sizes made this limit binding for the majority of households in a gradual manner. By 1988, about half the new mortgage originations were above this limit, and by 1995, two thirds were above the limit (Table 4). It was fortuitous that the ceiling was introduced towards the end of a recession and just before an upturn, as there was immediate upward pressure on house prices after the introduction of this policy. Thus, the success of the U.K. experience also demonstrates how timing is important.

Figure 4:
Figure 4:

United Kingdom: Limit on Deductibility Became Binding Over Time

Citation: IMF Staff Country Reports 2006, 284; 10.5089/9781451829556.002.A002

25. With the reform taking place over 25 years, the phasing-out of the tax deductibility of mortgage interest payments appeared to have a limited impact on house prices in the United Kingdom. Figure 5 shows changes to the tax deductible amount for a representative household in real terms and the cumulative real house price increases in the United Kingdom. For the calculation of the tax deduction of a representative household, a representative household is defined as a household that buys a house at the average house price in that year, using the average housing loan that has been recorded for that year. 37 Further, it is assumed that, as a new mortgage holder, the representative household pays interest on the mortgage loan according to the interest rate in that year. Hence, the mortgage interest a representative household pays and the tax deduction not only change due to changes in the tax law but are also sensitive to interest rate fluctuations in the market and the average price of a new house. Notice, for example, that from 1989 to 1993, the interest rate fell from 13.3 to 5.2 percent, and this fall reduced the interest burden of households; at the same time, it implied that the tax they could deduct declined. It is noteworthy that lowering the deductible rate after 1992 led to a sharp drop in the deductible amount. Further, while from 1989 to 1994, the drop in tax deductibility coincided with a drop in house prices, after 1996, house prices increased strongly despite a complete removal of tax deductibility by 1999—though, of course, the removal of tax deductibility was not the only factor at play. Thus, an econometric exploration was also tried.

Figure 5:
Figure 5:

United Kingdom: Tax Deduction, House Prices, and Key Economic Developments

Citation: IMF Staff Country Reports 2006, 284; 10.5089/9781451829556.002.A002

26. Econometrically, the impact of tax deductibility on U.K. house prices seems to have been limited (Table 5). Using disposable income, long-term interest rates, and tax deductibility as explanatory variables for real house price changes, we estimate the impact of having removed tax deductibility on real house prices. The coefficient on tax deductibility is small but significant at the 10 percent level. While these estimates may underestimate the impact of tax deductibility on housing prices because of the correlation of this variable with disposable income and interest rates, the impact nevertheless seems moderate. Though difficult to prove, this may well have reflected the long time period over which tax deductibility was phased out.

Table 5.

United Kingdom: Regression Results

article image
Source: IMF staff estimates.

Denotes significance at the 5 % critical level

Denotes signifiance at the 10% critical level.

The 10 percent critival level is 0.15.

Sweden

27. Sweden reduced the tax deductibility of mortgage interest payments over a short span of time (1985–91). Sweden reduced the maximum deductible rate that could be applied to mortgage interest payments to 50 percent in 1985 and 30 percent in 1991. Prior to this reform, the maximum deductible rate was 80 percent. After the reform, households that were in a higher marginal tax category than these maximum limits had to apply these limits when calculating their deduction, rather than their actual marginal income tax rate.

28. In the case of Sweden, it is difficult to isolate the impact of the partial phasing out of tax deductibility on housing prices, due to a number of important changes that happened at the same time. In addition to changes to the maximum deductible rate, Sweden went trough a number of significant tax reforms in the late 1980s and early 1990s. In 1991, taxes on labor income were reduced to a basic marginal tax rate of 30 percent and a top marginal tax rate of 50 percent (previously, the top marginal tax rate could be as high as 75 percent). In addition, in order to reduce inter-asset tax distortions and reduce overinvestment in housing and durable goods, various changes were made to the taxation of owner-occupied housing and other financial assets. 38 Further, the country went through a recession in the early 1990s. Thus, for Sweden, due to various data limitations, and the changes to several taxes over the same time period, it was decided that it would be impractical, if not implausible, to try to empirically isolate the impact of changes to tax deductibility of mortgage interest payments on house prices. 39

29. Nonetheless, a graphical analysis of house price developments suggests that the timing of the phasing-out of the tax deductibility of mortgage interest payments could have been better from the standpoint of the economic conjuncture. Many of the changes in taxes and the final reduction in the rate applied to the mortgage interest payment deduction (to 30 percent) were introduced in 1991. This was in the middle of a recession and during a time when house prices had already stagnated and started to fall. Thus it would seem reasonable to conclude that the sizable reduction in the tax deduction only compounded the impact of the recession in contribution to the drop in house prices during that period. By comparison, when Sweden first reduced the maximum rate that applied to the mortgage interest payment deduction (from 80 to 50 percent in 1985), this occurred in the midst of strong economic growth, the change only affected high-income households with very high marginal tax rates, and real house prices experienced spectacular growth, despite the partially offsetting effects of the reduction in deductibility.

Figure 6.
Figure 6.

Sweden: Real House Prices and Some Variables That Impact House Prices

Citation: IMF Staff Country Reports 2006, 284; 10.5089/9781451829556.002.A002

G. When and How to Phase out Mortgage Interest Deductibility in the Netherlands?

30. If policy makers were to decide to phase out the tax deductibility of mortgage interest payments, they would need to make decisions on certain aspects of this policy—with various implications for households and house prices. In particular, illustrated below for the Netherlands is the relevance of the timing of the phase-out (in terms of the economic conjuncture) and the speed of the phase-out (that is, the time period over which policies are binding), with some quantitative analysis. In the analysis, it is assumed that a loss in household income when tax deductibility is reduced can be treated in the same way as a drop in disposable income. The estimates of the determinants of real house prices in the Netherlands (Table 3) are then used to quantify the price impact of future changes in tax deductibility. Given the limitations of any modeling exercise of this kind, the numbers presented are not meant to be actual forecasts of what housing prices might be. Rather, they should be taken as being illustrative of the fact that policy options matter as do the economic circumstances under which policy is implemented.

31. Three policy options are analyzed. Policy 1: An immediate and complete removal of tax deductibility. Policy 2: Introduction of a nominal limit on the size of the mortgage loan that qualifies for tax a deduction. The nominal limit is set at the average size of a mortgage loan in 2005. Hence, because mortgage loans only qualify for a tax deduction up to that limit, this option becomes gradually more binding as house prices increase. Policy 3: A reduction of the deductible rate—meaning that the tax rate that would apply to the mortgage interest payment deduction would be below the marginal rate that would otherwise apply to income. Figure 7 summarizes the impact of these policy options on real house prices, based on one of three economic scenarios, namely “normal economic growth.” The assumptions behind this economic scenario are presented in Table 6, along with the assumptions behind the two other economic scenarios under which the different policy options are analyzed.

Figure 7.
Figure 7.

Netherlands: Average House Price Forecast Assuming Normal Economic Growth

(Real log house prices)

Citation: IMF Staff Country Reports 2006, 284; 10.5089/9781451829556.002.A002

Table 6.

Netherlands: Household Categories

(In euros; unless indicated otherwise)

article image
Source: Ministry of finance, March 20, 2006 ("Actualisatie varianten fiscale behandeling eigenwoningbezit"). Note: Age: all households <65; income categories: low <£30,000; average €30-60,000; high €£60,000.

32. Because of differences in the average loan size for and the actual marginal tax rate of different income levels, changes to tax deductibility affect low-, average-, and high-income households differently. Table 6 summarizes the key characteristics of different households in different income categories. For example, the average tax saving for a high-income household with an average loan size of £209,000 and a marginal tax rate of 50 percent is £5,585; whereas, the average tax saving of a low-income household, with an average loan of £127,000 and a marginal tax rate of 30 percent is £2,034. Hence, a complete removal of tax deductibility would lead to an immediate drop in disposable income for high-income households (low-income households) of £5,585 (£2,034). In the following analysis, taking into consideration the different loan sizes and different tax brackets for different income categories allows us to differentiate the price impact of phasing out tax deductibility for these three household categories.

33. Under Policy 1—an immediate and complete removal of the tax deductibility of mortgage interest payments—there is a significant drop in house prices for all households. For example, if the removal of tax deductibility were to coincide with the modest recession scenario, then real house prices for high-income households would decline by 20 percent in the first year after the policy change. If the modest recession were to continue, the real house price decline for high-income households would reach 36 percent at the end of the fifth year. A more severe recession, or higher interest rates, would lead to a more severe drop in real prices.

34. Under Policies 2 and 3—which show the tax deductibility of mortgage interest rates payments reduced over time (Figure 7 and Table 7):

Table 7.

Netherlands: Price Impact of Phasing Out Tax Deductibility

(Percent change in real house prices)

article image
Policy Scenarios: Policy 1: Immediate abolition of tax deductibility of mortgage interest rates. Policy 2: Mortgage loans that qualify for mortgage interest rate deduction are capped by a nominal limit. The nominal limit is set at €141,000, which is the average mortgage loan in the Netherlands in 2005. Policy 3: The marginal tax rate at which mortgage interest payments can be deducted from income is reduced to 30%. Households: HH1: Low-income household; HH2: Average-income household; HH3: High-Income household.
  • The drop in house prices during a recession is dampened compared with Policy 1. For example, one year after a policy change, the price drop for average income households during a modest recession would be -12.4, -3.0, and -5.2 under Policies 1, 2, and 3, respectively.

  • Inflation pressures are mitigated during an economic expansion compared to no change in tax deductibility. Without any policy change, there is a 29.1 percent (assuming normal economic growth) and 48.8 percent (assuming exuberant economic growth) real house price increases after six years. Introducing a nominal limit on mortgage loans that qualify for interest rate deduction (Policy 2) would moderate these expected price increases during a period of economic expansion. For example, after six years (assuming normal economic growth), real house prices increase by 26.6, 22.2, and 14.4 percent, for low-, average-, and high-income households, respectively, figures lower than under the no-policy-change scenario. Broadly similar results hold under exuberant growth, and the implications on real house prices under Policy 3 are similar to Policy 2.

35. Different policy choices also have additional consequences.

  • Policy 3 would be progressive but would have a procyclical element to it. In terms of distributional consequences, Policy 3 affects all tax-paying households in sufficiently high tax brackets as soon as it is introduced. Low-income households are, however, not affected by this policy change, whereas there is a large and immediate impact on average- and high-income households—thus, the progressive element to this approach. Under Policy 3, the foregone tax benefits are greater as interest rates rise. Since an increase in interest rates also puts downward pressure on house prices, the change in the deductible rate would compound the effect of an interest rate increase in putting downward pressure on house prices—thus having a procyclical element.

  • Policy 2 would be progressive, too, but in contrast to Policy 3, it would also have more of a countercyclical element to it. Households with high incomes and high loan values would be affected immediately. The larger a household’s loan, the larger would be the foregone tax benefit. Low- and medium-income households would have time to adjust to the situation more gradually, since the nominal limit would only become binding for them over time, as house prices increase. If rising interest rates were to put downward pressure on house prices, the nominal limit would become less binding as prices fall—thus having a countercyclical element. In addition, if interest rates rise, households would benefit from higher deductions offsetting somewhat higher interest payments.

H. Concluding Remarks

36. Low interest rates, a shift of bank lending toward the mortgage market, and new mortgage products have contributed to buoyant house price developments in the Netherlands in the last two decades. On standard analysis, price increases have been considerable, raising the possibility of a slowdown or even a temporary reversal if interest rates were to rise. A slowing or a reversal in house price increases has the potential to weaken consumption and growth.

37. House price increases and mortgage debt in the Netherlands have surpassed the experience of other industrial countries. Even though fundamentals explain a large part of the price increases, price developments, compared to historical values and on a cross-country basis, have been exceptional. Further, mortgage debt in the Netherlands is higher than in any other industrialized country, reaching unprecedented heights.

38. The development of new mortgage products in the Netherlands and the increased accessibility of loans are welcome from an efficiency perspective, but associated risks—well recognized by the Dutch officials—need to be well managed. Banks overall seem on a strong financial footing. However, a rise in interest rates is likely to impact households, in particular those with variable rate loans and the marginal households that were just able to afford a house may experience a cash flow squeeze. Generous interest deductibility has encouraged households to leverage themselves. Further, government-financed insurance schemes for mortgage debt (NHG) have shifted some of the risk from banks to the government, and thus the taxpayer, and banks have little incentive to scrutinize mortgage contracts under NHG guarantee.

39. The vulnerability of the economy and financial system to developments in real estate markets can be reduced. In the first place, adequately educating the public about the risks associated with different kind of mortgage loans, which is part of the code of conduct for mortgage lenders, is essential. With the previous real estate bust a memory of the distant past, households may not be taking sufficiently into account the possibility of a downturn in house prices. Households may also concentrate excessively on their budget for the near term and the affordability of a loan in the first year, rather than taking a sufficiently long-term view on the mortgage product they purchase. Second, since the NHG guarantee comes at no cost to households and banks, it introduces a price distortion. Finally, a phase-out of the tax deductibility of mortgage interest payments would reduce the incentive of households to leverage themselves, and some of the additional tax revenues as a consequence of the reform could be directed to well-targeted programs for low-income families. Additional revenues would also leave room for tax cuts, which could lessen the downward impact on housing demand from the phase-out and stimulate labor participation.

40. In thinking about ways to phase out the tax deductibility of mortgage interest payments, some key considerations would seem relevant. These are to: (i) have a progressive impact, with a larger share of the burden on the rich, especially in light of Dutch concerns about equity; (ii) avoid procyclicality whenever possible; and (iii) minimize, to the extent feasible, sudden disruptive effects. The latter concern points to considering a gradual phase-out of the deductibility of mortgage interest payments over time.

41. Judged against these considerations, the analysis in this paper indicates that introducing a nominal limit on the tax-deductible component of mortgage loans is an attractive option. As discussed earlier, particularly in paragraphs 34 and 35, it has progressive and counter-cyclical features. The quantitative analysis in Table 6 shows that it also has a gradual impact on households: the nominal limit tends to only become binding for low- and medium-income households over time, as house prices increase. And since this occurs more slowly than for high-income households, another progressive dimension is there. The option of reducing the deductible rate—meaning that the tax rate that would apply to the mortgage interest payment deduction would be below the marginal rate that would otherwise apply to income—also has positive features, but some elements of procyclicality make it less attractive. Of course, there is no guarantee with either option that forward-looking agents will not fully discount future changes, thus leading to sudden price drops. On the other hand, economic agents may put less weight on policy changes that will happen in the more distant future than those occurring more immediately, in which case a gradual phase-out will be less likely to have a sudden impact. In addition, with a gradual phase-out, households would continue to benefit from tax deductibility for a number of years, and it can be argued that they would therefore not adjust their portfolios suddenly. The quantitative analysis in this paper is in line with a more gradual decline in house prices when the tax deductibility is phased out over time and when implemented during a period of stronger economic growth. Further, while it is admittedly difficult to disentangle all the factors at play, the U.K. example is at least suggestive of the conclusion that the full impact of phasing out tax deductibility does not necessarily happen at once when the phase-out occurs over time. The U.K. example also underscores the importance of the timing of any policy initiative for avoiding disruptive effects.

APPENDIX I: Estimations for the United Kingdom

42. Augmented Dickey Fuller tests and Elliot, Rothenberg, and Stock tests were performed to identify unit roots (Tables A1 and A2). The nominal interest rate as expected is borderline stationary. In the Dickey Fuller test, if a constant and a trend is introduced in the unit root tests, this variable is stationary. We use a more powerful test, the Elliot, Rothenberg, and Stock to confirm that this variable is stationary. The Null hypothesis of unit root can be rejected at the 5 percent critical value.

Table A1.

United Kingdom: Augmented Dickey Fuller Test, 1970-2005

article image
Source: IMF staff estimates.

Denotes rejection of null hypothesis at 5% critical value. Null Hypothesis: series has unit root.

Optimal lag length is chosen using BIC and by looking at the autocorrelation in the residuals. Priority is given to removing autocorrelation.

Table A2.

United Kingdom: Elliot, Rothenberg, and Stock Test for Unit Roots, 1970-2005

article image
Source: IMF staff estimates.

Denotes rejection of null hypothesis at 5% critical value. Null hypothesis: series has unit root.

Optimal lag length is chosen using BIC and by looking at the autocorrelation in the residuals. Priority is given to removing autocorrelation.

The 5 percent critical value is -2.89.

43. We find no cointegration for the United Kingdom data in our sample (Table A3). It is often incorrectly assumed that for a cointegration relationship, all variables need to be I(1). However, near-integrated variables are often very important in establishing a sensible long-run relationship. If we include an I(0) variable in the cointegration relationship, the cointegration rank increases by one for each stationary variable included. Hence in the cointegration test, we are searching for two cointegration vectors. Using the Johansen-Juselius Maximum Likelihood Test for cointegration, we find no cointegration vector. Therefore, we use a model with first differenced variables and we include the interest rate in the regression as a detrended variable.

Table A3.

United Kingdom: Johansen-Juselius Maximum Likelihood Test for Cointegration, 1970-2005

article image
Source: IMF staff estimates.

Column r refers to the number of cointegration vectors.

The λ max and the trace statistics critical values are corrected for small samples using Cheung and Lai (1993).

Estimations for the Netherlands

44. Tables A4 and A5 summarize the results of the stationarity tests. Again, the nominal interest rate is a borderline case. Using the Elliott, Rothenberg test results we conclude that the nominal interest rate is a stationary variable after detrending the series.

Table A4.

Netherlands: Augmented Dickey Fuller Test, 1970-2005

article image
Source: IMF Staff estimates.

Denotes rejection of null hypothesis at 5% critical value. Null hypothesis: series has unit root.

Optimal lag length is chosen using BIC and by looking at the autocorrelation in the residuals. Priority is given to removing autocorrelation.

Table A5.

Netherlands: Elliot, Rothenberg, and Stock Test for Unit Roots, 1970-2005

article image
Source: IMF staff estimates.

Denotes rejection of null hypothesis at 5% critical value. Null hypothesis: series has unit root.

Optimal lag length is chosen using BIC and by looking at the autocorrelation in the residuals. Priority is given to removing autocorrelation.

The 5 percent critical value is -2.89.

45. The Results of the Cointegration test are as expected and including a stationary variable in the cointegration test leads to 1 additional cointegration rank. With two cointegration relationships and one stationary variable in the cointegration relationship, our system is correctly identified (Table A6).

Table A6.

Netherlands: Johansen-Juselius Maximum Likelihood Test for Cointegration, 1970-2005

article image
Source: IMF staff estimates.

Column r refers to the number of cointegration vectors.

The λ max and the trace statistics critical values are corrected for small smaples using Cheung and Lai (1993).

References

  • Agell, J., Berg, L., and Edin, P., 1995, “The Swedish Boom to Bust Cycle: Tax Reform, Consumption, and Asset Structure,” Swedish Economic Policy Review 2 (1995) pp. 271314.

    • Search Google Scholar
    • Export Citation
  • Bank for International Settlements, April 2005, “Real Estate Indicators and Financial Stability,” BIS papers, No. 21.

  • Ceron, J.A., and Suarez, J., 2006, “Hot and Cold Housing Markets: International Evidence,” Center For Economic Policy Research, Discussion Papers Series, No. 5411.

    • Search Google Scholar
    • Export Citation
  • Muellbauer, J., and Murphy, A., 1997, “Booms and Busts in the UK Housing Markets,” The Journal of the Royal Economic Society, No. 107, pp. 171127.

    • Search Google Scholar
    • Export Citation
  • Herring, Richard and Wachter, Susan (1999), “Real Estate Booms and Banking Busts,” Group of Thirty, Occasional Paper, No. 58.

  • Hendershott, P.H., Pryce, G., and White, M., “Household Leverage and the Deductibility of Home Mortgage Interest: Evidence from U.K. House Purchasers, Journal of Housing Research, Vol. 14, Issue.1.

    • Search Google Scholar
    • Export Citation
  • Hofman, D., 2005, “House Prices in the Netherlands,” Kingdom of the Netherlands—Selected Issues Paper, International Monetary Fund, pp. 25-33.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 2004, “The Global House Price Boom,” in World Economic Outlook, September, pp. 7189.

  • International Monetary Fund, 2003, “When Bubbles Burst,” in World Economic Outlook, April 2003, pp. 61-94.

  • Nadal Simone, Francisco. (2005), “Long-run Household Consumption Equilibrium in the Netherlands”, Kingdom of the Netherlands—Selected Issues Paper, International Monetary Fund, pp. 25-33.

    • Search Google Scholar
    • Export Citation
  • Verbruggen, J., Kranendonk, M. van Leuvensteijn, and M. Toet, 2005, “Welke factoren bepalen de ontwikkeling van de huizenprijs in Nederland?” CPB Document, No. 81, April.

    • Search Google Scholar
    • Export Citation
  • Scanlon, K., and C. Whitehead, 2004, “International trends in housing tenure and mortgage finance,” Council of Mortgage Lenders.

  • The Economist (2005), June 16, Special report on “The global housing boom: In come the wave.”

28

Prepared by Sibel Yelten.

29

The DNB estimated that housing price and asset increases added an average of 1–1½ percent to real GDP growth between 1998 and 2000, a key factor in explaining the comparative resilience of Dutch private consumption through 2001.

30

This follows on a selected issues paper from the 2005 Article IV Consultation, House Prices in the Netherlands by David Hofman. That paper did not find evidence of a deviation from fundamentals as of the second quarter of 2004 but also noted the limited comfort in the finding because the analysis also showed that the equilibrium price of housing was quite sensitive to interest rates.

31

The Kingdom of the Netherlands—Netherlands: Financial System Stability Assessment, including Reports on the Observance of Standards and Codes on the following topics: Banking Supervision, Securities Regulation, Insurance Regulation, Corporate Governance, and Payments Systems, and Anti-Money Laundering/Combating the financing of Terrorism, IMF Country Report N. 04/312, September 2004.

32

World Economic Outlook (April 2003) and Bank for International Settlements, BIS Papers, No. 21.

33

The special Report on “The global housing boom: In come the waves” published in The Economist on June 16, 2005 provides a good illustration of these views.

34

Ceron and Suarez (2006); Muellbauer and Murphy (1997); Herring and Wachter (1999); OECD Economic Outlook 75; World Economic Outlook, April (2003).

35

A good example that demonstrates this point is the study by Verbruggen and others (2005).

36

We estimate a cointegration relationship using the nominal interest rate and the disposable income per household. Since the nominal interest rate is stationary (with trend and constant), this variable is detrended in the short run relationship. See Appendix I for details.

37

The average tax advantage for this representative household is calculated in the following way. Before 1987, the average loan size is smaller than the deductible mortgage limit; therefore the average loan size, the maximum deductible rate, and the long-term interest rate in each year are used to calculate the tax advantage for an average household. After 1987, the cap on the mortgage loan size becomes binding for an average household, therefore the £30,000 limit is used to calculate the tax advantage, again using the maximum deductible percent and the interest rate in each year.

38

See Agell, Berg, and Edin (1995).

39

For example, the Swedish national income series for disposable income is very short and starts only in 1993; thus, it does not cover the time period when tax deductibility was reduced. To estimate the determinants of real house price developments in Sweden, it would be important to have a sufficiently long time series on disposable income, interest rates, and other data that would allow the quantification of the impact of each tax change in isolation.

  • Collapse
  • Expand
Kingdom of the Netherlands—Netherlands: Selected Issues
Author:
International Monetary Fund
  • Figure 1.

    The Netherlands: Percentage of Newly Provided Loans with LTV>100%

  • Figure 2.

    Inflation-Adjusted Residential Poperty Prices 1/

  • Figure 3:

    Netherlands: Real House Prices vs. Fitted Values from Estimation

  • Figure 4:

    United Kingdom: Limit on Deductibility Became Binding Over Time

  • Figure 5:

    United Kingdom: Tax Deduction, House Prices, and Key Economic Developments

  • Figure 6.

    Sweden: Real House Prices and Some Variables That Impact House Prices

  • Figure 7.

    Netherlands: Average House Price Forecast Assuming Normal Economic Growth

    (Real log house prices)