Denmark: Selected Issues
Author: International Monetary Fund. European Dept.
Publication Date:
27 Jun 2016
eISBN:
9781498319805
Language:
English
Keywords:
ISCR; CR; price; housing market; firm; housing stock; price divergence; supply issue; price overvaluation; developments in Denmark; Housing prices; Housing; Inflation; Property tax; Effective tax rate
Download PDF (592.7 KB)

This paper examines the selected issues related to the economy of Denmark: divergence in house prices, house prices in Denmark's cities, macroprudential policies, and product market reform and firm productivity. Recent house price developments in Denmark have been characterized by a growing divergence between different parts of the country, with big cities experiencing much more rapid price increases than other parts. House price booms and busts in Denmark, like in many other countries, are a big-city phenomenon. Macroprudential policies can help contain risks for households, the financial system, and the broader economy, but they should be carefully calibrated to avoid an undue drag on growth.

Abstract

This paper examines the selected issues related to the economy of Denmark: divergence in house prices, house prices in Denmark's cities, macroprudential policies, and product market reform and firm productivity. Recent house price developments in Denmark have been characterized by a growing divergence between different parts of the country, with big cities experiencing much more rapid price increases than other parts. House price booms and busts in Denmark, like in many other countries, are a big-city phenomenon. Macroprudential policies can help contain risks for households, the financial system, and the broader economy, but they should be carefully calibrated to avoid an undue drag on growth.

The Great Divergence: Regional House Prices in Denmark1

Recent house price developments in Denmark have been characterized by a growing divergence between different parts of the country, with big cities such as Copenhagen and environs experiencing much more rapid price increases than other parts. This paper examines the factors contributing to this price divergence. In an empirical analysis, municipal-level data are used to estimate the equilibrium house price level for each Danish city—the level explained by economic fundamentals—and to infer the extent to which municipal house prices may be overvalued or undervalued. The analysis points to emerging overvaluation in Copenhagen and Frederiksberg’s housing markets—particularly in the market for owner-occupied flats. The large difference in regional price trends may call for regional differentiation in macro-prudential and other housing policies to address housing market risks.

A. Introduction

1. There is a large divergence in house price patterns across Denmark. Following a large price correction after the 2006–07 housing boom, house prices started to recover in the Copenhagen region in early 2012, whereas in other regions (e.g., North and East Zealand) prices have started rising only recently. The average real price per square feet of housing2 in Copenhagen municipality increased by about 43 percent between 2012Q1 and 2015Q3—approaching its pre-crisis peak—compared to just over 11 percent in the median city. Over the first three quarters of 2015, prices also increased robustly in other major municipalities (e.g., Frederiksberg and Aalborg). The market for owner-occupied flats is generally seeing larger price increases compared to that for single-family homes—partly as flats are relatively concentrated in big cities. However, there is a wide distribution of price growth across the country, with some municipalities experiencing price declines and sluggish turnover, creating a ‘multi-speed’ housing market (Dam, Hvolbol, and Rasmussen, 2014; Pedersen and Isaksen, 2015).

A01ufig1

Real House Prices, 1992Q1-2015Q3

(index, 1992Q1 = 100)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A001

Sources: Statistics Denmark and Fund staff calculations.Note: Prices refer to the weighted average of prices (per square feet) of single-family houses and flats (weight is share of housing stock), deflated by CPI.

Growth of Real House Prices, 2015

(y/y percent)

article image
Sources: Statistics Denmark and Fund staff calculations.Note: 2015 figures cover the first three quarters.

Refers to detached/terraced houses

Refers to owner-occupied flats

2. For policymakers, it is important to monitor the extent to which house prices deviate from economic fundamentals. The housing market is of great importance to both financial and macroeconomic stability and a more stable housing market (without pronounced boom-bust cycles) would contribute to smoother economic development. While it is difficult to detect housing ‘bubbles’ in real time, it is helpful to gauge the degree of overvaluation or undervaluation in the housing market by comparing actual price levels to those that would be justified by demand and supply factors. One simple valuation measure—the price-to-income (PTI) ratio3—indicates that valuations in Copenhagen in 2014 were at a level comparable to that in 2004, right before the last price run-up. Meanwhile, median valuations have been much more subdued. The large regional differences in Denmark’s housing market may warrant a regional approach, both in terms of empirical analysis and policy implications.

A01ufig2

Price-to-Income Ratio in Copenhagen and Median City

(Index, 2000 = 100)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A001

Sources: Statistics Denmark and Fund staff calculations.

3. The paper is organized as follows: Section B documents the factors contributing to the regional differences in house prices across Denmark. Section C presents the analysis of equilibrium house prices using data from 96 Danish municipalities. Section D discusses the policy implications and concludes.

B. Factors Contributing to Regional Price Differences

4. The high demand for owner-occupied housing in Denmark’s large cities is underpinned by demographic trends. Population growth varies widely across the country; urban centers are becoming larger while some rural areas are losing population. For example, between 2012 and end-2015, the population of Copenhagen municipality increased by 42,500 inhabitants—equivalent to 7.7 percent growth—while the number of inhabitants in the municipality of Lolland was reduced by 2,600—or a decline of 6 percent. Much of the large influx of population into the capital is explained by immigration from outside Denmark, with internal migration playing a contributing but secondary role. According to Statistics Denmark’s projections, cities will continue to grow in size over the next few decades at the expense of rural areas. The age composition of large cities’ populations is also more conducive to home ownership. Currently, about 40 percent of Copenhagen’s residents belong to the 25–44 age group—those who are more likely to demand owner-occupied housing—compared to 25 percent for Denmark as a whole.

A01ufig3

Population Growth in Denmark’s Cities, 2015

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A001

Sources: Statistics Denmark and Fund staff calculations.
A01ufig4

Copenhagen: Contribution to Population Growth

(Number of people, in thousand)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A001

Sources: Statistics Denmark and Fund staff calculations.

5. Faster rising income and employment in big cities also exerts upward pressure on housing demand. Denmark’s labor market as a whole has been on a modest recovery track since 2012, but performance is uneven across regions, partly reflecting differences in the education and age composition as well as the types of jobs prevalent in each region. Over the period 2012–14 (2015 figure not yet available), real personal disposable income grew by 4.3 percent in the municipality of Copenhagen and 5 percent in Frederiksberg, compared to 3.5 percent for Denmark as a whole. Employment is also expanding more rapidly in the capital region than in other regions. Favorable economic and labor market trends reinforce the population trends by attracting more rural-urban migration as well as labor migrants from other parts of Europe.

6. Demand is also fueled by the favorable trends in the user cost of housing. The user cost depends on, among other factors, mortgage interest rates and housing taxes. Interest rates—both short and long term—have declined to very low levels in Denmark in recent years. The nominal freeze on valuations for the purpose of property value taxes introduced in 2002 implies that the effective tax rate has fallen relatively sharply in areas of rising house prices such as Copenhagen, since property taxes cannot increase with the rising market value of the property. Moreover, in 2016 the government has introduced a similar nominal freeze for land taxes, which will lower also the effective land tax rate in high price growth areas (Danmarks Nationalbank’s Monetary Review, 2015Q4).

7. Meanwhile, housing supply in big cities hardly keeps up with housing demand. Residential construction in Copenhagen—measured by either housing starts or completes—has picked up after the crisis and is now at a higher level than during the pre-boom years, but remains much lower than during the 2006–07 peak. Stricter zoning codes and land use regulations in Copenhagen combined with rapidly rising land prices are restricting new residential construction. As a consequence, demand for housing has outpaced supply in Copenhagen, with the housing stock expanding much slower than population (see also Chapter II which discusses housing supply issues).

A01ufig5

Copenhagen: Residential Construction

(1,000 sqm)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A001

Sources: Statistics Denmark and Fund staff calculations.
A01ufig6

Copenhagen: Population and Housing Stock

(index, 1992 = 100)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A001

Sources: Statistics Denmark and Fund staff calculations.

C. A Quantitative Perspective

8. This section presents a simple model relating municipal house prices to the demand and supply fundamentals discussed above. As in Meen (2002), house prices are often modeled using an inverted demand curve.4 The observed real price level in municipality i, Pit, is modeled as a function of the long-run equilibrium real house price, Pit*, which is determined by the housing stock Sit, demand shifters Xit, time-invariant unobserved municipal characteristics (e.g., amenities and locations—captured by the fixed effects αi), as well as common shocks (captured by time fixed effects μt):

Pit*=β1Sit+β2Xit+αi+μt+ɛit=Pit*+ɛit(1)

The demand shifters include municipal population in the 25-44 age group, municipal unemployment rate5, the real lending rate6, and the effective property tax rate7. An index of real construction costs is also included. All variables except for the lending rate and construction costs are available at the municipal level. Following Hort (1998), for robustness check, the baseline specification is augmented with the lags of the growth rates of all level variables to mitigate finite-sample bias.8 We estimate equation (1) separately for each market segment (single family houses or owner-occupied flats). The estimation sample covers 96 municipalities over the period 1996–2015.

9. Results indicate that these factors play significant roles in driving municipal house prices (Table 1). On the demand side, for example, a 10 percent increase in a municipal’s 25–44 aged population is associated with prices being 6.1 percent higher for single-family homes and 2.3 percent for flats. Prices are also higher in municipalities where the unemployment rate or the real user cost of housing (as captured by the lending rate and property tax rate) is lower. Lower property taxes also contribute to spurring housing demand and increasing prices of single-family homes; however, in our estimation property taxes do not have a statistically significant effect on the long-run prices of owner-occupied flats. On the supply side, a more limited housing stock or a rise in construction cost is associated with higher prices. Together, these factors explain about 80 percent of the variation in municipal house prices.

Table 1.

Denmark: Long-Run Determinants of Real House Prices

article image
Sources: Fund staff estimates.Note: Dependent variable is the logarithm of real house prices. Panel consists of about 96 municipalities over 1996-2015. A full set of year and municipality fixed effects are included. Specifications (2) and (4) control for the lags of the growth rates of all level variables. Statistical significance *** 1%, ** 5%, * 10%.

10. The model can be used to gauge the extent to which house prices deviate from long-run equilibrium levels. While indicators such as the price-to-income ratio are indicative, they are limited in the sense that they only consider one ‘fundamental’ variable, e.g., income, in assessing the degree of over- or undervaluation. Using our estimated model, we consider house prices in a particular municipality overvalued if the observed prices are higher than the estimated long-run level based on multiple determinants. In other words, using notations from equation (1), the degree of price deviation from long-run values is measured by:

ɛit=Pit−Pit*

The implied valuations from this exercise should be interpreted with caution, in the sense that the estimated equilibrium price levels are model-dependent, and can change considerably depending on the underlying fundamentals included in the model.

11. By our metric, current house prices in some large Danish municipalities are found to be moderately overvalued. The average prices of single-family homes in Copenhagen in 2015 are found to be about 9 percent higher than the estimated equilibrium value (Figure 1). While this is only about half of the estimated deviation during the 2006–07 peak, it suggests the beginnings of overvaluation and prices thus far continue to rise. Meanwhile, the average prices of owner-occupied flats are about 19 percent above the estimate value—which is comparable to the level during the 2006–07 peak (Figure 2). Results are mixed for other major municipalities, with little overvaluation in Odense and Esbjerg whereas Frederiksberg’s housing market is exhibiting even larger price overvaluation than Copenhagen’s.

Figure 1.
Figure 1.

Denmark: Estimated Equilibrium and Actual Prices in Selected Cities—Houses

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A001

Figure 2.
Figure 2.

Denmark: Estimated Equilibrium and Actual Prices in Selected Cities—Flats

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A001

D. Conclusion

12. Regional price divergence in Denmark’s housing market reflects a number of demand and supply factors. Rapid population and income growth in big cities pushes up demand for housing, especially as the housing stock does not expand at a commensurate pace due to land use regulations and zoning codes, which tend to be tighter in the already densely populated metropolitan areas. On top of this, historically low interest rates and tax incentives for home-ownership affect house prices in big cities disproportionately since the urban demographic structure tends to be skewed towards population of home-buying age (e.g., 25–44 age group).

13. The regional divergence in Denmark’s housing market may call for regionally-differentiated housing policy. Debt-to-income (DTI) caps, even if applied nationwide, would naturally have regionally-differentiated effects in Denmark given the high concentration of indebted households in high price growth areas. The authorities could also consider introducing other macro-prudential measures such as loan-to-value (LTV) limits with different degrees of stringency across regions to more effectively target high-risk areas such as Copenhagen without hampering the nascent recovery in other regions’ housing market (see also Chapter III). Such a strategy has recently been applied in, for example, New Zealand, where restrictions on high-LTV lending were tightened specifically for the Auckland housing market (Reserve Bank of New Zealand, 2015).9 In this vein, the Danish authorities have recently taken welcome targeted measures to strengthen mortgage lending standards, including issuing guidelines—“Seven Best Practices”—for banks in areas with rapid house price increases. Macroprudential measures aside, ending the freeze of property and land taxes—which contributes to property demand disproportionally in high price growth areas—would also be a key measure to address regional pressures. On the supply side, incentives for new residential construction could be improved by relaxing land use regulations and zoning codes in pressured urban centers to better align housing supply with population increase and housing demand.

References

  • Anundsen, A. and E. Jansen, 2013, “Self-Reinforcing Effects Between Housing Prices and Credit,” Journal of Housing Economics, Vol. 22, No. 3, pp. 192–212.

    • Search Google Scholar
    • Export Citation

  • Claussen, C. A., 2013, “An Error-Correction Model of Swedish House Prices,” International Journal of Housing Markets and Analysis, Vol. 6, No. 2, pp. 180–196.

    • Search Google Scholar
    • Export Citation

  • Dam N. A., T. S. Hvolbol, and M. H. Rasmussen, 2014, “A Multi-Speed Housing Market,” Danmarks Nationalbank Monetary Review, 3rd Quarter.

    • Search Google Scholar
    • Export Citation

  • Danmarks Nationalbank, 2015, Monetary Review, 4th Quarter.

  • Hort, K., 1998, “The Determinants of Urban House Price Fluctuations in Sweden 1968–1994,” Journal of Housing Economics, Vol. 7, pp. 93–120.

    • Search Google Scholar
    • Export Citation

  • Meen, G., 2002, “The Time-Series Behavior of House Prices: A Transatlantic Divide?” Journal of Housing Economics, Vol. 11, No. 1, pp. 1–23.

    • Search Google Scholar
    • Export Citation

  • Pedersen, E. H., J. Isaksen, 2015, “Recent Housing Market Trends,” Danmarks Nationalbank Monetary Review, 3rd Quarter.

  • Reserve Bank of New Zealand, 2015, Financial Stability Report, May.

  • Turk, R., 2015, “Housing Price and Household Debt in Sweden,” IMF Working Paper No. 15/276 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
1

Prepared by Giang Ho.

2

Weighted average of prices for single-family homes and owner-occupied flats.

3

The PTI at municipal level is calculated by dividing nominal property prices (weighted average of single-family houses and flats) by nominal average family incomes. The latter series is only available from 2000 onwards, and 2015 figure is not yet available.

4

Recent studies for Nordic countries using the same approach include Turk (2015), Claussen (2013) for Sweden, and Anundsen and Jansen (2013) for Norway. As we are most interested in the determinants of long-run equilibrium house prices in this study, a full-fledged error correction model with both long-run and short-run dynamics is not necessary.

5

The series for average personal disposable income at the municipal level is only available from 2002, so we use the unemployment rate series instead.

6

As mortgage interest rate is not available far back in time, we use instead an aggregate lending rate.

7

Calculated by dividing property tax payments by property value.

8

Asymptotically, the long-run equation may be consistently estimated by OLS even though the dynamics and endogeneity of some variables are ignored. In finite samples, however, ignoring the short-term dynamics may lead to substantial bias.

9

Specifically, starting in October 2015, the Reserve Bank of New Zealand introduced a new restriction on loans to property investors in the Auckland region with an LTV higher than 70 percent (i.e., to set a speed limit on such loans at close to zero). For all residential lending outside the Auckland region, the Bank proposed to increase the existing speed limit for loans with an LTV higher than 80 percent from 10 to 15 percent, to recognize relatively subdued housing market conditions outside Auckland.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Country Reports
Denmark: Selected Issues
Author: International Monetary Fund. European Dept.
Volume/Issue:
Volume 2016: Issue 185
Publisher:
International Monetary Fund
ISBN:
9781498319805
ISSN:
1934-7685
Pages:
54
DOI:
https://doi.org/10.5089/9781498319805.002

Real House Prices, 1992Q1-2015Q3

(index, 1992Q1 = 100)

Price-to-Income Ratio in Copenhagen and Median City

(Index, 2000 = 100)

Population Growth in Denmark’s Cities, 2015

Copenhagen: Contribution to Population Growth

(Number of people, in thousand)

Copenhagen: Residential Construction

(1,000 sqm)

Copenhagen: Population and Housing Stock

(index, 1992 = 100)

Denmark: Estimated Equilibrium and Actual Prices in Selected Cities—Houses

Denmark: Estimated Equilibrium and Actual Prices in Selected Cities—Flats