897
Views
503
Downloads |
An overview of the taxation of residential property: is it a good idea?
Marta Rodriguez-Vives* 
Miguel Angel Gavilan-Rubio*
Miguel Angel Gavilan-Rubio
Affiliation: Researcher at Analysis and Research Division, Bank of Slovenia, Ljubljana, Slovenia
 0000-0002-0001-9663
Preliminary communication | Year: 2021 | Pages: 283 - 303 | Volume: 45 | Issue: 2 Received: May 29, 2020 | Accepted: February 8, 2021 | Published online: June 6, 2021
|
FULL ARTICLE
FIGURES & DATA
REFERENCES
CROSSMARK POLICY
METRICS
LICENCING
PDF
Source: Own representation.
Note: For Greece, Japan and the OECD average the data refer to 2017. Recurrent taxes on net wealth include current taxes on capital; capital transactions include taxes on stamp duties; and “other taxes” include other non-recurrent taxes on property (capital levies) and other recurrent taxes on property (current taxes on capital). Own representation.Source: Authors based on OECD Revenue Statistics database and Eurostat.
Note: For Greece, Japan and OECD average the data refer to 2017. “Other taxes on property” includes taxes on net wealth, heritage and gift taxes, taxes on financial and capital transactions, other non-recurrent taxes on property and other recurrent taxes on property. Source: Authors based on OECD Revenue Statistics database and Eurostat.
Source: Eurostat, IMF database (IMF.org/housing).
* The last available years for Austria, Italy and Luxemburg are 2016, 2015 and 2013 respectively. Source: Barrios et al. (2019), Eurostat and ECB database.
|
|
Log Implicit Recurrent Tax Rate on
Immovable Property
|
|
Model 1 (RE)
|
Model 2 (RE)
|
Model 3 (RE)
|
Model 4 (RE)
|
Model 5 (RE)
|
Model 6 (RE)
|
|
Constant
|
-1.478***
|
-4.043***
|
5.356
|
1.078
|
0.0953
|
2.145
|
|
(0.28)
|
(0.58)
|
(3.51)
|
(3.23)
|
(3.04)
|
(3.14)
|
|
Growth of stock dwellings
|
-0.1250***
|
-0.0959***
|
-0.1310***
|
-0.1200***
|
-0.1320***
|
-0.0869**
|
|
(0.04)
|
(0.03)
|
(0.05)
|
(0.04)
|
(0.04)
|
(0.04)
|
|
Growth of price of m2
|
-0.0076***
|
-0.0011
|
-0.0081***
|
-0.0013
|
-0.0017
|
-0.0048
|
|
(0.00)
|
(0.00)
|
(0.00)
|
(0.00)
|
(0.00)
|
(0.00)
|
|
Log HH debt ratio
|
|
0.5590***
|
|
0.7150***
|
0.7480***
|
0.5560***
|
|
(0.11)
|
|
(0.18)
|
(0.18)
|
(0.17)
|
|
Log Gini index
|
|
|
-1.9890*
|
-1.7090*
|
-1.4280*
|
-1.7260**
|
|
|
(1.04)
|
(0.96)
|
(0.86)
|
(0.87)
|
|
Transfer tax rate
|
|
|
|
|
-0.0886*
|
-0.0893*
|
|
|
|
|
(0.05)
|
(0.05)
|
|
Interests income tax rate
|
|
|
|
|
0.0077***
|
0.0077***
|
|
|
|
|
(0.00)
|
(0.00)
|
|
Long term interest rate
|
|
|
|
|
|
-0.0471***
|
|
|
|
|
|
(0.01)
|
|
No of observations
|
338
|
337
|
256
|
255
|
255
|
255
|
|
No of countries
|
19
|
19
|
19
|
19
|
19
|
19
|
|
R2 overall
|
0.0258
|
0.2610
|
0.0168
|
0.2190
|
0.2140
|
0.1780
|
|
R2 within
|
0.0332
|
0.0795
|
0.0736
|
0.1200
|
0.1720
|
0.2040
|
|
R2 between
|
0.0367
|
0.2960
|
0.0277
|
0.2370
|
0.2130
|
0.1770
|
|
Hausman testa
|
0.8142
|
0.1058
|
0.8204
|
0.5624
|
0.2053
|
0.2578
|
Note: Standard errors in parentheses. Standard errors are obtained using robust estimations. * p<0.10, ** p<0.05, and *** p<0.01. a p-values reported.
|
|
Estimation of implicit recurrent tax
rate functions using stochastic frontier analysis
|
|
Model 7 (TI)
|
Model 8 (TI)
|
Model 9 (TI)
|
Model 10 (TI)
|
Model 11 (TI)
|
Model 12 (TI)
|
|
Constant
|
1.047***
|
-1.737***
|
7.9910***
|
3.330
|
2.427
|
4.475*
|
|
(0.31)
|
(0.65)
|
(2.41)
|
(2.43)
|
(2.34)
|
(2.41)
|
|
Growth of stock dwellings
|
-0.1250**
|
-0.0955**
|
-0.1310**
|
-0.1200**
|
-0.1310***
|
-0.0854*
|
|
(0.05)
|
(0.05)
|
(0.05)
|
(0.05)
|
(0.05)
|
(0.05)
|
|
Growth of price of m2
|
-0.0076**
|
-0.0012
|
-0.0080**
|
-0.0012
|
-0.0017
|
-0.0049
|
|
(0.00)
|
(0.00)
|
(0.00)
|
(0.00)
|
(0.00)
|
(0.00)
|
|
Log HH debt ratio
|
|
0.5530***
|
|
0.7200***
|
0.7410***
|
0.5450***
|
|
(0.11)
|
|
(0.15)
|
(0.15)
|
(0.17)
|
|
Log Gini index
|
|
|
-1.9590***
|
-1.6790***
|
-1.4390**
|
-1.7390***
|
|
|
(0.67)
|
(0.63)
|
(0.62)
|
(0.62)
|
|
Transfer tax rate
|
|
|
|
|
-0.0904***
|
-0.0904***
|
|
|
|
|
(0.03)
|
(0.03)
|
|
Interests income tax rate
|
|
|
|
|
0.0076**
|
0.0076**
|
|
|
|
|
(0.00)
|
(0.00)
|
|
Long term interest rate
|
|
|
|
|
|
-0.0476***
|
|
|
|
|
|
(0.02)
|
|
Log sigma2
|
|
|
|
|
|
|
|
Constant
|
0.4420
|
0.1680
|
0.3760
|
0.0957
|
0.1400
|
0.1950
|
|
(0.34)
|
(0.32)
|
(0.33)
|
(0.32)
|
(0.35)
|
(0.36)
|
|
Log t gamma
|
|
|
|
|
|
|
|
Constant
|
1.8300***
|
1.6020***
|
2.0310***
|
1.8320***
|
1.9530***
|
2.0610***
|
|
(0.40)
|
(0.39)
|
(0.38)
|
(0.38)
|
(0.41)
|
(0.42)
|
|
mu
|
|
|
|
|
|
|
|
Constant
|
2.4760***
|
2.2510***
|
2.7080***
|
2.3600***
|
2.2210***
|
2.1780***
|
|
(0.44)
|
(0.42)
|
(0.52)
|
(0.48)
|
(0.41)
|
(0.40)
|
|
No of observations
|
338
|
337
|
256
|
255
|
255
|
255
|
|
No of countries
|
19
|
19
|
19
|
19
|
19
|
19
|
|
AIC
|
537.7
|
506.6
|
369.4
|
339
|
328.8
|
322.4
|
|
BIC
|
560.6
|
533.3
|
394.2
|
367.3
|
364.2
|
361.3
|
|
Log likelihood
|
-262.8
|
-246.3
|
-177.7
|
-161.5
|
-154.4
|
-150.2
|
Note: Standard errors in parentheses. Standard errors are obtained using robust observed information matrix. * p<0.10, ** p<0.05, and *** p<0.01.
Source: Authors’ construction based on Model 12.
|
Policy tool
|
Definition
|
Primary
objective
|
Regulation
|
|
Taxation
|
|
|
ESA
2010
|
|
1. Buying property
|
Refers
to the one-off transaction tax for the sale of real estate either for the
first time or subsequent times. It typically takes the form of stamp duty,
VAT (first transaction), transaction tax, or capital acquisition tax.
Purchase costs usually are not deductible in the tax declaration. This tax is
widely applied in advanced countries as it is administratively appealing
since transactions can often be reasonably easily observed. These taxes could
also contribute to reducing asset price volatility. However, taxing real
estate transactions may also lead to undesirable outcomes, such as adversely
impacting labour mobility, reducing house affordability, and ownership
ratios.
|
|
Value added tax (VAT)
|
VAT on the value of the property, sold
for the first time
|
Generate government revenue
Reduce housing demand (especially
speculative house purchases)
|
D.211
|
|
Stamp taxes / transfer tax
|
One-off - Transaction taxes
All investors (domestic / foreign);
borrowers / equity investors
|
D.214 B, C
Stamp taxes
Taxes on financial and capital
transactions
|
|
2. Holding
property
|
This
involves recurrent taxes on
property, usually levied at sub-national/local level. It has
particular appeal as a source of local-government finance since property
values will reflect the benefits of local public spending. This type of tax
is somehow underutilised. It is attractive in the sense that the base is
relatively immobile and hard to hide, the tax comes at the top of the
hierarchy of long-run growth-friendliness mentioned earlier, and it can be
made progressive through a basic allowance or by varying the rate with the
value of the property. However, it requires some intense administrative
build-up and maintenance of infrastructure. Income from owner-occupied
property, whether in the form of imputed rental income or capital gains, is
typically untaxed or taxed at low rates. It also includes recurrent taxes on
net wealth (assets less liabilities), but they have been declining and only
four countries are currently using them.*
|
|
Recurrent taxes on immovable property
|
Taxes paid annually and linked to the
value of the property.
Households: Current taxes on capital,
other (D.59)
Other: Taxes on land buildings or other
structures (D.29)
|
Generate government revenue
Finance regional/local level services
|
D.59 Current taxes on capital
|
|
Recurrent taxes on net wealth
|
Assets minus liabilities
|
Inequality
|
D.59 Current taxes on capital
|
|
Other recurrent taxes on property
|
Other recurrent taxes on capital
|
D.59 Current taxes on capital
|
|
3. Renting
property
|
Residential rental property is
typically taxed on a comprehensive basis, in a similar way to interest
income. No tax relief is provided upfront, and returns are taxed as they are
earned-marked.
|
|
Rental income
|
Property Income and Personal Income Tax
(PIT)
|
Generate government revenue
|
D.45 and D.5
|
|
4. Selling
property
|
When people sell residential
property for more than the purchase price, they generally realize a capital
gain that is subject to taxation as income. The capital gain is defined as
the difference between selling and purchase prices, also known as the basis.
Capital gains taxes on the sale of residential properties are treated
differently depending on the holding period. Long-term capital gains relate
to dwellings held for more than a year, and are usually taxed with lower
rates and have a tax-free threshold for owner-occupied housing to protect
household net wealth. Short-term capital gains are generated when the sale of
the dwelling is made within a year, and are usually taxed with higher rates
to avoid housing bubbles.
|
|
Other non-recurrent taxes on property
|
Capital gains taxes (occasional or
exceptional levies on capital or wealth)
|
Inequality
Reduce housing demand and supply
|
D.91 Capital levies
|
|
5. Bequest
|
This includes taxes on inheritances and gifts. The yield is
usually relatively limited: rates are low, and exemptions and special
arrangements create multiple avoidance opportunities. The yield in the
countries with the highest returns (about 0.7% of GDP in Belgium and France)
suggests its potential as a revenue raising
tool. The main policy objective of inheritance taxes is to limit the
intergenerational transmission of inequality, but they also produce
distortions, which are difficult to assess. Theoretical results on optimal
bequest taxation differ widely, but Piketty and Saez (2013) find a positive
and relatively high rate.
|
|
Inheritance and gift taxes
|
Death duties or taxes on gifts inter-vivos,
to be levied on the capital of the beneficiaries
|
Inequality
Generate
government revenue
|
D.91 Taxes on capital transfers
|
* Spain, Iceland, Norway, and Switzerland.
|
Variable
|
|
Mean
|
Std. Dev.
|
Min
|
Max
|
Observations
|
|
Log implicit recurrent tax rate on immovable property
|
overall
|
-1.49
|
1.21
|
-3.97
|
0.78
|
N=364
|
|
between
|
|
1.13
|
-3.42
|
0.71
|
n=19
|
|
within
|
|
0.47
|
-3.75
|
-0.11
|
T=19.2
|
|
Growth of stock dwellings
|
overall
|
1.96
|
0.99
|
-1.35
|
11.89
|
N=430
|
|
between
|
|
0.62
|
0.27
|
2.75
|
n=20
|
|
within
|
|
0.80
|
-2.33
|
11.48
|
T=21.5
|
|
Growth of price of m2
|
overall
|
5.19
|
9.30
|
-37.39
|
54.96
|
N=407
|
|
between
|
|
2.14
|
0.92
|
9.16
|
n=20
|
|
within
|
|
9.07
|
-41.35
|
51.00
|
T=20.4
|
|
Log HH debt ratio
|
overall
|
4.42
|
0.82
|
1.06
|
5.83
|
N=454
|
|
between
|
|
0.69
|
2.90
|
5.59
|
n=20
|
|
within
|
|
0.46
|
2.50
|
5.54
|
T=22.7
|
|
Log Gini index
|
overall
|
3.42
|
0.12
|
3.14
|
3.66
|
N=315
|
|
between
|
|
0.12
|
3.21
|
3.60
|
n=20
|
|
within
|
|
0.04
|
3.27
|
3.57
|
T=15.8
|
|
Transfer tax rate
|
overall
|
3.68
|
3.00
|
0.00
|
15.00
|
N=448
|
|
between
|
|
2.82
|
0.00
|
10.76
|
n=20
|
|
within
|
|
1.23
|
0.72
|
16.11
|
T=22.4
|
|
Interests tax rate
|
overall
|
24.91
|
15.48
|
0.00
|
62.70
|
N=460
|
|
between
|
|
12.36
|
2.61
|
54.91
|
n=20
|
|
within
|
|
9.71
|
7.94
|
68.35
|
T=23
|
|
Long term interest rate
|
overall
|
4.16
|
2.16
|
0.09
|
14.00
|
N=424
|
|
between
|
|
0.58
|
3.37
|
5.46
|
n=20
|
|
within
|
|
2.09
|
-0.28
|
13.41
|
T=21.2
|
Note: "N" denotes number of observations, "n" number of countries and "T" average period observed per country.
|
Country
|
Average
Log-IRTRIP
|
Model 7
|
Model 8
|
Model 9
|
Model 10
|
Model 11
|
Model 12
|
|
Austria
|
-3.4233
|
0.0237
|
0.0301
|
0.0402
|
0.0550
|
0.0604
|
0.0592
|
|
Belgium
|
-1.7104
|
0.0891
|
0.1144
|
0.1717
|
0.2349
|
0.4260
|
0.4188
|
|
Czech Republic
|
-2.4327
|
0.0856
|
0.1272
|
0.0826
|
0.1381
|
0.1579
|
0.1443
|
|
Denmark
|
-0.3223
|
0.3748
|
0.2757
|
0.2940
|
0.2245
|
0.1795
|
0.2021
|
|
Finland
|
-1.5155
|
0.1036
|
0.1254
|
0.1463
|
0.1820
|
0.2047
|
0.2007
|
|
France
|
0.2522
|
0.5455
|
0.6615
|
0.5756
|
0.7111
|
0.8747
|
0.8814
|
|
Germany
|
-2.0309
|
0.0583
|
0.0695
|
0.0699
|
0.0923
|
0.0972
|
0.0983
|
|
Ireland
|
-2.2790
|
0.7220
|
0.8587
|
0.7228
|
0.8625
|
0.8586
|
0.8674
|
|
Italy
|
-1.0560
|
0.2788
|
0.3735
|
0.4415
|
0.5668
|
0.6894
|
0.6948
|
|
Latvia
|
-2.5971
|
0.3355
|
0.3913
|
0.8366
|
0.8588
|
0.8642
|
0.8650
|
|
Lithuania
|
-3.3523
|
0.1092
|
0.1470
|
0.1130
|
0.1722
|
0.1670
|
0.1651
|
|
Luxembourg
|
-2.6653
|
0.0541
|
0.0557
|
0.1003
|
0.1074
|
0.1466
|
0.1479
|
|
Netherlands
|
-0.9303
|
0.1735
|
0.1319
|
0.2258
|
0.1857
|
0.2338
|
0.2576
|
|
Portugal
|
-1.4678
|
0.0999
|
0.1085
|
0.2491
|
0.2589
|
0.2514
|
0.2909
|
|
Slovakia
|
-1.9490
|
0.1375
|
0.2019
|
0.1566
|
0.2600
|
0.2405
|
0.2327
|
|
Slovenia
|
-2.0983
|
0.1254
|
0.1848
|
0.1183
|
0.1972
|
0.2022
|
0.1889
|
|
Spain
|
-1.4739
|
0.1103
|
0.1231
|
0.1700
|
0.1939
|
0.2460
|
0.2482
|
|
Sweden
|
-0.3691
|
0.2843
|
0.2787
|
0.2614
|
0.2663
|
0.2530
|
0.2655
|
|
United Kingdom
|
0.7084
|
0.8440
|
0.8444
|
0.8711
|
0.8775
|
0.7489
|
0.8125
|
|
Average
|
-1.4934
|
0.2738
|
0.3013
|
0.3288
|
0.3689
|
0.3919
|
0.3989
|
Figure 1Housing dichotomy and the use of dwellings after purchase DISPLAY Figure
Figure 2Components of property tax, percentage of GDP (2018, in %) DISPLAY Figure
Figure 3Property tax revenue, as percentage of total government revenue (2018 vs. 2007, in %) DISPLAY Figure
Figure 4Residential property developments (in %) DISPLAY Figure
Figure 5Implicit property tax rate and supply of residential property (in %) DISPLAY Figure
Table 1Estimation of implicit recurrent tax rate functions using panel data analysis – random effects (selected EU countries, 1995-2017) DISPLAY Table
Table 2Estimation of implicit recurrent tax rate functions using stochastic frontier analysis (selected EU countries, 1995-2017) DISPLAY Table
Figure 6Estimated efficiency of recurrent tax on immovable property DISPLAY Figure
Table A1Taxing tools in the housing market DISPLAY Table
Table A2Summary of panel data variables (selected EU countries, 1995-2017) DISPLAY Table
Table A3Estimated tax efficiency by country (selected EU countries, 1995-2017) DISPLAY Table
* The opinions expressed herein are those of the authors and do not necessarily reflect the views of the ECB, Bank of Slovenia or the Eurosystem. We want to thank three anonymous referees for their suggestions for improving this work. Likewise, the authors want to express their gratitude to the participants and the discussants of the 4 th Public Sector Economics Conference (Zagreb, 24 October 2019) where a preliminary version of this article was discussed. Any errors are the sole responsibility of the authors.
1 For more details, OECD (  2019a) and its “Public policies towards affordable housing” section.
2 See Acosta-Ormaechea and Yoo ( 2012). According to Roeger and Veld ( 2010), property taxes are considered the most growth-friendly type of tax. The authors set a model with a tax on housing property, in which increments in this tax negatively affect housing investment. However, it does not directly distort the provision of the inputs to production and household consumption decisions. Moreover, by making investments in productive capital relatively more attractive than investment in housing, it leads to a higher stock of productive capital and more production.
3 Although taxing property refers to both residential and commercial property, this article focuses on residential property and the household sector.
4 The taxpayers can be households, individuals, or corporations. Also, the property can either refer to residential or commercial real estate. Although a more granular assessment is warranted, this is outside the scope of this paper. Instead, we refer to taxing property in a generic way, focusing on the household sector and residential real estate by default.
5 According to the third wave of the ECB’s Household Finance and Consumption Survey (HFCS), which refers to 2017 data – for more details see Household Finance and Consumption Network (2020).
6 Housing services provided by homeowner-occupiers are imputed as being equal to the rents they would have paid for comparable housing. It implies that the production of services (rental or imputed) from owner-occupied (imputed) and tenant-occupied dwellings (rental) are part of GDP. The arbitrariness lies in the imputation method. According to Lequiller and Blades ( 2014), a long-term upward trend in homeownership would automatically produce a downward trend in the total value of actual rents (and thus in GDP, all things being equal) and make it difficult to compare the output of different countries because homeownership rates vary across countries.
7 With an average increase of 4.2% in 2019Q4 in comparison to 2018Q4 – according to Eurostat’s house price data.
8 There are some patterns observed when considering the investment rate and prices together. We can observe some countries such as Spain and Greece, where both investment rate and price-to-income ratio have fallen, indicating a lack of demand in the market. By contrast, Germany, Japan and Austria have reached investment rates similar to pre-crisis levels and increases in prices with respect to their disposable incomes.
9 The 20 selected countries are Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Ireland, Italy, Latvia, Lithuania, Luxembourg, Netherlands, Portugal, Slovakia, Slovenia, Spain, Sweden and the United Kingdom.
10 We initially considered other variables, such as GDP per head (constant PPS) and inflation. These two variables never were significant. However, they did cause high group heterogeneity, serial correlation and collinearity, affecting the overall estimation and performance of the models. We finally decided not to include them in the models.
11 We performed the Hausman Test for each specification, resulting in Random Effects models in all of them. See the P-values of the Hausman Test at the bottom of the table.
12 For example, see Fatica and Prammer ( 2018) and European Commission ( 2015) for more details.
13 There is currently an excellent tax offshoring debate in the UK, as unknown ownership has severe consequences in the collection of Council Taxes, which could lead to potential revenue losses.
14 We estimated Time-invariant (TI) and Time-varying decay (TVD) frontiers and, in all feasible cases (some models did not converge when TVD specifications were used), the coefficient related to the Time-varying decay was insignificant, suggesting the use of Time-invariant models.
|
|
June, 2021
II/2021 |
