Francois Lequiller
OECD

• When one lives in a home, one consumes a « housing service ».
• Dwellings are owned by somebody (household, private company, government, special agency)
• In market economies, households can own dwellings
• Households have two « ways » of housing:
– Rent a home from an owner. The housing service is measured by the rent.
– Own their home and live in it.
• Owner Occupied Housing service is the estimate of the imputed housing service that households owning their home and living in it produce for themselves.

• GDP = sum of all productions (without double counting).
• Housing service is produced, and is thus part of GDP.
• National accounts aggregates should be « robust » to differences in
« institutional » settings.
• The fact that one can own one’s house or not is an institutional setting which should not affect the aggregate production figure.
• If OOH was not imputed:
– The comparison of the level of GDP between two countries would be affected by a difference in the rate of ownership by households of their houses.
– The change in GDP between two periods would be affected by the change in the rate of ownership of households of their own dwellings.
• History: OOH has always been imputed in national accounts

• France = higher rate of household ownership
• Germany = lower rate of household ownership
Final consumption of housing services, 2011, in % of GDP
Country
France
Actual rents Imputed Rents Total rents
2.3
7.9
10.2
Germany
3.8
5.3
9.0

18000
16000
14000
12000
10000
8000
6000
4000
2000
0
USA
GDP
GDP - OOH

• System of National Accounts excludes ownaccount production of services, except for
OOH (&6.34). Why?
– Size of the problem (housing services = 10% of
GDP)
– Use of GDP data in administrative context.
– Proximity of the « non-market » and the
« market » situation for housing.
– A good estimate can be made.

• The imputation of OOH is useful for statistics on revenue inequalities.
• Some (rare) countries tax OOH!
• How to compare the situation of a household which does not have to pay rent with one which pays a rent?
• In half of OECD countries, surveys on household revenues include an estimate of « imputed rents ».
• The method is based on assuming a rental value equivalent to the market value of similar dwellings.
• Benchmarking micro inequality indicators with macro inequality indicators needs to: (1) impute OOH, (2) social transfers in kind.

Final consumption of housing services, 2011, in % of GDP
Country Actual rents Imputed Rents Total rents
Canada
Czech Republic
Estonia
France
Germany
Hungary
Poland
Slovak Republic
Slovenia
Spain
United Kingdom
United States
2.7
2.0
1.2
2.3
3.8
0.4
0.8
0.6
0.4
1.5
3.3
2.6
8.1
6.4
5.9
7.9
5.3
6.2
3.5
4.4
6.1
7.0
9.7
8.1
10.8
8.4
7.0
10.2
9.0
6.6
4.3
4.9
6.5
8.5
13.0
10.7
Kazakhstan
7.0

• P12: Output from own final use: 100
• P2: Intermediate consumption (ordinary maintenance of dwelling typically not carried out by tenants + net insurance premiums): 10
• B1g: Gross value added: 90
• P51c: Consumption of fixed capital: 50*
• B1n: Net value added: 40
• B2g: Gross operating surplus: 90
• B2n: Net operating surplus: 40
IMPORTANT:
• OOH is not mixed income (B3).
• Operating surplus of households (B2) is exclusively Housing services (OOH and leasing services; &7.9)
* Parameters: Rent and User cost around 6% of value of capital, Consumption of Fixed capital around 2.3% of value of capital.=> CCF around 50% of housing service.

• B2g: Gross operating surplus: 90
• B2n: Net operating surplus: 40
• D29: Taxes on production: 5
• D41: Mortgage interest: 2
• B6g: Gross disposable income: 83
• B6n: Net disposable income: 33
• P31: Household final consumption expenditure
(housing service): 100
• B8g: Gross saving: -17
• Net saving: -67

• SNA has recommended inclusion of estimate for OOH in all its versions, including the first one.
• All OECD member countries include an estimate of OOH.
• OOH also important for PPP calculations (price approach/volume approach).
• In the EU, because of the use of GNI in budget calculation, strong legal approach.
• EU Commission Regulation 1722/2005
• Note: « rent » exclude water supply, refuse and sewage collection, electricity and gas, heating and hot water.

• Two major methods:
– Method 1: « Rental equivalence » or « stratification method ».
• Value OOH at the estimated rental that a tenant would pay for the same accomodation.
• EU Regulation’s preferred method. Described in very detailed manner in the
Regulation (including special cases).
• However this method assumes there is a wide and well-organised actual rental market.
– Method 2: « User Cost »:
• OOH using the « non-market » approach, based on costs.
• Second best applicable when: (1) the privately rented dwellings constitute less than
10% of the total dwelling stock by number and there is a large disparity between private and other paid rents (by a factor of three).
• Widely implemented in the EU for accession of East-European countries.
– Other minor methods: self-assessment, fiscal assessment.

• Method applied by BEA in USA: see presentation by
Michael Hayes
• SNA &6.117: « the output of rental services produced by owner-occupiers is valued at the estimated rental that a tenant would pay for the same accomodation, taking into account fators such as location, neighbourhood amenities, etc. as well as the size and
quality of the dwelling itself. »
• Basically: extrapolation of actual rents to owner- occupied situation
• Adapted when there is a wide and well organised rental market and extended information on characteristics of dwellings x rents.

• Obtain an estimate of rent by stratum of the owner occupied dwelling stock.
• Needs deep stratification to be reliable.
• Need good information on housing stock: rented/owned with detailed characteristics.
• Use large data sources:
– Population and/or Housing censuses
– Rental and Housing surveys
– Household Budget Survey
– Administrative sources
• Can use econometric techniques (hedonic regression) to test explanation potentiality of characteristics for rental variance
• Generally base year calculation extrapolated/interpolated

• Stratification: factors affecting the rental level
– Size of dwelling
– Amenities of dwelling
– Type of dwelling
– Environment characteristics (urban/rural, transport facilities)
– Socio-economic factors (type of owner)
– Use of capital price of dwelling as implicit principal stratification factor is possible.
– Delicate issue of using rents of government owned dwellings (implicit subsidies) to estimate privately owned rents.

• Method applied by CSO in Hungary: see presentation by
Zsuzsanna Boros
• Developped by Eurostat for EU accession countries.
• Adapted when no sufficiently developped actual rental market (two strict conditions in Reg 1722).
• Inspired by estimation of non-market output.
• Output = sum of costs = intermediate consumption + net taxes on production + gross operating surplus (GOS)
– Easy to measure: Intermediate consumption + net taxes:
– More difficult: GOS = consumption of fixed capital + net operating surplus (return to capital).

• Based on the following (simplified) formula:
GOS = i(r + d)S
• Where:
– i = inflation rate
– r = rate of return of capital (average effective mortgage rate; default Reg 1722 = 2.5%)
– d = rate of depreciation (geometric depreciation with no mortality function)
– S = Value of stock of dwellings at current price. In principle includes construction + land.

• What is the size of the rental market in Kazakhstan?
• Does Kazakhstan use the User cost method?
• What sort of information is available on the stock of owner occupied dwellings ?
• Should we separate land and construction?
• Should we include a measure of expected capital gains?
• Dwelling services produced by government: are there far from market price? If yes, should we introduce a market price?
• Dachas: can they be taken into account?