Dwellings repair & maintenance

Apart from some maintenance, investment in the repair and maintenance of dwellings is not included in National Accounts.

Dwellings repair and maintenance

The FIDIC-EFCA Consulting Engineering Industry Survey uses investment in physical assets given by Gross Fixed Capital Formation (GFCF) taken from National Accounts to estimate the demand for consulting engineering services.

Investments in Construction are classified by the latest (2008) UN System of National Accounts (SNA08) under two categories: Dwellings (called housing, residential construction, residential buildings, etc.) and Other Buildings (called non-housing or non-residential) and Structures (called infrastructure or works of civil engineering).

SNA08 GFCF covers produced assets that are used repeatedly or continuously in production over several accounting periods (i.e., for more than one year). It therefore assumes that the goods and services involved are not completely used up in the accounting period so the Construction GFCF is for New Build.

Construction New Build GFCF given in SNA08 for Dwellings and for Other Buildings and Structures includes not only fees for professional services but also items such as taxes, preparation and development costs, administration costs (including marketing and insurance costs) passed on to the client, finance costs assumed by the client (for example, insurance costs); and other costs (including acquisition costs).

Construction New Build GFCF therefore reflects the project-based nature of investments in physical fixed assets such as a building or a power plant since it includes most of the items that a client needs to consider for a specific project when making an investment decision on the basis of the total project cost.

It is for this reason that GFCF provides a reliable basis for estimating changes in the demand for consulting engineering services in the main areas of investment covered by the industry’s activity, namely construction, equipment (including defence systems), and intellectual property (mainly research and development).

Construction New Build GFCF includes major improvements such as reconstruction, major renovation, enlargements, and extensions which improve an asset’s performance, increase its capacity or prolong the expected working life.

The European System of National Accounts (ESA2010) that extends SNA08 also includes improvement (sometimes called renovation) in Construction New Build (e.g., the 2011 European Commission Panorama of the European non-residential sector report states that “renovation refers to all activities aimed at the improvement and maintenance of an existing building or parts thereof”).

The issue

A major problem that the FIDIC-EFCA Survey must address is Repair and Maintenance (R&M). Construction R&M is not included in Construction New Build GFCF but creates a demand for consulting engineering services.

The issue for the survey is to estimate the magnitude of the increase in New Build Construction GFCF that would result if all R&M for Construction is capitalised in National Accounts.

SNA08 points out that: “In practice it is not easy to draw the line between ordinary repairs and major improvements, although the SNA provides certain recommendations for this purpose”.

Consequently, while construction industry outputs for R&M and for Improvements are sometimes reported, this practice is not carried through to National Accounts. National Accounts include Improvement in New Build and do not separate truely new build construction (i.e., new construction) from improvements. Output data therefore has to be used to identify GFCF for new build construction, improvement and R&M.

Thus, Gruneberg pointed out in 2008 that R&M was not part of the Structures New Build GFCF reported in UK National Accounts but was part of total construction output reported in the UK Construction Statistics Annual. The output of Structures R&M which amounted to 46% of UK construction output for the period 1997 to 2005 was used to deduce Structures GFCF for R&M (or Structures GRM).

Approximately the same figure (43%) was estimated by Lopes ] as the output of R&M as a percentage of all construction output in the UK in 2000. A subsequent publication gave an updated percentage of 46% for 2003.

Following Lopes, the FIDIC-EFCA Survey assumes that there is GRM for all categories of Construction GFCF (namely both Dwellings and Other Buildings and Structures) rather than only for Structures as was assumed by Gruneberg.

The challenge is to obtain accurate and up-to-date estimates of the GRM for Dwellings, for Other Buildings and for Structures as these investments create a demand for consulting engineering services. This note discusses Dwellings R&M.


The construction industry’s output of residential buildings can be divided into the outputs for Output New Build (ONB), for Output Improvement (OI) and for Output R&M (ORM) where ORMI, the output for Repair and Maintenance and Improvement, is given by:

  • ORMI = ORM + OI

The outputs for New Build and Improvement are taken up by National Accounts and become part of the GFCF for Dwellings Construction New Build (GNB).

Assuming that Dwellings Construction GFCF is in the same proportion as residential building output, Dwelling Construction R&M GFCF (GRM) is given by:

  • GRM / (GRM + GNB - GSP) = ORM / (ORM + OI + ONB)

where GRM is the investment based on the GFCF of the dwelling without site preparation and improvement, i.e.,

  • GRM = (1 - y).GNB

where y is the factor (0.04, see by which GNB is reduced if land improvment is not included in GNB.

Manipulation gives:

  • GRM = (GNB - GSP) / [1 - ORMI * (1 - x)] = GNB (1 - y) / [1 - ORMI * (1 - x)]

where ORMI is expressed as a percentage of the total residential building output and x = OI / ORMI

A convenient representation of the model is to plot OI (the output for Improvement) as a function of ORMI (= ORM + OI), the output for R&M and Improvement) for various values of GRM, the investment in R&M.

R&M investment, GRM

There are a few cases where the R&MI output (ORMI) is split into R&M output (ORM) and Improvement output (OI) for residential building. This is because the improvement of buildings, both residential and non-residential, has assumed greater importance with the need to undertake renovation in accordance with energy saving and carbon emission reduction measures.

Two approaches are reported:

  • use national statistics data for ORMI and OI;
  • assume that OI corresponds to so-called "deep" investment in energy saving to give an upper limit for GRM (i.e., assumes that Improvement is ony for energy saving).

National statistics

Reported values of housing RMI output (ORMI) that definitely aim to include all R&M and Improvement in ORMI and to clearly distinguish ORMI from New Build output (ONB) are given in the table below, together with OI values and the GRM calculated using the model.

Estimates of GRM using national statistics data are relatively rare. Of the various estimates, those for Denmark and the UK are the most recent and the most reliable. The average GRM for the five countries in the table is 19% of GNB.

Country Year ORMI, % OI, % GRM (%GNB)
EC16 2009 59.3 39.0 20.5
DK 2015 48.7 27.6 21.6
DK 2016 49.2 28.9 20.5
DK 2017 42.8 27.5 13.3
UK 2012 52.7 37.5 13.2
UK 2013 50.6 35.8 12.8
UK 2014 46.5 36.0 7.3
FI 2000 46.0 23.5 23.9
NL 2012 52.1 27.3 27.7
DE 2010 77.8 69.2 5.0

Table 1 - GRM, the GFCF for Repair and Maintenance that is not included in the New Build Dwellings GFCF (GNB). This GRM is adjusted for site preparation civil works (estimated to average 4% of Construction Dwellings New Build GFCF in the EU (see ). EC18 refers to Euroconstruct data. ORMI and OI are given as percentages of the total residential building output. See Annex 1 for details for data entries

Fig. 1 - OI versus ORMI from Table 1, both plotted as a percentage of the total residential building output. The line is for the average of the six data sets and corresponds to GRM = 19% of GNB.

EU27 2015

For Euroconstruct's EC18 in 2015, the total building energy renovation market for construction has been estimated as bEUR 82.1 with 65% for residential buildings PDF. This estimate used the ACEEE 2008 proposal PDF that the renovation market corresponds to 15% of the residential building output. The total EC18 residential building renovation output was bEUR 646 (see Annex 2). The energy renovation output for the EC18 was therefore (100 x 82.1 x 0.65/387.6) or 8.26% of EC18 residential building renovation output. ORMI was 60% of the EC18 residential building output.

GRM is (79% of GNB compared with about 20% for the direct calculation of the investment in the R&M and Improvement of dwellings.

A more direct approach using the ACEEE 2008 proposal PDF is to assume instead that the energy efficient component of the renovation market corresponds to 15% of the residential building renovation output, OMRI. This 15% corresponds to “moderate” and “extensive” energy renovations of 10% and 5%, respectively PDF.

In 2015, the residential building renovation output was 60% of residential building output (see Annex 2). OI/ORMI was therefore 15/60 which gives a GRM of 75% of GNB assuming that renovation is only for energy renovation. The International Energy Agency (IEA) has challenged the use of this ACEEE factor of 15% (see PDF) and has in fact reduced the factor it to 5% to make estimates comparable to those derived from other IEA models.

In our case, given that both estimates of the GRM upper limit are several times the GRM of about 20% calculated using national statistics suggests that the renovation output on average throughout the EC18 countries includes a significant amount of output that is not for energy efficiency. Residential building energy renovation output therefore often cannot be used to estimate GRM. Indeed a 1996 UK survey indicated that expenditure on energy efficiency measures accounted for only about 23% of the total expenditure on housing renovation and repair, confirming that the investment in energy efficiency underestimates considerably the investment in improvement .

France 2011

In France in 2011, new housing construction output was bEUR 32.35 and "rehabilitation" was bEUR 44.8 (58.07% of total housing output) PDF.

In France in 2011, of the bEUR 38.76 output for housing "maintenance and improvement", 34.8% was for energy renovation PDF. Scaling "maintenance and improvement" to "rehabilitation", the output for energy renovation becomes (34.8 x 44.8/38.76) or bEUR 40.22 or 52.16% of the total housing output of bEUR 77.15. GRM is therefore only 2% of GNB suggesting that France's national statistics classify only a very small part of ORMI as Repair and Maintenance.

In addition, as for the UK, energy renovation only accounted for a relatively small percentage (35% in 2011 ) of the total expenditure on R&M and Improvement.


As there appear to be relative few EU countries for which national accounts probably aim to include all dwelling R&M and Improvement in the GFCF for New Build Construction it is reasonable to assume that the EU average for the GFCF for R&M (GRM) is 20% of the GFCF for New Build Construction (GNB).

It should be noted that GNB generally includes investment that is not included in construction output. In the case of the UK PDF, this investment is the dwelling improvement that is not carried out under contract (it is carried out by households or as “do-it-yourself” improvements). For the UK in 2013, ORMI is increased by 15.4% of Dwellings GNB (mSTR 8683) so OMRI is increased by 18% and GRM by 60% (from 17% of GNB to 27%). This is a substantial increase which cannot unfortunately be established for other countries.

The conclusion is that an appropriate average GRM for the European Union is 20% of GNB.

Annex 1: Table 1 data entries


Euroconstruct covering most of western Europe and several eastern European countries notes that R&M accounts for routine repair of broken, worn out elements or preventive replacement, and repairs forced by exceptional causes such as storms, floods, fires countries. In Euroconstruct statistics "Renovation" is for "repair and maintenance, refurbishment and reconstruction", i.e., it is equivalent to ORMI PDF . OI was an average of 39% of residential building renovation output for the 18 Euroconstruct countries in 2009. The residential building renovation output ONRI was about 57% of the residential building output.

Denmark 2015-7

“Major Repair” output for housing and “Maintenance” output for housing as given by Statistics Denmark Accessed Jan 2019 following a preliminary survey PDF. It is assumed that "Major Repair" output is included in GNB and corresponds to OI while “Maintenance” output is not included and corresponds to ORM.

UK 2012-4

In 2012, for Household Final Consumption Expenditure , household spending on dwellings’ alterations and improvements (namely ORMI) was mSTR 24155 with a total dwellings output of mSTR 45819.

From the same source, household expenditure for alterations and improvements (OI) represented 17.3 STR/week of the 24.3 STR/week spent on alterations and improvements and the maintenance and repair of dwellings (ORMI). OI/ORMI was therefore 0.71 assuming that construction-related spending is given by the average for spending on different types of activities (e,g, spending includes spending on equipment). The same source gives equivalent data for 2013 and 2014.

Ireland 2012

In 2012, 16.5% of Housing R&M output was reclassified and capitalised as RMI, so OI is greater than 16.5%. As there was probably some improvement included in the RMI output before the reclassification that was capitalised in Construction GFCF, the Housing OI of 16.5% is probably an underestimate.

In 2012, total housing construction output was bEUR 3.552 having been bEUR 2.967 before changes to accoints, so bEUR (3.552-2.967) of minor improvements were capitalised PDF. OI was at least bEUR 0.585. Housing construction ORMI was bEUR 2.16 PDF. The GRM is 72% of GNB and is clearly a very high upper limit that can be ignored.

Finland 2000

Finland reorganised its building renovation statistics in 2013, with extensions and renovation (improvements) capitalised in GFCF and R&M called refurbishment. Before this, for example for 2000, the separation of outputs for residential improvements and R&M was clear since "Construction of buildings output is calculated as the value of new construction and renovation combined. For investment purposes, this is calculated by means of building type (non-residential buildings, other buildings) and by type of construction (new work, refurbishment and yearly repairs). It is hard in practice to draw a line between refurbishment and yearly repairs. In print publications, refurbishment or improvement is often commonly defined as repairs which make a building more suited to its purpose and result in better quality or value in the building than before. Yearly repairs, or maintenance, are defined as an ordinary, regular procedure intended to keep the building at a quality level not higher than the original. For calculation purposes, refurbishment is classed as investment and yearly repair is classed as intermediate consumption." PDF In 2000, turnover for dwellings new construction was bEUR 3.955, with refurbishment broken up into bEUR 1.649 for yearly repairs and bERU 1.726 for refurbishment.

However, by 2019, the terminology had changed (refurbishment is now R&M). In 2019, for Statistics Finland, "renovation is broken down into basic renovation and refurbishment. In National Accounts, renovation refers to fundamental improvement of a building (after renovation the value of a building is comparable to a new building). Refurbishment is a less significant activity. Refurbishment refers to the regular repairs and maintenance of buildings". .

National statistics for 2016 give residential building construction outputs for "repairs and maintenance" as bEUR 7.3 and bEUR 6.7 for new building PDF. A separate report gives "building and apartment renovation" costs as bEUR 6.6 in 2016 PDF. It is likely that "repairs and maintenance" and "renovation" both refer to ORMI.

Netherlands 2012

In 2016, residential buildings construction output was 24.8% of total residential buildings construction output for "maintenance" and 27.3% for "restoration and renovation" PDF.

Germany 2010

In Germany in 2010, the output of residential construction was bEUR 108.66 for "modernisation" and bEUR 13.53 for maintenance. The residential building output was bEUR 122 for existing buildings and bEUR 35 for new buildings, for a total residential building output of bEUR 157 PDF. ORMI was therefore 122.19/157 = 77.8%. Assuming that the modernisation output corresponds to OI, OI is 108.66/157 = 69.2% of total residential building output and GRM is 5.0% of GNB.

Annex 2: Euroconstruct data

Year TOTAL, bEUR RES, bEUR RES, % TOTAL Res New, bEUR Res New, % RES Res Ren, bEUR Res Ren, % RES Ref
2009 32.6 2012a
2012 32.6 2012a
2013 1285 601.4 46.8 (231.6) 18/0.468 =38.5 28/0.468 =59.8 2013a,b
2014 (1283) 590 about 46 18/0.46 = 39.1 27.5/0.46= 59.8 2014a
2015 58 21/0.58= 36.2 37/.058= 63.8 2015a
2015 646 387.6 60 2015b
2016 1406 2016a
2017 1521 848.7 55.8 2017a

The table gives residential building outputs (RES) as a percentage of the total construction output (TOTAL) with the residential building new build (Res New) and residential building rennovation (Res Ren) outputs as a percentage of the residential building output (RES).

  • In 2012, residential construction in Euroconstruct countries (EC18) made up 32.6% of total output 2012a.

  • In 2013, Euroconstruct's total construction volume was bEUR 1285 with residential building construction accounting for just under 46.8% of total construction output 2013a. In 2013, of the nearly 46% of total construction output associated with residential building: new construction accounted for 18% and renovation and maintenance for 28% 2013b.

  • In 2014, residential construction amounted to bEUR 590 (new construction 18%, renovation and maintenance 27.5%), around 46% of total construction output 2014a.

  • In 2015, dwellings output was 58% of total output with new dwellings accounting for 21% of total output and the renovation of dwellings for 37% of total output 2015a.

  • In 2015, the EC19 residential output was bEUR 646, with renovations accounting for 60% 2015b.

  • In 2016, the total construction output was estimated as bEUR 1406, up by 2.6% from 2015 2016a, so the total construction output in 2015 was bEUR 1370.

  • In 2017, new residential construction accounted for 42.6% of total new construction spending, non-residential construction for 34.4% and civil engineering for 23%.

Updated February 2019