Construction output in Great Britain: Aug 2016

This is the second release of construction output covering data post-EU referendum. Following a revised increase in July 2016 of 0.5%, construction output decreased in August 2016 by 1.5%. You should note that we always warn against overly interpreting one month’s figures.

All new work, and repair and maintenance reported decreases, falling by 1.4% and 1.5% respectively, with the largest downwards contribution in all new work coming from infrastructure.

The underlying pattern as suggested by the 3 month on 3 month movement in output in the construction industry decreased by 1.3%.

The longer-term picture as suggested by the year-on-year movement shows output in the construction industry increased by 0.2% compared with August 2015 with all new work increasing by 1.1%, however, repair and maintenance decreased by 1.3%.

From January 2017, we are improving the way we publish economic statistics, with related data grouped together under new "theme" days. This will increase the coherence of our data releases and involve minor changes to the timing of certain publications. For more information see Changes to publication schedule for economic statistics.

The only period open for revision is July 2016 which has been revised upwards by 0.5 percentage points from flat to an increase of 0.5%. This was caused by the incorporation of late data with the largest revisions being reported in infrastructure and private new housing. More information on revisions can be found in the Quality and Methodology section of this release.

Output is defined as the amount charged by construction companies to customers for the value of work (produced during the reporting period) excluding VAT and payments to sub-contractors.

Construction output estimates are a short-term indicator of construction output by the private sector and public corporations within Great Britain and are produced from a monthly survey of 8,000 businesses in Great Britain. The estimates are produced and published at current prices (including inflationary price effects) and at chained volume estimates (with inflationary effects removed) both seasonally adjusted and non-seasonally adjusted.

Chained volume measures are also described as volume. Construction output is used in the compilation of the output approach to measuring gross domestic product (GDP).

Detailed estimates along with a longer run of time series data are available to download in the Output in the Construction Industry, August 2016 datasets. In these tables, you will find chained volume estimates back to Quarter 1 (Jan to Mar) 1997 and monthly estimates back to January 2010. Current price non-seasonally adjusted data are available back to Quarter 1 (Jan to Mar) 1955. More information on these statistics can be found in the "definitions and explanations" section in the background notes.

The data published in this release cover construction estimates for Great Britain. Construction output estimates for Northern Ireland can be obtained from the Department for the Economy website.

National Statistics status

On 11 December 2014 the UK Statistics Authority announced its decision to suspend the designation of Construction Price and Cost Indices due to concerns about the quality of these deflators. As a result the UK Statistics Authority announced its decision to suspend the designation of Output and New Orders as National Statistics in respect of the Code of Practice for Official Statistics.

We took responsibility for the publication of the Construction Price and Cost indices from the former Department of Business Innovation and Skills (BIS) on 1 April 2015, introducing an interim solution for measuring output prices in June 2015 for all periods from January 2014 onwards. We are currently developing a long-term solution for the deflation of construction statistics.  

In August 2016 all work:

• decreased by 1.5% compared with July 2016

• increased by 0.2% compared with August 2015

• in the 3 months (June 2016, July 2016, August 2016) compared with the previous 3 months (March 2016, April 2016, May 2016), decreased by 1.3%

• in the 3 months (June 2016, July 2016, August 2016) compared with the same 3 months a year ago, decreased by 0.5%

Figure 1 shows the 2 main components of all work. The monthly series began in January 2010 and shows the monthly path is volatile and dominated by all new work. The early part of the time series shows that after a rise in output in early 2010 the level remained fairly consistent until late 2011 when output started to fall.

During 2012, output remained fairly static before increasing steadily in 2013 and 2014, with all new work, and repair and maintenance contributing towards the overall growth. However, towards mid-2014, new work continued to increase while repair and maintenance remained fairly flat. In the most recent periods there has been a downturn in both all new work, and repair and maintenance, however, new work is now 20.0% higher than in January 2010 and repair and maintenance 11.0% higher over the same period.

Figure 2 looks at the main components of all new work. There was sustained growth in new housing from early 2013 to late 2014 and, after a contraction in mid-2015 there was a return to growth at the end of the year. The picture in 2016 is mixed, with periods of contraction and growth.

Infrastructure reported a decrease of 5.1% in August 2016 compared with July 2016 after an increase of 6.1% in July 2016 compared with June 2016. On the year there was a fall of 9.3%, which is the sixth consecutive month of year-on-year decreases. Infrastructure is a particularly volatile series due to the range of products that are included within this type of work; large movements are therefore not unusual.

Other new work has been fairly flat since early 2012. August 2016 continued this trend with an increase of 0.2% compared with July 2016, while there was an increase of 1.4% compared with the same period last year. Unlike new housing or infrastructure, other new work output has not reached the level seen in 2010.

Figure 3 looks at the main components of new housing. In the early part of the time series growth was dominated by public new housing, however, in the latter part of the time series private new housing has been the biggest contribution to growth. In August 2016, total new housing decreased by 1.3% compared with July 2016 with falls in both public and private new housing of 2.1% and 1.2% respectively.

Compared with August 2015, there was an increase of 7.9% in total new housing, which is the ninth consecutive period of year-on-year growth. There was an increase of 9.4% in private new housing, offset slightly by a fall of 0.8% in public new housing, which is the smallest contraction seen since April 2015. There have been 41 consecutive periods of year-on-year growth in private new housing, while public new housing has seen a run of 16 periods of contraction.

Figure 4 looks at the 2 main components of repair and maintenance. In the early part of the time series, housing repair and maintenance performed at a higher level than non-housing repair and maintenance. For a number of years both housing and non-housing repair and maintenance performed at a similar level, however, in more recent periods, housing repair and maintenance has shown stronger growth.

In August 2016, there was a fall of 1.5% in all repair and maintenance compared with July 2016. The level of £3.9 billion is the lowest since September 2013 when it was £3.8 billion. There were decreases in both housing and non-housing repair and maintenance of 2.1% and 0.9% respectively. The level of housing repair and maintenance at £2.0 billion was the lowest since February 2015, while the level of non-housing repair and maintenance of £1.9 billion was the lowest seen since March 2016.

Compared with the same period a year ago, repair and maintenance decreased by 1.3%. There was a fall of 3.9% in housing repair and maintenance, while non-housing repair and maintenance increased by 1.5%.

The monthly series for construction output is a fairly volatile one with erratic month-on-month movements. To remove the volatility the 3 month on previous 3 month path is shown alongside the monthly series in Figure 5.

In August 2016, there was a decrease of 1.5% compared with the previous month in all work, however, looking at the rolling 3 months of June 2016, July 2016 and August 2016, there was a fall of 1.3% compared with the rolling 3 month of March 2016, April 2016 and May 2016.

Table 1 provides a summary of growth rates across the different types of construction work in August 2016. Some main points from this table are as follows:

• there were month-on-month decreases in all work types except private commercial work

• within all new work there were year-on-year increases in private new housing, public other new work and private commercial work while public new housing, infrastructure and private industrial work showed decreases

• within all repair and maintenance there were year-on-year decreases in public housing repair and maintenance while private housing and non-housing repair and maintenance showed increases

Figure 6 shows the contribution of each sector to output growth in the construction industry between August 2016 and July 2016. In August 2016, all work types except private commercial work contributed to the decrease in construction output. The largest downwards contribution came from infrastructure.

Output in the construction industry estimates are produced from the Monthly Business Survey on the second Friday of the month, 2 months after the reporting month. Revised results, for previously published periods, are published in line with the National Accounts Revisions Policy. More information about the data content for this release can be found in the background notes.

Revisions are an inevitable consequence of the trade-off between timeliness and accuracy. The response rate in August 2016 was 70.8% of questionnaires, accounting for 82.0% of registered turnover in the construction industry. Therefore the estimate is subject to revisions as more data become available.

The monthly output in the construction industry time series now spans 80 months, however, you should note that 60 months is the minimum time span recommended by Eurostat for seasonal adjustment. While the seasonal pattern is generally established after 60 months in a monthly time series, there is still potential for increased revisions until the seasonal pattern has matured.

All estimates, by definition, are subject to statistical uncertainty and for many well-established statistics we measure and publish the sampling error associated with the estimate, using this as an indicator of accuracy. For construction output we publish sample and non-sample errors in Table 11 of the Output in the construction industry dataset. It should be noted that we are continually working on methodological changes to improve the accuracy of the construction output estimates. Progress on these can be found on the ONS continuous improvement page on our website.

Construction estimates are a main component of the output approach to measuring GDP along with the estimates of services, production and agriculture. To help you, the short-term economic indicator releases that directly feed into GDP include an additional table of the GDP components. It is anticipated that this table will inform you of the relationship between the individual components which comprise GDP output. The publication dates and the quarterly growths of the individual GDP components are shown below.

Each component of GDP has a weight within GDP based on its value in 2013. Construction has a weight of 59, which means that it is 59 parts of the 1,000 that make up total GDP.

To determine the effect each component has on GDP multiply the component growth by its weight in GDP.

An example using Quarter 1 (Jan to Mar) 2016 data:

Construction growth = -0.3
Weight in GDP = 0.059 (59/1000)
Effect on GDP = -0.3 * 0.059 = -0.02 or 0.0 to 1 decimal place (dp).

Revisions to components and the effect on GDP can be calculated using the same process. As a general rule there are no revisions to GDP when the component revisions are:

Index of Production (IoP) = between 0.3 and -0.3
Construction = between 0.9 and -0.9
Index of Services (IoS) = 0.0 (all values above or below 0.0 effect GDP due to the high weight of IoS in GDP).

Because;

IoP = 0.146*0.4 = 0.0584 or 0.1 to 1 dp
Construction = 0.059*0.9 = 0.0531 or 0.1 to 1 dp
IoS = 0.788*0.1 = 0.0788 or 0.1 to 1 dp

Table 2 shows the latest monthly and revised quarterly output figures that fed into the Quarterly National Accounts release for Quarter 2 (Apr to June) 2016 published on 30 September 2016.

The Quarterly National Accounts published on 30 September 2016 contained an estimate for quarterly construction of a decrease of 0.1%. This estimate has not been revised within this release.

The monthly estimate of construction output fell by 1.5% in August 2016, following 0.5% output growth in July. Weaker figures in August revert construction output to roughly the same level as August 2015 and some way below the average level seen across the first 6 months of 2016. The largest negative contributions this month came from infrastructure (minus 0.7 percentage points on the month), and repair and maintenance of public housing (minus 0.3 percentage points on the month).

In line with the weakness shown in this month’s ONS data, the Bank of England’s latest Agents’ summary of business conditions reported a slowdown in construction output, partly due to reduced growth among house builders. Likewise, the Markit/CIPS UK Construction Purchasing Managers’ Index for August recorded a mild reduction in business activity.

Similarly, the RICS UK Construction Market Survey for Quarter 2 (Apr to June) 2016 cited skills shortages, financial constraints and planning and regulatory delays as barriers to growth in construction (although this data source only covers the period up to the end of June 2016).

Housing

Construction of new housing output fell in August (by 1.3% compared with July), with broader indicators suggesting flat demand for housing in August: for example, the HM Revenue and Customs’ UK Property Transactions Statistics release reported a negligible change in the number of residential property transactions between July and August.

With respect to house prices, the ONS House Price Index for July 2016 reported an 8.3% increase in house prices in the year from July 2015, continuing the relatively strong growth seen since mid-2013. The average house price was £1,000 higher than in the previous month. More recently, Nationwide and Halifax house price data reported annual house price growth in the year to August 2016 of 5.6% and 6.9% respectively.

Output in the construction industry follows the Eurostat Short Term Statistics (STS) regulation for production in construction. Before any comparisons are made with the euro area or EU28, it is worth noting that the UK is the only member state to follow the A method for compiling production in construction statistics.

The latest release of production in construction published by Eurostat on 19 September 2016 showed that construction output in the euro area (EA19) increased by 1.8% and in the EU28 increased by 1.1% in July 2016 compared with June 2016. The Great Britain estimate for July 2016 showed that construction output increased by 0.5%. It should be noted that an accurate comparison cannot be made as Eurostat data are calculated on a 2010 = 100 basis, while Great Britain data are calculated on a 2013 = 100 basis.

An international estimate of construction output outside of the EU was published by the US Census Bureau on 3 October 2016 in its release Value of construction put in place. It should be noted that accurate comparisons cannot be made against these or other international statistics due to a variety of reasons, including differences in methodology.

1. Methods

Our monthly construction output survey measures output from the construction industry in Great Britain. It samples 8,000 businesses, with all businesses employing over 100 people or with an annual turnover of more than £60 million receiving a questionnaire by post every month.

Estimates are based on output data collected through the monthly Construction Output Survey. Response rates at the time of publication are included for the current month and the 3 months prior. The response rates for those historical periods are updated to reflect the current level of response, incorporating data from late returns. There are 2 response rates included, with 1 percentage for the amount of turnover returned and the other percentage for the amount of questionnaire forms.

Since the 1950s, new orders in construction data had been collected from a sample survey of businesses; originally monthly and then quarterly. There were some known quality issues with the survey data as:

• the coverage of the survey was unknown

• new orders allocated to regions were not always accurately recorded

The new orders data are now supplied under contract by Barbour ABI. Barbour ABI provide us with improved coverage and regional splits of new orders in construction data.

2. Quality

The latest Quality and Methodology report for the Output of the Construction Industry estimates can be found on our website.

The latest Quality and Methodology report for New Orders in the Construction Industry estimates can be found on our website.

These contain important information on:

• the strengths and limitations of the data and how it compares with related data
• users and uses of the data
• how the output was created
• the quality of the output including the accuracy of the data

3. Revision policy

Construction output conforms to the standard National Accounts Revisions Policy, which can be found on our website. In line with this, the construction output release for August 2016 has no revisions to earlier periods.

Figures for the most recent months are provisional and subject to revision in light of (a) late responses to the Monthly Business Survey MBS (b) revisions to seasonal adjustment factors which are re-estimated every period and (c) improved treatment of outliers.

4. Revisions

One indication of the reliability of the main indicators can be obtained by monitoring the size of revisions. Analysis of the previously published quarterly seasonally adjusted chained volume measure series has shown that revisions to construction data are small. Generally these quarterly revisions are less than 1 percentage point when compared with the final revised period 5 quarters after initial publication. This indicates that the published estimates are a reliable snapshot of the output in the industry at the date of publication.

The size and pattern of revisions for both output and new orders data which have occurred in the open period can be found in the 1 month and 3 month revisions triangles datasets on the construction web page. Please note that these indicators only report summary measures for revisions. The revised data may be subject to sampling or other sources of error. Details about this revisions material can be found on our revisions page.

It should be noted that due to seasonal adjustment taking place on a short span of data points used to interpret the seasonal effects, there is potential for increased revisions until the seasonal pattern is established within the time series. The seasonal pattern is generally established after 60 months in a monthly time series.

Please note that a monthly seasonally adjusted chained volume series is not available pre-2010. This is due to monthly data not being available for this period. These data are a requirement for creating previous year’s prices from which chain-linked volume measures are created:

• the coverage of the survey was unknown

• new orders allocated to regions were not always accurately recorded