Construction output in Great Britain: May 2020

• Construction output grew by a record 8.2% in the month-on-month all work series in May 2020 following the record decline of 40.2% in April 2020; the level of construction output is now down 38.8% on February 2020 before the impact of the coronavirus (COVID-19) pandemic.

• Construction output fell by a record 29.8% in the three months to May 2020, compared with the previous three-month period; this was driven by record falls of 30.3% in new work and 28.9% in repair and maintenance.

• The decrease in new work (30.3%) in the three months to May 2020 was because of record falls in most of the new work sectors; private new housing and private commercial were the largest contributors, falling by 42.5% and 29.5% respectively.

• The decrease in repair and maintenance (28.9%) in the three months to May 2020 was because of record falls in all repair and maintenance sectors; the largest contributor was private housing repair and maintenance, which fell by 39.8%.

• The Office for National Statistics (ONS) has released a public statement on COVID-19 and the production of statistics.

Construction output increased by 8.2% in May 2020 compared with April 2020, rising to £8,253 million, though output remains at a substantially lower level than normal compared with the all work construction output series prior to March 2020. This is shown by total construction output in May 2020 being 38.8% (£5,230 million) lower in comparison with the February 2020 level, which was before the impact of the coronavirus (COVID-19).

Following the large record monthly falls in April 2020 across all construction industry sectors, in May 2020 there was monthly growth across all sectors apart from public other new work and public housing repair and maintenance.

Figure 1 shows the monthly and quarterly indexed chained volume measure, seasonally adjusted series. The quarterly series provides a smoother and more comprehensive view of trends within the construction industry, compared with the more volatile monthly series.

March 2020 was significantly impacted by the coronavirus leading to then record falls in monthly growth, including in all work, which fell 5.4%. Construction output in April 2020 was significantly more affected by the coronavirus as large parts of the industry shut down as businesses adhered to the official government guidance. We received a large amount of anecdotal information regarding this from survey respondents, through qualitative comments to the survey along with information from the Business Impact of Coronavirus Survey (BICS). However, it is difficult to quantify the exact impact of the coronavirus alone on the industry.

For May 2020, anecdotal information noted that the partial recovery was because of the gradual easing in official guidance on restrictions in movement across parts of the UK but output remains 38.8% lower compared with February 2020. Table 1 shows the change in output for the types of construction work between February 2020 and May 2020. All types of work have seen large declines in output since the start of 2020.

Coronavirus in May 2020

We have worked closely with respondents and data suppliers and have used additional data sources to inform the estimates in this publication. We used qualitative information sourced from construction industry respondents to the Business Impact of Coronavirus Survey (BICS) to quality assure response we received for May 2020.

Anecdotal evidence from responders to both BICS and the Monthly Business Survey (MBS) suggests a limited increase in activity, though not equally across all construction sectors and all UK regions. Social distancing measures meant where businesses were returning to premises and sites the capacity and level of output were not at the same levels of work experienced prior to the coronavirus pandemic.

We also closely monitor business responses in each publication and use internationally recognised statistical methods to deal with non-response. To assist with this, the mode of data collection from April 2020 onward has moved to online data collection. This led to improved response rates for both April 2020 and May 2020, by both number of forms and turnover coverage of the industry when compared with March 2020, though response is still below levels seen prior to the coronavirus. Further information on this can be found in Section 9, Measuring the data, under the Impact of online data collection on response rates subheading.

For further information, we have released a public statement on COVID-19 and the production of statistics.

Table 2 shows different measures of growth for all types of work in May 2020. Illustrating the weakness in the industry since the coronavirus (COVID-19) pandemic, every type of work, apart from infrastructure, has seen a record decrease in the three-month on three-month series in May 2020.

It is also noticeable that while new housing saw record monthly increases of 21.4% and 42.1% in private and public work respectively in May 2020, both have experienced a significant decline in output over recent months. This is shown with both seeing record declines in the three-month on three-month series and three-month on three-month a year earlier series.

Contributions to growth

Construction output can be broken down by different types of work. These are categorised into all new work, and repair and maintenance, as shown in Figure 2. All new work accounts for approximately two-thirds of all work, while repair and maintenance accounts for approximately one-third of all work.

There was growth in both new work, and repair and maintenance in May 2020. New work saw a comparatively stronger rebound in May 2020 compared with repair and maintenance, though both remain at a low level in comparison with the periods before the coronavirus impacted.

Figure 3 shows the difference in three-month on three-month output from the different construction sectors, taken from our seasonally adjusted, chained volume measure series. Construction output fell by 29.8% (£12,160 million) in the three months to May 2020, compared with the previous three-month period.

There were record percentage declines in every construction sector in the three-month on three-month series in May 2020, apart from infrastructure, which fell 13.2% (£747 million) but was not a record. Both new work, and repair and maintenance contributed to the large fall in output in the three months to May 2020.

New work fell by a record 30.3% (£8,217 million) in the three months to May 2020, again driven by record falls in every new work sector, apart from infrastructure. As shown in Figure 4, the fall in the level of infrastructure output was not as sharp as in other components, with the level of output in all other sectors substantially lower in the three months to May 2020.

Repair and maintenance saw a record fall of 28.9% (£3,943 million) in the three months to May 2020, driven by record falls in all repair and maintenance sectors. Repair and maintenance has not seen growth in the three-month on three-month series for 12 consecutive months (since May 2019). The largest contributor to the fall in repair and maintenance was private housing repair and maintenance, which decreased 39.8% (£1,960 million).

As shown in Figure 5, despite monthly growth in private housing and non-housing repair and maintenance, the level of output remains substantially lower in all components of repair and maintenance in May 2020 compared with periods prior to February 2020.

The largest single contributor to the decline in all work output in the three months to May 2020 was private new housing, which fell 42.5% (£3,792 million) in the three-month on three-month series. From the anecdotal evidence received from contributors, housebuilders began to return to construction sites across the UK during May 2020, however, because of the social distancing measures these were not at full capacity. It was also highlighted that this was mainly to complete projects currently in progress rather than starting work on new projects.

Figure 6 shows the significant contribution to the fall in all work in April and May 2020, from total housing compared with the other construction sectors combined.

Construction output grew by 8.2% (£626 million) in May 2020 compared with April 2020. This was because of increases in all construction sectors, except public housing repair and maintenance, which fell 9.5% (£33 million) and public other new work, which fell 2.7% (£17 million).

The growth in May 2020 is the largest month-on-month growth in output since monthly records began in 2010 but it follows three consecutive periods of decline where construction output fell by a combined 44.5% (£6,108 million) between January 2020 and April 2020.

Figure 7 shows the month-on-month output growth in May 2020 for the different construction sectors, taken from our seasonally adjusted, chained volume measure series.

New work grew by a record 11.1% (£553 million) in May 2020 compared with April 2020, with repair and maintenance growing by 2.7% (£73 million). This is a record monthly growth in new work, beating the previous record of 6.9% (£468 million) in January 2011.

Figure 8 shows the weaker growth in repair and maintenance compared with new work in May 2020. It also illustrates that the level of output in both new work, and repair and maintenance remains relatively low when compared with periods prior to February 2020.

The record 11.1% (£553 million) growth in new work in May 2020 was driven by increases in all new work sectors except for public other new work, which fell 2.7% (£17 million). The largest contributors to this growth were private new housing, which grew a record 21.4% (£236 million), and infrastructure, which grew by 12.7% (£186 million).

The record growth in private new housing followed three consecutive periods of large declines, with the level of output still relatively low compared with months prior to February 2020. These declines were driven by a mixture of adverse weather and flooding in February 2020 and the coronavirus (COVID-19) in March and April 2020.

Repair and maintenance grew by 2.7% (£73 million) in the month-on-month series in May 2020. This was because of a 4.7% (£74 million) increase in non-housing repair and maintenance, and a 4.4% (£32 million) increase in private housing repair and maintenance. In comparison, public housing repair and maintenance saw a fall of 9.5% (£33 million). This comparatively weaker bounce back in repair and maintenance in May 2020 is worth noting, as repair and maintenance had seen similar falls to new work in recent months, as shown in Figure 9.

It is also worth noting that while businesses may not have undertaken repair and maintenance activity in May 2020 because of the official guidance on restrictions in movement across the UK, many households had increased time to undertake home improvements. This in turn is likely to have contributed to the 67.4% increase in the volume of sales for hardware, paints and glass stores as shown in the Retail sales, Great Britain: May 2020 bulletin (See Figure 6 in May 2020 Retail sales bulletin).

Business Impact of Coronavirus (COVID-19) Survey (BICS)

The Business Impact of Coronavirus (COVID-19) Survey (BICS) shows some evidence of increasing construction industry activity at the end of May 2020. Qualitative information sourced from this survey was used to quality assure responses we received for the Monthly Business Survey for construction and allied trades (MBS) for May 2020. Of construction industry respondents to BICS Wave 6, 12.6% responded that they had restarted trading after a pause in trading. This was more than any other industry and well above the 4.7% average for all industries, as shown in Table 3.

This is also described in the anecdotal evidence we have received from contributors to the Monthly Business Survey. Many businesses had described that they were recommencing work following a pause in activity in April and early May even if not at the same capacity as before the coronavirus pandemic.

BICS Wave 6 data also showed that construction industry respondents on average had more of their workforce returning from furlough (13.0%) than any other industry, which was more than double the average for all industries (6.1%), as shown in Table 4.

This is the first monthly release to incorporate the revisions made in the GDP quarterly national accounts, UK: January to March 2020 release, published on 30 June 2020. As a result, revisions have been made back to January 2019.

These revisions saw quarterly construction output growth revised for all quarters as shown in Table 5. Most notably Quarter 4 (Oct to Dec) 2019 is revised down by 0.9 percentage points, from negative 0.1% to negative 1.0%. In contrast, Quarter 1 (Jan to Mar) 2020 is revised up 0.9 percentage points from negative 2.6% to negative 1.7%.

These revisions have also led to a downward revision in 2019 annual growth by 0.4 percentage points, from 2.3% to 1.9%.

In addition to the revisions to construction output first published within the GDP quarterly national accounts, UK: January to March 2020 release, this bulletin also includes revisions to the April 2020 construction output estimates first published in Construction output in Great Britain: April 2020 on 12 June 2020. Table 6 shows the revised monthly and three-month on three-month paths as a result of these revisions from October 2019 to April 2020.

Monthly growth since the start of 2020 has been revised for all months. This is partly because of the lower response at the time of the first publication for all these months because of the challenges of data collection the coronavirus (COVID-19) poses. Table 7 covers the response rate based on turnover coverage at time of the first estimates against the latest response used to compile the estimates in this release.

Details of why revisions can be seen across the whole period are available in the Measuring the data section.

For further information on the revisions profile please see the output in the construction industry revisions triangles published on a one-month and three-month growth basis.

Construction output estimates

Construction output estimates are monthly estimates of the amount of output chargeable to customers for building and civil engineering work done in the relevant period, excluding Value Added Tax (VAT) and payments to subcontractors.

Seasonally adjusted estimates

Seasonally adjusted estimates are derived by estimating and removing calendar effects (for example, leap years such as this year) and seasonal effects (for example, decreased activity at Christmas because of site shutdowns) from the non-seasonally adjusted estimates.

Value estimates

The value estimates reflect the total value of work that businesses have completed over a reference month.

Volume estimates

The volume estimates are calculated by taking the value estimates and adjusting to remove the impact of price changes.

Construction output data collection

Our monthly Construction Output Survey measures output from the construction industry in Great Britain. The survey samples 8,000 businesses, with all businesses employing over 100 people, or with an annual turnover of more than £60 million, receiving an online questionnaire every month. The survey's results are used to produce non-seasonally and seasonally adjusted monthly, quarterly and annual estimates of output in the construction industry at current price and at chained volume measures (removing the effect of changes in price).

Data on new orders supplied by Barbour ABI are used to model the breakdown of the overall output figures for Great Britain into the lower level and regional data seen in Tables 1 and 2 of Construction output: sub-national and sub-sector.

Revisions to construction output data

Revisions in the release are a result of:

• late responses to survey returns replacing imputations, or revisions to original returns

• revisions to seasonal adjustment factors, which are re-estimated every month and reviewed annually

• HM Revenue and Customs (HMRC) Value Added Tax (VAT) returns replacing Monthly Business Survey (MBS) data for small- and medium-sized businesses for Quarter 4 (Oct to Dec) 2019 for the first time as well as potential revisions to previous VAT turnover data

• revisions to the input series for the Construction Output Price Indices

Quality and methodology

More quality and methodology information on strengths, limitations, appropriate uses, and how the data were created is available in the Construction output QMI and New orders in construction QMI.

Value Added Tax (VAT) data

Alongside the Monthly Business Survey (MBS), further information on output is gained from VAT turnover data, which are used to replace survey data for small- and medium-sized businesses. However, because of the delay in companies making VAT returns, these data are only taken on after a lag period. Currently, VAT turnover data are used for the period Quarter 1 (Jan to Mar) 2016 to Quarter 4 (Oct to Dec) 2019.

Further information on the use of VAT turnover in construction output estimates and its impact can be found in the following articles:

• VAT turnover implementation into national accounts

• VAT turnover data in National Accounts: background and methodology

• Quality assurance of administrative data (QAAD) report for VAT turnover data

Coronavirus (COVID-19) impact on ONS construction output in May 2020

Temporary ceasing of Output in the construction industry: sub-national and sub-sector data

The coronavirus (COVID-19) pandemic presents a significant challenge to the UK, and the Office for National Statistics (ONS) is working to ensure that the UK has the vital information needed to respond to the impact of this pandemic on our economy and society. This means we will need to ensure that information is provided faster, using new data sources and changing how our surveys operate, to ensure we provide the information necessary as the situation unfolds.

The effects of the outbreak on ONS capacity and capability during this period means we have reviewed the existing construction statistics releases and will be temporarily suspending the Output in the construction industry: sub-national and sub-sector dataset. This is to protect the delivery and quality of our remaining outputs as well as ensuring we can respond to new demands as a direct result of the coronavirus. This is also partially a reflection of the limitations of the model used to apportion new orders data to produce sub-level output data.

Impact of online data collection on response rates

As highlighted in Section 2, the coronavirus pandemic has significantly impacted construction output in recent periods. Official guidance on restrictions in movement for Great Britain, leading to the closure of work sites, also impacted response rates.

Data for the Monthly Business Survey for construction and allied trades (MBS) have been collected via online questionnaire since April 2020. This has meant that respondents can log on from any location and submit their data at an appropriate time. The paper questionnaire was moved to an online data collection platform, with minimal changes made to the questionnaire design. The only notable change was to address the treatment of housing associations, which are now classified as private housing rather than public housing, as previously on paper. For further information on this classification decision please see this statement for England and this article for Scotland, Wales and Northern Ireland.

Table 8 shows the response rates to the MBS at time of publishing, for each reference period. Response rates were relatively low in March 2020 and since then have improved when measured by both the turnover coverage of the industry and proportion of questionnaire forms returned. This illustrates the benefits of the move to electronic data collection, though response rates remain lower when compared with months before February 2020. It is worth noting that while response rates are low at the time of first publishing for the reference months in 2020, further response has been received and used in the compilation of the latest estimates.

To deal with non-response we impute for missing data using ratio imputation. This is a simple but effective method, used as a standard internationally. The method calculates the growth in the industry based on those businesses that did respond and applies it to the last known value for the non-responder. This means that if output notably reduces in an industry from one month to the next, the imputed values for non-respondents in that industry will also notably reduce when compared with the last known value.

Further information on the imputation methods for non-response is available.

While international best practice is used to impute for non-response, with the lower response rates highlighted in Table 8. It is important to note that the revisions to the months in 2020 may be larger than previous revisions profile prior to 2020, as actual data and revised data replace the larger than normal number of imputations for non-response at the time of the first monthly estimate.

Zero return responses to the Monthly Business Survey for construction and allied trades (MBS)

A zero return refers to when a survey respondent reports figures of zero across all types of work, meaning the total value of work done is zero for that reference month. Figure 10 shows zero returns as a proportion of all returns at the time of the first estimate for a reference month.

Prior to March 2020, we had a stable element of approximately 7% to 10% reporting zero returns. This partially increased in March 2020, but significantly increased into April 2020 as sites were closed because of the official guidance on restrictions of movement in Great Britain. We saw partial decline of zero returns in May 2020, however, this is still significantly higher than the months prior to April 2020.

Impact on seasonal adjustment of May 2020

The monthly chained volume measures are seasonally adjusted using a seasonal adjustment software tool (X-13-ARIMA-SEATS). The monthly series individual type of work series is then aggregated to form the quarterly seasonally adjusted chained volume measure series.

The seasonal adjustment parameters for output in the construction industry are reviewed annually. However, because of the volatility of these statistics, time series analysis experts are regularly asked to review the seasonal adjustment when required. This approach has been adopted for the latest months and has resulted in changes to seasonal adjustment specification files to ensure the seasonal adjustment parameters are appropriate. Most notably, because of the large decline since April 2020, all types of work are being treated as an additive outlier in these specification files for April and May 2020.

Coronavirus impact on the May 2020 bias adjustment

Typically, an adjustment to address any bias in survey responses for construction output is applied to the early construction output monthly estimates. See Improvements to construction statistics: Addressing the bias in early estimates of construction output, June 2018 published on 4 June 2018. The bias adjustment methodology is based on historical data. As the response rates for May 2020 are lower in comparison with months prior to February 2020 (Table 8) and no comparable historical data are available at the time of the first estimate for a reference month, we have not applied a bias adjustment for May 2020.

Links to additional ONS sources of coronavirus information

Since the last monthly output publication, various articles have been published, which help describe the ONS response to how the coronavirus might be seen in our estimates:

Coronavirus and housing indicators in England and Wales (published 2 July 2020)

Coronavirus and the effects on UK GDP (published 6 May 2020)

Real-time turning point indicators: a UK focus (published 27 April 2020)

Communicating gross domestic product (published 27 April 2020)

Our latest data and analysis on the impact of COVID-19 on the UK economy and population is also now available on a new webpage. This will be the hub for all special virus-related publications, drawing on all available data.

The Office for National Statistics (ONS) has released a public statement on the coronavirus and the production of statistics and any specific queries can be directed to the Media Relations Office.

Incorporating Construction Output Prices from Quarter 2 (Apr to June) 2020

As announced in the article Coronavirus and the effects on UK prices, Construction Output Price Indices data will form part of the quarterly Construction output in Great Britain bulletins. This change will be introduced from the April to June 2020 reference period, which will be published in August 2020 for the first time.

Exiting the EU

As the UK leaves the EU, it is important that our statistics continue to be of high quality and are internationally comparable. During the transition period, those UK statistics that align with EU practice and rules will continue to do so in the same way as before 31 January 2020.

After the transition period, we will continue to produce our national accounts statistics in line with the UK Statistics Authority's Code of Practice for Statistics and in accordance with internationally agreed statistical guidance and standards.

The Withdrawal Agreement outlines a need for UK gross national income (a fundamental component of the national accounts, which includes gross domestic product (GDP)) statistics to remain in line with those of other EU countries until the EU budgets are finalised for the years in which we were a member. To ensure comparability during this cycle, the national accounts will continue to be produced according to European System of Accounts (ESA) 2010 definitions and standards.

Data quality

These estimates are widely used by private and public sector institutions, particularly by the Bank of England and HM Treasury, to assist in informed decision-making and policymaking. Construction output is an important economic indicator and is also therefore used in the compilation of the output measure of gross domestic product (GDP).

National Statistics status

Great Britain construction output statistics and construction new orders are designated as National Statistics, in accordance with the Statistics and Registration Service Act 2007 and signifying compliance with the Code of Practice for Statistics.

Comparability

Output in the construction industry follows the Eurostat short-term business statistics (STS) regulation for production in construction. Headline volume estimates of construction output are assessed against Eurostat's handbook on price and volume measures in national accounts.

Construction output data used within this release are also used in the compilation of the GDP monthly estimate. While monthly data are available in the output in the construction industry back to January 2010, a longer time series back to 1997 can be obtained in the monthly GDP datasets. Data prior to 2010 are derived using statistical methods from the available quarterly construction output data and should therefore be treated with some caution.

Within this publication, a monthly, all work chained volume measure, seasonally adjusted series can be obtained back to January 1997 in index form to four decimal places. This can be found in the following datasets: Monthly GDP and main sectors to four decimal places and Monthly gross domestic product: time series.

Construction statistics recent engagement and development work

Further information on construction statistics development can be found in:

• Housing in construction output statistics, Great Britain: 2010 to 2019 (30 May 2020)

• Comparing ONS's economic data with IHS Markit and CIPS Purchasing Managers' Index surveys (published 21 October 2019)

• Conceptual and methodological differences between private housing construction output and gross fixed capital formation private sector dwellings (published 31 May 2019)

• Construction statistics development: improving the understanding of new orders in the construction industry and the gap between output and new orders (published 30 October 2018)

Further articles on other construction statistics development work and analysis are available.