Construction output for January 2011 was lower than December 2010 despite the latter month (December) experiencing several days of severe snow. It was suggested by some construction analysts that there could be a lag in the timing between work being carried out and its being reported on ONS’s monthly business survey. ONS contacted a sample of businesses to establish if there was any evidence to support this suggestion. The data collected by the office show that the total output estimate published by ONS is based on responses that are, on average, lagged by 0.24 weeks. This estimate is based on 187 responses from a total of 395 businesses contacted. On the basis of this analysis, ONS does not intend to make any change to the published estimates.
Construction output estimates in January 2011 were lower than in December 2010 (£7.8 bn and £8.5 bn respectively). Industry commentators expressed surprise at this as there was an expectation that January's output would 'bounce back' after a low December output that was affected by several episodes of heavy snow.
An exercise was undertaken to contact construction businesses to confirm:
a. how they compiled their output data; i.e. what source was used
b. if there was any lag and, if so, the size of the lag
c. whether January's output was greater than December
Almost 400 businesses were contacted by phone, although only half of these businesses supplied usable information.
Table 1 shows the results of the question; “On what basis do you calculate your business’ construction output?”
|Band (employment)||Work done/chargeable work||Invoices raised||Orders received||Turnover/income rec'd/invoices paid||Blank|
|<100 and >£60m||2||0||0||0||0|
Table 2 shows the result of the question; “Is the output you record for the reference month actually done in that period?”
|Work reported in reference period||Response rate|
|<100 and >£60m||2||0||0||100%|
Table 3 shows the response proportions to the above question
|Work reported in period (proportions)|
|<100 and >£60m||1.00||0.00|
Table 4 shows the results to the question; “If the work reported was not actually done in the reference period, by how much is it lagged?”
|Number of weeks lag|
|<100 and >£60m||0||0||0||0||0||0||0||0||0||0||0|
Table 5 shows the response proportions to the above question
|Number of weeks lag|
|<100 and >£60m||0||0.00||0.00||0.00||0.00||0.00||0.00||0.00||0.00||0.00||0.00|
Table 6 shows the results of calculations that determine a weighted average of lagged durations
|Band||Overall average lagging (weeks)||Proportion by sample size||Proportion by 2010 output||Weighted by sample size (weeks)||Weighted by 2010 output (weeks)|
|<100 and >£60m||0.00||0.01||0.01||0.00||0.00|
The follow up survey shows:
From table 1 - Most businesses compile their output data either from invoices raised or by calculating work done based on chargeable work. Most of the businesses that used turnover or invoices paid were small business where turnover is equivalent to output (little or no subcontracting or purchasing of services) and projects tend to be of very short duration so payment is made in the same period as the work done.
From table 2 and 3 - most businesses compile their data for the current reference period.
From table 4 and 5 - of the 45 businesses that reported a lag in their output figures, larger business reported smaller lags. Overall most business reported either a 2 week or 4 week lag, the larger lags being for small businesses.
From table 6 - the overall lagging by number of forms is 0.53 weeks. The overall lagging in the total output is 0.24 weeks.
Details of the policy governing the release of new data are available by visiting www.statisticsauthority.gov.uk/assessment/code-of-practice/index.html or from the Media Relations Office email: firstname.lastname@example.org