Short-term business statistics - revisions
From Statistics Explained
Revisions are defined as "any change in a value of a statistic released to the public".
This article provides an overview of revisions and revison policy of short-term business statistics (STS). It is part of a set of background articles treating various methodological aspects of short-term business statistics.
The first section presents the main reasons for revisions. The second section describes how the European statistical system (ESS) guidelines on revision policy have been implemented in short-term business statistics. Historical data on the revisions of short-term business statistics are presented in the third section. The article also provides some general conclusions on the quality of STS indicators revealed by the revision analysis.
Reasons for revising short-term business statistics
Routine revisions of short-term business statistics are necessary because of late incoming data, seasonal adjustment or regular benchmarking.
Methodological changes and changes of the reference and base year introduce main revisions that may be large in size but take place less frequently and regularly than routine revisions.
Correction of errors may take place at any moment.
- Late data
Short-term business statistics publish monthly and quarterly indicators within one to three months after the end of the reference period. To meet these deadlines STS requires early estimations which may have to be revised later when more complete data becomes available. Late data arrival can occur at least at two different levels.
1. Some respondents of statistical surveys in the Member States send their questionnaires after the national deadlines and the statistical authorities need to estimate the missing data when sending national data to Eurostat. Later, the estimated values are replaced by the more complete data that are re-transmitted to Eurostat.
2. Eurostat often releases its first estimates at the time when some Member States' data are still missing. In short-term business statistics, small countries have often 15 days more time to prepare their national contributions. Later, when all countries have sent their data and some have even revised their first estimates, Eurostat revises the European aggregates.
- Seasonal adjustment
Short-term statistics are adjusted for seasonal effects in order to allow comparing month-on-month or quarter-on-quarter changes whenever indicators are subject to seasonal variation. (For example retail sales in December are always higher than in November because of Christmas shopping. To compare December sales with those of November, the regular impact of Christmas has to be removed from the data.)
The estimation of seasonal effects requires long time series. The mathematical models used to identify seasonal effects work in a way that a new observation can change the seasonal factors applied to the whole time series. Consequently, when the new seasonal factors are applied to the older data, the whole time series may need to be revised. Additionally, an annual (or more frequent) overhaul of seasonal adjustment models is needed to maintain the quality of the models. When models are changed, the adjusted time series are also revised.
Benchmarking (also called data confrontation or consolidation) is defined as the adjustment of (generally) higher frequency data to take account of more complete lower frequency results, which become available only later. Short-term statistics can be confronted with structural business statistics, labour statistics or national accounts. Structural business statistics are also used as the basis for the weights of economic activities within the Member States and for the weights of the countries when compiling the European aggregates of short-term business statistics. They are released more than a year after the end of the reference year, and this has an impact on the updates of weights of the short-term statistics. Furthermore, several countries use quarterly national accounts, for example in estimating the labour productivity that is used for compiling monthly production in construction on the basis of hours worked.
- Methodological changes
Methodological changes, including changes of definitions and classifications, are a trigger for major revisions. They cause statistics to be redefined so that the results for the old and the new time series are different, if they are even comparable at all. In the latter case, the users need to be informed about the break in series. Short-term business statistics are classified according to Statistical classification of economic activities in the European Community (NACE). In 2005, NACE rev. 1.1 was replaced by NACE rev. 2.0 which entailed revisions of those short-term statistics series where the structure of the classification changed.
Currently the replacement of the Classification of types of construction (CC) by NACE and Classification of products by activity in the European Union (CPA) for construction statistics is under discussion. This change would require the recompilation of some construction time series.
- Changes in base and reference year
Revisions can also be caused by changes to the base and reference year, which, together with new weighting systems, can alter the paths of time series. The current base and reference year for short-term statistics is 2010. Changing the weighting system affects the growth rates of the European aggregates, if the weights of the different countries change considerably. The change of the reference year only means that all values of the time series are divided by the average of the new reference year. This can change the index levels for time series, but it does not have an impact on the growth rates.
Alterations to the data may also stem from correcting errors. It is not possible to entirely rule out errors during the compilation process in the Member States and in Eurostat. The correction of such errors requires revisions of the published statistical results. The correction of Member States' data leads also to corrections of the European aggregates. Opposite to all other revisions presented above, corrections of errors cannot be planned in advance, but it is possible to inform the users on how errors are dealt with. Significant errors need to be corrected as soon as possible.
Release and revision policy of short-term business statistics
In the past all STS data, including the EU and euro area aggregates, were revised whenever additional information from national statistical authorities became available. This procedure resulted in almost daily - although usually small - revisions since Member States sent their data on different days to Eurostat. In February 2012 the European Statistical System Committee (ESSC) approved guidelines on how to deal with the various types of revisions for the Principal European economic indicators (PEEIs), see below.
For short-term business statistics, the following changes were implememtend as of October 2012:
- European aggregates are generally released and revised only once per month. This new policy is applied for all Principal European economic indicators (PEEIs) and for the retail trade turnover indicator.
- Presently continuous revisions are still applied for all STS labour input indicators (numbers of persons employed, hours worked, gross wages and salaries), for construction costs and construction prices and for the industrial turnover indicator.
- National data are revised whenever new information becomes available.
For other STS PEEIs and retail trade turnover, the release and revision dates are identical and lie within a few working days after the end of the deadline for new data transmissions from the Member States. Hence data releases and revisions always take place in the beginning of each month.
The new revision policy concerns only routine revisions. Detected errors in national data or in European aggregates are corrected immediately and an error report is released.
Users will be informed about the forthcoming major revisions (e.g. due to methodological changes) in news releases and on Eurostat's website before the event takes place.
Revisions of STS PEEIs - Main statistical findings
In order to monitor the economic and financial development of the Euro area so-called Principal European economic indicators (PEEIs) have been selected. Eight of these 22 indicators are provided by STS:
- industrial production (volume)
- domestic industrial producer prices
- industrial import prices
- production in construction (volume)
- volume of sales in retail trade
- services turnover
- services producer prices
- building permits indicators (number of dwellings and useful floor area).
In this section, the following characteristics related to revisions are described for the above indicators (except for services producer prices):
- mean revision
- mean absolute revision
- range (difference between the biggest positive and smallest negative revision)
- revision tracks plotting the path of individual observations after their first release.
Mean revisions should be zero or close to zero. This means that upward and downward revisions cancel each other out and that there is no bias, i.e. that early releases are not systematically higher or lower than later releases.
Mean absolute revisions should be small in comparison with the revised data (e.g. an index or a growth rate). High mean absolute revisions indicate poor accuracy of early releases - even if mean revisions are small and there is no tendency to over or underestimate.
The range of revisions represents the difference between the largest and the smallest revisions of different observations.
The following revision analyses are based on index levels since spring 2009, i.e. after the implementation of the NACE rev. 2.0 classification. The figures show the most recent developments, starting from the beginning of 2012 for the monthly data and from the beginning of 2011 for the quarterly. In the figures, each line shows the development of an individual observation after its first release. An observation is the value of the index in a reference period.
Table 1 reveals bias and volatility of some short-term statistics PEEIs. In the following graphs, the revisions tracks will be used to explain the results of Table 1.
The index of industrial production shows some variations after the first release of the data (Figure 1). After the change of the seasonal adjustment method from direct to geographically indirect in March 2012, the revisions of the EU-27 aggregates have become quite moderate.
Compared to industrial production volume, domestic industrial producer prices (Figure 2) show very smooth revision tracks. The initial releases are generally not revised at all. The only exception was a short period in 2009 when the NACE rev. 2.0 was still being implemented at the same time as the new base year 2005.
The big revisions of extra-EU industrial import prices (Table 1) can be explained by the fact that Italy started providing this indicator only in December 2011. With its first data delivery, time series going back to 2005 were transmitted. As one of the big euro area countries, Italy has high weight in the EA-17 aggregates. Consequently, the euro area aggregates were revised systematically downwards for the whole period. Otherwise industrial import prices show generally a stable development after the early revisions following the first release (Figure 3).
In the case of the index of production in construction (volume) (Figure 4), there was a systematic increase of all index values in March 2012. This was a result of the change in the aggregation practice that took place at the same time as the change of the seasonal adjustment method: the monthly estimates based on the smaller countries' quarterly data were removed from the total and the monthly index is currently calculated on the basis of the true monthly data only. For this reason, the +2.3 index point change only appears in Table 1 when comparing the first and the latest values.
Volatility was identified as the major deficiency of the index of production in construction (volume) in the EFC's status report of 2008. Eurostat organised a task force in 2010-2011 that issued a report with recommendations aiming at improving the quality of IPC (Guidelines for compiling the monthly index of production in construction). The compilation of the index of production in construction was generally identified as a challenging task. Several countries have started the implementation of the guidelines in their production systems, but the results will not be immediately visible.
Compared to the volume measures of production in industry and construction, the volume of sales in retail sales appears to behave very well (Figure 5). Some revisions can be observed in the months following the initial release but afterwards the values seem to stabilise. The range of revisions is the smallest of all STS PEEIs (Table 1).
Services turnover (excluding turnover in retail trade) does not show significant bias, but there appear to be two level shifts of the series in the middle and at the end of 2011 (Figure 6).
Outside the major revisions, the short-term statistics price indicators are generally rather stable after the first months following the initial release. However, but volume indices of production in industry and construction and the building permits indicators show some volatility - even long after their first release.
When big countries revise their data, the European aggregates are also revised and this can be observed in the revision tracks of the European main aggregates. The major methodological changes, for example the change of classifications, the base year or the seasonal adjustment method understandably introduce revisions to the entire length of the concerned time series.