Short-Term Output Indicators Stakeholders Group Meeting Bank of ...

quiltamusedData Management

Nov 20, 2013 (3 years and 8 months ago)

75 views

Short-Term Output Indicators Stakeholders Group Meeting

Bank of England (BoE)

Monday 23 September 2013, 10:30 – 12:30

Attendees:
Deputy Director Office for the Chief Economic Advisor (ONS)
Assistant Deputy Director Business Indicators and Balance of Payments (ONS)
Grade 7 Retail Sales and Construction Team (ONS)
Grade 7 Index of Production Team (ONS)
SRO GDP(O) and Index of Services Team (ONS)
SEO Analysis & Methods Team (ONS)
Manager UK Team Bank of England (BoE)
Assistant Economist Office for Budget Responsibility (OBR)
Assistant Economist Office for Budget Responsibility (OBR)
Assistant Economist HM Treasury (HMT)
Minutes Retail Sales and Construction Team (ONS)

Apologies:
Deputy Director Business Indicators and Balance of Payments (ONS)



Minutes from the last meeting

Previous minutes were agreed as being a true and accurate record of the meeting.

It was asked if the minutes could be circulated sooner after the meeting as the current sets were
only circulated a week before the meeting.

Action 1 ONS to circulate draft minutes of the meeting within two weeks

Actions from the last meeting

Action Point 2 – Putting the estimates of GDP into the October press release
ONS spoke to the central GDP team to discuss the logistics and cannot envisage any problems with
it. It is possible that this will happen for November.

Action Point 3 – Web Data Explorer Demonstration
Action Point 4 – CDID’s
The web data explorer will be included within the seminar about measures to improve the output
measure of GDP. The date has for this has been confirmed as 2
nd
October.

BOE asked if this is the launch of the new interface or just an initial look at what it will look like in
the future.

ONS confirmed that this is just an initial look and that the launch is pencilled in for October and will
initially be focused on population and social statistics. Retail Sales will be added in for November.

BOE asked would there be any implications for the information they normally received, and would
this be the only chance to give feedback on the new interface.

ONS informed the group that this action originated from whether CDID’s could be increased in the
short term, the corporate objective and strategic direction is to remove CDID’s, however this is
several years away.
This will be the first opportunity to see how the data explorer works and to understand how to
extract data from there. As ONS continues to build more datasets into the data explorer that is
when it can be looked into whether there is a possibility of taking away the times series data.
There will be no change in what is published in its existing form, what will change, from RSI in
November, is an extra facility to access the data that can already be accessed through times series
data. That will be the opportunity to give feedback.

The timeframe to turn on the CORD system for Construction will be October with the aim to publish
in November. This means CDID’s will be provided from CSDB for Construction from November
onwards.

Action Point 5 & 6 - re’ National Accounts Data
These issues are still live and will be discussed after the 2
nd
Quarter of National Accounts is published
on Thursday.

Action Point 7 – Improvement Timetable
ONS will send out an update once the methodology team have confirmed. This will be something
that will be covered in the seminar on the 2
nd
October.

Agenda Items
Review of Q2

It was asked if any questions could be sent out in advance of the meeting in the future.


Forward look at Q3 2013

ONS stated that there have been weather effects in July and still some Olympic effects in August.
The weather affected the RSI figures for August, when looking at long-term comparisons there is an
Olympic effect going on in RSI.

HMT noted that it was interesting to see quite a sharp fallback in August in Retail Sales, from an
erratic effect after July being strong due to the weather; they also noted that Food was prominent.

ONS said that the figures were almost all driven by food. Feedback given suggested that sales went
back to normal levels.
An article being published on Tuesday 24
th
September will help to explain some of the effects that
have been seen in the data. The paper looks at what has happened in previous heat waves and
compared this heat wave to that of 2006, which spanned 2 periods. The food sector seems to boom
and falls back as the weather goes back to normal. The data shows that this was the same for 2006
and 2013 and that it was a food effect and nothing else. The data has been compared against the
non-food sector and how it behaves in heat waves and it is almost exactly the same pattern as 2006.

Action 2 ONS to circulate the article to the group

ONS also noted that the heat wave also affected IOP for July as well. Feedback from electricity and
gas and the food and drink sectors suggest an increase.

HMT stated that water supply, sewerage, then gas, and electricity seem to offset each other during
periods of extreme temperatures.

ONS confirmed that gas and electricity do fall and water increases because of the heat, explaining
the balance as a result.

HMT asked if there was an effect on the manufacturing output.

ONS replied that feedback received suggested that food and drink had increased. There was a boost
in July. This also ties in closely with what has been seen on RSI.

OBR asked if there were any differences between values and volumes.

ONS replied that are no differences on the effect of the prices, but it is key to note that driving the
store price inflation within RSI is the food sector which is up 3.4% and the store price inflation
slowed from 1.9% to 1.6% in August, but the food sector was still holding that up. Within data
returned, there was a marked difference in value data being returned, this can be seen within the
publication. It was noticed within the distribution analysis that the split was not just in large or small
stores, but spread across the two, the 0-9 employment band showed an increase.

BOE asked if the erratic effect is quantified in the article.

ONS said that it would be impossible to say how much of the effect was due to the weather. What it
does relate back to is the anecdotal evidence from retailers that suggested sales went back to their
normal levels in August. The boost seen in July was the hot weather effect across the board.
Averaged across the two months you would see a slight rise, but still showing some good growth in
the retail sector. The BRC’s footfall survey very much ties in with what we saw in August.

BOE asked if this follows across to IOP data and how much of a fall is expected in August.

ONS said that electricity, gas, and water supply would balance each other out, but nothing in total
production, just an offset within water and utilities.

Retail Sales are broken down by MCI, which is dominated by supermarkets. That showed that we
were all eating less, buying less clothes and household goods. There was no one category that
changed pattern, across supermarkets it was all goods that had decreased. Feedback in July saw
that all goods were boosting sales; barbecues, barbecue food and garden furniture. In July,
everything went up and in contrast, in August, everything has fallen.
This is not in the article but ONS can produce a summary of this it it is required.

HMT were happy to see Output and New Orders together in the last publication, it raised the profile
of the New Orders Survey. How much of the rise in construction figures is coming from house
building?

ONS replied that there was not a great deal of feedback, but it has been noticed that there is some
strength in the numbers. The figures were not as strong in June, there was a big difference in the
increase compared to July, and big changes were seen in other sectors.
There was a boost in figures within the repair and maintenance sector and its difficult to quantify
whether it was that impact or not. This is where it is quite helpful to have the New Orders data
along side Construction so that you can see an increase in new orders and new housing. The lag
between housing new orders is far smaller than the lag between new orders in construction. These
figures suggest that there is more new housing being seen and the Trade Press confirmed that new
housing was picking up.

Within Retail sales, there has been a stronger DIY sector, less strong in the furniture side, which
suggests there is work going through within there, which backs up the increase in repair and
maintenance within Construction. A lot of the shopping that is being done is being done in London.
Therefore, there was speculation that there was a ‘London Factor’. It may be worth looking back at
the regional data to see it is a regional impact rather than a GB impact.

HMT said that there is a sense that from New Orders data there is an underlying improvement in
housing. Infrastructure is a volatile series that is driven by something that comes out the following
quarter, but this is not seen within housing, there is no special or big number of orders.

ONS reported that the increase in infrastructure was from wind turbines and solar energy, so that
will be a while before those figures are seen in the output data. This does need to be treated quite
carefully as it was a big governing factor in the overall increase. In addition, it is key to stress that
this was a new data source from Barbour ABI, which was a far more complete data source than
previously published, but looking at comparison data that ONS tested against Barbour ABI there was
very little difference.

HMT inquired as to why the tables are still being sent out in the publication with 2005 prices.

Currently only the lock-in journalists are still interested in this data for their own templates, and they
were to be streamlined and removed. If the group do not want the 2005 data anymore then we can
remove it and publish only CVM data, this could tie in with the November changes.

Action 3 ONS to send article on house building, whether it is a London or GB
factor
Action 4 Stakeholders to send a request to ONS to discontinue the 2005 data

HMT asked about the DECC forecasts. There are presumptions of normal outputs being distorted by
other mining and quarrying. This is moving away from where we would expect, is that coming to an
end or was is more of an erratic thing?

ONS responded that during April and May the oil industry was picking up, then in June, there was a
dip and this was due to two plants shutting down. Other mining and quarrying did increase during
April and May as did associated gas production, but this levelled out in July. The industry is volatile.

BOE asked if there was any impact from the helicopter crash and the surrounding issues.

ONS have not had that level of information from DECC. There was a meeting with DECC and it was
asked if it was possible to have feedback with the pattern we have seen in the industry.


Construction Development Programme

ONS told the group that in recent months they have had two new data suppliers, Barbour ABI for
New Orders and Davies Langdon.
Davies Langdon has taken over the contract for supplying price deflators for output in the
construction industry from the Building Cost Information Service (BCIS). BCIS are still producing the
prices under a subcontractor agreement while work is continuing with Davies Langdon to ensure
that better tender price indices are produced for the construction industry.
ONS said they should be able to influence what their preferred methodology for different parts of
the construction industry would be in terms of measuring prices. As it is, it is quite easy to measure
the output price for a house but it is less easy to do it for infrastructure. Industry expert Jim Meikle
is working with Davies Langdon is to look at models for this in the future.
There should be a complete new set of deflators by the next blue book publication for output in the
construction industry.

BOE asked if this would cause more revisions for blue book in 2014.

ONS said that there could possibly be revisions as there will be substantial methods changes,
however, they do not know what the effect on prices will be. Initial investigations have found some
of the deflators have very low samples and it is unclear how accurate they are. ONS do not have
control of those deflators as they are sourced in from BIS. ONS are working with BIS to try to
improve that and create new deflators that work for them.

HMT asked what aspect of the deflators is not working.

ONS replied that the sample sizes are not working as the basket of goods that we use to price for
output has not been changed for a considerable number of years, so new technologies have not
been brought on board. Current deflators are a good quality but they could be better and that is
what is being worked towards.

HMT asked if it is possible to set out what aspects of the construction price indices feed into GFCF.

Action 5 ONS to send out document detailing which aspects of the
construction price indices feed into GFCF

BOE asked if some information could be set out regarding where the data is coming from and what
the shortcomings could be with it and feeding into how this is going to improve it. There is a lot
attention on the GFCF building component and people will ask if this will push it in the right
direction. That way we can get a sense of how big the changes are likely to be.

ONS said, that there isn’t a great deal of information that BCIS have been willing to publish about the
methodology but they can find out what they have and what Davies Langdon are willing to share and
they will put it in the public domain.

BOE asked what the implications of this for GDP are.

There will be implications perhaps on the output side, but it these will not be known until the new
methodology is in place and how it will look.
Davies Langdon are very keen that when the new price inflators are produced, they are very
considerate of National Accounts needs. Therefore, National Accounts needs and objectives are the
first to be considered in the development.

The steering group meeting being held on Monday 30
th
September is about ensuring that the
construction industry are engaged with the changes we plan to make. A lot of work has been done
to produce initial papers shortly after that meeting and ONS are hopeful that by the middle of
October they will have a much better understanding of what is planned to do through published
information rather than just being able to talk about it.

ONS are currently working towards June to be able to provide and idea of quantification. They want
to test on at least one book worth of data, so for September’s blue book they will be looking to test
in June and hope to be able to backdate everything in the test as well. Testing will just take place on
the construction output side either, it will test across the whole piece, and National Accounts are on
board with this as well.

Action 6 ONS to produce notes on what series are affected and to extend this
to other parts of National Accounts so that there is a picture for the
3 measures of GDP.

HMT asked have there been any new developments on Seasonal Adjustment.

From June, the quarters are now being formed for the new seasonally adjusted series. There have
been no big fluctuations yet, but it is a new series in a volatile industry. There will be a normal
annual seasonal adjustment review unless we can see that the data is not looking good, and then it
will be reviewed earlier.

Property Developers

ONS are undertaking an industry review of real estate. They have been looking at the coverage and
at the moment property developer growth is not recorded separately. There is a note to this effect
in the bulletin. This will be talked about in the seminar on the 2
nd
October.


Communication of uncertainty within the short term estimates

Within Retail Sales, we publish a few paragraphs on standard errors in the background notes. It has
been highlighted that this may not be the most transparent text we have anywhere in the bulletin.
What we currently have is much geared to statistical users than any other user.

The background notes are not greatly accessed and it takes a long time to produce the standard
errors. We have to find a way to present this information so that it satisfies the needs of all users.

ONS would like feedback from the group regarding this text. Are they happy with how it is being
published or would they like to see something different?

Some initial feedback from HMT & OBR was that they were aware the standard errors were at the
back of the bulletin but they did not necessarily look at them unless there was an issue with them.

ONS said that they were thinking about making them more prominent in the bulletin. There have
been discussions about what figures to lead on within the bulleting and it is not always appropriate
to have the monthly indicator as a headline figure as it is too volatile. This is the same for all
monthly short-term outputs. Should we report on it more upfront in the bulletin and not in the key
points or would a better option be to adopt fan charts so that you can see how the components fit
together?

BOE stated that it depends on what is trying to be achieved, as there are bits of useful information
you could want to get across,
1 – Which survey, should I, as an analyst place more weight on
2 – How has that weight changed over time?
3 – What is news worthy and what is not?
These 3 things are to be communicated with this sort of analysis. Confidence intervals do not always
help because they are so wide and there is not a time series, it is impossible to get a sense across the
numbers of which ones you should place more weight on.

ONS agreed that this is about getting the right level of communication, and are considering the right
way of doing it. Different options of presenting this paragraph are being explored with in the
background notes.

Framework Regulation for Integrated Business Statistics

FRIBS is the Framework Regulation for Integrated Business Statistics.
This is being discussed in Eurostat now and this is just a short overview to raise awareness of it.
Some of it will involve changes in short term statistics and the way we use statistical units.
We will have to provide absolute values as well as indices, for the Short Term Indicators. One impact
is an increase in detail for wholesale and retail. Non Seasonally Adjusted, Calendar Adjusted and
Seasonally Adjusted estimates will have to be produced.
This has not gone through regulation yet but we are looking at somewhere between 2016 and 2018.
It will be worked on over the next few years.

GDP(O) Improvement Seminar

The date of the seminar has been confirmed for the 2
nd
October at Church House, with a 1pm start
and finish around 5pm. The final agenda will be sent out Thursday 26
th
September.

The agenda will include a section on strategy; the National Accounts work plan and revisions
performance. There will be another section on cross cutting issues such as un-registered businesses,
e-commerce and a construction review.
There will also be sections on industry reviews, wider improvements and experimental statistics.
There is a new policy in place on experimental statistics and we are hopeful that we can remove the
experimental label on a number of industries on IOS.
The day will finish off with a session on web data access.

ONS are hoping to consult on recommendations for real estate in December, and then consult on
Health and Public Admin in spring 2014.
Property Developers will be dealt with separately and will be considered at the project board mid
October.
Water Transport, Social Care and Professional Services will also be taken forward and timelines for
these will be outlined at the seminar.

There is also a lot of wider work around rebasing 2010 happening. Also expanding current price data
for a number of industries on MBS and looking at VAT feasibility in a number of areas, there is a very
tight deadline to get these through blue book.
ONS will also look at elaborating on the VAT data in the New Year. ONS recently has access to better
quality VAT information and we are now able to link up between HMRC and the ONS IDBR. We now
are able to apportion VAT at reporting unit level and put VAT turnover data at industry level.


There are three potential uses of this information,
1. Firstly, as an indicator in its own right – there are some places within GDP(O) where the data
sources are not as strong as preferred. VAT could offer an opportunity to broaden the current
price estimates.
2. Secondly, as a benchmark. HMRC put estimates out for public use and these can sometimes be
misconstrued in GDP(O) but there is a benefit as a benchmark
3. Thirdly, is about augmenting the MBS sample over time. There is a possibility a different view
could be taken about how the MBS is constructed and how the mixture of VAT turnover data
and MBS survey data could produce new estimates that combine the two. We are very clear on
the governments desire to use more admin data and get compliance down on businesses, this
would be one of the key things we could to achieve that. There are advisors in Eurostat advising
us on how that could be achieved.

The methodology surrounding productivity adjustments will be looked at. This is an adjustment we
do that uses wages or employment data, which should come out some time in 2014.

BOE asked how big a deal the productivity adjustment is as for example, there would be a big
implication if pre-prices productivity estimates were still being used in the latest data.

ONS are unclear as to how big they would be. The breadth of industries affected was outlined in the
article published. For a number of industries we used employment data, our strategy is to remove
that data, as we are very keen to expand current price in MBS and VAT. We also use wage data and
again this has a big impact on productivity, this work has not been done yet but it has been
identified as a key area and we are aware of some of the weaknesses of using that data. We hope to
be able to do work on this later in 2014.

Action 7 ONS to send details of the GDP(O) seminar to the BOE

Any Other Business

Some short-term inquiries are going through the Statistics Authority re-assessment, IOP, IOS, RSI,
Construction, including New Orders and Preliminary Estimates of GDP. Preliminary GDP was covered
in National Accounts last time, this time it will be included within the short-term indicators.
Emails will shortly be being received from the Statistics Authority asking for feedback. These will be
progressing over the next few months and finalising early next year. These are routine re-
assessments and will be done every few years. The Quarterly and Annual National Accounts will be
assessed in the future too.

HMT said that Rob Doody did some adhoc analysis request on an indices level detail current constant
price in £m which is updated on the website regularly. Some people have asked how the analysis
can be used. What is the ONS position on it and how does it fit in with the Short-term Output
Indicators. It is consistent with balanced annual data, what can we infer from this?

ONS replied that it is a difficult area because the indices published for output are not balanced, the
£m figures are constrained to the balanced GVA figures once they have been through the National
Accounts process. There has been work looking at nominal GDP and how to do that on the output
side as well. It is still at an early stage, currently the unbalanced output are the best look at what the
industry is doing.
The £m figures are generated for a Eurostat requirement and are not published and are only
produced as an adhoc analysis. These may not be the best estimates of the industries because of
the effects of National Accounts balancing. They will not be vastly different because output leads
GDP balancing, but there will be a small discrepancy. These are released during month 3 and then
there is a second estimate and will be revised after Q&A.

HMT currently use blue book data to talk about the contribution of various industries in the
economy in certain years, best practice is talk about the financial services. For example, the financial
services sector is £xbn and you look at the annual figure, but the financial services industry is falling
and the question being asked is how much is the financial services industry in £bn.
What would be the advice on this be?

In terms of annual data, up to and including the last supply use balance sheet of 2011, it is fully
balanced and there are no discrepancies. This is the best data to use to look at a percentage of GVA
from any given industry. The balancing process is not as detailed for the short-term indicators as it is
for the annual data and the best advice would be to use the annual 2011 data.

OBR said that it is important to clarify this position as some users are using it a lot and it can be quite
confusing.

Action 8 ONS to produce a note as to what the figures are and what they
should be used for so that users can understand what they are




Date of next meeting -
















Action Points
No.

Responsibility

Action

1 ONS ONS to circulate draft minutes of the meeting within two weeks

2 ONS ONS to circulate the article to the group

3 ONS ONS to send article on house building, whether it’s a London or GB factor

4

Stakeholders

Stakeholders to send a request to ONS to discontinue the 2005 data


5 ONS ONS to send out a document detailing which aspects of the construction
price indices feed into GFCF
6 ONS ONS to produce notes on what series are affected by the new price
inflators, and to extend this to other parts of National Accounts so that
there is a picture for the 3 measures of GDP.
7 ONS ONS to send details of the GDP(O) seminar to the BOE
8 ONS ONS to put together a note as to what the figures are and what they
should be used for so that users can understand