Slides presented at The Bank of England , London for the Economic Forum on Monday 21 October 2019 to provide delegates an understanding of economic statistics.
2. Agenda
09:15 – 10:00 Registration with tea and coffee
10:00 – 10:05 Welcome and introduction – Richard Heys, Deputy Chief Economist (ONS)
10:05 – 10:10 Opening Remarks – James Bell, Director of Monetary Analysis, Bank of England
10:10 – 10:25 Transforming of National Accounts – Rob Kent-Smith (ONS)
10:25 – 10:40 Comparison of ONS indicators – Henry Duquemin (ONS)
10:40 – 10:50 Question and answer session – Grant Fitzner, Chief Economist (ONS)
10:50 – 11:05 Refreshment break
11:05 – 11:20 Economic Growth and carbon Emissions – Amina Syed and Obinna Agbugba (ONS)
11:20 – 11:35 Country and Regional Gross Domestic Product (GDP) – James Scruton and Charlotte Richards (ONS)
11:35 – 11:50 SDGs:The goals have the economy in their targets – Fiona Dawe (ONS)
11:50 – 12:00 Question and answer session – Grant Fitzner, Chief Economist (ONS)
12:00 – 12:10 Closing remarks – Richard Heys Deputy Chief Economist (ONS)
6. The journey so far …
• Integration of VAT data
• New publication model for GDP moving to monthly GDP releases and 2
rather than 3 quarterly estimates
• Quarterly Regional GDP – more later!
• Blue Book 2019
7. A framework fit for the future
Nominal GDP Output
Input (of goods
& services)
output / production
approach
Real GDP Nominal GDP Prices
1 2
3
1
2
3
New data on the output of goods and services across the economy;
unprecedented detail on services
New data on the goods and services businesses use
New approaches to deflation (accounting for price change): we are going to improve
the consistency between real and nominal GDP by doing both stages of the
calculation at the same time.
Improving the
production
approach
Improving
Deflation
13. Blue Book 2020 – Emerging Priorities
1. International Comparability:
Improvements outlined following in-depth review of the methodologies
applied during the compilation of the accounts
2. Meeting Domestic User needs:
Further improvements to the Trade dataset
Revisions to data in the financial account for Bonds & Loans
3. Continuous Improvement:
Producer Price Indices (PPI) annual chain-linking
Continuation of work to deliver Double Deflation
14. Enabling double deflation
More work to do:
• Work to improve congruence of the microdata sources
• Continuation of our deflator development programme
• Undertaking an evaluation of our deflator set pre 2008
• A review of the level of detail we use to compile GDP.
15. Double Deflation Roadmap
2020: Experimental estimates
2021: Integrate estimates into National Accounts
Beyond 2021: Further improvement
But – time series and industry detail to be research led
16. Next steps
• Outline Blue Book 2020 scope - later this year
• Share our research on double deflation, establishing likely time series
and level of granularity - first half of 2020
• Publish experimental double deflated estimates - second half of 2020
17. Henry Duquemin
Office for National Statistics
Comparing ONS’s
economic data with IHS
Markit and CIPS UK
Purchasing Managers'
Index surveys
21 October 2019
18. Why are we comparing ONS data with IHS
Markit’s Purchasing Managers’ Index (PMI)?
•PMIs are closely watched indicators that are published in advance of official estimates
of growth
• Therefore it is important to understand how PMIs relate to official estimates
•PMIs have total sample size approximately 1,500, ONS sample is approximately
41,000
•PMIs are diffusion indexes, these combine movements in the underlying data to
provide a single indicator which can then be monitored and tracked over time
•We explore some of the strengths and weaknesses of diffusion indexes as a method
and how they relate to official estimates
•To help accomplish this we construct a diffusion index from ONS data
19. Overview of presentation
•We make comparisons between three different series
• IHS Markit PMIs
• ONS official estimates of output
• Diffusion indexes constructed from data from the Monthly Business Survey (MBS)
•These comparisons look at ONS data over three periodicities
• Month on month changes
• Three-month on three-month changes
• Three-month on year ago changes
•Assess the performance of the different indicators during economic shocks
20. Constructing a monthly business survey
based diffusion index
•ONS sector coverage is adjusted so that it matches the coverage of Markit
as closely as possible
•Percentage change of turnover for each business in the MBS is calculated
•This percentage change is compared with a threshold value (±5%) to
determine whether the business is classified as having increased, the
same or decreased revenue
•The diffusion index is then calculated using:
I = (1× PU ) + (0.5× PS ) + (0 × PD)
21. Measuring correlation between series
•We need a way to objectively assess the strength of the relationship between
series
•Most of the time series in this analysis are non-stationary
•This makes traditional methods of assessing the level of correlation (like Pearson
coefficient) inappropriate
•Instead we use detrended cross-correlation analysis (DCCA) which is accurate
even for non-stationary series
•DCCA allows the correlation to be investigated over periods of different lengths
•Produces a graph of the DCCA correlation coefficient vs time scale
23. Three-month on three-month a year ago index of
services growths, services MBS diffusion index and
services PMI
24. DCCA coefficient between annual three-months
MBS DI and PMI
•Significant correlation between
PMIs and diffusion indices based
on MBS data for GDP and all three
individual sectors
•Correlations becomes significant
when 4-8 months or more of data is
used depending on the sector
25. DCCA coefficient between annual three-months
official estimates of growth and PMI
•Both PMIs and diffusion indices based on
MBS show significant correlation with
official estimates of growth for:
• All sector/GDP
• Manufacturing sector
• Construction sector
•No significant correlations between either
diffusion index with official estimates of
growth for services sector
•7/8 comparisons between series based on
ONS data and PMIs show significant
correlation
27. Correlation between PMI and three-month on three-
month MBS diffusion indices
•Significant correlation between PMIs
and diffusion indices based on MBS
data for:
• All sector/GDP
• Services sector
• Construction sector
•No significant correlation for
Manufacturing sector
28. CORRELATION BETWEEN OFFICIAL ESTIMATES OF
THREE-MONTH ON THREE-MONTH GROWTHS AND
DIFFUSION INDICES
•Significant correlation between PMIs and official estimates of growth for all
sector/GDP and manufacturing sector
•No significant correlation between PMIs and official estimates of growth for
services and construction sector
•Significant correlation between MBS based diffusion indexes and official estimates
of growth for all three individual sectors and GDP
•Significant relationships exist between ONS three-month on three-month data and
PMIs, but the relationship is not as strong or extensive as the one between PMIs
and ONS three-month on three-month a year ago data with only 5/8 pairs of series
between the two significantly correlated
29. Month on month GDP growths, all sector MBS
diffusion index and all sector PMI
30. Correlation between official estimates of growth
and diffusion indices for month on month data
•No significant, sustained correlation between PMIs and ONS MBS diffusion
indices
•No significant correlation between PMIs and official estimates of month on month
growth
•Significant correlation between MBS based diffusion index and ONS official
estimates of growth for services, construction and GDP but not for manufacturing
•In this case the manufacturing sector is an example of official estimates of growth
and diffusion indexes sharing the same base data but still having no significant
correlation
31. Performance during economic shocks
•Previous work in 2012 comparing ONS data with Purchasing Managers Index
(PMI) has noted that: “with unexpected and prolonged shocks, the Markit/CIPS
trend exaggerates the magnitude; but with expected/short ‘shocks’, Markit/CIPS
under-estimates”
•This work did not include the performance of an ONS diffusion index during these
shocks
•We look at one example of an “expected” economic shock and one example of an
unexpected
32. Queens diamond jubilee – June 2012
Sector PMI (Standard
deviation from
mean)
MoM ONS
diffusion index
(Standard
deviation from
mean)
MoM ONS
headline
growths
(Standard
deviation from
mean)
Services -1.2 -1.6 -3.7
Manufacturing -1.5 -1.9 -2.9
Construction -1.4 -2.7 -2.8
GDP (all sector) -1.4 -1.2 -4.2
33. Aftermath of the EU referendum result – July
2016
Sector PMI (Standard
deviation from
mean)
MoM ONS
diffusion index
(Standard
deviation from
mean)
MoM ONS
headline
growths
(Standard
deviation from
mean)
Services -2.6 -1.1 0.27
Manufacturing -1.6 -1.3 -0.58
Construction -1.9 -1.5 0.62
GDP (all sector) -2.6 -1.1 0.32
34. Conclusions
•Purchasing Managers Index (PMI) show the strongest relationship with ONS annual
three-month data, with significant and sustained correlations found between the two
data sets
•Despite asking about month to month variations, PMIs have no significant correlation
with ONS month on month MBS based diffusion indices or official estimates of growth
•This may suggest that PMI respondents take a wider and longer term view of business
conditions than just month to month variations as well as allowing for seasonal and
other distorting factors
•Even when official estimates and diffusion indices share the same base data, the
difference in methodologies can produce two series with no significant correlation
•The MBS diffusion indices show similar tendencies to over/underestimate during
economic shocks that is seen in the PMIs
37. Dr Amina Syed
Economic Adviser
George Agbugba
Assistant Economist
The decoupling of economic
growth from carbon
emissions: UK evidence
21st October 2019
38. Contents
The relationship between environment and the economy
UK’s structural change
Technical progress and improvements in energy efficiency
International trade of carbon emissions
Conclusion
40. Environmental Kuznets Curve
The environmental Kuznets curve suggests that economic
development initially leads to a deterioration in the environment
But after a certain level of economic growth, a society begins
to improve its relationship with the environment and levels of
environmental degradation reduces.
52. Source: (Eora, 2019); (WRI,2019) and (BEIS,2019)
Different measures of CO2 emissions, 1970 to 2015, UK
0
100
200
300
400
500
600
700
800
900
1000
(MillionTonnes)
Consumption-based emissions Territorial-based emissions
53. -6
-4
-2
0
2
4
6
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Tonnes per capita
China Russia UK USA India Japan
Net trade of carbon emissions, per capita, selected countries and regions, 1992 to 2015
54. -120
-100
-80
-60
-40
-20
0
20
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Milliontonnes
China India Russia USA EU (Minus UK)
Net import of UK’s carbon emissions, by main countries and regions, 1990 to 2015
Source: (Eora, 2019)
55. Source: (World Bank, 2019) and (WRI,2017)
Correlation between real GDP per capita and CO2 emissions (tonnes) globally, 1960 to 2014
56. Conclusions
While UK CO2 emissions peaked in 1972, once imported emissions are considered, UK emissions
peaked in 2007.
The biggest source of these ‘imported’ emissions is China, followed by the EU.
Directly UK produced emissions declined due to a combination of environmental policies and a
shift of the UK economy from manufacturing to service-based industries.
UK’s directly produced emissions, continue to fall, with the energy generation (-67%) sectors
showing the biggest falls in emissions between 1990 and 2017.
The switch from coal to renewable energy has seen UK CO2 emissions continue to fall.
57. Country and Regional
Gross Domestic
Product (GDP)
Office for National Statistics
James Scruton and Charlotte Richards
59. To meet user demand for faster statistics to
monitor sub-national economic activity
Independent Review of Economic Statistics
by Professor Sir Charles Bean
• More timely regional economic statistics
• Greater granularity in regional economic statistics
• More use of administrative data sources
60. Already exist
• Scotland GDP
• Northern Ireland NICEI (~GDP)
• Welsh STOI (not GDP)
New coverage
• GDP for 9 English regions
• England GDP
• Wales GDP
Completing the set of quarterly
GDP measures across the UK
62. Data sources
Survey data (MBS) v Administrative data (VAT)
• Both collect turnover for a wide range of industries
• Monthly Business Survey is faster (UK results in 3 months)
• VAT has far wider coverage (c1.9 million records)
Other sources used for other industries (e.g. direct volume
measures, government expenditure etc.)
63. Methodological issues
Regional activity: RU v LU
• Reporting Unit (RU) classifies company activity by industry
• Local Unit (LU) classifies activity at each individual site
Regional apportionment using LU employment
Deflation using national prices
64. Consistency with other estimates
Regional Accounts – annual structural measure of GVA
• Benchmark historic time series to ensure consistency
UK monthly and quarterly GDP
• Constrain latest quarters to be consistent with UK GDP
Scottish and Northern Irish GDP
• Used in constraining methodology but unchanged by it
65. Publication plans
Experimental Statistics – so expect further changes
Same format as UK monthly GDP
• 20 industry sections (A to T)
• Broad sectors (Production, Construction, Services)
First release was on 5 September 2019 (2012Q1 to 2018Q4)
Next release in October (2019Q1)
• Then quarterly thereafter (6 months after end of quarter)
66. Future developments
• User consultation – does this meet your needs?
• Fine-tune our methods (e.g. constraining)
• Improvements to data sources (e.g. agriculture)
• Longer-term, combine with flexible geography project to
provide other areas users want
68. All countries: Qtr on Qtr % Real GDP growth
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2017 Q1 2017 Q2 2017 Q3 2017 Q4 2018 Q1 2018 Q2 2018 Q3 2018 Q4
England Wales Scotland Northern Ireland UK
69. -2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
North East North
West
Yorks &
Humber
East
Midlands
West
Midlands
East of
England
London South East South
West
Wales Scotland Northern
Ireland
Real GDP growth; Qtr on Qtr 1 year ago; Whole Economy
2018 Q1 2018 Q2 2018 Q3 2018 Q4
72. -1.0
0.0
1.0
2.0
3.0
4.0
5.0
North East North
West
Yorks &
Humber
East
Midlands
West
Midlands
East of
England
London South East South
West
Wales Scotland Northern
Ireland
Real GDP growth; Qtr on Qtr 1 year ago; Services
2018 Q1 2018 Q2 2018 Q3 2018 Q4
73. SDGs: The Goals have the economy in their targets
Fiona Dawe
Head of Sustainable Development Goals
@FionaDaweONS
90. ONS Economic Forum, on the Road:
24 October 2019, GLA, London
13 November 2019, Manchester
29 November 2019, Belfast
Special ONS Economic Forum:
Economic Statistics Transformation Programme - UK flow of funds (enhanced financial accounts)
26 November 2019, London
ons.gov.uk/economicevents
ONS Events
91. ONS Consultation – Human Capital
Reviewing how we measure Human Capital in the UK
Closing date: 18 November 2019
www.consultations.ons.gov.uk
92. The Economic Statistics Centre of Excellence (ESCoE) will hold its annual conference, organised in partnership with the UK Office
for National Statistics (ONS), at Kings’ Business School, King’s College London, 20-22 May 2020.
We invite submissions of papers on all aspects of the measurement and use of economic statistics, including: the productivity puzzle,
the digital economy, National Accounts and ‘Beyond GDP’, regional statistics, measurement using big data and administrative data,
international trade flows and the location of economic activity.
Please submit a full paper or an extended abstract (full papers are preferred) by 12 January 2020.
We also welcome proposals for Special Sessions of up to three papers. These should be sent to:
economicmeasurement2020@niesr.ac.uk, and not submitted through Conference Maker. Each delegate may be an author of more
than one paper, but each paper should have a different presenter.
Confirmed keynote speakers are:
Anil Arora (Statistics Canada)
John Van Reenen (London School of Economics)
Anna Vignoles (University of Cambridge)
ESCoE Conference on Economic Measurement 2020
Call for papers
www.escoe.ac.uk
Editor's Notes
on the whole of the Welsh economy to compliment the Northern Irish and Scottish figures.
But what does this mean for society?
Ambient anthropogenic air pollution – manmade air pollution
Remained relatively flat – despite the amount of CO2 emissions reducing over the same time period
But who is being effected? – regional? who?