Академический Документы
Профессиональный Документы
Культура Документы
Bulletin No. 8
N J Bennett
PAGE
NO.
List of Tables and Figures iv
Acknowledgements v
Abstract 1
Introduction 2
Background 3
Conclusions 13
References 15
iii
LIST OF TABLES AND FIGURES
iv
ACKNOWLEDGEMENTS
The author would like to recognise the assistance provided by Linden Holder
Simon Lott, Nick Sibery and Laura Gadsby of DASA (Economic Statistics)
with the supporting analysis contained within the Annexes, the preparation of
many of the diagrams and for peer reviewing the main document.
I am also grateful for the comments made on an earlier draft of this note by
Neil Davies and Tony Turner, Director and Principal Consultant respectively,
within the DASA Directorate for Economic Statistics and Advice.
v
ABSTRACT
There have been significant problems with the definition, consistency and
coverage of defence related exports data over recent years. DASA has
endeavoured to ensure that defence exports statistics are tackled on a cross-
governmental basis. The Defence Trade Statistics Working Group (DTSWG)
with representatives from across the relevant Government departments was
convened in order to review the quality of the data, the issue of trade
suppressions and to ensure presentational consistency across Government.
Recently a number of key developments have emerged which have had (or
have the potential to have) a significant impact on the quality of these
statistics. These are:
In this bulletin we first set out some of the history behind the methodology
used to construct the defence exports statistics published in UKDS. We then
examine the potential to improve the quality of the export data and provide an
update of progress on recent developments in this area. Some conclusions
regarding the future production of defence export statistics are aired.
1
UK Trade & Investment is the Government organisation that supports companies in the UK
doing business internationally and overseas enterprises seeking to set up or expand in the
UK. It is part of BERR.
2
BERR was formerly known as the Department for Trade and Industry (DTI)
1
Recent Developments relating to the improvement of Defence
related Exports Statistics
Introduction
2
b. Large increases in the aerospace sales reported by the Society of
British Aerospace Companies (SBAC). In collating the 2005 survey
SBAC discovered that their turnover and new orders data had been
overstated in 2005. Changes in the survey format resulted in some new
order and sales data from outside the UK being incorrectly attributed to
the UK. SBAC has amended their results and conducted a thorough
review of the 2006 data. They are confident that the aerospace survey
provides the most accurate and robust picture of the UK aerospace
industry.
4. In this bulletin we first set out some of the history behind the
methodology used to construct the defence exports statistics published
in UKDS. We then examine the potential to improve the quality of the
export data and provide an update of progress on recent development
in this area. Some conclusions regarding the future production of
defence export statistics are aired.
Background
a. HMRC Her Majestys Revenue & Customs. HMRC gather data from
the Single Administrative Document (SAD) which is completed by
exporters for exports of goods outside the EU (European Union), and
from returns from (VAT registered) traders for dispatches of goods (via
the Intrastat survey) within the EU. The level of HMRC data capture
has changed with time. With the introduction of Intrastat and new EU
commodity codes in the early 1990s (see DASA Defence Statistics
Bulletin No.4, dated July 1994) the level of detail available for capture
has diminished. This has continued recently with the creation of six
new dual use aerospace codes (from 2006) and the amalgamation of a
3
further three. The largest change in terms of the value of deliveries has
occurred to the new 88033000 code (more on this later).
Within this document, for ease of exposition we shall use the term
exports to indicate the aggregation of true exports and dispatches.
4
Radios, electronic items and vehicles also may be dual use: military
exports of these are difficult to identify.
c. The returns rely to a large extent on the accuracy of the data that are
entered. HMRC perform quality assurance checks to improve the
quality of the data.
Recent Developments
11. Comparisons between the 2005 and 2006 datasets revealed a very
large increase in the aggregate export deliveries total of some 670M
(1391M in 2005 to 2059M in 2006 See Annex C). Allowing for the
compensating yearly fluctuations recorded against the other tariff
codes in the dataset, by far the largest increase was recorded in a
single tariff code - the newly amalgamated code 88033000. This
code is one of several used to construct the Military Aircraft & Parts
category in UKDS Table 1.13. Export deliveries in this dual use code
alone had increased by some 750M between 2005 and 2006 (509M
3
http://www.dasa.mod.uk/natstats/quality/NSQR_Series_Report_32_Report_Part_3.pdf
5
under the old 88033090 code to 1260M against the new 88033000
code. The effect of the increase in trade reported against 88033000 is
clear when comparing the aggregate numbers under the Military
Aircraft & Parts category at Annex C.
13. The challenge then was to determine how much of the 750M increase
recorded under the single tariff code (88033000) constituted a real
increase in military trade and how much was in fact civil and was being
erroneously included under the new code.
14. Following the statistical tests briefly outlined at para 12, the following
validation exercises took place:
c. Where it was clearly evident from the time series data that most of the
trade had historically been recorded as civil, the values now recorded
under that company and/or country were abated by that element which
was previously recorded to be civil. In addition if a company was
deemed to be a wholly civil exporter, the value of export trade was
removed completely.
6
d. Further validation was then carried out by comparing the level of
defence orders (surveyed and collected by the DESO) over time with
that recorded as a delivery in 2006 under 88033000. Some simplifying
assumptions had to be made concerning the lag between orders and
deliveries, the corporate ownership of the companies concerned, and
in the coverage of the datasets. For instance, the services and
government to government sales (e.g. defence export orders relating to
Al Yamamah4) elements of the orders data were removed in order to
improve the comparability between the two datasets. In practice
though, a direct comparison between orders and deliveries is extremely
difficult, particularly given the differing methods of product classification
employed by the DESO and HMRC.
e. DASA also sought to determine from the HMRC data capture systems
if there were any other fields/variables which we could employ to
further distinguish the military/civil nature of the export (e.g. such as
product type, end user or license indicator). HMRC confirmed that this
was not the case and that although licence data were collected for
exports, it would be very resource intensive to reconcile the deliveries
data with licenses as both were collected and analysed on two
separate systems5. Also, the application of a BERR export licence
does not necessarily mean that the goods in question are for military
use. Therefore the quality of employing licences as an indicator of
military trade is not certain.
4
Al Yamamah (Arabic for The Dove) is a major government to government arms deal,
signed in 1985, for the supply of aircraft, support services, equipment, weapons and
electronic warfare systems by Britain to the Saudi Royal Air force.
5
The feasibility of using export license data is currently being explored following recent
advice from HMRC.
7
available. For the remaining 6 companies, either the value of export
trade was removed completely due to civil nature of the exporter (e.g. a
commercial airline or holiday company) or was not adjusted due to the
inconclusive nature of the supporting data. The companies where the
data were inconclusive were small in value.
f. For all other countries recording export deliveries under 88033000 tariff
codes, the aggregate civil/military abatement ratio (44:56) was applied
across the total value recorded under each country.
Split by Commodity:
Split by destination:
8
b. The country level breakdowns which were published as Table 4.4 in
ARSEC 2006 should be revised to the numbers shown in Annex D.
c. The current set of adjustments draws heavily on time series data which
ceased after 2005. This approach is, therefore, not sustainable for
more than a couple of years. The development of alternative methods
for positively identifying defence export trade needs to be faced
quickly.
Table 2: Estimates of Total Export Deliveries and Orders: Defence Equipment and Services
9
(2005) that we see a spike and the HMRC plus SBAC additions figures
(driven by the surge in sales data) diverge from the orders data which
ultimately led to a breakdown in this loose relationship.
8000
7000
6000
5000
M
4000
3000
2000
1000
0
2000 2001 2002 2003 2004 2005
Year
Graph 1: Relationship between HMRC deliveries (plus SBAC additions) and the DESO
orders data
21. DASA and DESO staffs met to discuss the apparent disparity between
the exports orders and sales data. DESO had identified a similar surge
in SBAC orders data when compared to the data they collect in their
survey of industry. In collating the 2005 survey SBAC discovered that
their turnover and new orders data had been overstated. Changes to
the survey format resulted in some new orders and sales data from
outside the UK being incorrectly attributed to the UK. This had
occurred where sales/orders won by the overseas subsidiaries of
certain defence suppliers had been counted as a UK export by the UK
based head offices of these suppliers. SBAC has now amended the
2005 results and conducted a thorough review of 2006 data. They are
confident that the aerospace survey provides the most accurate and
robust picture of the UK aerospace industry. Table 3 details how the
revised sales figures have resulted in a major adjustment to the 2005
estimates of total export activity contained with Table 1.14 of UKDS
2007.
10
Table 3: Estimates of Total Export Deliveries and Orders: Defence Equipment and Services
22. HMRC advised the DTSWG, early in 2004, that they were to undertake
a mini review of defence trade statistics (to validate the data they
provide). Pending the outcome of this review, the exports and imports
figures reported in UKDS were de-badged as National Statistics as it
was considered that their quality could not be endorsed whilst the
ongoing data quality problems persisted.
24. Given that the HMRC pilot survey had limited success, the remaining
options open to us narrowed. We have also explored the option to
expand the scope of questions contained in the current survey of
defence suppliers administered by the Defence Export Services
Organisation (DESO). Recent announcements concerning the transfer
of the DESO to the UKTI in the Department for Business Enterprise
and Regulatory Reform (BERR) are likely to have a further impact on
any firm plans to take this option forward though. Indeed, at a recent
meeting of the DTSWG (November 2007) it was agreed that
arrangements should be put in place to transfer the responsibilities for
producing defence trade statistics to BERR.
25. We, however, continue to strive to improve the quality and coverage of
the export statistics. Most recently, and with agreement from the
11
DTSWG, DASA have tasked HMRC with examining the potential for
analysing the value of export licenses alongside current methods used
for the data extraction. Where possible, we will continue to build on the
validation checks introduced during this years exercise.
26. At the EU level plans to simplify the system used to collect intra EU
trade are likely to have a further impact on the potential to collect
consistent and accurate trade data.
27. Intrastat is the system for collecting detailed trade in goods statistics
within the EU. It is a monthly survey, which in the UK covers around
35,000 VAT registered businesses and accounts for nearly half of the
statistical burdens placed on businesses. Previous attempts to reduce
the scale of the survey and the burdens on data providers have not
generated any radical reductions, with member states often supporting
their own requirements for detailed data rather than Commission
proposals for simplification. With the current focus on reprioritisation of
the European Statistical System, endorsed by the UK in its EU
presidency, there is renewed commitment and opportunity to deliver
changes.
29. Each has its advantages and disadvantages, which have been covered
extensively in recent studies7 and are not repeated here. Whereas
6
The Statistical Office of the European Communities (Eurostat) is the statistical arm of the
European Commission, producing data for the European Union and promoting harmonisation
of statistical methods across the member states of the European Union.
12
previously the disadvantages have prevented implementation, the
current climate requires a reform of Intrastat to be achieved and
recognises that some negative consequences will be inevitable. The
ECOFIN8 Council have recently agreed the Commissions proposal to
reduce the coverage threshold in the short term whilst seeking to move
to single flow in the longer-term.
Conclusions
30. In this bulletin we have explored some of the history behind the
methodology used to construct the defence exports statistics published
in UK Defence Statistics. We then examined the potential to improve
the quality of the current export data and provided an update of
progress on recent developments in this area.
13
transferring the responsibilities for producing defence trade statistics to
BERR. At the EU level, plans to simplify the system used to collect
intra EU trade flows are likely to present further difficulties for the
collection of consistent and accurate trade data.
14
REFERENCES
15
ANNEX A - Tables 1.13 & 1.14 UK Defence Statistics 2007
The data in this table are outside the scope of National Statistics because they do not meet all of the high professional quality
assurance standards set out in the National Statistics Code of Practice.
16
Table 1.14 Estimates of Total Export Deliveries and Orders: Defence Equipment and
Services
This table provides an estimate of total defence export activity relating to the UK. It uses data on additional aerospace equipment
and services from a survey undertaken by the Society of British Aerospace Companies. Aerospace services include training,
consultancy and project support related to the export activity. The second part of the table provides data on identified export orders
of defence equipment and services. This illustrates the relationship between export orders and actual export deliveries.
The data in this table are outside the scope of National Statistics because they do not meet all of the high professional quality
assurance standards set out in the National Statistics Code of Practice.
Identified Export Orders for Defence Equipment 4 970 5 540 4 737 4 160 5 041 4 882 4 546 3 989 5 527
and Services3
Split by Equipment Type:
Air Sector 3 456 3 193 3 501 3 245 3 553 3 526 3 119 2 491 4 133
Land Sector 535 656 616 341 509 303 475 584 670
Sea Sector 71 368 475 50 464 252 209 369 280
Not Specified 908 1 323 145 524 515 801 663 546 444
Sources: HM Revenue & Customs, Society of British Aerospace Companies, MOD Defence Export Services Organisation (DESO)
1
Changes to the internationally agreed tariff codes used for recording goods exports (see Annex C of the Annual Report on
Strategic Export Controls) had originally contributed to an apparent increase in the deliveries figures from 2005 to 2006. The
further amalgamation of military and civil codes which has resulted in a discontinuity in the code set used to compile these data has
been investigated. It was not clear how far the increase reported reflected a true increase in the value of military goods exported as
opposed to the inclusion of civil goods previously excluded. Further investigation of the HMRC dataset revealed a large element of
this reported increase in 2006 (recorded against one of the dual use codes) as being probably civil in nature. The 2006 figure has
therefore been revised downwards accordingly.
2
In collating this years survey SBAC discovered that their turnover and new orders data had been overstated in 2005. SBAC has
amended 2005 results and conducted a thorough review of 2006 data. They are confident that the aerospace survey provides the
most accurate and robust picture of the UK aerospace industry. Changes in the 2006 survey format resulted in some new order
and sales data from outside the UK being incorrectly attributed to the UK.
3
Figures for export orders are taken from the Defence Export Services Organisations survey of known Defence Contractors.
17
ANNEX B - Summary of HM Revenue and Customs Combined
Nomenclature (CN) Codes as published in ARSEC
Part 1 - Tariff codes used to compile data on the number of small arms and light
weapons
Code Description1
93011100 Artillery weapons (for example, guns howitzers and mortars): self propelled
93011900 Artillery weapons (for example guns howitzers and mortars): other
93012000 Rocket launchers; flame throwers; grenade launchers; torpedo tubes and
similar projectors
93019000 Military weapons, other than revolvers, pistols and the arms of heading 9307:
other etc
930200002 Revolvers and pistols, other than those of heading 9303 or 9304
Part 2 - Additional tariff codes used to compile data on the value of defence
exports
Code Description1
87100000 Tanks and other armoured fighting vehicles, motorised, whether or not fitted
with weapons, and parts of such vehicles
88021100 Helicopters: of an unladen weight not exceeding 2000kg
88021200 Helicopters: of an unladen weight exceeding 2000kg
88022000 Aeroplanes and other aircraft, of an unladen weight not exceeding 2000kg
88023000 Aeroplanes and other aircraft, of an unladen weight exceeding 2000kg but not
exceeding 15000kg
88024000 Aeroplanes and other aircraft, of an unladen weight exceeding 15000kg
880310003 Propellers and rotors and parts thereof
880320003 Under-carriages and parts thereof
880330003 Other parts of aeroplanes or helicopters
880510103 Aircraft launching gear and parts thereof: deck-arrestor or similar gear and
parts thereof: aircraft launching gear and parts thereof
880510903 Aircraft launching gear and parts thereof: deck-arrestor or similar gear and
parts thereof: other
88052100 Ground flying trainers and parts thereof: air combat simulators and parts
thereof
88052900 Ground flying trainers and parts thereof
89061000 Warships
93051000 Parts and accessories of articles of headings 9301 to 9304: of revolvers or
pistols
93059100 Parts and accessories of articles of headings 9301 to 9304: other: of military
weapons of heading 9301
93063010 Other cartridges and parts thereof: for revolvers and pistols of heading 9302
and for submachine guns of heading 9301
93063030 Other cartridges and parts thereof: for military weapons
93069010 Other [munitions and ammunition] for military purposes
93070000 Swords, cutlasses, bayonets, lances and similar arms and parts thereof and
scabbards and sheaths therefore.
Further information on classification is available in the Integrated Tariff and the ICN (available
free online at www.uktradeinfo.com).
Details on the compilation of overseas trade in goods statistics are available in GSS
(Government Statistical Service) Series No. 10: Statistics on the Trade in Goods, available
online from the ONS (Office for National Statistics) website.
1
Descriptions taken from IntraStat Classification Nomenclature
2
Information available for non-EU destinations only.
3
Dual use (military/civilian) codes. Information from Customs Procedure Code and
knowledge of trader used to apportion military trade element.
18
ANNEX C - Total Defence Export Deliveries: 2005 Vs 2006 Pre
Adjusted
Split by Commodity:
19
ANNEX D - Table showing revisions to 2006 ARSEC country breakdowns
plus differences from original dataset.
Table 4.4: Value of exports of military equipment from 1 January 2006 to 31 December 2006.
20
Mexico 0.48 0.40 -0.08 South Korea 17.10 9.21 -7.89
Moldova 0.01 0.01 -0.01 Spain 41.91 11.56 -30.35
Mongolia 0.20 0.15 -0.05 Sri Lanka 0.82 0.38 -0.44
Morocco 0.11 0.09 -0.02 Swaziland 0.00 0.00 -0.00
Mozambique 0.01 0.01 Nil Sweden 24.44 20.20 -4.24
Nepal 0.08 0.06 -0.03 Switzerland 79.72 67.43 -12.29
Netherlands 11.74 8.08 -3.66 Syria 0.09 0.07 -0.02
New Zealand 6.44 6.35 -0.09 Taiwan 1.37 0.84 -0.53
Nigeria 0.27 0.25 -0.02 Thailand 8.13 7.24 -0.89
Norway 37.28 26.02 -11.27 Togo 0.00 0.00 Nil
Oman 20.59 12.92 -7.68 Tokelau Islands 0.02 0.01 -0.01
Pakistan 27.58 27.25 -0.33 Tonga 0.00 0.00 Nil
Papua New Guinea 0.00 0.00 Nil Trinidad & Tobago 0.03 0.02 -0.01
Philippines 0.34 0.19 -0.15 Tunisia 0.34 0.25 -0.09
Poland 2.62 1.86 -0.76 Turkey 35.22 33.93 -1.29
Portugal 9.01 6.44 -2.57 Turkmenistan 0.04 0.02 -0.02
Puerto Rico 0.15 0.07 -0.08 Uganda 0.03 0.01 -0.01
Qatar 2.37 1.37 -1.00 Ukraine 0.15 0.11 -0.04
Romania 1.37 0.84 -0.53 United Arab Emirates 19.41 8.46 -10.94
Russia 0.21 0.16 -0.06 United States 610.77 333.61 -277.16
San Marino 0.00 0.00 Nil Uruguay 0.11 0.09 -0.01
Saudi Arabia 44.81 43.26 -1.55 Venezuela 0.18 0.18 Nil
Senegal 0.22 0.10 -0.12 Vietnam 0.02 0.01 -0.01
Serbia & Montenegro 0.02 0.01 -0.01 Yemen 0.13 0.08 -0.05
Seychelles 0.02 0.01 -0.01 Zambia 0.03 0.03 Nil
Singapore 68.22 9.40 -58.82 2,058.70 1,357.89 -700.81
Slovakia 0.22 0.21 -0.00 Nil indicates a nil value.
0.00 in Difference column indicates a difference of less
Slovenia 0.12 0.09 -0.03 than 5,000.
South Africa 10.92 5.25 -5.67
21