E-commerce: measuring, monitoring and gross domestic product

E-commerce: measuring, monitoring and gross domestic product
Jacqui Jones
Office for National Statistics
7 August 2014
Section 1: Introduction
Figure 1: E-Commerce as a percentage of UK turnover in 2008 and 2012
Source: Office for National Statistics
Did you know?
Data sources
There are several sources of official e-commerce estimates from the: annual e-commerce survey, the
annual Internet access questions (on the Opinions and Lifestyle survey), and the monthly retail sales
inquiry. E-commerce and Internet access estimates are also produced by other European Economic
Area (EEA) countries (see Section 3).
Internet access and internet purchases
In 2013 (the latest year for comparable country data):
UK households with Internet access was 9 percentage points higher (88%) than the EU27
average (79%); and
the percentage of UK adults who made an Internet purchase in the last 12 months was 30
percentage points higher (77%) than the EU27 average (47%).
(Note EU28 data are not yet available). (See Section 4.1)
Between 2010 and 2013, the UK was amongst the five European Economic Area (EEA) countries with
the highest percentage of individuals making an Internet purchase in the last 12 months (see Section
4.1).
1
Business e-commerce
In 2012, e-commerce accounted for 18% (£492 billion) of UK business turnover1. 21% of UK
businesses in 2012 made e-commerce sales to their own country; 9% of UK businesses made ecommerce sales to EU countries and 7% to non-EU countries.
In 2012, the percentage of large businesses (employing 1,000+ employees) making EDI sales (34%)
and website sales (47%) continued to be higher than for small businesses (employing 10 to 49
employees) making EDI sales (4%) and website sales (17%) (see Section 4.2).
Gross Domestic Product
There are three key points in relation to e-commerce from a measuring GDP perspective (see Section
5):
1. GDP is measured by three approaches: production, expenditure and income; and there are
numerous data sources that are used to measure the three approaches. Where e-commerce
is important is in providing information to inform the measurement of imports, especially in
relation to e-services such as digital downloads and applications.
2. To ensure that the value added of unregistered businesses is included in the measurement
of GDP, adjustments are made as part of the annual national accounts Supply and Use
balancing process. These adjustments are based on periodic analysis using additional data
from Her Majesty’s Revenue and Customs (HMRC).
3. ONS have an on-going programme of work to improve coverage of e-services exported and
imported by households. The first developments in this area will be improvements to online
gambling data in the UK National Accounts 2014 (Blue Book 2014).
There is a need to measure and monitor e-commerce from the perspectives of both policy and
measuring the economy; to do this you need to have source data. From a policy perspective there is
growing interest in measuring and monitoring domestic and cross-border e-commerce. In Europe
this interest is largely driven by the European Commission’s Digital Agenda. In relation to the
measurement of the economy the increasing transition to the purchase of e-services, such as digital
downloads and applications (apps) (many of which are imported) necessitates the need to
continuously review and improve data sources.
This paper looks at official sources of e-commerce data, estimates produced from these sources, and
e-commerce from a measuring Gross Domestic Product (GDP) perspective. E-commerce includes
electronic transactions over computer networks, e.g. electronic data interchange (EDI) and website
sales. This paper includes the following sections:
Section 2, which looks at the definition of e-commerce.
Section 3, which looks at e-commerce data sources.
Section 4, which looks at estimates of e-commerce.
Section 5, which looks at the measurement and compilation of GDP and where the
measurement of e-commerce is important.
Section 6, which provides some concluding comments to the paper.
1
For non-financial businesses with 10 or more employees excluding the following industries: agriculture,
mining and quarrying; finance and insurance; veterinary activities; public administration; education; health and
social work; arts and recreation; and other service activities.
2
A companion article to this one, entitled ‘Monitoring e-commerce’, was published on 7 August 2014
and proposes the use of an e-commerce indicator dashboard.
Section 2: Defining E-Commerce
Typically, when people talk about e-commerce (electronic commerce) they talk about website sales
and purchases but e-commerce is broader than this (see Figure 2) and includes any form of
electronic business transaction e.g. electronic data interchange (EDI). E-commerce can be carried
out: business to business; business to consumer; and consumer to consumer. EDI is typically carried
out business to business.
Figure 2: Organisation for Economic Co-operation and Development (OECD) e-commerce definition
E-commerce transactions are defined as ‘the sale or purchase of goods or services, conducted over
computer networks by methods specifically designed for the purpose of receiving or placing or
orders’. It is important to note, under this definition, that ‘the goods or services are ordered by
these methods, but the payment and the ultimate delivery of the goods or services do not have to
be conducted online’ (OECD, 2011, p. 72).
The statistical office of the EU (Eurostat) has a glossary in which EDI is defined as follows:
“EDI, is the exchange of data in electronic format, usually compatible between sender and receiver.
EDI offers businesses the opportunity to retrieve information electronically from their internal
systems and to send that information to trade partners/suppliers/customers/government through a
communications network. An example might be putting data from one type of database
management system into a sequential format and then moving the data to a second location where
they are stored in a format different from the original database management system.
Using EDI, business data is exchanged from one computer to another in a standard format. The
information is organised to allow a fully automated computer transaction that needs no human
intervention during the whole process. The information that is contained in an EDI document is the
same as that in a conventional hard copy (printed document)”.
E-commerce includes transactions of goods and services; these will include goods and services that
can also be bought and sold using other transaction modes such as telephone and face-to-face.
However, there are now a number of services available that are only available as e-services, such as
digital downloads, apps, and e-books.
It is also important to distinguish the difference between websites and the Internet in the context of
sales. The Internet is a global system of inter-connected computer networks, while a website is a set
of related web pages consisting of information. Internet and website sales are therefore different
concepts. For example, e-commerce transactions made using EDI over the Internet would be
excluded from website sales.
3
Section 3: E-commerce data sources
3.1
The annual e-commerce survey
Since 2000, a comparable annual e-commerce survey has been run in all European Union (EU)
countries and also in some non-EU countries to meet the requirements of the EU Regulation
808/2004. The e-commerce survey is designed to measure the extent of the use of Information
Communication Technologies (ICT) (i.e. computers, Internet connection, websites) by businesses,
and the extent of electronic trading by businesses (e.g. Internet trading). Eurostat plays a key role in
the annual e-commerce survey and each year leads a review of data requirements. This ensures
that the survey questions are updated on an annual basis to reflect changes and developments in
the use of ICT and e-commerce. However, these changes result in breaks in the times series, for
example, changes were made in the 2008 e-commerce survey, which mean that estimates prior to
2008 are not comparable with those produced since.
Annually, the UK e-commerce survey samples approximately 8,000 businesses from all industries,
apart from agriculture; mining and quarrying; finance and insurance, veterinary activities; public
administration; education; health and social work; arts and recreation; and other service activities.
Currently, businesses with fewer than 10 employees are excluded from the sample2. It is not possible
to estimate what the current e-commerce survey results would be if businesses with fewer than 10
employees were still covered in the survey. However, in 2004, the last year that businesses of this
size were included in the survey, ONS estimated that their value of Internet sales represented 6.6%
of the total value of Internet sales. In comparison with large businesses, businesses with fewer than
10 employees had lower rates of ICT adoption and usage.
In the UK, the results of the e-commerce survey are published as part of an annual ‘E-commerce and
ICT Activity’ statistical bulletin. Total e-commerce sales are estimated by aggregating sales received
over a website and sales received over EDI. The latest available estimates are for 2012, with the next
publication, for 2013 estimates, scheduled for November 2014.
3.2
Internet access sources
ONS has collected data on Internet access since 1998; Internet purchasing has been collected since
2001. The data are collected in response to questions included on the Opinions and Lifestyle Survey
(to produce annual estimates) and the Labour Force Survey (LFS) (to produce quarterly estimates3).
The Opinions and Lifestyle Survey achieves approximately 3,000 interviews and the LFS
approximately 41,000 interviews.
On an annual basis, comparable Internet access questions are included in surveys in all EU countries
and also in some non-EU countries. However, it is important to note that the majority of estimates
published by Eurostat are on a slightly different basis to the official estimates produced by ONS. This
is because in Eurostat’s Internet access estimates, households are only included if they contain at
least one resident in the 16 to 74 years old age group. In contrast, the ONS includes all households
and adults. This means that the ONS estimates are usually lower than the published Eurostat totals.
Eurostat also publish estimates that represent the whole of the UK (incorporating Northern Ireland
2
Until the 2004 survey, businesses with fewer than 10 employees were included in the survey; subsequently
the survey was refocused on the coverage required under the EC regulation as funding could not be prioritised
to maintain a wider coverage.
3
A consultation looking at future statistical releases in terms of internet access (August 2014) can be found on
the ONS consultation pages.
4
estimates since 2013 and prior to that estimating using UK weights); while some ONS estimates only
cover Great Britain (GB).
The latest available annual Internet access survey estimates are for 2014 (published on 7 August
2014).
3.3
The monthly Retail Sales Inquiry
In measuring the output of the retail sector, the value of retail sales are collected from GB registered
retailers in the ONS monthly Retail Sales Inquiry (RSI). Internet sales of retail goods have been
collected as part of the ONS monthly RSI since March 2008. The survey excludes services such as
insurance, ticket sales, and digital downloads.
The RSI samples approximately 5,000 retailers on a monthly basis. The sample includes all large
retailers and a representative sample of medium and small retailers. The RSI provides data to
comply with the European Short Term Statistics regulation.
The results of the survey are published monthly in the Retail Sales Index statistical bulletin.
3.4
Other e-commerce sources
Data on purchases of goods and services are included in responses to the ONS Living Costs & Food
Survey, where households are asked to keep a diary of their expenditure. Separate data on ecommerce purchases are not available but are included in the collected expenditure data for goods
and services.
Section 4: E-commerce estimates
This section looks at some of the estimates produced from the annual Internet access questions, the
annual e-commerce survey and the monthly retail sales inquiry. It should be noted that, as with any
estimates produced from sample surveys, there will inherently be uncertainty surrounding the
estimates caused by sampling and non-sampling error. To minimise these potential sources of error
ONS ensures that the most robust methods are used in the collection of data and production of
estimates.
The ‘Monitoring e-commerce’ article published alongside this article provides an interactive demo of
how the main e-commerce indicators might be presented.
4.1
Internet access and purchasing estimates
In 2014, 38 million adults in GB accessed the Internet every day, 22 million more than in 2006, when
directly comparable records began. In GB, 22 million households (84%) had Internet access in 2014,
with the vast majority making use of a broadband connection over a Digital Subscriber Line (DSL),
fibre optic or cable line.
Of the 4 million households without Internet access, the majority (53%) said that they did not have a
connection because they 'did not need it'.
5
Figure 3: Adults who accessed the Internet in 2014
Comparable 2014 estimates for other European Economic Area (EEA)4 countries are not currently
available but looking at the 2013 estimates they show that in 2013 (Figure 4), Iceland had the
highest percentage of households with Internet access (96%) and Turkey the lowest percentage of
households with Internet access (49%). Internet access in the UK (88%) was 9 percentage points
higher than the EU27 average (79%) (Note that EU28 data are not yet available).
4
The EEA includes Norway, Iceland and Lichtenstein, and 27 of 28 member states of the European Union (EU).
6
Figure 4: Percentage of households with Internet access and percentage of individuals who made an
Internet purchase in the last 12 months (2013)
Turkey
Bulgaria
Greece
Romania
Portugal
Lithuania
Cyprus
Croatia
Italy
Spain
Hungary
Latvia
Poland
Czech Republic
Slovenia
Slovakia
EU27
Malta
Estonia
Belgium
Austria
France
Ireland
Germany
United Kingdom
Finland
Sweden
Denmark
Norway
Luxembourg
Netherlands
Iceland
% of individuals who made an
Internet purchase in the last 12
months
% of households with Internet
access
0
20
40
60
80
100
120
Source: Eurostat
Notes: data for the Former Yugoslav Republic are not available.
Percentage of households who have Internet access at home. All forms of Internet use are included. The population
considered is aged 16 to 74. Countries are ranked based on proportion of households with Internet access.
Still looking at Figure 4, in 2013, the percentage of UK individuals who made an Internet purchase in
the last 12 months was 30 percentage points higher (77%) than the EU27 average (47%). In 2013,
individuals in the UK and Denmark made greater use of Internet purchasing than the other countries
shown in Figure 4.
7
Figure 5 shows, the percentage point increase, between 2010 and 2013, of individuals that made an
Internet purchase in the last 12 months. Here you can see that all countries had an estimated
increase but this varied from 15 percentage points in Latvia and Lithuania to 2 percentage points in
the Netherlands and Norway. Between 2010 and 2013, the UK had a 10 percentage point increase
in individuals that made an Internet purchase in the last 12 months. For comparison purposes Figure
6 shows the 2013 proportions of individuals that made an Internet purchase in the last 12 months, in
the same order as the percentage point chart (Figure 5).
Throughout the period from 2010 to 2013, the UK was amongst the five countries with the highest
percentage of individuals making an Internet purchase in the last 12 months, along with Norway,
Denmark and Sweden. Luxembourg replaced the Netherlands in the top five in 2012 and 2013.
Figure 5: Between 2010 and 2013, the
percentage point increase of individuals that
made an Internet purchase in the last 12
months
Norway
Netherlands
Poland
Romania
Italy
France
Finland
Estonia
Sweden
EU27
Cyprus
Bulgaria
Spain
Malta
Slovenia
Germany
Denmark
Czech Republic
United Kingdom
Portugal
Luxembourg
Ireland
Hungary
Belgium
Slovakia
Iceland
Croatia
Austria
Greece
Lithuania
Latvia
Figure 6: In 2013, the percentage of individuals
that made an Internet purchase in the last 12
months
percentage point
increase
between 2010
and 2013
0
5
10
15
20
Norway
Netherlands
Poland
Romania
Italy
France
Finland
Estonia
Sweden
EU27
Cyprus
Bulgaria
Spain
Malta
Slovenia
Germany
Denmark
Czech Republic
United Kingdom
Portugal
Luxembourg
Ireland
Hungary
Belgium
Slovakia
Iceland
Croatia
Austria
Greece
Lithuania
Latvia
2013
0
20
40
60
80
100
Source: Eurostat
8
Note: data for the Former Yugoslav Republic and Turkey are not available for all years and have therefore been excluded.
In Figure 5, countries are ranked based on their percentage point increase between 2010 and 2013. For comparative
purposes Figure 6 shows the percentage of individuals that made an Internet purchase in the last 12 months, in the same
country order as Figure 5.
4.2
E-commerce survey estimates
In 2012, the estimated value of e-commerce transactions, for non-financial businesses with 10 or
more employees, was £492 billion; this represented 18% of UK turnover (turnover as measured by
the ONS Annual Business Survey) (see Figure 7).
Figure 7: E-commerce sales in 2012
Source: Office for National Statistics
Figure 8 shows the percentage of UK turnover derived from e-commerce sales from 2008 to 2012.
Over these years, the value of EDI sales has remained substantially higher than website sales. In
2012, the value of EDI sales was £328 billion and website sales £164 billion.
9
Figure 8: Percentage of UK turnover derived from e-commerce sales, 2008 to 2012)
20
Website sales
EDI sales
Total e-commerce sales
18
16
14
12
10
8
6
4
2
0
2008
2009
2010
2011
2012
Source: Office for National Statistics
Note: Businesses with 10 or more employees.
The percentage of businesses selling over a website increased from an estimated 13% in 2008 to
19% in 2012. Figure 9 shows the percentage of UK businesses making e-commerce sales, by size of
business (defined by the number of employees). In 2012, the percentage of large businesses (those
employing at least 1,000 people) making website sales continued to be higher (47%) than small
businesses (those employing 10 to 49 employees) (17%) (see Figure 9).
The percentage of businesses selling via EDI was 6% in 2012. As with website sales, in 2012 the
percentage of large businesses making EDI sales continued to be higher (34%) than small businesses
(4%) (see Figure 9).
Looking at Figures 9 it is evident that in 2012, there continued to be a smaller percentage of
businesses selling via EDI (6%) compared with the percentage of businesses selling via a website
(19%), but it must be remembered that the total value of EDI sales (£328 billion) continued to be
higher than the value of website sales (£164 billion). It should be noted that some businesses make
sales via a website and EDI.
10
Figure 9: Percentage of UK businesses making e-commerce sales, by size of business (2008 to 2012)
50
10 - 49 employees
50 - 249 employees
250 - 999 employees
1000+ employees
All size bands
45
40
35
30
25
20
15
10
5
0
2008
2009
2010
2011
2012
2008
2009
Website sales
2010
2011
2012
EDI sales
Source: Office for National Statistics
Figure 10 shows the percentage of businesses making EDI and website sales by industry. It should be
noted that some businesses make sales via EDI and a website. Here you can see that in 2012, the
construction industry had the lowest percentage of businesses selling over a website (7%) and
accommodation & food services the lowest percentage of businesses selling via EDI (1%). In
contrast, the retail industry had the highest percentage of businesses selling over a website (34%)
and the wholesale industry the highest percentage of businesses selling via EDI (15%).
Figure 10: Percentage of UK businesses (with 10 or more employees) making EDI and website sales
by industry (2012)
Construction
Other services
Accommodation & food services
Manufacturing
EDI sales
Website sales
Information & communication
Retail
Wholesale
0
5
10
15
20
25
30
35
40
Source: Office for National Statistics
Note: estimates for the Utilities and Transport & storage sectors are not included in this table due to concerns about data
quality. Proportions are not additive as businesses may make both EDI and Internet sales.
11
4.2.1 E-commerce sales as a percentage of turnover
Figure 11 shows the percentage of businesses (with 10 or more employees) where at least 1% of
turnover was from online sales (EDI and website). In 2012, Denmark (27%) had the highest
percentage of businesses where at least 1% of turnover was from e-commerce sales. The UK, at 19%,
was 5 percentage points above the EU275 average (14%).
Figure 11: Percentage of businesses (with 10 or more employees) where at least 1% of turnover was
from e-commerce sales (2012)
Italy
% of businesses selling online
(at least 1% of turnover)
Bulgaria
Latvia
Cyprus
Greece
Romania
Poland
Hungary
Estonia
Slovenia
France
Malta
Spain
Austria
Netherlands
EU27
Portugal
Luxembourg
Finland
Slovakia
Croatia
United Kingdom
Lithuania
Belgium
Germany
Ireland
Iceland
Sweden
Norway
Czech Republic
Denmark
0
5
10
15
20
25
30
Source: Eurostat
Notes: data for Turkey are not available. Data for the financial sector are not included.
5
EU28 data are not yet available.
12
4.2.2 Cross-border e-commerce sales
It must be remembered that the European definition of e-commerce includes the sale or purchase of
goods or services conducted over computer networks using methods specifically designed for the
purpose of receiving or placing orders, including via EDI or website. It must also be noted that under
the European definition the goods or services must be ordered using e-commerce methods but the
payment and the ultimate delivery of the goods or services do not have to be conducted using ecommerce.
Using the above European definition of e-commerce6, Figure 12 shows the percentage of businesses
(with 10 or more employees) that made e-commerce sales in their own country, and Figure 13 the
percentage of businesses that made e-commerce sales to other EU countries and to non-EU
countries in 2012.
Figures 12 and 13 show, unsurprisingly, that in 2012, all countries shown had a higher percentage of
businesses making e-commerce sales to their own country compared with e-commerce sales to
other EU countries and non-EU countries. Denmark had the highest percentage of businesses
making e-commerce sales to their own country (29%). In contrast, only 10% of Danish businesses
made e-commerce sales to other EU countries (3 percentage points higher than the EU27 average of
7%); and the percentage of Danish businesses making e-commerce sales to non-EU countries was 6%
(compared with an EU27 average of 4%).
In 2012, 21% of UK businesses made e-commerce sales to their own country, which was 5
percentage points higher than the EU27 average (16%). 9% of UK businesses made e-commerce
sales to other EU countries and 7% of UK businesses made e-commerce sales to non-EU countries.
In 2012, Luxembourg and Iceland had the highest percentage of businesses making e-commerce
sales to other EU countries (Luxembourg: 15%, Iceland: 14%) and to non-EU countries (both at 12%).
6
The UK definition of e-commerce is consistent with the European definition.
13
Figure 12: Percentage of businesses (with 10 or
more employees) that made e-commerce sales
in their own country (2012)
% of businesses making e-commerce sales to their
own country
Italy
Bulgaria
Cyprus
Ireland
Romania
Latvia
Greece
Poland
Hungary
Estonia
Slovenia
Spain
France
Luxembourg
Portugal
Malta
EU27
Austria
Slovakia
Finland
Belgium
Croatia
United Kingdom
Netherlands
Lithuania
Sweden
Norway
Czech Republic
Iceland
Denmark
Figure 13: Percentage of businesses (with 10 or
more employees) that made e-commerce sales
to other EU countries and to non-EU countries
(2012)
% of businesses making e-commerce sales to non-EU
countries
% of businesses making e-commerce sales to other EU
countries
Italy
Bulgaria
Cyprus
Ireland
Romania
Latvia
Greece
Poland
Hungary
Estonia
Slovenia
Spain
France
Luxembourg
Portugal
Malta
EU27
Austria
Slovakia
Finland
Belgium
Croatia
United Kingdom
Netherlands
Lithuania
Sweden
Norway
Czech Republic
Iceland
Denmark
0
5 10 15 20 25 30 35
0
5 10 15 20 25 30 35
Source: Eurostat
Note: data for Germany and Turkey are not available. Countries are ranked based on proportion of businesses making ecommerce sales to their own country.
4.3
Estimates of Internet sales from the retail sales inquiry
Figure 10 showed estimates from the ONS e-commerce survey, which indicated that the highest
percentage of businesses making Internet sales was in the retail industry. Retail Internet sales have
grown in recent years. In June 2009, the average weekly value (non-seasonally adjusted) for Internet
retail sales was £303 million, which was 5.6% of the value of all weekly retail sales (£5,391 million).
In June 2014, the average weekly value for Internet retail sales was £664 million, which was 10.5% of
the value of all of all retail spending (excluding automotive fuel).
14
Looking at the different retail industry sectors, Figure 14 shows their percentage of Internet sales in
June 2014. Non-store retailing (which includes mail order, catalogues, and stores selling
predominantly over the Internet) had the highest percentage of Internet sales in June 2014, with an
average weekly value of £315 million. In contrast, in June 2014, the food sector continued to have
the lowest percentage of Internet sales (3.7%), with an average weekly value of £109 million.
Figure 14: Percentage of GB Internet sales in each retail sector (June 2014)
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
All retailing
excluding
automotive fuel
Food stores
Department stores Textile, clothing and
footwear stores
Household goods
stores
Non-store retailing
Source: Office for National Statistics
Note: Value of non-seasonally adjusted Internet sales as a proportion of all retailing excluding automotive fuel.
Section 5: E-Commerce from a measuring Gross Domestic Product perspective
From the perspective of measuring the economy, the concept of Gross Domestic Product (GDP) is
measured using three different approaches, in turn these need to be exhaustive and when balanced
give a single estimate for the UK economy. The primary focus in the measurement of GDP is not on
the mode of transaction (e.g. e-commerce), but on total UK economic activity. There are three key
points in relation to e-commerce from a measuring GDP perspective:
1. GDP is measured by three approaches: production, expenditure and income; and there are
numerous data sources that are used to measure the three approaches (see Table 1).
Where e-commerce is important is in providing information to inform the measurement of
imports, especially in relation to e-services such as digital downloads and applications.
2. To ensure that the value added of unregistered businesses is included in the measurement
of GDP, adjustments are made as part of the annual national accounts Supply and Use
balancing process. These adjustments are based on periodic analysis using additional data
from Her Majesty’s Revenue and Customs (HMRC).
15
3. ONS have an on-going programme of work to improve coverage of e-services exported and
imported by households. The first developments in this area will be improvements to online
gambling data in the UK National Accounts 2014 (Blue Book 2014).
This section looks first at the measurement and compilation of GDP and then specifically at ecommerce in relation to GDP.
5.1
Measurement and compilation of GDP
The standards used in the production of national accounts and the associated economic indicators
(e.g., GDP, gross national income (GNI)) are set out in: the System of National Accounts (SNA) and
the European System of Accounts (ESA). UK national accounts are currently compiled on an ESA
1995 basis but will move to an ESA 2010 basis in September 2014.
GDP measures and combines in a single figure:
“all the output (or production) carried out by all the firms, non-profit institutions, government
bodies and households in a given country during a given period, regardless of the type of goods and
services produced, provided that the production takes place within the country’s economic
territory” (Lequiller and Blades, 2006)
In the UK three different approaches are used in the estimation of GDP: production, often referred
to as the output (O) approach, expenditure (E), and income (I).
GDP (O) measures the sum of the value added created through the production of goods and services
within the economy (our production or output as an economy). This approach provides the first
preliminary estimate of quarterly GDP and shows the different contributions of industries (for
example, services) within the economy.
GDP (E) measures the total value of final expenditures on all finished goods and services produced
within the economy; this is by consumers, profit and non-profit institutions and government. It
includes gross capital formation (spending on capital assets, inventories and valuables) and exports
of goods and services less imports of goods and services.
GDP (I) measures the total income earned by individuals and businesses in the production of goods
and services within the economy. The estimates split income into, for example, income earned by
businesses (corporations), employees and the self employed.
The three GDP approaches are calculated, in current price terms, by:
using data from a variety of sources including surveys, secondary sources and forecasts.
applying industry weights derived from the annual national accounts supply use process,
which includes an adjustment for the under-coverage of unregistered self-employed people
and businesses (see Section 5.3.1).
seasonally adjusting the current price data.
In volume terms the three GDP approaches are calculated by:
by deflating the current price data to remove the effects of price changes.
seasonally adjusting the data.
GDP is therefore a complex estimate to produce and is often difficult for people to understand how
it is produced. To assist in understanding GDP data sources, a list is included in the article by
Andrew Walton, published on 30 May 2012.
16
5.2
E-commerce and GDP
Section 4 provided an overview of e-commerce estimates, which showed that:
in 2012, e-commerce accounted for an estimated 18% (£492 billion) of UK business
turnover. 21% of UK businesses in 2012 made e-commerce sales to their own country; 9% of
UK businesses made e-commerce sales to other EU countries and 7% to non-EU countries;
the wholesale industry had the highest percentage of businesses selling via EDI (15%); and
the retail industry had the highest percentage of businesses selling over a website (34%).
In the measurement and compilation of GDP all of this e-commerce activity is included (domestic,
exports and imports) but in different ways. Table 1 shows the GDP scenarios that need to be
considered for the purchasing and selling of goods and services, alongside their GDP classification
and examples of data sources. What is apparent from the GDP scenarios is that the value of
domestic output, expenditure and exports are included in the measurement of GDP; whereas the
value of imports of goods and services are deducted in the compilation of GDP(E) as:
GDP(E) = (final consumption) + (gross capital formation) + (exports – imports)
Figure 15, provides a pictorial overview of the compilation of GDP. In the measurement and
compilation of GDP:
“The distinction between resident and non-resident institutional units is crucial to
the definition and coverage of GDP” (System of National Accounts, 2008, p. 105).
Figure 15: Pictorial overview of the compilation of GDP
From a GDP perspective, what is important is not that the purchases or sales were undertaken
electronically but that purchases and sales from non-UK businesses and households are accurately
measured.
17
Table 1: GDP scenarios, classifications and examples of data sources
GDP scenarios
GDP classification
Examples of data sources
Goods and/or
The goods and/or
Range of ONS business surveys, household surveys
services sold by a UK services produced are and administrative data are used to estimate the
producer and sold to classified as domestic
various components of GDP. Often the same source
UK based consumers output and
is used for different components of GDP.
such as businesses,
expenditure and
government or
included in UK GDP.
ONS business surveys are based on samples drawn
households.
from a comprehensive business register of UK
The value of the goods producers.
and/or services
produced (i.e. output) ONS monthly business survey (covering most
less the goods and
industries in the economy) collects sales of goods
services used up to
and services by UK producers – feeds into GDP (O)
produce the output
and GDP (E).
(i.e. intermediate
consumption)
ONS Annual Business Survey (covering most
contribute to the
industries in the economy) collects a variety of
production (output)
detail including turnover, purchases, capital
approach to
expenditure, labour costs, etc. – feeds into GDP (O),
measuring GDP – refer (E) and (I).
to as GDP (P).
ONS quarterly surveys collecting details such as
businesses capital expenditure, inventories and
profits – feed into GDP (O), (E) and (I).
ONS conducts a range of household-type surveys
such as the International Passenger Survey
(collecting expenditure details of travellers in and
out of the UK) – these feed into GDP (O) and (E).
ONS Living Costs and Food Survey collecting details
on expenditures by households – feeds into GDP
(E).
A variety of other sources are also used such as
government expenditure outturns for central
government expenditures and VAT-based sources
from HMRC as well as data for the agricultural
industry from the Department for Environment,
Food and Rural Affairs and Bank of England
covering all banking activity in the UK.
Goods and/or
services sold by a UK
producer to a nonUK based consumer
(e.g. a business or
household)
These are all classified
as exports. The export
value is included in
GDP (E).
Exports of goods are collected from UK producers
as part of Intrastat (for the EU) and Extrastat (for
non-EU) by HMRC. Various adjustments are made
to the foreign trade statistics to move to a balance
of payments basis (for use in the National Accounts)
including coverage type adjustments.
Exports of services are collected from a number of
18
surveys like the ONS International Trade in Services
Survey (covering exports from UK producers), ONS
International Passenger Survey, Chambers of
Shipping and other data sources.
Goods and/or
services bought
from a non-UK
producer by a UK
consumer (either
business,
government or
household).
These are classified as
imports. The import
value is deducted in
the estimation of
GDP(E) as they do not
contributed to UK GDP
but the country where
the goods and/or
services were
produced
Imports of goods are collected from UK producers
as part of Intrastat (for the EU) and Extrastat (for
non-EU) by HMRC. Various adjustments are made
to the foreign trade statistics to move to a balance
of payments basis (for use in the National Accounts)
including coverage type adjustments as well as
issues like smuggling.
Imports of services are collected from a number of
surveys like the ONS International Trade in Services
Survey (covering exports from UK producers), ONS
International Passenger Survey, Chambers of
Shipping and other data sources.
Imports of services is an area where ONS are
seeking to improve coverage.
5.3
Current coverage
5.3.1 Unregistered businesses
The ONS Interdepartmental Business Register (IDBR), used as the sample frame for ONS and some
other government departments’ business surveys, includes all UK businesses registered for either
Value Added Tax (VAT) or Pay As You Earn (PAYE). In March 2013, the registered business
population was estimated to be 2.17 million enterprises, compared with 2.15 million in March 2012
(ONS, October 2013). In contrast, the estimated number of unregistered businesses was higher than
registered businesses. At the start of 2013, there were an estimated 2.7 million unregistered
businesses. Between the start of 2012 and 2013, estimates of unregistered businesses increased by
89,000 (3.3%) (Department for Business, Innovation and Skills (BIS), October, 2013).
Unregistered businesses (e.g. smaller non-employing businesses including a large proportion of self
employed sole proprietors and partnerships) are not easily identifiable in administrative systems. It
would therefore be difficult to maintain a register of unregistered businesses to be used as a
sampling frame. Including unregistered businesses in surveys would also pose an additional burden
on them.
To overcome this issue and to ensure that the value added of unregistered businesses is included in
the measurement of GDP, adjustments are made as part of the annual national accounts Supply and
Use balancing process. These adjustments are based on periodic analysis using additional data from
HMRC. ONS are currently repeating this analysis.
The analysis is undertaken at the detailed UKSIC class level and adjustments are aggregated and
applied to unbalanced Supply and Use industry data at the UKSIC 2007 2- or 3-digit level; this is
because individual industries have different proportions of unregistered businesses.
19
Tables 2 and 3 show examples of the current largest and smallest unregistered business industry
adjustments, the figures represent the percentage uplifts applied to both market sector industry
output and intermediate consumption data from the Annual Business Survey during the Supply and
Use process. The contribution of the adjustments to final balanced estimates of output and
intermediate consumption and hence gross value added can vary due to other coverage, conceptual
and coherence adjustments applied later on during the balancing process.
Table 2: Examples of industries with estimated high contributions from unregistered businesses
(based on Supply and Use analysis)
Industry
49.3-5
59
93
96
56
55
Land transport services and transport services via pipelines,
excluding rail transport
Motion picture, video and TV programme production services,
sound recording & music publishing
Sports services and amusement and recreation services
Other personal service activities
Food and beverage serving services
Accommodation services
Estimated missing output
and intermediate
consumption from
unregistered businesses (%)
17.3
7.3
6.8
5.7
5.5
4.8
Table 3: Examples of industries with estimated low contributions from unregistered businesses
(based on Supply and Use analysis)
Industry
53
29
10.6
06
11.01-6
Postal and courier services
Motor vehicles, trailers and semi-trailers
Manufacture of grain mill products, starches and starch products
Extraction of crude petroleum and natural gas
Manufacture of alcoholic beverages
Estimated missing output
and intermediate
consumption from
unregistered businesses (%)
0.8
0.3
0.1
0.0
0.0
In the estimation of GDP, what is more important is not the number of unregistered businesses but
the value of their economic activity. In relation to some speculation that the contribution of
unregistered businesses has increased since the financial crisis, BIS estimates show relatively stable
proportions of unregistered business turnover to total turnover between 2006 and 2013 (see Table
4). Any under coverage will most probably impact on GDP levels rather than GDP growth.
20
Table 4: Number and proportions of unregistered businesses and their turnover
At the
start of
year
2006
2007
2008
2009
2010
2011
2012
2013
Estimated total
number of
businesses
Estimated number
of unregistered
businesses
Estimated value of
unregistered
business turnover
(million)
Unregistered
businesses as a
proportion of all
businesses
%
(£billion)
Estimated
unregistered business
turnover as a
proportion of all
business turnover
(million)
4.1
4.3
4.3
4.3
4.4
4.5
4.8
4.9
2.0
2.1
2.1
2.2
2.4
2.5
2.7
2.7
49%
50%
49%
51%
53%
55%
55%
56%
88
94
98
101
88
84
91
91
3.4%
3.4%
3.3%
3.1%
2.7%
2.8%
2.9%
2.8%
See note
below
Source: BIS*
Note: turnover estimates after 2009 are not strictly comparable to the earlier estimates – improved methodology used for
estimating the number of unregistered businesses was introduced in 2010 and it has only been possible to update the
historical series for the number of businesses, which was reduced by about 400,000 when the methodology changed.
*BIS do not directly collect information on the turnover of unregistered businesses, though estimates have been derived by
comparing against turnover data for zero-employee VAT/PAYE registered businesses for each 2 digit SIC division.
Comparable estimates of unregistered turnover are only available between 2010 and 2013.
5.3.2 Trade
Measuring exports and imports of goods is relatively straight forward as they are products that have
to be physically moved. In the UK, exports and imports of goods are measured via Intrastat (for the
EU) and Extrastat (for non-EU), which are collected by Her Majesty’s Revenue and Customs (HMRC).
More challenging is the measurement of exports and imports of services, as they do not have to be
physically moved and are therefore, more difficult to measure.
5.3.2.1 Trade in services
The main sources of trade in services are surveys of UK businesses. In the UK, exports and imports
of services are measured by the ONS International Trade in Services (ITIS) survey, and other sources
such as Bank of England surveys. ITIS includes businesses known to engage in overseas trade in
services (purchases and sales), either from their past responses to ITIS or from their responses to a
trigger question in the ONS ABS. Exports of services are covered (though there remain challenges in
ensuring that multinational businesses identify and include their intra-trade). Imports are covered
insofar as a UK business is a customer for the service in question.
ITIS has a sample size of 14,000 businesses, including all large businesses and a representative
sample of medium and small businesses. The survey excludes businesses classified to travel,
transport, banking and other financial institutions, higher education, charities and most activities
within the legal profession; as there are other sources of data for these e.g. International Passenger
Survey.
Where a household purchases services from a non-UK business, thereby importing a service, there
are currently limited data sources for the value of the import. To improve this ONS has initiated a
programme of work to identify additional sources of data (please see section 5.3.2.2 which provides
an overview of work that has been undertaken to improve online gambling estimates).
Household expenditure on imported services is collected in responses to the ONS Living Costs and
Food Survey, which collects data on total household expenditure and is included in GDP(E).
21
However, households will not typically know that the service has been imported e.g. purchased from
a non-UK registered business, so the explicit value of UK household service imports is not identified.
Currently therefore:
including household expenditure on imported services correctly increases the expenditure
component of GDP; but
this is not offset by the identification of the value of household service imports, which would
reduce the expenditure component of GDP as the impact of ‘trade’ on GDP is calculated as
exports less imports (see Figure 16).
There is a risk therefore that the current estimates of services (particularly e-services) directly
imported by households is marginally understated in the trade estimates and marginally overstated
in GDP; this is why an ONS programme of work has been initiated.
Figure 16: Measurement and coverage of GDP(E)
Good coverage
from business and
household data
sources
Good coverage
from business data
sources
GDP(E) = (final consumption) + (gross capital formation) + (exports – imports)
Good coverage of goods
and services exported and
imported by businesses
Ongoing programme of
work to improve coverage
of services exported and
imported by households
It is currently not possible to estimate the impact to the level of GDP from imported services to
households. However, it is likely to be negligible in relation to annual GDP, which was estimated to
be £1.6 trillion in 2013 in market prices (2010=100). Looking crudely at the potential impact on
trade estimates, in 2013 imports of goods was estimated at £413 billion and exports of goods was
£305 billion. Table 5 shows the 2013 current price (seasonally adjusted) values for imports and
exports of goods (e.g. books and videos) that could potentially also be purchased as services (i.e.
traded online as e.g. downloads). In 2013, these goods represented only 0.6% of total imported
goods and 1.1% of total exported goods. It is unlikely that all of these physical products would be
transformed to e-services, so, in terms of total trade and GDP, the total value of e-services exported
and imported is an important but most probably negligible value.
22
Table 5: 2013 current price (£million, seasonally adjusted) imports and exports of goods that could
also be traded as services
Printed
matter
Packaged
computer
software
Films and
videos
Imports
1648
515
Audio
Total
recordings
and
printed
music
443
130
2736
Exports
2771
223
313
172
3479
Source: ONS MQ10 Trade in Goods Industry BOP
Note: Printed matter includes: printed books, books on disk, tape of other physical media, directories and mailing lists
printed on physical media, printed newspapers and printed journals and periodicals.
Note: Packaged computer software includes: packaged computer games.
Note: Films and videos includes: cinematographic film, films and other video content on disk, tape or other physical media.
Note: Audio recordings and printed music includes: musical audio disks, tapes or other physical media, printed music
(sheet music etc).
The Entertainment Retailers Association (ERA) in their 2014 Yearbook estimate that in 2013 the
majority of entertainment market sales (value), despite the rise of downloads and streaming, are still
from physical formats (56.6%); although some of these will be sold online. In relation to the
proportion of sales via online and physical stores, they estimate that in 2013, 60% of sales were from
‘Internet derived sales – including home delivery, digital download and streaming and other access
services’ and the ‘remaining 40% of revenue was generated by physical stories’ (ERA, 2014). They
also note that e-services that allow consumers to “access content rather than purchasing it outright
(i.e. products like discs or downloads)”is the fastest growing segment (ERA 2014). However, it
should be noted that many of these e-service providers are non-UK registered businesses and the
purchased e-services, in relation to GDP would be treated as imports.
5.3.2.2 Improvements to online gambling data
As part of the ONS programme of work to improve measurement of the value of imported and
exported e-services, improved online gambling data will be included in the UK National Accounts
Blue Book 2014.
Online gambling is a fast-growing e-service. Research by the ONS has shown that the abolition of
the UK betting tax in 2001, replacing it with a tax on operators gross profits, along with deregulation
of the UK gambling industry through the Gambling Act 2005, were catalysts for increased online
gambling and operators moving offshore (or transferring the online arm of their business offshore).
The online gambling market comprises of four sectors; poker, sports betting, bingo and casino.
More than 50% of the current market is dominated by five companies; three of these have moved
offshore in recent years. A Parliamentary business paper (2012) states that 90% of online gambling
consumed in the UK is now supplied from outside the UK.
This equates to an under coverage of imports, as online gambling activity is not currently measured,
although household expenditure on online gambling is included. In the measurement of GDP, the
gambling service charge (i.e. the amount bet minus the amount paid out in winnings (i.e. the amount
lost by the gambler per bet)) needs to be measured. This is equal to the gross gambling yield (GGY)
23
from the operator’s perspective. Data for GGY is available from the Gambling Commission which
publishes estimated figures obtained from H2 gambling capital.
From a trade perspective there is interest in companies operating offshore. An adjustment is made
to GGY figures to account for the percentage of companies operating offshore each year. Data are
available from 2009; a back series to 1997 has been calculated using information on companies’
relocation abroad in conjunction with GGY growth rates 2009 to 2012.
Section 6: Concluding comments
This paper has looked at available sources of e-commerce data, the e-commerce estimates and ecommerce from a measuring GDP perspective. From a policy perspective the internet access and
purchases, the e-commerce survey, and the internet sales series of the retail sales inquiry provide
good sources of estimates to measure and monitor e-commerce activity by UK businesses.
From a measuring the economy perspective, it is essential that fit for purpose data sources are
maintained. ONS has therefore commenced a programme of work to improve the measurement of
services exported and imported by households. The first improvements will be included in Blue Book
2014 with the measurement of online gambling.
Acknowledgements
I would like to thank the following people for their contributions to this paper: Darren Morgan, David
Matthews, Hazel Clarke, Pete Lee, Stephen Curtis, Russ Pierce, Ciara Williams, Julie Griffiths, Kate
Davies, David Howells, Richard Wild, Sanjiv Mahajan and Heather Bovill. As well as those who
produced the infographic to accompany this paper.
References
Business Innovation and Skills (2013) Business Population Estimates, Business Innovation and Skills.
Entertainment Retailers Association, 2014 Yearbook.
European Commission European System of Accounts, 1995, Eurostat
European Commission European System of Accounts, 2010, Eurostat
Lequiller, F., and Blades, D. (2006) Understanding National Accounts, OECD publication.
OECD (2011) Guide to measuring the information society, OECD.
Office for National Statistics (2014) Retail Sales, June 2014, Office for National Statistics.
Office for National Statistics (2014), Internet access – households and individuals 2014, Office for
National Statistics.
24
Office for National Statistics (2013), E-commerce and ICT activity of UK businesses, 2012, Office for
National Statistics
Office for National Statistics (2013), UK Business: Activity, Size and Location, Office for National
Statistics.
Parliamentary business paper (2012), The Gambling Act 2005: A bet worth taking? Culture, Media
and Sport Committee.
United Nations System of National Accounts, 1993 and 2008, United Nations.
Walton, A. (2012) Updated Analysis: Why is GDP revised, Office for National Statistics.
25