Department of Trade and Industry
Office of Science and Technology
PSA target metrics for the UK research base
December 2005
Crown Copyright
The material featured in this report is subject to Crown copyright protection unless otherwise indicated.
2
Using OST PSA target metrics for Research or Private Study:
All the material featured in this report may be copied or downloaded to file or printer for the purposes of research and private study without requiring specific prior
permission. Where OST material is being published or copied to others the following statement must be shown:
Source: Office of Science and Technology, PSA target metrics 2005
Crown copyright material is reproduced with the permission of the Controller of HMSO
Individuals and organisations wishing to reproduce material, other than for the purposes of research or private study, require a Licence. This must be sought from the
Office of the Governmentâs Chief Scientific Adviser.
Copyright of a Third Party
The permission to reproduce Crown protected material does not extend to any material in this report which is identified as being the copyright of a third party. This
exclusion extends to all those data that are the property of Thomson Scientific
ÂŽ
. Authorisation to reproduce such material must be obtained from the copyright holders
concerned other than for the purposes of research and private study.
Contact details
The reporting organisation is
Evidence
Ltd
103 Clarendon Road, Leeds LS2 9DF
t/
0113 384 5680
f/
0113 384 5874
e/
enquiries@evidence.co.uk
Evidence
Ltd is registered in England, Company no 4036650, VAT registration 758 4671 85
http://www.evidence.co.uk
3
Contents
Section
Page
Summary
4
Introduction
6
Definition and description of indicators
8
Research FootprintsÂŽ
12
Thematic commentary
16
Indicator summary pages
20
Theme 1 Inputs (including expenditure on research)
22
Theme 2 Outputs (including people and publications)
Theme 3 Outcomes (research recognition, citations; training and research quality)
Theme 4 Productivity - financial (outputs and outcomes related to inputs)
Theme 5 Productivity - labour (outputs and outcomes related to other measures)
Theme 6 People
Theme 7 Business expenditure
Background to the indicators
142
Glossary
156
4
Summary
This is our third report on the performance of the UK research base. It confirms the UKâs strong relative international performance in terms of achievement, productivity
and efficiency. The UK sustains a more consistent performance across fields than most countries and is strongest overall in the natural sciences. On many indicators it
has been second only to the USA, but the situation is dynamic: it has moved into first place this year on some indicators and it has been overtaken in other areas. The
UKâs strong international performance in terms of quality has been achieved with lower investment compared to its competitors. Thus far, this has led to a high level of
productivity in the research base, for both research publications and trained people. The UK has a relatively lower general availability of people with research training
although there is a shift to a more highly skilled profile for the research workforce.
Theme 1 - Inputs (including expenditure on research)
The UKâs share of OST comparator group R&D expenditure is about 4.5%. The
UK is spending less on research (Gross Expenditure on R&D, GERD) as a
proportion of its total economic activity (GDP) than its competitors. GERD, at
about 1.8% of GDP, has increased only marginally, and the UK is seventeenth
of the twenty-one OST comparator group nations for which data are available
(the average is about 2.25% of GDP) and seventh among G8 nations. Total
publicly performed R&D (PUBERD = Government + Higher Education sectors)
increased in real terms over the last decade (with a slight recent drop) but the
increase is less than the average for the OST comparator group.
Theme 2 - Outputs (including people and publications)
The UKâs share of OST comparator group PhD awards and publications is
about 9%. The UKâs share of PhD awards is broadly the same as other
countries for which data are available but much less than Germany. Growth is
higher in SE Asia and some smaller and research competitive European
nations. The UKâs share of world journal article publications is ranked second
behind the USA. A recent plateau in publication output appears to be
consolidation rather than actual contraction.
Theme 3 - Outcomes (research recognition, citations; training and
research quality)
The UKâs share of world citations is about 12% and is second highest to the
USA. It remains well ahead of most OST comparator group competitors.
Germany has not improved on last year, but China and some smaller nations
are increasing their global share rapidly and this has effects on larger countries.
The UK is also second to the USA in all of ten main research fields except
mathematics, where it remains third, and in physical sciences and in
engineering, where it is fourth. UK share of citations has been maintained in
most fields. Smaller nations have displayed sometimes substantial but
inconsistent increases. Other measures reflect the high impact of UK
publications. The UK has now moved ahead of the USA on impact for pre-
clinical & health and biological sciences. There are marked improvements
elsewhere but next yearâs data will reveal whether these are statistical or
sustained. The UK also benefits from a relatively low frequency of uncited
material and a consistently good performance across disciplines. Correlation
between training volume and impact is positive through weak for the UK
whereas it is absent for most other G8 countries.
5
Theme 4 â Productivity - financial (outputs and outcomes related to
inputs)
The UK is highly productive by the measures applied in this analysis but it is not
consistently quite as strong as in the past. The UK produces relatively more
PhDs per unit HERD (Higher Education R&D spend) than most OST
comparator group nations. It is ranked fifth where the USA is ranked eleventh.
In terms of publicly performed R&D (PUBERD), UK productivity ranks first in the
G8 and well ahead of the OST comparator group average. The UK is still
ahead of other G8 nations in terms of citations per unit GERD but has dropped
to fifth place in the OST comparator group. Its share of citations (Theme 3)
compared to share of funds (Theme 1) is twice the OST comparator group
average. The UK has been overtaken by Switzerland in terms of citations per
unit HERD, although it has improved on other G8 nations.
Theme 5 â Productivity - labour (outputs and outcomes related to other
measures)
The UK is highly productive in terms of labour productivity. The UK is second
behind and has gained further on Germany in terms of PhDs awarded per
researcher; only Switzerland and Spain are more productive. The UK has a
lead position in G8 countries and is third overall behind Switzerland and the
Netherlands on relative productivity (papers published per researcher) and
effectiveness (citations acquired per researcher). Indeed, its slight fall in
papers but gain on citations suggests a shift to quality over quantity. [NB Italy
data are anomalous.]
Theme 6 - People
For the UK, the general availability of highly skilled people with research
training is lower than among its competitors, although the balance within the
research workforce suggests that there is increasing professionalisation. Only
0.3% of the UK population as a whole and only 0.6% of the workforce would be
classified as a researcher, which is less than the OST comparator group
average. The UKâs rank position in the low âteens among the OST comparator
group has changed little over the period. Researchers are becoming more
frequent as a proportion of R&D personnel, however, which may reflect
increasing research professionalisation. This is good for the UK where change
is faster than average for the OST comparator group.
Theme 7 - Business expenditure
The UK has performed relatively well in terms of business investment in the HE
research base as defined by OECD but its position is now weakening.
Business spending as a part of HERD increased in the UK in the mid-1990s but
it is now falling and the UK is no longer ahead of OST comparator group
average. The UK profile for business spend as a proportion of HERD tends to
track that of the USA. At the level of research fields, the fall is most noticeable
in the natural sciences, its core area of research strength.
6
Introduction
This is a report about indicators of the UKâs relative international research
performance in science, engineering, the social sciences and the humanities
and arts. It is the third report with these indicators and in this format. The
information content, analysis, commentary and overall structure have
developed in successive cycles. The main change this year is the inclusion of
data on research in the humanities and arts. This has extended the number of
sub-fields covered by some indicators.
The
Research Footprint
ÂŽ
diagrams summarise the outcomes of analyses for six
leading indicators, comparing the research profile of the UK, the G8 and a
number of other leading research economies. The
thematic commentary,
following the Footprints, gives a broad overview of the UKâs performance in
terms of the selected indicators.
Background
The objective is to support a system for assessing outputs, outcomes and
impacts related to the Public Service Agreement (PSA) target to âimprove the
relative international performance of the UK research baseâ. This target is
challenging. Many studies have indicated that successive advances in
research become increasingly expensive (the âsophisticationâ factor, Advisory
Board for the Research Councils (ABRC), âStrategy for the Science Baseâ,
1986) and the costs of improving relative performance rise in parallel.
The Atkinson Review of âMeasurement of Government Outputâ (2005) for the
Office of National Statistics affirmed that âthe measurement of quality is central
to our concernsâ. Sir Tony Atkinson recommended that all assessments of
output and productivity should account for this. This is what the OST sets out
to do, capturing information not only on what the research base produces for
the Science Budget investment but how its output is perceived internationally.
The UK is widely acknowledged to be an extremely effective research
performer. It is therefore difficult to improve significantly on this relative level of
achievement. Indeed, it will be difficult in some fields to maintain the UKâs
international status without, for example, additional investment that meets the
growing competition from technologically specialist research nations in the
Asia-Pacific economies. This yearâs report particularly highlights the impact
that China is having as its research base expands.
Until 2002, the UK Office of Science and Technology (OST) employed a core
set of indicators that demonstrated the position of the UK and reflected
effectiveness in the use of research funding. This report describes an extended
basket of indicators established in 2003. Plurality in an indicator system is a
desirable feature, because over-dependency on any one indicator can be
misleading. A balanced set can take account of differences in the pattern of
performance between research disciplines, the interaction between inputs and
outputs and possible measures of efficiency and effectiveness, and year on
year fluctuations in any one indicator.
Assessing excellence is as important as measuring system average. The peak
of research excellence, however defined, includes those highly innovative
outcomes that are most likely to impact on economic performance. The
indicators in this report allow for disaggregation, to throw light on changing
patterns of selectivity and concentration within the UK science base. That work
is being developed in a parallel OST study to be published in 2006.
Other countries and communities â such as the EU, the NSF in the USA,
CWTS in the Netherlands and the OST in Paris - already publish reports about
national science and technology indicators on a regular basis. This report has
taken note of the good practice established elsewhere.
Data and Indicators
Every piece of research data should have three attributes: subject area, time
and location. Each attribute works at varying levels of detail and we need to
identify the best level for analysis. Data about research usually measure
something in one of three primary categories: input (usually financial), activity
(or proxies such as staff numbers) and outputs. Secondary indicators describe
the relationship between them. Sometimes, outputs can be followed through
into outcomes and impacts. The UK indicators include both primary and
secondary indicators and focus on impacts where possible. They are listed in
the table (below) on the â
Definition and description of indicatorsâ.
7
Bibliometric data play a key part in these indicators. There are reasons for
being cautious about some uses of such data particularly with respect to social
science and to humanitiesâ and artsâ research (see
Background
sections after
the indicator pages). There is also great value and applicability. Crucially,
these data uniquely provide us with international comparisons of research
quality for most countries and by subject area.
The
Background
sections that come after the indicator data describe the main
data sources,
list the range of the OST
comparator group of countries
, the level
of
subject disaggregation
and the
time frames
used for comparisons. There is
also a discussion about the significance and interpretation of
bibliometric
indicators
and some cultural aspects of publication and citation behaviour in
different countries and disciplines.
International comparisons
are made across an OST comparator group of 25
countries. This includes the full G8 (UK, USA, Canada, France, Germany ,Italy,
Japan, Russia). In addition there is a combination of some larger and OECD
countries from different continents with research bases both similar and
contrasting in structure to the UK, and a spread of smaller nations with active
and rapidly growing research bases with specific strengths. These are
Belgium, Denmark, Finland, Netherlands, Poland, Spain, Sweden, Switzerland,
Australia, Brazil, China, India, Iran, Israel, Singapore, South Africa, South
Korea, and Taiwan. A separate line of analysis for a group of 15 EU countries
(EU15 = member countries in 2003) is also included where feasible and
appropriate.
Many of the graphs that illustrate performance indicators use short codes for
these countries, for clarity. These codes are linked to their countries in a table
in the
Background
sections.
Subject disaggregations
used in this report employ two systems of
categorisation. First, there are five main OECD categories (medical sciences,
natural sciences, engineering and technology, social sciences and humanities
[which includes the arts]). Second, subjects based on 69 UK Units of
Assessment (UoAs) are grouped by publication similarity into ten Super-UoAs
(Clinical, Pre-clinical/health, Biological sciences, Environmental sciences,
Mathematics, Physical sciences, Engineering, Social sciences, Business,
Humanities).
International R&D databases have historically focussed on science and
technology and are therefore have some deficits in social science and
humanities data. This does affect some analyses, and this is discussed further
in the
Background
sections.
The humanities and arts are included in the subject spread for the first time.
Not all the research indicators used in the natural sciences are well suited to
analysing research performance in these disciplines. Account can be taken of
feedback from readers and users of this report so as to âtuneâ these
presentations for later editions.
Details of the countries and subjects are given in the relevant part of the
Background
section after the indicators.
8
Definition and description of indicators
Indicator number
Description of performance
indicator
Condition signalling
improvement
Level of disaggregation
Primary data sources
THEME 1
INPUTS including expenditure on
research
OECD MSTI 2005-1
1.01
GERD relative to GDP (R&D
intensity)
Increased proportion of R&D
specific spend
System
1.02
Publicly performed R&D
(PUBERD) as proportion of GDP
Increased proportion of R&D
specific spend
System
OECD MSTI 2005-1
THEME 2
OUTPUTS including people and
publications
2.01
Number and share of OECD PhD
awards
Increased count and increased
share by comparison with
competitors
System
OECD Education Database
OECD MSTI 2005-1
2.02
PhDs awarded per head of
population
Increased ratio
System
OECD Education Database
2.03
Number and share of world
publications
Increased count and increased
share by comparison with
competitors
System
ISI National Science Indicators 2004
THEME 3
OUTCOMES including research
recognition and citations; training
and research quality
3.01
Number and share of world
citations
Increased count and increased
share by comparison with
competitors
System
ISI National Science Indicators 2004
3.02
Number and share of world
citations in ten main research
areas
Increased national count and
share
SUoA
ISI National Science Indicators 2004
3.03
Rank of share of world citations
by nine main research fields -
frequency in top 3
More frequent presence in top
three among fields
System
ISI National Science Indicators 2004
9
Indicator number
Description of performance
indicator
Condition signalling
improvement
Level of disaggregation
Primary data sources
3.04
Share of citations relative to
share of publications
Increase in citation share
compared to source share within
field
SUoA, using NSI5 for
constant time frame
ISI National Science Indicators 2004
3.05
Number and proportion of uncited
publications
Decrease in ratio of uncited to
total sources
SUoA, using NSI5 for
constant time frame
ISI National Science Indicators 2004
3.06
Number and share of cited
publications
Increase in ratio of cited to total
sources
System
ISI National Science Indicators 2004
3.07
National share of papers in top
1% by citation count
Increase in share
System
ISI Essential Science Indicators
3.08
Citation impact (citations per
publication) relative to world
baselines
Increased impact compared to
world
System
ISI National Science Indicators 2004
3.09
Citation impact relative to world
baselines in ten main research
fields
Increased impact in main
research fields
SUoA
ISI National Science Indicators 2004
3.10
Variation and consistency of
research strength
Reduced ratio between variance
and average. Maximised ratio of
average/variance.
SUoA
ISI National Science Indicators 2004
ISI National Science Indicators 2004
3.11
Relationship between distribution
of research training across
subjects and research quality
Improved match in distribution;
improved research training power
(product of volume and quality)
Data related across
research fields (OECD
level)
OECD Education Database
THEME 4
PRODUCTIVITY â FINANCIAL
including outputs and outcomes
related to inputs
OECD Education Database
4.01
PhDs awarded relative to HERD
Increased ratio
System
OECD MSTI 2005-1
OECD Education Database
UK SET statistics and HESA data
4.02
PhDs awarded relative to HERD
in five main research areas
Increased ratio
OECD fields
OECD RDS 2004-2
10
Indicator number
Description of performance
indicator
Condition signalling
improvement
Level of disaggregation
Primary data sources
OECD Education Database
4.03
PhDs awarded relative to
PUBERD
Increased ratio
System
OECD MSTI 2005-1
ISI National Science Indicators 2004
4.04
Citations relative to GDP
Improved ratio of citations per
GDP compared to recent past
System
OECD MSTI 2005-1
ISI National Science Indicators 2004
4.05
Number and share of citations
relative to GERD
Increased citations per unit
expenditure
System
OECD MSTI 2005-1
ISI National Science Indicators 2004
4.06
Citations relative to PUBERD
(GOVERD + HERD)
Increased citations per unit
expenditure
System
OECD MSTI 2005-1
ISI National Science Indicators 2004
4.07
Citations relative to HERD
Increase in citations per unit
spend
System
OECD MSTI 2005-1
ISI National Science Indicators 2004
UK SET statistics and HESA data
4.08
Citations relative to HERD in five
main research areas
Increase in citations per unit
spend at the OECD macro-
category level
OECD fields
OECD RDS 2004-2
THEME 5
PRODUCTIVITY â LABOUR
including outputs and outcomes
related to non-financial measures
OECD Education Database
5.01
PhDs awarded per researcher
Increased ratio
System
OECD MSTI 2005-1
ISI National Science Indicators 2004
5.02
Publications per researcher
Increased relative output
System, NSI5 averaging
OECD MSTI 2005-1
ISI National Science Indicators 2004
5.03
Citations per researcher
Increase in citation ratio
System, rolling 5 year
averages
OECD MSTI 2005-1
THEME 6
PEOPLE
6.01
Researchers per thousand
population
Increased ratio
System
OECD MSTI 2005-1
11
Indicator number
Description of performance
indicator
Condition signalling
improvement
Level of disaggregation
Primary data sources
6.02
Researchers per thousand
workforce
Increased ratio
System
OECD MSTI 2005-1
6.03
R&D personnel per thousand
population
Increased ratio
System
OECD MSTI 2005-1
6.04
R&D personnel per thousand
workforce
Increased ratio
System
OECD MSTI 2005-1
6.05
Researchers per R&D personnel
Change in ratio
System
OECD MSTI 2005-1
THEME 7
BUSINESS EXPENDITURE
7.01
Business R&D investment in
publicly performed R&D (BE-
PUBERD as a proportion of
PUBERD)
Increased ratio
System
OECD MSTI 2005-1
7.02
BE-GOVERD as a proportion of
total GOVERD
Increased ratio
System
OECD MSTI 2005-1
7.03
BE-PNPERD as a proportion of
total PNPERD
Increased ratio
System
OECD RDS 2004-2
7.04
BE-HERD as a proportion of total
HERD
Increased ratio
System
OECD MSTI 2005-1
7.05
BE-HERD as a proportion of total
HERD in five main research
areas
Increased ratio
OECD fields
OECD RDS 2004-2
UK SET statistics and HESA data
12
Research Footprints
ÂŽ
There are over 30 research indicators grouped under seven themes. This complex body of data provides an informative and comprehensive view of many aspects of
the comparative international performance of the research base, but it is not readily absorbed. We have used a Research FootprintÂŽ to illustrate key data.
Each country has a distinctive Research Footprint
ÂŽ
of its international research competitiveness. Our figure uses six key indicators and provides a direct graphical
comparison of the performance of select comparator countries with the OST comparator group average. The shaded area is the âfootprintâ of a stated country, which can
be compared directly with the dotted line that marks the average footprint for the group. Each axis measures a specific indicator, with the lowest level of performance
(low rank or zero activity) at the origin near the centre and the maximum value at the outer end of the axis. The area of the footprint has no statistical significance.
1.02 PUBERD
per
GDP
Theme: Inputs
Full title: Publicly performed R&D (PUBERD) as proportion of GDP
Description: Volume of publicly funded R&D relative to general economy
2.01
Share of OECD PhDs
Theme: People (research degree output)
Full title: Number and share of OECD PhD awards
Description: Highly skilled people: research degree output
2.03
Share of world publications
Theme: Outputs
Full title: Number and share of world publications
Description: Relative output volume
3.01
Share of world citations
Theme: Outcomes
Full title: Number and share of world citations
Description: Esteem measured by share of world citations
3.03
Lead citation share by research field
Theme: Outcomes
Full title: Frequency in top three for rank of world citation share by nine main
research fields
Description: Breadth of research strength measured by spread of dominance
6.02
Researchers per thousand workforce
Theme: People
Full title: Workforce research capacity
Description: Skilled R&D capacity within national workforce
3.01 Share of world citations
2.03 Share of world
publications
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
13
Research Footprint
ÂŽ
of comparative UK research performance
The Research Footprint
ÂŽ
for the UK is compared in the next two pages, first
with other G8 nations and with the pattern for the EU15 as a whole and second
with a set of other leading research nations in the OST comparator group.
The display uses absolute values, not ranked position. Because of data and
analysis changes, the comparator average Research Footprint
ÂŽ
â essentially
the reference point for others - covers a slightly different area to that of 2004
and relative performance of specific countries may therefore appear to expand
or retreat. The data coverage â for countries, years and fields â has improved
again since last year. This has identified a number of exceptional performers for
particular indicators, some of which appear to behave inconsistently and may
be amended in later reports. There are also some anomalous values (e.g.
those involving GDP for Russia).
The status of the USA is quite clear. It will continue to be a strong performer
across the board and contributes the maximum volume performance on most
indicators because of its sheer size, although its efficiency is certainly less
impressive and its effectiveness is being challenged. One example of an
exception is in terms of public expenditure on R&D as a proportion of GDP
(indicator 1.02). On that indicator there are smaller nations, such as Sweden,
Denmark and the Netherlands, that all spend relatively more. Other indicators
show that they all have relatively highly skilled populations and rising
performance in niche research areas.
The EU has no calculated value on indicator 3.03 [which would be a summation
of specific countries rather than an integrated figure] where it would score as
highly as the USA. It has more PhDs than the USA (indicator 1.01) and higher
average PUBERD per GDP (indicator 1.02) but relatively fewer researchers in
its total workforce (indicator 6.02), although the data are patchy in this area.
The UK has a substantial and growing share of world publications (indicator
2.03) and citations (indicator 3.01). Because it ranks second on average to the
USA, it also has a very good position on average ranking by major research
area (indicator 3.03). The UKâs share of OECD PhD awards (indicator 2.01)
matches its share of outputs but is less than its achievement on other
indicators. It has a relatively low (below average) concentration of researchers
within its workforce (indicator 6.02) by comparison with some of the smaller
countries with rising profiles.
Overall, the UK can be seen to have a good all round performance by
comparison with most of the G8 and all of the smaller countries, but the well
balanced performance of Germany â good business investment, strong PhD
output, above average proportion of research workers - is also notable. Each
report has confirmed that it is the major research competitor for the UK in
Europe. Japanâs strength is in its research work-force (indicator 6.02), as it is
for Sweden. While Switzerland has a strong average bibliometric performance
in many fields, in the Research Footprint
ÂŽ
presentation its relatively small
research base and capacity becomes clear.
China has been given a Research Footprint
ÂŽ
this year, replacing Belgium.
China has exceptional growth in its research base and is rapidly expanding its
volume of activity with rapid growth in GERD, growth in outputs in most fields of
science and technology and a very large R&D workforce with a relatively high
proportion of researchers. The Research Footprint
ÂŽ
shows the extent to which
this latent strength has yet to be translated into quality outcomes. On these
criteria, China remains a âsleeping giantâ but the weight of its footfall can be
expected to change this picture significantly in the next few years.
14
Research Footprints
ÂŽ
for UK, G8 countries (except Russia) and the EU15
14
17
20
23
UK
USA
EUROPEAN UNION
CANADA
FRANCE
GERMANY
ITALY
JAPAN
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
15
Research Footprints
ÂŽ
for other leading comparator nations
77
50
53
56
SWITZERLAND
SWEDEN
SPAIN
SOUTH KOREA
NETHERLANDS
DENMARK
CHINA
AUSTRALIA
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
3.01 Share of world citations
2.03 Share of World Sources
6.02 Researchers per
thousand workforce
1.02 PUBERD per GDP
2.01 Share of OECD PhDs
3.03 Lead citation share
by research field
16
Thematic commentary
Theme 1 - Inputs (including expenditure on research)
The UK is spending less on research as a proportion of its total economic
activity than its competitors and less than the EU15 average. It share of
comparator group R&D spend is about 4.5%, which is barely half its share of
group outputs (
Theme 2
and
Indicator 4.05
). Its ranked position on R&D spend
is also declining on current data.
GERD as a proportion of GDP indicates research intensity in the economy. UK
GERD was about 8% higher in 2003 compared with the average for the recent
past (
Indicator 1.01
) but it remains at only about 1.8% of GDP and has fallen in
regard to the OST comparator group average. Business expenditure in UK
R&D has fallen across the public sector (
Theme 7
). The UK is 6
th
among G8
nations and ranks only 17
th
among the twenty-one OST comparator group
nations for which data are available (average around 2.3% GDP and
progressively increasing). The most rapid increase in GERD is for China, which
now exceeds all but the USA and Japan.
GOVERD and HERD refer to expenditure on R&D performed in the
GOVernment and Higher Education sectors. The sum of these (PUBERD)
forms the bulk of the publicly funded research base. This has generally
increased in real terms for the UK over the last decade, but dropped in 2001-02
and the increase now is less than the average for the OST comparator group.
The UK is 7
th
among G8 nations (
Indicator 1.02
). UK HERD has increased in
real terms (see table in
Indicator 4.07
) but at a slower rate than the OST
comparator group.
Theme 2 - Outputs (including people and publications)
The UKâs share of global research people and publication outputs are both
around 9% The UK is ranked 3
rd
globally in terms of people output and 2
nd
in
terms of papers, despite a recent fall in volume. This contrasts with its rank on
inputs (
Theme 1
).
The UKâs share of PhD awards (now over 9%) is similar to that of similar
countries for which data are available but much less than Germany (
Indicator
2.01
). In relation to population size, UK output has been in line with a general
trend whereas Germany is a high outlier. Looking to the future, growth is higher
in South East Asia, and some smaller and research competitive European
nations also outrank the UK on this measure (
Indicator 2.02
).
The UKâs share of world journal article publications (slightly less than 9%) was
overtaken by Japan in 2003 but has now moved back to 2
nd
behind the USA
(
Indicator 2.03
). A plateau in publication output over the last few years followed
a period of annual increases in many countries. The drop in UK share is likely
to be due to consolidation rather than contraction as it has not affected quality
(
Theme 3
).
Theme 3 - Outcomes (research recognition, citations; training and
research quality)
The UK is very efficient in terms of research outputs (as the contrast between
Theme 1 and 2
shows) and, generally, has fewer lower quality papers than
most of its competitors. The UK relative international research performance is
second behind the USA in terms of overall research recognition but it now
performs better than the USA in some specific areas though less well in others.
It has a good balance of strength in its performance, however, and has fewer
areas of relative weakness than some competitor nations. The rapid growth of
outputs from the China research base is likely to make the international picture
more dynamic over the next few years.
The UKâs total citation count is 2
nd
highest to the USA and has risen to just over
12% of world citations (
Indicator 3.01
). It remains ahead of all other OST
17
comparator group competitors. Germany was closing but has not improved its
relative position in recent years. China is rapidly increasing its global share, as
are some smaller nations and this has effects on larger countries.
The UK is clearly ahead of competitors (albeit 2
nd
to the USA) on citation count
in all of ten research fields including the new humanities field, but not in
mathematics, where it is 3
rd
, and in physical sciences and in engineering, where
it is 4
th
(
Indicator 3.02
). Its share of global citations has increased in most fields
but not in environmental sciences. It is becoming clear that there are some
inconsistent annual oscillations and that longer term trends are more important.
Citation growth among smaller nations does âattackâ the global share of larger
countries but the trend among individual small countries is erratic.
A measure of consistency in research strength is found in the frequency with
which a country appears in the top tier of nations, ranked by citation count for
each of nine main fields (
Indicator 3.03,
see also
Indicator 3.10
). The USA is in
the top tier in all fields. The UK performs strongly on this indicator and is in the
top three in seven of nine fields. Germany is the only other nation that performs
well across many fields. It had improved to place in the top three in six fields
but has slipped to only five fields. No nation outside the G8 is placed in the top
three on this measure in any field.
Citation count usually increases with numbers of papers published (source
count). So, although numbers of citations (
Indicators 3.01, 3.02
) give one
measure of esteem, it is also important to look at share of global citations
compared to share of sources so as to account for this scaling.
Indicator 3.04
shows that the UK has a positive difference between share of citations less
share of outputs in eight of ten research areas, particularly biological, physical
and environmental sciences. Other countries have multiple positive balances
but these are smaller in scale than for the UK or concentrated in niche areas.
Not all papers are subsequently cited, so their impact on research remains
obscure. About 35% of UK papers remain uncited in any 5-year period but this
proportion and the UKâs share of the world's uncited papers has declined â so
its performance has improved - compared to the recent past (
Indicator 3.05
).
The decline is seen across all fields; it has the fewest uncited among G8
countries in Biological, Environmental and Social Sciences and Humanities.
The UKâs ratio of cited papers to total outputs is greater than for the OST
comparator group average and improving in all fields (
Indicator 3.06
).
The UK has the 2
nd
greatest - and a growing - share (13.2%) of the worldâs
most highly cited papers (
Indicator 3.07
). Its average research impact
(measured by the number of citations per paper for these data) is ranked 8
th
but
the volume of this high impact material is greater than for countries ahead of
the UK such as Switzerland, the Netherlands and Scandinavia. UK share has
again increased whereas that for the countries ahead of the UK on impact,
including the USA, has fallen.
Research impact is usually indexed by the average ratio of citations to sources.
Indicator 3.08
shows that, for the research base as a whole, the UKâs average
impact increased progressively in the last ten years and has stepped up
between the recent past (the five years to 2003) and 2004. This has moved the
UK even closer to the USA and further ahead of most other G8 countries but
Germany has also improved and is a close competitor. The UK now stands 5
th
in the world where it was 7
th
.
How does the pattern of research impact vary between disciplines? It is
reviewed at the level of the ten main research fields in
Indicator 3.09
. The data
show marked improvements for the UK in some areas where it is now ranked
ahead of the USA in pre-clinical & health sciences and in biological sciences.
Although ranked further back among the OST comparator group as a whole, its
impact has improved progressively in many fields over the last decade. The UK
has almost closed the gap on the USA in mathematics, and has a marked
improvement in physical sciences and a sustained upturn in engineering.
Is research strength broad based and diverse or does it peak only in specific
disciplines? Average research impact and variation in quality between fields
both contribute to the overall performance of the research base (see also
Indicator 3.03
). A more consistent performance provides more even capacity
and hence enables flexibility in response to opportunities.
Indicator 3.10
links
information on field specific impact and considers both average across fields
and variation between fields for leading research countries. The UK has both a
strong average and a relatively even performance. The UK position has been
consistent between 2000 and 2004 whereas the USA has dropped back and
18
the Netherlands has improved its average only at the cost of more concentrated
excellence.
Does people quality match publication quality? It is not possible directly to
measure the quality of highly skilled people trained in the research base, but we
can test the broad correlation between relative training volume and relative
research impact. There is a correlation between training and impact for the UK
(
Indicator 3.11
) whereas this is absent for most G8 countries. UK PhD awards
are concentrated in the natural sciences, which is where
Indicator 3.09
suggests that the UK has relative research strength. Thus the UKâs
advantageous productivity pattern (
Theme 1, Theme 2
) is matched by a very
positive quality pattern.
Theme 4 â Productivity - financial (outputs and outcomes related to
inputs)
The UK is highly productive across the research base by the measures applied
in this analysis, as has been indicated above. However, its position is
challenged in some fields and is not as uniformly strong as it has been in the
past.
Highly trained people are an important product of the research base.
Indicator
4.01
shows relative productivity as PhDs awarded in relation to general spend
in the HE sector (HERD) and by main research areas (
Indicator 4.02
). This is
discussed above in terms of output (
Indicator 2.01
).
Total UK PhD awards are similar to most other G8 nations, though less than
Germany and much less than the USA. Compared to the level of spend on
R&D performed in the HE sector (HERD,
Indicator 4.01
) the UK produces
relatively more PhDs per unit spend than most OST comparator group nations
and is ranked 5
th
, above the EU average. The USA is ranked 11
th
. Other
measures do indicate that UK quality is high, so the assets of the research base
appear to be used effectively although it is, of course, not necessarily the case
that higher people output per unit research spend is a quality measure.
At the level of five main research fields, the UK ranks 3
rd
to Germany and the
USA in volume in the natural sciences but has moved ahead of Germany in
terms of PhDs per HERD, and is well ahead of OST group average.
Productivity has fallen back markedly in engineering. The correlation between
level of HERD and PhD awards is less clear in medical sciences, where the UK
is lower ranked and has productivity well below OST group average; Germany
remains well ahead. The UK is below the OST comparator group average in
the social sciences, although there are issues about data quality in this area,
but well ahead in the humanities (
Indicator 4.02
).
We can also index PhD output in terms of total PUBlic expenditure on R&D
(PUBERD,
Indicator 4.03
) and the UK then ranks 1
st
in the G8 ahead of
Germany and well ahead of the OST comparator group average against which
it is improving its position.
Research performance in relation to the general economy (GDP) is of less
direct significance. Although productivity is falling, the UK is 1
st
among G8
countries on citations per GDP (
Indicator 4.04
) but lies 7
th
overall behind
Switzerland and Scandinavian nations. Its performance is about 30% better
than the OST comparator group average whereas other G8 apart from Canada
are below that average.
The UK has dropped from 3
rd
to 5
th
in the OST comparator group in terms of
citations per unit GERD but has maintained its performance over the recent
past and is well ahead of other G8 nations. As would be predicted from
examination of
Theme 1
and
Theme 2
, it share of citations per share of spend
is better than twice the OST comparator group average (
Indicator 4.05
).
UK PUBERD has fallen compared to the OST comparator group average over
the last ten years.
Indicator 4.06
shows that citations per unit spend have
improved over the period, while the USA and Canada have fallen on this
measure and the UKâs lead within the G8 has therefore increased. The UK is
now 3
rd
to Switzerland and Denmark in the OST comparator group.
In terms of citations per unit spend as HERD (
Indicator 4.07
), the UK had led
the OST comparator group in the past and fell back slightly last year but is 2
nd
to Switzerland. Its performance has remained well ahead of OST comparator
19
group average and improved on other G8 nations while the USA has dropped.
At the level of OECD fields (
Indicator 4.08
) the UK is now 2
nd
to the Netherlands
in natural sciences. Denmark has moved further ahead in engineering, but the
USA has dropped behind the UK. In the social sciences the UK has improved
its performance relative to the OST comparator group average but the USA,
though declining, remains far ahead on this measure. In the humanities the UK
has a clear lead, although the significance of this is open to interpretation.
Theme 5 â Productivity - labour (outputs and outcomes related to other
measures)
The UK is highly productive in terms of labour productivity.
Skilled people are a key contribution to future resources for the knowledge
economy. The UK is 2
nd
in the G8 but has gained further on Germany in terms
of PhDs awarded per researcher (
Indicator 5.01
). Only Switzerland and Spain
are more productive. The UK has maintained its level of output during a period
when the USA slipped back.
On relative productivity (papers published per researcher,
Indicator 5.02
) and
effectiveness (citations acquired per researcher,
Indicator 5.03
) the UK has a
strong lead position in G8 countries (excepting anomalous data for Italy) and
has been 3
rd
overall behind Switzerland and the Netherlands. In the last year,
UK paper productivity has fallen slightly compared with the recent past
(previous five years) but citation acquisition has improved suggesting that
people have produced fewer but better papers.
Theme 6 - People
The UK has a weaker availability of highly skilled people with research training
â in relation to population and to the workforce as a whole - than its
competitors. However, there may be some shift in the skills balance among
research personnel.
We can index each countryâs researchers and R&D personnel in relation to the
general population and the national workforce. As we move towards an
increasingly technology and knowledge based economy, these indicators may
take on added significance as a measure of national capacity to absorb and use
knowledge based opportunities. The Scandinavian countries are strong in this
regard.
In the UK, only 0.3% of the population as a whole would be classified as a
researcher (
Indicator 6.01
). This is less than the OST comparator group
average, where the UK is ranked 15
th
, and is lowest in the G8 where no nation
is much higher than 0.5%. The USA has reached 0.5% and is consistently
improving whereas there has been little change in the last decade for the UK.
Frequency of R&D personnel in the population is higher, typically 0.45%
(
Indicator 6.03
) but the UK is again below OST comparator group and EU
average and is the lowest except for Italy in the G8.
If comparison is made to the size of the national labour-force (
Indicator 6.02,
6.04
), the UK rank has slipped to 17
th
in the OST comparator group for
researchers (0.6% of labour-force) and 15
th
for R&D personnel (0.88% down
from 0.92% last year). In both cases, the UK is the lowest ranked except Italy
among G8 nations.
Change in the structure of research personnel in the labour-force is reflected in
a measure of researchers as a proportion of R&D personnel (
Indicator 6.05
).
This may be an indication of increasing professionalisation of research work.
The UK now ranks 8
th
in the OST comparator group, ahead of Denmark and
Sweden, and 2
nd
to Japan in the G8. Its share of researchers as a proportion of
R&D personnel is increasing faster than average.
Theme 7 - Business expenditure
The UK has performed relatively well in the past in terms of business
investment in the HE research base as defined by OECD. Recent data suggest
that this investment is not being sustained.
20
The Business Enterprise (BE) sector funds some of the research performed in
the public sector. While business investment options are obviously affected by
the general economic cycle, increases in BE within R&D may reflect relevance
of public research to commercial objectives and confidence in the ability of the
research base to produce returns on investment in one form or another.
Total PUBlic sector Expenditure on R&D is PUBERD. Business expenditure on
R&D that is performed in the public sector (BE-PUBERD) is correlated with total
spend (
Indicator 7.01
). The UK has performed well on this measure but has
dropped to 8
th
overall and is now only 3
rd
in the G8 group. Among generally flat
trends UK BE-PUBERD appears to be decreasing but it is difficult to forecast.
Business spend as a proportion of GOVERD (
Indicator 7.02
) is less well
correlated. The UK has had a strong position and is still ranked 4
th
in the OST
comparator group but BE as a proportion of GOVERD is falling from twice the
OST comparator group average to less than one-third better than that. UK
business spend in regard to Private Non-Profit R&D (PNPERD,
Indicator 7.03
)
is typical of the OST comparator group. There is little spread among the G8
apart from an exceptionally high result for Japan.
BE spending as a part of HERD (
Indicator 7.04
) increased in the UK in the mid-
1990s but the UK has now fallen behind both OST comparator group and EU
average and has fallen to 10
th
in rank. The UK profile for business spend as a
proportion of HERD has consistently tracked the USA but there has been
growth in Germany and some smaller countries whereas in other countries
there have been marked falls. At the level of four OECD fields, a marked
decrease for the UK is in the natural sciences â an area of strong performance -
whereas there has been an improvement in the social sciences (
Indicator 7.05
).
Note that HERD as defined by the OECD covers more than just universities and
the basket may vary between countries.
21
Indicator summary pages
The body of this report is a page by page summary of the detailed quantitative analyses for each indicator. The layout for each page follows a similar pattern:
â˘
Report on indicator and the headline results.
â˘
Table of key results (actual values and ranked performance among comparators) for the latest year for which data are available and the average value for the
previous 5 years. The Table also shows current change in performance (ratio of activity between latest and recent, or difference in rank) and then ranks that
change among the OST comparator group and G8 competitors to reflect the relative as well as absolute shift in current UK performance.
â˘
Charts of data for UK and competitors (usually G8 plus select others) showing trends.
Additional explanatory notes are in the
Background
section at the end of this document.
Description of performance indicator
2 Citations in main research fields (SUoAs): national count and share of OST comparator group total
Condition signalling improvement
Increased national count and share
Biological Sciences
Average
1997 - 2001
Actual
2002
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
168,098.8
10,323
0.0614
OST average citation count
65,102.2
3,814
0.0648
UK citation count - Rank among OST group
2
2
0
UK citation count - Share of OST group
0.100
0.104
1.046
9
2
UK citation count - Share of world
0.123
0.133
1.083
Note that this table ranks both performance and change in performance
Ranked change in perf
Citation share among OST group for G8 nations (USA omitted) for SUoA Biological Sciences
0
0.02
0.04
0.06
0.08
0.1
0.12
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Year
C
o
u
n
tr
y'
s sha
re of
t
o
ta
l
O
S
T
c
o
mpar
a
tor
gr
oup
ci
ta
ti
o
n
s w
ith
in
S
U
o
A
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Indicator Headline
In Biological Sciences, the UK citation
share overall is 2nd to the USA.
The UK's rank citation count remained
2nd highest to the USA for both 1997-
2001 and 2002 periods (difference = 0 -
remaining the same).
The change in UK citation share ratio of
1.046 between periods 1997-2001 and
2002 is second among G8 nations to
Germany which had a ratio of 1.092,
indicating that Germany's citation share is
growing at a slightly faster rate than the
UK.
The UK ranked position of 9th on change
in OST citation share is because other
countries, notably Poland and Spain in
this case, are increasing their citation
share at a faster rate than the UK - even
though their citation counts are smaller
than the UK. The average citation share
ratio of the 8 countries ahead of the UK is
1.331.
Between the periods 1997-2001 and 2002
the USA has shown a decrease in citation
share by a ratio of 0.978.
Although there were reductions in UK
citation share in both 1996 and 2001
there is a slight but progressive rise
across the period.
Graphs may omit USA &
EU 15 if inclusion would
distort vertical axis
Key results for UK (or OST group)
actual and ranked performance in
latest data year and average for
previous 5 years
Indicator identification and
description
One or more graphs illustrating
data scatter and performance
trend of UK and OST group or G8
Text may include commentary on additional issues arising from data or indicators
Change in UK performance
ranked among total OST group
and G8
Change (ratio or difference)
between last year and recent
Headline outcome of
analyses describing UK
status and recent
performance
Description of performance indicator
1.01 GERD relative to GDP (R&D intensity)
Condition signalling improvement
Increased proportion of R&D specific spend
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK GERD as a percentage of GDP - Actual
1.85
1.84
0.99
OST average GERD as percentage of GDP
2.15
2.29
1.06
UK GERD per GDP - Rank
16
17
-1
UK GERD per GDP / OST average GERD per GDP
0.86
0.80
0.93
GERD as a percentage of GDP - G8 nations (Russia not included)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
GE
RD as a percentage of GDP
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
EUROPEAN
UNION
Data: OECD (MSTI)
Indicator Headline
The relative size of GERD (the most general
measure of the share of total GDP that is spent
on research and development, R&D) is an
indication of the relative research intensity of
the economy as a whole. GERD is an input
measure rather than a measure of research
performance. Nonetheless it is an important
contextual indicator alongside output measures
(Theme 2).
UK GERD makes up about 4% of the
comparator group total (this contrasts with UK
outputs which make up 9% of the comparator
group total outputs - see Indicator 2.03). In
absolute terms, adjusted for PPP, it is just over
10% of that of the USA and is smaller than that
of both France and Germany. In relative terms,
it is a lower percentage of GDP at 1.84% (as
measured in million 2000 $ - constant prices
and PPP, this compares with 1.89% when
measured in million current PPP $) than for the
OST group average and has fallen in recent
years by comparison to that average. It has
also fallen below the average for the EU15
nations.
The most rapid rate of increase in GERD is that
for China, for which no time series data were
available until this year. Although it has only
reached 1.3% of GDP, this is a relative
doubling over the decade and a fourfold
increase in absolute value. China's GERD
($80Bn PPP 2000) now exceeds all but the
USA and Japan and will soon approach 50% of
the EU15 total ($190Bn). We draw attention to
increasing China research activity on a number
of indicators.
cont./
22
Description of performance indicator
1.01 GERD relative to GDP (R&D intensity)
Condition signalling improvement
Increased proportion of R&D specific spend
GERD relative to GDP - OST comparator nations, 2003
EU15
TWN
SKO
SGP
ISR
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
USA
UK
1000
10000
100000
1000000
100000
1000000
10000000
100000000
GDP (2000 $M PPP)
GE
RD (2000 $M
P
P
P
)
Data: OECD (MSTI)
Check Background section of the report for country short codes
Indicator Headline cont.
GERD is progressively increasing for the
comparator group as a whole, although the
increase to 2003 was the smallest for ten
years.
Sweden, Finland and Switzerland all have
GERD over 2.5% as a share of GDP as do
Israel and South Korea. Most other G8 nations
also have a flat or declining profile, although
Italy has maintained its GERD and Canada has
moved ahead of the UK.
The figures in this report are affected by the
inclusion of some countries where data were
previously unavailable, including China. All the
data are now rebased to year 2000 purchasing
power parity instead of 1995.
GERD includes both public and private sector
spend and is thus related to a country's
industrial R&D structure, the balance of
different industries and their investment in
research. It is only a partial measure as
regards the influence of domestic policy on
R&D. However, the tendency of industry to
invest is likely to be affected by public policy
and its relationship to both general
competitiveness and the specific value of the
public sector research base.
Specific data on business expenditure (BERD)
are considered further in Theme 7.
23
Description of performance indicator
1.02 Publicly performed R&D (PUBERD) as a proportion of GDP
Condition signalling improvement
Increased proportion of R&D specific spend
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK PUBERD as a percentage of GDP - Actual
0.60
0.57
0.96
OST comparator group average PUBERD as percentage of GDP
0.68
0.72
1.07
UK PUBERD per GDP - Rank
15
18
-3
UK PUBERD per GDP / OST comparator group av'ge PUBERD per GDP
0.88
0.79
0.90
PUBERD as a percentage of GDP - G8 nations (Russia not included)
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
P
U
BE
RD as a percentage of GDP
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
EUROPEAN
UNION
Data: OECD (MSTI)
Indicator Headline
PUBERD is the sum of GOVERD (expenditure
on R&D performed in the Government sector)
and HERD (R&D in the higher education
sector). This thus accounts for the bulk of
public sector R&D expenditure. This is an
input measure, not an index of performance.
UK PUBERD is only just over 4% of the OST
comparator group total. In terms of PUBERD
relative to GDP, the UK has dropped three
places since 1998 and its PUBERD has fallen
slightly while the group average has increased.
It is therefore now at 80% of group average
whereas it had been ahead of the OST
comparator group average in 1995.
The UK ranks last in the G7 group on PUBERD
relative to GDP. It slipped behind the USA in
2001 and Italy last year. It is well below the
EU15 average.
The main trend over the period has for UK
PUBERD to be relatively static, as it has for
France and Germany. Elsewhere, there have
been significant increases for the USA,
Canada, Scandinavian nations, China, and
other Asian nations.
China is included in this data analysis for the
first time. It has clearly the fastest rate of
increase and its PUBERD has trebled in value
since 1995, increasing by 50% as a share of
GDP. It now stands at $30Bn (2000 PPP)
slightly les than half the total for the EU15
($65Bn). China is on track to overtake the UK
on PUBERD/GDP (this indicator) in the next
few years.
24
Description of performance indicator
1.02 Publicly performed R&D (PUBERD) as a proportion of GDP
Condition signalling improvement
Increased proportion of R&D specific spend
PUBERD compared to GDP - OST comparator nations, 2003
EU15
TWN SKO
ISR
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
USA
UK
1000
10000
100000
100000
1000000
10000000
100000000
GDP (2000 $M PPP)
P
U
BE
RD (2000 $M
P
P
P
)
Data: OECD (MSTI)
Check Background section of the report for country short codes
25
Description of performance indicator
2.01 Number and share of OECD PhD awards
Condition signalling improvement
Increased count and increased share by comparison with competitors
Average
1998 - 2001
Actual
2002
Ratio or
difference
UK PhDs count - Actual
12012
14232
1.18
OST comparator group average PhDs count
8948
9372
1.05
UK PhD count - Ranked performance
3
3
0
UK PhD count - Share OECD comparator nations (%)
8.64
9.76
1.13
Note that this table ranks both performance and change in performance
G8 nations: PhD share among OST comparator group (USA & Canada omitted; no data for Russia)
0
4
8
12
16
20
24
28
32
36
1999
2000
2001
2002
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group PhD
s
UK
USA
FRANCE
GERMANY
ITALY
JAPAN
Data: OECD Education Database
UK - HESA data collection
procedure changed in
2001 (see headline note)
Indicator Headline
There are only 5 years' data on PhD awards in
the OECD Education database. There are also
some data gaps among the OST comparator
group countries. Data are present for G7
countries, although recent years are missing
for Canada. The data have proved stable and
the trends appear fairly level.
The UK's count and share have risen over the
period although this has not affected its ranked
position. Its share of PhD awards is similar to
its share of publication outputs (Indicator 2.03)
and much greater than its share of input
funding (Theme 1).
The USA is the major producer of PhD
students with a slightly declining 44,000 annual
out-turn, which is about 30% of the OST
comparator group total. This compares with
about 50% for the EU15.
Germany (16% down from 18%) ranks 2nd to
the USA, with the UK 3rd at just under 10% of
the group total. A number of countries do not
yet return OECD data on this indicator,
including China and India.
This is an important indicator because people
are a key output from the research base.
Absolute numbers indicate sustainable
capacity, but the share of the group total is
also valuable as a comparator with other
measures of input and output. Highly skilled
people reflect the capacity to make use of
knowledge, where other indicators indicate its
generation and impact. The numbers of people
available to the public and private sector
research base may also be critical to economic
innovation.
26
This page is left blank intentionally
27
Description of performance indicator
2.02 PhDs awarded per head of population
Condition signalling improvement
Increased ratio
Average
1998 - 2001
Actual
2002
Ratio or
difference
UK PhDs awarded per 1000 population - Actual
0.20
0.24
0.04
OST comparator group average - PhDs awarded per 1000 population
0.20
0.21
0.01
UK PhDs awarded per 1000 head of population - Rank
6
5
1
UK PhDs awarded per 1000 population / OST comparator group average
PhDs awarded per 1000 population
1.01
1.12
0.10
PhDs output per capita, G8 nations (no data for Russia)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
1998
1999
2000
2001
2002
Year
P
h
Ds awarded per 1000 popul
ati
on
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
EUROPEAN UNION
Data: OECD (MSTI), OECD Education Database
UK - HESA data collection
procedure changed in 2001
(see headline note)
Indicator Headline
The UK is ranked 2nd behind Germany within
the G8 in terms of PhD awards per head of
population (Germany produced 15% more
PhDs per head of population than the UK,
0.29 c.f. 0.24 PhDs per 1000 population), and
5th amongst all OST comparator group
nations for which there are data. The UK
remained close to the OST comparator group
and EU average throughout the period. Its
apparent improvement in the last two years is
attributable partly to data revision and partly to
cuts elsewhere.
Outside the G8, the OST comparator group
rankings are led by Sweden (average 0.39
PhDs per 1000 population), Switzerland
(0.38), and Finland (0.35). There are no data
for China or India. The EU average is 0.19
and the USA average is 0.15, so Europe is
producing more PhDs and many of these
come from leading research nations.
People are a key output from the research
base and form a key input of highly skilled
people to the workforce. The numbers of
PhDs awarded relative to population size as a
whole is a broad measure of the relative
training capacity and productivity of different
countries. Comparison between trainee
output and the size of the training population is
made in indicator 5.1.
cont./
28
Description of performance indicator
2.02 PhDs awarded per head of population
Condition signalling improvement
Increased ratio
PhDs awarded per head of population - OST comparator group nations, 2002
EU15
SKO
ISR
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
JAP
ITA
GER
FRA
USA
UK
100
1000
10000
100000
1000
10000
100000
1000000
10000000
Population in thousands (log scale)
P
h
Ds awarded (l
og scal
e)
Data: OECD (MSTI), OECD Education Database
Check Background section of the report for country short codes
Indicator headline cont.
This indicator is relevant to both this 'outputs'
theme and to the 'people' theme as discussed in
the thematic commentary to this report.
UK data in 2001 were affected by a change in
HESA data collection (see Background) but have
now settled at a new level. There are no Canada
data after 2000. There are now 5 yearâs data for
PhD awards available on the OECD Education
database. Trend analyses using these data have
not been as informative as for some other
indicators but the pattern is now becoming clear
and the levels of reported activity are evidently
stable.
29
Description of performance indicator
2.03 Number and share of world publications
Condition signalling improvement
Increased count and increased share by comparison with competitors
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK publication count - Actual
70348
69419
0.99
32219
33783
1.12
UK publication count - Ranked performance
3
2
1
8.41
7.90
0.94
22
4
UK publication count - % share of world
9.26
8.81
0.95
22
4
Note that this table ranks both performance and change in performance
Sum of comparator nations exceeds actual world total.
World total is corrected for international co-authorship, which creates duplication in OST total.
Ranked change in performance
UK publication count - % share of OST comparator
group
OST comparator group average publication count
Publication share among OST comparator group for G8 nations (USA omitted)
0
1
2
3
4
5
6
7
8
9
10
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
P
ercentage share of total
OS
T com
p
arator
group publicat
ions
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Indicator Headline
Publication share is a measure of research
activity relative to the rest of the world.
Volume is a scale measure of significance and
changes in share are an important indicator of
performance.
The UK is ranked 2nd behind the USA in
national share of global publications for 2004,
the most recent year. This means it has
regained its position relative to Japan.
The UKâs publication count has not increased
markedly, however, since it reached 70,000
outputs per year in 1999. There is some
periodicity in UK output associated with
assessment cycles. From 2001 there may be
some shift from greater volume to better quality
(see later indicators on citations).
Since 1998, the 5 year rolling average growth
for the OST comparator group has been
around 3%. In that period, the UK publication
total has been flat and therefore growth is
consistently less than the group as a whole,
but the USA's annual growth has also fallen
from 1% to 0.2%, and Japan from 5.4% to
2.3%. The UK ranked change in performance
(publication share 2004 compared with 1999-
2003 average) is 22nd among all the nations in
the OST comparator group, up from 25 last
year (France, Germany and Japan rank lower).
The UK had just under 8% of OST comparator
group publications, or 8.8% of global
publications, for 2004 [this difference is
because the sum of the OST comparator group
publications includes some duplication so the
individual national shares of the total are
uniformly depressed compared to true global
values (see methodology notes in
Background)].
30
Description of performance indicator
2.03 Number and share of world publications
Condition signalling improvement
Increased count and increased share by comparison with competitors
Indicator headline cont.
China has more than trebled its output from
13,500 publications in 1995 to 46,000
publications in 2004 and is ranked 2nd on
growth (Iran is ranked 1st and has increased
nearly ten-fold). Italy (at 10) and Canada (13)
are the only G8 counties ranked better than
20th on performance change. This indicates
that the smaller countries in the OST
comparator group are increasing their
publication share at a faster rate than the G8.
The consequence of the relative changes in
output, and the growth of new research
economies, is that the UKâs outputs will
progressively index less than in the past as a
share of world, although it may be greater
absolutely. It is inevitable that UK share will
continue to fall if growth elsewhere is
sustained, but this need not affect quality.
Chinaâs total is now almost the same size as
France at just over 46,000. Together with
South Korea (19,000), Taiwan (13,000) and
Singapore (5,000), this Asia-Pacific group now
exceeds any one European country in terms of
publication volume. The opportunities for
international collaboration and the implications
for the balance of research endeavour are
likely to be very significant.
In 1998 the EU15 total outputs passed that of
the USA and now clearly exceeds it (2004: EU
= 37.9% of world, USA = 33.6%).
Publication share among fast-growing nations (Russia and China on second axis)
0
0.5
1
1.5
2
2.5
3
3.5
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
P
ercentage share of total
OS
T com
p
arator
group publicat
ions
0
1
2
3
4
5
6
SPAIN
INDIA
SOUTH
KOREA
BRAZIL
TAIWAN
POLAND
SINGAPORE
IRAN
RUSSIA
CHINA
31
Description of performance indicator
3.01 Number and share of world citations
Condition signalling improvement
Increased count and increased share by comparison with competitors
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
555364
33426
0.06
OST comparator group average citation count
232286
13955
0.06
UK citation count - Ranked performance
2
2
0
UK citation count - % share OST comparator group
9.14
9.21
1.01
16
4
UK citation count - % share of world
11.52
12.23
1.06
16
4
Note that this table ranks both performance and change in performance
Sum of comparator nations exceeds actual world total.
World total is corrected for international co-authorship, which creates duplication in OST total.
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA & EU15 omitted)
0
1
2
3
4
5
6
7
8
9
10
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Percent
age share of
t
o
ta
l OST com
p
arat
or group
cit
at
ions
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson
ISIÂŽ National
Science Indicators
Indicator headline
Relative citations counts are the most general
measure of research excellence because
esteem is related to citation frequency. The
more times an article is cited by others, the
greater its influence is deemed to be.
The UK has improved on this indicator relative
to 2003. There is significant continuing growth
in China.
The UK is ranked 2nd to the USA in terms of
national share of global citations for 2004, the
most recent year. It has about 12% of global
citations (but 9% of OST comparator group for
reasons noted below).
UK rank has remained 2nd to the USA both as
an average for the recent period (1999-2003)
and for 2004 (difference in table = 0). Citation
counts are fewer in more recent years, but the
rate at which the count falls may vary between
nations.
UK citation share rose slightly between the
period 1999-2003 and 2004. The change (ratio
= 1.01) is better than that of other leading
research nations. Germany had been catching
up to the UK but has now plateaued while
France and Japan have fallen back. The
EU15âs combined share has declined slightly,
but the USA is also dropping â by about 0.5%
per year for several years.
One of the biggest increases is that for China,
which has accelerated to a more than four-fold
change over the decade (from 0.92% to
3.78%) and approaches the citation volume of
Italy (but the publication volume of France -
see 2.03). Ranked 18th in 1995 it now ranks
8th by volume.
cont./
32
Description of performance indicator
3.01 Number and share of world citations
Condition signalling improvement
Increased count and increased share by comparison with competitors
Indicator headline cont.
India is growing too, but much more slowly and
has not changed its rank position. Some
smaller countries, such as Iran (0.02 to
0.15%), Singapore (0.15 to 0.49%) and South
Korea (0.44 to 1.68%), have substantially
increased their citation share over the period
but their citation counts remain small
compared to G8 competitors.
Total citation count might seem to provide the
simplest measure of recognition, but total
citation count is dependent on output or source
volume. An increase in citations might
therefore be due to an increase in output rather
than esteem. World levels of publication have
also risen over the years.
A measure that helps to benchmark national
citation counts is to consider not numbers but
the share that each nation has of the world
total. Even this may be misleading, however,
because some nations that have recently
become more scientifically active are
increasing their share of world cites. It may
therefore be appropriate to look at annual
change in global share for each country by
comparison with close competitors. Where all
drop in global share â because of emerging
research nations â those that drop less than
others are clearly suffering less from
competition elsewhere.
The analysis is carried out by year for the
national research system across all subjects.
The sum of OST comparator group citations
includes some duplication (see notes on
methodology) so individual national shares of
total are uniformly depressed compared to true
global values.
33
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Clinical Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
415,732
22,616
0.05
OST comparator group average citation count
158,355
8,342
0.05
UK citation count - Rank among OST comparator group
2
2
0
UK citation count - % share of OST comparator group
10.05
10.43
1.04
14
2
UK citation count - % share of world
12.82
13.86
1.08
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Clinical
0
2
4
6
8
10
12
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group cit
at
ions wit
h
in SU
oA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Indicator Headline
See 3.01 for background and 3.03 for
summary.
In Clinical Sciences, the UK citation share
overall is 2nd to the USA. Its share of world
citations has improved compared to the
previous year.
The UK's ranked citation count remained 2nd
highest to the USA for both 1999-2003 and
2004 periods.
UK citation count reduced at a ratio of 0.05
between the periods 1999-2003 and 2004.
This is very similar to the OST comparator
group average. (Citations are always fewer in
more recent years).
The change in UK citation share ratio between
periods 1999-2003 and 2003 (1.04) is second
only to Russia within the G8 nations.
Germany's citation share had been growing at
a slightly faster rate than the UK but this trend
seems to have levelled off while some other
nations have declined slightly.
The USA has shown little change in citation
share between the periods 1999-2003 and
2004 and seems to be stabilising after a long
period of reducing share.
The UK ranked position of 14th on change in
OST comparator group citation share is
because other countries, notably Poland, India,
Singapore and China in this instance, are
increasing their citation share at a faster rate
than the UK - although their citation counts are
smaller than the UK. Chinaâs rapid growth
seems to be levelling off in this field.
Data: Thomson ISIÂŽ National Science Indicators
34
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Pre-Clinical and Health-Related Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
54,882
3,341
0.06
OST comparator group average citation count
18,441
1,051
0.06
UK citation count - Rank among OST comparator group
2
2
0
UK citation count - % share of OST comparator group
11.31
12.23
1.08
12
3
UK citation count - % share of world
14.66
16.60
1.13
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Pre-Clinical and
Health-Related Sciences
0
2
4
6
8
10
12
14
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group cit
at
ions wit
h
in SU
oA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator He
adline
See 3.01 for background and 3.03 for
summary.
In the Pre-Clinical and Health-Related
Sciences, the UK citation share overall is 2nd
to the USA.
There is a considerable gap between UK
citation share 1995 - 2004 and the next closest
competitor (Germany). Consequently, the
UK's rank citation count remained 2nd highest
to the USA for both 1999-2003 and 2004
periods (difference in rank = 0).
UK citation count reduced at a rate similar to
the OST comparator group as a whole over the
period 1999-2004.
The improvement in UK citation share ratio of
1.08 between periods 1999-2003 and 2004
places it third among G8 nations behind
Russia and Japan, indicating that these
nations citation share are growing at a faster
rate than the UK.
The USA had shown a decrease in citation
share over the period since 1995 but recovered
slightly in 2004.
The UK ranked position of 12th on change in
OST comparator group citation share is
because other countries, notably China, India
and South Korea in this case, are increasing
their citation share at a faster rate than the UK -
even though their citation counts are less than
the UK. The average citation share growth
ratio of the countries ahead of the UK is +1.44.
35
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Biological Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
208,262
11,724
0.06
OST comparator group average citation count
78,877
4,496
0.06
UK citation count - Rank among OST comparator group
2
2
0
UK citation count - % share of OST comparator group
10.14
10.03
0.99
14
4
UK citation count - % share of world
14.36
14.58
1.02
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Biological Sciences
0
2
4
6
8
10
12
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group cit
at
ions wit
h
in SU
oA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
See 3.01 for background and 3.03 for
summary.
In Biological Sciences, the UK citation share
overall is 2nd to the USA. The UK's rank
citation count remained 2nd highest to the USA
for both 1999-2003 and 2004 periods
(difference in rank = 0). The overall picture for
the G8 has changed little this year.
The marginal change in UK citation share ratio
of 0.99 between periods 1999-2003 and 2004
is fourth among G8 nations behind the USA,
Canada and Russia. Germany, which is third
to the UK and USA in citation share, had a ratio
of 0.98, indicating that its citation share is
affected by similar factors. Canada has
recovered slightly from a period of decline.
The UK rank of 14th on change in OST
comparator group citation share is because
other countries, notably China (three-fold over
ten years), Singapore and Iran in this case, are
increasing their citation share at a faster rate
than the UK - though their citation counts are
smaller than the UK. China (1.6%) now
exceeds Belgium and Denmark and is of
similar scale to Sweden.
The USA has stabilised after a period of
decrease in citation share between 1995 and
2000 and has remained level at just under 40%
since 2001.
Although there are fluctuations in UK citation
share across the period, the trend is for a
consolidated performance at just over 10% of
OST group, but its share of world has actually
improved marginally over the ten year period.
This may reflect increasing UK collaboration
within the G8.
36
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Environmental Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
37,647
2,265
0.06
OST comparator group average citation count
13,554
920
0.07
UK citation count - Rank among OST comparator group
2
2
0
UK citation count - % share of OST comparator group
10.50
9.47
0.90
21
7
UK citation count - % share of world
13.40
13.10
0.98
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Environment
0
2
4
6
8
10
12
14
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group cit
at
ions wit
h
in SU
oA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
See 3.01 for background and 3.03 for
summary.
In Environment the UK citation share overall is
2nd to the USA. The UK's rank citation count
remained 2nd highest to the USA for both 1999-
2003 and 2004 periods (difference = 0).
The change in UK citation share ratio of 0.90
between periods 1999-2003 and 2004 is 7th
among G8 nations and reflects a progressively
declining share since 2000. Germanyâs share
also decreased where it had been growing and
seemed likely to overtake the UK. Canada is
now improving markedly.
The UK ranked position of 21st on change in
OST comparator group citation share is
because other countries, notably South Africa,
Poland and Iran in this case, are increasing
their citation share at a faster rate than the UK -
even though their citation counts are only a
tenth of the UK. The average citation share
ratio of the countries ahead of the UK is 1.1.
The USA has shown a marked decline in
citation share by over 2%, or a ratio of 0.95
between the periods 1999-2003 and 2004.
China citation share has been level at 2.5%
since 2001.
37
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Mathematics
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
5,743
339
0.06
OST comparator group average citation count
2,852
156
0.05
UK citation count - Rank among OST comparator group
3
3
0
UK citation count - % share of OST comparator group
7.63
8.36
1.10
6
1
UK citation count - % share of world
9.03
10.02
1.11
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Mathematics
0
1
2
3
4
5
6
7
8
9
10
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group cit
at
ions wit
h
in SU
oA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
See 3.01 for background and 3.03 for
summary.
In Mathematics, the UK citation share overall
has been 3rd to the USA and (since 1998)
France. It has significantly improved its
position in 2004, with other G8 nations
generally dropping back in share. However, it
is now 3rd to the USA and China which has
moved up to 2nd - with 9.6% of OST group
citations - from 8th in 1995.
The change in UK citation share ratio of 1.10
between periods 1999-2003 and 2004 is 1st
among G8 nations Its position of 6th on
change in OST comparator group citation
share is a marked change from recent years (it
had been 15th last year). Only Belgium and
China were substantially better in improvement.
The USA had suffered a decreasing citation
share since 1995 but this pattern appears to
have stabilised between the periods 1999-2003
and 2004.
The UK's citation share has been more variable
over the period than for some other fields.
From the mid-1990s UK share has fluctuated
whilst France had increased steadily to
overtake the UK. This is an area which needs
further monitoring to verify that the UK's
changed position is real and sustained rather
than a statistical aberration.
38
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Physical Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
154,044
11,671
0.08
OST comparator group average citation count
77,813
5,788
0.07
UK citation count - Rank among OST comparator group
4
4
0
UK citation count - % share of OST comparator group
7.58
7.76
1.02
12
2
UK citation count - % share of world
10.65
12.09
1.13
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Physical Sciences
0
2
4
6
8
10
12
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group cit
at
ions wit
h
in SU
oA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
See 3.01 for background and 3.03 for
summary.
In Physical Sciences, the UK citation share
overall is 4th to the USA, Germany and Japan.
The UK's rank citation count remained 4th
highest to the USA, Germany and Japan for
both 1999-2003 and 2004 periods (difference =
0). The UK appears to have lagged
significantly on these nations in the recent past,
but its relative impact has been well ahead of
Japan and moved ahead of Germany. There is
sustained evidence of slow but continuing
growth in UK citation share, and it is ranked
2nd on change among G8 nations. It has
moved well ahead of France (at 5th in the
group) and seems likely to overtake Japan.
China has moved up from 13th to 6th globally
on share in this area. moving ahead of Italy
with 4.8% of group citations. The UK ranked
position of 12th on change in OST comparator
group citation share is because other countries,
notably China, India and Poland in this case,
are increasing their citation share at a faster
rate than the UK.
USA share has dropped from over 35% to less
than 29%. The EU15, at 38%, has broadly
maintained its total share over the last ten
years.
39
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Engineering
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
58,774
3,082
0.05
OST comparator group average citation count
32,026
1,732
0.05
UK citation count - Rank among OST comparator group
4
4
0
UK citation count - % share of OST comparator group
6.90
6.85
0.99
17
4
UK citation count - % share of world
8.51
8.69
1.02
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Engineering
0
2
4
6
8
10
12
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group cit
at
ions wit
h
in SU
oA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
See 3.01 for background and 3.03 for
summary.
In Engineering, the UK citation share overall is
4th to the USA, Japan and Germany.
The UK's rank of citation count remained 4th
highest to the USA, Japan and Germany for
both 1999-2003 and 2004 periods (difference =
0).
While the UK's ranked position is unchanged, a
trend of falling UK share of world citations over
the period from 1994 has been halted and
some uplift is emerging. A growing gap
between the UK and Germany has narrowed.
The flat UK citation share ratio is average
among G8 nations, and ahead of USA, Japan
and France.
China has moved up to 5th in OST group share
with 6.3% of citations, up from 11th in 1995,
and thus just behind the UK.
The UK ranked position of 17th on change in
OST comparator group citation share is
because other countries, notably Singapore,
China, Denmark and Finland in this case, are
increasing their citation share at a faster rate
than the UK - even though their citation counts
are smaller than the UK.
UK average research impact in Engineering
has improved and leading units are
outstanding. We have previously noted the
significant diversity in performance within the
UK research base.
40
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Social Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
9,703
600
0.06
OST comparator group average citation count
3,401
212
0.06
UK citation count - Rank among OST comparator group
2
2
0
UK citation count - % share of OST comparator group
11.08
10.87
0.98
12
5
UK citation count - % share of world
11.98
11.90
0.99
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Social Science
0
2
4
6
8
10
12
14
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group cit
at
ions wit
h
in SU
oA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
See 3.01 for background and 3.03 for
summary.
Bibliometric data for the Social Sciences
cannot be compared directly with other
disciplines (see Background). Recent work
has shown that European comparisons must
be made with caution because Anglophone
journals are over-represented.
In the Social Sciences, the UK citation share
overall is 2nd to the USA. It remained 2nd
highest to the USA for both 1999-2003 and
2004 periods (difference = 0).
The change in UK citation share ratio between
periods 1999-2003 and 2004 is fifth among G8
nations and follows an earlier period of rapid
UK expansion. Evidence from research
assessment suggests that journal publication
is becoming increasingly important in this field.
The UK ranked position on change in OST
comparator group citation share is only 12th.
A number of other countries, notably Germany
but also Belgium and India in this case, are
increasing their citation share at a faster rate
than the UK - even though their citation counts
are smaller than the UK.
The USA has shown a decrease in citation
share of over 10% between 1995-2004. This is
partly accounted for by substantial shifts in the
database coverage to include a wider range of
European journals. This will make the data
more valuable in the future although it makes
current trends less clear.
China has less than 1% of citations in the OST
comparator group.
41
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Business
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
4,683
246
0.05
OST comparator group average citation count
1,625
76
0.05
UK citation count - Rank among OST comparator group
2
2
0
UK citation count - % share of OST comparator group
11.19
12.45
1.11
13
3
UK citation count - % share of world
13.13
15.25
1.16
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Business
0
2
4
6
8
10
12
14
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
C
ount
ry's percent
age share of
t
o
ta
l OST
com
p
arat
or group cit
at
ions wit
h
in SU
oA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
See 3.01 for background and 3.03 for
summary.
Bibliometric data for Business, like the Social
Sciences, should be treated with caution and
cannot be compared directly with other
disciplines.
In Business, the UK citation share overall is
2nd to the USA.
The UK's rank of citation count remained 2nd
highest to the USA for both 1999-2003 and
2004 periods.
The UK ranked position of 13th on change in
OST comparator group citation share is
because other countries, notably Singapore,
South Africa and Switzerland in this case, are
increasing their citation share at a faster rate
than the UK - even though their citation counts
are smaller than the UK.
USA citation share has fallen from 65% to just
over 51% in this area. EU15 share has risen
from less than 20% to 27% over the same
period, of which the UK has been a significant
part. China has more than doubled its share,
but only to 2.6%.
42
Description of performance indicator
3.02 Number and share of world citations in ten main research fields
Condition signalling improvement
Increased national count and share
Humanities
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK citation count - Actual
1,915
169
0.09
OST comparator group average citation count
533
45
0.08
UK citation count - Rank among OST comparator group
2
2
0
UK citation count - % share of OST comparator group
14.33
14.58
1.02
16
4
UK citation count - % share of world
15.19
15.34
1.01
Note that this table ranks both performance and change in performance
Ranked change in performance
Citation share among OST comparator group for G8 nations (USA omitted) for SUoA Humanities
0
2
4
6
8
10
12
14
16
18
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Country's percentage
share of total O
S
T
comparator group citations w
ithin SUoA
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
This is the first year in which any bibliometric
data for the Humanities and Arts have been
analysed in this report. It is necessary to treat
these data with great caution as other analyses
have confirmed the degree to which these
subjects continue to make use of other modes
of output, notably books, for their key
publications.
With that caveat in mind, and recognising that
citation counts are very low compared to the
numbers of researchers working in these
disciplines, we can nonetheless see that the
UK has a strong and improving position in the
database.
That said, it must be noted that non-
Anglophone journal coverage is inevitably
rather poor, and the literature is said to be
parochial where journal content and usage
would be more international for natural
sciences and technology.
We will continue to monitor the UK's position
and will work with other agencies, including the
newly established UK Arts & Humanities
Research Council, to explore the value and
relevance of these and other research
performance indicators in these fields. For the
present, we take the outcome reported here as
a very positive reflection of the health of the UK
research base in this area.
43
Description of performance indicator
3.03 Rank of share of world citations by nine research fields
Condition signalling improvement
More frequent presence in top three among fields
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK frequency (out of 9) in top 3
6.6
7
1.061
Indicator Headline
The comparative level of national share of
world citations (3.02) is a good measure of
strength in a research field. This may be an
isolated peak of strength, however, or it may be
typical of general performance.
This indicator extends indicator 3.02 by looking
across the level of nine main research fields to
assess consistency of performance. The fields
roughly correspond to a University âFacultyâ or
group of Schools, and are defined by grouping
cognate Units of Assessment (i.e. those units
using similar literature). UoAs are subject
categories used in UK research assessment.
Nine main fields are used rather than ten
because we are as yet uncertain about the
value of national ranked position in regard to
the Humanities.
The measure is simply a count of the
occurrence of a country in the top 3 by rank
citation share. The UK ranks in the top 3 by
citation count for 7 of the nine main research
fields (indicator 3.02). This is a reflection of
strength in depth in comparative international
research performance. This has placed the UK
second to the USA, which wins in all nine
fields, in almost every year of the last ten. It
improved over Germany compared to 2003.
Only G8 countries appear in this ranking, since
smaller research nations are unable to
compete on volume. The performance profile
of Germany is notable, improving from placing
in 4 to placing in 6 fields last year but only 5 in
2004. Canada, by contrast, slipped back and
now ranks in 2-3 fields.
If the EU15 grouping were to be analysed in
the same way as the USA, it too would be
placed in the top 3 in 9 out of 9 categories.
Leading nations for ranked citation volume in nine research fields
0
1
2
3
4
5
6
7
8
9
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Fre
q
ue
nc
y
of oc
c
u
rre
nc
e
i
n
top 3
na
ti
ons
UK
USA
CANADA
FRANCE
GERMANY
JAPAN
Data: Thomson ISIÂŽ National Science Indicators
44
This page is left blank intentionally
45
Description of performance indicator
Condition signalling improvement
Increase in citation share compared to share of papers within field
Country
Year
Clinical
Pre-Clinical
Biological
Sciences
Environment Mathematics
Physical
Sciences
Engineering
Social
Science
Business
Humanities
UK
2004
0.38
1.18
2.02
1.43
0.94
1.27
0.27
-1.7
-0.88
0.87
USA
2004
5.39
7.07
8.59
4.63
6.34
8.09
5.36
5.53
11.63
6.73
CANADA
2004
0.37
1.08
-0.62
0.27
-0.34
0.44
-0.32
0.12
-0.71
-0.86
FRANCE
2004
-0.21
-0.73
0.13
0.37
0.45
-0.06
0.44
-0.22
-0.3
-3.98
GERMANY
2004
-0.4
-0.32
0.88
0.92
0.23
0.98
1.49
-1.25
-0.62
-2.46
ITALY
2004
-0.35
-0.49
-0.67
-0.56
-0.07
0.17
-0.35
0.18
-0.62
-0.43
JAPAN
2004
-2.03
-3.48
-1.34
-1.28
-1.89
-1.79
-0.72
-0.45
-0.85
-0.02
RUSSIA
2004
-0.35
-0.13
-1.04
-1.99
-1.35
-2.98
-1.43
-0.18
-0.03
-0.41
AUSTRALIA
2004
-0.26
-0.36
-0.69
0.14
0.23
0.09
-0.15
-0.45
-0.92
-0.4
BELGIUM
2004
0.13
0.26
-0.15
-0.07
0.25
-0.1
0.17
-0.1
-0.02
-0.15
BRAZIL
2004
-0.5
-1.16
-1.15
-0.46
-0.29
-0.51
-0.34
-0.35
-0.2
-0.15
DENMARK
2004
0.15
0.25
-0.06
0.26
0.14
0.26
0.26
0.03
-0.1
0.39
FINLAND
2004
0.1
0.16
-0.11
0.07
0.01
-0.02
0.01
0.16
-0.15
0.23
NETHERLANDS
2004
0.25
0.52
0.19
0.26
0.34
0.67
0.48
0.19
-0.27
0.98
POLAND
2004
-0.16
-0.28
-0.79
-0.44
-0.48
-0.63
-0.46
-0.06
-0.03
-0.09
SPAIN
2004
-0.38
-0.36
-1.18
-0.62
-0.36
-0.1
0.09
-0.07
-0.93
-0.92
SWEDEN
2004
-0.02
0.11
0.08
0.47
0.06
0.13
0.28
0.43
-0.12
-0.02
SWITZERLAND
2004
0.41
0.83
0.61
0.65
0.18
0.66
0.64
0.17
0.13
-0.33
CHINA
2004
-0.56
-1.03
-1.22
-1.55
-2.05
-3.34
-2.16
-0.52
-0.85
0.45
INDIA
2004
-0.52
-1.24
-1.93
-1.13
-0.83
-1.35
-0.97
-0.27
-1.37
-0.02
IRAN
2004
-0.08
-0.16
-0.14
-0.1
-0.23
-0.17
-0.15
-0.02
-0.06
0.01
ISRAEL
2004
-0.21
-0.06
0.14
-0.13
0.06
0.18
0.09
-0.41
0
0.07
SINGAPORE
2004
-0.1
-0.04
-0.03
-0.08
0.19
-0.25
-0.37
-0.15
-0.39
0.13
SOUTH AFRICA
2004
-0.09
-0.13
-0.47
-0.19
-0.07
-0.08
-0.12
-0.16
-0.25
-0.06
SOUTH KOREA
2004
-0.49
-0.85
-0.63
-0.46
-0.94
-0.93
-1.01
-0.23
-0.72
0.4
TAIWAN
2004
-0.48
-0.61
-0.45
-0.41
-0.49
-0.61
-1.05
-0.25
-1.41
0.05
3.04 Share of citations relative to share of publications
Percentage difference between share of citations and share of publications
within OST group of countries for most recent 5 years
The analysis assumes that each country 'uses' journals in a field in the same way. It is, in practice, evident that the general assumption is incorrect for the Social
Sciences and Humanities. Some countries' national publishing bases are less well covered by ISI and only more international work is included in the common
database. In these areas the USA also has an exceptional positive balance and comparative outcomes may thus be further flawed.
Bibliometric data for Social Sciences and Humanities should always be treated with caution and cannot be compared directly with other disciplines. In other
indicators, the analysis for these disciplines focuses essentially on the year to year trend. Only here is the comparison solely between countries.
Indicator Headline
The difference between a country's share of
world papers (sources) and its share of world
citations is like a balance in trading. There is an
investment in publications and there is an
acquisition of international recognition in
citations. Stronger research is reflected in a
greater citation share than source share.
The UK has a positive difference between its
share of world citations and its share of world
sources in eight of ten areas.
The analysis is carried out for a recent five-year
window at the level of Super-UoAs, with
Humanities introduced for the first time this year.
The USA has a strong positive balance in all
areas. Some other countries (Switzerland, the
Netherlands, Denmark, Sweden) also have
positive differences across multiple subject
areas but the absolute values of these
differences are generally less than that of the
UK. Germany has a marked positive balance in
Biological, Environmental and Physical Sciences
and Engineering. China has a growing output
but is generally in citation 'deficit'.
For smaller countries the trading balance is
mostly negative. This reflects the general
scaling relationship between output volume and
citation share, also seen at institutional level
within countries.
The balances in social sciences, business and
humanities are skewed towards the USA. The
balance of journal coverage in these areas is
not as representative as in the natural sciences.
The intention of this indicator is to identify more
effective research systems by comparing each
countryâs share of OST group citations with its
share of outputs within a field. Citation impact
(cites per paper) is analysed in 3.09.
46
Description of performance indicator
Condition signalling improvement
Increase in citation share compared to share of papers within field
3.04 Share of citations relative to share of publications
2004 Share of Sources by Share of Citations
Pre-Clinical
USA
UK
0.0001
0.001
0.01
0.1
1
0.0001
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
ar
e of
Cit
at
ions
w
it
h
in O
S
T c
o
m
p
ar
at
or grou
p
2004 Share of Sources by Share of Citations
Clinical
UK
USA
0.0001
0.001
0.01
0.1
1
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
ar
e of Ci
ta
ti
ons
wi
thi
n
OS
T c
o
mpa
ra
tor
gr
oup
2004 Share of Sources by Share of Citations
Biological Sciences
USA
UK
0.0001
0.001
0.01
0.1
1
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
ar
e of
Cit
at
ions
w
it
h
in O
S
T c
o
m
p
ar
at
or grou
p
2004 Share of Sources by Share of Citations
Environment
USA
UK
0.0001
0.001
0.01
0.1
1
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
ar
e of Ci
ta
ti
ons
wi
thi
n
OS
T c
o
mpa
ra
tor
gr
oup
Data: Thomson ISIÂŽ National Science Indicators
47
Description of performance indicator
Condition signalling improvement
Increase in citation share compared to share of papers within field
3.04 Share of citations relative to share of publications
2004 Share of Sources by Share of Citations
Mathematics
USA
UK
0.0001
0.001
0.01
0.1
1
0.0001
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
a
re
of
C
it
a
ti
ons
wi
th
in
OS
T c
o
m
p
a
ra
tor group
2004 Share of Sources by Share of Citations
Physical Sciences
UK
USA
0.0001
0.001
0.01
0.1
1
0.0001
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
ar
e of
Cit
at
ions
w
it
h
in O
S
T c
o
m
p
ar
at
or grou
p
2004 Share of Sources by Share of Citations
Engineering
USA
UK
0.0001
0.001
0.01
0.1
1
0.0001
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
ar
e
of Ci
ta
ti
ons
wi
thi
n
OS
T c
o
mpa
ra
tor
gr
oup
2004 Share of Sources by Share of Citations
Social Science
USA
UK
0.0001
0.001
0.01
0.1
1
0.0001
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
ar
e of
Cit
at
ions
w
it
h
in O
S
T c
o
m
p
ar
at
or grou
p
Data: Thomson ISIÂŽ National Science Indicators
48
Description of performance indicator
Condition signalling improvement
Increase in citation share compared to share of papers within field
3.04 Share of citations relative to share of publications
2004 Share of Sources by Share of Citations
Business
USA
UK
0.0001
0.001
0.01
0.1
1
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
ar
e of Ci
ta
ti
ons
wi
thi
n
OS
T c
o
mpa
ra
tor
gr
oup
Data: Thomson ISIÂŽ National Science Indicators
2004 Share of Sources by Share of Citations
Humanities
UK
USA
0.0001
0.001
0.01
0.1
1
0.0001
0.001
0.01
0.1
1
Share of Sources within OST comparator group
S
h
ar
e of Ci
ta
ti
ons
wi
thi
n
OS
T c
o
mpa
ra
tor
gr
oup
49
Description of performance indicator
3.05 Number and share of uncited publications
Condition signalling improvement
Decrease in ratio of uncited to total sources
Part 1 - UK SUoA analysis
NSI5 2000
NSI5 2004
Ratio
UK uncited papers
130544
122771
0.94
UK published papers
338273
351230
1.04
UK uncited papers as a percentage of all sources
38.59
34.95
0.91
NSI5 data make use of papers in a five year period and citations to those papers within the same specified period.
Volume is therefore about five times the annual average in Indicator 2.03
G8 share of world uncited papers (USA & EU15 omitted)
0
1
2
3
4
5
6
7
8
9
10
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
P
e
rc
e
n
ta
ge
of tota
l world unc
ite
d
pa
pe
rs
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
The status of uncited papers is disputed.
Papers that receive no subsequent attention are
different in kind to papers that are even
infrequently cited. It is argued that negative
reports may be uncited yet can be important in
identifying blind alleys and other unfruitful areas.
Nonetheless, since uncited work has no
measurable influence, it may be inefficient to
invest undue resources in research that has no
apparent value to other researchers. It is
assumed that research management should
work towards a minimisation in the proportion of
papers that are uncited.
The UK has been successful in progressively
reducing both the proportion of its papers that
remain uncited in a five-year period (35% down
from 38.5%) and its share of the world's uncited
papers. Its total output increased but a smaller
number of these papers remained uncited at the
end of the period. It has improved its
performance in this regard relative to the EU
competitors and to Japan.
Note that in this indicator, a time factor needs to
be considered (see background notes). The
proportion of papers published in any one year
that remain uncited in later years will decline
with time. Thus a progressive year by year
decrease would be observed in the proportion of
any cohort that remains uncited. Fixed citing
windows are therefore used here for analyses of
uncited papers in different periods, to allow the
appropriate comparisons to be made. We
compare papers published in a 5 year period
with citations over the same years rather than
the paper's lifetime.
50
Description of performance indicator
3.05 Number and share of uncited publications
Condition signalling improvement
Decrease in ratio of uncited to total sources
Clinical
Pre-Clinical
& Health
Biological
Sciences
Environment Mathematics
Physical
Sciences
Engineering
Social
Science
Business
Humanities
Total UK
uncited
papers
2000
95220
13039
37691
11622
6251
43883
44802
9223
5511
9444
130544
2003
94139
11991
36651
11797
6237
43072
42360
9695
5632
9358
124904
2004
93175
11665
35761
11576
6261
41891
41119
9803
5753
9585
122771
Percentage change 2000 - 2004
2.19
11.78
5.40
0.40
-0.16
4.76
8.96
-5.92
-4.21
-1.47
6.33
UK - numbers of uncited papers in main subject areas (SUoAs) in overlapping 5 year windows
Indicator Headline cont.
The analysis is carried out both at a national
level for comparisons between countries and at a
more detailed SUoA level to develop our
understanding of the distribution of uncitedness.
Within specific SUoA fields: Pre-Clinical &
Health, Biological Sciences, Physical Sciences
and Engineering all saw a decrease in the
number of uncited papers between the 5 years
preceding 2000 and the 5 years preceding 2004.
This indicates an improvement in performance.
In the Social Sciences, Business and
Humanities' SUoAs there was an increase in the
number of uncited papers but this is
accompanied by increases in volume. The
proportion uncited decreased in all areas and the
UK now has the lowest proportion uncited in
Social Sciences among G8 countries.
Bibliometric data for the Social Sciences and
Humanities should be treated with caution and
cannot be compared directly with other
disciplines.
51
Description of performance indicator
3.05 Number and share of uncited publications
Condition signalling improvement
Decrease in ratio of uncited to total sources
Part 2 - Comparison between UK and other countries
Percentage of uncited papers by SUoA (NSI5 data)
0
10
20
30
40
50
60
70
80
90
100
Cli
nic
al
Pr
e-C
linic
al
Bi
olog
ical Sc
ie
nc
es
En
viro
nm
ent
M
at
hem
at
ics
Ph
ys
ical
Sc
ien
ces
En
gine
erin
g
So
cial
Sc
ienc
es
Bu
sine
ss
Hu
m
an
ities
SUoA
U
n
c
ite
d pa
pe
rs
a
s
a
pe
rc
e
n
ta
ge
of tota
l s
our
c
e
s
UK 2000
UK 2004
OST Grp 2000
OST Grp 2004
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
Across almost all subject areas, the UK
produces fewer papers that remain uncited
than does the OST comparator group on
average. The UK had more uncited papers
than average for the first time in Pre-clinical
and health sciences in 2003 and this increased
in 2004.
Changes in the UK's proportion of uncited
papers follow the OST comparator group trend
across most SUoA fields. For both the UK and
OST comparator group average there is
usually a fall between the two five-year periods
to 2000 and to 2004. The exception is the
Humanities, but citation rates are
fundamentally lower in this area. Here too,
however, the UK is more often cited than the
group as a whole or any other G8 country.
52
Description of performance indicator
3.05 Number and share of uncited publications
Condition signalling improvement
Decrease in ratio of uncited to total sources
G8 - Percentage of papers remaining uncited within period
Clinical Pre-Clinical
Biological
Sciences
Environment Mathematics
Physical
Sciences
Engineering
Social
Science
Business
Humanities
Total
uncited
papers
Five year period ending in 2000
UK
51.8
55.0
40.3
41.3
56.6
46.1
58.3
60.2
60.2
79.4
130544
USA
48.4
48.0
40.1
43.7
55.3
45.2
58.2
58.0
57.5
79.4
466361
CANADA
49.6
53.4
42.3
40.9
58.5
45.6
58.5
59.7
60.8
82.3
64556
FRANCE
52.8
56.0
42.9
41.6
58.7
48.1
57.8
71.1
63.3
92.2
94260
GERMANY
51.4
52.2
43.1
43.5
59.9
45.0
56.8
72.4
64.0
86.3
122440
ITALY
53.0
54.5
46.1
48.8
58.9
46.8
59.6
66.5
64.3
87.0
56572
JAPAN
53.2
47.2
48.4
49.5
67.9
52.3
61.5
72.0
68.6
81.6
138398
RUSSIA
69.7
53.5
67.1
66.1
73.1
62.7
71.2
84.3
80.3
93.3
82266
Five year period ending in 2004
UK
48.9
52.9
37.8
37.3
53.3
42.3
54.5
55.4
57.2
77.5
122771
USA
46.3
45.8
39.1
42.0
53.2
42.3
55.4
56.7
55.0
78.8
447314
CANADA
47.5
52.9
41.3
41.1
58.3
44.7
56.4
56.2
60.8
81.6
63910
FRANCE
50.7
53.9
40.7
38.3
54.7
44.5
53.6
64.3
60.8
92.3
89049
GERMANY
47.7
50.3
40.8
38.9
54.9
41.6
52.8
64.7
60.4
83.9
116494
ITALY
49.7
50.1
44.3
44.9
56.7
43.8
56.4
61.7
64.4
83.6
59796
JAPAN
48.6
43.7
46.5
46.7
64.0
48.6
57.9
67.8
67.7
79.1
133342
RUSSIA
68.5
50.4
63.8
66.0
68.9
59.1
65.2
82.8
63.6
91.5
72761
Difference between earlier and later periods (2000 value - 2004 value, +ve result = decrease in uncitedness)
UK
3.0
2.1
2.5
4.0
3.4
3.8
3.8
4.8
2.9
1.9
7773
USA
2.1
2.3
1.0
1.7
2.1
2.9
2.9
1.4
2.5
0.6
19047
CANADA
2.1
0.5
1.0
-0.2
0.2
0.9
2.1
3.5
0.1
0.7
646
FRANCE
2.2
2.1
2.2
3.2
4.0
3.6
4.3
6.7
2.5
-0.1
5211
GERMANY
3.7
1.9
2.3
4.6
5.0
3.3
4.0
7.7
3.6
2.4
5946
ITALY
3.3
4.4
1.8
3.8
2.2
3.0
3.2
4.7
-0.1
3.4
-3224
JAPAN
4.6
3.5
1.9
2.8
3.9
3.7
3.6
4.2
0.9
2.5
5056
RUSSIA
1.2
3.1
3.2
0.1
4.2
3.6
6.0
1.5
16.8
1.8
9505
53
Description of performance indicator
3.06 Number and share of cited publications
Condition signalling improvement
Increase in ratio of cited to total sources
Part 1 - UK SUoA analysis
NSI5 2000 NSI5 2004
Ratio
UK cited papers
207729
228459
1.10
UK published papers
338273
351230
1.04
UK cited papers as a percentage of all sources
61.4
65.0
1.06
NSI5 data make use of papers in a five year period and citations to those papers within the same specified period.
G8 share of world cited papers (USA & EU15 omitted)
0
2
4
6
8
10
12
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
P
e
rc
e
n
ta
ge
of tota
l world c
ite
d pa
pe
rs
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
Number and share of cited papers is the positive
counterpart of indicator 3.05, the number and
share of uncited papers.
Whereas uncited papers are a well defined
category, cited papers are a rather more diffuse
group including both low and high impact papers.
They are therefore more difficult to define in
terms of a performance indicator.
The UK's cited papers have increased recently at
a relatively faster rate than the total national
output (ratio 2004 vs. 2000).
The UK's cited papers as a proportion of all
outputs has consequently improved from 61 to
65%. The UK's share of world cited papers
increased to 2002 but has now fallen back
slightly.
The general G8 trend has been for an increase
in citedness between the 5 year periods to 2000
and to 2004, with improvements for most EU
countries. The immediate trend is for some
levelling in this pattern.
Citedness for China outputs has improved
substantially between the two five year periods
but remains significantly below the OST group
averages for the present.
54
Description of performance indicator
3.06 Number and share of cited publications
Condition signalling improvement
Increase in ratio of cited to total sources
Clinical
Pre-Clinical
Biological
Sciences
Environment
Mathematics
Physical
Sciences
Engineering Social Science
Business
Humanities
Total UK
uncited
papers
2000
88451
10669
55803
16534
4786
51299
32057
6098
3651
2452
207729
2003
96798
10527
58942
19218
5490
56411
34536
7651
4210
2689
226837
2004
97398
10379
58828
19449
5493
57152
34304
7889
4299
2788
228459
Percentage change 2000 - 2004
10.1
-2.7
5.4
17.6
14.8
11.4
7.0
29.4
17.7
13.7
10.0
UK - numbers of cited papers in main subject areas (SUoAs) in overlapping 5 year windows
Percentage of cited papers by SUoA (NSI5 data)
0
10
20
30
40
50
60
70
Cli
nic
al
Pr
e-C
linic
al
Bi
olog
ical Sc
ie
nc
es
En
viro
nm
ent
M
at
hem
at
ics
Ph
ys
ical
Sc
ien
ces
En
gine
erin
g
So
cial
Sc
ienc
es
Bu
sine
ss
Hu
m
an
ities
SUoA
Ci
te
d paper
s as a per
cent
age of
t
o
ta
l sour
ces
UK 2000
UK 2004
OST Grp 2000
OST Grp 2004
Data: Thomson ISIÂŽ National Science Indicators
55
Description of performance indicator
3.07 Number and share of publications in top 1% by citation count
Condition signalling improvement
Increase in share
Country
Papers
exceeding
threshold
Rank by
Papers
Papers as %
share OST
Group
Citations for
these papers
Rank by
Citations
Average
impact of
highly cited
papers
OST group
rank by
impact
G8 rank by
impact
USA
25,746
1
61.4%
2,208,042
1
85.76
8
4
UK
5,528
2
13.2%
482,795
2
87.34
6
2
GERMANY
4,715
3
11.3%
384,599
3
81.57
15
7
JAPAN
2,921
4
7.0%
251,576
4
86.13
7
3
FRANCE
2,847
5
6.8%
240,515
5
84.48
10
5
CANADA
2,435
6
5.8%
214,811
6
88.22
5
1
ITALY
1,878
7
4.5%
156,234
7
83.19
14
6
SWITZERLAND
1,683
8
4.0%
148,561
8
88.27
4
NETHERLANDS
1,678
9
4.0%
141,250
9
84.18
11
AUSTRALIA
1,353
10
3.2%
113,567
10
83.94
12
CHINA
1,251
11
3.0%
66,017
13
52.77
25
SWEDEN
1,102
12
2.6%
101,435
11
92.05
3
SPAIN
1,100
13
2.6%
93,051
12
84.59
9
BELGIUM
775
14
1.8%
62,757
14
80.98
16
DENMARK
712
15
1.7%
59,708
15
83.86
13
ISRAEL
607
16
1.4%
56,291
16
92.74
2
RUSSIA
586
17
1.4%
44,730
18
76.33
19
8
SOUTH KOREA
560
18
1.3%
35,043
19
62.58
23
FINLAND
478
19
1.1%
46,109
17
96.46
1
POLAND
374
20
0.9%
29,984
20
80.17
17
INDIA
354
21
0.8%
26,291
21
74.27
20
BRAZIL
321
22
0.8%
22,805
22
71.04
21
TAIWAN
290
23
0.7%
17,652
23
79.35
18
SINGAPORE
169
24
0.4%
9,878
24
58.45
24
SOUTH AFRICA
129
25
0.3%
8,193
25
63.51
22
IRAN
14
26
0.0%
405
26
28.93
26
EUROPEAN UNION
15,972
0
0
1,267,368
0
0.00
0
0
Indicator Headline
Some publications have exceptional citation
rates compared to others in their field. The
national share of the most cited 1% of papers
is an indicator of interest. The data reported
here cover the five-year period 2000-2004.
The citation counts are not directly comparable
to last year because the point at which we
sampled the papers was later in the cycle
(November instead of January). Paper counts
and rank position are comparable, although
the total of papers in the set is greater as
Thomson journal coverage has increased.
The UK has increased its share from 12.9% to
13.2%, which is a further improvement over
each previous 5-year window. The UK is
second in volume of highly cited papers to the
USA. Its share (13.2%) is high relative to its
share of all papers (about 9%, see Indicator
2.03).
The UK has moved from 10th to 6th in the
OST comparator group in terms of average
impact of highly cited papers, and is 2nd to
Canada in the G8. The absolute volume of
other countries is, however, much lower.
On volume, the USA has fallen further back
from its 62.7% last time. The biggest
improvement among the G8 is for Germany, by
a further 0.5% on top of +0.4% last year.
China has also significantly improved its
position, up from 1% two years ago to 3% now.
It continues to have an atypical distribution of
peak activity, in relatively low citation areas,
thus resulting in a low average impact.
cont./
56
Description of performance indicator
3.07 Number and share of publications in top 1% by citation count
Condition signalling improvement
Increase in share
Numbers of highly cited papers to 2000-2004 for OST comparator group
GER
EU15
UK
USA
100
1,000
10,000
100,000
1,000,000
10,000,000
10
100
1,000
10,000
100,000
Highly cited papers (log scale)
Ci
tati
on count (l
og scal
e)
U
Data: Thomson ISIÂŽ Essential Science Indicators
Indicator headline cont.
The USA has also dropped from 6th on
average impact to 8th in the OST group and is
behind the UK, Canada and Japan. Finland
has moved up to 1st overall on average impact
although only 18th on volume.
The USA has an exceptional share of highly
cited papers (62.7% compared with just over
33.7% of global output). The EU total (de-
duplicated to take account of co-authorship
between countries within the EU) shows that
Europe remains behind the USA in this respect
despite now exceeding it in terms of total
outputs at 35.9%.
Being above world average is a good initial
sign of quality, whatever the specific indicator.
As a general rule, however, the research that
lies in the far upper end of the quality
distribution is most likely to have a long lasting
impact on its research field and - perhaps - the
economy more widely.
âExceptionalâ could have a number of
definitions. We use the Thompson ScientificÂŽ
benchmark of the top 1% of papers by citation
count for each field and year for international
research comparisons. The ISI data are
compensated for differences between fields in
their size and citation behaviour.
57
Description of performance indicator
3.08 Citation impact (citations/publication) relative to world baselines
Condition signalling improvement
Increased national impact
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
7.91
0.48
0.06
OST comparator group average impact
6.23
0.36
0.06
UK impact - Rank
7
5
2
UK rebased impact - Actual
1.25
1.37
1.10
6
3
0.98
1.02
1.05
Note that this table ranks both performance and change in performance
Ranked performance change
OST comparator group average rebased
impact
Rebased average impact (world = 1.0) for G8 nations
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
The most frequently used index of research
performance is that of impact, measured as
citations per paper. This is widely accepted
internationally as an index of research quality.
The UK has improved its relative international
performance (Rebased Impact (RBI) of 1.37
[rebased against world = 1.0, see below] in
2004, 1.25 recent average) and its overall rank
(now 5th from 7th). It is 2nd to the USA (1.45
average recent impact) among G8 nations,
and is evidently closing the gap as it has done
throughout the last decade.
Germany has significantly improved its
position, from around world average in 1993 to
a 1.34 average RBI in 2003 and will continue
to challenge the UK and the USA.
The USA lies 2nd globally to Switzerland (1.74
average recent impact). Denmark lies 3rd
overall with the Netherlands 4th. Sweden is
6th but with a slightly higher recent average
impact than the UK.
China, which has had a major increase in
investment and output, has an average impact
of around 0.63 and lies 24th in the OST
comparator group. Its average RBI had
improved over the decade but has plateaued in
the last few years.
Citation rates vary between fields, and older
papers have more time to accumulate
citations. A common baseline is therefore
created to compare impact data. This baseline
is the world average figure for the stated field
and year. Setting actual performance figures
against a common reference point is called
ârebasingâ (or normalisation) and the figure
usually quoted will be ReBased Impact (RBI).
[Data by field are shown in indicator 3.09.]
58
This page is left blank intentionally
59
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Clinical Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
10.91
0.60
0.06
OST comparator group average impact
8.83
0.46
0.05
UK impact - Rank
8
4
4
UK rebased impact - Actual
1.18
1.30
1.10
8
2
0.96
1.00
1.04
Note that this table ranks both performance and change in performance
Ranked performance change
OST comparator group average rebased
impact
Clinical - Rebased Impact
0.8
1.0
1.3
1.5
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
For Clinical Sciences, the UK is ranked 4th in
terms of impact for 2004, the most recent year.
This is a significant uplift from last year and
places it 2nd among G8 nations to the USA.
The UK has moved ahead of other
competitors, has caught up with Canada as
predicted last year, but remains behind
Switzerland and Belgium within Europe.
UK rebased impact rose from 1.09 in 1995, to
1.18 for the period 1999-2003 and 1.3 for 2004
(world average = 1.0). The step up in 2002
seemed exceptional but is now seen to be part
of a sustained trend. This degree of
performance change (ratio = 1.10) is first
among G8 nations, except Russia where
performance is currently erratic. It is 8th
among the OST comparator group, up from
12th last year, where improvement is led by
South Africa and Poland. These
improvements among smaller nations are
rather variable from year to year.
USA impact has been more or less static.
China had risen above world average in 2003
but has slipped back to an RBI of 0.85 (0.81
over the recent past).
60
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Pre-Clinical & Health-Related Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
12.05
0.79
0.07
OST comparator group average impact
9.28
0.55
0.06
UK impact - Rank
8
3
5
UK rebased impact - Actual
1.30
1.54
1.19
6
3
1.01
1.07
1.06
Note that this table ranks both performance and change in performance
Ranked performance change
OST comparator group average rebased
impact
Pre-Clinical - Rebased Impact
0.6
0.8
1.0
1.2
1.4
1.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
In Pre-Clinical & Health-Related Sciences, the
UK is ranked 3rd in terms of impact for 2004,
the most recent year. Among G8 nations, it is
now ahead of both the USA and Canada. The
UK's performance has improved markedly
throughout the recent period (1999 to present).
UK rebased impact rose from 1.2 in 1996, to
1.3 on average for the period 1999-2003 and
was 1.54 in 2004 (world average = 1.0). The
degree of rebased performance change (ratio
= 1.19) is 6th among OST comparator group
nations, and 3rd to Japan and Russia within
the G8 group.
USA impact has been fairly flat through the
last ten years. The EU15 has shown a slow
but progressive average improvement. China
does not have a major research presence in
this area.
In 2003, a number of less research intensive
countries showed an exceptional improvement
in citation impact in this area. In Europe, the
Netherlands, Denmark and Switzerland are
leaders, but only the latter sustained this
improvement into 2004.
Japan's improvement is particularly marked
while France has performed rather less well in
this area than would be expected for its
research base generally.
61
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Biological Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
10.89
0.64
0.06
OST comparator group average impact
7.24
0.40
0.06
UK impact - Rank
3
3
0
UK rebased impact - Actual
1.28
1.40
1.09
6
3
0.84
0.88
1.04
Note that this table ranks both performance and change in performance
Ranked performance change
OST comparator group average rebased
impact
Biological Sciences - Rebased Impact
0.6
0.8
1.0
1.2
1.4
1.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
In Biological Sciences, the UK is ranked 3rd.
It has sustained the trend of the last ten years
and has caught up with the USA in terms of
performance. It remains behind Switzerland.
A possibly maverick result in 2004 (based on
a small number of papers) is that of
Singapore which has moved into 2nd place,
from an average of 9th in the recent past.
This unexpected result will need to be
reviewed next year.
There has been a relative international
improvement in UK rebased impact which
rose from 1.28 to 1.4 between the period
1999-2003 and 2004. UK average of 1.28 for
the recent period compares to 1.34 for the
USA and 1.44 for Switzerland. However,
neither of these countries has shown the
trend of improvement that the UK has
achieved. Only Germany's relative position
has improved as consistently over the period.
This is an area of strength in terms of volume
and performance for the UK. It is well ahead
of OST comparator group average and only
those nations mentioned above have
consistently competitive performances. It
makes a significant contribution to the overall
performance of the EU15 which has
otherwise slipped back slightly.
62
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Environmental Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
6.09
0.37
0.06
OST comparator group average impact
4.96
0.30
0.06
UK impact - Rank
8
9
-1
UK rebased impact - Actual
1.23
1.36
1.10
12
4
0.992
1.10
1.10
Note that this table ranks both performance and change in performance
Ranked performance change
OST comparator group average rebased
impact
Environment - Rebased Impact
0.6
0.8
1.0
1.2
1.4
1.6
1.8
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
Research performance in the Environmental
Sciences has been quite variable between
years. The UK has improved relative to the
USA and Germany, but its ranked impact in the
OST comparator group has slipped to 9th.
UK rebased impact has risen from below 1.1 in
the early 1990s to 1.23 for the period 1999-
2003 and 1.36 in 2004. Nonetheless, the
degree of rebased performance change (ratio =
1.10) is only 12th among OST comparator
group nations. This is because a number of
smaller nations (including South Africa, for
which we noted last year that the values
appear exceptional) have improved their
performance to an even greater degree. These
include the Netherlands and the Scandinavian
group.
While the UK has shown consistent
improvement in relative international
performance over the period, and has thus
overtaken the USA, this is a field in which a
number of countries have similar performance
and the degree of competition is therefore
marked. The Scandinavian countries have
better impact than the G8 lead group.
Germany's recent performance had taken it
well ahead but it has now been pegged back.
EU networks will, of course, allow the research
elsewhere in Europe to be accessed by other
nations in global environment networks.
63
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Mathematics
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
2.46
0.15
0.06
OST comparator group average impact
2.14
0.11
0.05
UK impact - Rank
8
4
4
UK rebased impact - Actual
1.17
1.38
1.18
2
1
1.02
1.00
0.98
Note that this table ranks both performance and change in performance
Ranked performance change
OST comparator group average rebased
impact
Mathematics - Rebased Impact
0.6
0.8
1.0
1.2
1.4
1.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
In Mathematics, the UK is ranked 4th in terms
of impact for 2004, the most recent year.
There has been a notable improvement in the
last two years and the UK's rank has changed
from 8th for the recent period (1999-2003).
UK rebased impact has fluctuated in this area.
It dropped to and moved around 1.2 for much
of the last ten years but has risen steeply
since 2002. The degree of rebased
performance change this year is 2nd among
OST comparator group nations, and 1st within
the G8 group. This suggests something of a
welcome renaissance amongst UK
mathematics.
Mathematics is a field in which smaller
countries can be competitive. Singapore had
improved to achieve a similar impact to the UK
but this year has fallen back. Australia has
also shown a sawtooth performance. Belgium,
Denmark and Switzerland have all been
ranked 1st in the recent past. China has
improved to around world average.
The USA was consistently around 4th among
the OST comparator group. Within the G8,
France improved its performance to 2001 but
then fell back while Germany has shown
significant recent improvement.
64
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Physical Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
7.76
0.60
0.08
OST comparator group average impact
6.22
0.43
0.07
UK impact - Rank
7
5
2
UK rebased impact - Actual
1.37
1.67
1.22
4
2
1.09
1.21
1.11
Note that this table ranks both performance and change in performance
Ranked performance change
OST comparator group average rebased
impact
Physical Sciences - Rebased Impact
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
In Physical Sciences, the UK is ranked 5th in
terms of impact for 2004. The UK's rank has
improved from 9th in the mid-1990s to 7th on
average for the recent period (1999-2003).
There has been a sustained rise in UK
rebased impact from 1.14 in 1994 to 1.37 in
the period 1999-2003 and an exceptional 1.67
in 2004, overtaking Canada (1.45) which fell
back steeply this year. The UK's recent
rebased performance change is 4th among
OST comparator group nations. Whether the
upturn in performance for the Netherlands is
real or artefactual will need examination next
year, as will the upward move for the UK,
Germany, France and Italy.
Physical Sciences are an important part of the
underpinning research for engineering and
technology development. The UK's
improvement in performance and its relative
citation impact should be seen against a
backdrop of a relatively low share of world
citations (Indicator 3.02), which may reflect a
limited capacity.
China has increased capacity but not
performance in this area and it remains at
around 0.7 RBI against the world average.
Denmark, Israel and Australia have been
strong performers in recent years.
65
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Engineering
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
3.78
0.21
0.06
OST comparator group average impact
3.62
0.19
0.05
UK impact - Rank
11
9
2
UK rebased impact - Actual
1.06
1.21
1.14
5
2
1.01
1.07
1.06
Note that this table ranks both performance and change in performance
Ranked performance change
OST comparator group average rebased
impact
Engineering - Rebased Impact
0.8
1.0
1.2
1.4
1.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
In Engineering, the UK is ranked 9th in terms
of impact for 2004, the most recent year. The
UK's rank has improved from 14th in the early
1990s and an average of 11th for the recent
period (1999-2003).
UK rebased impact rose from close to world
average in the mid-1990s to 1.21 for 2003
(world average = 1.0). The degree of rebased
performance change (ratio = 1.14) is 5th
among OST comparator group nations, and 1st
within the G8 group (excepting Russia).
There has been a relative international
improvement in the UK's performance, which
has been behind not only the USA among G8
nations but also its close European
competitors. The UK is now well ahead of
EU15 average and has clearly overtaken
France, with a sustained improvement over
three years. Its performance remains behind
some smaller European nations, however,
including the Scandinavian countries (Denmark
ranks 1st with RBI > 2), Belgium and the
Netherlands as well as Switzerland and
Germany.
China has improved its RBI from about 0.7 to
about 0.8 but remains at 20th in the OST
comparator group.
66
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Social Sciences
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
2.86
0.17
0.06
OST comparator group average impact
2.78
0.13
0.05
UK impact - Rank
11
9
2
UK rebased impact - Actual
0.92
0.94
1.02
9
5
0.88
0.74
0.84
Note that this table ranks both performance and change in performance
Ranked performance change
OST comparator group average rebased
impact
Social Science - Rebased Impact
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
As we have noted in Indicator 3.02, bibliometric
data for the Social Sciences should be treated
with caution and cannot be compared directly
with other disciplines. While the comparisons
here are like-for-like, journal coverage for non-
Anglophone European nations is probably in
deficit so international comparisons must be
interpreted with caution.
The UK is ranked 9th in terms of impact for
2003, the most recent year. Rank position is
more volatile in this field (e.g. see Italy) than in
the Natural Sciences but the UK's position as
well as performance has been consistent over
the decade.
UK rebased impact was below world average,
although it improved between the period 1999-
2003 and 2004. It is placed behind a number of
smaller nations which have impact above world
average. Coverage for these nations may be
selective, Less international non-Anglophone
journals and their contents may be absent from
the database. Thus, whereas the UK has a
diversity of journal articles in the database, for
some smaller countries only exceptional items
are covered and only the peak of performance
is analysed.
Overall, the UK's trend against this backdrop is
more important than its position. To that extent,
the improvement in its impact and rank are a
positive.
67
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Business
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
2.40
0.12
0.05
OST comparator group average impact
2.14
0.09
0.04
UK impact - Rank
8
8
0
UK rebased impact - Actual
1.00
1.16
1.16
7
2
0.85
0.93
1.09
Note that this table ranks both performance and change in performance
OST comparator group average rebased
impact
Ranked performance change
Business - Rebased Impact
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
The UK is ranked 8th in terms of impact for
2004. The UK's rank is similar to that for the
recent period (1999-2003). While the UK has
been affected by a volatile performance profile
with a peak in 1998 and a dip in 2000, it now
seems to be on an upward trajectory.
UK rebased impact rose from below world
average in 1994 to 1.1 in 1998 (world average
= 1.0). The degree of rebased performance
change (ratio = 1.01) is 7th among OST
comparator nations, and 2nd only to Russia
within the G8 comparator group.
There are similar fluctuations to those of the
UK in some other G8 countries (see chart) and
some very extreme variations in smaller
countries (possibly due to the effects of small
numbers of papers). Elsewhere, Belgium and
Switzerland have performed well on average
but otherwise only Sweden performs above
world average for any extended period.
The USA has had a consistently strong position
and dominates business and management
coverage. This reflects the USA's dominance
of the recorded and indexed literature in these
fields and its influence on journal coverage.
68
Description of performance indicator
3.09 Citation impact relative to world baselines in ten main research fields
Condition signalling improvement
Increased impact in main research fields
Humanities
Average
1999 - 2003
Actual
2004
Ratio or
difference
UK - all OST
UK - G8
UK impact - Actual
0.78
0.06
0.08
OST comparator group average impact
0.81
0.06
0.08
UK impact - Rank
13
11
2
UK rebased impact - Actual
1.17
1.15
0.99
15
5
1.15
1.15
1.00
Note that this table ranks both performance and change in performance
OST comparator group average rebased
impact
Ranked performance change
Indicator Headline
This is the first time that the Humanities and
Arts have been covered for bibliometric
analyses in this report. The data must be
interpreted with caution.
Paper counts are often rather low in the
national samples. We know that journal
outputs are not usually a main mode of
publication in these fields and that citation
behaviour is different to the sciences. We are
also aware that journal coverage is dominated
by Anglophone sources and that there may be
only limited input from non-European and non-
American nations, perhaps only the very peak
of their country's output.
In the context of those caveats the UK's
position in the chart accompanying this text
seems unexceptional. Its impact is above
world average, is ahead of many G8
competitors and compares well with the USA
which has much more diverse database
coverage.
No immediate conclusions should be drawn but
this indicator will be of widespread interest and
we will monitor its change for future reports.
Humanities - Rebased Impact
0.2
0.6
1.0
1.4
1.8
2.2
2.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Year
Rebased I
m
pact
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
NETHERLANDS
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators
69
Description of performance indicator
3.10 Variety and consistency of research strength
Condition signalling improvement
Reduced ratio between variance and average. Maximised ratio of average/variance.
Country
5 years to
Average
1/Variance
Country
5 years to
Average
1/Variance
UK
2000
1.14
40.30
UK
2004
1.18
36.52
USA
2000
1.33
67.01
USA
2004
1.31
62.96
CAN
2000
1.05
32.52
CAN
2004
1.06
27.00
FRA
2000
0.86
11.03
FRA
2004
0.88
9.54
GER
2000
0.94
9.13
GER
2004
1.01
11.68
ITA
2000
0.87
19.57
ITA
2004
0.96
34.51
JAP
2000
0.87
38.85
JAP
2004
0.89
37.13
NED
2000
1.33
22.32
NED
2004
1.39
9.56
SUI
2000
1.12
2.97
SUI
2004
1.23
6.10
Rebased bibliometric impact across SUoAs
Analysis of rebased impact across SUoAs for select members of OST comparator group
Performance is maximised by combining high average impact with low variability
SUI
NED
JAP
ITA
GER
FRA
CAN
USA
UK
SUI
NED
JAP
ITA
GER
FRA
CAN
USA
UK
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
0.70
0.80
0.90
1.00
1.10
1.20
1.30
1.40
Average of rebased impact across SUoAs
1/
V
ari
ance i
n
reb
ased i
m
pact across
SU
oA
s
2000
2004
Data: Thomson ISIÂŽ National Science Indicators
Indicator Headline
See below for Methodology.
It is argued that a balanced research economy
has both strength and evenness. That is, high
average quality across fields would be
enhanced by low variation between them.
This is a desirable combination because
research strength should not be unduly
concentrated. Investing in priorities is sensible
but research is long term; capacity takes years
to build. Not all needs and opportunities can
be foreseen, so the flexibility to shift into new
areas is worth paying for. It contributes to a
national capacity to appraise research
developments in other countries and to
respond to research opportunities.
We can visualise this characteristic of the
research base by looking simultaneously at
average impact and the reciprocal of variance
across fields. (We use the reciprocal because
we are interested in systems that minimise
variation; this is one of several alternative
measures of variability). We then look for
countries that maximise the net product.
For the UK, average impact is relatively high
compared to most G8 countries but less than
the USA. It is also less than the Netherlands
(NED) and Switzerland (SUI). Both the smaller
countries have greater variance between
disciplines, however, which suggests that their
research economy is relatively concentrated.
In the graph, the UK shows a positive shift in
average impact with a decrease in variance for
the two 5 year periods to 2000 and to 2004.
For the Netherlands impact also increases (as
it does for Switzerland), but variance increases
markedly (decrease in 1/variance), reducing
the net product.
70
Description of performance indicator
3.10 Variety and consistency of research strength
Condition signalling improvement
Reduced ratio between variance and average. Maximised ratio of average/variance.
Indicator headline cont.
The USA has increased variance and reduced
average impact and has thus declined on both
dimensions compared to other research
economies. Germany is improving on both
counts and has an improved net outcome.
Methodological note
To index consistency, average national impact
across fields (SUoAs) has been compared with
the variation between fields. Variation is
measured here by calculating variance, but other
statistical indexes may be preferred in later
iterations of this indicator. Normally, variance will
increase with the absolute value of the statistical
mean so a relatively low mean/variance ratio
reflects poor consistency.
Data are displayed as average national impact
plotted against the inverse of variance. In this
plot the more balanced economies will be those
that maximise their mean/variance ratio, or
optimise the product of mean and 1/variance and
lie in the upper right sector of the graph.
There are, of course, a number of solutions to the
same net product but generally an exceptionally
high average impact will be required to offset
high variability. It is very difficult to conceive of
one country being able to establish such a global
predominance in any one field that it could
achieve this. More commonly, an optimal
management solution is found by reducing
variation in performance between fields.
71
Description of performance indicator
3.11 Relationship between distribution of research training across subjects and research quality
Condition signalling improvement
Improved match in distribution; improved research training power (product of volume and quality)
Medical
Sciences
Natural
Sciences
Engineering
and
Technology
Social
Sciences
Humanities
UK average rebased impact 1998 - 2002
1.18
1.32
1.08
1.07
1.16
UK average PhDs awarded 1998 - 2002
1825
4823
2004
2179
1582
0.94
0.96
0.99
0.91
0.97
1415
2356
1132
1858
937
OST comparator group average PhDs awarded
1998 - 2002
OST comparator group average rebased impact
1998 - 2002
Indicator Headline
What is the quality of the researchers we train?
People are a key output from the research base and a
major route for the transfer of knowledge and know-
how from producers to users. Highly trained people
with PhDs are a specialised part of that output. Direct
measurement of people quality is unfeasible, but we
can measure the association between the distribution
of training and research quality.
This year, Humanities & Arts data have been added to
the analytical coverage.
For the UK, there is a relative concentration of PhD
awards in Natural Sciences where average research
impact rebased against world benchmarks is also
highest. The data points are too few to calculate a
correlation but the general relationship between
concentration of training and performance appears
positive.
For the USA, the relationship between training volume
and research performance is positive only for sciences
and technology and a relative excess of PhDs is
awarded in the Social Sciences and, to a lesser extent,
in the Humanities and Arts.
There is no clear relationship between training
concentration and research quality for other G8
countries for which data are available and there is a
negative relationship for the OST comparator group
average. For Germany and Japan, too, there appear
to be relatively more PhDs awarded in fields where
rebased impact is relatively lower.
If the benefits of training are related to research
quality, then this positive relationship for the UK should
bring wider benefits to the research base. This is
therefore a marker of relative consistency in the
structure of the research base.
Indicator Headline cont.
We cannot assert that a country that manages its
training so as to ensure that highly qualified people
benefit from a rich research environment will
necessarily benefit at an economic level, but it seems
reasonable to suppose that where training is
associated with research quality this is at least unlikely
to be a disbenefit.
For this indicator, data are taken from the OECD
Education Database. These data are disaggregated
by fields that are most readily mapped to other OECD
fields rather than SUoAs.
Detailed exploration and consideration of the data may
reveal additional features. It is possible that a national
policy might actually advocate increased PhD output in
areas of relative weakness, to increase the pool of
trained people.
72
Description of performance indicator
3.11 Relationship between distribution of research training across subjects and research quality
Condition signalling improvement
Improved match in distribution; improved research training power (product of volume and quality)
UK 2002 - rebased impact vs. PhDs awarded by OECD field
Humanities
Social Sciences
Medical Sciences
Engineering &
Technology
Natural Sciences
0
1000
2000
3000
4000
5000
6000
7000
8000
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
Rebased Impact
PhDs awarded
USA 2002 - rebased impact vs. PhDs awarded by OECD field
Humanities
Social Sciences
Medical Sciences
Engineering &
Technology
Natural Sciences
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
Rebased Impact
P
h
D
s
aw
ar
ded
France 2002 - rebased impact vs. PhDs awarded by OECD field
Humanities
Social Sciences
Medical Sciences
Engineering &
Technology
Natural Sciences
0
1000
2000
3000
4000
5000
6000
7000
8000
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
Rebased Impact
PhDs awarded
Data: Thomson ISIÂŽ National Science Indicators,
OECD Education Database
73
Description of performance indicator
3.11 Relationship between distribution of research training across subjects and research quality
Condition signalling improvement
Improved match in distribution; improved research training power (product of volume and quality)
Germany 2002 - rebased impact vs. PhDs awarded by OECD field
Humanities
Social Sciences
Medical Sciences
Engineering &
Technology
Natural Sciences
1500
2500
3500
4500
5500
6500
7500
8500
9500
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
Rebased Impact
PhDs awarded
Italy 2002 - rebased impact vs. PhDs awarded by OECD field
Humanities
Social Sciences
Medical Sciences
Engineering &
Technology
Natural Sciences
0
1000
2000
3000
4000
5000
6000
7000
8000
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
Rebased Impact
P
h
D
s
aw
ar
ded
Japan 2002 - rebased impact vs. PhDs awarded by OECD field
Humanities
Social Sciences
Medical Sciences
Engineering &
Technology
Natural Sciences
0
1000
2000
3000
4000
5000
6000
7000
8000
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
Rebased Impact
PhDs awarded
OST comparator group average 2002 - rebased impact vs. PhDs
awarded by OECD field
Humanities
Social Sciences
Medical Sciences
Engineering &
Technology
Natural Sciences
0
1000
2000
3000
4000
5000
6000
7000
8000
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
Rebased Impact
PhDs awarded
Data: Thomson ISIÂŽ National Science Indicators,
OECD Education Database
74
This page is left blank intentionally
75
Description of performance indicator
4.01 PhDs awarded relative to HERD
Condition signalling improvement
Increased ratio
Average
1998 - 2001
Actual
2002
Ratio or
difference
UK PhDs awarded per $M HERD - Actual
2.15
2.15
1.00
OST comparator group average - PhDs awarded per $M HERD
1.85
1.77
0.96
UK PhDs awarded per unit $M HERD - Rank
5
5
0
G8 nations and EU15: PhDs per HERD (no data for Russia)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
1998
1999
2000
2001
2002
Year
P
h
Ds awarded per HE
RD $M
2000 P
P
P
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
EUROPEAN UNION
Data: OECD (MSTI), OECD Education Database
UK - HESA data collection
procedure changed in 2001
(see headline note)
Indicator Headline
This indicator compares the output of people
with research degrees with the specific
research spend in HE (HERD).
The UK produces more PhDs relative to HERD
than the OST comparator group on average
(for the 17 countries for which data were
available). It is ranked 5th, behind Germany,
South Korea and Spain, with a similar output to
Switzerland and Finland. The EU15 is
significantly more productive than the USA,
which is ranked 11th.
The UK produces slightly in excess of 8% of
the world's PhDs, rather more than most of the
G8 nations (Indicator 2.01). The exceptions
are Germany and the USA who produce twice
and four times as many PhDs than the UK
respectively.
The relative volume of PhDs for each nation is
otherwise broadly similar to that for publication
output. Note that the general trend for
decreasing relative output is despite the data
being adjusted for PPP.
Highly skilled postgraduates are a key output
of the higher education sector. They transfer
knowledge to users and, more generally, they
transfer know-how and technological
advances. It should not be assumed that all
science and engineering postgraduates are
necessarily employed in science and
technology let alone research. More generally,
the gain to the economy is in having a trained
and technological workforce capable of
assessing and responding to technology
related opportunities and issues.
cont./
76
Description of performance indicator
4.01 PhDs awarded relative to HERD
Condition signalling improvement
Increased ratio
PhDs awarded relative to HERD - OST comparator nations, 2002
EU15
SKO
ISR
SUI
SWE
UK
USA
FRA
GER
ITA
JAP
AUS
BEL
DEN
FIN
NED
POL
ESP
100
1000
10000
100000
100
1000
10000
100000
HERD $M 2000 PPP (log scale)
P
h
Ds awarded (l
og scal
e)
Data: OECD (MSTI), OECD Education Database
Check Background section of the report for country short codes
Indicator Headline cont.
Total expenditure in the higher education
sector is a broad overall measure of the cost of
producing trained people. It might be
reasonable to assume that there is also a
research element to postgraduate training,
either directly or through enhancement of the
training environment, and this will be
influenced by research specific expenditure.
It is not necessarily the case that low relative
research spend (many PhDs awarded per $M
R&D expenditure) is a good thing, since it may
imply poor quality training.
UK PhD data are subject to a change in HESA
data collection from 2001 (see Background).
77
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Medical Sciences
Average
1998 - 2001
Actual
2002
Ratio or
difference
UK PhDs awarded per HERD ($M 2000 PPP) - Actual
1.09
1.28
1.17
1.92
1.81
1
UK PhDs awarded per unit HERD ($M 2000 PPP) - Rank
9
7
2
OST comparator group average - PhDs awarded per unit HERD
($M 2000 PPP)
Indicator Headline
The indicator compares HERD expenditure
allocated to the Medical Sciences with the
numbers of PhDs awarded in this category.
Medical Sciences include both clinical and pre-
clinical areas, nursing and health.
The UK produces fewer PhDs per unit spend
than the OST comparator group average.
This is a much lower level of productivity per
unit spend than Germany but similar to that of
Finland, Australia and, amongst the G8
nations, Japan. The USA is substantially
below the group average.
UK PhD data are subject to a change in HESA
data collection from 2001 (see Background).
In this field, the UK continues its recent trend
of increasing PhD awards (rising from 2001 by
more than 10%). The UK remains the 4th
largest - 2,300 (behind Germany - 8,400,
Japan - 4,300 and USA - 3,800) by volume.
There is only a poor correlation across
countries between funding and PhD output,
implying that although both may scale with an
underlying size factor such as staff numbers,
there are other factors affecting training and
research volumes.
Medical Sciences - PhDs per HERD, G8 nations (no data for Canada, France, Italy or Russia)
0.0
1.0
2.0
3.0
4.0
5.0
6.0
1998
1999
2000
2001
2002
Year
Ph
Ds aw
ard
e
d
p
e
r HERD $M 2000 PPP
UK
USA
GERMANY
JAPAN
EUROPEAN UNION
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
78
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Medical Sciences - PhDs awarded relative to HERD - OST comparator group nations, 2002
SUI
EU15
SWE
ESP
POL
NED
FIN
DEN
AUS
JAP
GER
USA
UK
100
1000
10000
100000
100
1000
10000
100000
HERD $M 2000 PPP (log scale)
Ph
D
s
award
e
d
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
79
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Natural Sciences
Average
1998 - 2001
Actual
2002
Ratio or
difference
UK PhDs awarded per HERD ($M 2000 PPP) - Actual
2.49
2.59
1.04
1.83
1.76
1
UK PhDs awarded per unit HERD ($M 2000 PPP) - Rank
2
3
-1
OST comparator group average - PhDs awarded per unit HERD
($M 2000 PPP)
Natural Sciences - PhDs per HERD, G8 nations (no data for Canada, France, Italy or Russia)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
1998
1999
2000
2001
2002
Year
Ph
D
s
award
e
d
p
e
r H
E
R
D
$M
2000 PPP
UK
USA
GERMANY
JAPAN
EUROPEAN UNION
Indicator Headline
The indicator compares HERD expenditure
allocated to the Natural Sciences with the
numbers of PhDs awarded in this category.
Here, the Natural Sciences include biological,
physical, environmental and agricultural fields.
The UK produces about 1.5 times as many
PhDs per unit spend as the OST comparator
group average. This is a similar level of
productivity per unit spend to Germany. The
USA is just below the group average and
Japan are just above the group average.
UK PhD data are subject to a change in HESA
data collection from 2001 (see Background).
In this field, the UK currently awards over
5,500 PhDs per year and is the 3rd largest
(behind USA - 11,000 and Germany - 7,000)
by volume.
There is a clear correlation across countries
between funding and PhD output, implying
that probably both scale with an underlying
size factor such as staff numbers, which affect
training and research volumes.
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
80
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Natural Sciences - PhDs awarded relative to HERD - OST comparator group nations, 2002
SUI
EU15
SWE
ESP
POL
NED
FIN
DEN
AUS
JAP
GER
USA
UK
100
1000
10000
100000
100
1000
10000
100000
HERD $M 2000 PPP (log scale)
Ph
D
s
award
e
d
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
81
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Engineering and Technology
Average
1998 - 2001
Actual
2002
Ratio or
difference
UK PhDs awarded per HERD ($M 2000 PPP) - Actual
1.99
1.69
0.85
1.56
1.35
1
UK PhDs awarded per unit HERD ($M 2000 PPP) - Rank
2
5
-3
OST comparator group average - PhDs awarded per unit HERD
($M 2000 PPP)
Indicator Headline
The indicator compares HERD expenditure
allocated to Engineering and Technology with
the numbers of PhDs awarded in this
category.
The UK has around one quarter more PhDs
per unit spend than the OST comparator
group average (for the 5-year period). This is
a similar level of productivity per unit spend to
Finland and Sweden and above Germany
(who are slightly below the OST comparator
group average in this measure). The USA is
similar to Germany, whilst Japan is
substantially below the OST comparator group
average.
The UK is close to the EU average, but its
productivity has dropped over the period.
In Engineering and Technology, the UK
awards about 2,000 PhDs per year (there has
been a drop of about 10% between 2001 and
2002). It is the 4th largest (to USA - 5,400,
Japan - 3,100 and Germany 2,400) by volume.
There are no available data for France, but its
research in this area has historically had good
impact.
UK PhD data are subject to a change in HESA
data collection from 2001 (see Background).
There is a correlation across countries
between funding and PhD output, implying
that probably both scale with an underlying
size factor such as staff numbers, which affect
training and research volumes although other
factors may also play a role.
Engineering and Technology - PhDs per HERD, G8 nations (no data for Canada, France, Italy or
Russia)
0.0
0.5
1.0
1.5
2.0
2.5
1998
1999
2000
2001
2002
Year
Ph
D
s
award
e
d
p
e
r H
E
R
D
$M
2000 PPP
UK
USA
GERMANY
JAPAN
EUROPEAN UNION
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
82
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Engineering and Technology - PhDs awarded relative to HERD - OST comparator nations, 2002
SUI
EU15
SWE
ESP
POL
NED
FIN
AUS
JAP
GER
USA
UK
100
1000
10000
100
1000
10000
HERD $M 2000 PPP (log scale)
Ph
D
s
award
e
d
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
83
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Social Sciences
Average
1998 - 2001
Actual
2002
Ratio or
difference
UK PhDs awarded per HERD ($M 2000 PPP) - Actual
3.01
2.78
0.92
3.69
3.31
1
UK PhDs awarded per unit HERD ($M 2000 PPP) - Rank
5
5
0
OST comparator group average - PhDs awarded per unit HERD
($M 2000 PPP)
Indicator Headline
The indicator compares HERD expenditure
allocated to Social Sciences with the numbers
of PhDs awarded in this category. Social
Sciences includes business and management
fields.
The UK produces rather fewer PhDs per unit
spend than the OST comparator group
average. The range is much wider in this field
than the science-technology-based categories
and other analyses have thrown doubt on data
consistency for some countries.
The UK's average level of productivity per unit
spend is roughly half that of Germany, and
one third that of the USA. In this field, the UK
awards more than 2,000 PhDs per year and is
the 3rd largest (to the USA - 17,000 and
Germany - 4,000) by volume.
UK PhD data are subject to a change in HESA
data collection from 2001 (see Background).
There is a correlation across countries
between funding and PhD output, implying
that probably both scale with an underlying
size factor such as staff numbers, which affect
training and research volumes although other
factors may play a role.
There are no data available for Japan HERD
in Social Sciences.
Social Sciences - PhDs per HERD, G8 nations (no data for Canada, France, Italy or Russia)
0.0
2.0
4.0
6.0
8.0
10.0
12.0
1998
1999
2000
2001
2002
Year
Ph
D
s
award
e
d
p
e
r H
E
R
D
$M
2000 PPP
UK
USA
GERMANY
EUROPEAN UNION
UK - HESA data collection
procedure changed in 2001
(see headline note)
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
84
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Social Sciences - PhDs awarded relative to HERD - OST comparator nations, 2002
EU15
SWE
ESP
FIN
AUS
GER
USA
UK
100
1000
10000
100000
100
1000
10000
HERD $M 2000 PPP (log scale)
Ph
D
s
award
e
d
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
85
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Humanities
Average
1998 - 2001
Actual
2002
Ratio or
difference
UK PhDs awarded per HERD ($M 2000 PPP) - Actual
3.96
3.72
0.94
2.49
2.35
1
UK PhDs awarded per unit HERD ($M 2000 PPP) - Rank
3
3
0
OST comparator group average - PhDs awarded per unit HERD
($M 2000 PPP)
Indicator Headline
The indicator compares HERD expenditure
allocated to Humanities with the numbers of
PhDs awarded in this category. This is the
first time that Humanities data have been
analysed in this indicator. The data coverage
is sparse, reducing the feasibility of
interpretation
The UK produces a much larger number of
PhDs per unit spend than the OST comparator
group. This is a similar average level of
productivity per unit spend to Finland and
approximately twice that of Germany, which
itself is below the OST comparator group
average.
In this field, the UK awards close to 2,000
PhDs per year and is the 3rd largest (after the
USA - 6,000 and just behind Germany also
with 2,000) by volume. The UK recently
started producing more humanities PhDs than
France.
UK PhD data are subject to a change in HESA
data collection from 2001 (see Background).
There are no data available for the USA
HERD in Humanities.
Humanities - PhDs per HERD, G8 nations (no data for USA, Canada, France, Italy or Russia)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
1998
1999
2000
2001
2002
Year
Ph
D
s
award
e
d
p
e
r H
E
R
D
$M
2000 PPP
UK
GERMANY
EUROPEAN UNION
UK - HESA data collection
procedure changed in 2001
(see headline note)
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
86
Description of performance indicator
4.02 PhDs awarded relative to HERD in five main research fields
Condition signalling improvement
Increased ratio
Humanities - PhDs awarded relative to HERD - OST comparator nations, 2002
UK
GER
AUS
ESP
SWE
EU15
100
1000
10000
100
1000
10000
HERD $M 2000 PPP (log scale)
Ph
D
s
award
e
d
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS), OECD Education Database, UK SET statistics and HESA data
87
Description of performance indicator
4.03 PhDs awarded relative to PUBERD
Condition signalling improvement
Increased ratio
Average
1998 - 2001
Actual
2002
Ratio or
difference
UK PhDs awarded per PUBERD (2000 $M PPP) - Actual
1.36
1.54
1.14
OST comparator group average - PhDs awarded per PUBERD
(2000 $M PPP)
1.17
1.15
0.99
UK PhDs awarded per PUBERD (2000 $M PPP) - Rank
6
4
2
UK PhDs awarded per PUBERD (2000 $M PPP) / OST comparator group
average - PhDs awarded per PUBERD (2000 $M PPP)
1.16
1.34
PhDs per PUBERD (2000 $M PPP), G8 nations (no data for Russia)
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
1998
1999
2000
2001
2002
Year
PhD
s awarded per PU
B
E
R
D
(
2000 $M PPP)
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
EUROPEAN UNION
OST GRP AVGE
Data: OECD (MSTI), OECD Education Database
UK - HESA data collection
procedure changed in 2001
(see headline note)
Indicator Headline
The indicator compares the output of people
gaining research degrees with the total public
research spend (PUBERD, the sum of
GOVERD and HERD).
The UK is ranked 1st within the G8 nations in
terms of PhD awards per PUBERD, and 4th
amongst the OST comparator group (behind
Poland, Switzerland and Spain). The UK
remains significantly above the OST
comparator group average throughout the
period 1998-2001.
Over the 5-year period only one other G8
nation has improved the PhD/PUBERD ratio -
Japan by 30%. The ratios for the USA and
France have continued to decline. A UK
decline in 2000 is obscured by the change in
data collection in 2001.
Outside the G8, the PhD/PUBERD ratio of
South Korea has levelled off after an increase
in the previous 4 years of more than 13%.
The EU15 average is fairly level.
UK PhD data are subject to a change in HESA
data collection from 2001 (see Background).
88
Description of performance indicator
4.03 PhDs awarded relative to PUBERD
Condition signalling improvement
Increased ratio
PhDs awarded per PUBERD (2000 $M PPP) - OST comparator group, 2002
EU15
SKO
ISR
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
JAP
ITA
GER
FRA
USA
UK
100
1000
10000
100000
100
1000
10000
100000
PUBERD (2000 $M PPP) (log scale)
P
h
Ds awarded (l
og scal
e)
Data: OECD (MSTI), OECD Education Database
Check Background section of the report for country short codes
89
Description of performance indicator
4.04 Citations relative to GDP
Condition signalling improvement
Improved ratio of citations per GDP compared to recent past
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK citations per GDP (2000 $M PPP) - Actual
0.49
0.12
0.24
OST comparator group average citations per GDP (2000 $M PPP)
0.35
0.09
0.26
UK citations per GDP - Rank
7
7
0
UK cites per GDP / OST comparator group average cites per GDP
1.38
1.27
0.92
Indicator Headline
This is the first of a series of indicators that
measure citation accumulation (hence research
esteem) in relation to the components of
investment.
This indicator measures the association
between total GDP and national citation count.
The UK is the leading country among the G8
group. The UK is well ahead of the USA which
is ranked 11th in the OST group overall (the UK
holds the 7th position). Several countries,
including the UK, Canada and the USA have
declined in performance relative to the OST
average. Germany's performance has recently
levelled, having improved slightly since 1995
throughout the time period.
UK relative international performance remains
well ahead of the average for the OST
comparator group, although it declined slightly
in 2001 and again in 2003 compared to the
OST group average.
The UK is ranked 7th overall for recognition
compared to general expenditure, behind
Israel, Switzerland, the Netherlands and the
Scandinavian countries.
The EU15 combined overall performance is
increasingly similar to that of the USA,
especially since the late 1990s.
Data for Russia are anomalous, because of the
rapid fall in the OECD figures for Russian GDP.
cont./
Citations relative to GDP rebased to OST comparator group average - G8 nations
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
Ci
tati
on /
GDP
rebased to OS
T com
p
arator group average
(=1.
0)
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN
UNION
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
90
Description of performance indicator
4.04 Citations relative to GDP
Condition signalling improvement
Improved ratio of citations per GDP compared to recent past
Citations relative to GDP - OST comparator group nations, 2003
EU15
TWN
SKO
ISR
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
USA
UK
10000
100000
1000000
100000
1000000
10000000
100000000
GDP (2000 $M PPP) (log scale)
Ci
tati
on count (l
og scal
e)
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
Check Background section of the report for country short codes
Indicator Headline cont.
At a gross level, it is not clear that all research
performance should be attributed solely to
specific investment in R&D. There are other
reasons why some economies may tend to
support a greater level of innovation and
inventiveness and these may be linked more
loosely to the overall level of economic activity.
Conversely, some nations may have rather a
low level of research performance compared to
the size of the economy.
Citations accumulate with time, so citation
count for earlier years is greater than for the
more recent years. Direct comparison between
countries is therefore informative but data need
to be rebased for comparison between years.
91
Description of performance indicator
4.05 Number and share of citations relative to GERD
Condition signalling improvement
Increased citations per unit expenditure
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK GERD (2000 $M PPP)
27792
30203
1.09
UK citations / GERD (2000 $M PPP) - Actual
26.24
6.31
0.24
OST comparator group average citations / GERD (2000 $M PPP)
16.04
4.10
0.26
UK citations / GERD (2000 $M PPP) - Rank
3
5
2
2.09
2.08
0.99
UK share of OST comparator group citations/spend - Rank
3
5
2
UK share of OST comparator group citations / UK share of OST
comparator group GERD (2000 $M PPP)
OST comparator group citation share / OST comparator group GERD share for G8 nations
0
0.5
1
1.5
2
2.5
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
Rati
o ci
tati
on share /
GE
RD share
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
Indicator Headline
This indicator compares national citation
counts to Gross Expenditure on R&D (GERD)
including both public and private sector spend.
The private sector generally publishes far less
than the public sector and its expenditure is
therefore less likely to generate citations. As a
consequence, national economies that have a
balance towards the private sector will
generate fewer citations per unit GERD.
Within the OST comparator group the UK has
a strong performance in terms of relative share
of citations compared to relative expenditure.
The UK continues to lead among the G8
nations although its position is perhaps not as
strong as it has been previously.
UK GERD has increased in real terms over the
last ten years and again was greater in 2003
than the average for the recent past. The OST
comparator group average rate of increase is
dominated by USA figures: USA GERD
increased by over one third in the last decade.
Citations accumulate with time and are always
fewer in more recent years. The UK has
dipped very slightly in its relative international
performance on this indicator (an average
change in citations per GERD - ratio between
2003 and recent average is UK = 0.240, OST
comparator group average = 0.255).
Switzerland still leads in this indicator (32.2
cites/$M GERD average in recent 5 years) but
in the last 3 years Poland has risen through the
ranks from around 12th to 2nd or 3rd. The UK
is ranked 5th behind Switzerland, Poland, the
Netherlands and Denmark.
cont./
92
Description of performance indicator
4.05 Number and share of citations relative to GERD
Condition signalling improvement
Increased citations per unit expenditure
Citation counts relative to GERD - OECD countries, 2003
EU15
TWN
SKO
ISR
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
USA
UK
10000
100000
1000000
1000
10000
100000
1000000
GERD (2000 $M PPP) (log scale)
Ci
tati
on count (l
og scal
e)
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
Check Background section of the report for country short codes
Indicator Headline cont.
The EU15 combined overall performance
appears to have stabilised against that of the
USA. The USA has about 0.80 cites/GERD
compared to the EU15 total. The USA remains
larger in absolute terms of GERD (USA =
$M268k in 2003; EU15 = $M189k) and total
citations (USA = 800,000 in 2003; EU15 =
680,000).
93
Description of performance indicator
4.06 Citations relative to PUBERD (GOVERD + HERD)
Condition signalling improvement
Increased citations per unit expenditure
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK GOVERD + HERD (2000 $M PPP)
8930
9379
1.05
UK citations / GOVERD + HERD (2000 $M PPP) - Actual
81.08
20.31
0.25
48.05
12.12
0.25
UK citations per GOVERD + HERD (2000 $M PPP) - Rank
2
3
1
OST comparator group average citations / GOVERD + HERD (2000 $M
PPP)
Indicator Headline
It is useful to know about national research
performance in relation to levels of public
spending on the research base. While GERD
indicates total national expenditure on R&D,
and it is reasonable to anticipate some
interaction and gearing between public and
private sectors, private R&D investment is not
generally aimed at producing papers or people.
Indicator 1.02 shows that UK expenditure on
R&D performed in the public sector has risen in
real terms compared to the recent past. This
increase is less, however, over the last ten
years than for the OST group average.
The UK has maintained its strong relative
international performance in the G8 and the
wider OST comparator group.
Most countries have experienced a reduction in
the absolute number of citations/$M PUBERD
but the relative positions are unchanged.
Thus, the UK (81.1 citations/$M PUBERD on
average for recent years) is ranked 2nd
recently and 3rd in 2003 to Switzerland (129.3
cites/$M) and just behind Denmark. The USA
(55.2 cites/$M) is ranked 9th.
Of the G8, Germany (41.5 cites/$M) has
maintained its upward trend and Japan, though
not increasing as in previous years has
remained stable whilst other countries are
experiencing declines in this measure.
cont./
Citations per GOVERD + HERD rebased to OST comparator group average (= 1.0) - G8
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
Ci
tati
o
n
s p
er GOV
E
RD + HE
RD
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN
UNION
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
94
Description of performance indicator
4.06 Citations relative to PUBERD (GOVERD + HERD)
Condition signalling improvement
Increased citations per unit expenditure
Citation count compared to GOVERD + HERD - OECD countries, 2003
EU15
TWN
SKO
ISR
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
USA
UK
10000
100000
1000000
1000
10000
100000
GOVERD + HERD (2000 $M PPP) (log scale)
Ci
tati
on count (l
og scal
e)
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
Check Background section of the report for country short codes
Indicator Headline cont.
Relative positions within the OST comparator
group average using this indicator show little
change over the ten year period. The
exceptions to this are Denmark and Sweden
who have risen by 4 places (from 6th to 2nd)
and fallen by 4 places (from 2nd to 6th)
respectively.
In this indicator, research recognition as
citation count is compared with R&D carried
out specifically in the Government (GOVERD)
and Higher Education (HERD) sectors
(together designated as PUBERD). These are
the bulk of public sector R&D. It is feasible but
challenging and costly to separate the citations
attributable to papers published by different
sectors in each country. The indicator
therefore currently reflects the relationship
between national research performance and
public sector spend.
Citations accumulate over time and are always
fewer in more recent years. Compared to other
countries, UK performance indexed as citations
per unit PUBERD has remained broadly similar
to the OST comparator group average.
95
Description of performance indicator
4.07 Citations relative to HERD
Condition signalling improvement
Increased citations per unit expenditure
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK HERD (2000 $M PPP)
5786
6462
1.12
UK citations per HERD (2000 $M PPP) - Actual
128.85
29.48
0.23
OST comparator group average citations per HERD (2000 $M PPP)
80.07
18.87
0.24
UK citations per HERD (2000 $M PPP) - Rank
4
2
-2
Citations per unit HERD rebased to OST comparator group average (= 1.0) for G8 nations
0.0
0.5
1.0
1.5
2.0
2.5
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
Ci
tati
ons per HE
RD - (OS
T
com
p
arator group average =
1.
0)
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
Indicator Headline
HERD is Expenditure on R&D performed in the
Higher Education sector. Here, citation counts
are compared with HERD.
UK HERD has increased markedly, over one
third in real terms over the last ten years and
by nearly a quarter between 1999 and 2002
alone. OST comparator group average has
increased by a similar amount over ten years.
Citations accumulate over time and are always
fewer in more recent years. UK actual citations
per HERD naturally decrease since expenditure
is greater and older papers have had more time
to accumulate citations. By comparison to the
average for the OST comparator group, the UK
has about 1.6 times as many citations per unit
expenditure.
The UK's position in 2003 is 2nd (29.5
cites/$M), it has apparently lost some ground to
Switzerland (36.0 cites/$M) and is now only
slightly ahead of Denmark (27.7 cites/$M). The
UK is ranked 4th with regard to average recent
performance in this indicator behind
Switzerland, Denmark and Russia. However,
the UK is still a clear leader amongst the G8
nations.
The EU15 has continued to improve its overall
performance relative to the USA. The ratio of
USA cites/$M HERD in 1995 to that of the
EU15 in the same year was 1.41 and for the
most recent year, 2003, this ratio had dropped
to 1.08. Average HERD for the most recent
five year period for the USA and the EU15 is
close to $40BN.
cont./
96
Description of performance indicator
4.07 Citations relative to HERD
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Citation counts relative to HERD - OECD countries, 2003
EU15
TWN
SKO
ISR
CHI
SUI
SWE
ESP
NED
BEL
AUS
JAP
ITA
GER
FRA
CAN
USA
UK
10000
100000
1000000
1000
10000
100000
HERD (2000 $M PPP) (log scale)
Ci
tati
on count (l
og scal
e)
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
Check Background section of the report for country short codes
Indicator Headline cont.
Russia's performance is affected by volatile
changes in GERD and its components.
The OECD definition of what can be included in
HERD is relatively broad and may include activity
outside universities and colleges in some
countries. There is no implication here that HE is
the only agent acquiring citations. In many
research economies the major driver of
innovative and original research is the higher
education sector. However, the research institute
sector is also an important part of the research
base in the UK and even more so in France,
Germany and Japan. The OECD HERD
definition is flexible enough to permit broad
comparability.
97
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Medical Sciences
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK HERD (2000 $M PPP)
1607
1821
1.13
UK citations per HERD (2000 $M PPP) - Actual
357.79
79.77
0.22
OST comparator group average citations per HERD (2000 $M PPP)
306.98
69.95
0.23
UK citations per HERD (2000 $M PPP) - Rank
3
4
-1
UK cites per HERD / OST comparator group average
1.17
1.14
Indicator Headline
For some countries, both HERD and citation
counts can be disaggregated by OECD field for
analysis in this indicator.
For the Medical Sciences, the UK acquires
more cites per unit HERD than the OST
comparator group average and is ranked 3rd
behind Denmark (on its own at over twice OST
average, but with a somewhat anomalous blip
in 2002) and Finland in recent years. Ranked
1st among G8 nations, UK performance is
broadly constant over the period. From 2001
Spain replaces Finland in second place, and in
2003, the Netherlands edges the UK into 4th
position.
The USA is ranked 7th, at 0.85 OST
comparator group average.
The data for Denmark show a general level of
performance that is exceptional and where the
profile departs substantially from that of other
countries. This is due to relatively low amounts
of HERD expenditure until a marked rise in
2002 which accounts for the drop in 2002.
The EU15 line should be regarded with caution.
The OECD data do not accurately reflect a true
sum and the index is therefore over-inflated.
HERD is Expenditure on R&D performed in the
Higher Education sector. Here, citation counts
are compared with HERD. It is not implied that
HE is the only agent acquiring citations, but in
many research economies it is the major
player.
Medical Sciences - Citations per unit HERD rebased to OST comparator group average (= 1.0) for
OECD nations
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
C
it
at
ions per H
E
R
D
-
(
O
ST com
p
arat
or group
average = 1.
0)
UK
USA
GERMANY
JAPAN
AUSTRALIA
DENMARK
FINLAND
NETHERLANDS
POLAND
SPAIN
SWEDEN
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
98
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Medical Sciences - Citation counts relative to HERD - OECD Countries, 2003
NED
JAP
EU15
SWE
ESP
POL
FIN
DEN
AUS
GER
USA
UK
1000
10000
100000
1000000
10
100
1000
10000
100000
HERD (2000 $M PPP) (log scale)
Ci
tati
on count (l
og scal
e)
Check Background section of the report for country short codes
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
99
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Natural Sciences
Average
1998 - 2002
Actual
2002
Ratio or
difference
UK HERD (2000 $M PPP)
1918
2080
1.08
UK citations per HERD (2000 $M PPP) - Actual
254.94
61.66
0.24
OST comparator group average citations per HERD (2000 $M PPP)
190.60
49.53
0.26
UK citations per HERD (2000 $M PPP) - Rank
2
2
0
UK cites per HERD / OST comparator group average
1.34
1.24
Natural Sciences - Citations per unit HERD rebased to OST comparator group average (= 1.0) for
OECD nations
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
C
it
at
ions per H
E
R
D
-
(
O
ST com
p
arat
or group
average = 1.
0)
UK
USA
GERMANY
JAPAN
AUSTRALIA
DENMARK
FINLAND
NETHERLANDS
POLAND
SPAIN
SWEDEN
Indicator Headline
For some countries, both HERD and citation
counts can be disaggregated by OECD field for
analysis in this indicator.
For the Natural Sciences, the UK acquires
more cites per unit HERD than the OST
comparator group average and is ranked 1st
among G8 nations, and 2nd behind the
Netherlands overall. Recent UK performance
is steady after a period of decline.
The USA is placed bottom of the OST nations
in this indicator, and its performance continues
to decline.
HERD is Expenditure on R&D performed in the
Higher Education sector. Here, citation counts
are compared with HERD. It is not implied that
HE is the only agent acquiring citations, but in
many research economies it is the major
player.
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
100
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Natural Sciences - Citation counts relative to HERD - OECD Countries, 2003
EU15
SWE
ESP
POL
NED
FIN
DEN
AUS
JAP
GER
USA
UK
1000
10000
100000
1000000
100
1000
10000
100000
HERD (2000 $M PPP) (log scale)
Ci
tati
on count (l
og scal
e)
Check Background section of the report for country short codes
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
101
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Engineering and Technology
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK HERD (2000 $M PPP)
1045
1202
1.15
UK citations per HERD (2000 $M PPP) - Actual
58.17
12.25
0.21
OST comparator group average citations per HERD (2000 $M PPP)
45.35
11.38
0.25
UK citations per HERD (2000 $M PPP) - Rank
2
2
0
UK cites per HERD / OST comparator group average
1.28
1.08
Indicator Headline
For some countries, both HERD and citation
counts can be disaggregated by OECD field for
analysis in this indicator.
For Engineering and Technology, the UK
acquires more cites per unit HERD than the
OST comparator group average and is ranked
1st among G8 nations, and 2nd behind
Denmark overall. Its recent performance
shows a steady decline, however.
The USA is ranked 3rd, with Germany in 4th
place, showing improvement over the period.
HERD is Expenditure on R&D performed in the
Higher Education sector. Here, citation counts
are compared with HERD. It is not implied that
HE is the only agent acquiring citations, but in
many research economies it is the major
player.
Engineering and Technology - Citations per unit HERD rebased to OST comparator group average (=
1.0) for OECD nations
0.4
0.9
1.4
1.9
2.4
2.9
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
C
it
at
ions per H
E
R
D
-
(
O
ST com
p
arat
or group
average = 1.
0)
UK
USA
GERMANY
JAPAN
AUSTRALIA
DENMARK
FINLAND
NETHERLANDS
POLAND
SPAIN
SWEDEN
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
102
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Engineering and Technology - Citation counts relative to HERD - OECD Countries, 2003
EU15
SWE
ESP
POL
NED
FIN
DEN
AUS
JAP
GER
USA
UK
1000
10000
100000
10
100
1000
10000
HERD (2000 $M PPP) (log scale)
Ci
tati
on count (l
og scal
e)
Check Background section of the report for country short codes
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
103
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Social Sciences
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK HERD (2000 $M PPP)
738
860
1.16
UK citations per HERD (2000 $M PPP) - Actual
119.93
22.47
0.19
OST comparator group average citations per HERD (2000 $M PPP)
63.05
11.89
0.19
UK citations per HERD (2000 $M PPP) - Rank
2
2
0
UK cites per HERD / OST comparator group average
1.90
1.89
Indicator Headline
For some countries, both HERD and citation
counts can be disaggregated by OECD field for
analysis in this indicator.
Bibliometric data for the Social Sciences
cannot be compared directly with other
disciplines. Recent work has shown that
European comparisons must be made with
caution.
For the Social Sciences, the UK acquires more
cites per unit HERD than the OST comparator
group average and is ranked 2nd behind the
USA overall. Its recent profile shows a steady
performance, while the USA is declining.
Germany is ranked 3rd, showing improvement
over the period.
The EU15 line should be regarded with caution.
The OECD data do not accurately reflect a true
sum and the index is therefore over-inflated.
HERD is Expenditure on R&D performed in the
Higher Education sector. Here, citation counts
are compared with HERD. It is not implied that
HE is the only agent acquiring citations, but in
many research economies it is the major
player.
Social Sciences - Citations per unit HERD rebased to OST comparator group average (= 1.0) for OECD
nations
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
C
it
at
ions per H
E
R
D
-
(
O
ST com
p
arat
or group
average = 1.
0)
UK
USA
GERMANY
AUSTRALIA
DENMARK
FINLAND
POLAND
SPAIN
SWEDEN
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
104
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Social Sciences - Citation counts relative to HERD - OECD Countries, 2003
EU15
SWE
ESP
FIN
DEN
AUS
GER
USA
UK
1000
10000
100000
10
100
1000
10000
HERD (2000 $M PPP) (log scale)
Ci
tati
on count (l
og scal
e)
Check Background section of the report for country short codes
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
105
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Humanities
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK HERD (2000 $M PPP)
405
498
1.23
UK citations per HERD (2000 $M PPP) - Actual
5.24
1.14
0.22
OST comparator group average citations per HERD (2000 $M PPP)
1.09
0.27
0.25
UK citations per HERD (2000 $M PPP) - Rank
1
1
0
UK cites per HERD / OST comparator group average
4.82
4.18
Indicator Headline
This is the first time that data on the
Humanities have been analysed for this
indicator.
For some countries, both HERD and citation
counts can be disaggregated by OECD field for
analysis in this indicator.
For the Humanities, the UK acquires roughly
three times more cites per unit HERD than the
OST comparator group average and is ranked
1st. Its recent performance shows a modest
decline, but no other nation's performance is
close.
As is widely understood, citations in this field
are extremely low compared with other fields.
HERD is Expenditure on R&D performed in the
Higher Education sector. Here, citation counts
are compared with HERD. It is not implied that
HE is the only agent acquiring citations, but in
many research economies it is the major
player.
Humanities - Citations per unit HERD rebased to OST comparator group average (= 1.0) for OECD
nations
0.0
1.0
2.0
3.0
4.0
5.0
6.0
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
C
it
at
ions per H
E
R
D
-
(
O
ST com
p
arat
or group
average = 1.
0)
UK
GERMANY
AUSTRALIA
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
106
Description of performance indicator
4.08 Citations relative to HERD in five main research fields
Condition signalling improvement
Increase in cites per unit spend at the OECD field level
Data: Thomson ISIÂŽ National Science Indicators, OECD (RDS), UK SET statistics and HESA data
Humanities - Citation counts relative to HERD - OECD Countries, 2003
EU15
SWE
ESP
FIN
DEN
AUS
GER
UK
10
100
1000
10000
10
100
1000
10000
HERD (2000 $M PPP) (log scale)
Ci
tati
on count (l
og scal
e)
Check Background section of the report for country short codes
107
Description of performance indicator
5.01 PhDs awarded per researcher
Condition signalling improvement
Increased ratio
Average
1999 - 2001
Actual
2002
Ratio or
difference
UK PhDs awarded per researcher - Actual
0.07
0.08
1.11
OST comparator group average - PhDs awarded per researcher
0.06
0.06
0.99
UK PhDs awarded per researcher - Rank
5
4
1
UK PhDs awarded per researcher / OST comparator group average PhDs
awarded per researcher
1.21
1.36
1.12
PhDs per FTE researcher, G8 nations (no data for Russia)
0.00
0.02
0.04
0.06
0.08
0.10
0.12
1998
1999
2000
2001
2002
Year
P
h
Ds awarded per FTE
researcher
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
EUROPEAN UNION
OST AVERAGE
Data: OECD (MSTI), OECD Education Database
Indicator Headline
The labour productivity of the research base is
measured by the volume of outputs compared
to the numbers of highly skilled people
contributing to research.
This indicator is also one of sustainability. It
measures the relationship between the volume
of 'researchers' (see definition in Glossary)
already in the research base and the output of
highly skilled people gaining research
degrees.
The UK is ranked 2nd behind Germany within
the G8 in terms of PhD awards per
researcher, and 4th within the OST
comparator group (behind Switzerland and
Spain). The UK has remained well ahead of
the OST comparator group average
throughout the period.
The number of PhD awards per researcher in
the UK has declined in the most recent year.
Over the 4-year period it has shown apparent
growth of 15%, but this is largely due to data
adjustment. UK data for 2001 are affected by
a change in HESA data collection.
Both Japan and Belgium have seen increases
in this indicator of nearly 30% and more than
40% respectively. Japan trails the G8,
however, by a significant margin.
Both the EU average and the USA average
are falling over the period. This may have
implications for planning as it will affect the
availability of highly skilled people.
There are no data for China.
108
Description of performance indicator
5.01 PhDs awarded per researcher
Condition signalling improvement
Increased ratio
PhDs awarded per researcher - OST comparator group nations, 2002
EU15
SKO
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
JAP
ITA
GER
FRA
USA
UK
100
1000
10000
100000
1000
10000
100000
1000000
10000000
FTE researchers (log scale)
P
h
Ds awarded (l
og scal
e)
Data: OECD (MSTI), OECD Education Database
Check Background section of the report for country short codes
109
Description of performance indicator
5.02 Publications per researcher
Condition signalling improvement
Increased relative output
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK papers per researcher - Actual
2.17
2.08
0.96
1.38
1.35
0.97
UK papers per researcher - Rank
3
4
-1
0.62
0.63
1.02
OST comparator group average papers per
researcher
OST comparator group average compared to
UK (UK=1)
Ratio of publications to FTE researchers for G8 nations
0
0.5
1
1.5
2
2.5
3
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
P
ublicat
ions per
r
esear
cher
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN UNION
OST GRP AVGE
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
Indicator Headline
Research culture is reflected in the numbers of
publications in international journals that are
produced by each researcher active within the
research base.
The UK's recent (2.17 papers/FTE) and current
(2.08 papers/FTE) performance is 3rd to
Switzerland (2.69, 2.61) and the Netherlands
(2.32, 2.22) as an average for the recent
period (1998-2002) and for 2003. [Italy - see
below].
The UK's ratio of papers per researcher has
fallen between the recent period and 2003.
For the OST comparator group, the average
number of papers per researcher also fell
slightly during this period.
UK output per researcher remains much higher
than the OST comparator group average. In
absolute terms the UK relative output has
improved to more than twice that of the USA.
The UK ranks first in the G8 for the average
recent period whereas several of the G8 are
below the OST comparator group average.
The performance of the EU15 has been fairly
stable over the 10-year period at around 1.4 to
1.5 papers per researcher whilst the USA
output has declined from 1.2 down to 0.95
papers/FTE.
China's productivity has doubled over the ten
years, but is only one-tenth that of the UK.
Italy's indexed papers/FTE researcher has
increased. This may be an artefact, however,
because of a diminishing researcher base.
That decline has stabilised, so a continuing
upward trend would indicate genuine
improvement.
110
Description of performance indicator
5.02 Publications per researcher
Condition signalling improvement
Increased relative output
Publications on ISI databases and numbers of researchers
SAF
EU15
TWN
SKO
SGP
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
USA
UK
10000
100000
1000000
10000000
10000
100000
1000000
10000000
FTE researchers, 2003
P
ubl
icati
ons (sum
m
ed over 5 years 1999-2003)
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
Check Background section of the report for country short codes
Indicator Headline cont.
This is a simple indicator that compares the
number of national publications in journals
recorded on Thomson ScientificÂŽ database
with the numbers of 'researchers' recorded on
OECD databases. Productivity increases with
the number of publications per capita, by
definition. In fact, because publications here
are indexed only from international databases,
this is also a reflection of quality since there
will be other publications for each country that
appear only in more specialist, local and
national-quality journals.
Unfortunately, these data cannot be
disaggregated below national level.
'Researcher' information held by OECD is only
available at the level of the major research
fields within the HE sector, and even here, UK
data is not supplied to OECD in this form as
HESA does not collect data about managers
and technicians in HE. This index is calculated
only for the research base as a whole.
111
Description of performance indicator
5.03 Citations per researcher
Condition signalling improvement
Increase in citation ratio
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK citations per researcher - Actual
10.29
11.09
1.08
OST comparator group average citations per researcher
5.96
6.45
1.08
UK citations per researcher - Rank
3
4
-1
OST comparator group average compared to UK (UK=1)
0.56
0.56
1.00
Numbers of citations per researcher for G8 nations
0
2
4
6
8
10
12
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
Ci
tati
ons per FTE
researcher
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN UNION
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
Indicator Headline
As with indicator 5.02, this is a simple and
readily understood index calculated at the level
of the research base as a whole.
Gross output per capita in international
journals is a good measure of volume of
research activity. It is also important to
consider the effectiveness of that output by
measuring the citations they generate: in other
words, we need a measure of quality as well
as quantity.
For the UK, there has been a clear and
substantial trend of improvement in citations
per researcher throughout the 10-year time
period, indicating a relative improvement in
research quality as measured by peer esteem.
The UK's relative international position has
improved in line with that of Switzerland and
the Netherlands. Along with Denmark, these
nations have an average of more than 10
citations per researcher per year over the 10-
year period.
China's rapidly increasing output (see 2.03) is
generated by a huge researcher population.
Its impact still lags (see 3.01) and its
effectiveness by this indicator is less than 0.5
citations per FTE researcher at present.
Italy has nominally overtaken the UK this year,
but reservations about the apparent Italian
performance (see Indicator 5.02) apply here
also.
cont./
112
Description of performance indicator
5.03 Citations per researcher
Condition signalling improvement
Increase in citation ratio
Numbers of citations and researchers
EU15
TWN
SKO
SAF
SGP
CHI
SUI
SWE ESP
POL
NED
FIN
DEN BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
USA
UK
10000
100000
1000000
10000000
10000
100000
1000000
10000000
FTE researchers, 2002
Ci
tati
ons (sum
m
ed 1998-2002)
Check Background section of the report for country short codes
Indicator Headline cont.
The UK output of citations per FTE researcher
has consistently increased by around 6% per
annum. This is similar to the improvement
shown by Germany. However, the UK accrues
almost twice as many citations per researcher
as either Germany or the USA, and performs
better in this indicator than the G8 nations and
compared to OST comparator group average.
The EU15 citations/researcher have increased
over the 10-year time period (from 4.8 to 6.6)
whilst the USA index has not (remaining
steady at around 5.7 to 5.8).
Citations accumulate over time and on
average are usually fewer in more recent
years. Constant 5-year windows for citation
counts are used in this indicator to adjust for
appropriate comparisons between periods.
Data: Thomson ISIÂŽ National Science Indicators, OECD (MSTI)
113
Description of performance indicator
6.01 Researchers per thousand population
Condition signalling improvement
Increased ratio
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK researchers as a percentage of total population - Actual
0.28
0.30
1.07
0.31
0.34
1.11
UK researchers as a percentage of total population - Rank
14
15
1
0.91
0.88
OST comparator group average - researchers as a percentage of total
population
UK researchers as a percentage of total population / OST comparator
group average researchers as a percentage of total population
Researchers compared with total population, G8 nations
0
0.1
0.2
0.3
0.4
0.5
0.6
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
Researchers as a percentage of total
popul
ati
o
n
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
Data: OECD (MSTI)
Indicator Headline
The numbers of researchers in the population
as a whole provides a measure of the resource
capacity of each country in regard to research
innovation, and an indication of the
technological orientation of the country.
Researchers (see Glossary definition)
constitute less than 0.5% of the population in
those countries in the OST comparator group
for which data are available (22 countries).
For the UK, researchers as a percentage of
the total population has again risen slightly in
2003 compared to the average for the recent
past. This is part of a gradual longer term
trend and is similar to but slightly lower than
the average rise for the OST comparator group
as a whole.
The UK proportion of researchers in the
population is much lower than for most of the
G8 nations and below that of the OST
comparator group average. The UK rank
amongst the OST comparator group in this
indicator has remained steady at around 14.
Many research economies elsewhere in
Europe have increased their relative
researcher population by a greater percentage
than the UK and other G8 nations.
China has a much smaller relative researcher
population but this has risen from 0.04% in
1995 to 0.07% in 2003. The total is about half
that of the USA.
114
Description of performance indicator
6.01 Researchers per thousand population
Condition signalling improvement
Increased ratio
Numbers of researchers compared with total population - OST comparator nations, 2003
EU15
TWN
SKO
SAF
SGP
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
USA
UK
10000
100000
1000000
10000000
1000
10000
100000
1000000
10000000
Population in thousands (log scale)
Researchers (l
og scal
e)
Data: OECD (MSTI)
Check Background section of the report for country short codes
115
Description of performance indicator
6.02 Researchers per thousand workforce
Condition signalling improvement
Increased ratio
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK researchers per 1000 workforce - Actual
5.61
5.92
1.05
OST comparator group average - researchers per 1000 workforce
6.21
6.85
1.10
UK researchers per 1000 workforce - Rank
15
17
-2
0.90
0.86
UK researchers per 1000 workforce / OST comparator group average
researchers per 1000 workforce
Researchers per 1000 workforce, G8 nations
0
2
4
6
8
10
12
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
Researchers per 1000 workforce
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN UNION
Data: OECD (MSTI)
Indicator Headline
The numbers of researchers in the workforce
provides a measure of the resource capacity
of each country in regard to research
innovation.
The UK continues to perform poorly compared
to the other G8 nations in this indicator. The
USA proportion of researchers in the
workforce continues to rise (9.1 per 1000
workforce, recent average) and is well ahead
of the UK (5.6 per 1000 workforce). Finland
remains the leader for this indicator with 13.4
researchers per 1000 workforce. Japan and
Sweden are both around the same recent
average proportion as the USA (9.7 per 1000
workforce in both countries).
The UK's volume of researchers as a
proportion of the workforce has increased from
5.1 per 1000 workforce in 1995 to 5.9 per 1000
workforce in 2003. The ratio between the
latest year and the recent past is in line with
this trend.
The UK has been consistently ranked around
15th over the longer period. Since 2001,
however, both South Korea and Taiwan have
moved ahead of the UK. Recent change for
the UK is similar to but less than the OST
comparator group average. The UK remains
more than 10% below average for the OST
comparator group of 22 countries for which
data were available.
Researchers (see Glossary definition) make
up a specific subcategory of R&D personnel,
and provide a particular measure of research
capacity. Both these terms are somewhat
broadly defined and there may be
inconsistency between national figures.
116
Description of performance indicator
6.02 Researchers per thousand workforce
Condition signalling improvement
Increased ratio
Comparison between numbers of researchers and size of total workforce - OST comparator
nations, 2003
EU15
TWN
SKO
SAF
SGP
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
USA
UK
10000
100000
1000000
10000000
1000
10000
100000
1000000
Workforce in 1000s (log scale)
FTE
Researchers (l
og scal
e)
Data: OECD (MSTI)
Check Background section of the report for country short codes
117
Description of performance indicator
6.03 R&D personnel per hundred population
Condition signalling improvement
Increased ratio
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK R&D personnel as a percentage of total population
0.46
0.45
0.98
0.50
0.53
1.07
UK R&D personnel as a percentage of total population - Rank
14
15
-1
0.93
0.85
OST comparator group average - R&D personnel as a percentage of total
population
UK R&D personnel as a percentage of total population / OST comparator
group average R&D personnel as a percentage of total population
R&D personnel as a percentage of total population G8 nations (no data for USA)
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
R
&
D
Personnel as a percent
age of
t
o
ta
l populat
ion
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN UNION
Data: OECD (MSTI)
Indicator Headline
The numbers of R&D personnel in the
population as a whole provides a measure of
the highly-qualified human knowledge base of
each country in regard to research and
development, and an indication of the
technological orientation of the country.
The UK proportion of R&D personnel has
remained at around 0.45%. Amongst the G8
nations (the USA is not included in this
indicator), the UK has a lower proportion of
R&D personnel in the population than every
member, except Italy. Within the OST
comparator group as a whole the UK is ranked
15th, a decline of 3 positions since 1995. The
UK has also dropped below EU15 average.
R&D personnel account for just over 0.5% of
the workforce, on average, in the 21 OST
comparator group countries for which data are
available (there is no data available for the
USA).
The OST comparator group average has
improved steadily from 0.45% in 1995 to
0.53% in 2003.
R&D personnel (see Glossary definition)
provide a general measure of research
capacity, of which researchers (see Glossary
definition) are a particular set. Both these
terms are somewhat broadly defined and there
may be inconsistency between national
figures.
118
Description of performance indicator
6.03 R&D personnel per hundred population
Condition signalling improvement
Increased ratio
Numbers of R&D personnel compared with total population - OST comparator nations, 2003
EU15
TWN
SKO
SAF
SGP
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
UK
10000
100000
1000000
10000000
1000
10000
100000
1000000
10000000
Total population in 1000s (log scale)
R&D P
ersonnel
(l
og scal
e)
Data: OECD (MSTI)
Check Background section of the report for country short codes
Indicator Headline cont.
For some countries in the OST comparator
group this index has shown consistent year-on-
year growth resulting in substantial overall
improvements across the 10-year period for
Singapore, the Scandinavian countries and to
a lesser extent, Belgium and Spain. In 2003,
these countries, with the exception of Spain,
all had more R&D personnel per 1000
population than the UK.
119
Description of performance indicator
6.04 R&D personnel per hundred workforce
Condition signalling improvement
Increased ratio
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK R&D personnel as a percentage of workforce - Actual
0.92
0.88
0.96
1.01
1.07
1.06
UK R&D personnel as a percentage of workforce - Rank
14
15
-1
0.91
0.83
UK R&D personnel as a percentage of workforce / OST comparator group
average R&D personnel as a percentage of workforce
OST comparator group average - R&D personnel as a percentage of
workforce
R&D personnel as a percentage of workforce, G8 nations (no data for USA)
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
R
&
D
Personnel as a percen
ta
ge of
workf
o
rce
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN UNION
Data: OECD (MSTI)
Indicator Headline
The numbers of R&D personnel in the
workforce provides a measure of the highly-
qualified human knowledge base of each
country in regard to research development.
The UK has less than 1% of its workforce
classified as R&D personnel in OECD data
and its relative concentration has declined
while the average has improved. The UK
percentage has fluctuated around 0.9% in the
last 10 years but has not changed significantly.
The pattern has generally been slightly better
for other G8 nations, except Russia.
The UK is below the OST comparator group
average, which increased steadily over the last
decade. The OST comparator group average
(over 1% of workforce) is for a group of 21
countries, excluding the USA, for which data
are available. Within this group the UK has
been ranked 14th in most years. This index
has changed little for the UK whilst the OST
comparator group average and the EU15
average have increased, reducing the UK
'share' of the total. Meanwhile smaller
countries including Singapore, the
Scandinavian countries and Belgium have
moved ahead of the average.
It might be expected that R&D personnel
should increase as a proportion of workforce in
a knowledge-based economy, but issues of
classification may affect data reporting.
R&D personnel (see Glossary definition)
provide a general measure of research
capacity, of which researchers (see Glossary
definition) are a specific part. Both these
terms are somewhat broadly defined and there
may be inconsistency between national
figures.
120
Description of performance indicator
6.04 R&D personnel per hundred workforce
Condition signalling improvement
Increased ratio
Numbers of R&D personnel compared with workforce - OST comparator nations, 2003
EU15
TWN
SKO
SAF
SGP
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
UK
10000
100000
1000000
10000000
1000
10000
100000
1000000
Workforce in thousands (log scale)
R&D P
ersonnel
(l
og scal
e)
Data: OECD (MSTI)
Check Background section of the report for country short codes
121
Description of performance indicator
6.05 Researchers as a proportion of R&D Personnel
Condition signalling improvement
Increased ratio
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK researchers as a proportion of R&D Personnel - Actual
0.61
0.67
1.10
0.58
0.61
1.04
UK researchers as a proportion of R&D Personnel - Rank
12
8
4
1.05
1.10
OST comparator group average - researchers as a proportion of R&D
Personnel
UK researchers as a proportion of R&D Personnel / OST comparator
group average researchers as a proportion of R&D Personnel
Researchers compared with R&D personnel, G8 nations (no data for USA)
0.3
0.4
0.5
0.6
0.7
0.8
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
Researchers as a proporti
on of R&D P
ersonnel
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
RUSSIA
EUROPEAN UNION
Data: OECD (MSTI)
Indicator Headline
Although numbers of both researchers and
R&D personnel have appeared somewhat
static for the UK as a proportion of workforce,
the data in this indicator suggest that a shift
has taken place. This shift was in line with
other G8 nations but may have accelerated.
What these data show is that researchers are
increasing as a proportion of R&D personnel.
There is no overall decrease in R&D
personnel; for the UK these numbers
increased slightly although they have now
plateaued. This outcome therefore seems to
reflect the growing professionalisation of the
research workforce in a knowledge-based
economy, as traditional technical roles become
more complex and sophisticated.
The UK researchers/R&D personnel ratio has
improved in the last few years both in absolute
and relative terms. There was an increase last
year and an increase of 10% between the
recent and current year figures. In addition,
the UK has improved its ranking from 12th to
8th in the most recent year.
The researcher proportion is highest for
Singapore and South Korea, both countries
with more than three quarters of their
researchers as a proportion of R&D personnel.
China also performs strongly in this regard and
ranks 3rd. Italy, the Netherlands and
Switzerland have the lowest researcher
proportions, all with less than half. There are
no USA data.
Researchers (see Glossary definition) make up
a specific subcategory of R&D personnel, and
provide a particular measure of research
capacity. Both these terms are somewhat
broadly defined and there may be
inconsistency between national figures.
122
Description of performance indicator
6.05 Researchers as a proportion of R&D Personnel
Condition signalling improvement
Increased ratio
Numbers of researchers compared with R&D personnel - OST comparator nations, 2003
EU15
TWN
SKO
SAF
SGP
CHI
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
RUS
JAP
ITA
GER
FRA
CAN
UK
10000
100000
1000000
10000000
10000
100000
1000000
10000000
R&D Personnel (log scale)
Researchers (l
og scal
e)
Data: OECD (MSTI)
Check Background section of the report for country short codes
123
Description of performance indicator
7.01 Business R&D investment in publicly performed R&D (BE-PUBERD as a proportion of PUBERD
Condition signalling improvement
Increased ratio
Average
1998 - 2002
Actual 2003
Ratio or
difference
UK - percentage of BE-PUBERD within total PUBERD - Actual
9.16
6.61
0.72
6.82
6.78
0.99
UK - percentage of BE-PUBERD within total PUBERD - Rank
5
8
-3
1.34
0.98
OST comparator group average percentage of BE-PUBERD within total
PUBERD
UK percentage of BE-PUBERD within total PUBERD / OST comparator
group average percentage of BE-PUBERD within total PUBERD
BE-PUBERD as a percentage of PUBERD - G8 nations (no data for Russia)
0
2
4
6
8
10
12
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
BE
-P
UBE
RD as p
ercen
tag
e o
f P
U
BE
RD
UK
USA
CANADA
FRANCE
GERMANY
ITALY
JAPAN
EUROPEAN
UNION
Data: OECD (RDS & MSTI)
Indicator Headline
BE-PUBERD is R&D performed in the publicly
funded sector (Government plus Higher
Education) that is funded by the Business
Enterprise sector. Business expenditure via
investment in other sectors may reflect
confidence in the research and relevance to
business objectives.
The UK had ranked ahead of other G8 nations
but has dropped behind Germany and Canada.
It is now close to OST comparator group and
EU15 average, but has fallen to 8th among
OST comparator group nations in terms of
business investment in publicly funded R&D.
This puts it behind Poland (which bounced
back from a 'data' drop last year), the
Netherlands, Belgium and Finland.
The biggest increases on the recent 5 year
average are Italy (28%), Switzerland and
Denmark (17%) and the Netherlands, Finland
and Germany (9%) compared with a stable
OST comparator group average.
The long term trend for the EU15 seems to be
an increase on this indicator. By contrast, the
USA lags on Europe and is declining.
China ranks 19th in the OST comparator
group.
Some 2003 data are produced by forecasting.
Forecasted data are 'sense checked' by
comparing forecasts in previous reports with
outcomes.
124
Description of performance indicator
7.01 Business R&D investment in publicly performed R&D (BE-PUBERD as a proportion of PUBERD
Condition signalling improvement
Increased ratio
Business Enterprise expenditure in PUBERD - OST comparator nations, 2003
EU15
SWE
ESP
POL
NED
FIN BEL
AUS
JAP
GER
FRA
CAN
USA
UK
100
1000
10000
1000
10000
100000
PUBERD (2000 $M PPP) (log scale)
BE
-P
UBE
RD (2000 $M
P
P
P
) (l
og scal
e)
Data: OECD (RDS & MSTI)
Check Background section of the report for country short codes
125
Description of performance indicator
7.02 Business R&D investment in GOVERD (BE-GOVERD as a percentage of GOVERD)
Condition signalling improvement
Increased ratio
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK percentage of BE-GOVERD within total GOVERD - Actual
13.58
8.95
0.66
7.88
7.99
1.01
UK percentage of BE-GOVERD within total GOVERD - Rank
4
6
-2
1.72
1.12
UK percentage of BE-GOVERD within total GOVERD / OST comparator
group average percentage of BE-GOVERD within total GOVERD
OST comparator group average percentage of BE-GOVERD within total
GOVERD
BE-GOVERD as a percentage of GOVERD - G8 nations, not USA (no data for Russia)
0
2
4
6
8
10
12
14
16
18
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
BE
-GOV
E
R
D as a percentage of GOV
E
R
D
UK
CANADA
FRANCE
GERMANY
ITALY
JAPAN
EUROPEAN UNION
Data: OECD (RDS & MSTI)
Indicator Headline
BE-GOVERD is R&D performed in the
Government sector that is funded by the
Business Enterprise sector.
For the UK, R&D expenditure from business
sources as a proportion of GOVERD remains
volatile, having dipped to a low-point in the mid-
1990s, peaked in 1999 and is now dipping
again. Nonetheless, UK BE-GOVERD is still
above the OST comparator group average.
The UK remains ahead in the G8 group and
ranks 4th to the Netherlands (which is of a
similar scale in terms of BE-GOVERD), Finland
and Poland overall, declining to 6th in 2003.
There have been substantial improvements in
both Finland and Poland over the period.
There are no data for the USA or China.
Some 2003 data are produced by forecasting.
Forecasted data are 'sense checked' by
comparing forecasts in previous reports with
outcomes.
126
Description of performance indicator
7.02 Business R&D investment in GOVERD (BE-GOVERD as a percentage of GOVERD)
Condition signalling improvement
Increased ratio
BE-GOVERD compared to GOVERD - OST comparator nations, 2003 not USA
EU15
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
JAP
ITA
GER
FRA
CAN
UK
1
10
100
1000
10000
100
1000
10000
100000
GOVERD (2000 $M PPP) (log scale)
BE
-GOV
E
R
D (2000 $M
P
P
P
) (l
og scal
e)
Data: OECD (RDS & MSTI)
Check Background section of the report for country short codes
127
Description of performance indicator
7.03 Business R&D investment in PNPERD (BE-PNPERD as a percentage of PNPERD)
Condition signalling improvement
Increased ratio
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK percentage of BE-PNPERD within total PNPERD - Actual
12.84
10.50
0.82
18.07
18.02
1.00
UK percentage of BE-PNPERD within total PNPERD - Rank
10
10
0
0.71
0.58
OST comparator group average percentage of BE-PNPERD within total
PNPERD
UK percentage of BE-PNPERD within total PNPERD / OST comparator
group average percentage of BE-PNPERD within total PNPERD
BE-PNPERD as a percentage of PNPERD - G8 nations (no data for Germany, Italy or Russia)
0
10
20
30
40
50
60
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
BE
-P
NP
E
RD as a percentage of P
N
P
E
R
D
UK
USA
CANADA
FRANCE
JAPAN
EUROPEAN
UNION
Data: OECD (RDS)
Indicator Headline
BE-PNPERD is R&D performed in the Private-
Non-Profit sector that is funded by the
Business Enterprise sector.
This indicator has seen substantial change
since last year due to revision to UK PNPERD
data. Total PNPERD has seen a substantial
increase, particularly in recent years, leading to
an apparent relative decrease in the proportion
funded by the Business Enterprise sector.
UK business source expenditure as a
proportion of PNPERD dropped by 50%
between 1998 and 2002, beginning to show
growth again only in 2003.
Among G8 nations, the UK is 4th with Canada
and France well behind Japan. Japan is
exceptional on this indicator. Among the OST
comparator group, the UK's rank position has
remained steady at 10.
In terms of absolute scale of BE-PNPERD,
apart from Japan and the US, only France and
South Korea (another strong performer in this
indicator) approach the UK level.
The OST comparator group average has
remained almost static, but there are marked
variations between countries and greater
fluctuations within some such as Sweden.
Spain, in particular, has markedly improved its
position.
Some 2003 data are produced by forecasting.
Forecasted data are 'sense checked' by
comparing forecasts in previous reports with
outcomes.
128
Description of performance indicator
7.03 Business R&D investment in PNPERD (BE-PNPERD as a percentage of PNPERD)
Condition signalling improvement
Increased ratio
Business Enterprise expenditure in PNPERD - OST comparator nations, 2003
EU15
SUI
SWE
ESP
POL
NED
FIN
DEN
BEL
AUS
JAP
FRA
CAN
USA
UK
0.1
1
10
100
1000
10000
1
10
100
1000
10000
100000
PNPERD (2000 $M PPP) (log scale)
BE
-P
NP
E
R
D (1995 $M
P
P
P
) (l
og scal
e)
Data: OECD (RDS)
Check Background section of the report for country short codes
129
Description of performance indicator
7.04 Business R&D investment in HERD (BE-HERD as a percentage of HERD)
Condition signalling improvement
Increased ratio
Average
1998 - 2002
Actual 2003
Ratio or
difference
UK - percentage of BE-HERD within total HERD - Actual
6.65
5.56
0.84
6.82
6.83
1.00
UK - percentage of BE-HERD within total HERD - Rank
7
10
-3
0.98
0.81
UK percentage of BE-HERD within total HERD / OST comparator group
average percentage of BE-HERD within total HERD
OST comparator group average percentage of BE-HERD within total
HERD
Business R&D investment in HERD - G8 nations (no data for Russia or Italy)
0
2
4
6
8
10
12
14
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
BE
-HE
RD as p
ercen
tag
e o
f HE
RD
UK
USA
CANADA
FRANCE
GERMANY
JAPAN
EUROPEAN
UNION
Data: OECD (RDS & MSTI)
Indicator Headline
BE-HERD is R&D performed in the Higher
Education sector that is funded by the
Business Enterprise sector. There is a general
but rather weak correlation between volume
HERD and the level of BE-HERD (see graph).
In other words, BEHERD as a share of HERD
is fairly consistent across a range of countries.
For the UK, the overall increase in relative
business R&D funding within the HE sector in
the mid-1990s plateaued, dipped at the turn of
the millennium, and is now below the EU15
average. There has been a very similar dip for
the USA which is now well below the EU.
BE-HERD has fallen back in South Korea,
where it was high, but is strong and continues
to be maintained in Germany. Both Canada
and Germany are ahead of the UK in absolute
and relative terms. Their data cannot,
unfortunately, be disaggregated at the more
detailed level of Indicator 7.05. Belgium is a
srong performer in this indicator, and
Switzerland, Finland and the Netherlands all
show strong recent growth.
The OST comparator group average fluctuates
from year to year. It generally fell from the
early 1990s but picked up in 2001 because of
simultaneous increases in a number of
European countries. The UK is consequently
now below group average and its rank position
has fallen from 7th to 10th.
Some 2003 data are produced by forecasting.
Forecasted data are 'sense checked' by
comparing forecasts in previous reports with
outcomes.
130
Description of performance indicator
7.04 Business R&D investment in HERD (BE-HERD as a percentage of HERD)
Condition signalling improvement
Increased ratio
Business Enterprise expenditure in HERD - OST comparator nations, 2003
EU15
SWE
ESP
NED
BEL
AUS
JAP
GER
FRA
CAN
USA
UK
100
1000
10000
1000
10000
100000
HERD (2000 $M PPP) (log scale)
BE
-HE
R
D (2000 $M
P
P
P
) (l
og scal
e)
Data: OECD (RDS & MSTI)
Check Background section of the report for country short codes
131
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Medical Sciences
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK - Percentage of BE-HERD within total HERD - Actual
6.51
5.47
0.84
5.85
6.00
1.03
UK - Percentage of BE-HERD within total HERD - Rank
3
5
-2
1.11
0.91
OST comparator group average percentage of BE-HERD within total
HERD
UK - Percentage of BE-HERD within total HERD / OST comparator group
average percentage of BE-HERD within total HERD
Indicator Headline
BE-HERD is R&D performed in the Higher
Education sector that is funded by the
Business Enterprise sector.
Data for OECD field Medical Sciences BE-
HERD are available only for 8 countries.
These do not include USA, Canada, France
and Germany. This explains the somewhat
unusual mix of countries shown in the attached
graph.
UK Medical Sciences BE-HERD is second
highest to Japan in absolute terms and third
highest to Poland and Spain as a proportion of
HERD among OST comparator group
countries.
The UK's steady decline reflects a fall-off in BE-
HERD since 2000 in the context of (sometimes
rapidly) rising HERD. Both BE-HERD and
HERD decreased in 2003.
Denmark has been very volatile and Finland
and Poland have seen significant fluctuations
although on a smaller volume than the UK.
Increases have not always been sustained.
Some 2003 data are produced by forecasting.
Forecasted data are 'sense checked' by
comparing forecasts in previous reports with
outcomes.
The EU15 figure is based on a limited data set.
Medical Sciences - BE-HERD as a percentage of HERD
0
2
4
6
8
10
12
14
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
BE
-HE
RD a
s
a
pe
rc
e
n
ta
ge
of HE
RD
UK
JAPAN
AUSTRALIA
DENMARK
FINLAND
POLAND
SPAIN
SWEDEN
EUROPEAN UNION
Data: OECD (RDS & MSTI), UK SET statistics and HESA
132
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Medical Sciences - Business Enterprise expenditure in HERD - OST comparator nations,
2003
UK
JAP
AUS
DEN
FIN
POL
ESP
SWE
EU15
1
10
100
1000
10
100
1000
10000
100000
HERD (2000 $M PPP) (log scale)
B
E
-H
ER
D
(
2000 $M
PPP)
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS & MSTI), UK SET statistics and HESA data
133
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Natural Sciences
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK - Percentage of BE-HERD within total HERD - Actual
6.70
5.25
0.78
4.30
4.27
0.99
UK - Percentage of BE-HERD within total HERD - Rank
1
3
-2
1.56
1.23
OST comparator group average percentage of BE-HERD within total
HERD
UK - Percentage of BE-HERD within total HERD / OST comparator group
average percentage of BE-HERD within total HERD
Natural Sciences - BE-HERD as a percentage of HERD
0
1
2
3
4
5
6
7
8
9
10
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
BE
-HE
RD a
s
a
pe
rc
e
n
ta
ge
of HE
RD
UK
JAPAN
AUSTRALIA
DENMARK
FINLAND
POLAND
SPAIN
SWEDEN
EUROPEAN UNION
Indicator Headline
BE-HERD is R&D performed in the Higher
Education sector that is funded by the Business
Enterprise sector.
Data are available only for 8 countries at the
OECD field level. These do not include USA,
Canada, France and Germany. This explains
the somewhat unusual mix of countries shown
in the attached graph.
UK Natural Sciences BE-HERD is still highest
in absolute terms (by 1/3rd over Spain) but is
now 3rd as a proportion of HERD among the
group of comparator countries. This element of
UK business funding declined in 2001 after
substantial growth in the last decade.
UK research impact is also strong in this area
and research volume is high in Life Sciences.
Relatively high levels of business investment
may reflect confidence in UK research.
Spain, Finland and - to a lesser extent -
Australia have seen significant relative
increases.
The average OST comparator group figure for
this indicator is fairly stable.
Some 2003 data are produced by forecasting.
Forecasted data are 'sense checked' by
comparing forecasts in previous reports with
outcomes.
Data: OECD (RDS & MSTI), UK SET statistics and HESA data
134
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Natural Sciences - Business Enterprise expenditure in HERD - OST comparator nations, 2003
UK
JAP
AUS
DEN
FIN
POL
ESP
SWE
EU15
1
10
100
1000
100
1000
10000
100000
HERD (2000 $M PPP) (log scale)
B
E
-H
ER
D
(
2000 $M
PPP)
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS & MSTI), UK SET statistics and HESA data
135
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Engineering and Technology
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK - Percentage of BE-HERD within total HERD - Actual
10.85
9.57
0.88
9.34
8.98
0.96
UK - Percentage of BE-HERD within total HERD - Rank
4
5
-1
1.16
1.06
OST comparator group average percentage of BE-HERD within total
HERD
UK - Percentage of BE-HERD within total HERD / OST comparator group
average percentage of BE-HERD within total HERD
Indicator Headline
BE-HERD is R&D performed in the Higher
Education sector that is funded by the
Business Enterprise sector.
Data are available only for 8 countries at the
OECD field level. These do not include USA,
Canada, France and Germany. This explains
the somewhat unusual mix of countries shown
in the attached graph.
UK Engineering and Technology BE-HERD is
highest in absolute terms (just ahead of Japan)
among OST comparator group countries but
lower as a proportion of HERD than that of
Finland. It is of a similar proportion to Spain
and Australia. This element of business
funding has fallen only marginally in absolute
terms and as a proportion of HERD in the UK
in the last decade.
Denmark has seen a significant increase
moving ahead of the UK in 2003, while
Sweden has fallen back somewhat on this
indicator.
UK research impact had been less strong in
this area but has recently improved (see
Indicator 3.09). The pattern of business
investment could reflect wider issues, but the
high absolute level should be noted as an
expression of confidence. Annual variations
may reflect other economic issues faced by the
engineering industry in the UK.
Some 2003 data are produced by forecasting.
Forecasted data are 'sense checked' by
comparing forecasts in previous reports with
outcomes.
Engineering and Technology - BE-HERD as a percentage of HERD
0
5
10
15
20
25
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
BE
-HE
RD a
s
a
pe
rc
e
n
ta
ge
of HE
RD
UK
JAPAN
AUSTRALIA
DENMARK
FINLAND
POLAND
SPAIN
SWEDEN
EUROPEAN UNION
Data: OECD (RDS & MSTI), UK SET statistics and HESA data
136
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Engineering and Technology - Business Enterprise expenditure in HERD - OST comparator
nations, 2003
EU15
SWE
ESP
POL
FIN
DEN
AUS
JAP
UK
1
10
100
1000
10
100
1000
10000
HERD (2000 $M PPP) (log scale)
B
E
-H
ER
D
(
2000 $M
PPP)
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS & MSTI), UK SET statistics and HESA data
137
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Social Sciences
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK - Percentage of BE-HERD within total HERD - Actual
4.43
3.44
0.78
2.82
2.82
1.00
UK - percentage of BE-HERD within total HERD - Rank
2
3
-1
1.57
1.22
UK percentage of BE-HERD within total HERD / OST comparator group
average percentage of BE-HERD within total HERD
OST comparator group average percentage of BE-HERD within total
HERD
Indicator Headline
BE-HERD is R&D performed in the Higher
Education sector that is funded by the
Business Enterprise sector.
Data are available only for 7 countries at the
OECD field level. These do not include USA,
Canada, France, Germany or Japan. This
explains the somewhat unusual mix of
countries shown in the attached graph.
UK Social Sciences BE-HERD as a proportion
of HERD continues to be the highest in
absolute terms among comparator countries,
marginally ahead of Spain.
The UK is third highest to Spain and Finland as
a proportion of HERD. Finland has improved
significantly over the period, and overtook the
UK in 2001. The Poland line drops because
BE-HERD declines to almost zero while HERD
steadily rises in that country.
This element of business funding has grown
slowly in absolute terms, in line with total
HERD in the UK in the last decade. This
pattern is typical of the rest of the group.
Overall across the OST comparator group
there is a steady rise in business expenditure
in these HE research areas in line with the
general level of public investment.
Some 2003 data are produced by forecasting.
Forecasted data are 'sense checked' by
comparing forecasts in previous reports with
outcomes
Social Sciences - BE-HERD as a percentage of HERD
-1
0
1
2
3
4
5
6
7
8
9
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
BE
-HE
RD a
s
a
pe
rc
e
n
ta
ge
of HE
RD
UK
AUSTRALIA
DENMARK
FINLAND
POLAND
SPAIN
SWEDEN
EUROPEAN UNION
Data: OECD (RDS & MSTI), UK SET statistics and HESA data
138
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Social Sciences - Business Enterprise expenditure in HERD - OST comparator nations, 2003
UK
AUS
DEN
FIN
ESP
SWE
EU15
1
10
100
10
100
1000
10000
HERD (2000 $M PPP) (log scale)
B
E
-H
ER
D
(
2000 $M
PPP)
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS & MSTI), UK SET statistics and HESA data
139
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Humanities
Average
1998 - 2002
Actual
2003
Ratio or
difference
UK - Percentage of BE-HERD within total HERD - Actual
1.30
1.07
0.82
1.76
2.00
1.14
UK - percentage of BE-HERD within total HERD - Rank
3
3
0
0.74
0.54
OST comparator group average percentage of BE-HERD within total
HERD
UK percentage of BE-HERD within total HERD / OST comparator group
average percentage of BE-HERD within total HERD
Indicator Headline
This is the first time that data for this indicator
have been analysed for Humanities and Arts.
BE-HERD is R&D performed in the Higher
Education sector that is funded by the
Business Enterprise sector.
Data are available only for 7 countries at the
OECD field level for Humanities & Arts. These
do not include USA, Canada, France,
Germany or Japan. Data for Sweden only
cover some years. This explains the
somewhat unusual mix of countries shown in
the attached graph.
UK BE-HERD is about half that of Spain but
much more in absolute terms than other
countries. As a proportion of HERD, the UK is
broadly in the main group of countries where
data are available. The level of investment
has been stable over the period.
The index for Spain is extremely high and this
is due to high BE-HERD not low HERD. The
Poland figures are slightly erratic.
Excepting Spain, the UK appears to attract a
relatively good level of BE-HERD into a well
supported research base in these disciplines.
Some 2003 data are produced by forecasting.
Forecasted data are 'sense checked' by
comparing forecasts in previous reports with
outcomes.
Humanities - BE-HERD as a percentage of HERD
0
1
2
3
4
5
6
7
8
9
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
BE
-HE
RD a
s
a
pe
rc
e
n
ta
ge
of HE
RD
UK
AUSTRALIA
DENMARK
FINLAND
POLAND
SPAIN
SWEDEN
EUROPEAN UNION
Data: OECD (RDS & MSTI), UK SET statistics and HESA data
140
Description of performance indicator
7.05 Business R&D investment in HERD (BE-HERD as %ge of HERD) in five main research fields
Condition signalling improvement
Increased ratio
Humanities - Business Enterprise expenditure in HERD - OST comparator nations, 2003
EU15
ESP
POL
FIN
DEN
AUS
UK
0.1
1
10
100
10
100
1000
10000
HERD (2000 $M PPP) (log scale)
B
E
-H
ER
D
(
2000 $M
PPP)
(
lo
g
scale)
Check Background section of the report for country short codes
Data: OECD (RDS & MSTI), UK SET statistics and HESA data
141
142
Background to the indicators
The following pages provide background information on data sources,
international coverage, subject level disaggregation, time frames and the theory
and methodology used in bibliometric analyses.
Codes and abbreviations for countries and for fields of research are defined in
the appropriate sections.
There is also a glossary for other terminology and abbreviations.
143
Data and sources
With the exception of the UK, the main data sources used for OST indicators
are:
â˘
Finance and people â OECD
â˘
Publications â Thomson Scientific
ÂŽ
For the UK, OST has drawn our attention to the most recent data available from
SourceOECD and ONS.
The OECD is the main provider of internationally comparable data on research
and development. Its two products on the measurement of science and
technology, Main Science and Technology Indicators (MSTI, 2005-1, first
edition) and Research and Development Statistics (RDS, 2004-2, second
edition) provide the basis for much of the data used in these analyses. The
latest editions of each of these products were used to generate the indicators
listed in the table below. In addition, OECD provides the only reliable
international comparisons of educational data via its online Education
Database.
The OECD provides comments on a number of the data points in RDS and
MSTI, explaining their derivation or discussing their accuracy. These
comments have not been reproduced here but are available to the interested
reader when referring to the original data. Sources can be found at:
http://www1.oecd.org/dsti/sti/stat-ana/index.htm
There are some points of difference between MSTI and RDS. MSTI has been
the preferred database for most of the analyses in this report as it provides data
on a greater range of countries. RDS, however, provides data disaggregated at
the level of fields of science and also covers PNPERD.
Data are presented for the years 1995 to 2003/4. Gaps of one or two years in
the time series have been filled by interpolation, whilst missing values at either
end of the time series were created by extrapolation where appropriate.
Rolling 5 year averages have been created for researchers, population and
labour-force data in order to enable like-for-like comparisons with ISI data. The
average is produced from the value for the year in question, and the four years
which precede it.
In this report, OECD data are usually available for 21 countries. Coverage for
the 17 OECD nations is broadly complete, but data for some countries are
missing from some tables. This may be because there were no data available,
or that there were so many missing data points in the data available that no
meaningful attempt to interpolate and extrapolate could be made.
Where necessary and feasible, OECD data has been supplemented by data
sourced from EUROSTAT, the UN, the Higher Education Statistics Agency
(HESA), and the statisticsâ portals of individual national governments.
Financial data is given in units of Million constant US$ at 2000 prices and
corrected for Purchasing Power Parity (PPP). This is a change from last year,
where PPP-1995 was used.
The interpretation of OECD science and technology data is governed by the
Frascati Manual, which has become the internationally recognised methodology
for collecting and using R&D statistics. Some basic definitions from the
Frascati manual appear below; detail is in the Glossary.
The OECD Education Database provides internationally comparable data on
key aspects of education systems. It makes use of data collected by UNESCO,
OECD and EUROSTAT. The interpretation of OECD education data is
governed by the OECD publication âData Collection on Education Systems:
Definitions, Explanations, and Instructionsâ.
Indicator Basic
source
BE-GOVERD
OECD Research and Development Statistics
BE-HERD
OECD Research and Development Statistics
BE-HERD by field of science
OECD Research and Development Statistics
BE-PNPERD
OECD Research and Development Statistics
GDP
OECD Main Science and Technology Indicators
GERD
OECD Main Science and Technology Indicators
GOVERD
OECD Main Science and Technology Indicators
HERD
OECD Main Science and Technology Indicators
HERD by field of science
OECD Research and Development Statistics
National populations
OECD Main Science and Technology Indicators
PhD graduates
OECD Education Database
PhD graduates by field of science
OECD Education Database
PNPERD
OECD Research and Development Statistics
R&D personnel
OECD Main Science and Technology Indicators
Researchers
OECD Main Science and Technology Indicators
Labour (work) force
OECD Main Science and Technology Indicators
144
There has been a change in UK postgraduate data because of changes in data
collection by HESA, the UK Higher Education Statistics Agency. This affects
the status of awards made from dormant registrations and increases the
numbers of PhD awards from 2001 onwards by about 4.5% compared to
previous data. More information is available in an article published by HESA at
the time, see:
http://www.hesa.ac.uk/holisdocs/pubinfo/student/changes.htm
All publication and citation data are provided by Thomson ScientificÂŽ. The
NSICCOD2004 was the specific database from which figures were taken for
these analyses. Two main methods are used in analysing these data:
â˘
NSI1
: analyses based on data from the most recent (or any specific)
calendar year use the ISI NSI1 data frame, looking at the numbers of
articles published and the citations they have accumulated to date;
â˘
NSI5
: analyses based on a select period are most effective if a five-
year window is taken, using the ISI NSI5 data frame. This takes the
publications for a stated five-year period (e.g. NSI5 for 2004 is the 5-
year window 2000-2004) and the citations to those articles in the same
five-year period.
Frascati manual data definitions (
see also Glossary
)
GERD: Gross domestic expenditure on R&D
BERD: Total R&D performed in the business sector
GOVERD: Total R&D performed in the government sector
PNPERD: Total R&D performed in the private non-profit sector
HERD: Total R&D performed in the higher education sector (which is very
broadly defined and may in some countries cover more than
universities and colleges)
Where the prefix BE- is used, this denotes that portion of the R&D within the
sector concerned that is financed by the business enterprise sector (a
table below illustrates this)
Researchers: professionals engaged in the conception or creation of new
knowledge, products, processes, methods and systems
R&D personnel: all persons employed directly on R&D, and those providing
direct services such as managers, administrators, and clerical staff
Labour force (workforce): Total number of persons available for work, whether
in employment or not
Other data definitions
PUBERD: the sum of GOVERD and HERD, equating to R&D performed in the
publicly funded sectors
Notes on data manipulation
Extrapolation
was achieved using Excelâs FORECAST function. Forecasted
data are 'sense checked' by comparing forecasts in previous reports with
outcomes. Where outcome clearly differed from forecast, e.g. because of
marked inflections in trend, then forecasting has been omitted for the latest
report.
Interpolation
was achieved by adding the difference between the available
upper and lower values divided by the number of missing years to the lower
figure.
OECD field of science categories:
a single category covers both agriculture
and natural sciences
GDP
(Million constant $ 2000 prices and PPPs) was translated to GDP (Million
current PPP$) using OECDâs Implicit GDP Price Indices (2000 = 1.00)
UK HERD
was rebuilt by field of science using:
â˘
HESA data on Total HEI Research Grant & Contract Income (from
Resources of Higher Education, Table 4: Research Grants and
Contracts Income by Cost Centre, Source and Institution). HESA cost
centre codes were mapped to OECD fields of science, and agricultural
sciences combined into natural sciences, and income allocated to
administration and services (<1% of the total) was pro-rated across
OECD fields of science.
â˘
OST data on HEFC R&D Expenditure by subject area (ONS
Government R&D survey reproduced on the OSTâs SET statistics
website: Table 5.3 Higher Education Funding Councils R&D and SET
expenditure by subject area
(
http://www.ost.gov.uk/setstats/index.htm
).
This accounted for 95% of HERD on average, and the shortfall was pro-rated
across OECD fields of science.
145
Proportions of HERD by field of science were then calculated, and values
extrapolated for 1995 and 2003. These were then used to split the available
totals for these years.
UK BE-HERD
was rebuilt by field of science using:
â˘
HESA data on Total HEI Research Grant & Contract Income (UK
industry, commerce, public corps) (from Resources of Higher
Education, Table 4: Research Grants and Contracts Income by Cost
Centre, Source and Institution). HESA cost centre codes were
mapped to OECD fields of science, and agricultural sciences
combined into natural sciences, and income allocated to administration
& services (<1% of the total) were pro-rated across OECD fields of
science.
The following table is adapted from Table 6.1 of the Frascati manual. It shows the distinction between funding and performing sector in establishing the composition of
HERD and sets out the disaggregation of HERD in order to illustrate the nature of BE-HERD, business funded research performed by the HE sector.
Business enterprise
Private non-profit
Government
Higher education
Business enterprise
BE-BERD, i.e. private
sector R&D financed
by companies
BE-PNPERD
BE-GOVERD
BE-HERD, e.g.
industrial research
contracts to universities
Total domestic performance
financed by the business
enterprise sector
Government
GOV-BERD, i.e.
Government R&D
contracts and grants to
industry
GOV-HERD e.g.
contracts from
Government
departments
Total domestic performance
financed by the government
sector
Public general university funds (GUF)
GUF, i.e. from DfES via
HEFCs
Total domestic performance
financed by public general
university funds (GUF)
Higher education
HE-HERD, i.e. from
own funds incl.
endowments
Total domestic performance
financed by the higher
education sector
Private non-profit (PNP)
PNP-HERD
Total domestic performance
financed by the private non-
profit sector
Abroad
HERD other
Total domestic performance
financed by abroad
Total
Total performed in the
business enterprise
sector
Total performed in the
private non-profit
sector
Total performed in the
government sector
Total performed in the
higher education sector
BERD
PNPERD
GOVERD
HERD
<
----G
E
R
D--
-
-
>
Sector of funding source
< - - - - GERD - - - - >
< - PUBERD (OST category) - >
Total
Sector of performance
146
OECD Indicators
were created from the following source files, fields and criteria:
Indicator element
OECD source filename
Fields & criteria
BE-GOVERD
RDS2004-2 Table 1. Gross domestic expenditure on R&D --
GERD -- by sector of performance and source of funds
MEASURE=Million constant $ 2000 prices and PPPs;
SECTFIN=Gov. fin. by Bus. enter.
BE-HERD
RDS2004-2 Table 1. Gross domestic expenditure on R&D --
GERD -- by sector of performance and source of funds
MEASURE=Million constant $ 2000 prices and PPPs;
SECTFIN=Higher educ. fin. by Bus. enter.
BE-PNPERD
RDS2004-2 Table 1. Gross domestic expenditure on R&D --
GERD -- by sector of performance and source of funds
MEASURE=Million constant $ 2000 prices and PPPs;
SECTFIN=PNP fin. by Bus. enter.
PNPERD
RDS2004-2 Table 1. Gross domestic expenditure on R&D --
GERD -- by sector of performance and source of funds
MEASURE=Million constant $ 2000 prices and PPPs;
SECTFIN=TOTAL PRIVATE NON-PROFIT (PNP)
HERD by OECD Field of Science
RDS2004-2 Table 18. Higher education intramural
expenditure on R&D -- HERD -- by field of science
MEASURE=Million constant $ 2000 prices and PPPs;
BE-HERD by OECD Field of Science
RDS2004-2 Table 20. Higher education intramural
expenditure on R&D -- HERD -- by field of science and source
of funds
MEASURE=Million constant $ 2000 prices and PPPs;
HE/FIN=Business enterprise
GERD
MSTI2005-1 Indicator 3. GERD -- (Million 2000 dollars --
constant prices and PPPs
Researchers
MSTI2005-1 Indicator 7. Total researchers (FTE)
R&D personnel
MSTI2005-1 Indicator 9. Total R&D personnel (FTE)
R&D personnel per thousand labour force
MSTI2005-1 Indicator 10.a. Total R&D personnel per
thousand labour force
HERD
MSTI2005-1 Indicator 47. HERD -- (Million 2000 dollars --
constant prices and PPPs
GOVERD
MSTI2005-1 Indicator 54. GOVERD -- (Million 2000 dollars --
constant prices and PPPs
147
GDP
MSTI2005-1 Indicator A.2. Gross Domestic Product (Million
Current PPP$)
Converted to Million 2000 dollars -- constant prices and
PPPs using MSTI2005-1 Indicator B. Implicit GDP Price
Indices (2000 = 1)
Population
MSTI2005-1 Indicator E. Total Population (Thousands)
Labour force
MSTI2005-1 Indicator H. Labour Force (Thousands)
PhDs awarded
OECD Education Database: Number of graduates by field of
study, level of education, programme orientation, duration of
programme and sex
Country=[ALL]; Duration of programme=All educational
programmes - 900000; Sex=Total males + females - 90;
Level of education=Advanced research programmes - 60;
Type of programme=All educational programmes - 900000;
Year=[ALL]; Educational/labour market destination of
programme=Total - 900000; Field of study=[ALL]
PhDs awarded by OECD Field of Science
As above
As above
The most recent UK data on GERD, GOVERD and PNPERD have been taken from SourceOECD supplemented by data from ONS, HESA and OSTâs own SET
statistics, and have been adjusted for Million constant $ 2000 prices and PPPs.
148
International comparisons and data coverage
There are 25 countries (the OST comparator group) covered in this report in
addition to the UK.
The OST comparator group is spread by geography and type, and are thus of
value for comparisons with any national research base.
The combined output of the selected countries in the OST comparator group
accounts for more than 95% of the worldâs relatively highly cited papers over
the last twenty years. Highly cited papers are, in this context, those that have
been identified by Thomson Scientific
ÂŽ
as the most cited 1% by field and year
of publication. The group covers similar proportions of total world outputs.
The EU15 non-national group was introduced in the 2004 report to summarise
research activity in Europe, because of increased interest in the development of
the European Research Area. The EU15 are the countries that were members
of the EU for most of the period covered by the report; new member states
increased the EU to 25 in 2004.
The EU15 is not included in the aggregate statistics for the OST comparator
group. The EU15 bibliometric data generally reflect true aggregate figures and
do not duplicate activity that is collaborative between member states. This is
not always true, however, of the OECD data where some countriesâ data are
missing from some variables.
The OST group includes the full G8, a combination of some larger and OECD
countries from different continents with research bases both similar and
contrasting in structure to the UK, and a spread of smaller nations with active
and rapidly growing research bases with specific strengths.
Country groups
Some countries would form the normal core of any international reference set.
These are major economies with a strong and diverse research base. They
include countries with University-based research systems very similar to that of
the UK and others with systems that are based more strongly on research
institutes outside Universities. Additional performance factors related to
research system can thereby be examined.
European countries provide a fuller regional economic context. Those in the
OST group include medium to large research economies, have active and well
established research bases and interact substantially with the UK.
Social and economic change in the former Soviet Union and among EU
candidate countries suggests that monitoring research developments in this
Country group
Country name
Short code
G8
UK UK
USA USA
Canada
CAN
France
FRA
Germany
GER
Italy ITA
Japan
JAP
G8/E Europe
Russia RUS
Other W Europe
Belgium BEL
Denmark
DEN
Finland
FIN
Netherlands
NED
Spain
ESP
Sweden
SWE
Switzerland
SUI
Other E Europe
Poland POL
Other Europe
EU15 group
EU15
Other World
Australia AUS
Brazil
BRA
China
CHI
India IND
Iran IRA
Israel ISR
Singapore
SGP
South
Africa
SAF
South
Korea
SKO
Taiwan
TWN
149
area will extend information gained from the core European analysis. It should
be noted, however, that post-Soviet economic changes produce somewhat
anomalous indicators if estimated GDP changes rapidly.
A spread of leading research economies in other continents provides a broad
overview of the UKâs relative international standing. This year, the rapidly
evolving research performance of China has made it central to any international
research comparison. India is developing more slowly but is likely to become a
key focus within a few years.
Finally, smaller research economies are active in specific ânicheâ areas often
related to key technologies of economic significance. The countries of interest
in the OST comparator group are likely to change from time to time. Those
initially included continue to show rapid recent growth and a significant increase
in research impact.
Reference benchmarks
Two baselines have been created as reference benchmarks, and they are used
for each indicator and field. The first reference benchmark is the global total or
average. The second reference benchmark is the total or average for the OST
comparator group. Within the report, the specific benchmark that has been
used is specified. [The relevant one depends on the availability of data for each
indicator.]
Note that summed bibliometric data for the OST comparator group may appear
to exceed world totals, because of joint publications between countries. This is
discussed in a methodological note (below).
International data coverage
Finance and workforce data may be limited for some countries and some
subject areas, particularly in the social sciences and in the arts and humanities.
Work carried out for the Economic & Social Research Council highlighted some
deficits and some inconsistencies with regard to postgraduate training data for
some smaller countries. Data for the G8 appear generally sound.
Bibliometric data are generally available for all countries. For the social
sciences, while some larger fields appear to be reasonably well covered
internationally there are other specific disciplines in which there are clear
deficits for non-Anglophone countries. This means that comparisons between
the USA, UK and Canada may be sound but the relative position of e.g. France
and Germany would be less certain.
The research base varies in structure between countries (as noted above) and
there are also differences â possibly but not necessarily as a consequence - in
research culture and thus in activities such as publication and citation
behaviour. We comment below on some possible factors that arise from this.
150
Subject disaggregation
Three principal levels of subject disaggregation are used in this report:
System
(i.e. country level);
OECD
; and UK-
SUoA
.
The subject disaggregation used here is nested and hierarchical. âSystemâ
breaks down into five âOECDâ categories some of which are then broken down
into the ten âSUoAsâ.
Mapping data at a subject level
Research data can be grouped at a system level (total national papers, total
science and arts expenditure) or at levels of detail described as fields, subjects
or disciplines. A balance needs to be struck between a coarse level of analysis
and too fine a level, both of which can obscure information.
For analyses of output performance patterns, the UKâs SUoAs (see below) can
be used, but it is also feasible to use finer levels of discrimination.
Evidence
Ltd
has developed a number of methodologies for mapping data from different
sources to a common set of categories.
System
(Country)
System refers to the country as a whole. This gives a national overview of
research activity and performance.
System is often the only available level because data are not attributed to any
specific subject category. It is not entirely satisfactory because of the innate
cultural differences between major research fields. The relative size of different
fields may swamp important differences between fields within countries.
OECD categories
OECD coarse-level categories are broad fields used for categorising much of
the OECD database. This provides a satisfactory separation between major
parts of the research base, but still obscures some performance detail.
For this OST report we have combined the OECD data for natural and
agricultural sciences. The category for agriculture is useful for measuring the
specific economic activity in this sector but it is of much less significance as a
separate grouping for research base analyses.
The five OECD categories used here are
1. Medical Sciences
2. Natural and Agricultural Sciences
3. Engineering and Technology
4. Social Sciences
5. Humanities [including Arts where data permits]
Units of Assessment
Units of Assessment (UoAs) are the 68 subject categories established in the
UK for the cyclical Research Assessment Exercises up to 2001.
A list of these categories is available from the HEFCE website
http://www.hefce.ac.uk/Research/assessment/
These categories are generally too fine and numerous for international
comparisons, other than those focussing on a single discipline.
SUoAs (Super-UoAs)
SUoAs are grouped
Unit of Assessment
(UoA) subject categories. This usefully
separates some of the major sub-divisions within the OECD categories, such as
biological, physical and environmental sciences within the OECD Natural
Science and Agricultural Sciences category.
The groups are based on an analysis of similarity of journal usage by
researchers submitting to the UK Research Assessment Exercise in 1996 and
2001. Some of the groups are substantially larger than others and might be
identified as âmajorâ fields, but this designation refers to size only rather than
policy significance.
The ten SUoA categories used here are
Clinical (major) = OECD category 1
Pre-clinical and health = OECD 1
Biological sciences (major) = OECD 2
Environment = OECD 2
Mathematics = OECD 2
Physical sciences (major) = OECD 2
Engineering (major) = OECD 3
151
Social science (major) = OECD 4
Business = OECD 4
Humanities, languages and arts = OECD 5
Economic and social research
The application of some research indicators to the economic and social
sciences is disputed, as we have noted elsewhere in this report.
Recent work for the Economic & Social Research Council has confirmed that
bibliometrics must be used with caution in this area. The economic and social
coverage of the Thomson Scientific journal databases is not balanced in the
same way as natural science disciplines. The lower language diversity results
in a deficit in coverage for some large European research economies.
The bias towards Anglophone journals may affect the UK in two ways: it is
relatively less well covered than the USA, so the database has less utility, but
more âaverageâ material may be covered than for other European competitors,
so its net indexed impact may be reduced.
It is also noteworthy that a high proportion of the material cited by articles in
social science journals is not covered by the Thomson databases, although this
varies between disciplines.
Although the defects of existing bibliometrics are familiar to social science
researchers, many of them make extensive use of journal, article and citation
information in reaching judgements about research quality. However, they do
so in an âexpertâ fashion alongside other data and it is not possible readily to
translate their approach into systematic evaluation.
The use of journal articles as a preferred output mode for economic and social
research appears to be increasing, as judged by RAE data and survey
outcomes. Bibliometrics are likely to be of increasing importance and
bibliographic databases and indices are likely to be of increasing value to social
scientists over the next few years.
Humanities, languages and arts
Previous reports have focused on the fields of Natural and Social Science
research covered by the Research Councils then funded by the Office of
Science and Technology. They therefore excluded the broad-based Arts and
Humanities.
In 2005, a new Arts and Humanities Research Council took responsibility from
the former Arts and Humanities Research Board (AHRB) and has become part
of the UKâs national Research Council system.
New indicators appropriate to the different research paradigms in these
disciplines are likely to be required. While research funding and research
training are clearly common to all disciplines their relation to performance is not
the same in all cultures. Publication and citation behaviour also differs
markedly, perhaps more so in the Humanities than in the Social Sciences.
Background data is being gathered by relevant agencies to support the
development of new indicators and the AHRC will be exploring the options that
arise. Their staff are in regular contact with the OST on this.
In the interim, it has been decided that, where the data allow, the existing
indicators should be extended to capture information about humanities
research. Data on the language disciplines and on the visual and performing
arts is very sparse but this has been included where available.
The international databases are often much weaker on humanities and arts
research activity. Many countries make no returns in this area and others, with
significant research bases, supply data only in some years. This further
reduces the capacity for analysis.
It is acknowledged that indicators in this report, and elsewhere, have been
developed principally for use alongside the âscience research paradigmâ. Their
relationship to âresearch performanceâ in the humanities is only partly
understood. This yearâs presentation is therefore one that should stimulate the
wider debate on the assessment of research in the humanities but should not
be taken to provide any grounded or authoritative measure of the UKâs recent
standing.
152
Time frames
This report uses analyses of:
â˘
Current performance, in the latest year (or five-year period) for
which data are available.
â˘
Comparisons of recent performance with an average for the
previous five-year period.
â˘
Trends in performance over the last ten years.
The emphasis in performance analysis indicators is on the current position of
absolute and relative indicators for one or a group of countries.
Current performance can only be fully understood, however, if it is also set
against recent and longer-term trends
Some data series only make more sense in a longer time context because of
missing values or exceptional year-to-year variation.
Time windows
Five-year windows address annual activity fluctuations within subjects, smooth
out marked annual changes in inputs and outputs, help to compensate for
missing values in a data series and present a more readily understood profile of
research performance.
Thomson ScientificÂŽ data make use of overlapping five-year windows for
appropriate comparisons of e.g. citation counts across time. Because citation
counts are less on average for more recent years a direct comparison between
two years is sometimes meaningless. If the citations that accumulate over a
fixed period of years are used then this provides a sensible reference point
between publications from different years or periods.
Thomson recommends using a five-year (NSI5, National Science Indicators
over 5 years) period for papers and the citations that are attributed to them.
Thus the NSI5 for 1996 is the set of papers published in the years 1992-1996
and the citations to those papers that had accumulated by the end of 1996.
The NSI5 for 1997 will overlap with the last four years of papers and include the
next later year, with the citations that accumulate for those papers to the end of
1997.
Evidence
groups data into five-year windows using the same convention. The
average annual performance for a five-year window labelled 2001 will be the
average for the years over the period 1997-2001.
Moving five-year windows also help to overcome the problems of missing years
in OECD data.
Current performance
The last calendar year (2004) has been used for many of the indicators.
In some instances there are as yet no data for the last year, so the most recent
year for which data are available is used instead. This is usually 2003.
Where five-year windows are employed, the current performance is usually
based on data for 2004 or the 2004 âwindowâ which covers the average
performance for 2000-2004.
Recent performance
When ârecentâ performance is calculated, this is done using the latest available
data. Because some data from earlier years will be revised later, this means
that the ârecentâ value in on report may differ from the calculation for the same
value given in an earlier report.
Recent data for the UK include selectively updated figures from the ONS.
If ârecentâ data are changed then rankings may be revised as a consequence.
Thus, the UK may in one report be ranked 10
th
recently and currently, yet in the
next report be said to have improved from 12
th
to 10
th
. This will be because
either the UK or another countryâs data has changed so that the UKâs relative
position for past years has fallen.
Current performance is usually compared with the average performance in
recent years.
For this report, recent usually means the previous five years. If the current data
refer to 2004 then the recent data refer to the average for 1999-2003.
For five-year windows, the window used for the recent comparator is specified
in the particular analysis.
153
Longer-term patterns
Trends
are important where year on year variation can only properly be
interpreted in the context of the longer term. Different forms of trend analysis
may require annual data or rolling five-year windows. Each can help to
establish, first, whether the current snapshot is a good reflection of performance
and, second, whether any projection can be made of likely future performance.
Lags between inputs and outputs
The timing (or phase) relationships between different types of data are
important for SEB indicators. For example, inputs precede outputs. A specific
project grant will precede the publications that report on the project outcomes
by some years.
A three-year lag has sometimes been inferred in UK policy studies, mostly
because this fits with a long established three-year project structure where
funding is allocated in year 1 for activity that starts immediately and begins to
show substantive results in year 2 leading to articles being written in year 3 and
later. Publication may occur 12-18 months after an article is written.
The time lag between input and output may vary between indicators and
change over time and there may be other, less transparent, links to elucidate.
There is therefore no simple, universal time lag that could readily be applied to
this indicator system.
We could also consider not three- or five-year lags but the longer term. For
example, we could explore patterns at institutional level over ten-year or even
longer periods that take into account investment through capital as well as
recurrent spend.
Furthermore, there is no evidence either that all national systems have the
same time lags or that these differ. We do know that there are differences in
citation behaviour between countries (we discuss this in more detail below)
which sometimes leads to a âspikeâ in relative UK citations soon after publication
at the same time as a relative âtroughâ in Japanese citations. Later analyses
shows the Japanese tend to pick up but at a slower rate while some UK papers
may peak early.
To summarise, no time lag has been applied to the secondary indicators in the
first cycle of OST SEB indicators because we have no clear and uniform basis
on which to make general assumptions. Output data are therefore compared
with input data for the same year, although these inputs cannot have funded
these outputs. More specific analyses with different time lags may be used in a
future indicator cycle, but this will depend on exploring alternative scenarios to
throw light on this aspect of research performance.
154
Bibliometrics
Bibliometrics are important in indexing research performance. Bibliometric data
have particular characteristics of which the user should be aware, and these
are considered here.
The data come from Thomson ScientificÂŽ databases, a single source collated
to the same standard and therefore providing a level of comparability not found
in other data. The data are also valuable because they can readily be
disaggregated by field, by year and for most countries.
Journal papers (publications, sources) report research work. Papers refer to or
âciteâ earlier work relevant to the material being reported. New papers are cited
in their turn.
Papers that accumulate more citations are thought of as having greater
significance or influence in their field. Citation counts are therefore recognised
as a measure of impact, which can be used to index the excellence of the
research from a particular group, institution or country.
Most impact measures use average citation counts from groups of papers,
because some individual papers may have unusual or misleading citation
profiles. These are diluted in larger samples.
Time factors
Citations
accumulate
over time. Older papers therefore have, on average,
more citations than more recent work. The following Figure shows the pattern
of citation accumulation for a set of 30 journals in Geological Sciences. Papers
less than 8 years old are, on average, still accumulating additional citations.
Only for older sources has the citation count plateaued.
Papers are also
more likely
to be cited at all over time. The Figure shows that
the percentage of papers that have never been cited drops over about five
years. Beyond five years, some 10% or more of papers continue uncited.
Account must be taken of these time factors in comparing current research with
historical patterns. For these reasons, it is sometimes more appropriate to use
a fixed 5-year window of papers and citations to compare two periods than to
look at the longer term profile of citations and of uncitedness for a recent year
and an historical year.
Citation accumulation for papers in Geological Sciences
0
5000
10000
15000
20000
25000
1
981
1
982
1
983
1
984
1
985
1
986
1
987
1
988
1
989
1
990
1
991
1
992
1
993
1
994
1
995
1
996
1
997
1
998
1
999
2
000
2
001
2
002
Year of publication
C
ita
tion c
o
un
t t
o
end-
20
02
0
10
20
30
40
50
60
70
80
90
P
e
rc
ent
age
pa
per
s unci
ted by
end
-2002
Citations to date
Uncited papers
155
Discipline factors
Citation rates vary between disciplines and fields. On the whole, citations
accumulate more rapidly and plateau at a higher level in Biological Sciences
than Physical Sciences, and Natural Sciences generally cite at a higher rate
than Social Sciences.
There is no intention that the indicators reported here should be used for
disciplinary comparisons within countries. Account must be taken of
disciplinary factors in comparing impact indices. For example, a direct
comparison of citations per paper between Biochemistry and Economics is
inappropriate and would be misleading. The world average in a given field,
however, can provide a useful reference point for comparisons between
countries. It is more informative if the values for each country within any
discipline are normalised, or REBASED against a world average for that field.
Rebased impact factors in this report refer to a world average index of 1.0 for
that field
All comparisons made in this report are âlike-for-likeâ. Citation rates may be
less informative about performance in some fields because they may be lower
or citation behaviour (the reasons why people cite other work) may be poorly
understood. Nonetheless, so long as we use fair comparisons we should
expect that such variations do not unbalance our conclusions. For example,
UK Natural Science is compared with USA and Germany Natural Science, and
UK Social Science in 2003 is compared with UK Social Science in 1998-2002.
Only if behaviour within a discipline differs significantly between countries or if
the data for one country is unrepresentative compared to others would the
comparisons become invalid.
Bibliometric data for Social Science should always be regarded with caution.
Recent analyses confirm that, while they may be both valid and useful, there
are issues about national imbalances â especially at a disaggregate field level â
and any publication analyses must be interpreted against the background
context of other indicators and detailed commentary.
Location factors
Citations accumulate for each author on a paper and for each institution and
country included in the authorsâ addresses. The world total of citations is
consequently less than the sum of national citations.
â˘
As an example, imagine a set of four papers. One has a German author, one
has a UK author and two have both UK and German authors.
â˘
Each paper is cited twice. There are a total of eight (8) citations.
â˘
There are six UK citations: two to the UK only paper and two to each of the
jointly authored papers.
â˘
The (UK + Germany) citations = 12, because there are similarly six German
citations. This exceeds the actual total of 8.
While it is feasible to create an overall total for numbers of world papers and
citations, from which duplication can be removed, it is onerous to do this for a
changing sub-set of countries for each data analysis. De-duplication has been
done for the EU15.
Data are only available for some countries in the OST comparator group for
some analyses, (e.g. data on researchers are a sub-set). Consequently, where
the sum of papers or citations is calculated for the sub-set (e.g. to index
citations per researcher), then the total includes duplicates for joint papers.
The value of the UK activity in relation to both the OST comparator group and
the world total is given for indicators involving only publications data. In these
cases, it will be seen that the UK is apparently smaller as a proportion of the
OST comparator group than of the world, because of the duplication between
countries. Nonetheless, this has no effect on comparative values such as rank
or ratios of activity.
National factors
The volume of papers on Thomson databases for G8 countries is not
disproportionate in the Natural Sciences, although there is said to be an
Anglophone bias and some of these countries do not have English as a first
language. The imbalance in some â if not all - the Social Sciences and in the
Humanities is greater.
There is some selectivity in publication behaviour in some countries. For
example, a study of Spanish Earth scientists (J Rey-Rocha, Scientometrics
(2002) 55, 377) showed that they publish parochial reports in Spanish journals
not indexed by Thomson. The effect of this on Spanish citation indices is not
clear but it may mean that only higher impact work is indexed. If a similar
pattern is true for other countries, there would then be a consistent sampling
bias in favour of more citable publications for non-Anglophone countries (i.e.
lower volume but higher average quality).
Citation behaviour also differs between countries. UK researchers tend to
access new work and cite it more rapidly than researchers do elsewhere. This
means that some high UK relative citation rates may dip later. This does not
distort overall perceptions of relative national performance but it is important to
be aware that this is a background component.
156
Glossary
AHRC
The Arts and Humanities Research Council funds research and
postgraduate study within the UK's
HEIs
in traditional humanities subjects,
such as history, modern languages and English literature, and in the
creative and performing arts. It also provides funding for museums,
galleries and collections that are based in, or attached to, HEIs in England.
The AHRC was established on 1 April 2005, and replaced the Arts and
Humanities Research Board.
BBSRC
The Biotechnology and Biological Sciences Research Council receives
money from
OST
through the Science Budget to fund academic research
and training in biosciences. It was formed in 1994 by the merger of the
former AFRC with the biotechnology division of the former SERC.
BERD
is the total R&D performed in the business sector. Where the prefix BE- is
used for other R&D spend (e.g. BE-HERD), this denotes that portion of the
R&D within the sector concerned that is financed by the business enterprise
sector.
Bibliographics
is used as a term for descriptive data referring to publication
activity or submissions that do not provide a direct measure of performance.
Bibliometrics
are measures of research activity and performance derived from
databases of journal articles and of citations of those articles. There are
associated secondary measures based on relative journal and article
citation rates.
BSTS
refers to OECDâs Basic Science and Technology Statistics. These are
disaggregated further than
MSTI
but cover fewer countries. In 2004, BSTS
was succeeded by Research and Development Statistics (
RDS
).
Chief Scientific Adviser
is the head of
OST
and provides advice to the
Government on science, engineering and technology matters.
Citations
are the formal references made in a journal paper or other publication to
earlier work. These citations (or cites) usually indicate that the earlier work
supports the publications methods, data or claims in some way. Negative
citations may also occur.
DG-RC
is the Director General of
Research Councils
, a senior member of the
OST
who advises on the allocation of the UK
Science Budget
.
Efficiency
in the context of
Evidence
Ltd reports is the relationship between the
volume of outputs from the system and a stated volume of inputs.
Effectiveness
in the context of
Evidence
Ltd reports is the relationship between
the volume of outputs and their average quality.
EPSRC
The Engineering and Physical Sciences Research Council is the UK's
main agency for funding research and related postgraduate training in
engineering and the physical sciences. It emerged from the former SERC
in 1994.
ESRC
The Economic and Social Research Council is the UKâs leading research
funding and training agency addressing economic and social concerns. Its
predecessor until 1983 was the Social Science Research Council,
established in 1965.
Eurostat
is the Statistical Office of the European Communities situated in
Luxembourg. It had a budget of âŹ140 million in 2000. Established as a
directorate of the European Community in 1959, its modern task is to
provide the European Union with a high-quality statistical information
service at European level that enables comparisons between countries and
regions.
Expected citation rate
â see Journal Average Impact factor
Frascati
Manual
was first published as the outcome of an OECD meeting in June
1963 with national experts on R&D statistics at the Villa Falcioneri in
Frascati, Italy. The result was the first official version of the
Proposed
Standard Practice for Surveys of Research and Development
, now
commonly known as the Frascati Manual. The Working Party of National
Experts on Science and Technology Indicators (NESTI) has now developed
a âFrascati Familyâ of methodological manuals, including publications on
innovation (
Oslo Manual
), human resources (
Canberra Manual
) and the
technological balance of payments and patents.
FTEs
Full Time Equivalents. Many research and other posts are filled on a
fractional basis and there are also a significant number of part-time
research students. The balance of full and part-time posts and students
varies between institutions and a direct head-count may therefore be a poor
indication of the actual volume of activity. To account for this, head-count
numbers may be converted to full-time equivalents (e.g. two 0.5 FTE posts
equate to 1.0 FTE). In other cases the actual head count may be more
relevant.
157
G8
refers to a group of eight leading economies. This comprises the UK, USA,
Canada, France, Germany, Italy Japan and Russia. The G7 is an earlier
version of the same group, without Russia.
GERD
is Gross Expenditure on R&D
GOVERD
is total R&D performed in the government sector.
HE
is higher education in the broad sense.
HEIs
are higher education institutions. In the UK specifically they are the
universities and colleges funded for teaching and research by the regional
HEFCS.
HERD
is total R&D performed in the higher education sector (which is very broadly
defined by OECD and may in some countries cover much more than
universities and colleges). That part of HERD funded by the business
enterprise sector may be denoted as BE-HERD.
HESA
The Higher Education Statistics Agency was established in 1993 and is the
central source for HE statistics. It seeks to standardise data collection
processes and formats.
Impact
is the average citation rate of the outputs for a specified source (country,
organisation, author). This is a simple and direct measure of research
performance since citations usually reflect acknowledgement by later
authors of the value of a published item. The impact figure can be taken as
a local measure of the 'worth' of publications. In this report, impact figures
are
rebased
to take account of the world average figure in the field. In this
way, comparisons can be made between fields that have different raw
impact values to judge their effectiveness.
ISI
is the older name for the Philadelphia based division of
Thomson Scientific
.
The former Institute for Scientific Information was founded by Eugene
Garfield in 1958. It is the world's premier source of information on journal
outputs and their citations. ISI provides a range of commercial information
products designed to support research and research management,
including 'Current Contents' and the Science and Social Science Citation
Indexes. ISI indexes over 8,000 journals in 35 languages, which is agreed
to represent most or all of the material likely to be recognised as having
significant value to others for most science fields. ISI data may under-
represent new and emerging fields and so disbenefit interdisciplinarity and
is less rich in coverage of the Social Sciences. It covers the Arts and
Humanities less well.
Journals
Research findings are published in journals, conference proceedings
and books. Journals are the main mode of rapid output for most scientific
fields. The first research journal was reputedly the
Journal des Scavans
,
inaugurated in 1665. It was published by Denys de Sallo in Paris. By 2000
there were estimated to be about 20,000 journals carrying over one million
research papers per year.
Journal Average Impact Factor (JAIF)
can be calculated as the average number
of citations received by the papers in a stated journal in a particular year.
JIF varies between journals: those such as âNatureâ and âScienceâ tend to
publish papers that receive many citations and they have a high JIF.
Publication in a journal with high impact is often seen as a mark of prestige.
JIF for any one journal varies between years, because more recent years
have obviously had less time to accumulate citations.
JIF
is also calculated through a more complex algorithm by ISI. The ISI
ÂŽ
impact
factor system is a commercial product available through Thomson
Scientific.
MRC
The Medical Research Council was founded in 1913 (initially as a Medical
Research Committee, under the provisions of the National Health Insurance
Act, 1911). It promotes research into all areas of medical and related
science with the aims of improving the health and quality of life of the UK
public. It funds research both in Universities and through its own institutes
and units.
MSTI
refers to OECDâs Main Science and Technology Indicators. These are at a
summary level compared to
RDS (BSTS)
but cover more countries.
NERC
The Natural Environment Research Council was established by the
Science and Technology Act (1965)
with responsibilities transferred from
the Nature Conservancy and the National Oceanographic Council. It now
promotes and support research, survey, long-term environmental
monitoring and related postgraduate training in terrestrial, marine and
freshwater biology and Earth, atmospheric, hydrological, oceanographic
and polar sciences and Earth observation. It funds research in Universities
and in its own institutes.
NSI
refers to Thomson ISI ÂŽâs National Science Indicator product. The NSI5 is the
standard five year grouping of bibliometric data used in the NSI to provide
constant time windows for trend analysis, because citations accumulate
over time and comparisons between years would otherwise be problematic.
158
OECD
is the Organisation for Economic Cooperation and Development and is a
major source of data for international R&D statistical R&D statistical
analyses. It evolved in 1961 from the former Organisation for European
Economic Co-operation which was formed to administer American and
Canadian aid after World War II. It now has 30 member and 70 associate
countries. Its members account for about two-thirds of global goods and
services.
Office of Science and Technology
â see OST
ONS
is the Office for National Statistics. It was created in April 1996 when the
Central Statistical Office merged with the Office for Population, Censuses
and Surveys. It is the government department that provides statistical and
registration services. The Director of ONS is the National Statistician who
is also Registrar General for England & Wales. ONS is responsible for
producing economic and social statistics used by Government to create
evidence-based policies and monitor performance against them. The Office
builds and maintains data sources both for itself and for customers.
OST
is the UK Governmentâs Office of Science and Technology. It was created in
1992 by the amalgamation of the Cabinet Officeâs Science and Technology
secretariat and the Science Branch of the former Department of Education
and Science. The head of OST is the
Chief Scientific Adviser
. OST is
also the home of the
Director General of Research Councils
Output
is specifically the numbers of journal articles recorded on the databases of
ISI but is used generically to refer to other outputs from research, including
patents and highly trained people.
Output volume
in research journals world wide was estimated in 2000 to be about
one million research papers per year in some 20,000 titles.
PDRAs
are Post-Doctoral Research Assistants, the non-permanent research
workers in the transition between PhD training and full independence. They
are usually employed on shortâterm, e.g. 3 year, research grants and
contacts.
Performance
in regard to research is frequently indexed as the impact of outputs.
In
Evidence
reports there are a wider range of performance indicators, and
the ratio between research input and output as well as impact can be an
important measure.
Period
is used for various time windows:
â˘
The period for which ISI data on outputs and impact are available, 1981-
2000
â˘
The period to present from the first Research Selectivity Exercise in 1986.
â˘
The period between RAEs, e.g. a recent 1996 and current 2001 RAEs.
PGRs
are Post-Graduate Research students.
Along with journal articles they are
one of the key outputs from the research base.
PNPERD
is the total R&D performed in the private non-profit sector
PPARC
The Particle Physics and Astronomy Research Council funds research
and training in particle physics, astronomy, solar system science and
particle astrophysics. It supports international scientific facilities in
Edinburgh, La Palma and Hawaii.
PPP
Purchasing Power Parity states that exchange rates between currencies are
in equilibrium when their purchasing power is the same in each of the two
countries. This means that the exchange rate between two countries
should equal the ratio of the two countries' price level of a fixed basket of
goods and services. The simplest way to calculate PPP between two
countries is to compare the price of a "standard" good that is identical
across countries. Sophisticated versions of PPP look at a large number of
goods and services. One of the key problems is that people in different
countries consume very different sets of goods and services, making it
difficult to compare purchasing power.
PSA
refers to the Public Service Agreement system. This was introduced in 1998
with the intention of setting out publicly clear objectives and targets showing
what Government departments aimed to achieve in terms of public service
improvements.
PUBERD
is the sum of GOVERD and HERD, equating to R&D performed in the
publicly funded sectors
Purchasing Power Parity
, see PPP.
R&D
is defined by the
OECD
as Research and Development.
R&D
personnel is defined by
OECD
/
Frascati
as all persons directly employed on
R&D, as well as those providing direct services such as R&D managers,
administrators and clerical staff.
Ranking
refers to the position an institution holds relative to others in the same
field. The data may be ranked according to output volume (numbers of
159
papers produced in a given period) or impact (average of citations per
paper in some given basket of publications).
RBI ReBased (or relative) Impact
compares performance to a world average for
that discipline and year. At a fine level this relative impact can be assessed
for specific journals. Science papers tend to attract more citations than
social sciences, and there are variations within science. Older papers
naturally have more citations than new papers. Unless these factors are
taken into account it is not reasonable to compare citation rates. Reference
to the appropriate world average allows this comparison.
RDS
refers to the OECDâs Research and Development Statistics which replaced
the former
BSTS
in 2004. They are disaggregated further than
MSTI
but
cover fewer countries.
Relative citation rate
â see rebased impact.
Research Assessment Exercise
The RAE is the cyclical process of assessing
UK University research. RAE grades are used as weighting factors to
determine the allocation of research resources. RAEs have taken place in
1986, 1989, 1992, 1996 and 2001. The next is in 2008.
Research Councils
are independent Non-Departmental Public Bodies (NDPB),
established by Royal Charter and accountable to Parliament through the
DTIâs Office of Science and Technology (
OST
).
Researchers
is an
OECD
/Frascati definition used to denote professionals
engaged in the conception or creation of new knowledge, products,
processes, methods and systems and also in the management of the
projects concerned
Royal Society, The
, is the UKâs national academy of science. It was founded in
1660, is independent of UK government (although receiving a grant-in-aid
through the
Science Budget
) and has some 1300 Fellows and Foreign
Members. It is the worldâs oldest scientific academy in continuous
existence.
Science Budget
is the money allocated to the
Research Councils
, which funds
their own institutes and HEIs, usually in the form of peer-reviewed grants for
specific research projects as part of the dual support system and through
research studentships. There are also directed programmes, initiatives and
centres
Science Citation Index
is a main ISI database of scientific journal publications
and their citations and can be searched electronically.
SEB
is the national Science and Engineering Base (the acronym also refers to the
Society for Experimental Biology).
SET
refers to Science, Engineering and Technology.
Share
is the fraction or percentage of e.g., outputs published by the peak
compared to the UK total. It is also used for other research activity
measures.
Sources
are the publications (papers, articles) in journals tracked by the ISI
database. In this report, sources are presented at the UoA level where
possible. For example, when reporting on UoA 3 (Hospital-Based Clinical)
all relevant sources for the institution are reported.
SUoAs
(Super-UoAs) are disciplinary groupings of cognate UoAs with similar
publication profiles.
Super-UoAs
see SUoAs.
Thomson Scientific
is the current name of the former
ISI
.
Thomson Corporation
is a leading international business with annual revenues of
approximately US$6billion and employing about 35,000 people worldwide.
The Corporation's common shares are listed on the Toronto (TSE: TOC)
and London stock exchanges.
Evidence
has a strategic alliance with
Thomson ScientificÂŽ.
UoAs
are Units of Assessment, the disciplinary units used as subject categories
for research assessment. In 1992 there were 72 UoAs, but in the 1996 and
2001 RAEs a system of 69 UoAs was used, not all of which were active on
both occasions.
UK average impact
is the average number of citations per paper attributable to an
UK address for publications in that field. For UK HEIs the average impact is
the average of the total HEI dataset and not the average of the individual
HEIs.
Workforce
(labour force) is an OECD term used to denote the total number of
persons available for work, whether in employment or not