An Innovation Index Based on
Knowledge Capital Investment:
Definition and Results for the UK
Market Sector
Tony Clayton, Mariela Dal Borgo, Jonathan
Haskel
j.haskel@ic.ac.uk
www.ceriba.org.uk
Sheffield, 2008
Popular discussion: the
transformation of the economy
• The new/knowledge economy
– Technology changes
• Software, hardware, ICT
– Trade changes
• Move to knowledge-intensive activities
• Rise of the service sector
• Innovation as a key economic driver
– DIUS “Innovation Nation” White Paper:
commission an “innovation index”
Where do we see Innov/New Economy
in the macro statistics?
• The New Economy
– GDP growth: stable
– Labour prod growth/TFP growth: stable/slowdown
– Investment shares: stable
• Innovation
– Most think of R&D (input) or TFP (output)
– Some innovation-type spending measured and
capitalised e.g. software
– Most treated as an intermediate
• R&D
• Apple
– Relegated to TFP
– Limits what Economists can say about innovation
LPG and TFPG in the UK
4
TFP
Capital deepening
Human capital deepening
3
0.81
0.43
0.28
2
1.35
1.55
2.00
1
1.00
0.65
0.48
0
1990-1995
1995-2000
Note: LPG=Cap deep + human cap deep + TFPG
2000-2004
Our main claim
• Innovation and the New Economy
– needs better measures of innovative investment
– Incorporated into National Accounts
• Some new economy is about investment in tangible
assets
– Computer hardware
– Telecoms equipment
– Tangible investment: included in GDP
• But some about
– investment in intangible “knowledge” assets
– Broader than R&D
• Computerised information: software, databases
• Creative property: scientific R&D, non-sci R&D (e.g. design)
• Firm competencies: training, branding, organisational capital
Research programme
• Settle on list of intangible assets
• Measure investment in them
• Incorporate into National Accounts so
consistent with GDP and explain growth
• Hopefully gives
– Better GDP measure
– Better account of drivers of GDP
– Better account of innovation
• Current method
The iPod
– Inputs: Count scientific R&D only.
– Output/value added:
• No effect by assumption: all R&D used up in a year, no
enduring asset created, so only an intermediate
• Intangibles approach
– Inputs:
• Upstream: more than just R&D
– Upstream spending also on design, software
• Downstream: need associated coinvestment
– Marketing, organisational change
– Output/value added
• Rises: spending is an enduring knowledge asset so its
investment
• Shows up as innovation spending on intangible
assets
How does the iPod show up in (US) National
Accounts?
• iPod: Designed in California, made in China
• Apple
– nominal value added =
• Sales – value of imports – (design+marketing+R&D)
– (Industry classification may be wholesaler)
– Real value added: depends on deflators
– Innovation effect of iPod: TFP growth if real value
added rises
• Alternative model
– Apple builds knowledge capital asset by investing in
R&D but also design, marketing
– Real GDP rises with more investment
– Innnovation measured by more knowledge
investment and (maybe) increased TFPG
Growth and innovation accounting,
outline
•
•
•
•
1. Data on invest in intang assets
2. Deflate and build real intang asset stocks
3. Recalculate GDP and factor shares to include intang
4. Build Hall/Jorgenson capital services to add assets
together
• 5. Build labour quality adjustment
• 6. Calculate LPG, capital and labour deepening, TFPG
• 7. Innovation
– accounting is spending on intang assets
– Index is share-weighted knowledge capital growth plus TFPG
Questions we can answer
1.
2.
3.
4.
5.
6.
Is investment increasingly in know assets?
Is GDP or GDP growth understated?
What has happened to investment?
What has happened to the capital share?
What has happened to Y/L growth?
What are the contributions of labour, tangible
capital, knowledge/intangible capital and TFP to
growth?
7. What does innovation look like
1. What intangible assets are being spend on?
2. What is the contribution to growth?
Other work
• Corrado, Hulten and Sichel (CHS) for the US
• Findings interesting:
– Increasing importance of investment in intangible
assets
– Y and I/Y understated
– Capital share has been rising
– Labour prod growth understated (although mid 90s
acceleration similar)
– TFP growth has accelerated, but less TFP growth
than before
• Giorgio Morrano, Haskel, Wallis for UK
• Van Ark, Manole, Hao forEU countries
Results
• GHW paper, May 07
– Followed CHS method
– Growth a/c for whole market sector
• CDH paper, Sept 08
– 6 market sector industries
– Improved design data using OECD software
methodology
Summary of results
• Innovation accounting
– 1997: Tang invest = £81bn, knwlg invest = £76bn
– 2005: Tang invest = £96bn, knwlg invest = £117bn
• Of which R&D = 7%, software = 14%, training = 25%
• Innovation index
– 2000-05 mkt sector lab prod. growth= 2.74 pppa,
• Contrib of knowledge capital= 1.24 pppa.
• Contrib of mfring knowledge capital = 0.75pppa
Algebra of including intangibles
Excluding intang:
Expenditure : V  C  I TAN
Income : pV V  wL  pKTAN K TAN
Growth a / c :  ln V  sL  ln L  sK TAN  ln K TAN   ln TFP
Innovation : II   ln TFP
Including intang:
Expenditure : V  INTAN  C  I TAN  I INTAN
Income :
pV V  INTAN  wL  pKTAN K TAN  pKINTAN K INTAN
Growth a / c :  ln V  INTAN  sL  ln L  sK TAN  ln K TAN  sK INTAN  ln K INTAN   ln TFP  INTAN
Innovation : II  sK INTAN  ln K INTAN   ln TFP  INTAN
Details of g a/c method
1. Collect nominal investment in intangibles time
series
•
•
•
What investment measures?
Availability of data over time e.g. training
Measures for market sector only: some interp needed
2. Deflate to get real investment series
•
•
Choice of deflator. Use mkt sector deflator
Software and hardware:
•
•
Soft: purchased and own account (own a/c pro’y adjust)
Hardware
Method contd
3. Build real capital stock using perpetual
invent method
•
•
•
•
Starting point (1970=0)
Depreciation rates: use CHS rates. Experiment.
Adjustment for annual data (assumes investment
comes on stream in mid-point of year)
Assets:
– Tang: buildings, plant, vehicles, computers
– Intang: see below, main are software, R&D, advertising,
training
Method contd
4. Re-calculate market sector GVA to include
intangibles
•
Adjusted Nominal GVA
–
•
Unadjusted nominal m.sector GVA + nominal investment
Adjusted Real GVA growth
–
–
Weighted growth in real market sector GVA + weighted growth
in real investment
Weights are shares of nominal market sector GVA and
nominal investment in unadjusted market sector GVA
5. Adjust operating surplus of m.sector GVA
–
–
Adj op surpl =adjusted GVA – m.sector labour compensation
Mixed income: allocated prorata (note BofEng, to wages)
Method contd
6. Build Hall/Jorgenson VICS measures
•
Rental rates and rate of return assume:
– equalised rates of return across all assets
– rental rates times stock =
– Tax adjustment
•
•
Generate VICS for each asset as share of asset
in payments in  times real asset growth rate
Generate economy-wide VICS summing VICS
for each asset
Method,contd.
7. Build quality-adjusted labour index
•
BofE index adjusts hours for
– Education
– Gender
– Age
8. Do growth accounting
Types of intangible, details
A. Computerized information
• Computer software (bought in, own account)
• Computer databases
B. Scientific and creative property
• Science and Eng R&D spending, usually leading to a patent/licence
• Mineral exploration (mostly R&D in oil and minerals)
• Artistic originals (mostly R&D in creating artistic originals)
• Other product development, design, research, usually not leading to a
patent/licence (I.e. non-scientific R&D spend)
– product devel costs in fin svcs
– architect and eng design
– R&D in soc sci and humanities
C. Economic competencies
• Brand equity (to develop reputation capital via branding or trademarks)
• Firm-specific human capital
• Organizational structure (organisational capital)
A. Computerized information
Method and data sources
Computerized information
Computer software
Purchased: via surveys. Own account: labour force data on
software professionals. Source: National Accounts
Computerized databases
Included
B. Scientific and creative property
Innovative property
Scientific R&D
Current expenditure on R&D from BERD. R&D in computer
industry subtracted
Mineral exploration
Mainly R&D in mining and spending on mineral exploration.
National Accounts
Copyright and license costs
Spending on creation of originals e.g. TV, movies, books.
National Accounts
Other product development, design and
research
20% of all intermediate purchase by Financial Services
New product development costs in the
industry, ONS data. Intermediate purchases reduced by
financial industry
purchases of adv, software, consulting and design.
New architectural and engineering
designs
Sales of architecture and design industry SIC 742, ABI data.
Own account using labour costs data.
R&D in social science and humanities
No broad statistical information. Estimated as twice industry
revenues of social science and humanities R&D industry
C. Economic competencies
Brand equity
Advertising expenditure
Total spending on advertising as reported by Advertising
Association, less expenditure on classified ads
Market research
Twice revenues of the market and consumer research industry
as reported in ABI.
Firm-specific human capital
NESS05, survey of employer provided training. Includes: A)
Direct firm expenses (in-house trainers, outside trainers, tuition
reimbursement, and outside training funds) B) Wage and
salary costs of employee time in training.
Organizational structure
Purchased
Data on revenues of managment consulting industry from
Management Consulting Assocation.
Own account
No broad statistical information. Estimated as 20% of value of
executive time using ASHE data on wages in executive
occupations, excluding software occupations.
Industries
TABLE 1
Assignment of SIC codes and NACE1 sections to our 6 industries
Proposed sector categories
SIC code
1
Agriculture, Fishing and Mining
1 - 14
2
3
4
Manufacturing
Electricity, Gas and Water Supply
Construction
15 - 37
40 - 41
45
5
Wholesale and Retail Trade, Hotels and
Restaurants,
Transport
and
Communications
50 t- 64
Financial
Services
65 - 74
6
Intermediation
and
Business
A
B
C
D
E
F
G
H
I
J
K
NACE1 sections
Agriculture, hunting and forestry
Fishing
Mining and quarrying
Total manufacturing
Electricity, gas and water supply
Construction
Wholesale and retail trade
Hotels and restaurants
Transport
and
storage
and
communication
Financial intermediation
Real estate, renting and business
activities
Tan and intang invest by ind
T ot al t angible and int angible invest ment by indust ry, 1997-2005 (£ bn).
Agricult ure, F isihing &
Mining
Manufact uring
Elect ricit y, Gas &
Wat er
Const ruct ion
T rade, Hot els & Rest .,
T ransport & Comm.
F inancial & Business
Services
T ot al Market Sect or
T angibles
Int angibles
T angibles
Int angibles
T angibles
Int angibles
T angibles
Int angibles
T angibles
Int angibles
T angibles
Int angibles
T angibles
1997
7.2
2.5
19.8
26.1
5.3
1.1
1.9
3.2
30.8
17.6
16.1
25.1
81.0
Int angibles
75.6
2005
6.5
1.8
13.0
31.4
5.4
1.5
2.5
6.4
37.0
31.2
31.9
45.0
96.3
117.3
Intan investment by type (% total)
Software
14.3
Scientific R&D
7.9
9.6
0.4
Mineral exploration
Copyright licenses
16.0
0.3
0.2
0.2
Financial services innovation
5.2
6.4
6.3
Purchased architectural & engineering design
6.1
2000
11.1
Own-account architectural & engineering design
2004
12.0
0.4
R&D in social sciences and humanities
0.3
Advertising
9.1
10.4
1.8
Market research
2.6
22.4
Firm-specific human capital
15.0
Organizational structure
0.0
5.0
10.0
15.0
24.3
17.8
20.0
25.0
30.0
%
Intan investment by ind (% indVA)
20
18
16
14
%
12
10
8
6
4
2
0
1997
1998
1999
2000
2001
2002
2003
2004
2005
Agricult ure, Fishing & Mining
Manufact uring
Elect ricit y, Gas & Wat er
Const ruct ion
Trade, Hot els & Rest ., Transp. & Comm.
Financial & Business Services
Mkt sector intang invest, by asset
type
30
Soft ware
R&D
25
Mining
Copyright
20
£ bn
Finance
Design (purch.)
15
Design (ownacc.)
Non-scient ific
10
R&D
Advert ising
5
Market research
Training
0
Org. St ruct ure
1997
1998
1999
2000
2001
2002
2003
2004
2005
Decomposition of ALP
Excluding Soft ware
Including All
Int angibles
Growth Rates
2.81
2.80
Average labour product ivit y
2.74
2.74
Hours
0.08
0.07
Aggregat e Value-Added
Contributions
Average labour product ivit y
2.74
2.74
Capital Deepening
1.15
2.23
ICT T angible Capital Deepening
0.52
0.47
Non-ICT T angible Capital Deepening
0.63
0.58
-
1.19
Labour Qualit y
0.52
0.45
Aggregat e TF P
1.07
0.05
Int angible Capit al Deepening
Innovation index
All industries
Cont ribut ion of
Int angible
Capit al
Deepening
1.19
Cont ribut ion of
Aggregat e T F P
0.05
T ot al
1.24
1
2
3
4
5
6
Agricult ure, F ishing and Mining
Manufact uring
Elect ricit y, Gas and Wat er Su pply
Const ruct ion
T rade, Hot els & Rest ., T ransport & Comm.
F inancial Int ermediat ion and Business Services
0.01
0.54
0.02
0.08
0.29
0.25
-0.19
0.18
-0.02
-0.02
0.18
-0.08
-0.18
0.72
0.00
0.06
0.47
0.17
1
% o f co lum n t o t a ls
A griculture, Fishing and Mining
1%
-15%
2
Manufacturing
45%
58%
3
Electricity, Gas and W ater Supply
2%
0%
4
Construction
7%
5%
5
T rade, Hotels & R est., T ransport & Comm.
24%
38%
6
Financial Intermediation and Business Services
21%
14%
6
T otal
100%
100%
Future work: extended survey
Intangible
investment
type
Computerised
information
Innovative
property
Includes the following
intangibles
(1) Computer software
(2) Computer databases
(1) Scientific R&D
(2) Mineral exploration
(3) Copyright and license costs
(4) New product development costs in
the financial industry
(5) New architectural and engineering
designs
(6) R&D in social science and
humanities
Economic
(1) Brand Equity
competencies (2) Firm-specific human capital
(3) Organisational structure
Status
√
√
√
√
√
X
Part
√
Part
Part
X
Initial feedback from summer pilot
• ‘Extended R&D’ survey
– Firms understand technical/non-technical
innovation input
– Important who we talk to, need more than one
contact
– ‘Boundary of firm’ issues – with MNEs
– Some industry differences in approach
– Hard areas
• Own account organisational capital
• Training – including ‘opportunity cost’
Extra slides
GHW Results for 2004
Innov spend
category
Intangible
goods
Software
Scientific and
non-sci R&D
Patents,
copyrights,
mineral
exploration
Design
Intangible
competencies
Product
reputation
Human
competencies
Organisational
competencies
Total
Measurement
Amount
2004, £bn
% tot
intan
invest
ONS surveys
Sci R&D:ONS surveys. Non-sci
R&D, guesstimate
ONS
22
18
18%
15%
3
2%
Turnover of design ind
18
15%
Adv and mkt research spending
19
8%
Employer-training
29
24%
Purchased: management consultancy.
Own-account: 20% of value of
management time
22
18%
130.7
100%
Intangible investment by asset type, % of MGVA
15
Brand Equity
12
Firm-specific resources
UK
9
Scientific R&D
6
Nonscientific R&D
Computerised
information
3
0
1945
1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
US
Invest share of MGVA
28
26
24
Including intangibles
22
20
18
Including software
UK
16
14
Excluding software
12
10
8
1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
US
LPG and TFPG
4
TFP
Capital deepening
Human capital deepening
0.38
3
0.81
0.73
0.43
0.28
0.25
2
2.27
1.35
1.55
2.00
1.71
1.90
1
1.00
0.65
0.48
0.46
0.57
0.65
1990-1995
1995-2000
2000-2004
0
1990-1995
1995-2000
2000-2004
Existing National Accounts
Including all intangibles
Spares
Output concepts in the National
Accounts
Elect
Cars
Output
Value added
Firm A
(Electricity
provider) £
100
100
100
Firm B (Car producer) £
100
200
200
100
Output with different treatment of
intangibles
Advertising
Cars
Output
Value added
Firm A
(Advertising
services
provider) £
Firm B (Car producer)
with advertising treated
as intermediate
consumption £
Firm B (Car producer)
with part of advertising
treated as investment £
100
100
200
200
100
50
200
200
150
100
Treatment of intangibles in national
accounts
Category
Type of knowledge capital
Computerized information
•Computer software
•Computer databases
Scientific and creative property
•R&D
•Mineral exploration
•Copyright and license costs
•Other product development
Economic competencies
•Brand equity
•Firm-specific human capital
•Organizational structure
Knowledge embedded in computer
programs and computerized
database
Treatment in accounts
Major component, computer
software, mostly expensed
Knowledge acquired through
scientific R&D and non-scientific
inventive and creative activities
Most scientfic and non-scientific R&D
is expensed
Knowledge embedded in firmNo items recognised as assets of the
specific human and structural
resources, including brand names
firm
Should intangibles/knowledge
assets be included in investment?
• Definition of investment:
– Spending today to create an asset that
generates income in the future
• Examples:
– R&D: creates scientific knowledge stock,
future returns
– Software: creates computerised knowledge,
future returns
– Training: creates knowledge in workforce,
future returns
Practical problems
• Definitions: do intangibles generate an asset for the
firm?
– Brands, yes
– Workforce training, belongs to workers
• Measurement of intangible expenditure
– Own account/purchased e.g. software
– Valuation of intangibles e.g. brand names, organisational capita
• How much expenditure is investment?
– Advertising: long term brand building versus short term publicity
• Redefining national accounts
– Intangible investment is income to firms
– So labour share of output changes
Our key argument to measure innovation
• Imagine an economy with no innovation
• How could we get more output?
– Lay on another plane and another crew
– Deepening physical capital and labour = duplication
• How would we get more output from innovation?
– Faster boarding, turnaround: better software
– Deepening knowledge capital = innovation
• Summary: innovation is
– extra output over and above that from use of additional
physical capital and labour
– Or, the extra output from use of new knowledge capital
– Implications for measurement
• Measure growth in output (GDP), and in inputs: physical
capital, knowledge capital, labour
• Innovation accounts are data on knowledge investment
• Innovation index is contribution of knowledge investment to
output growth
What our measure is and is not
• Output
– output of more goods and services.
– Not number of inventions or ideas
– GDP as currently measured does not fully capture this
• Inputs
– knowledge capital
• Needs investment e.g. R&D
• Comes for free e.g. imitating
– knowledge capital investment is more than just R&D
• R&D, software, design, firm-funded training, organisational capital
• Investment is not just purchases from the “creative inds” since in
house too. So covers hidden innovation
• Inputs to outputs
– Needs assumptions on life lengths of knowledge stock, deflators
• Relation to other innovation definitions, see paper
Intangibles and innovation
• Axioms
– Innovation is the implementation of
something new
– It needs investment (unless pure
spillover)
• Consistent framework: accounting for
intangibles
– Activity
• Innovation needs investment in knowledge
• Knowledge investment is more than just
R&D, the creative industries, an
intermediate input
• Knowledge investment produces
knowledge or “intangible” assets
– Impact
• Investment raises value added
• Service flow from kn assets contributes to
growth