[ 1 ]
Knowledge accounting: an economic approach
Francisco Pérez and Eva BenagesIvie and Universitat de València
2012 WORLD KLEMS ConferenceAugust 8-10, 2012Harvard University
[ 2 ]
MotivationHow can we measure the knowledge intensity of economies?Despite this term being frequently used, there is no metric that measures it accurately…
Industry classification according to technology intensity, on the basis of R&D expenditure and the use of high-skilled labor.
Limitations: - It focuses on the current creation of knowledge rather than how the productive system uses it.- Normally based on a single factor: R&D expenditure in the case of manufacturing and human capital in services.- It doesn’t take into account intra-industry heterogeneity among countries.
Synthetic indices that group multiple variables or indicators.Limitations: -They are usually partial and have ambiguous meaning, because
aggregation criteria are not economically clearly founded.
Knowledge Satellite AccountsLimitations: - Complexity and data requirements.
[ 3 ]
Motivation
How can we measure the knowledge intensity of economies?
Our proposal:
• Knowledge accounting seems viable on account of the theoretical and empirical advances achieved in the measurement of physical and human capital.
• Focus: to measure the weight of knowledge in GVA, by calculating the market value of a set of inputs which are knowledge-based and incorporate knowledge into the production processes.
• Application: the methodology can be applied to economies whose national accounts systems offer industry data on various types of labor and capital services and their corresponding compensation (EU KLEMS database).
[ 4 ]
Methodology• Lij is the amount of labor of type i used in sector j
• Khj is the amount of capital of type h used in sector j
• PijL is the unitary wage paid for labor of type i in sector j
• PhjK is the user cost of type h capital in sector j
• Vj is the quantity of sector j value added
• PjV is the price of sector j value added
The value added (GVA) of sector j is distributed among the different inputs included in the production process so that,
We decompose the value of labor and capital services into two parts, one that incorporates knowledge and another one that does not:
assuming that there are f types of low-skilled labor and g assets that don’t incorporate knowledge significantly.
1 1
m nV L K
j j ij ij hj hji h
V P L P K P
1 1 1
fm mL L L
ij ij ij ij ij iji i i f
L P L P L P
1 1 1
gn nK K K
hj hj hj hj hj hjh h h g
K P K P K P
[ 5 ]
MethodologyAssumption of non-separability: the knowledge of a qualified worker is a contribution to the production process non-separable from the contribution of unqualified labor by the same worker (similarly, in knowledge intensive capital the contribution of non knowledge components of this capital, like the iron of the machine, is not separable from the contribution of knowledge-intensive capital).
Value of knowledge intensive labor (KIL) Value of knowledge intensive capital (KIK)
Knowledge-intensive value added (KIV):
Knowledge intensity (ç) of an activity j:
Knowledge intensity (Ç) of the economy depends on the weight of the various industries in the aggregate, on the basis of the value-added generated:
1
nns Kj hj hj
h g
KIK K P
1
mns Lj ij ij
i f
KIL L P
ns ns nsj j jK IV K IL K IK
/ ( )ns ns Vj j j jç KIV V P
1
1
( )Vq
j jnsqj
Vjj j
j
V PÇ ç
V P
[ 6 ]
Methodology
Knowledge-intensive value-
added (KIV)
Human capital (KIL)(high and medium-
skilled labour)
KNOW
LEDG
E-IN
TENS
IVE
INPU
TS
Highlyknowledge-
intensivevalue-added
High-skilledlabour
ICT capital services
Machinery and equipment capital
services (KIK)
HIGH
LY K
NOW
LEDG
E-IN
TENS
IVE
INPU
TS2 x 2 categories of knowledge inputs used in productive activities
[ 7 ]
Data
• EU KLEMS database
• 18 countries
• 28 industries
• 1980-2007
Table 3. Countries and period coverage
Countries Period covered
Germany 1980‐2007
Australia 1980‐2007
Austria 1980‐2007
Korea 1977‐2005
Denmark 1980‐2007
United States 1970‐2007
Slovenia 1995‐2006
Spain 1980‐2007
Finland 1970‐2007
France 1980‐2007
Ireland 1995‐2005
Italy 1970‐2007
Japan 1970‐2006
The Netherlands 1979‐2007
Portugal 1995‐2005
United Kingdom 1970‐2007
Czech Republic 1995‐2007
Sweden 1993‐2007
[ 8 ]
Industry results: knowledge intensity by industries and countriesFIGURE 1. GVA knowledge intensity by industries, 2007. (percentage)a) Germany b) US
0 20 40 60 80 100
EducationMining and quarryingTransport equipment
Renting of m&eq and business…Electrical and optical equipment
Manufacturing, nec; recyclingPulp, paper, printing and…
Public Admin.Health and social work
Financial intermediationRubber and plastic products
Coke, refined petroleum, nuclear…Basic metals and fabricated…
Hotels and restaurantsOther non-metallic mineral…
ConstructionChemicals
Mahinery, necWood and products of wood and…
Wholesale and retail tradeAgriculture and fishing
Food, beverages and tobaccoTransport and storage
Textiles, leather and footwearTOTAL INDUSTRIES
Post and telecommunicationsOther comm., social and…
Electricity, gas and water supplyReal estate activities
0 20 40 60 80 100
Renting of m&eq and business…Electrical and optical equipment
EducationHealth and social work
Transport equipmentMahinery, nec
Pulp, paper, printing and…Transport and storage
ConstructionPublic Admin.
Rubber and plastic productsManufacturing, nec; recycling
Wholesale and retail tradeWood and products of wood and…
Financial intermediationOther comm., social and…
Basic metals and fabricated…Textiles, leather and footwear
Other non-metallic mineral…TOTAL INDUSTRIES
Post and telecommunicationsFood, beverages and tobacco
Hotels and restaurantsCoke, refined petroleum, nuclear…
ChemicalsAgriculture and fishing
Electricity, gas and water supplyMining and quarryingReal estate activities
c) South Korea
0 20 40 60 80 100
Renting of m&eq and business…Education
Wood and products of wood and…Pulp, paper, printing and…
Other comm., social and personal…Electrical and optical equipment
Other non-metallic mineral…Health and social work
Mahinery, necTransport equipment
ConstructionTextiles, leather and footwear
Agriculture and fishingWholesale and retail trade
Hotels and restaurantsRubber and plastic products
Transport and storageManufacturing, nec; recycling
TOTAL INDUSTRIESChemicals
Public Admin.Post and telecommunicationsFood, beverages and tobacco
Basic metals and fabricated metal…Financial intermediation
Coke, refined petroleum, nuclear…Mining and quarrying
Electricity, gas and water supplyReal estate activities
Highly-qualified human capital Medium-qualified human capital ICT Machinery and equipment (non-ICT)
d) Spain
0 20 40 60 80 100
EducationRenting of m&eq and business…Post and telecommunications
Health and social workPublic Admin.
Financial intermediationElectrical and optical equipment
Transport equipmentMahinery, nec
Other comm., social and…Wood and products of wood and…Coke, refined petroleum, nuclear…
Transport and storagePulp, paper, printing and…
Rubber and plastic productsOther non-metallic mineral…
TOTAL INDUSTRIESChemicals
Electricity, gas and water supplyBasic metals and fabricated…Wholesale and retail trade
Food, beverages and tobaccoTextiles, leather and footwear
Mining and quarryingManufacturing, nec; recycling
ConstructionHotels and restaurantsAgriculture and fishing
Real estate activities
Note: 2005 for South Korea. Source: EU KLEMS and own calculations.
[ 9 ]
Industry results: dispersion of knowledge intensityFIGURE 2. Boxplot of GVA knowledge intensity by industries, 2005
Note: The Czech Republic, Ireland, Portugal, Slovenia and Sweden have not been considered .Source: EU KLEMS and own calculations.
[ 10 ]
Industry results: dispersion of knowledge intensityFIGURE 2. Boxplot of GVA knowledge intensity by industries, 2005
Note: The Czech Republic, Ireland, Portugal, Slovenia and Sweden have not been considered .Source: EU KLEMS and own calculations.
[ 11 ]
Industry results: temporal evolution of dispersionFIGURE 3. -convergence in GVA knowledge intensity by industries,1980-2005. (coefficient of variation)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
Agriculture, hunting, forestry and fishingMining and quarryingCoke, refined petroleum products and nuclear fuelElectricity, gas and water supply
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
Food, beverages and tobaccoTextiles, leather and footwearWood and products of wood and corkPulp, paper, printing and publishing
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
ChemicalsRubber and plastic productsOther non-metallic mineral productsBasic metals and fabricated metal prod.Machinery, nec.
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
Electrical and optical equipmentTransport equipmentManufacturing, nec; recyclingConstructionWholesale and retail trade
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
Hotels and restaurantsTransport and storagePost and telecommunicationsFinancial intermediationBusiness activities
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
Public AdministrationEducationHealth and social workOther social and personal servicesReal estate activities
Note: The Czech Republic, Ireland, Portugal, Slovenia and Sweden have not been considered .Source: EU KLEMS and own calculations.
[ 12 ]
Industry results: knowledge contribution to productivity
d) Spain
FIGURE 4. Labour productivity by components and indutries, 2007. (percentage)
a) Germany
0% 20% 40% 60% 80% 100%
TOTAL INDUSTRIESAgriculture and fishingMining and quarrying
Food, beverages and tobaccoTextiles, leather and footwear
Wood and wood and cork productsPulp, paper, printing and publishing
Coke, refined petrol. and nuclear fuelChemicals
Rubber and plastic productsOther non-metallic mineral prod.
Basic metals and fabric. metal prod.Machinery, nec.
Electrical and optical equipmentTransport equipment
Manufacturing, nec; recyclingElectricity, gas and water supply
ConstructionWholesale and retail trade
Hotels and restaurantsTransport and storage
Post and telecommunicationsFinancial intermediation
Real estate activitiesBusiness activities
Public AdministrationEducation
Health and social workOther social and personal services
b) US
c) South Korea
Note: 2005 for South Korea. Source: EU KLEMS and own calculations.
0% 20% 40% 60% 80% 100%
TOTAL INDUSTRIESAgriculture and fishingMining and quarrying
Food, beverages and tobaccoTextiles, leather and footwear
Wood and wood and cork productsPulp, paper, printing and publishing
Coke, refined petrol. and nuclear fuelChemicals
Rubber and plastic productsOther non-metallic mineral prod.
Basic metals and fabric. metal prod.Machinery, nec.
Electrical and optical equipmentTransport equipment
Manufacturing, nec; recyclingElectricity, gas and water supply
ConstructionWholesale and retail trade
Hotels and restaurantsTransport and storage
Post and telecommunicationsFinancial intermediation
Real estate activitiesBusiness activities
Public AdministrationEducation
Health and social workOther social and personal services
0% 20% 40% 60% 80% 100%
TOTAL INDUSTRIESAgriculture and fishingMining and quarrying
Food, beverages and tobaccoTextiles, leather and footwear
Wood and wood and cork productsPulp, paper, printing and publishing
Coke, refined petrol. and nuclear fuelChemicals
Rubber and plastic productsOther non-metallic mineral prod.
Basic metals and fabric. metal prod.Machinery, nec.
Electrical and optical equipmentTransport equipment
Manufacturing, nec; recyclingElectricity, gas and water supply
ConstructionWholesale and retail trade
Hotels and restaurantsTransport and storage
Post and telecommunicationsFinancial intermediation
Real estate activitiesBusiness activities
Public AdministrationEducation
Health and social workOther social and personal services
Human capital Machinery and equipment capital Non-qualified labour Real estate capital
0% 20% 40% 60% 80% 100%
TOTAL INDUSTRIESAgriculture and fishingMining and quarrying
Food, beverages and tobaccoTextiles, leather and footwear
Wood and wood and cork productsPulp, paper, printing and publishing
Coke, refined petrol. and nuclear fuelChemicals
Rubber and plastic productsOther non-metallic mineral prod.
Basic metals and fabric. metal prod.Machinery, nec.
Electrical and optical equipmentTransport equipment
Manufacturing, nec; recyclingElectricity, gas and water supply
ConstructionWholesale and retail trade
Hotels and restaurantsTransport and storage
Post and telecommunicationsFinancial intermediation
Real estate activitiesBusiness activities
Public AdministrationEducation
Health and social workOther social and personal services
[ 13 ]
Industry results: summary• Knowledge intensity in the same industry varies notably among
countries and warns us of the risk of classifying industries in categories by technological intensity.
• Human capital is more determinant in establishing differences in knowledge intensity than machinery, among sectors and within sectors.
• Data of activities specialized in the production of human capital and machinery and equipment confirm that inputs which incorporate knowledge are also produced with knowledge.
• The time series show that the use of knowledge is an increasingly common feature of the countries in the sample.
• Differences among countries in labor productivity for each sector are primarily associated with the contribution of highly qualified labor.
[ 14 ]
FIGURE 5. GVA knowledge intensity by countries, 1980, 1995 y 2007. (percentage)
0 20 40 60 80 100
South Korea
United Kingdom
The Netherlands
US
Japan
Sweden
Slovenia
Italy
Czech Republic
Denmark
Finland
Austria
Germany
France
Ireland
Australia
Spain
Portugal
Note: 2005 for Ireland, Portugal and South Korea and 2006 for Japan and Slovenia.Source: EU KLEMS and own calculations.
a) Knowledge-intensive inputs b) Highly knowledge-intensive inputs
0 20 40 60 80 100
South Korea
US
Finland
Slovenia
Japan
United Kingdom
Spain
Australia
Portugal
Sweden
Czech Republic
Ireland
France
The Netherlands
Austria
Germany
Denmark
Italy
2007 1995 1980
Aggregate results: knowledge intensity by countries
[ 15 ]
Aggregate results: dispersion and convergence
FIGURE 6. Boxplot of GVA knowledge intensity bycountries, 1985-2005
FIGURE 7. -convergence in GVA knowledge intensity. Total industries, 1980-2005. (coefficient of variation)
0.00
0.05
0.10
0.15
0.20
0.25
Note: The Czech Republic, Ireland, Portugal, Slovenia and Sweden have not been considered .Source: EU KLEMS and own calculations.
Note: The Czech Republic, Ireland, Portugal, Slovenia and Sweden have not been considered .Source: EU KLEMS and own calculations.
[ 16 ]
Aggregate results: determinants of differences
-40 -30 -20 -10 0 10 20 30 40
PortugalSpain
AustraliaIrelandFrance
GermanyAustriaFinland
DenmarkCzech Rep.
ItalySloveniaSweden
JapanThe NetherlandsUnited Kingdom
South Korea
Note: 2006 for Japan and Slovenia and 2005 for Ireland, Portugal and South Korea. 2005 for Figure 3c.Source: EU KLEMS and own calculations.
FIGURE 8. Knowledge-intensity shif-share analysis. 2007(absolute differences in percentage points on GVA)
a) Reference country: US b) Reference country: Germany
-40 -30 -20 -10 0 10 20 30 40
PortugalSpain
AustraliaIrelandFranceAustriaFinland
DenmarkCzech Rep.
ItalySloveniaSweden
USJapan
The NetherlandsUnited Kingdom
South Korea
c) Reference country: South Korea d) Reference country: Spain
-40 -30 -20 -10 0 10 20 30 40
PortugalSpain
AustraliaIreland
GermanyFranceAustria
DenmarkFinland
Czech Rep.Italy
SwedenSlovenia
JapanThe Netherlands
USUnited Kingdom
-40 -30 -20 -10 0 10 20 30 40
PortugalAustralia
IrelandFrance
GermanyAustriaFinland
DenmarkCzech Rep.
ItalySloveniaSweden
JapanUS
The NetherlandsUnited Kingdom
South Korea
Country Effect Specialization Effect Allocation Effect Total Effect
[ 17 ]
Aggregate results: knowledge and productivityFIGURE 10. Labour productivity by components, 2007. (2000 PPS euros per hour worked)
0 10 20 30 40 50
US
The Netherlands
Sweden
Germany
Finland
France
Ireland
Denmark
Italy
Australia
Austria
United Kingdom
Spain
Japan
Slovenia
South Korea
Portugal
Czech Rep.
Human capital Machinery and equipment capitalNon-qualified labour Real estate capital
Note: 2006 for Japan and Slovenia and 2005 for Ireland, Portugal and South Korea. Source: EU KLEMS and own calculations.
-30 -25 -20 -15 -10 -5 0 5
Czech Rep.
Portugal
South Korea
Slovenia
Japan
Spain
United Kingdom
Austria
Australia
Italy
Denmark
France
Ireland
Finland
Germany
Sweden
The Netherlands
Knowledge inputs Non-knowledge inputs Total
FIGURE 11. Decomposition of the differences in labourproductivity. Reference country: US, 2007(absolute differences in percentage points on GVA)
Note: 2006 for Japan and Slovenia and 2005 for Ireland, Portugal and South Korea. Source: EU KLEMS and own calculations.
[ 18 ]
Aggregate results: knowledge and GVA growthFIGURE 13. Knowledge contribution to the annual growth of GVA. 1980-1995 and 1995-2007 (percentage points)
-1%
0%
1%
2%
3%
4%
5%
6%
7%
8%
Sout
h Ko
rea
Japa
n
Aust
ralia US
Spai
n
Aust
ria
The
Net
herla
nds
Ger
man
y
Uni
ted
King
dom
Ital
y
Fran
ce
Den
mar
k
Finl
and
Czec
h Re
p.
Irel
and
Port
ugal
Slov
enia
Swed
enNote: 1990-2006 for Japan and Slovenia and 1990-2006 for Ireland, Portugal and South Korea in figure 10b.Source: EU KLEMS and own calculations.
a) 1980-1995
-1%
0%
1%
2%
3%
4%
5%
6%
7%
8%
Irel
and
Slov
enia
Sout
h Ko
rea
Finl
and
Spai
n
Aust
ralia US
Swed
en
Uni
ted
King
dom
The
Net
herla
nds
Port
ugal
Aust
ria
Czec
h Re
p.
Fran
ce
Den
mar
k
Japa
n
Ger
man
y
Ital
yKnowledge-based GVA Non knowledge-based GVA Total GVA
b) 1995-2007
[ 19 ]
Aggregate results: knowledge and productivity growthFIGURE 14. Knowledge contribution to the annual growth of labour productivity. 1980-1995 and 1995-2007 (percentage points)
-1%
0%
1%
2%
3%
4%
5%
6%
7%
8%
Sout
h Ko
rea
Japa
n
Finl
and
Spai
n
Aust
ria
Uni
ted
King
dom
Fran
ce
Ger
man
y
Den
mar
k
Ital
y
Aust
ralia
The
Net
herla
nds
US
Czec
h Re
p.
Irel
and
Port
ugal
Slov
enia
Swed
en
Note: 1990-2006 for Japan and Slovenia and 1990-2006 for Ireland, Portugal and South Korea in figure 14b.Source: EU KLEMS and own calculations.
a) 1980-1995
-1%
0%
1%
2%
3%
4%
5%
6%
7%
8%
Slov
enia
Irel
and
Sout
h Ko
rea
Swed
en
Finl
and
Japa
n
Czec
h Re
p. US
Uni
ted
King
dom
Ger
man
y
Aust
ralia
Fran
ce
Port
ugal
Aust
ria
The
Net
herla
nds
Den
mar
k
Spai
n
Ital
yKnowledge activities Non-knowledge activities Total
b) 1995-2007
[ 20 ]
Aggregate results• Recently, in several advanced economies the intensity of the use of
knowledge has been higher than 75%. The increasing intensity is associated primarily with higher education and ICT.
• Country differences in knowledge intensity decrease, due to the positive evolution of less developed countries.
• The main differences among countries are based on intra-industry specialization, rather than inter-industry specialization.
• The disadvantages in country productivities in comparison with the US stem from the high use of knowledge-intensive factors in the American economy.
• From 1980 to 2007 knowledge has played a crucial role in the growth of GVA and productivity in almost all developed countries.
[ 21 ]
Knowledge accounting: conclusionsThe results of applying the proposed metric of the use of knowledge-based factors in productive activities show that:
• Today advanced economies are extensively based on knowledge: more than half of the GVA remunerates human capital and machinery, factors which incorporate knowledge to the production processes.
• Knowledge is an increasingly common feature of all industries: the factors which incorporate it are the basis for competitiveness and growth.
• The differences among countries in knowledge intensity stem primarily from intra-industry differences: due to the very different activities being carried out within them and to their different weight in each economy.
[ 22 ]
Knowledge accounting: policy implications
• The conventional country classification by categories of technological intensity prevents us from showing that the same industry can differ in the intensity of knowledge among countries what is critical for promoting the use of knowledge is the renewal of the productive structure in each industry.
• The differences among industries in the intensity of use of human capital and machinery do not respond to the dichotomy manufacturing vs. services manufacturing is not superior in the use of knowledge.
• The highest contributions of knowledge to productivity lie on a better use of human capital, particularly, of the workers with higher education if education does not meet the needs of industry or if employment offered by firms does not exploit the potential of education, more value added will not be generated nor wages will be able to increase.
•
[ 23 ]
Extensions
• Case study: Spain• Pérez y Benages (2012): “El PIB basado en el
conocimiento: importancia y contribución al crecimiento”, Ivie, forthcoming.
• GVA vs Total Production knowledge intensity• Knowledge intensity of intermediate inputs• Knowledge intensity of imports
• Knowledge intensity as determinant of productivity and growth: the role of different components
[ 24 ]
Knowledge accounting: an economic approach
Francisco Pérez and Eva BenagesIvie and Universitat de València
2012 WORLD KLEMS ConferenceAugust 8-10, 2012Harvard University