the importance of measuring business use of ict - itu… · the importance of measuring business...
TRANSCRIPT
11
Third Workshop on Information Society Measurement in Latin America and the Caribbean, Panama, 23 November 2006
THE IMPORTANCE OF MEASURING BUSINESS USE OF ICT
Martin Schaaper
Economic Analysis and Statistics Division, Directorate for Science, Technology and Industry
22
Why measuring business use of ICT?
Assess take up and diffusion– … and compare with other countries
How does the Internet change the way businesses run their operationsImpacts on overall economic performance (productivity, growth, etc.)– … and compare with other countries
Policymakers need this information to make sound decisions
44
ICT diffusion among OECD-and non-OECD countries
Diffusion can be measured by activity and financial statistics.OECD has been collecting activity statistics on ICT use and e-commerce from member countries for several years and first published data in 2001 (“OECD Science, Technology and Industry Scoreboard: Towards a Knowledge-based economy”).The most recent version of this publication has been the 2005 edition.
55
Business use of the Internet, 2004, Percentage of businesses with 10 or more employees
(1) All size classes
(2) 16+ employees
0
20
40
60
80
100
Japa
n (20
03)
Denmark
Finlan
dSwed
enGerm
any
Korea
Canad
aAus
tralia
(4)
Mexico
(200
3)
Netherl
ands
Italy
Greece
Norway
Poland
Chines
e Taip
ei (1)
Hong K
ong (
China)
Hunga
ryPort
ugal
Singap
ore (2
003)
(1)
Slovak
Rep
ublic
Thail
and (
2)Chil
e (1)
% Have Internet access Have own Web site Broadband (2)
66
Internet selling and purchasing, 2004, Percentage of businesses with 10 or more
employees
- 40- 30- 20- 10 0 10 20 30 40 50 60
Canada
Hong Kong, China (1)
United Kingdom
Germany
Sw itzerland (2002)
Australia
Singapore (2003) (1)
Japan (2003)
Korea (3)
Chinese Taipei (1)
Mexico (2003)
Chile (1)
Thailand (2)
Purchasing Selling
(1) All size classes(2) 16+ employees(3) Data refer to enterprises with 1 or more employees.
77
Total e-commerce transaction value (including via the Internet), 2002 to 2004.
As a percentage of total enterprise turnover
0
5
10
15
20
25
Irelan
dUnit
ed King
dom
Denmark
German
yFinl
and
Norway
Belgium
Austria
Czech
Rep
ublic
Portug
alSlov
ak Rep
ublic
Italy
Luxe
mbourg
Spain
Poland
Greece
%
2002 2003 2004
99
Business perceptions of the benefits of buying and selling over the Internet
Australian Bureau of Statistics data show that businesses which buy and sell over the Internet perceive benefits from doing so.The longer a business has been buying or selling over the Internet, the more likely it is to report benefits– first mover advantage?– survival of the fittest (those which are still selling are those
which have succeeded and therefore are more likely to see benefits)?
– or just getting better at it over time?
1010
AUSTRALIA (ABS): BUSINESS PERCEPTIONS OF THE IMPACT OF RECEIVING ORDERS (SELLING) VIA THE INTERNET, 2000–01
Quality
of cu
stomer
servi
ceBus
iness
costs
Sales o
r cus
tomers
Efficien
cy
Compe
titive
ness
Tradin
g hou
rsTra
ding a
rea
% o
f bus
ines
ses
selli
ng o
ver
the
Inte
rnet
Not applicableNo changeIncreasedDecreased
1212
OECD case study work: Electronic Commerce Business Impact Project (EBIP)
During 2001-2002, the OECD co-ordinated a case study project on the impacts of electronic commerce on business. The study involved nearly 220 firms across 11 countries and a common methodology was used across 14 broad sectors. Participants: Canada, France, Italy, Korea, Mexico, the Netherlands, Norway, Portugal, Spain, Sweden, the United Kingdom. The aim of the project was to improve understanding of impacts of electronic commerce on business. EBIP used common methodology for firm-level case studies to improve cross-country and cross-sector comparability.Previous case study information is anecdotal, fragmented, not comparable across sectors or countries.
1313
What is more important the ‘commerce’factor or the ‘e’ factor?
Successful e-commerce strategies led by commercial considerations.
E-commerce part of larger business and economic transformations. Successful application and use are embedded in broader business strategies with major emphasis on both e-commerce and ICT skills.
But e-commerce a major business innovation that most firms will have to adopt.
1414
Overall the Internet is having large impacts on how firms conduct business
Expected/actual e-commerce impacts by business function
0%10%20%30%40%50%60%70%
Adverti
sing
Ca ta logues and st
ock list
s
Informatio
n servi
cesNegotia
tion
Orderin
g
Billing &
Paym
entFinanceDe liv
eryInform
ation C
ap ture
Inform
ation M
anagement
Market Ana lysis
Market Development
E x pec ted E ffec ts A c tua l Im pac ts
T RA NSA C T IO NPREPA RA T ION
T RA NSA C T IONC O M PL ET IO N
PRO DUC T IONSUPPO RT
1515
.. Reshaping many business processesExpected/actual e-com process impacts by business process
0 %
5 %
1 0 %
1 5 %
2 0 %
2 5 %3 0 %
3 5 %
4 0 %
4 5 %
5 0 %
Diversi
ficati
onDiffe
rentia
tion
Custom
isatio
nBun
dling
Design
Logis
tics
Produc
tion
Co-ordi
natio
n
Expan
sion
Segmen
tation
Trust
Loya
lty
E xp e c te d E ffe c ts A c tu a l Im p a c ts
P R O D U C TIN N O V A T IO N
P R O C ES SIN N O V A T IO N
R EL A T IO N A LIN N O V A T IO N
1616
.. With generally positive benefits:one-third of firms had positive impacts on
turnover or profitability
0% 20% 40% 60% 80% 100%
Employment
Profitability
Turnover
Increase
Decrease
No Change
NA
1818
OECD Growth ProjectA two year study on the reasons for differences in underlying economic growth of OECD member nations during the 1990s.Final report released in 2001 - “The New Economy: Beyond the Hype”.General findings were that:– ICT is a key technology with the potential to transform economic
and social activity.– But it is not the only factor, others are:
• quality of human capital• providing more scope for entrepreneurs to explore business
opportunities• support for innovation• getting the fundamentals right – sound macro-economic
management, openness to trade and investment, efficient markets, well functioning economic and social institutions.
1919
OECD Growth Project (continued)
ICT plays three roles– Through capital deepening, as ICT is an important asset
in overall business investment.– Through multi-factor productivity growth in the
production of ICT goods and services (e.g. technological progress in semi-conductors).
– Through MFP growth thanks to the use of ICT, either through efficiency gains in individual firms, or through network/spillover effects from its use.
2020
OECD Growth Project (continued)
Findings in respect of ICT considered both ICT production and consumption and were as follows:– Productivity in the ICT sector can improve economic
productivity overall, but– … successful economies were more likely to have rapid
diffusion of ICT, particularly in service industries.– Deregulation of ICT industries encourages competition and
greater investment in ICT.– Policies which build confidence – an appropriate regulatory
and legal environment and government leading by example –are important.
2121
OECD Growth Project (continued)
In respect of ICT particularly, the policy recommendation from the final report was that:– “ICT is an enabling technology, that is transforming
economic activity. – Governments should:
• focus policy efforts on increasing the use of new technology
• increase competition and continue with regulatory reform in the telecommunications industry to enhance the uptake of ICT
• ensure sufficient competition in hardware and software to lower costs
• build confidence in the use of ICT for business and consumers; and, make e-government a priority.”
2323
Much better evidence on the economic impacts of ICT at three levels of analysisthrough macro-economic evidence on the role of ICT investment in capital deepeningthrough sectoral analysis showing the contribution of ICT-producing sectors and ICT-using services to productivity growthand through detailed firm-level analysis that has demonstrated the wide-ranging impacts of ICT in the economy, even in sectors where sectoral data suggested that little was happening – this result is thanks to work in 13 OECD countries.
2525
The data on ICT have improved
Early 1990s: mainly private data, sometimes with poor coverage, sample bias, and unknown quality.
Now:– Surveys on ICT use by households and businesses in most OECD
countries, with considerable detail on uses and technologies.– Official estimates of ICT investment, including software.– Industry data for many countries.– Growing comparability of data.– Efforts to establish longitudinal (linked) databases in many
countries.
Greater prospects for empirical research on ICT.
2626
Prices have fallen…(producer price index for PCs and workstations, 1998=100)
Source: BLS.
010
20304050
607080
90100
Dec-98
Apr-99
Aug-99
Dec-99
Apr-00
Aug-00
Dec-00
Apr-01
Aug-01
Dec-01
Apr-02
Aug-02
Dec-02
Apr-03
Aug-03
Dec-03
Apr-04
2727
But countries have not invested to the same extent…(ICT investment as % of non-residential investment)
0
5
10
15
20
25
30
35
Spain
Portug
alIre
land
Austria
France
Japa
nIta
lyGerm
any
Netherlan
dsGreec
eAus
tralia
Canada
Denmark
Belgium
United King
domFinl
and
Sweden
United Stat
es
1980
1990
2002*
Source: OECD Database on Capital Services, March 2004.
2828
Why the differences in growth from ICT?Returns (& investment) in ICT depend on other factors:– Skills, innovation, organisational changes.– Not all firms succeed – experimentation and selection play a
role – can successful firms gain market share?– The scope for change depends on the business environment;
impacts in the US seem larger, perhaps because of greater scope for innovation and organisational change.
Competition matters:– Competition forces firms to enhance efficiency.– Firms that invest first/most in ICT are often new or foreign.– Competition reduces the costs of ICT & fosters diffusion.
Structural factors – share of services and of large firms.
2929
Aggregate impacts of ICT investment
Mainly analysed through standard growth accounting approach, based on capital services:– Requires long time series of ICT investment.– For international comparisons – use of hedonic deflators for
all countries to enhance comparability of results.
Official data on ICT investment are now available for 15-20 OECD countries.Still some measurement problems, e.g. as regards software investment.
3030
The US and small EU countries have had a large contribution of ICT investment, France, Germany and Italy a small one
(contribution to GDP growth, in percentage points)
0.0%
0.1%
0.2%0.3%
0.4%
0.5%
0.6%
0.7%0.8%
0.9%
1.0%
France
German
yIta
lySpa
inPort
ugal
Finlan
dGree
ceJa
pan
Canad
aIre
land
Austra
liaNeth
erlan
ds
United
Kingdo
mBelg
iumDen
markUnit
ed Stat
esSwed
en
90-95
95-2002*
Source: OECD Productivity Database, May 2004.
3131
Finding: ICT investment and diffusion mainly differ because of lack of demand
Demand has been held back by a business environment that was not sufficiently geared towards effective use of ICT – e.g. regulatory barriers, lack of skills, difficulty to change organisational set-ups, lack of innovation, etc.Competition has not been equally strong in all countries – costs of ICT and communications differ, as do pressures to improve performance.Demand has been held back by lack of security and trust.Structural factors also matter – ICT is not suited to all markets, sectors, or business models.
3232
Sectoral approach to impacts of ICTFocus on ICT production – what role does it play, does it explain US-EU differences:– Requires definition of ICT production:
• Developed by OECD (both ICT manufacturing and ICT services)• But not easily applicable for constant price series.
– Typically focus on some key components.
Focus on ICT-using industries:– Sectors that use ICT most (or have the greatest potential for
ICT use) might first observe productivity impacts – these are mainly services industries.
– Definitions differ and are not used the same across studies.
3333
ICT manufacturing is only important for some countries(average annual contribution to labour productivity growth, in %)
-0.10.00.10.20.30.40.50.60.70.80.91.01.1
Spain
Italy
Mexico
Netherl
ands
Canad
aGerm
any
Denmark
United
Kingdo
mSwitz
erlan
dBelg
iumFranc
eJa
pan
United
States
Sweden
Finlan
dIre
land
Korea
1990-95
1996-2002*
3434
Productivity growth in ICT services (telecom and software) also plays a role
Annual average contribution to labour productivity growth, in %
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Switzerl
and
Belgium
Norway
Denmark
Austria
France
Spain
United
States
Netherl
ands
Japa
nMex
ico Italy
Canad
aSwed
en
United
Kingdo
mIre
land
Korea
Luxe
mbourg
Finlan
dGerm
any
1990-95
1996-2002*
3535
Countries with less product market regulation have seen a stronger pick-up in productivity in ICT-using services
Australia
AustriaBelgium
Canada
Denmark FinlandFranceGermany
Ireland
ItalyJapan
SwedenSpainNetherlands
United Kingdom
United States
-1.0
-0.5
0.0
0.5
1.0
1.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0Index of Product Market Regulation, 1998
Incr
ease
in p
rodu
ctiv
ity g
row
th, I
CT-u
sing
indu
strie
s, 19
90-9
5 to
199
6-20
01 (%
poi
nt) Correlation = -0.71
T-statistics = -3.80
3636
Countries with strict employment protection legislation and product market regulation have invested less in ICT
Australia
Austria
Belgium
Canada
Denmark
Finland
France
Germany GreeceIreland
ItalyJapan
Netherlands
Portugal
Spain
Sweden
United Kingdom
United States
0
5
10
15
20
25
30
0 1 2 3 4Employment protection legislation index
ICT
inve
stm
ent i
n 19
98 (a
s %
of G
FCF)
Correlation = -0.65T-statistics = -3.46
3737
Countries with strong innovation in applications and software have invested more in ICT
Austria
BelgiumDenmark
France
GermanyGreece
Ireland
Italy
Netherlands
Portugal
Spain
Sweden
United Kingdom
AustraliaCanada
Japan
United States
10
12
14
16
18
20
22
24
26
28
30
0 10 20 30 40
Share of ICT patents in all patents, 1998
ICT
as a
% o
f non
-resi
dent
ial i
nves
tmen
t, 19
98
Correlation = 0.59T-statistic = 2.84
3838
Countries with an abundance of high-skilled workers have invested more in ICT
Sweden
France
United States
Greece
Italy
GermanyDenmark
Belgium
Netherlands
United Kingdom
Austria
Spain
Finland
Luxembourg*
0
1
2
3
4
5
6
0 1 2 3 4
Share of high-skilled ICT workers in total occupations
ICT
inve
stm
ent (
%G
DP
)
Correlation = 0.68T-statistic = 3.04
3939
And countries with a rapid increase in investment in ICT have had more rapid MFP growth
Ireland
Finland
Denmark
Sweden
Australia
CanadaUnited States
Netherlands
Austria
Italy
Japan
France
Germany
Spain0
2
4
6
8
10
12
14
-1.5 -1.0 -0.5 0.0 0.5 1.0Change in MFP growth from 1980-90 to 1990-2000
Cha
nge
in IC
T in
vest
men
t as
% o
f GFC
F, 1
990-
2000
4040
Problems with the industry approachBreakdown in ICT-producing and ICT-using industries is not always very satisfactory.Industry data are not always comparable, e.g. for ICT production and certain services.The impact of ICT on productivity is not formally analysed, ideally:– Regression analysis of impacts ICT use/ICT capital at the
industry level.– Growth accounts at the industry level.
Unfortunately, industry data on ICT investment are still lacking for many countries (no time series).
4141
A firm-level perspective on ICTOECD project involving researchers and
statisticians in 13 countries
4242
Large variety of methods and dataData based on combination (linking) of sources:– Economic performance often from production surveys;– ICT from investment surveys, specialised ICT and e-
commerce surveys, innovation surveys (US from computer network use survey).
– Additional data, e.g. on organisation, skills & innovation.Methods include:– Labour productivity regressions.– Estimates of production functions.
Variety of methods and data can help strengthen evidence.The studies point to many important interactions.
4343
Example 1 Hollenstein: What determines ICT uptake by firms?
Study for Switzerland, shows that ICT is linked to other firm-level factors.Firms that adopt ICT:– Anticipate benefits from improved customer-orientation
and lower costs.– Have overcome problems in financing and skills.– Have a high capacity to absorb new knowledge (linked to
human capital and innovation).– Are faced with strong competition and in markets with a
high level of uptake of ICT.– Also introduce new forms of work organisation.
4444
Example 2 Gretton, et. al: The impacts of ICT in Australia
The use of ICT raised aggregate multi-factor productivity by about 0.2% points – over and above impacts of ICT capital.Impacts are largest in finance and insurance.Impacts taper off over time – they are largest a few years after adoption.Complementary factors play a large role in explaining firm’s success with ICT – human capital, experience with innovation, use of advanced business practices and intensity of organisational restructuring.Early users of ICT are large firms with skilled managers.Firms with high openness to trade use more ICT.
4545
Estimated contribution of ICT to MFP growth in Australia (% points)
0.18
0.07
0.13
0.17
0.14
0.11
0.06
0.29
0.090.09
0.140.14
0.04
0.14 0.14
0.12
0.28
0.09
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Man
ufac
turin
g
Con
stru
ctio
n
Who
lesa
le tr
ade
Ret
ail t
rade
Acco
m.,
cafe
s &
rest
uara
nts
Tran
spor
t &st
orag
e
Cul
tura
l &re
crea
tiona
lse
rvic
es
Pro
perty
&bu
sine
ssse
rvic
es
Gro
up to
tal
Computer use & Internet aloneComputer use, Internet & computer use interactions
Source: Gretton, et al. 2004.
4646
Example 3 Hempell, et al. : The link between ICT and
innovation
Comparative study with comparable data for Germany and the Netherlands from innovation surveys.Key question: do firms that engage in complementary innovation benefit more from investment in ICT?In both countries, returns to ICT investment are much higher when accompanied by innovation.Continuous innovation leads to higher returns than occasional innovation efforts.This is also the case in the services sector.
4747
Example 4 Maliranta and Rouvinen: The impacts of ICT in
FinlandThe additional productivity of ICT-equipped workers ranges from 8 to 18% (after controlling for many other factors).This effect is higher in young (new) firms and in the ICT-producing sector.Manufacturing firms benefit mostly from internal networks (e.g. local area networks), whereas service firms benefit most from external networks (e.g. the Internet).Organisational factors are important, as suggested by the greater productivity effects in new firms.Selection also plays a role – not all firms succeed – having a dynamic business sector is important.
4848
Example 5 Clayton, et al.: Impacts of e-commerce in the UK
Computer networks for trading have positive impacts on productivity.E-buying has positive impacts on output growth, whereas E-selling has negative impacts.This is likely due to pricing effects – e-buyers can source more competitively, e-sellers may loose out.In some industries, e-commerce seems to have contributed to lower prices, perhaps due to greater price transparency and more rapid reactions of suppliers.
4949
UK: Productivity analysis and ITInvestment + Employees + Transactions + Communications
Purpose of analysis is to understand different effects
Labour Productivity: by Employee Use of ICT in Manufacturing and Services
0
10
20
30
40
50
Les s than 25 percent
25 per cent to 49 percent
50 per cent to 74 percent
75 per cent and over
Pe rce ntage of e mploye e s using compute rs / inte rne t
£thousand
Using computers Using Internet
5050
UK: Productivity analysis and IT
– Hardware and software capital show significant productivity gains, both levels and yearly changes
– Gains attributable to IT capital across service sectors higher than those in manufacturing
– Returns to IT capital across high and low user sectors, and for large and small firms
– IT capital reinforces productivity effects of other capital investment, and of skills
– Returns to IT investment ‘explains’ productivity advantage of US multinationals in UK, concentrated in high IT using sectors
Significant results – IT investment
5151
UK: Productivity analysis and ITSignificant results – IT use
– Employee use adds significantly to productivity, over and above effects of hardware and software capital,
• after controlling for sector, size, ownership etc
– Manufacturing firms achieve an extra 2.2% output for each additional 10% employees using computers
– For employees using internet the extra is 2.9%
– Productivity effect greater for more recently created firms
– Productivity effects associated with communications useUK firm level productivity results can be found in full athttp://www.statistics.gov.uk/cci/article.asp?ID=1235 (Web link to ONS summary article)and are broadly consistent with firm level results for Chinese manufacturing athttp//www.mo.rcast.u-tokyo.ac.jp/index-eng.html (analysis by Kazu Motohashi)
5252
UK: Productivity and broadband
No clear link from broadband availability to productivity, but:
– Investment in IT higher in regions / sectors with broadband available, especially for UK firms in non-IT intensive industries.
– Early adopters of broadband experienced high initial telecommunication costs but these have declined over time
– Broadband users more likely to have multiple e-business links.
– Multiple links plus broadband improve labour productivity.
– Firms with a high broadband equipped labour share have significantly higher productivity
Full results in OECD paper DSTI ICCP IIS(2006)9 Shikeb Farooqui and Raffaella Sadun
Significant results – broadband use
5353
Example 6 Atrostic et al. : Impacts of computer networks in the
United States, Japan and Denmark
Computer networks have positive impacts on productivity in all three countries.In the US manufacturing sector, plants using computer networks have 5-11% higher productivity than plants not using networks.In Japan, both interfirm and intrafirm networks are linked to strong firm performance.
5454
The firm-level studies all show that ICT use can contribute to improved firm performance
Positive impacts of ICT on labour productivity (and MFP, where measured) or market shares in all countries:– But these are conditional on other factors and firm
characteristics (skills, innovation, organisational change).– Not all firms succeed – selection plays a role.– ICT is only part of a broader strategy to improve firm
performance.Networking technologies are particularly important.The impacts of ICT are also found in the service sector (despite lack of evidence at the industry level).
5555
Why the difference between firm-level and aggregate evidence on ICT’s impacts?
Aggregate impacts ICT in some countries may still be disguised by other factors.Lags – the US invested earlier and more – aggregate impacts in other countries might still come.US firm-level impacts may be larger – because of scope for reorganisation and complementary innovation.Successful US firms may be able to gain more market share (e.g. Walmart) – more re-allocation.Spill-over effects? - Perhaps already some of this in the US (and Australia), not much in other OECD countries.
5656
Conclusions
Better data have enabled a wide range of empirical research on ICT and its economic impacts.This research has had important policy implications:– It has shown that ICT plays an important role in economic
growth.– ICT is no panacea, and linked to many other factors – only
focusing on ICT diffusion is not sufficient.– Countries have not equally benefited from ICT – there is scope
for learning and improvements in the policy environment.
There is scope for much more analysis, in particular with firm-level data, and in the context of international comparisons.
5757
Policy implications from work on ICT and growth
Fostering a business environment for effective use of ICT– Competition matters for ICT uptake and effective use.– Regulatory burdens may limit investment and
experimentation.– Enable firms to make the necessary organisational
changes.– New skills are needed, requiring changes to education
and training systems.– Need for coherent innovation policies.– Management matters – ICT is no panacea.
5858
Policy implications (continued)
Competition in ICT goods and services– Evidence of the benefits of competition is convincing and
liberalisation should be pursued.– Crucial for broadband development.– Regulators must be vigilant for anti-competitive
behaviour.– Technology continues to develop – policy frameworks
and regulations should be technology neutral.
5959
Policy implications (continued)
Boost Security and Trust– Uncertainty remains regarding the security of electronic
commerce.– Attention is required to security of information systems
and networks, authentication, protection of privacy and consumer protection.
– Cross-border issues and enforcement.– E-government can help build confidence – and is
important to improve government efficiency and the delivery of public services.
6060
Policy implications (continued)
Unleash growth in the services sector– particularly important because of the potential
benefits– policies to reduce regulation and reduce barriers
to entryHarness the potential of innovation and technology diffusion– ICT linked to, and assists, innovation– ICT fosters networking and knowledge transfer.
6161
Measurement issuesBetter measurement has clarified the picture– Quality-adjusted price indices have helped
demonstrate the impact of ICT on growth.– Software was included in 1993 SNA as an asset.– Measures of R&D are being improved:
• New (2002) edition of OECD Frascati Manual.• R&D may be included in new SNA.
– Measures of innovation are being improved.– There is more attention at statistical offices for
analysis at the firm level -linking existing data sources can help.
– Data on ICT demand are improving with adoption of OECD standards.
6262
However, some measurement problems remain
Skills data (apart from formal education statistics) are poor.Organisational factors are difficult to capture.Software investment is not yet fully picked up in national accounts.Deflators for ICT production differ considerably.Measurement of services output is poorly in some areas e.g. finance, health, education. Cross-country firm-level studies are still scarce, although much work is underway. OECD is trying to act as forum for exchange of best practices.
6363
THANK YOU!THANK YOU!www.oecd.org/sti/measuring-infoeconomy