Information and Communications Technology ICT, Economic Growth and Development

1
Information and Communications Technology (ICT),
Economic Growth and Development
Matti Pohjola HSE Helsinki School of Economics
2
Contents of the lecture

• Facts about economic growth
• Lessons from economics and economic history
• How is ICT different from electricity as a GPT
• Analyses of growth and productivity impacts of
  ICT
• Studies on ICT diffusion
• Policy conclusions

3
1. Facts about economic growth

• Economic growth is a rather recent phenomenon
  (about 200 years old)
• Income inequality has grown
• GDP per capita is 25-30 times higher in the
  industrial countries than in the least developed
  countries
• Latecomers to economic growth have been able to
  catch up with the leading countries
• The Nordic countries
• Japan and the East Asian Tigers
• China and India ?

4
2. Lessons from economics and economic history

• GDP growth arises from improvements in labour
  productivity
• which are driven by advances in technology
• i.e. in ideas about how to combine inputs to
  produce outputs
• Economic history teaches that
• growth is not a smooth process but is subject to
  sharp accelerations and decelerations
• which are associated with the arrival,
  diffusion and exhaustion of new General Purpose
  Technologies
• steam, electricity, ICT

5
The components of the standard of living (GDP
per capita)
or
GDP per capita labour productivity ? labour
input per capita
It follows that, in the long run
Growth of GDP per capita growth of labour
productivity
6
Example GDP per capita growth in Finland,
1900-2005
GDP per capita and its components (indices 1900
1)

• During 1900-2005
• GDP per capita has increased 13-fold
• although hours worked per capita have declined
  by 10
• because GDP per hour worked has increased
  14.5-fold

7
The sources of labour productivity (Y/L)

• Human capital per worker (h)
• Capital intensity (k)
• Technology A
• ideas or knowledge about how to combine inputs to
  produce outputs
• product and process innovations
• the most important source
• Production function Y A F(K, hL)
• gt labour productivity Y/L A F(k, h)

8
The theory of economic growth
Skilled workforce, hL
Knowledge A
Patents
Machinery K manufacturing
Growth
Consumers
9
General purpose technology (GPT)

• All technologies are not equally important for
  economic growth
• GPT is the driving force of productivity growth
• It has the following characteristics
• pervasiveness it spreads to most sectors
• improvement it gets better and becomes cheaper
  over time
• innovation spawning it makes easier to invent
  new products and processes
• Examples steam, electricity, ICT
• GPT first decreases and then increases
  productivity growth
• by making prevailing business models inefficient
  and by creating new, more productive ones

10
Industrial Revolutions and General Purpose
Technologies

• ICT 1975-
• microprocessor and PC
• telecommunications
• Internet
• automation of white-collar work
• tools that amplify brain power

3.
2.

• Electricity 1890
• combustion engine
• road network
• automation of blue-collar work
• tools that amplify muscle power

1.

• Steam 1780
• railways
• industrial production
• mechanization of agriculture
• tools that amplify muscle power

11
The three phases of productivity improvement
resulting from the arrival of a GPT

• Manufacturing of the new technology
• product and process innovations
• Use of the new capital goods
• capital deepening the new capital goods are
  substituted for the old ones
• increasing capital intensity improves labour
  productivity
• Reorganization of work
• which the new technology makes possible
• Example the reorganization of factories made
  possible by the electric unit drive and the
  assembly line

12
The GDP growth contributions of the three GPTs
(percentage points per year)
Sources growth accounting studies by Crafts
(2002, 2004), Oliner and Sichel (2000)
13
ICT vs steam and electricity

• The output contribution from steam was quite
  small
• The maximum output contribution from electricity
  was about 1 percentage point per year
• it took 30-50 years for the productivity impact
  to show up
• most of it (70 ) arose from the increase in
  total factory productivity in the other sectors
  than the electricity producing sectors
• improvement of machine control, factory design
  factory design etc
• reorganization of blue-collar work (through, e.g.
  assembly line)
• The measured contribution of ICT outweighs both
  steam and electricity
• no observed TFP growth yet outside ICT production
• gt the best is yet to come!
• reorganization of white-collar work (e.g. in
  offices)

14
3. How is ICT different from electricity as a GPT

• Its diffusion has been faster
• Its relative price decline has been more rapid
• ICTs productivity growth contributions through
  production and use have been larger
• ICT investment is heavily concentrated in service
  industries whereas electricity investment was
  concentrated in mining, manufacturing and
  transport sectors
• services are drivers of economic growth, business
  services especially

15
ICT will be at least as pervasive as electricity
has beenPercentage of US households with
electric service and PCs
Source Jovanovic and Rousseau (2006), General
purpose technologies
16
The shares of electricity and ICT in the U.S.
nonresidential capital stock, 1920-2005 ()
17
ICT intensive industries in the US
Note ICT intensity share of industry ICT
capital stock in total industry capital stock
Source Kevin Stiroh
18
4. Analyses of Growth and Productivity Impacts of
ICT

• Growth accounting studies for advanced countries
  (USA and EU)
• about one half of labour productivity growth can
  be attributed to the use and production ICT
• Comparison of productivity growth in
  ICT-producing and ICT-using sectors
• fast productivity growth in ICT production
• pick-up of labour productivity growth in ICT
  using industries
• Correlation between ICT-investment and economic
  growth across countries
• correlation is weakand hard to measure
• Correlation between ICT-investment and
  productivity across firms
• correlation is strong but the direction of
  causality is not clear

19
1) Growth accounting studies
Production function Y A F(K, L) A(AICT, AO)
F(KICT, KO, L) Growth accounting dY/Y vICT
dKICT/KICT vO dKO/KO vL dL/L a dAICT/AICT
ß dAO/AO Here O refers to other than ICT
capital, technology etc, vs denote income shares
and a,ß are output shares. Practical problem
vICT and AICT are not observable but have to be
estimated. ICTs growth contributions vICT
dKICT/KICT and a dAICT/AICT
20
Growth accounting result regarding the impacts of
ICT in the United States, Finland and United
Kingdom
Conclusion About one half of observed labour
productivity growth can be attributed to ICT
Sources Jalava and Pohjola (2005), Jorgenson,
Stiroh and Ho (2003), Oulton (2002)
21
2) Comparison of productivity growth in
ICT-producing and ICT-using sectors
Contributions to labour productivity growth in
1995-99
Source van Ark, 2001. Will be updated by the EU
KLEMS Project
22
3) Correlation between ICT-investment and
economic growth across countries

• Difficult to find any statistically significant
  correlation (Pohjola 2001, 2002)
• Possible explanations
• Most countries have not yet invested much in ICT
• Poorer countries lack ICT enhancing
  infrastructure (Kraemer and Dedrick in Pohjola
  (2001))
• Economic structures are quite different
  manufacturing vs services
• The benefits of ICT are not to be found in the
  supply side but in the demand side (Quah in
  Pohjola (2001))

23
The studies are cross-country growth regressions
The level of labour productivity is increasing in
the savings rates in physical, human and ICT
capital.
24
The growth rate of labour productivity is
increasing in the savings rates in physical,
human and ICT capital but decreasing in the
initial level of labour productivity. This
equation can be estimated from cross-section data.
25
4) Correlation between ICT-investment and
productivity across firms

• there exist positive relationships between ICT
  investment and various measures of economic
  performance across firms in the US (Bresnahan,
  Brynjolfsson, Hitt et al.)
• the positive correlation between computer
  investment and labour productivity is much weaker
  over time than across firms
• similar but weaker impacts have been found for
  firms in France (Greenan, Mairesse and
  Topiol-Bensaid in Pohjola (2001))
• causality is difficult to prove it may be the
  case that the high-productivity firms (i.e. the
  firms on the technology frontier) invest in
  latest technology to retain their leading
  position

26
Conclusions from production function estimations

• output f(ICT-capital, other capital, labour)
• the average output elasticity of ICT- capital is
  about 0.05 in a sample of 20 studies
• this is roughly the same as the income share of
  ICT capital
• gt ICT capital earns normal returns!

27
5. Studies on ICT diffusion

• Explain differences between countries in the
  adoption and use of ICT
• Kiiski and Pohjola in Pohjola (2001) 75
  countries, 1995-2000
• Pohjola (2003) 49 countries, 1993-2000
• Factors having positive impact
• GDP per capita
• relative price of ICT
• Internet access cost
• education
• Factors which are not significant (although are
  often thought to be important)
• openness
• regulation or competition

28
6. Policy conclusions

• In the long run, the benefits from ICT use are
  greater than the benefits from ICT production
• Technology by itself is not a solution to any
  development problem, it only provides an
  opportunity
• Complementary investments in learning and works
  practices are needed
• Business strategies are at least as important as
  competence in technology
• a problem in many developing countries

29
Policy conclusions (contd)

• The price of ICT is the most important factor in
  explaining the diffusion of ICT
• Deregulation and competition matter if they lower
  prices
• But providing access to ICT in not enough,
  participation in the information society matters
  even more
• Education, learning, acceptance
• Intellectual property rights
• Weak rather than strong to promote diffusion

30
The future?

• The mechanisms of value creation are different in
  the digital economy from those in the industrial
  economy
• Information and networks are the sources of value
• Services industries are the drivers of economic
  growth
• software, ICT services, other business services
  
• Automation, outsourcing and offshoring of
  white-collar work is the new source of
  productivity growth
• This is estimated to increase the growth rate of
  labour productivity by half a percentage point
  per year in the United States
• Where will innovations locate?

31
Global deployment of work
High wage countries
Low wage countries
Product Design Business Strategy Engineering
Strategic Marketing Personal services
Basic manufacturing Routine Programming
Customer Service Impersonal services
Cross National Production System

• Routine and low-skill jobs tend to leave for
  developing countries
• Creative and high-skill jobs may stay in
  developed countries

32
The future?

• Two possible ways to benefit from the ICT and
  digital revolution
• to become a producer
• to become a user
• The production of both ICT equipment and services
  is shifting from high-wage to low-wage countries
• China is the largest computer producer
• India is the largest offshored ICT services
  market area
• both are becoming sophisticated ICT users
• New technology can work to reduce inequality in
  the world during the coming decades!

33
Indias ICT sector
34
Own research work
Pohjola, M. (ed) (2001), Information Technology,
Productivity and Economic Growth, Oxford
University Press, 2001. Pohjola, M. (ed) (2002a),
Special issue on The New Economy, Information
Economics and Policy 142 Pohjola, M. (2002b),
The New Economy in growth and development, Oxford
Review of Economic Policy 183 Pohjola, M.
(2003), The adoption and diffusion of ICT across
countries, in D. Jones (ed), The New Economy
Handbook, Elsevier. DCosta, A.P. (ed), (2006),
The New Economy in Development, Palgrave. -
based on UNU- WIDER conference organized by
Matti Pohjola