Transcript Keynes ja suur majanduskriis - IPF

Beyond the Internet:
Lessons from Estonia and Slovenia
Meelis Kitsing
International Policy Fellow
Center for Policy Studies
(affiliated with the Open Society Institute
and Central European University)
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Outline









Why does Internet diffusion matter?
Methodology and theoretical framework
Why are Estonia and Slovenia chosen for cases?
Outcomes in Estonia and Slovenia
Characteristics of Estonia
Characteristics of Slovenia
Political economy explanations
Lessons for other countries
Policy options
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Why does the Internet diffusion matter?
Formal reasons




Enlargement of the European Union
The eEurope+ 2003 initiative “The EU candidate countries …
avoid a digital divide with the EU.”
Other countries are engaged in promoting the diffusion of ICTs;
it is crucial not to be left behind.
World Bank, UNDP and other international organization give aid
and encourage countries to develop ICTs.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Why does the Internet diffusion matter?
Political reasons

“Technology will make it increasingly difficult for the state to
control the information its people receive. The Goliath of
totalitarianism will be brought down by the David
of the microchip.” --Ronald Reagan



Freer flow of information
Reduced transaction costs for accessing information and public
services
New means for civic activism (organizing protests, participating
in the organization et al)
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Why does the Internet diffusion matter?
Economic reasons

Productivity gains

“Progress can be measured by how few workers - not how
many - are needed to get the job done. The joke is that the
factory of the future will employ only one man and one dog.
The man is needed to feed the dog. The dog's job is to keep the
man from touching the computer.” -- Robert D McTeer,
(president of Federal Reserve Bank of Dallas)

Productivity gains are quite visible on the micro level.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Why does the Internet diffusion matter?
Economic reasons



Productivity gains are quite visible on the micro level.
No proof of productivity gains on macro level outside of the
United States (Daveri 2002, Pohjola 2001, Stiroh and Jorgenson
2002)
“Although there is increasing evidence that IT investment is
associated with an improvement in company performance in
industrial countries, studies that look at the bigger
macroeconomic picture find little correlation between IT
investment and overall productivity – and some studies even
find a negative correlation.” -- Matti Pohjola
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Why does the Internet diffusion matter?
Political economy reasons



Network externalities - Metcalfe’s law states that the value of a
network grows with the square of the number of people
connected to it.
The law is grounded in basic microeconomics: There are
externalities of being connected to the certain classes of goods
and “bads” (negative goods).
Value of telephone networks and the Internet is higher when all
people can use them for communicating with each other, as
opposed to only a limited number of people being able to use
them.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Why does the Internet diffusion matter?
Political economy reasons



Many negative externalities occur with the use of ICT networks
as increasing dependence makes individuals and organization
more vulnerable (privacy concerns, cyber crime, cyber terror,
emails from “bad” person).
The law indicates that positive network externalities prevail in
the case of ICT networks.
It is crucial to find a fine balance between open and smart
networks.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Why does the Internet diffusion matter?
Political economy reasons



An open network fosters innovation and competition because it
is by nature a dumb network, i.e. applications are hosted at the
edge of the network by anyone.
A smart network decreases both negative and positive
externalities by imposing higher levels of control, as the
applications are hosted at the network’s core.
Positive externalities and openness determine the value and
importance of the Internet, and it makes sense to diffuse ICT as
networks rather than specific ICTs. The concepts of network
externalities and open network capture both political and
economic benefits
of Lecture
the Internet.
Public
at Popper Room of
Central European University on
March 7, 2003
Telecosm - No longer just a fantasy

“At wavelengths running from the millimeters of microwaves
down to the nanometers of visible light -- is the new frontier
of the millennium, empires of air and fiber that
command some 50,000 times more communications
potential than all the lower frequencies we now use put
together. A purely human invention, they provide
the key arena of economic activity for the new century.”
-- George Gilder
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Methodology and theoretical framework

A qualitative approach. Two detailed case-studies.

Dependent variable is Internet penetration.



Explanatory variables are institutions – both informal and
formal.
The theoretical framework is heavily based on the theories of
neo-institutional economics.
Institutions are the rules of the game in society, and can be
divided into formal (laws, regulations) and informal institutions
(customs, traditions).
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Methodology and theoretical framework



Institutions determine the cost of transactions. When cost of
transacting is too high, people may not undertake particular
activities (e.g. connect to the Internet, start an ISP).
Transaction costs do not matter only at the corporate level
(Coase 1937); they hold importance for the whole development
of society (Coase 1960). Effective and low-cost enforcement of
contracts are the key to progress.
Performance of economies is a consequence of the incentive
structures, which are in turn determined by the institutional
framework.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Why were Estonia and Slovenia chosen for
case-studies?

Estonia and Slovenia have achieved similar outcomes in Internet
diffusion

Estonia and Slovenia have different political economy models

Different paths of transition from socialism to a market economy

Both are small, open economies close to the European Union

Both countries were part of a larger union and have had to
build-up a nation-state
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Outcomes in the Internet diffusion
Country
Internet users in 1999
(per 10 000 inhabitants)
Czech Republic
682
Estonia
1387
Greece
705
Hungary
597
Italy
1430
Latvia
430
Lithuania
279
Poland
542
Portugal
1000
Romania
267
Slovakia
1112
Slovenia
1257
Spain
703
2000
2001
971
2721
947
715
2304
619
609
725
2494
357
1203
1507
1343
1363
3005
1321
1484
2758
723
679
978
3494
447
1203
3008
1828
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Explaining outcomes geography, size and wealth



Geography – it matters but certainly not as much as one would
expect
Size – also matters, but many examples exist of “small”
countries with low levels of Internet diffusion
Wealth – important indicator in general terms, but Estonia is 2.5
times poorer than Slovenia, and many countries with a similar
per capita GDP to that of Estonia and Slovenia have lower levels
of Internet diffusion.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Explaining outcomes civil liberties and democracy



Civil liberties and democracy – statistical research by Dimitrova
has indicated that the status of civil liberties and
democratization is strongest predictor of country-level Internet
use.
Extremely important factor for post-communist countries known
for violating civil liberties and authoritarian rule.
But this finding is not helpful for explaining outcomes in
comparing these two cases to the other EU candidates.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Explaining outcomes - computers and main lines



Number of personal computers – regression analysis
demonstrates that Estonia is a clear outlier as it has more
Internet users than availability of personal computers would
predict. Slovenia is an outlier in opposite sense as many of the
PCs are not connected to the Internet.
Main telephone lines – it matters in the case of less-advanced
countries where main line penetration is extremely low.However,
most EU candidate countries have a fairly similar penetration of
main telephone lines.
Regression analysis demonstrates that Estonia and Slovenia are
clear outliers – it means that Internet penetration is significantly
higher than the number
of main telephone lines would predict.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Estonia - Soviet heritage



Until the break-up of the Soviet Union, Estonia benefited from
an extremely limited diffusion of Western ICTs. COCOM imposed
tight export controls on the exports of ICTs.
However, the infrastructure of main telephone lines was more
advanced than in most other parts of the Soviet Union.
Presence of Soviet light industry and the demand for electronics
skills benefited the development of the “epistemic IT
community.” The Soviet IT specialist developed a pirated version
of Macintosh called AGAT and a pirated IBM called EC. Estonian
IT specialists developed a PC called JUKU, which was not
pirated.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Estonia - Government




In the early 1990s, the local IT community became crucial in
setting government policies in terms of IT spending,
procurement and use (a certain percentage of general budget
category is dedicated for IT spending).
Use of Internet in government affairs and making services
available for citizens.
Spending on Internet access in the schools and public Internet
points.
Active promotion of use of Internet and e-government.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Estonia - Private sector



Establishment of local PC assemblers, such as Microlink and
Astrodata, and later Ordi, which gained a significant market
share and benefited from government spending on IT.
Mid-1990s establishment of Internet banking services by the
leading banks.
Active use of Internet by companies for communication with
outside world and within companies
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Estonia - Telecomm market
liberalization



Monopoly in the fixed lines over voice telephony until the end of
2000 (competition in mobile telephony)
Free market in data transmissions, ISPs and backbone
providers.
Leased lines and alternative uses of infrastructure use was
partially liberalized before the end of 2000.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Estonia - Privatization of the
incumbent telecomm company




A concession agreement between Sonera/Telia and government
in 1992, which granted a monopoly position for a limited time.
In return, government got a strategic shareholder and
investment in infrastructure
In 1997, 24 percent of the shares owned by the government
went for sale through initial public offering.
Government’s stake is reduced to 27 per cent and it has a
golden share. Telia/Sonera have 49 per cent and other investors
have 24 per cent.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Estonia - Trade and FDI




Estonia radically liberalizes its trade barriers by moving to
unilateral free trade in 1995 and re-orientating its trade flows to
the West.
Hence, it makes it possible to exploit the benefits of
liberalization of COCOM regime to the full extent.
It achieves the highest trade to GDP ratio in the Central and
Eastern Europe.
Open foreign direct investment regime attracts investors mainly
from Finland and Sweden by creating a positive spillover for the
Lecture at Popper
of
diffusion and usePublic
of Internet
withinRoom
Estonia.
Central European University on
March 7, 2003
Characteristics of Slovenia - Yugoslav heritage



Limited technology transfer was possible before the
break-up of Yugoslavia. COCOM regime impact was not
as strong as in the other East bloc countries (Yugoslavia
was not a member of COMECON).
Technology transfer was made possible due to the
extensive trade relations with the Western Europe
(mainly Germany). 50 percent of the exports went to the
West in the 1980s.
IT-related research and education was established in the
mid-1970s. In the early 1980s the secondary schools
began to installPublic
mainframe
computers.
Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Slovenia - Yugoslav heritage



Development of local “epistemic IT communities” was
possible due to the existence of local technology industry
(Iskra Delta).
Decentralized control of economy and “social ownership”
of companies created incentives for development of
creative solutions (compared with the Soviet command
economy). Iskra Delta was broken up before the
dissolution of Yugoslavia.
Foreign direct investments in high-tech were made
usually in Slovenia. Siemens started a joint-venture with
Iskratel in 1989.Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Slovenia - Government



Government has pursued a pro-active industrial policy
which has benefited the IT industry.
Government has encouraged the use of Internet in
government affairs and made the crucial information
available on the Internet. However, e-services are
currently not available for the citizens. Ministry of
Information Society was created in 2001.
Government has also carried out specific projects to
increase the diffusion of Internet at schools, public
libraries and research institutions.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Slovenia - Private sector



Slovenia has a wide range of medium-sized hardware and
software companies.
Many multinationals (Siemens, Cisco, Microsoft) have invested in
Slovenia or partnered with Slovenian companies.
Slovenian companies are in relatively high positions in the value
chains of Western multinationals.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Slovenia - Telecomm market
liberalization



Originally, Slovenia planned to open the market for competition
by the end of 2000. However, Slovenia formally ended the
monopoly in fixed lines over voice telephony in 2001.
Implementation of the decision was constantly postponed.
Slovenia had officially liberalized the market in data
transmissions but it was a monopoly in reality. ISP services were
partially liberalized but licenses were required.Leased lines and
alternative infrastructure use was partially liberalized.
Market opening has been subject to constant pressure by the
EU.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Slovenia - Privatization of the
incumbent telecomm company




State agency was split into Telecom and Post in 1995
Government owns 74 percent, employees 13 percent, and state
investment funds the rest of the share of Telekom Slovenije.
The Slovenian government engaged in building a “national
champion” of the telecomm company.
Privatization has been constantly discussed with the EU, and
currently, the decision has not been made.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Characteristics of Slovenia - trade and FDI



In the 1990s Slovenia has gradually increased the share of the
EU in its trade.
Step-by-step, the trade barriers have been liberalized but further
reduction are required for entering into the EU.
The ratio of FDI to GDP remains well below average in the
region. WTO, OECD and other international organization see
main barriers in an open privatization policy and a stable
regulatory environment.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Political economy explanations




Path-dependence: Slovenia certainly had a better initial starting
position than Estonia. (infrastructure, skills, trade openness =
technology transfer)
Slovenia’s and Estonia’s policies towards the telecomm sector
also reflect the general path of reform and chosen political
economy models in both countries.
Estonia had a radical, shock therapy-type approach to the
transition and relatively free market economy.
Slovenia chose a gradual reform path and a political economy
system similar
Public Lecture at Popper Room of
University
on
to Central
socialEuropean
democratic
corporatism.
March 7, 2003
Political economy explanations



However, incertainty in market opening and privatization of
telecomm in Slovenia reflects a higher degree of regulatory
capture and the prevalance of an “old boy network” than in
Estonia.
Establishing a Ministry for Information Society with the aim of
dealing with telecom market issues is a good example. The
current inability to deliver an e-government in Slovenia indicates
a lack of incentives in the cozy government relationships.
Both countries achieved a similar outcome in Internet
penetration under the conditions of monopoly in the fixed lines.
Internet access costs
wereatsignificantly
Public Lecture
Popper Room of lower than in the Czech
Central
European
University Poland
on
Republic, Hungary,
Latvia,
Lithuania,
and Slovakia.
March 7, 2003
Political economy explanations



Before and after the market was liberalized in Estonia, Slovenia
was able to maintain somewhat lower prices for Internet access
than in Estonia, especially so in the peak hours. However, under
monopoly conditions local calls are usually subsidized by
international calls.
This indicates that, in general, the Slovenian and Estonian
governments were able to keep the incumbent telecom
company to the account.
This can be explained by the interest group pressure: “social
ownership” in Slovenia and FDI in other ICT sectors (backbone,
mobile telephony,Public
ISPs)
in Estonia.
Lecture
at Popper Room of
Central European University on
March 7, 2003
Political economy explanations



Hence, trade openness and FDI is crucial for Estonia; it
increased the competitive pressures and helped overcome the
gap in the initial starting position with Slovenia.
Estonia has been more radical in changing its formal institutions
for reducing the transaction costs in the diffusion of the
Internet.
Slovenia’s change of formal institutions has been more gradual.
However, informal institutions embedded in the system and
Yugoslav heritage has facilitated the reduction of transaction
costs.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy lessons for other countries



Estonia and Slovenia demonstrate that it is possible to achieve a
similar outcome with different ways of transition.
Both cases indicate it is hard to change telecomm policy without
a change in the broader political economy framework. Hence,
liberalization of telecoms without liberalization of other
economic policies may not work.
Internet diffusion may spread under the conditions of monopoly
in the fixed lines when a country’s government is able to keep
control over the incumbent and regulatory capture is avoided.
However, such a situation may not be sustainable in the long
run.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy lessons for other countries


However, Slovenia’s outcome is strongly explained by the pathdependence derived from the former system.
Hence, the countries without such advantages may find it
difficult to implement Slovenia’s model.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy option I




Partially opening up a telecom market. Setting a firm deadline
for a complete market opening.
Sustaining a monopoly fixed lines or partial monopoly (e.g. rural
areas) for returns in investment into infrastructure.
Privatization of incumbent telecom company combined with a
Liberal foreign direct investment regime
Independent regulatory agency, by naming key decision-makers
for a long period of time (like an independent central bank)
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy option I


Actively promote the use of Internet by government bodies,
offer public services online
Unintended consequences of such policy could be



unpredictability of technological developments, which may
undermine the monopoly;
difficulty in determining the exact period of monopoly;
or possible lobbying by the monopoly to change the rules of the
game.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy option II




Open telecom market for competition
Privatize incumbent telecomm company and Ensure a liberal
foreign direct investment regime
Build infrastructure with government resources in the areas
where is a market failure (e.g. rural areas).
Create an independent regulatory agency by naming key
decision makers for a long period of time (like an independent
central bank)
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy option II


Actively promote the use of Internet by government bodies,
offer public services online
An unintended consequence of such policy might be

hidden in the definition of a market failure, which may lead to the
possibility that government competes with private structures.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy option III




Open telecom market for competition
Privatize incumbent telecomm company and ensure a liberal
foreign direct investment regime
Create an independent regulatory agency by naming key
decision-makers for a long period of time (like an independent
central bank)
Actively promote the use of Internet by government bodies,
offer public services online
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy option III

An unintended consequence of such policy might be

inequality of access in urban and rural areas.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy option IV




Open the telecomm market for competition
Do not privatize the incumbent telecom company but set a firm
long-term deadline for privatization and lock it in (through
international organizations and/or constitutional means).
Ensure a liberal foreign direct investment regime
Create an independent regulatory agency by naming key
decision-makers for a long period of time (like an independent
central bank)
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Policy option IV


Actively promote the use of Internet by government bodies,
offer public services online
Unintended consequences of such policy might be



difficulties in determining the length of monopoly,
unpredictability of technological developments that may undermine
the monopoly,
credibility of lock-in mechanism, and the perception that a
government-backed company hinders competition.
Public Lecture at Popper Room of
Central European University on
March 7, 2003
Thank you

Working papers are available. Presentation and policy
paper will be made available at www.policy.hu/kitsing
Public Lecture at Popper Room of
Central European University on
March 7, 2003