Contradiction, Confusion
and Hubris:
A Critical Review of European Information Society Policy
by Nicholas Garnham
"Let every student of nature take this as a rule, that
whatever the mind seizes upon with particular satisfaction is to
be held in suspicion."
Francis Bacon
EuroCPR was founded as the EU began the move from state monopoly
of telecommunication networks and services to regulated competitive
market provision – a process usually referred to as liberalisation
or de-regulation.
The purpose of EuroCPR was to match the telecommunication policy
research effort in the US by developing a European research capability,
in part to avoid simply importing US regulatory models and their
accompanying politico-cultural ideological assumptions. The aim
was also to bring that capability into a dialogue with policy makers
with the objective of ensuring a more rational policy-making process
with policy based upon the best available evidence and dispassionate
analysis, rather than the crude interplay of economic and political
interests.
As one of the parties involved, I think I can say how impossibly
naive that aim now looks. Indeed, I would suggest that, in the light
of nearly 20 years experience, the results have not been very encouraging.
I want today to explore the reasons for this and what useful conclusions
we might draw for both research and policy making.
Two years ago Eli Noam asked, and rightly, where the voices of
critical, or even sceptical, researchers could be heard during the
‘telecoms mania’ of the late 1990s. In trying to explain
the telecom/dot.com boom and bust Martin Fransman described it in
terms of a ‘consensual vision’, which, although it turned
out to be wrong in almost every respect and although there were
good reasons to dispute it at the time, was accepted by both market
actors and analysts to be an incontrovertible Truth. I want to follow
Noam and Fransman by asking where are the voices of the sceptical
researcher? Are policy makers now being gripped by the new consensual
vision of the information society? Have we learnt nothing?
So far, from the abundant research, and the increased use of this
research by policy makers leading to a firmer grasp of realities
and better policy formation, I would argue that as policy has moved
through its three cycles, liberalisation, convergence and the information
society, its ambitions have grown. It has progressed from aims of
shaping a competitive telecommunications market to bringing into
existence a new social order, or at least a new economic order –
the Lisbon goal of making Europe ‘the most competitive and
dynamic knowledge-based economy in the world by the year 2010’
– but at the same time its research-based grasp of reality
has weakened. In short the story now is one of policy over-reach
and research under-reach.
The problem in my view lies in both the policy-making and research
processes. Before going on to explain how and why I believe this
to be the case I will share with you my conclusions – conclusions
that I confess I find uncomfortable. Here I should note that I have
arrived by a somewhat different route at conclusions similar to
those reached by Johannes Bauer in his paper for this conference.
(It is good to know that one is not alone.) These conclusions are
that researchers (and I certainly include myself here) at least
in their roles as policy analysts and advisers, do not well understand
the processes they are analysing and exaggerate the possible scope
of their understanding, but are unwilling to admit to this limitation.
Policy makers are unwilling, maybe congenitally unable, to accept
either the very limited scope of their powers or admit to their
failures. Both researchers and policy makers thus tend to suffer
from severe historical amnesia.
In short, on both sides the story of the last 20 years is one of
hubris.
Why is this historical amnesia both important and debilitating?
Because it tends to make both research theories and models, and
the policies that draw on them for support, appear much more coherent
and tidy than in fact they are. In judging both the applicability
and relevance of an economic theory or model, or the efficacy of
a policy, we need to know the problem it was designed to address.
In discussing the research here I will focus on economics because
I think it uncontroversial to say that, for better or worse, it
is economic theories and models that have driven, or at least been
mobilised to legitimate, the telecommunication/industrial society
policy process
There is a tendency in all academic disciplines, and economics
is no exception, to search for maximal internal logical coherence
and scope in theories. In my view it is more useful to see theories
and models as responses to real world problems and applicable only
to that set of problems. It follows from this that theories and
related policies designed to address such problems are likely, when
faced by the extremely complex social system that is the contemporary
global economy, to throw up new problems requiring new theoretical
and policy responses elsewhere in the system. In such a situation,
and in my view this has been the situation in the telecommunications/information
society field over the last two decades, to cling to a theoretical
position in the name of intellectual consistency, or a policy in
the name of political consistency is a grave error. This has particularly
applied in recent years to models of the market and competition.
Thus, to understand the strengths and weaknesses of policy in the
telecommunication/information society field and of the research
underlying them in recent years, we need to be clear about the problems
these policies and theories were designed to address. The difficulty
is that there is agreement neither about the problems or their relative
importance, nor about the appropriate theoretical way to approach
them. This in particular has led policy makers, who of course want
to please everyone, to formulate sets of mutually contradictory
policies and/or to give regulatory agencies the invidious task of
balancing conflicting goals. Thus they mobilise what they do not
seem to appreciate or hope no one else will notice are mutually
contradictory theories.
In the current information society climate I have a strong sense
that in the intellectual and political excitement about detailed
analysis and policy implementation we have been distracted from
the underlying issues. The result, in my view at least, is that
too much policy research and formulation is now operating in a self-justifying
world of its own. This reminds me of the old Irish joke when on
being asked the way to get somewhere the reply comes: ‘If
I wanted to get there I wouldn’t start from here’.
Where then would I start? If we examine the development of research
and policy as it has moved from telecommunications, to converged
media, to the information society, the conclusion one reaches is
not so much that either the economic theories and models or the
policies were wrong (although some clearly were) as that they were
partial and often contradictory. Policy has been driven by a range
of different interests with differing definitions of the problem,
different aims and different supporting economic models and theories.
At the start of the liberalisation process we can identify a range
of policy goals and supporting analysis each of which involved different
conceptions of the role of telecommunications in the economy, different
models of the economy and of markets and, therefore, different definitions
of the problem that liberalisation and re-regulation were designed
to solve.
First there is the industrial policy approach. Here telecommunication
operators were seen primarily as investors in networks and associated
technologies. At national level the problem was seen as insufficient
investment in network modernisation and an associated inadequacy
of network facilities and services to meet demand. At the European
level it was the fragmented nature of the European equipment manufacturing
industry caused by national monopsonies seen as the cause of US
and Japanese domination of the global equipment business that was
the focus of attention. It was this policy perspective and analysis
that led directly into the series of Framework programmes and the
development of European standards, especially for mobile telephony.
There was an approach that stressed the role of the telecoms network
as a key business infrastructure. This was a response to pressure
from the corporate sector, particularly from multinationals, for
competitive supply of both networks and services, and at a European
level to facilitate cross border operation and, thus, harmonisation
of regulation. Here consumer welfare was seen to derive primarily
from increased business efficiency.
Another approach stressed innovation, and saw telecommunications
as primarily a field for ICT development and the regulated monopoly
structure as a barrier to the development of the ICT industry and
services. Here liberalisation was in part a response, as it had
been in the US, to pressure from the ICT industry.
At the European level there was an approach which stressed the
importance of cross national infrastructures both as unifying mechanisms
and as contributing to the efficiencies derived from market scale.
Here it was not the monopoly nature of the existing industry so
much as its national nature that was the problem.
Finally there was an approach that stressed the general role of
telecommunications in economic and social development. It is here
that we find the seeds of what became information society policy
with its vision of telecommunications as the infrastructure of a
knowledge society leading to Internet mania. This approach had close
links with the innovation approach if only as part of the marketing
discourse of the ICT industry. Here the network modernisation and
service innovation debate took on an aspect of religious faith,
thus often bypassing rigorous economic or cost-benefit analyses.
Across these policy perspectives ran a series of different and
often conflicting theories of market and market power.
1. Neo-classical competitive equilibrium. In the political rhetoric
it was this market model that was largely mobilised to both describe
and justify liberalisation. Re-regulation was to create easy market
entry and price competition which was supposed ipso facto to increase
consumer welfare. Here the arguments were about natural monopoly,
cost based pricing and so on.
2. Schumpeterian competition. Here the goal was not price competition
but entrepreneurial innovation, which in its turn depended upon
the monopoly rents derivable from successful innovation.
3. The Hayekian model of the market as a search mechanism in the
face of necessary uncertainty with prices and price competition
not seen as driving productive and allocatory efficiency, but as
signalling choices in the face of information overload.
Let me now look at each of these policy strands and their underlying
justificatory arguments in more detail.
· Telecommunication network access and services as a retail
commodity. The problem was seen as a loss of consumer welfare through
monopoly rents, and the solution as the creation of fully competitive
markets. The market model used was one of price competition with
low barriers to entry. The debate was over the extent of natural
monopoly, the position of public service, and the degree to which
network facilities and services could be efficiently unbundled.
The expectation was that sector specific regulation could be phased
out in favour of general competition. It is within this thematic
that most regulatory debate and policy development has taken place.
· Telecommunications as a business service and investment
good. This is often confused with theme (a), and this confusion
continues to bedevil information society policy, but in fact both
the characteristics of the market and the definition of the problem,
and possible policy responses are very different. Here, the drivers
of the policy were major corporate users and their allies in government.
Attention was focused on leased lines and advanced services. The
problem was seen to be the slowness of response, the assumption
that one size fits all, and the price averaging policies of monopoly
incumbents. At a more general level the problem was one of corporate
efficiency, productivity and inward investment rather than consumer
welfare directly.
In Europe there was a particular version of these problems, namely
the lack of European level operators or unified regulations in the
face of the desire of multinationals for one-stop shopping.
· Industrial policy focusing on telecommunications as both
a market for equipment, a major investor in, and therefore driver
of, technological development and as a provider of network capacity
for upstream hardware and software developers. Because of the attention
given to competitive retail and wholesale markets this crucial driver
of liberalisation policy has been overlooked. It is important to
note here that this has been a consistent concern that has been
reflected in papers presented at successive EuroCPR meetings, including
this one. It is important because, as we shall see, it has returned
to haunt information society policy.
We can see here two sets of arguments and two relevant interest
groups.
On the one hand, the problem, from the view of both national governments
and the Commission, was the nationally based monopsonistic relationship
between incumbents and equipment suppliers. There was a need both
to consolidate manufacturing (and associated research and development
– R&D) on at least the European scale to challenge North
American and Japanese competitors, while at the same time ensuring
continuing investment by operators in both R&D and equipment
purchase. We can witness in current policy and the recurrent versions
of the Framework programme the continuing strength of this industrial
policy strand and of the equipment manufacturers’ lobby. The
problem however was that this policy was quite at odds with a policy
aimed at squeezing operator margins in the name of consumer welfare.
In practice regulators were often left to balance the two.
The second set of arguments saw the whole field of telecommunications
as one of technological innovation in two senses.
1. It was necessary to have both network and service competition
in order that new network technologies and new software could enter
the market and ‘fight it out’. Thus the telecommunications
market was seen primarily not as a competitive market for a homogeneous
range of services, whether voice telephony or data transmission,
but as a market for technological and service innovation. Here the
main policy driver was the ICT industry. We should not forget that
the liberalisation movement in the US started with the desire of
the computer industry to have greater access to and control over
the telecommunication network and the services that ran over it.
2. The telecommunication network was the necessary infrastructure
for innovation not only in ICTs and software, but across the whole
field of the economy. It is here that we find the seeds of one strand
of information society policy and its stress on broadband and e-business
within a general innovation agenda. I will return to this later.
From this position the goal was network and service innovation.
The problem was and still remains who will pay for this innovation.
Here, two different economic models have been mobilised, each quite
different from the price based, low entry barrier consumer market
that has dominated regulatory debate.
The first, from which is derived the principle of technological
neutrality in regulation, is the Hayekian model of the market as
the only available tool for selecting technological options in the
face of radical uncertainty. In this model prices are signals that
aggregate a wide range of choices between technological solutions
across a myriad of users with different assessments of their needs.
It has much to be said for it within its own terms as a critique
of state planning. But it is quite incompatible with industrial
policy of the current EU type. It is not useful in the face of network
technologies with long lead times, and constant or growing returns
to scale, and where it is very difficult to identify a user that
is setting a price. In particular, it is incompatible with the influential
path-dependency and endogenous growth models, which drive much information
society policy. It is also important to stress that it is totally
incompatible, for better or worse, with a policy to encourage broadband
or use broadband penetration as a benchmark for policy success,
to set targets for R&D expenditure or target specific areas
for research support.
The second model, perhaps now the most influential of all, is the
Schumpeterian model of competition through innovation. Because the
concept of innovation is now so central to economic and regulatory
policy and the key driver of information society policy it is important
to outline the main features of the Schumpeterian model and the
problem it was addressing. Schumpeter, like Keynes, was faced with
the stagnation of the capitalist industrial economies. In his view
the equilibrium model of inter-firm price competition as the driver
of capitalist development led inevitably to squeezed rates of profit
and sectoral oligopolies and thus a fall in rates of investment.
As a way out he argued that dynamic market development was driven
not by price competition, but on the contrary by entrepreneurial
product and process innovation. The risk of such innovation was
covered by the possibility of monopoly rents for successful innovation.
It is essentially this argument that has been used to defend Microsoft
against anti-trust action. Crucially, it is incompatible with a
price competition, low entry barrier competitive market model.
Within our discussion two further problems arise here. First the
regulatory structure needed to encourage such entrepreneurial innovation
is quite different to that required to maximise consumer welfare
in established product and service markets. The difficulty in recent
years in the telecommunication/ICT field has to been to know which
situation we face. Thus, we have witnessed, and continue to witness
in broadband, a constant policy fluctuation between the need to
support network and service innovation and the need to maximise
consumer welfare.
Second, as in the telecommunication/dot.com bust, you can have
too much innovation and the process can then lead to serious overshoot
and over investment. In this case no innovator can establish the
monopoly rents required to make the process profitable and consumers
suffer because all investment is taken from current consumption,
as a bet on the future. It is important to stress that the telecommunication/Internet
boom of the late 1990s was essentially a Schumpeterian boom. Analysts,
investors and managers bought the Schumpeterian argument. The result
was speculative competition between investors for a share of the
future excess monopoly profits that the winners in this winner-takes-all
market would achieve, and it was upon this that valuations were
based.
It is important to stress here that the Schumpeterian analysis
has developed in two divergent directions. One sees the process
of innovation as a competition between entrepreneurs to produce
distinct products and services. The central dilemma here is one
of incentives to innovate and the risk/reward ratio. The other,
within a more general evolutionary economics framework, places the
emphasis on innovation in general purpose technologies and the problem
then becomes one of the long run adoption process.
We can see the different policies to which these strands lead if
we look at the field of intellectual property (IP). The first strand
favours a strengthening of IP protection as an incentive to innovate.
The second sees the situation as one of the optimum sharing of knowledge
in a socio-economic learning process and thus favours Open Source,
and minimal IP protection.
Against this background let me now turn back to EU information society
policy and the consensual vision underlying it. Francis Lorentz
gave us a very clear and concise presentation of this vision in
his opening remarks at this conference.
As with EU telecommunications policy, of which it is in part a
development, information society policy is characterised by a major
cleavage and contradiction between an industrial policy, a state
intervention strand and a competitive market strand. To take one
example, after years of championing the market competition model
and technological neutrality in telecommunication networks and services,
policy now, when the market has not delivered the supposedly desired
goods, has flipped to an obsession with broadband roll-out, the
targeting of a specific range of so-called information society technologies
(IST) through the Framework programmes and even an attempt to legislate
at state level for a specific level of R&D expenditure. Indeed,
my main criticism of the EU information society policy label is
that it serves as cover for the continuation of a largely failed
industrial policy and a protection at EU level of the budgets associated
with it.
I think one of the best ways to understand telecommunication and
then information society policy in Europe is by tracing back to
the original European coal and steel community, the twists and turns
of an industrial policy and state planning approach – a state
planning approach that owed its prestige and strength and a significant
proportion of its manpower to the success of the French Commissariat
du Plan in the 1950s and 1960s – as they grappled with the
pro-competitive and anti-state provisions of the Treaty of Rome.
From the point of view of policy makers and lobbyists the information
society brand or consensual vision has three advantages.
1. It is sufficiently vague as to hide these contradictions.
2. It is sufficiently forward looking to distract attention from
the past failures of what is in effect the same policy.
3. It conveys an illusion of socially useful activity and relevance
while distracting attention from the analysis of the problems to
which these policies are a supposed solution.
The consensual vision consists first of a definition of the economic
problems facing the EU to which information society policy is a
response, and second of a set of analyses of these problems from
which appropriate policy solutions flow.
The definition of the policy problem has very old roots in EU policy
thinking and lies at the heart of the Lisbon goal: it is the Colbertism
underlying Lorentz’s presentation; it is that we are in competition
with the US and Japan in a zero-sum economic game. Since Japan has
temporarily dropped out of the race this is the Defi Americain all
over again. In the old days we had to have a European computer industry
or an aerospace industry. Now we have to have European broadband
penetration and e-business at the same levels as or better than
those of the US. Let us just note in passing, the fundamental argument
of economists that while firms compete on global markets, nations
and regions do not. And the fact that at the same time as the EC
was, within this analysis, constructing Esprit, Framework and related
programmes to compete with a supposedly superior US, the European
economies were consistently out-performing the US.
In short, while in the nineteenth and early twentieth centuries
the great powers competed by building bigger and better fleets,
and during the cold war bigger and better missiles, the new national
virility symbols are broadband networks and e-business start-ups.
Progress of a sort I suppose.
The crucial point to be made at this juncture is that in so far
as there has been in recent years a widening of the GDP per capita
gap this has little if anything to do with ICTs or a knowledge economy,
or with a superior endogenous competitive environment, á
la Michael Porter, but is very largely due to an increase in the
working age population through increased rates of immigration.
In my judgement the economic problem is quite other and it is shared
by the US and Europe. In spite of the euphoria associated with the
telecommunication and Internet booms and with claimed increased
sustainable rates of productivity growth associated with ICTs, which
is one of the core components of the consensual vision, the real
underlying problem is the stagnation of the major industrial economies
and the accompanying three decades of decline in corporate profitability.
This is due to saturated markets, the completion of the sectoral
consolidation process on a global scale and, thus, the end of returns
to economies of scale in mature markets and crucially to the shift
in the centre of economic gravity to the service sector.
This is central to arguments about the information society because
it can be seen as a re-labelling of the service or post-industrial
societies that were the favoured terms in the 1960s, ‘70s
and ‘80s, and because one version of information society theory
saw ICTs as the productivity-enhancing machines of the service sector,
and service workers have been renamed knowledge workers without
any change in their functions. Some of you probably remember how
in the 1980s within the OECD, countries measured their percentage
of knowledge or information workers instead of broadband or e-business
penetration as tokens of success or failure in a race once again
to catch up with the US. Plus ça change.
The whole liberalisation, de-regulation movement can be seen as
a response to the problem of service sector productivity and especially
of that significant part that lay and still lies in the public sector.
Now what is significant in recent years is that whatever gains in
productivity growth rates have been achieved have been not in the
service sector, nor even generally in the manufacturing sector,
but in the manufacture of the chips themselves.
In my view one of the major weaknesses of current information society
research and policy is a misunderstanding of the potential relationship
between ICTs and services and thus an exaggeration of the contribution
they could ever make to enhancing either productivity or service
quality.
As with the original telecommunications policy the consensual vision
of the information society, while in using a common label it enables
people to believe they share a common vision, in fact contains within
it a number of distinct visions or strands of analysis leading to
distinct policy interventions, each of which needs to be judged
on its own terms rather than taken as a single package. This is
important for us because these differing visions have different
views of the role of ICTs – if any – within the information
society. Indeed it is important to stress the ‘if any’
because the ICT and telecommunications industries and their associated
policy makers have succeeded in propagating the erroneous view that
the information society is primarily about them.
The core of the information society vision, derived mainly from
Daniel Bell, can be simply stated. Those capitalist economies are
shifting from fixed to human capital as the key source of value
added and economic growth. There are, however, different versions
of this vision.
The Bell version places the emphasis on the contribution of trained
scientific labour power in the process of production. Here the ICT
and pharmaceutical industries are key cases. ICT growth is the result
of this process, but not its cause. The policy problems are the
production and efficient deployment of highly skilled labour power.
The benchmarks are numbers of university graduates and R&D intensity.
1. It should not however be confused with the more general knowledge-
or symbolic-worker vision. This holds that an ever-increasing proportion
of employment involves not working with machines to manipulate matter
– the classic industrial model, but in creating, manipulating
and distributing information. This is also sometimes dubbed the
immateriality thesis. The problem is that this takes different forms
with different economic implications, and different relations to
ICTs. On the one hand it describes the growth of the service sector
where humans are dealing directly with other humans and the value
exchanged is embedded directly in human labour. Assessment of the
impact of ICTs on either productivity, quality or market growth
in services remains highly controversial. On the other hand it describes
the bureaucratic overhead that results from growth in market size
and complexity. It represents the costs of coordination. Thus its
increase is not a sign of economic efficiency. It is here that the
long held out, but continually dashed, hopes for the impact of ICT
investment on productivity largely lie. It is worth pointing out
that one of the sources of the information society lies in Japan’s
response in the early 1970s to the problem of its dependency on
imported oil. The purpose of the information society was to create
an economy that was no longer producing and distributing things,
but was producing symbols. This also in their view had environmental
advantages. Against this background it is worth noting that the
US, supposedly the world’s leading information society, is
so dependent on increasing oil consumption that it cannot even ratify
the Kyoto protocol.
2. A variant of the immateriality vision is the death of distance.
This holds that the move from atoms to bits, allied to the rapidly
falling cost of high capacity digital networks, removes the transport
cost barriers to market size that firms, particularly service firms,
can efficiently serve. This is true, but can be exaggerated and
needs analysing market by market, service by service.
3. The symbolic economy/content vision assumes the growth of the
media sector and the positive impact of so-called new media. This
version of the vision is really just the leisure society thesis
revisited and does not stand up well to even cursory empirical investigation.
The successful communication developments have been those that sell
connectivity, with people supplying their own content. The media
content sector has grown in nominal terms, but this is largely due
to a relative price effect Consumption time has not increased, indeed
with rising wealth it tends to decline.
4. Finally, we have the Schumpeterian innovation vision. Here innovations
may or may not be knowledge intensive. ICTs are simply one among
many cases of innovation. Information in this model is much more
about creative cluster and their modes of intercommunication, and
about the interaction between information and markets in terms of
innovation choice and risk. The policy focus here is the context
for entrepreneurial activity and the creation of innovation clusters.
The policy dilemma is the tension in the innovation process between
sharing and property rights and capturing returns, exemplified,
above all, in the intellectual property area.
In conclusion the key division for policy debate is between those
who focus on ICT and communication as a new growth sector in its
own right and thus what can be done, if anything, by public bodies
to optimise that growth and capture those markets, and those who
see communication networks and services as the essential infrastructural
underpinning beneficial economic developments elsewhere. From this
perspective the old questions of the appropriate role of public
finance and management in infrastructural provision have not gone
away. In particular, the current policy field is riven by contradiction
between liberalisation/market competition and public intervention/regulation,
between technological neutrality and industrial and research policies.
In short the problems with information society policy and related
research are firstly that they are based on a faulty analysis of
the underlying difficulties facing the EU in terms of competition
with the US. Second, even if we accept the underlying analysis and
the Lisbon goal there is no evidence that we in fact have the policy
instruments that might produce the desired result. Thirdly, this
is in part because the theories and models underlying policy formulation
and implementation are much more controversial, partial, doubtful
and contradictory than either theorists or policy makers are prepared
to admit. And this is in large part, as Bauer has stressed, because
what the world theorists are trying to understand and policy makers
to steer is inherently more complex than either researchers or policy
makers feel comfortable in admitting to.
This then leads me to my final conclusion that we must take all
grand visions, plans and theories with a very large pinch of salt.
This is not, I must stress, an argument against public intervention
and in favour of letting the market rip, although versions of it
have been used as such. Market messianism is subject to the same
critique. We know that markets are of different types, are messy
mixtures of private actions and public rules and institutions and
have extremely unpredictable results.
Now the message is that both researchers and policy makers need
to be much more humble. We cannot, I think, avoid policies and regulatory
interventions. No human society or social group can be unplanned
in that extreme sense. The best we can hope for is messy, short
term interventions in specific areas to solve specific problems.
There are no general answers or rules and the results of intervention
are likely to be very different from what was planned.
|
