Corporate governance, public governance and global governance: the common
thread
Diane Coyle1
Working Paper, Institute of Political and Economic Governance, University of
Manchester, http://www.ipeg.org.uk
December 2003.
The question of governance has catapulted to prominence in recent years. A
search of news databases for references to ‘governance’ shows the number
of uses of the word has been on an upward trend through the late 1990s but
has exploded during the past 12 months or so [FIG 1]. However, the
governance question arises in a number of different contexts: it is applied to
corporations, governments and international agencies. The aim of this paper
is to explore the links – if any – between corporate, public and global
governance.
In each context the growth of interest stems from the sense that there has
been a governance crisis. This is perhaps clearest in the case of corporate
governance where the timing of the preoccupation with governance questions
in public debate is related to a series of high-profile corporate scandals.
The most notorious are the huge US bankruptcies related to alleged frauds
and undoubted executive greed – such as Enron (bankruptcy filing December
2001), Tyco (CEO indicted Jun 2002), WorldCom (bankruptcy filing July
2002) and Qwest (accounts restated by $1bn July 2002). However, the
corporate misbehaviour and mismanagement of the late 1990s have not been
confined to the US, as examples such as Philip Holzmann in Germany
(bankruptcy filing March 2002), Vivendi in France (chief executive forced
out July 2002), and the numerous corporate and financial bankruptcies in
Japan and Korea since the late 1990s demonstrate. Equally, there has been a
widespread regulatory reaction to such events, including the Sarbanes-Oxley
Act in the US and the Turnbull, Myners and Higgs reviews, whose
1
Enlightenment Economics and IPEG, University of Manchester. I am grateful to the
participants in a seminar at the University of Manchester on 22 October 2003 for their
comments. All errors and infelicities are my own. Further comments welcome at
Diane@enlightenmenteconomics.com
Diane Coyle
Page 1
3/3/2016
implementation is self-regulatory, alongside new corporate legislation in the
UK.
In the case of global governance, the sense of crisis dates back earlier to
the global financial crisis of late 1997-98 and the failure of the WTO talks
in Seattle in December 1999. Initially the debate subsequent to these
events was framed in terms of reform of the international financial
architecture, but the same debate is now more commonly described in terms
of governance: international governance on the one hand; and on the other,
good governance within individual countries as an aspect of their potential
within the global economy. All international financial institutions and
agencies now proclaim their own governance and governance within their
member countries to be a priority, and have put considerable institutional
efforts and resources into codifying and measuring aspects of good
governance. The emphasis on governance has official manifestations ranging
from the adoption by the Organisation for Economic Co-operation and
Development of a convention outlawing bribery of public officials by
businesses to a programme of research on the economic effects of
corruption at the World Bank and International Monetary Fund.2
It is harder to pinpoint any particular trigger for the interest in governance
in the public sector, but there is nevertheless a vigorous debate on public
sector reform in many countries, not just the UK, where improving public
services was brought to the fore as a policy priority of the New Labour
government elected in 1997. In the United States a prominent commission on
public service reform chaired by Paul Volcker reported in January 2003 3,
and there is a continuing active state and local level debate on education
reform. In France controversy has recently centred on education, with a
rapidly increasing proportion of families opting for private schools, and also
lately on health given the extraordinary level of excess deaths amongst the
elderly during the summer 2003 heatwave. In Germany the government of
Gerhard Schröder launched an ambitious programme of reform explicitly
calling into question some of the assumptions of the country’s welfare state,
under the heading Agenda 2010, in March 2003. In each case the
2
See http://www.oecd.org/dataoecd/50/33/1827022.pdf and George Abed and Sanjeev
Gupta, Governance, Corruption and Economic Performance, IMF, 23 September 2002,
details at http://www.imf.org/external/pubs/nft/2002/govern/index.htm
3
Available at http://www.brook.edu/dybdocroot/gs/cps/volcker/volcker_hp.htm
Diane Coyle
Page 2
3/3/2016
presumption that governments are not matching up to what people want
appears to have gained fresh political salience.
Of course, perceptions of a crisis as filtered through media reports do not
amount to sufficient evidence of the existence of a governance crisis in any
of these contexts, even if there are policy adjustments taking place as a
result. However, the idea of a crisis of governance, interpreted broadly, is
also consistent with widespread polling evidence indicating a loss of public
trust in many of the institutions concerned, from big corporations to
government agencies and politicians. According to a Mori poll published in
April 20034, “while trust in individual professions has generally remained
static, trust in institutions has declined, in some cases quite significantly.”
There had been significant declines in trust in politicians and government,
big companies, the police and the judiciary, journalists and local authorities –
but not much decline in trust in professionals such as doctors and teachers according to Mori.
Polls in Canada, Britain, the US, Italy, Spain, Belgium, the Netherlands,
Norway, Sweden, and Ireland have produced similar results. In the United
States, the University of Michigan’s index of trust in the federal
government has fallen from 52% percent in 1964 to a low of 26% in 1994
before recovering to 34% in 1998.5 A survey (covering 46 countries) by
Gallup International (on behalf of the World Economic Forum)6 in late 2002
reported low levels of trust in many institutions, notably governments and
politicians, large companies and international organisations such as the WTO.
(The armed forces, non-governmental organisations and educators were
amongst the most trusted in this survey.)
The question is why failures of governance are a hot issue in so many
countries and so many contexts at this time (not that there need be one
single cause). After all, the range of actors involved is immense – business
executives, public officials, international officials, politicians - in a wide
http://www.mori.com/pubinfo/rd-trust.shtml
http://www.umich.edu/~nes/nesguide/toptable/tab5a_5.htm. The index bounced to 43% in
2002, in common with many similar post-9/11 polls... The National Election Studies index of
efficacy shows a more pronounced decline from 74% in 1960 to a low of 33% in 1994, before
showing the same pattern of modest recovery then post 9/11 bounce. See
http://www.umich.edu/~nes/nesguide/toptable/tab5b_4.htm
6
http://www.weforum.org/pdf/AM_2003/Survey.pdf
4
5
Diane Coyle
Page 3
3/3/2016
range of institutions and organisations around the world. It strains credulity
to suppose that all of these organisations have become, coincidentally, less
well managed since the late 1990s, and more natural to hypothesise that the
context in which all organisations are operating has changed in
systematically important ways during the past decade or so.
My aim is to explore what that systematic change might be. I suggest it is
related to the dramatic fall in the costs of processing and communicating
information, due to the radical improvements in information and
communication technologies. At the most straightforward level, the
increased availability of information and greater ease of monitoring
institutions must have contributed to the loss of deference noted by so
many observers. In other words, it makes it easier to see the warts. Access
to information has reduced information asymmetries in many areas of public
service; and those where there remain substantial asymmetries such as
medicine and education are in many countries exactly the areas where public
trust has declined the least (see the poll evidence cited above).
At the same time, public expectations of levels of service have increased –
this is one of the contributing factors highlighted by the Mori report. In
addition, I argue here that the optimal boundaries and internal structures of
organisations – whether private or public sector – can be expected to change
in response to a reduction in information costs and to any thinning of the
information fog in which we all operate. This is a hypothesis which has been
raised in the context of the use of ICTs by businesses, and there is a small
amount of mainly case study evidence to support it. However, the same
argument applies to other kinds of organisation including public sector and
international bureaucracies.
I.
Organisations of all kinds are entities which get to the places markets
cannot reach. This includes private sector companies and government
department or agencies. Writers from J.M.Keynes to Daniel Bell have
recognised the similarities between large organisations in the private and
public sectors. Keynes classed certainbig companies as part of the public
sector even in a 1924 lecture, well before many of them were nationalised.7
Bell writes:
7
See Robert Skidelsky, John Maynard Keyes: The Ecconomist as Saviour 1920-37, pp227ff.
Diane Coyle
Page 4
3/3/2016
“A business corporation, like a university or a government agency or a
large hospital – each with its hierarchy and status system – is now a
lifetime experience for many of its members. .... if we set up a
continuum, with economizing at one end of the scale (in which all
aspects of the corporation are single-mindedly reduced to becoming
means to the goals of production and profit) and sociologizing at the
other (in which workers are guaranteed lifetime jobs and the
satisfaction of the workforce becomes the primary levy on resources)
then in the last thirty years the corporation has been moving steadily,
for almost all its employees, towards the sociolgizing end of the
scale.” 8
In economic theory the study of the boundary and form of the firm dates
back to the famous essay by Coase9, in which he explained the existence and
scope of companies in terms of the costs of transacting in a world of
incomplete information. If the transaction costs of market exchange are too
high, the transactions will take place within the boundaries of an
organisation instead. More recently the New Institutional Economics10 has
once again made transactions costs and informational asymmetries central to
the theory of the firm (and other types of organization). So when might
market transactions be too costly?
Most economic theory has concentrated on the impossibility of drawing up
complete contracts to cover all contingencies – for example, if a supplier is
required to make a costly investment on behalf of a customer but cannot be
sure that the customer will not try to renegotiate or walk away if
circumstances later change, then vertical integration is likely. However,
industries where this situation arises, such as automobile production, display
very different structures in different countries, so other forms of
transaction cost are also clearly relevant.
8
‘The Coming of Post-Industrial Society’, (1974), p294.
Ronald Coase, ‘The Nature of the Firm’, Economica, 4, November 1937, 386-405.
10
See for example Oliver Williamson, The Economic Institutions of Capitalism (1985)
9
Diane Coyle
Page 5
3/3/2016
These could include agency problems, difficulties in transferring complex or
tacit knowledge, benefits of detailed market monitoring and so on11. The
transactions involved might rely too much on information that is hard to
codify and capture in a price. It might be too difficult to monitor effort or
compliance because of inadequate or asymmetric information. There might
be incompatible incentives or moral hazard. There could be significant
externalities or free rider problems. Or the amount and granularity of the
information involved might just be too high, leading to prohibitive costs of
co-ordination.
So organisational structure will vary depending on the nature and size of the
transactions costs in each specific context. Thus if detailed local market
information becomes more valuable or less costly, greater delegation or even
spin-offs might occur. Innovative technologies can also change optimal firm
structure. Holmström and Roberts write:
“Information and knowledge are at the heart of organizational design,
because they result in contractual and incentive problems that
challenge both markets and firms. Indeed, information and knowledge
have long been understood to be different from goods and assets
commonly traded in markets ........ We think that knowledge transfers
are a very common driver of mergers and acquisitions and of
horizontal expansion of firms generally, particularly at times when new
technologies are developing or when learning about new markets,
technologies or management systems is occurring.”
The typical form of both private and public sector organisations indeed
changes dramatically over long periods of time. The half-century after the
second world war saw the rise to dominance of the hierarchical, centralised
form in large-scale units. Industrial concentration ratios rose in many
countries through to the late 1970s. In the private sector this ‘Fordist’
model was associated with increasing concentration due to high capital
requirements and economies of scale in the mass market. The same
structure was paralleled in the public sector, both in nationalised industries
and also in the organizational form of administration and public services.
Bengt Holmstrom and John Roberts, ‘The Boundaries of the Firm Revisited’, Journal of
Economic Perspectives, 12(4) Fall 1998, 73-94; Kenneth Arrow, ‘Vertical Integration and
Communication’, Bell Journal of Economics, 6, 1975, 173-182
11
Diane Coyle
Page 6
3/3/2016
Some economists have linked the hierarchical and bureaucratic form to a
cluster of early 20th century technologies such as rail, telegraph and
electricity which made operations on a large scale both possible and
profitable.12
However, the hierarchical, centralised organisation is also a product of high
information costs – at least relative to the benefits of transferring
knowledge. Hierarchies can minimise communications costs because they
minimise the number of links required to connect multiple individuals,
compared to more decentralised structures.13 Hub-and-spoke transport
networks are analgous. Similarly, standardised products in long production
runs offer an efficient way to use an inflexible and scale-intensive
manufacturing technology, so greater flexibility thanks to computerisation
allows flexible production.14
It is reasonable to expect that lower information costs or higher benefits
from the sharing of knowledge because of innovation will lead to an optimal
organisational structure that is flatter and more decentralised – exactly the
kinds of change that have characterised many private sector companies
since around 1990. Best15 identifies an earlier shift away from the classic
Fordist model of production to a more complex multi-product ‘just-in-time’
(but still vertically organised) production system pioneered by Toyota,
followed by a more recent shift in some US industries to a network model of
horizontal inter-firm organisation. He too links this shift to technology
transfer: “With open systems the American PC industry has developed a
networking model of inter-firm organization that has demonstrated an
unprecedented rate of technological innovation and diffusion.”16 In
management-speak, these kinds of change towards flatter companies and
horizontal organisational forms are variously described as outsourcing, reengineering, delayering and networking.
For example, Chandler, The Visible Hand, 1977.
See for example Thomas Malone, ‘Modelling Co-ordination in Organizations and Markets’,
Management Science, 33(10), 1987, pp1317-1332; and Thomas Malone and John Rockart,
‘Computers, Networks and the Corporation’, Scientific American 265(3), 1991, 128-136.
14
Paul Milgrom and John Roberts, Economic Organization and Management, Prentice Hall
1992.
15
Michael Best, The New Competitive Advantage: The renewal of American Industry,
Oxford University Press 2001.
16
Ibid p53
12
13
Diane Coyle
Page 7
3/3/2016
There has clearly been less change in the public sector, where the absence
of pressure applied by, say, disappointing profits or share prices means
organisational change is very much harder to achieve. Nevertheless, the
rhetorical themes and patterns of change are similar – decentralisation and
devolution have been a common theme in many countries including the UK and
France, as has the delegation of authority to separate agencies (such as the
outsourcing of interest rate decisions to the Bank of England), and the
trimming of some layers of bureaucracy. In the case of the international
institutions too there has been an emphasis on delegation and delayering,
whether the EU’s principle of subsidiarity or the IMF’s conversion (in
principle) to encouraging its borrowers to draw up their own poverty
reduction and economic adjustment plans.
II.
It is worth giving an indication here of the scale of the decline in
information and communication costs during the recent past. For the pace of
technical progress in ICT, and the consequent decline in costs, is without
historical precedent. The figures are simply extraordinary.
Moore’s Law is well-known. In 1965 Gordon Moore of Intel predicted the
processing power of computer chips would double roughly every two years –
later accelerated to every 18 months. The current technological platform of
silicon-based semiconductors is predicted to allow Moore’s Law to hold until
at least 2015, beyond which horizon new technological platforms are
expected to sustain or increase the pace. [FIG 2] The increase in computing
power has been so spectacular that it is usually shown on a log-linear scale.
Here it is presented on a linear scale to emphasise timing of the exponential
‘take-off’ in computational power, in the late 1990s, as the timing is relevant
to the argument here.
There has been a correspondingly large decrease in the price of
computational power, of the order of 30-40% a year for the past 30 years,
or a 99.99%+ decline in three decades. A cent’s worth of computer power
today would have cost $10,000 (in today’s money) in 1970. The prices of
phone calls and internet connections have been declining at almost equally
rapid rates. These numbers are very large, in themselves and by comparison
to earlier innovative technologies. For example, the real price of illumination
Diane Coyle
Page 8
3/3/2016
is estimated to have fallen from 40 cents to 0.1 cent per thousand lumen
hours between 1800 and 1990, a modest decline of 99.75% in 190 years
(rather than less than 30 years in the case of computers) and by 7-8% a
year during the 30 years of peak decline, by comparison.17 Car prices fell
12% a year in real terms during the peak decades of innovation in
automobiles from 1900-1930. The bottom line is that the acquisition and
exchange of any information that can be codified (in words or pictures) has
the potential to become virtually free to almost everybody, almost
everywhere.
As a result many indicators point to the extremely rapid diffusion of the
relevant new technologies since 1995. This is the year in which web browsers
becamse commercially available. It also marks the point at which the full
exponential power of Moore’s Law took hold, putting cheap and powerful PCs,
low-cost phone calls and mobile telephones within reach of very many more
pockets.
So, for example, the proportion of UK households with internet access at
home has climbed from zero in 1994 to 14% in 1999 and 48% on the latest
official figures (2003Q2). [FIG 3] Other indicators of rapid diffusion of
ICTs include the number of web hosts [FIG 4], and the rising share of ICT
investment in GDP [FIG 5]. The rapid diffusion of these technologies has
not been confined to the developed countries. Two have made particular
strides in the developing world – satellite TV and mobile telephony. The
number of mobile subscribers now exceeds the number of fixed telephone
line subscribers in about 100 countries. [FIG 6]
While in many cases the use of the technology had increased prior to the
mid-1990s – and some of the basic technological steps date back to around
1970 – there does in many cases seem to have been a marked acceleration
within the past decade. What’s more, the diffusion and use of the
technological potential will depend on whether social structures change to
accommodate it – after all, about a third of the world’s population still has no
access to electricity nearly a century after almost every household and
William D. Nordhaus, ‘Do real output and real wage measures capture reality? The history
of lightsuggests not,’ Yale University, Cowles Foundation Discussion Paper No. 1078,
September 1994.
17
Diane Coyle
Page 9
3/3/2016
every workplace in the leading economies was connected to the grid.
Institutions matter as well as technological possibilities.
III.
In this context it is worth asking whether the sense of crisis in governance
in such a wide range of institutions could be linked in any way to the new
pervasiveness of low-cost information and communication. It is technically
and economically feasible for almost anybody in an organisation to acquire
and use vast quantities of information – although not necessarily
organisationally feasible.
For as well as creating internal pressures to reorganise in order to capture
efficiency gains, the new technologies also put pressure on organisational
forms through their impact on consumers or users. Whether willingly or not,
organisations are becoming more transparent. Internal information, and the
internal use of external information, have become more readily available to
customers or users as well as employees and suppliers.
In the case of businesses, the radically reduced cost of processing and
communicating information – alongside continuing innovations in
manufacturing technology which are reducing the unit cost of short
production runs - has created the scope for innovation at ever-finer
aggregations of demand for their products and services. In other words
greater customisation is leading to an explosion in choice. Most of the
product differentiation in the new goods now takes the form of intangible
attributes or service-like qualities – these are where value is added.
After all, nobody ever said “quantity is the spice of life.” When our basic
needs are met, we seek to meet additional wants in individually varied ways.
[FIG 7] Henry Ford famously said of the Model T: “The customer can have
any colour he wants so long as it’s black.” The choice of new vehicle models
in US is now nearly 300; Ford offers 46 colours. The paradigm now isn’t Ford
but Dell. Michael Dell says: “Companies that are successful today are those
that can get closest to their customers’ needs.” It gives customers ordering
online 16 million theoretical possible combinations of specifications for a
desktop PC.
Diane Coyle
Page 10
3/3/2016
Conventional measures of economic welfare such as GDP almost certainly
underestimate the welfare benefits of greater choice18. To give a recent
example, Brynjolfsson et al19 estimate that the benefits of the increased
choice due to online book sales (2.3m titles on amazon.com compared to
250,000 in the Barnes and Noble superstore in New York City) are roughly
twice as large as the gains due to reduced book prices. Only the latter
effect is conventionally measured as a real income gain, through the
deflation of nominal GDP.
With only the slow and indirect mechanism of electoral change exerting
pressure, it’s no surprise that the public sector is lagging behind leading
private sector companies in responding to increasingly fragmented demands
or in exploiting the potential of the technologies for efficiency gains. 20 The
public sctor is offering services that are considerably more complex than
was the case two or three decades ago, so responding to technological
potential is not easy. It is not made any easier by the institutional path
dependence of public sector organisations – given their history, it is
genuinely hard to see how they could move from where they are to where
they might optimally be. At the same time, it is hard to see the public sector
being able to avoid organisational change when, as private consumers,
citizens are experiencing greater choice (including choice over quality) in
everything they buy. Public providers of services – especially those which are
also provided competitively by private sector organisations, which includes
health, education, transportation – are lagging behind in the innovation arms
race which is characteristic of modern capitalism. 21 People will most likely
not be satisfied with Model T Ford choices in public services when they are
increasingly getting Dell or amazon choices in their purchases of private
services and products.
It is important to stress that this is not an argument in support of the
privatisation of public services – it says nothing about the location of the
public-private boundary. Nor does it have any implications for the
See Nordhaus, cited above. See also Diane Coyle, Paradoxes of Prosperity, chapter one.
Brynjolfsson, Erik, Michael D Smith & Yu Hu, ‘Consumer Surplus in the Digital Economy:
Estimating the Value of Increased Product Variety at Online Booksellers,’ MIT Working
Paper, April 2003, http://ebusiness.mit.edu/erik/CSDE%202003-04-061.pdf
20
I do not want to exaggerate the sharpness of the public-private boundary in drawing this
distinciton, however.
21
William J Baumol, The Free Market Innovation Machine, Princeton University Press 2002.
18
19
Diane Coyle
Page 11
3/3/2016
universality of public services. It touches, rather, on the organisation of the
public sector, the degree of vertical integration in the provision of services
and the extent to which users are given choice and control over their
experience of public services. To clarify the distinction with an example, I
am not arguing that the parental desire for variety means they should have
any choice over the school their children attend; on the contrary, there are
important social and economic reasons for encouraging more children to
attend their local ‘bog standard’ school. But I am arguing that the school
system should be reorganised to move away from batch-processing pupils in
groups of 25 or 30 in accordance with a standardised curriculum. Instead
they should give individual children (and parents) far greater choice over the
kinds of options they select within school, over timetabling, over styles of
learning, more top-up options outside the school and so on. For my argument
is that the contrast between the range of options available in the public and
private sectors could be an important source of users’ disappointment with
public services.
There is obviously no exact parallel in the global context to the pressure of
cheap information on organisational form. However, it is reasonable to think
that the new availability of information is nevertheless causing tensions and
pressures in terms of global governance. “In earlier eras, a culture was
transmitted across national boundaries by migration, travel and reading.
Since leisure travel and literacy were often limited to the rich, the
understanding – and exploitation – of other cultures was often enjoyed only
by elites. Television and the post-1970s media, along with cheaper and more
rapid transportation via jet airplanes, changed all that. Culture could move
with nearly the speed of sound and reach billions of people.”22
Looking at the jump in global access to television from the mid-1990s and
mobile telephony from the late 1990s, transmitting to hundreds of millions
or billions of people hitherto unavailable information – on any subject from
prices in nearby urban markets, practical know-how from networks of
overseas migrants through to images of lifestyles to which viewers could not
previously have aspired because they didn’t know about them - we should not
underestimate the likely impact on culture, politics and business. There is
relatively little authoritative research on the social and economic impact on
developing countries of television and telephony in particular (as opposed to
22
Walter LaFeber, Michael Jordan and the New Global Capitalism, Norton 2000, p18.
Diane Coyle
Page 12
3/3/2016
ICT in general, which has generated a far more substantial research
literature). However, it seems likely that the impact will be profound.
IV
These arguments might seem plausible but they raise the obvious question of
whether there is in fact any evidence that reduced information costs are
changing organisational structures in any significant ways. There have been
surprisingly few studies trying to assess this question, even though the
theory of the firm rests to a large degree on hypotheses about transactions
and information costs, and no research that I am aware of looking at public
sector organisations as opposed to businesses.
Yet there is some interesting, if patchy, evidence on the corporate front.
There have been some case studies, for example, mainly in the US. They
suggest that the scope for reorganisation afforded by new technologies has
allowed some companies to harness their investments in ICT to raise
productivity levels by as much as 25-30%.23 But their experience has
demonstrated that improved efficiency is not possible without organisational
change. Brynjolfsson and Hitt have argued that the complementary
investments in internal organisation, supplier networks and production
processes dwarf recorded spending on ICTs.24 They cite many studies which
document the substantial changes in work practices needed to make
efficient use of the new technologies. Bresnahan, Brynjolfsson and Hitt,
surveying about 400 large firms, found that higher IT spending is associated
with increased delegation of authority, and a more highly educated and
skilled workforce.25 After all, individual employees cannot make use of
almost freely available information and flexibility in production if no
decision-making authority has been delegated to them. If key decisions still
have to be referred up a managerial hierarchy, then the economic benefits
of fast access to very fine-grained information (about customer preferences
or cost differentials or logistical hiccups, for instance) cannot be realised.
See Erik Brynjolfsson and Lorin Hitt(1996 and 2000a) , McKinseys (2001).
Erik Brynjolfsson and Lorin Hitt, ‘Beyond Computation: Information Technology and
Organisational Transformation’, Journal of Economic Perspectives, 14(4), Fall 2000, 23-48.
25
Timothy Bresnahan, Erik Brynjolfsson and Loren Hitt, ‘IT, Workplace Organisation and
the Demand for Skilled Labor: A firm-level analyisis,’ MIT working paper, 2000.
23
24
Diane Coyle
Page 13
3/3/2016
In addition to this literature, which explores the impact of ICTs on
productivity at the firm level, and reports en passant that productivity gains
are linked to the reorganisation of the firm, some studies have looked
specifically at the effects of the technologies on firm structure. [Motohashi
(2001) reports that firms which have invested more in their computer
networks tend to outsource more. More details].26 The results are therefore
consistent in finding that the trend in business is towards decentralisation
and reduced vertical integration.
Another kind of evidence for the incentives to reorganisation offered by
the new information technologies is simply the massive mergers and
acquisitions boom of the late 1990s [FIG 8]. M&A waves are pro-cyclical but
the merger boom of the late 90s clearly swamps the previous boom of the
late 80s in magnitude (itself the biggest, relative to relevant GDP measures,
to that date). The number of acquisitions in a wide range of US and EU
industries, which dominate the global figures, point to a large-scale
reorganisation of corporate structures in the leading economies.
Another possible indicator is tentative evidence for the polarisation in the
size of companies in several countries. We might expect the technologies to
hollow out the size distribution of companies. Earlier this paper argued that
cheap and pervasive information would encourage local or decentralised
decision-making. This would suggest a shift towards smaller sized firms. On
the other hand, there are vast economies of scale in many leading industries
either because of high fixed capital requirements (autos, aerospace) or high
R&D costs (pharmaceuticals, software) or increasingly network effects
(software, telecoms) and winner-takes-all effects in demand (cinema, music).
Thus there are merits in being very large or in being very small but no
compelling advantage in being medium-sized.
The size distribution of companies (measured by market capitalisation or
number of employees) has in fact been found to follow a power law, with
various coefficients. Thus there are few very large companies and more
medim sized ones and many more small ones: firm sizes follow a Zipf
distribution.27 Thus in the UK, 0.4% of firms have 100 or more employees,
1.7% have 20-99 employees, 8.5% have 5-19, 39.4% 1-4 and 56.2% are sole
26
27
OECD (2003), Hitt (1999), Motohashi (2001).
Axtell (1999), Ijiri and Simon (1977).
Diane Coyle
Page 14
3/3/2016
traders.28 The proportions differ somewhat for other countries but in
general the pattern is the same.
This pattern occurs in several kinds of economic data, including city sizes,
and domestic income distributions. However, in the case of the global income
distribution, there is evidence that the power law relation holds only in the
middle of the range, between the 30th and 80th centiles, and there are more
very rich and very poor countries than implied by a power law. 29 This is
equivalent to the now well-known observation that there are ‘convergence
clubs’, or in other words that individual countries tend to converge to either
a high or low level of GDP per capita.30
There are figures – although not at all comprehensive or decisive suggesting that both very small firms and very large ones are becoming
increasingly important, at least in the US and UK economies. Researchers at
the OECD, using industry surveys for three benchmark years (1967, 1977
and 90) and five countries, report that the average size of companies by
number of employees has fallen sharply in the US and UK and somewhat less
dramatically in France and Germany. Only in Japan was there no increase in
the employment share of the smallest companies or decrease in the
employment share of the largest companies.
At the same time, the gap by which average labour productivity of the
largest firms exceeds that of the smallest has risen over time. The
consequence is that the output shares of the biggest firms have tended to
rise over time. 31 In addition, the OECD data set covers establishments
rather than enterprises for the US, so cannot distinguish a move to more
establishments within one firm from a move towards smaller firm size.
Separate research for the US indicates that the average size of firm and
the relative importance (in terms of employment or value added) of the
largest firms increased during the 1980s and 90s – alongside a decrease in
28
DTI data
Di Guilmi et al (2003)
30
Quah (1996), Baumol (1986)
31
Bart van Ark and Erik Monnikhof, ‘Size Distribution of output and employment: a data set
for five OECD countries 1960s-1990’, Economics Department Working Paper no 166,
OECD1996.
29
Diane Coyle
Page 15
3/3/2016
average industry concentration ratios due to a simultaneous rise in the
importance of very small firms.32
This evidence hinting that very large companies and very small ones may be
becoming more important in their contributions to output, relative to the
past, suggests convergence clubs could be developing in this context too.
(Axtell’s orginal model explained the decreasing numbers of firms as firm
size increases in terms of the increased incentive for individuals within
firms to free ride as their own effort becomes less and less important to
the firm as a whole. Free riding becomes harder the easier it is to monitor
individual effort, which might well occur with a reduction in information
costs. So an increase in the relative importance of large firms is consistent
with this model, as each individual’s reward could become more sensitive to
the individual level of effort. )
To return to the main argument, a further area of relevant evidence for the
impact of ICT on firm organisation is a decline in the longevity of companies.
Recent technical change could be expected to decrease corporate life
expectancy. Value-added is increasingly reliant on intangible assets and
investments, such as expenditure on R&D or marketing, brand values,
management know-how and employee skills.33 The result is that almost all
the increase in real GDP in both the US and UK during the 1990s was
‘weightless’, that is literally involved no increase in the physical mass of
economic output.34 As the US Federal Reserve Chairman Alan Greenspan
explained in a recent speech: “A firm is inherently fragile if its value added
emanates more from conceptual as distinct from physical assets......Trust and
reputation can vanish overnight. A factory cannot.” Thus we might expect a
step up in the corporate death rate and decreased average longevity.
Long-lived companies have always been rare: of the original 1917 Forbes 100
corporations, 61 had ceased to exist by 1987 and another 21 had fallen out
of the list of the 100 biggest US companies. (The 18 survivors included Ford
and General Motors, GE, Kodak and DuPont.) But the average company
‘What’s been happening to aggregate concentration in the United States (and should we
care?)’, Lawrence J White, Stern Business School working paper, 2001.
33
Diane Coyle, The Weightless World, Capstone 1996/MIT Press 1997.
34
Sheerin, Caroline, ‘UK Material Flow Accounting’, Economic Trends no. 583 June 2002,
Office for National Statistics.
32
Diane Coyle
Page 16
3/3/2016
lifetime has fallen substantially. In the United States, Richard Foster and
Sarah Kaplan calculated a time series for the average lifetime of companies
listed in the S&P500 index, finding that it moved pro-cyclically (because
recessions lead to a higher corporate death rate) but also dsiplayed a strong
downward trend from 75 years in 1939 and 25-35 years through the 60s,
70s and 80s to a little over 15 years in 2000.35 Arie de Geus36 calculated
that the average life of all EU companies in the mid-1990s (ie including nonlisted small companies) was 12.5 years. Although he did not calculate past
life expectancies, this figure for the late 1990s is certainly lower than one
might have expected.37 With a lifespan likely to be only 15 years, it is not
surprising to find rooted hierarchies vanishing from many private businesses.
In the case of governments, it could be argued that reputation based on
intangibles is in fact the only relevant source of value. While the issue of
longevity does not arise in the case of public sector institutions in modern
democracies, the importance of reputational factors in today’s 24-hour
media culture could certainly help account for the loss of trust in public
bureucracies, both national and international.
V.
The evidence cited here has covered the corporate sector but a shift in the
boundary between the transactions that can and can’t occur in the market,
mediated by price, is fundamental to understanding the impact of ICT on
organisational form and why this should raise questions of governance across
the board. On the one hand, the greater availability of information means
more transactions can take place in markets because it is easier to monitor
and enforce contracts – hence phenomena such as outsourcing, and the
reallocation of production between nations which is at the heart of
globalisation. This is a source of tremendous efficiency gains due to the
increased scope for specialisation.
This raises questions about the boundary but, much more importantly, the
form of the public sector. For example, the Bank of England is a state
organisation – but need it be? The scope for the exchange of information
Foster and Kaplan, Creative Destruction, 2001.
De Geus, The Living Company, 1997.
37
Japan may be an exception. Takashi Shimizu reports (2002) that the average coporate
life expectancy in Japan is 100 years and has been rising.
35
36
Diane Coyle
Page 17
3/3/2016
and monitoring of contract compliance in the case of setting interest rates
to meet an inflation target is clear. This is a case where the task could
readily be contracted out. In general, where monitoring is relatively
straightforward, the public sector could seek efficiencies from more use of
outsourcing and a reduced degree of vertical integration. This would also
help it offer more choice. To emphasise an earlier point, the issue is not
privatisation per se but rather the structure of production of public
services.
On the other hand, the value of tacit, non-marketable knowledge has been
increasing as the economy becomes increasingly weightless. This kind of
knowledge is extremely difficult to monitor and difficult to transact in a
market. This arguent suggests that big public sector computer systems –
which have in many cases been contracted out – should not have been
outsourced because monitoring and enforcing contract compliance is
impossible. (Equally, private sector companies should not have outsourced
this function either. Their experience has often been just as disastrous as
the notorious public sector examples.)
Intangible value also makes greater demands on the quality of organisations
– it raises the returns to organisational capital. The difference in
productivity levels between one firm and a competitor may lie almost entirely
in their organisational structures and behaviour rather than their spending
on ICTs or any physical capital. The failure to reorganise to capture the
potential efficiency gains is a failure of governance – in any context, public
or private, it is a failure to manage as efficiently as possible.
Diane Coyle
Page 18
3/3/2016
REFERENCES
Abed, George and Sanjeev Gupta, Governance, Corruption and Economic
Performance, IMF, 2002, details at
http://www.imf.org/external/pubs/nft/2002/govern/index.htm
Arrow, Kenneth, ‘Vertical Integration and Communication’, Bell Journal of
Economics, 6, 1975, 173-182
Axtell, Robert, ‘The Emergence of Firms in a Population of Agents: Local
Increasing Returns, Unstable Nash Equilibria and Power Law Size
Distributions’, CSED Working Paper No 3, June 1999, available at
www.brookings.edu
Baumol, William J., ‘Productivity Growth, Convergence and Welfare’,
American Economic Review, 76(5), pp 1072-1085, December 1986.
Baumol, William J, The Free Market Innovation Machine, Princeton
University Press 2002.
Bell, Daniel,The Coming of Post-Industrial Society, (1974).
Best, Michael, The New Competitive Advantage: The Renewal of American
Industry, Oxford University Press 2001.
Bresnahan, Timothy , Erik Brynjolfsson and Loren Hitt, ‘IT, Workplace
Organisation and the Demand for Skilled Labor: A firm-level analyisis,’ MIT
working paper, 2000.
Brynjolfsson, Erik, and Lorin Hitt, ‘Paradox Lost? Firm-Level Evidence on the
return to Information Systems Spending’, Management Science, 42:4 pp541558, 1997.
Brynjolfsson, Erik, and Lorin Hitt, ‘Computing productivity: Are Computers
Pulling Their Weight?’ MIT Working Paper 2000. (2000a)
Brynjolfsson, Erik, and Lorin Hitt, ‘Beyond Computation: Information
Technology and Organisational Transformation’, Journal of Economic
Perspectives, 14(4), Fall 2000, 23-48 (2000b).
Brynjolfsson, Erik, Michael D Smith & Yu Hu, ‘Consumer Surplus in the
Digital Economy: Estimating the Value of Increased Product Variety at
Online Booksellers,’ MIT Working Paper, April 2003,
http://ebusiness.mit.edu/erik/CSDE%202003-04-061.pdf
Chandler, Alfred, The Visible Hand: The Managerial Revolution in American
Business, Harvard University Press 1977.
Coase, Ronald, ‘The Nature of the Firm’, Economica, 4, November 1937, 386405.
Coyle, Diane, The Weightless World, Capstone 1996/MIT Press 1997.
Diane Coyle
Page 19
3/3/2016
Coyle, Diane, Paradoxes of Prosperity, Texere 2001.
De Geus, Arie, The Living Company, Nicholas Brealey, 1997.
Di Guilmi, Corraldo, Edoardo Gaffeo and Mauro Gallegati, ‘Power Law Scaling
in the World Income Distribution’, Working Paper August 2003,
http://www.economicsbulletin.com/2003/volume15/EB-03O40003A.pdf
Foster and Kaplan, Creative Destruction, Doubleday, 2001.
Hitt, Lorin, ‘Information technology and Firm Boundaries: Evidence from
Panel Data’, Information Systems research, 10:9, pp 134-149, June 1999.
Holmstrom, Bengt and John Roberts, ‘The Boundaries of the Firm Revisited’,
Journal of Economic Perspectives, 12(4) Fall 1998, 73-94.
Ijiri, Y, and H.A.Simon, Skew Distributions in the Size of Business Firms,
North Holland 1977.
Walter LaFeber, Michael Jordan and the New Global Capitalism, Norton
2000.
Malone, Thomas, ‘Modelling Co-ordination in Organizations and Markets’,
Management Science, 33(10), 1987, pp1317-1332.
Malone, Thomas, and John Rockart, ‘Computers, Networks and the
Corporation’, Scientific American 265(3), 1991, 128-136.
McKinsey Global Institute, US Productivity Growth 1995-1000:
Understanding the Impact of Information Technology Relative to Other
Factors, October 2001.
Milgrom, Paul, and John Roberts, Economic Organization and Management,
Prentice Hall 1992.
Motohashi, K, ‘Economic Analysis of Information Network Use:
Organisational and Productivity Impacts on Japanese Firms’ , METI working
paper, 2001.
Nordhaus, William D, ‘Do real output and real wage measures capture
reality? The history of lightsuggests not,’ Yale University, Cowles Foundation
Discussion Paper No. 1078, September 1994.
OECD, ICT and Economic Growth: Evidence from OECD Countries, Industries
and Firms, 2003
Quah, Danny, ‘Twin peaks: Growth and Convergence in Models of Distribution
Dynamics’, Centre for Economic Performance Discussion Paper no. 280, LSE,
February 1996.
Sheerin, Caroline, ‘UK Material Flow Accounting’, Economic Trends no. 583
June 2002, Office for National Statistics.
Shimizu, Takashi, ‘The Longevity of Japanese Big Business’. Annals of
Business Administration Sciences, Vol 1 No 3, October 2002, 39-46)
Diane Coyle
Page 20
3/3/2016
Simon, Herbert, ‘Organizations and Markets’, Journal of Economic
Perspectives, Vol 5, No 2 (Spring 1991), pp25-44.
Skidelsky, Robert,John Maynard Keynes: The Ecconomist as Saviour 192037, Macmillan 1992.
van Ark, Bart, and Erik Monnikhof, ‘Size Distribution of output and
employment: a data set for five OECD countries 1960s-1990’, Economics
Department Working Paper no 166, OECD1996.
Volcker, Paul, et al, Urgent Business For America: Revitalizing the Federal
Government for the 21st Century, available at
http://www.brook.edu/dybdocroot/gs/cps/volcker/volcker_hp.htm
White, Lawrence J, ‘What’s been happening to aggregate concentration in
the United States (and should we care?)’, Stern Business School working
paper, 2001.
Williamson, Oliver, The Economic Institutions of Capitalism, Free Press
1985.
Diane Coyle
Page 21
3/3/2016
FIGURE 1
Number of us es of 'governance' in a s ample of newspapers
2500
2000
1500
1000
500
0
1991
1992
1993
1994
1995
1996
1997
1999
2000
2001
2002
2003
Source: Neon database, adjusted to keep sample of newspapers covered unchanged
throughout the period.
Diane Coyle
Page 22
3/3/2016
FIGURE 2
Moore's Law: number of transis tors per integrated circuit, Intel
45000000
40000000
35000000
30000000
25000000
20000000
15000000
10000000
5000000
0
1965
Diane Coyle
1970
1975
1980
Page 23
1985
1990
1995
2000
2005
3/3/2016
FIGURE 3
Proportion of UK households with internet access at home
60
50
40
30
20
10
0
1998
Diane Coyle
1999
2000
Page 24
2001
2002
2003
3/3/2016
FIGURE 4
Diane Coyle
Page 25
3/3/2016
FIGURE 5
Investment in ICT as a share of GDP (%)
Source: OECD 2002
5
4
3
1980
2
1990
2001
1
0
Diane Coyle
France
UK
Page 26
US
3/3/2016
FIGURE 6
Global diffusion of technologies
tectechnology
all figures per 1000 people (UN DP)
20
0
18
0
16
14
0
0
12
0
10
08
0
6
0
4
2
0
00
Diane Coyle
TVs
Fixed telephone
Mobile
lin e s
s u bs cr ib er s
1970
0
1990
0
1995
5
2000
0
Page 27
2001
1
3/3/2016
FIGURE 7
Number of products available, key supermarket lines
Source: Federal Reserve Bank, Dallas
Ever-increasing choice...
... number of supermarket products availab
Pr o duct
19 80
19 98
Br e ad p ro duc ts
95
324
Bo ttl ed wa ter
12
125
Se as onin gs
61
40 3
Che w in g g um
47
16 7
W a shi ngpo wde r
12
48
Air fre shene r
53
372
Diane Coyle
Page 28
le
3/3/2016
FIGURE 8
Global M&A, $ billions
Source: UNCTAD, 2003
Diane Coyle
Page 29
3/3/2016