SOC IE,.Y 

..........................................................


.

.

.

.

. 

FABIAN PAMPHLET 544

.

.

. ................................................... . 


.

. 

.

.

.

. 

. 
. .. Targeting

. 

.

.

. 
. Competitive


. 

.

.

. 

. 
... Industries 
•

~~ 
••• 

.

.

.. 
by P.A. Geroski 
and K. G. Knight


.

.

.

.

.

. 

.

.

.

. 


.


.

. 


. 


.

. 

. . . . . . . . . .. . . . .. . . . .. . . .. . . . . . . . . . . . . .. . . . . . . .. . . .. . . .. . . . 



FABIAN PAMPHLET 544 

2 1 0037737 5 TELEPEN 

TargetingcompetitivE 
industries 

1. Markets and market failure 
2 Market infrastructures 
3 Targeting growth points 
4 Local industrial policy 
5 What should Labour do? 
7 


BLPES 

1 

5 
9 
13 
17 

Paul Geroski is Associate Professor of Economics at the 
London Business School. 

Kenneth Knight is Senior Lecturer in Economics at the 
University of Warwick. 

This pamphlet, like 
all publications of the 
Fabian Society, 
represents not the 
collective views of the 
Society but only the 
views of the authors. 
The responsibility of 
the Society is limited 
to approving its 
publications as 
worthy of 
consideration within 
the Labour movement. 

Design: Tony Garrett 

Aprill991 

ISBN 0 7163 0544 5 

I SN 0307 7523 

Printed by The College 
Hill Press Limited <TU l, 
London and Worthing 

Publi hed by the Fabian 
Society, 11 Dartmouth 
St, London SW!H 9B 


Markets and market failure 

1 

One of the more curious ofMrs Thatcher's 
achievements in eleven years of rule is that she 
made almost all of us into believers in the 
efficiency of markets. Or did she? 

M
M
ost now accept the presumption that markets will handle certain 
kinds of problems far more effectively than administrative action, 
and some of us are willing to design administrative procedures 
that mimic the operations of markets in dealing with 
problems for which markets do not or cannot exist. The hoary traditions of 
paternal and planner socialism have (by and large) given way to new conceptions 
ofmarket socialism which promise to be both more democratic and more 
effective than their predecessors. Of course, some believe in markets a little 
more fervently than others, and many see limits to the efficiency of market 
operations that others will not accept. The interesting question is, then, not 
so much one of'markets or not' as of'how much market, how much socialism'. 

A natural way to start the search for an answer to this question is to look 
for systematic types of market failure, such as monopoly, uncertainty or the 
existence of significant externalities. The difficulty with using the market 
failures paradigm as a basis for designing a coherent industrial (and regulatory) 
policy is that one is almost inevitably reduced to thinking about policy 
on a case by case basis. Every market has its own peculiar mix of market 
failures, some more serious than others, and the appropriate solution to any 
particular type ofmarket failure is bound to depend on the nature and severity 
ofthe other failures that occur in that market. However, any policy that begins 
with the presumption that every case is unique and requires its own unique 
solution will lack coherence, and is bound to degenerate into political expediency 
and opportunism sooner or later. What one requires to design a coherent 
policy programme is a set ofpresumptions about which types ofmarket failure 
cause serious problems in a wide variety of settings. Our view is that the 
process of making or creating markets is fraught with market failures of the 
type discussed above, and, in particular, that it can result in an inadequate 
development ofthe infrastructure ofmany markets, leading to unsatisfactory 
market performance. 

Properly functioning markets require the existence of certain supporting 
institutions, and the efficiency with which they operate depends on the quality 

1 


ofthis infrastructure. Firms operating in any particular product market rely 
on various input suppliers to provide the wherewithal to manufacture an 
attractive product, and on various distributors to get it to those consumers 
who wish to purchase it. The success with which a firm can produce and sell 
its product obviously depends on its relations with suppliers and distributors, 
as well as on their basic competitive abilities. Public policy problems arise 
because a market's infrastructure helps to support all ofthe firms who operate 
in it, and no one firm has an incentive to create the necessary infrastructure 
alone ifthis will benefit its rivals. Ofcourse, many markets do exist, and that 
means that it has proved to be in someone's interest to create them. The 
problem is that many entrepreneurs may find it profitable to create a market 
only when their control over the infrastructure gives them market power in 
that market. That is, the conditions under which many markets are created 
may more or less permanently blight the efficiency oftheir operations. 

Fixed costs 

To appreciate the. problems of creating markets, consider what might be 
involved in setting up even the simplest ofmarkets-an exchange market (i.e. 
one where no production occurs). One needs a large building, often located in 
a high-rent area of town, to house the exchange, office buildings to house the 
individuals involved in each transaction, an elaborate record-keeping system 
to keep track ofall the exchanges, and a system ofmonitoring and enforcement 
to ensure that the rules of the exchange are obeyed. One must also meet the 
cost ofthe time ofall the people involved in the activity. These fixed costs are 
not inconsiderable, and it is not clear how an entrepreneur that tries to set up 
such a marketmight recoup these costs and make a d~centprofit. The problem 
is that a market like this creates positive externalities, providing valuable 
information and an opportunity for participants to realise gains from trade. 
All traders are likely to benefit from the existence of, for example, a futures 
market, but not so much that it is worth the while of any single one of them 
to create the market. The consequence is that very few organised futures 
markets actually exist, and many of those that are created fail. A study of 
futures markets in the U.S. suggests that about 40% of all newly-created 
futures markets fail within the first five years oflife. 

An example ofthe type of problem that may emerge in creating a market 
can be drawn from the computer industry. It is important to ensure that the 
basic hardware ofa machine is compatible with both the kinds ofsoftware that 
users require and the various peripherals or add-ons they may wish to use. 
The existence of a basic standard machine and/or software operating system 
is, therefore, crucial to the development of the market simply because the 
existence ofa standard enormously enhances consumer choice and the ability 
to make choices quickly and effectively. The more users there are of a basic 
software system, the more incentive software developers have to develop 

2 


packages (particularly specialist packages); the more standardisation there is 
in hardware, the more competitors there are likely to be on the market, and 
the less likely it is that consumers will get 'locked in' to any one of them. An 
'open standard' is, therefore, an important part of the development of the 
market. However, it is rarely the case that firms participating in the market 
will have an incentive to develop an open system. Rather, it will be in each 
firm's interest to develop a 'proprietary system' that locks consumers into its 
product, restricts the number of competitors a firm faces and gives it leverage 
in bargaining with software or peripherals producers, that wish to gain access 
to its standard and sell in its capitve market. When a single firm dominates 
the market in question (such as IBM in the case of computers) it is likely to 
be the case that the firm will develop a proprietary standard. The consequence 
can be under-developed infrastructure -less software packages, peripherals 
and add-ons are produced than could or ought to be -and a market in which 
the dominant market power becomes entrenched, perhaps for long periods of 
time. 

The way in which market infrastructure develops plays an important role 
in determining the kinds of competitive strategies used by firms which operate 
in thatmarket. The competitive skills which firms acquire are often developed 
co-operatively with upstream suppliers or downstream distributors and, as a 
consequence, the ability offirms to acquire certain types ofskills depends on 
the strength and depth ofsupporting institutions upstream and downstream. 
Firms that operate in particular markets in different countries are bound to 
develop comparative advantages that reflect the different capabilities in the 
infrastructures in the national markets in which they operate. Differences in 
their performance can often be traced back to strengths and weaknesses in 
these infrastructures. 

Kitchen manufacturers 

For example, the U.K. kitchen manufacturing industry suffers from the lack 
of a well-developed domestic infrastructure of machine tools and materials 
and parts suppliers. U.K. manufacturers are disproportionately represented 
in the low quality, low value-added segment; German producers dominate the 
top, high quality segment. Kitchen furniture manufacture in the U.K. is a 
distinctly low-tech, low skill operation, but in Germany even small manufacturers 
use highly-sophisticated machines, almost all of which are made in 
Germany. Those British manufacturers who do use sophisticated machinery 
generally experience more breakdowns than their German counterparts, and 
are far less productive. The problem is that the U.K. users get no maintenance 
assistance from local agents, so need to carry large stocks of spare parts. By 
contrast, with machine manufacturers sometimes only half an hour away by 
car, German manufacturers call on their suppliers' specialised service engineers 
fairly frequently. The U.K. industry suffers from a structural weakness 

3 


that no amount ofmanagerial talent or incentives applied to kitchen furniture 
manufacturing are likely to ameliorate. 

It follows from this that those concemed with creating national comparative 
advantages in particular sectors should concentrate on developing 
the infrastructures of those sectors. Further, since the process by which 
market infrastructures develop is fraught with problems associated with 
monopoly, uncertainty and extemalities, the promotion and development of 
market infrastructure should be a leading industrial policy concem for all but 
the most obdurate laissez-faire enthusiasts. 

The challenge for public policy makers is to create markets which can then 
be used by entrepreneurs who do not enjoy monopoly power. Failure to develop 
such infrastructure can mean that certain types of markets fail to come into 
existence, that others are dominated by their creator in a way that inhibits 
subsequent market development, and, finally, that others fail to develop their 
full ,potential because their infrastructure is too poor to enable the firms that 
operate in them to develop genuine comparative advantages which enable 
them to operate competitively in world markets. Markets may work well as a 
rule, but 'the market for making markets' almost certainly does not, and it is 
the failure ofthis market above all that calls for the attention of public policy 
makers. 

4 


Market infrastructures 


Competitive advantage is generally thought to 
depend on the possession of some 'special asset' 
or 'special skill'. 

I 
I
f the special asset or skill is both unusual and results in a product or 
service that creates considerable value added for consumers, then it will 
yield high rents to its possessor. If, in addition, it is an asset or skill that 
is difficult to imitate, then its possessor will enjoy these high rents in 
perpetuity. Itfollows, then, that to help restore the comparative advantage of 
U.K. industry in world markets, industrial policy must be designed around 
the problem of stimulating the development of such special assets or skills. 
Most discussions that stress the need for a supply side policy to rejuvenate the 
industrial infrastructure ofthe U.K. (which is nearly all ofthem) focus mainly 
on the need for more investment on R&D and training, and they typically 
advocate national policies that are general enough to benefit all firms and 
workers. Although it is clearly the case that market failures bedevil private 
sector efforts to make these types of investments (and have done so to a 
noticeable degree in the U.K.), our view is that this kind of public policy 
response is too limited in scope to be of much use in compensating for these 
market failures. Better targeted and more locally based policies are also 
needed. 

In very general terms, the special assets or skills that create competitive 
advantage for a firm are those which enable it to produce better designed 
products more efficiently. The ability to do this depends (in part) on the firm's 
consumers, suppliers and on its rivals. Consumer needs can only be met ifthey 
are articulated, and firms cannot be expected to produce sophisticated, high 
technology products if consumers are unprepared or unwilling to use them. 
Although the size ofthe market is important in stimulating innovation, what 
matters more is the 'pull' that sophisticated buyers exert on innovators by 
demanding new and innovative products. To meet this demand, firms must 
develop efficient production processes, and their ability to do this depends (in 
part) on the quality of the suppliers of materials, machinery and labour that 
they deal with, and often on how closely they are located to the producing firm. 
Home-based suppliers create advantages in downstream industries when they 
deliver cost-effective inputs quickly and effectively. More important, supporting 
industries often stimulate innovation through a kind ofsupply 'push' that 

2 


5 


is based on close working relationships that facilitate a quick and constant 
flow ofinfonnation, and an ongoing exchange ofideas. Finally, it is one thing 
to have opportunities, but it is quite another thing to take full advantage of 
them. No matter how sophisticated their buyers and how technically adept 
their suppliers are, finns that enjoy protected positions of monopoly power 
usually fail to fully develop their skills simply because they lack the incentive 
to act that competition brings. Competitive advantages almost always emerge 
from the crucible of active product market competition. 

Clusters 

In short, the competitive ability of finns depends in large measure upon the 
infrastructure of the market that they operate in. Buyers make demands 
which, when reinforced by active competition amongst close rivals, leads finns 
to try to create new and innovative products. Their ability to respond to their 
buyers' needs depends on whether or not they can successfully manufacture 
the product efficiently, and this depends on the skills and abilities ofsuppliers 
(and on how effectively they are honed by the pressures of competition 
amongst suppliers). Itfollows, then,that the special assets or skills that create 
competitive advantage in any one sector depend (at least in part) upon those 
developed by finns in upstream, downstream and neighbouring sectors. The 
result is a whole cluster of(often geographically proximate) activities which 
has a competitive strength far in excess ofthe competitive strength ofthe sum 
of its parts. In a sense, the proposition that competitive advantage emerges 
from clusters ofmutually reinforcing innovative activity means that economic 
development is inherently uneven: nations rarely develop comparative advantages 
across the whole range of manufacturing industry, and sectors that 
develop competitive advantages which enable them to grow rarely expand in 
a regionally balanced fashion. 

The proposition that nations develop comparative advantages mainly in 
clusters of highly-related activities and not across the board is fairly evident 
to even a causal observer. Recent work at Harvard directed by Michael Porter 
has documented this pattern in some detail. For example, his team observed 
the U.K. to have competitive advantages in alcoholic beverages, confectionery 
and biscuits, cigarettes, cosmetics, perfume, various household products (such 
as toothpaste and soaps), financial services, consumer goods retailing, and 
petroleum and chemicals. Competitive advantage was seen to be weakest in 
textiles, office products, telecommunications equipment, consumer electronics 
and mechanically basedconsumergoods (such as watches).Japan, bycontrast, 
was observed to have competitive advantages in transportation equipment, 
office machines, consumer electronics, steel and fabricated metals, electronic 
components and computing equipment, cameras and film. Relatively weak 
Japanese sectors included chemicals and plastics, food and beverages, packaged 
consumer products and nearly all types ofservices. West Gennan finns, 

6 


to take a final example, were observed by Porter's team to have substantial 
competitive advantages in chemicals, metal and metal working, optical-related 
products and in a wide range of household goods (such as porcelain-related 
products, fumiture and appliances). Relatively weak sectors were found 
to include all types of services, semiconductors and computers, electronics 
products and telecommunications equipment. 

Both the rich interaction ofupstream, downstream and neighbouring firms 
which defines clusters and their unbalanced regional development is evident 
in a wide range of cases discussed by Porter and others. For example, consider 
the world printing industry. West German dominance of the production of 
printing presses dates from the 19th century, and today the industry accounts 
for 35% of world production of printing presses. Ithas an export share ofjust 
over 50% and it exports to more than 120 countries. The six leading firms in 
the sector are all located within 150 miles ofeach other in southem Germany, 
and are part ofa cluster that includes world leaders in the production ofpaper 
making machinery, paper producers and ink producers. That these several 
upstream and downstream producers form a richly interacting cluster is 
evident from even the most casual inspection of the process of innovation in 
this sector. The development of high-speed printing machines, for example, 
required the cooperation of paper makers (who developed paper that could 
withstand the pressures ofhigh speed printing) and of paper machine producers 
(who developed the machinery necessary to product that type of paper). 
Similarly, ink producers played a major role in the development of inks that 
could be used in several generations ofnew printing presses. The whole cluster 
is supported at one end by the high standards and sophistication of German 
printers (the German market is only the sixth largest in the world), and, at 
the other end, by the work done in the German chemical industry on synthetic 
indigo and by a series ofvocational schools set up by the leading firms to train 
skilled labour. 

Robots 

Japan produces more than 50% of the world's robots, exporting at least 20% 
of its domestic production. The industry started in 1970 with a licensing 
agreement between Kawasaki Heavy Industries (a major user ofrobots and a 
producers of related products and services) and one of the American firms 
which first developed robots (Animation). By 1980, there were 130 robot 
producers in Japan at the core of a cluster of activities driven by users and 
supported by a wide range of supplier industries. The major users of robots 
are automotive and domestic appliance producers, and much of the development 
of this sector can be traced to their actions. Cooperative trade unions 
and managements with solid engineering backgrounds made Japanese firms 
much more· willing to investin robots than their (oreign competitors in a range 
of sectors. Nissan, for example, was an important early customer, helping 

7 


Kawasaki to design and service robots. In due course competitive pressures 
in the automotive industry led the other major Japanese car producers to help 
develop and install robots, and the growth and development of the robot 
market piggybacked (in part) on the growth of the Japanese car industry, 
providing a base market from which further applications ofrobotic technology 
could be developed (such as the electronics, plastics processing, general machinery 
and metal working sectors). Needless to say, the success ofJapanese 
robotics producers in meeting this demand was due in no small part to the fact 
that the technologies employed in producing robots (numerically controlled 
machine tools, motors, optical sensors, and the like) have, in the main, been 
developed by Japanese firms who, as often as not, are leaders in these sectors 
as well. 

The fact that economic development tends to be rather uneven, that nations 
tend to develop comparative advantage in clusters of activities rather than 
across the board, carries several important implications for the design of 
industrial policy. In particular, it means that policy must be targeted, that it 
mustbe designed and implemented locally, and that regional policy mustfocus 
on building up 'growth points' rather than on dispersing economic activity. We 
consider each of these in turn. 

8 


Targeting growth points 

It seems clear that creating competitive 
advantage in particular sectors often requires 
developing the infrastructure of those sectors, 
building up clusters ofmutually supporting 
activities which jointly develop a range of 
special skills or assets. 

T
T
he needs ofeach sector are idiosyncratic. Itfollows that policy must 
be custom-designed and sector-specific in application. The infrastructure 
suitable for textiles has little in common with that suitable 
for cars, and any initiative that tries to address both simultaneously 

will almost inevitably be too general to be useful. 
Although it does not seem to be particularly remarkable taken atface value, 
the observation that policy needs to be targeted cuts against the grain ofrecent 

U.K. industrial policy practice. In fact, most industrial support in the U.K. is 
not sector-specific, and much ofit comes in the form ofcapital allowances and 
stock relief. The sole virtue ofgeneral non-discretionary support systems (such 
as those implemented through the tax system or general subsidies disbursed 
by the DTI) is that they are easy to administer, support being given on the 
basis ofeligibility rather than need (which is, of course, much harder for civil 
servants to determine). As a consequence, some firms get support that they 
don't need, while others don't get what they need or don't get it in the form 
that they need it. Tax incentives encouraging investment are not needed by 
firms who elect to pay high dividends, and they provide little assistance for 
firms whose investment programme is disrupted by failures in capital goods-
supplying industries. R&D tax credits do not stimulate the investment in 
human capital that is often necessary to exploit new technology, and encouraging 
R&D in technologically barren areas is pointless. In fact, most of the 
rather limited sector-specific support that has been given over the years in the 
U.K. has flowed to one of a small number of crisis sectors (such as steel, 
automobiles, shipbuilding, and so on), and often it has done so in response to 
short-term political pressures. As a result the application ofindustrial policy 
in the U.K. has often been a little confused, and many ofthe policies directed 
at crisis sectors have, in fact, been designed to slow adjustment rather than 
to facilitate it. 
3 


9 


That support ought to be sector-specific is often described -and then 
dismissed -as requiring civil servants to 'pick winners'. In fact, as an 
alternative to the current policy of being 'led by losers', picking winners has 
something to be said for it. Rather more subtly, 'picking winners' is a characteristic 
of a type of industrial policy that contents itself with handing out 
subsidies and then standing back to watch what happens. The only decisions 
involved in implementing this kind of policy are, of course, 'who' and 'how 
much'. In fact, there is plenty of evidence to suggest that subsidies are both 
expensive and extremely inefficient. They encourage firms to develop lobbying 
skills rather than competitive skills, and they create a climate ofdependency 
which saps their competitive vigor. Where government can play a positive 
industrial policy role that targets strategically important sectors is in the 
marketplace, subject ofcourse to the potential and unexplored restrictions of 
the Single European Market after 1992. Governments are enormously large 
buyers of a wide range of goods and services ranging from electronics to 
pharmaceuticals, from office equipment to transportation equipment. Many 
of the sunrise sectors of the 1960s and 1970s started with help from government 
purchasing contracts (typically defence contracts), and government 
purchasing is likely to play a large role in stimulating the emergence of a 
number ofsunrise sectors in the 1990s. As buyers ofa range ofsuch products, 
governments are in a position to act as a stimulant to supply by encouraging 
competition amongst suppliers, demanding products that are technologically 
sophisticated and innovative, and providing a large enough market for the 
new products to amortise most of the R&D expenditures needed to develop 
them. Although procurement polices can go disastrously wrong (as in the case 
ofSystemX), when they are well managed they can stimulate the development 
of a range of highly-interrelated sectors. An intelligently managed procurement 
policy does not pick winners so much as create them. 

Semi-conductors 

A classic example of the successful use of procurement policy to develop a 
sector can be found in the early development of the U.S. semi-conductor 
industry. The development of both the early silicon transistor and the integrated 
circuit emerged at the prompting of the U.S. armed forces (although 
the original transistor appeared independently). The clarity ofdefence needs 
focused the innovative efforts of supplying firms. The size of military orders, 
the forces' willingness to pay, and the high premium put on quality by defence 
procurement bodies enabled producing firms to learn about new products and 
processes in a way that simply would not have been possible had more 
market-driven (and, therefore, more short-term) commercial considerations 
been applied to product and production choices. Many of the early transistors 
were incredibly expensive to produce, and the sector would never have developed 
as it had if the sole source of demand had been the private sector. 

10 


Military demand enabled firms to scamper up the learning curve, and thus, 
in effect, subsidised the emergence ofa huge private sector market (one whose 
size now far exceeds the defence-related market for semi-conductors). Similar 
effects ofprocurement policy can be observed in the US computer industry, in 
commercial aircraft and other sectors. 

Top down or bottom up? 

Ofcourse, procurement policy is not the only form oftargeted industrial policy 
available to policymakers. Government can facilitate the development of 
various types ofinfrastructure institutions or the restructuring of particular 
sectors, and government can ensure the provision ofcertain types ofgoods or 
services (such as training). There are several ways in which such policies can 
be implemented. 'Top down' policy strategies involve spending decisions made 
by policymakers that directly create market infrastructure institutions, while 
'bottom up' policies are those in which such spending decisions are made by 
private sector agents in response to incentives created (or manipulated) by 
policy markers. Either way, the success of such policies depends on whether 
they can stimulate the development of competitive advantage; that is, on 
whether they help insure the emergence of a cluster of mutually supporting 
activities. The point is easy to see in the contrast between U.K. and Japanese 
industrial policies in the machine tools industry. 

Japan is now the world leader in machine tools, having risen to a share of 
world production of about 25% in 1982 from less than 1% in 1955. Over the 
same period, the U.K. share declined from just under 9% to just over 3%. Both 
countries followed interventionistindustrial policies. InJapan, R&D subsidies 
were provided to the industry, together with the usual range of soft loans, tax 
breaks and protectionist measures. However, the important policy choice was 
made in the early 1970s when one company, Fujitsu FANUC, was encouraged 
to become the dominant supplier of control units to the industry. This effectively 
standardised the parts used, making it very easy for producers to 
manufacture low-cost numerically-controlled lathes and machining centres. 
In the U.K., the policy of the 1960s (implemented through the IRC) was one 
ofencouraging mergers, although no attempt was made to ensure that the new 
national champions realised any of the efficiencies promised by such rationalisation 
(Alfred Herbert was run as 24 separate companies for quite some 
time after rationalisation had turned it into the world's largest machine tools 
producer). R&D was supported, although in the U.K. money was channeled 
into highly ambitious products and very little ofit diffused through the sector. 
Finally, although standardization was recognised as being important, import 
substitution was accorded far greater weight, and firms were encouraged to 
produce products that could be used as substitutes for those produced abroad 
and imported into the U.K.. 

It is clear that the success of the Japanese policy arose because policyma


11 


kers were able to help articulate desirable product choices for the industry as 
a whole, and they managed to support and sustain thatchoiceby a very shrewd 
intervention in the infrastructure of machine tools production. 'By contrast, 

U.K. policy (which used many of the same instruments as were used by the 
Japanese) was dominated by attempts to develop over-ambitious 'big science' 
projects, as well as by a quite unwarranted beliefin the proposition that mere 
size confers competitive advantage. Efforts to standardise crucial inputs and 
to develop products suited to the current competitive strengths of U.K. 
producers were neglected in favour ofrather short-sighted efforts to ameliorate 
the U.K.'s chronic balance of payments problem. What U.K. policy lacked 
was a sense of the relationship between machine tools producers and both 
their suppliers upstream and their buyers downstream. 
Research projects that corresponded to no real needs, and a total lack of 
effective interest in standardisation left U.K. policymakers with little to do 
but to create large giants in the 1960s and bury them in the 1970s. 

12 



Local industrial policy 


Since clusters ofcompetitive advantage develop 
in geographical proximity to each other, and 
specific firms in these clusters often require 
quite highly specialised, idiosyncratic 
infrastructures the design and implementation 
of industrial policy must inevitably be localised. 

A
A
ny targeted policy worthy of consideration requires a good deal of 
highly detailed information about the target sector if it is to be 
properly designed and implemented, and there is no doubt thatlocal 
policymaking units operating close to the targeted firm or sector 
will have a comparative advantage (relative to civil servants in Whitehall) in 
gathering such information. Further, broadly designed policies administered 
from afar tend to become more and more unwieldy as the number of agents 
involved increases, complicating administration in a way that often blunts the 
effective application ofpolicies. Smaller, locally-based policies can often be far 
more effective in this respect, and, in addition, they may also prove to be more 
sensitive to local needs and preferences. 

What is required to create and sustain an infrastructure in a particular 
sector is often little more than the coordinated mobilisation of local institutions, 
and it may require no more than helping to bring into fruition plans 
made by the firms in the sector themselves. Links with local universities and 
various vocational training institutions are needed to provide training for 
skilled workers ofvarious types, something that these institutions may not do 
in the absence ofa well-defined, clearly expressed local demand for these skills. 
Entrepreneurial activities in a wide range of related sectors need to be 
stimulated both by the development of clear mechanisms for expressing 
demand, and by the development of institutions to help overcome bottlenecks 
in supply. Most sectors are surrounded by a busy, buzzing hive of support 
activities, many of which are provided by small firms. To stimulate the 
development ofthese support activities and to keep them competitive, finance 
and hands-on managerial assistance needs to be made available and premises 
need to be provided. Private sector institutions are often uninterested in 
supporting this type of economic activity (because they are unable to appropriate 
enough of the gains to make their investments pay), and this means 

4 


13 


that there is likely to be an important role to be played by enterprise boards 
of various types. 

Indeed, practical experience in the U.K. and abroad (notably in Italy) has 
shown that local enterprise boards can often serve as a focus around which a 
number of richly interacting layers of local economic activity swirl, and one 
senses thatthis type of institution has a richer potential than has been realised 
thus far. Greater London Enterprise, for example, provides venture capital 
funding (often in partnership with private sector agents and occasionally on 
a sector-specific basis) coupled with hands on management of their portfolio 
of projects. Sector-specific funding projects (targeting tourism, film production, 
and so on) are more or less explicitly designed to create clusters of 
mutually supporting activities. More subtly and more important, bundling 
funds together into general or sector specific programmes helps to bridge the 
well-known financial black hole that affiicts small firms: the amounts that 
they wish to borrow are too small for most financial institutions, whose 
minimum lending levels can run to 250,000 or more. Greater London Enterprise 
effectively bundles a range of small loans together into a package large 
enough for private sector financial institutions to handle, and then manages 
the resulting portfolio. Part of the hands-on management component of the 
services provided to small firms consists of providing basic business skills to 
entrepreneurs who have a good idea but little business sense or experience, 
and part of it is delivered via a variety of training and consultancy services 
direct towards small local enterprises (and to various local authorities). 
Finally, Greater London Enterprise helps to develop industrial and commercial 
property in London (again, often in partnership with private sector firms), 
providing (amongst other things) managed works paces for small firms. These 
are integrated workplaces that provide a range of premises (allowing individual 
firms to expand or decline without changing their business address) and 
associate support services (including creches, in-house printing and so on). 
Many of the newly formed TECs ought to be able to provide similar, complementary 
services on a local basis. 

Networks 

Regional Councils in Italy have proved to be successful in stimulating the 
growth of small and medium-sized firms, often creating networks (or 'industrial 
districts') ofmutually supporting activities. The services provided include 
administrative activities like bookkeeping, auditing, industrial relations support, 
legal services and insurance provided in exchange for a subscription fee. 
More interesting have been a number ofsector-specificinitiatives. In clothing, 
for example, the Centro Informazione Tessile dell'Emilia Romagna offers a 
range of design, marketing and manufacturing services (including market 
research and forecasting, training and so on) to its 500 or so member firms, 
while other institutions offer quality certification, financial assistance, infor


14 


mation on subcontractors and so on. Finally, public policy has included a 
range ofland use policies that have stimulated the development ofindustrial 
parks, and the provision ofspecialised premises for small artisan firms. 

These various services -which help to co-ordinate the development of a 
sector and its infrastructure at a particular location -are analogous to 
activities which help create markets (such as futures markets). They convey 
positive externalities to all participants,but it is often difficult for the provider 
ofthese externalities to obtain a fair reward for his/her activities or a reasonable 
compensation for the risk involved. That is to say, the market for 
providing this kind of service is riddled with market failures, and this opens 
up a window of opportunity for a carefully designed pro-active industrial 
policy. 

Regional policy 

Maximising the benefits ofclustering through the creation ofgrowth areas of 
industrial development has been a feature ofboth French and Italian regional 
policy. In those countries, land use planning is perceived as being complementary 
to industrial policy, with a resulting recognition of the importance of 
'economies of association'. These arise when the efficiency of a complete 
complex of firms is crucial to the performance of each individual firm in the 
group, exactly the effects that we have described as resulting from clustering. 
In France, this is expressed in the notion of a 'pole de croissance', a growth 
point at which interrelated firms are encouraged to locate in order to secure 
the efficiency gains from association that are available. 

In Britain there has been some awareness ofthe potential value ofclustering 
firms at growth points, and as long ago as 1940, the Barlow Commission 
argued that new industry should be located in 'key points'. However, there has 
been little practical implementation of the idea even in the halcyon days of 

U.K. regional policy in the 1960s and 1970s. In fact, the traditional U.K. 
regional policy stance is in part a legacy ofthe 1930s when high unemployment 
was highly localised, and it has led to a strategy that focused on palliatives for 
above average unemployment, one dedicated to dispersing activities evenly 
throughout the country. As a result, regional policy has encouraged the 
diversification offirms as an assurance against unemployment, and not their 
concentration to maximise joint efficiency. Not surprisingly, appraisers of 
regional policy in the U.K. have noted its favourable short-run impact on 
unemployment, but many have also commented on its disappointing effect on 
the long-run performance of the slow-growing regions at which it has been 
directed. Despite years ofrelatively lavish policy attention, regional imbalances 
in the U.K. have persisted over time (and, in some cases, worsened). If, as 
we have argued, the process of economic development is inherently uneven, 
then some regional imbalances must be expected. The policy failure comes 
from the fact that some regions have stagnated over long periods of time while 
15 


others have persistently prospered. The policy of dispersing the activity of 
successful enterprises or industries has not only led to the scattering ofseeds 
on infertile ground, but also created a range of political hostages that have 
distorted industrial policy making in these sectors throughout the country (as 
the Wilson government's response to the Chrysler crisis illustrates). 

There are other reasons for the neglect of the clustering approach in the 
U.K.. While there is general recognition of the right of public authorities to 
determine where industrial development occurs, there is a conspicuous reluctance 
to interfere (except within broad limits) with the rights oflandowners to 
determine what development actually takes place. Land-use planning in the 

U.K. is, as a result, almost entirely reactive. The effect of this has been a 
pattern of haphazard land use for which it is difficult to find any economic 
justification. This passive policy stance also makes the implementation of a 
clustering strategy well nigh impossible. The only pro-active element in 
location policy in practice is the competition thattakes place between planning 
authorities wanting to attract developers to their own area, a practice which 
tends to encourage excessive passivity in the public response to the pressure 
from private developers. It also prevents the realisation of the 'economies of 
association' that are so important to competitive advantage and economic 
performance. To regulate the competition between local authorities, the DoE 
issues Strategic Planning Guidance for Structure Plans. These are prepared 
by local authorities who are forced by this procedure to co-operate and agree 
to region-wide constraints on their own individual land-use planning. Although 
it does offer a promising framework for the kind oflocation planning 
that is required if market failures are to be overcome and the benefits of 
clustering are to accrue, it has been less than wholly successful in practice. 
On its own, however, Strategic Planning Guidance is not enough. Local 
planning authorities need also to have the power and the responsibility to plan 
actively. One way forward is to give local authorities an explicit economic 
development role. To do this they need to be free of the financial restraints 
that prevent them acquiring development land and force them to dispose of 
thatwhich they currently own. Alternatively and perhaps more appropriately, 
a region-wide agency could perform this vital role of land acquisition and 
disposal, while the detailed planning ofthe land could then be devolved to the 
lower tier authorities. Where land is in private hands, developers must not be 
in a position (as they are now) to win almost any application for planning 
consent on appeal, irrespective ofthe economic case for the development they 
propose. At the very least, they need to be required to demonstrate that their 
proposals do not prevent desirable clustering from taking place. 

16 



What should Labour do? 

5 

The Conservative years have been bad for 
British industry. A triumphalist adulation of 
free markets coupled with a succession of 
Industry Ministers lacking in vision and drive 
has done nothing to reverse the long-term 
decline of the U.K.'s manufacturing base or its 
international competitiveness. 

I
I
t will not be hard to improve on this rather dismal performance. Indeed, 
the natural place to start is with one of the deeper flaws in the Tory 
conception of what governments can and ought to do. Insofaras Mrs 
Thatcher's government had a positive industrial policy, much of it 
centred around the perceived need to alter the environment in which business 
operates. Promoting something called the 'enterprise culture' and turning 
various conspicuously successful entrepreneurs into folk heros were initiatives 
that sprang from the view that many of the U.K.'s industrial policy 
problems are cultural and have nothing to do with the way that markets 
operate or with the infrastructure that they operate in. 

However plausible this view might, at first glance, seem to be, the hard 
truth is that one cannot change a culture or a business environment simply 
by renaming the DTI and lionising Richard Branson. Cultures and attitudes 
persist as they do because they are supported by institutions ofvarious sorts, 
and real change requires that one remoulds these institutions. 

The Labour Party can (and should) take the lead in trying to rejuvenate 
the U.K.'s manufacturing base, and thatitmust do this not only by introducing 
a coherent and active industrial policy, but, more deeply and more subtly, by 
changing the intellectual environment in which industrial policy discussions 
are staged. Industrial policy will never succeed where civil servants and 
politicians are so overawed by the apparent potency of market forces that they 
readily concede to doubts about their competence and ability to act. The 
analysis discussed above seems to us to provide a firm basis upon which to 
erect a productive policy programme, and a sturdy platform from which to 
assail the self-doubts that have long crippled the development and implementation 
ofindustrial policy in this country. 

17 


Training 

Labour has already given explicit recognition to the fact that investment in 
R&D and investment inhuman skills through training are deeply complementary. 
There is simply no point in building scientific laboratories (or in giving 
tax breaks to firms who promise to do so) ifthere are no scientists capable of 
using them effectively. Policy in this area can be thought of in terms of 
creating a market for skills ofdifferent kinds. To create such a market, policy 
needs to focus and then articulate demand. It needs to organise and mobilise 
the sources of supply needed to meet that demand, and to set standards and 
provide the certification that individuals need ifthey are going to be active in 
a market for skills of different kinds. 

The major problem in this regard is that training must have a major 
on-the-job component. The skills needed by firms often contain a large 
'firm-specific' component-general educational establishments cannot provide 
that kind of training. The market failures that bedevil the provision of 
firm-specific training are well-known, and it is highly unlikely that they will 
be solved by TECs which are dominated by employers. TECs have the 
potential to stimulate the provision oftraining, but there must be more public 
sector representation on their boards, and they must be given responsibility 
to monitor and maintain the standards of any training activity that they 
support or that occurs in their area and is supported by public funds. Finally, 
trade unions have a major role to play, and also a responsibility to play that 
role. The market failures that inhibit private sector profit-seeking organisations 
from acting do not, and should not, always inhibit private sector nonprofit 
organisations from acting. 

Important as it is, however, training alone is not enough. The way forward 
might be summarised in the form of two further broad policy commitments. 

Sector-specific policies 

Both common sense and most of the available evidence suggests that sector 
specific policies must be the way forward. Itis hard to believe that macroeconomic 
policy alone will have much effect on the U.K.'s international competitiveness, 
and there is little reason to think that pulling a few strings in the 
form oftax incentives orgeneral subsidies will have much impacteither. Thus, 
a commitment to sector-specific policies carries with it a commitment to avoid 
an exclusive reliance on certain types ofpolicy tools, and to begin working with 
others. Our view is that procurement policy is an extremely potent industrial 
policy tool, one that involves government action dictated by the 'selfinterest' 
of its needs as a consumer within the context of an ongoing market process. 
In a sense, procurement policy provides one partial butrather natural solution 
to the question of'who picks the winners', a solution that does not require civil 
servants to second guess private sector demand. Other policy ·initiatives that 

18 



are consistent with the spirit of employing sector specific policies involve 
creating or encouraging the growth of institutions that facilitate the emergence 
of sophisticated buyers and the development of supplier capabilities, 
encouraging the growth ofenterprise boards that provide hands-on assistance 
to small and medium-sized firms, resisting the temptation to reduce the degree 
ofcompetition in markets by creating 'national champions', and so on. Finally, 
no matter how well-conceived and well-designed a policy is, the success of its 
implementation will always depend on the abilities of those doing the work. 
A commitment to sector specific policies involves not only a commitment to 
breathing life into the DTI, but also a commitment to investing in the skills 
and capabilities of those who work there. The weakness of British management 
is not confined to the private sector. 

Local and regional policy 

The analysis discussed above suggests not only that policy should be sector 
specific, but that it ought to be conceived and implemented at as local a level 
as possible. There is little to gain and almost everything to lose from an 
excessive centralisation of policymaking in Whitehall. Our view is that local 
institutions ought to be given an explicit responsibility for encouraging economic 
development, and the means with which to do so. These powers and 
responsibilities might be vested in existing local authorities, or in a set of 
regional development associations. We prefer the latter arrangement for 
several reasons. Local authorities are, in a sense, 'too local', and even the most 
tight geographical clustering of economic activities is bound to span several 
local authority jurisdictions, creating unnecessary tensions and coordination 
problems. Further, local authorities often lack the expertise to take on an 
economic development role, and it seems to us sensible to consider clustering 
that expertise together into a smaller number ofrather broader-based institutions. 
Finally, regional development authorities can (and should) operate at 
one remove from the political process, having the freedom to resist the 
politically convenient and the resources to do what many politicians cannot, 
namely, think about the long term.Ofcourse, local institutions cannot operate 
simply as they wish. We see two roles for a central body (such as the DTI) to 
play in the policy process. First, regional development agencies must account 
for their actions on a regular basis, and steps need to be taken in order to 
prevent them from being captured or corrupted by local businesses. Second, 
the various plans and policies initiated by each region must be woven together 
to form a coherent whole that is in line with national priorities. 

It goes without saying that creating a few regional development agencies 
will not solve the U.K.'s economic problems. There are a range of different 
activities which constitute a market's infrastructure, ~nd many ofthese must 
be provided by specialised institutions. There are two models that might be 
used by policymakers to encourage the growth ofthese. One involves providing 

19 



tax (and other) incentives that encourage private agents to undertake desired 
activities (this might be called a 'bottom up' approach), while the otherinvolves 
using subsidies (and other devices) to create these institutions directly (a 'top 
down' approach). For example, to encourage the development of financial 
institutions which support small-scale, high-tech firms, one might provide tax 
breaks for large firms or financial institutions that are willing to invest in 
small firms, or one might simply create local development banks and provide 
them with an initial capital base to work with (together with a Board of 
Directors drawn from the public and private sector to ensure that that capital 
base is used in a businesslike manner). Although we do not know of enough 
evidence to make a clear choice between these two types of policy, our 
inclination is to prefer the 'top down' approach. It seems to us that there are 
often situations in which habitual practice and uncritically examined conventional 
attitudes make private sector agents unwilling to undertaken certain 
types of activities. If the lack of private sector activity really is caused by 
market failure, leading by example is the only way forward. 

There is no doubt that regional policy must remain near the centre of any 
industrial policy programme introduced into the U.K.. However, it seems to 
us that policymakers must accept the unevenness of the process of economic 
growth by encouraging the clustering of economic activity, not resisting it. 
This, in turn, means that change in crisis sectors must be fa-::ilitated and not 
slowed, no matter how pressing the short-run concern with preserving jobs 
and rectifying the balance of payments. Clusters must be encouraged to grow 
and flower and then they must be allowed to decline and give way to other 
activities, and there is an important role for regional policy in both the birth 
and death process. What is more, it is a role that regional policymakers are 
unlikely to be able to play effectively unless they have the freedom to operate 
at some remove from short-term political pressures. For these reasons, we see 
some sense in delegating many of regional policy responsibilities to regional 
development associations that have some independence of action and are not 
directly controlled by ministers or civil servants in Whitehall. 

This decentralised model roughly corresponds to what currently happens 
in the area ofland-use planning, but we think that land-use planning must be 
much more explicitly tied into industrial policy programmes than has been 
the case so far. What this means is that the DTI must play a much greater 
role in the development and overseeing of planning decisions, and that this 
integration of the activities of the DOE and the DTI must extend down to the 
regional level. Again, regional development agencies seem to be the appropriate 
level at which to decentralise policy action, but it seems clear that ifland 
use planning is to be used to help stimulate economic activity, then regional 
development agencies must have a stronglocal presence on their management 
boards, and they must be held accountable to, though not dictated by, local 
interests. 

20 


Related publications from the Fabian Society 

A Public Services Pay Policy. William Brown and Bob Rowthorn. November 
1990. £3.00 

No British government since the war has adopted a coherent policy on public 
sector pay. A Public Services Pay Advisory Commission, on the model of ACAS, 
should be created, to provide the data on which comparability exercises and 
pay negotiations can be carried out. 

Controlling Inflation: two views. Gerald Holtham and Neil MacKinnon. 
July 1990. £3.00 

To avoid an anti-inflationary policy which places an unacceptable burden 
on British manufacturing industry, Holtham proposes that tax brackets 
should be indexed to target rather than real inflation, and MacKinnon argues 
for 'Euro-index-linking' for prices charged by private monopolies. 

Sustainable Development: greening the economy. Michael Jacobs. July 
1990. £3.00 

Sustainability requires a new set of economic objectives, and constraints 
on the market economy. Government action is required to achieve change on 
the scale required, and to mitigate inequitable effects. 

Income and Wealth in the 1980s (1990 edition). Thomas Stark. July 1990. 
£15.00 
Stark uses government statistices to contradict claims that the poor have 
done well out of economic growth under the Conservatives. 

The Reform of Direct Taxation. Fabian Taxation Review Committee. 
February 1990. 

This report, the first major enquiry into the British tax system since the 
1970s, dispels the myth that the Conservatives have delivered real tax cuts, 
and argues for practical and attractive alternatives. 

Available from the Fabian Society, 11 Dartmouth St, London SW1H 9BN 
Please make cheques payable to the Fabian Society. We are not able to invoice 
orders less than £25.00. Credit card orders accepted during office hours. 


Targeting competitive industries

....................................................... 


Paul Geroski and Kenneth Knight argue that to develop 
competitive advantage, clusters of geographically-concentrated 
industrial activity should be encouraged. The unevenness of 
economic development must be accepted and encouraged by· 
national government, not resisted. 

Policy must also be sector specific. This is often described-and 
then dismissed -as 'picking winners'. A genuine industrial 
policy does not so much pick winners as back them by creating 
the right market infrastructure. 

Policy needs to be implemented locally-Whitehall cannot know 
local conditions and needs. The authors recommend that much 
greater emphasis be placed on the role of local and regional 
agencies, following the succesful practice in Italy and France. 

The Fabian Society 
brings together those 

who wish to relate 
democratic socialism 
to practical plans for 
building a better society 
in a changing 

world. It is affiliated 
to the Labour Party, 
and anyone who is 
eligible for membership 
of the Labour 
Party can join; others 
may become associate 
members. For details 
of Fabian membership, 
publications and activities, 
write to: 
Simon Crine, General 
Secretary, Fabian Society, 
11 Dartmouth 
St, London SWlH 9BN.

£3.50