Oxford Development Studies, Vol. 29, No.

1, 2001

Technology Support for Small-scale Industry in

Developing Countries: A Review of Concepts and
Project Practices

HENNY ROMIJN

ABSTRACT The paper is a review of approaches towards institutional technology support for
small-scale manufacturing enterprises in developing countries since the early 1970s. Early
programmes tended to suffer from a number of weaknesses, stemming from a limited conceptu-
alization of technology and an inadequate understanding of the role of the small-scale sector in
industrial development more broadly. There was also a lack of practical experience with project
implementation. However, in recent years important advances have been made on all these
fronts. Four features of recent technology assistance programmes that have tended to be
associated with success are discussed, and illustrated with evidence from different projects.
Broadly, successful projects: (a) embrace the notion that durable competitiveness of small
producers in a competitive economic environment requires that they develop internal capabilities
to effectively assimilate, use and adapt product and process technologies; (b) are demand-
driven; (c) target the assistance to groups of producers with common interests and problems,
and help them to organize themselves in collective bodies that can evolve into self-help
institutions; and (d) include appropriate incentive structures based on market principles.

1. Introduction
Small enterprises typically make a large contribution to manufacturing employment in
poor countries. However, the developmental contribution of most of them is limited to
generating subsistence employment ª of last resortº . Hence, in the face of fast labour
force growth and limited employment absorption in other sectors, developing country
governments have mounted efforts to improve productivity and earnings in these ® rms.
This has spawned a plethora of policies and programmes, and an almost boundless
literature documenting them.
This article is a review of one speci® c subset of that literature, namely studies that
shed light on measures aimed at improving technological performance. This aspect has
recently received less attention than other forms of assistance. Although much is being
written about so-called business development services (BDS), of which technology
support is a component, most of that literature has a broader focus, dealing with
management, organization, sales, employment, income and general quality issues. Few

Henny Romijn, Department of Technology and Policy, Faculty of Technology Management , Eindhoven
University of Technology, PO Box 513, 5600 MB Eindhoven, The Netherlands.
The author is grateful to Mike Albu, Jeffrey James and three anonymous referee s for useful comments on earlier
drafts, and to Ruud Picavet for inspiring discussions and literature guidance.

ISSN 1360-0818 print/ISSN 1469-9966 online/01/010057-20 Ó 2001 Internationa l Development Centre, Oxford
DOI: 10.1080/13600810120016209
58 H. Romijn

publications contain technical details about upgrading of products, processes and

organization, the support needed to bring about such improvements and the effective-
ness of delivery mechanisms.
Yet technological competence is an especially important determinant of small
manufacturers’ ability to hold their own in a context of liberalization and increasing
integration of manufacturing into global networks. Many of their markets, even tra-
ditional ones, are changing rapidly. In this situation, a lack of capability to produce
ef® ciently, meet deadlines, or upgrade product quality and design spells defeat.
The main objective is to identify important common factors behind the success and
failure of technology assistance projects of public agencies and non-governmental
organizations (NGOs).1 Insight into the question of what constitutes ª best practiceº in
this ® eld is still sorely lacking. The rationale for technology support for small ® rms is
elaborated in Section 2. Lessons from past debates and interventions are drawn in
Section 3. Section 4 treats recent developments in the debate about small industry
promotion more broadly, as these hold important lessons for current technical assist-
ance practices. This sets the scene for the discussion of recent approaches to project
design and implementation in Section 5, focusing on key principles behind the success
of new approaches that are likely to have more general validity. These are illustrated
with project case studies in Section 6. Conclusions are given in Section 7.
The emphasis of this review is on small-scale ª workshopsº , so common throughout
the developing world. They typically employ between ® ve and 50 people, including
some hired labour. They have some division of labour and use basic machinery, but
their managerial practices and technological characteristics are worlds removed from
those of modern large companies. They tend to be engaged in well-established or even
traditional activities, making: basic wooden furniture; simple metal products such as
tractor-trailers, ploughs and window frames; leather goods; local construction materi-
als; or processed foodstuffs such as tofu and pasta. The customer base usually includes
large numbers of the poor and lower middle class. Except in some Asian newly
industrialized countries (NICs), few are in the forefront in new high-tech sectors and
successful exporters. Only a small minority engages in formal R&D.2 Self-employed
workers such as traditional blacksmiths, potters and weavers, and very small family-run
ª micro-enterprisesº operating in the informal sector are not part of this review. The
support programmes mounted for them are focused more on poverty alleviation than
on business growth, and a discussion of these programmes raises different issues.3

2. The Importance of Technology Support

The importance of institutional (i.e. public and NGO) technology support to industry
in general derives from the argument that investments in technological upgrading and
learning are subject to widespread market failure. Without corrective intervention,
private companies will under-invest in technological effort relative to the social opti-
mum (Stoneman, 1995; Dasgupta, 1987). First, new knowledge created by a company
tends to leak out to competitors, reducing its incentive to innovate. Second, invest-
ments in technological upgrading are subject to risk due to uncertainty. Third, new
knowledge often becomes substantially useful only in fairly large amounts, so that the
required minimum investment outlay and risk will sometimes exceed the capacity of
individual investors. Scale advantages also occur in the form of bene® ts from interactive
technological learning, which creates positive externalities by sparking new knowledge
and ideas. Co-ordinated action can: balance innovators’ and imitators’ interests; help
decide which industries and projects should receive priority in the national research
 Technology Support 59

effort; help mobilize suf® cient resources to that end; get private actors to co-operate
towards the achievement of those goals; and reduce risk through improved access to
information and knowledge.
Support policies are particularly vital in less-developed countries (LDCs), where
problems of appropriability, co-ordination and information are especially severe. The
® nancial, research and human capital base of ® rms tends to be weak. Most of their
technological efforts are relatively easy to copy because they are, for the most part,
adaptive and incremental, not amenable to patenting. Inter-® rm communication and
co-operation are cumbersome because of poorly developed physical infrastructure, and
risk is high in an information-poor environment. Even relatively easy tasks become
dif® cult, costly and risky (Lall & Teubal, 1998, p. 1371).
Small ® rms generally ® nd it more dif® cult to cope than medium and large ones.
Their technological capabilities are weaker, and they are usually not in a position to get
funding for innovation on reasonable terms through the regular ® nancial system. Owing
to resource constraints, their information search efforts and investments in training and
education tend to be quite restricted. Lack of ® nance, skill and expertise combined with
high uncertainty also lead to risk-averse behaviour, depressing investments in techno-
logical effort. In some industries, modern, ef® cient techniques of production suitable
for a small scale of operation are also lacking. Moreover, problems associated with
economies of scale affect small ® rms worse than large ones. While the latter can to some
extent overcome scale problems and capture externalities from inter-personal interac-
tion within the con® nes of their own walls, small ® rms can create critical mass only
through inter-® rm co-operation and market exchange, which is often precluded by
intense competitive rivalry, lack of trust, poorly functioning factor markets and under-
developed private-sector services. The large majority of the workshops that are the
subject of this paper are caught in a vicious circle of low-tech, low-productivity
production, unable to embark on substantial technological upgrading efforts without
help.
Potential bene® ts of institutional support to small ® rms are large. By improving
access to information, ® nance and suitable technologies, support can help ® ll the
ª missing middleº observed in the manufacturing size structure in poor countries.
Dynamic small ® rms could also contribute signi® cantly to national exports. Moreover,
clusters of geographically close producers may create signi® cant positive dynamic
externalities (Levy et al., 1999).

3. Early Programmes
The small enterprise sector was ® rmly put on the mental map of LDC policy-makers
in the early 1970s as part of a general disenchantment with industrialization strategies
favouring top-down modernization through expansion of the modern large-scale sector
that had been pursued in the 1950s and 1960s. As the bene® ts of ª trickle-downº were
apparently limited, income-creation approaches based on direct targeting of poorer
sections of the population gained widespread favour. The International Labour Organi-
zation took the lead in documenting the precariousness of those working in small
enterprises, and the serious constraints they faced (e.g. Sethuraman, 1981).
Early technology support programmes predominantly adopted a ª supply-pushº
approach (UNDP et al., 1988). It was thought that the provision of a variety of services
could overcome producers’ resource constraints and thereby foster competitiveness.
Many countries set up state-run small and medium industry development organizations
(SMIDOs) charged with providing services such as technical and management training,
60 H. Romijn

marketing assistance, advice about technology choice, assistance with technology pro-
curement and provision of subsidized ® nance. The scope of the programmes was
generally broader than technological upgrading alone. However, there were also a
number of bodies, both state and NGO, focusing speci® cally on technology. These
programmes achieved some good results, but they were also beset with a large number
of problems. Here, we only discuss some issues that bear speci® cally on technological
support. It is inevitably a broad-brush approach that does not do justice to the
speci® cities associated with particular programmes across different countries.

3.1 Underlying Conceptual Problems

In the early years of enterprise promotion, the precarious existence of many workshops
was thought to stem from larger ® rms being more ef® cient as a result of economies of
scale and the use of more productive techniques. The question of whether small
enterprises would be able to operate ef® ciently at all was hotly debated (e.g., Little et
al., 1987; Goldar, 1988).
With hindsight, the scope of this debate appears somewhat limited. The technologi-
cal problem of small producers was predominantly seen to be one of lack of suitable
machinery and equipment, in line with the literature about technology and develop-
ment in the 1970s (e.g. Sethuraman, 1977; Harper, 1984). Thus, their lack of
competitiveness was attributed to high relative unit cost stemming from lack of
appropriate hardware. Little attention was paid to the possibility that human skills,
knowledge and organizational capacity to use, adapt and improve the hardware
ef® ciently might also be in short supply and might constitute an equally crucial
constraint on competitiveness. Moreover, the focus was on cost levels in individual
small units versus those in larger ® rms, i.e. there was no recognition of potential
competitive advantages that might accrue through exploitation of cluster or network
synergies involving groups. Meanwhile, the preoccupation with the problematic
horizontal (competitive) relations of small ® rms with their larger counterparts pre-
vented most researchers and policy-makers from exploring growth possibilities through
development of complementary (vertical) relations.4

3.2 Lessons Learned

3.2.1 `Appropriate’ technologies and their limitations. In the debate on aid for small ® rms,
some claimed that ef® cient technologies suited to small-scale operation did not exist.5
Others suggested that such technologies did indeed exist in several industries, but that
there were serious disincentives to their adoption due to unfavourable macroeconomic
policies favouring use of large-scale modern techniques and boosting demand for their
products (Stewart, 1987; Bhalla, 1985; Stewart & Ranis, 1990; Haggblade et al., 1990).
Both arguments carried some weight. The ® rst led policy-makers to promote
research in institutes aimed at the development of small-scale ef® cient technologies.
Various strategies were identi® ed (Bhalla et al., 1984; James, 1989). Foreign donors
were much involved in such projects; it was the golden age of the appropriate
technology movement.
This approach achieved a degree of success, though many so-called ª appropriateº
technologies failed at the commercialization stage. One big lesson was that technology
development in the public domain and its subsequent diffusion in top-down, ª supply-
pushº fashion was not an effective model.6 Technology institutes often had competent
engineers, but they knew little about the requirements of poor producers and communi-
 Technology Support 61

ties, were usually located far from them and had little awareness about social, economic
and cultural issues. Even where that was not the case, signi® cant communication
barriers between developers and prospective users tended to preclude effective infor-
mation exchange.
The successes (mainly in East Asian countries, especially Japan, Taiwan and China)
demonstrated that technologies have to be developed in close collaboration with the
prospective users through a process in which the users can take signi® cant control over
the direction of the project and, in effect, assume ownership of the technology. Another
essential condition is that equipment producers (i.e. local capital goods makers) have to
be involved at an early stage of development, since they (rather than the users or the
technology institutes) have to take care of repair, maintenance, replication and
modi® cation in the light of practical experience by the users. Technologies are rarely
perfect when they come ª off-the-shelfº . Often, several rounds of feedback between
developers and users are needed to improve and adapt them in iterative fashion. For
this reason, the best technology development model involves close and ongoing interac-
tion between users, institutes and producers as equal partners with complementary
knowledge and skills.7

3.2.2 Lack of incentives and competitive pressures. Early programmes had major design
and implementation ¯ aws. Supplying crucial missing ingredients to small companies
is a good thing in underdeveloped economies where well-functioning markets for
essential services rarely exist, but in practice it often bred complacency among
recipients. Sometimes small entrepreneurs took the assistance for granted simply
because they belonged to an ª underprivileged classº . This was especially the case when
projects providing assistance with product design and quality simultaneously provided
an assured public outlet for the produce, leading to the removal of competitive
pressure.
There were other incentive problems. Supporting small companies is not as reward-
ing and glamorous as assistance to bigger business. It is less visible, does not bring
political in¯ uence or important contacts and needs much effort to achieve good results.
Not surprisingly, the most successful projects involved committed individuals who were
not primarily driven by high monetary rewards.8 In addition to weak incentives, many
programmes lacked effective sanctions for inadequate performance of assistance agen-
cies. State-run or parastatal organizations functioning with ª soft budget constraintsº
did not have to rely on commercial sources of revenue for their continued existence.
The technological assistance of the SMIDOs has had even more modest results than
the programmes run by specialized technology development institutes and NGOs, and
it is not worth dwelling much on it.9

3.2.3 An unconducive macro-environment. The ª macro-incentives approachº to technol-

ogy improvement mentioned earlier gained considerable in¯ uence in the early 1980s, a
bit later than the appropriate technology movement. Studies were undertaken that
linked inappropriate technology choices to biased incentive structures caused by ad-
verse macro-level and ª mesoº -level policies commonly pursued in developing econom-
ies. Much light was shed on why the effectiveness of the early programmes was so
limited. It is probably true that their lack of success can be traced as much to problems
in the general economic and institutional environment in which they had to function as
to ¯ aws in conception, design and implementation at the micro-level.
In particular, the problem of lack of market opportunities faced by small enterprises
62 H. Romijn

was at least partly caused by the general economic malaise and the lack of growth
possibilities for small companies in many countries pursuing import-substitution strate-
gies that favoured modern, large-scale forms of production. Overvalued local currencies
and cheap credit for large import-substituting companies led them to establish import-
intensive integrated production facilities. There was little incentive to establish back-
ward linkages to local companies. The impact of policies such as tax and ® nancial
incentives and local content regulations was generally modest.10
To the extent that the incentives approach advocated ª getting the prices rightº , it
tied in well with the sort of reforms that countries had to introduce when they
embarked on structural adjustment. However, it went well beyond pointing towards
biased factor prices, drawing attention to major institutional, legal and structural
constraints preventing the small industry sector from ¯ ourishing, and that could not be
remedied quite so easily.

3.2.4 Lack of integration of small ® rm aid with national industrialization strategies

The macro-incentives approach was particularly useful in raising awareness about how
small industries were linked to, and affected by, what was happening in the wider
economy. Curiously, the studies about small-scale enterprises and the informal sector
generated in the 1970s largely ignored wider industrialization strategy issues, while
studies about countries’ industrialization in general tended to disregard the small
industry sector. This lack of integration in research had its effect at the policy level.
Policy-makers generally did not properly integrate the promotion of their small
industry sector within broader industrialization strategies and objectives, almost as if
it existed in a vacuum. Many countries still suffer from such legacies, even those that
have made a serious effort at industrial policy reform (see, e.g. King, 1996; Oladeji,
1998).

4. The 1990s: New Directions in Concepts and Practices

The practice of technology support has moved forward in recent years, especially in
the 1990s. This partly re¯ ects changes in the broader debate about small industry
development and its role in industrialization in general. This section highlights some
important emerging lessons for the design and implementation of technology sup-
port projects. Sections 5 and 6 focus on innovative project approaches that have
absorbed these lessons, along with the ® ndings from earlier project practices discussed
above.
The debate has clearly been in¯ uenced by the major changes in the policy climate
since the late 1970s. The rise of neo-liberal thinking de-emphasized state involvement
in the economy, spelled the demise of inward-looking approaches to industrialization
and advocated increased openness to trade and foreign investment. The effects of these
policies on local industrialization have varied. Some of the relatively advanced develop-
ing countries in East and Southeast Asia and Latin America are increasingly being
integrated into large regional and/or global trade and production networks that are
steadily growing in importance. In contrast, in many of the truly low-income econom-
ies, where liberalizatio n took place in the context of heavy structural adjustment
programmes, industrial development has suffered (Lall, 1999). Sub-Saharan Africa’ s
share in global manufacturing, already very low at 0.4% in 1985, declined to 0.3% 10
years later (UNIDO, 1996, p. 22).11
In both cases, however, the old notion that economic prospects for small industries
 Technology Support 63

would derive mainly from their ability to achieve competitive cost levels with occasional
injections of appropriate equipment proved increasingly inadequate in the context of
the big shift in the macroeconomic climate and the forces pushing for global integration
and marginalization . There are important common elements in the changing literature
about small industry development across major regions. In particular, the old view that
producers could improve their competitiveness by absorbing technical improvements
designed elsewhere is beginning to be replaced by a more dynamic notion of competi-
tiveness, one that depends on small ® rms’ own internal capacity to make an independent
and unique contribution to local technical progress on an ongoing basis. This is
happening in Asia, Latin America and Africa.

4.1 Advanced LDCs: Small Firms, Technical Capability and Systemic Competitiveness
The change is perhaps most evident in the most advanced and fast-industrializing
economies of Asia and Latin America (especially the East Asian NICs), where there is
much potential for small ® rms to join in export-oriented industrialization with increas-
ing participation in international trading and production networks (Levy et al., 1999).
In a dynamic industrial environment, small ® rms’ competitiveness begins to be per-
ceived as an integral part of national competitiveness. This is mainly due to the
perceived complementarity of the activities of large and small undertakings. The empha-
sis is on the need for local specialized subcontractors to customers downstream in the
ª value chainº , who can react quickly and ¯ exibly to their changing requirements and
begin to play a role in the design and adoption of technological improvements. Thus,
the competitiveness of small companies is increasingly being perceived in terms of their
internal capabilities to choose, use, adapt and develop technology. Such capabilities are
essential to foster competitiveness in a fast-evolving environment which continuously
places new demands upon large and small ® rms alike (Meyanathan, 1994; UNIDO,
1996, pp. 53± 56).
Consequently, the small enterprise debate in these countries is beginning to link up
with the debate about acquisition of technological capability in industrial develop-
ment.12 The latter has largely replaced the old static ª choice of techniqueº framework
that underpinned the early small ® rm support programmes already discussed. Its point
of departure is that the existence of adequate local skills and knowledge for incorporat-
ing more advanced technologies in developing countries cannot be taken for granted.
Whereas technological hardware (equipment, blueprints) can be transferred, the capa-
bility to make use of that hardware has to be developed through a gradual learning
process resulting from purposive efforts to assimilate, adapt and modify the new
technology. Many of these efforts take the form of small ª shop-¯ oorº improvements
rather than formal R&D. Unlike the earlier choice of technique material, the capability
approach sees technological constraints primarily in the lack of human knowledge, skills
and organization rather than machines, and it views technological progress as endoge-
nous to ® rms and within developing countries more broadly. This approach holds
important lessons for the design of technology support projects for small ® rms.
The trend towards a more dynamic interpretation of small ® rm technological
performance and competitiveness is evident even in countries like South Korea and
Singapore, whose earlier policies were biased towards the promotion of large-scale
® rms. There has been a remarkable shift in their policy-stance since the early 1980s,
when they began to experience the dif® culties associated with advancing into higher
technology-based manufacturing without an extensive local subcontractor network
(Wong, 1994; Lee, 1992; Leipziger & Petri, 1993; Chon, 1996; Chung & Park, 1998).
64 H. Romijn

4.2 Middle-income LDCs: Dynamism in Small Industry Clusters

Another important contribution with implications for technology support has come
from studies emphasizing the ef® ciency of groups of small ® rms and their interactions in
networked geographical clusters which may foster collective competitiveness (e.g.
Humphrey & Schmitz, 1996; Schmitz, 1995; Nadvi, 1996; Tewari, 1996; Rabellotti,
1995; Ceglie & Dini, 1999; World Development, September 1999). Much of this
literature is focused on middle-income developing countries in Asia and Latin America,
such as Pakistan, India, Indonesia, Brazil and Mexico, but there are also some
contributions dealing with lower-income African countries. The main concern is again
the internal economic dynamism of small producers, but the emphasis is on inter-® rm
dynamics rather than intra-® rm learning. Drawing on writings about ¯ exible specializa-
tion in industrial districts in developed countries (e.g. Piore & Sabel, 1984; Best, 1990;
Pyke & Sengenberger, 1992), the suggestion is that clustering in developing countries
can help sustain long-term competitiveness among participating ® rms. The writers
appear to presuppose that clustering may also enhance capability to initiate and diffuse
technological improvements, though this claim has remained largely unsubstantiated.
The innovation and learning effects have not been investigated systematically.
ª Learningº crops up in some studies, but what this entails in practice remains unclear
(Albu, 1997).13

4.3 Poor LDCs: Technological Capability Building by Individual Producers

A ® nal important contribution has come from studies about the importance of internal
technological capability acquisition in individual small industrial units. Many are set in
lower-income countries with poor growth prospects, mostly but not exclusively in
Africa. Here, the concept of small ® rm competitiveness is evolving more under duress
than as a result of emerging opportunities because of international competition from
more advanced economies. The large-scale industrial sector has all but collapsed and
foreign investors have failed to ® ll the gap in a big way. Small companies are being seen
as a potential force for industrial regeneration and increased attention is being devoted
to questioning whether they can function as an engine of industrial growth and become
a source of competitive advantage in their own right. There is growing concern that
such regeneration cannot, and should not, be built on low wages, dismal working
conditions and paltry pro® ts, at least not beyond the short term. Acquisition of more
advanced technological capabilities is beginning to be perceived as a major requirement
for escaping from the low-wage, low-skill scenario. It is also recognized that this will
require a supportive policy environment, one in which the small ® rm sector is fully
integrated in the design and implementation of industrial support programmes and
technology policy (King & McGrath, 1999; Wangwe, 1993; Oladeji, 1998; Oyelaran-
Oyeyinka, 1997; King, 1996; Maldonado & Sethuraman, 1992; Massaquoi, 1995;
Romijn, 1997, 1999; Smillie, 1991; Appropriate Technology, 1997).

5. Innovative Approaches at the Project Level

The insights emerging from the recent conceptual debate about the role of small ® rms
in development are beginning to be incorporated into the design of technology assist-
ance projects, along with lessons learnt from accumulating hands-on experience with
project implementation by practitioners (Section 3). There is reason to believe that the
 Technology Support 65

success record of some of the more recent programmes is better than those of earlier
ones.
Unfortunately, even now, only a few project studies have systematically pinpointed
the main features that are believed to underpin their success. Hence, the discussion
cannot go much beyond identifying some broad principles which can be distilled from
the reviews in Section 3 and 4, and which at the same time can be associated with
success in assistance projects in practice.

5.1 Acquisition of Indigenous Technological Capability

Perhaps the most important common factor central to improved outcomes of technical
assistance projects is that sustained improvement in the competitive position of small
producers must come from their acquiring internal technological capability to initiate
and pursue adaptations and improvements to products, processes and production
organization on an ongoing basis. Humphrey & Schmitz (1996) hint at this requirement
when they state that competitiveness of small producers is a process rather than a state.
Therefore, they argue, one-off improvements are of limited use; projects need to aim
for ª cumulative bene® tsº .
The implication for the design of technological assistance projects is that they must
initiate and facilitate a process of change which creates opportunities for producers to
engage in continuous development of their technological knowledge, skills and organi-
zation. The Donor Committee on Small Enterprise Development, an umbrella group
of big aid agencies which has organized much of the ongoing discussion, refers to this
as ª indigenous technology developmentº (Donor Committee on Small Enterprise
Development, 1997, pp. 39± 40).
In this approach, the ultimate project objective is not the one-off design and
adoption of improved technological ª hardwareº . That is not to say that its introduction
should become unimportant or redundant. Making a mechanical lathe available to a
woodworking shop, or showing local metalworkers a prototype of an improved cooking
stove for low-income households, obviously constitutes a developmentally bene® cial
policy intervention in its own right. However, rather than viewing the supply of
equipment or designs as ® nal objectives, their introduction into a local business
community should also, and perhaps even primarily, be seen as a means through which
small producers can master new technical and organizational skills that will strengthen
their ability to introduce other product and process innovations on their own initiative at
a later stage (Jeans, 1999).14
The adoption of this notion has obvious implications for the delivery of assistance.
Rather than using external experts to come up with a perfect design for a new artefact,
an appropriate adaptation of a foreign artefact to local conditions, or a quick ® x to a
technical production bottleneck in a small workshop, projects that aim to stimulate
indigenous technology development must use those experts as teachers and facilitators.
Producers should participate in design, adaptation and problem-solving processes so
that these will create possibilities for them to build their own design skills (Jeans, 1999).
Another way in which projects can help producers to learn is by creating an
ª information-rich environmentº , facilitating access to knowledge and information that
can form inputs into the learning process (Levy et al., 1999, p. 243). Examples include
projects sponsoring courses, organizing ® eld trips by producers to more advanced
factories, facilitiating the use of specialized consultants, helping producers to participate
in trade fairs and promoting information-sharing among ® rms.
66 H. Romijn

5.2 Demand Orientation

An effective way in which assistance agencies can trigger and sustain indigenous
technological capacity building is by taking a demand-led approach. That is, project
interventions have to start by identifying a new market channel for the target producers
and establishing concrete possibilities for them to establish themselves in that new
market (Humphrey & Schmitz, 1996; Tendler & Amorim, 1996; Dawson, 1999).
There are different ways in which this has been achieved. Most of the evidence
concerns successful negotiation by assistance agencies to help small producers (or
groups) to become suppliers in public procurement schemes, but assistance agencies
have also successfully linked small producers to new private sector clients, not only at
home but also abroad (Levy et al., 1999).
This customer-focused approach has at least two major advantages over the earlier
supply-side project interventions reviewed in Section 3. First, access to a dynamic
market in which producers are required to deliver products with improved designs and
quality, to pay attention to standardization, to meet deadlines, to control their unit cost
of production, and so on, provides a ® nancial incentive for small ® rms to invest in
efforts to make the improvements needed to live up to the expected standards. Second,
it tends to focus assistance quite tightly around supplying the critical missing inputs
required to overcome producer bottlenecks experienced in the process of trying to meet
customer demand and responding to their complaints. The assistance can be applied at
once; hence there is an immediate test of its practical relevance. This avoids the danger
of projects supplying a broad array of services that will ultimately not be useful in
practice. A particularly effective way to ensure that small producers get the right type
of technical assistance has been an arrangement in which a representative of (typically
large) clients will agree to engage in ª buyer-mentoringº for a while, training small
suppliers and providing technical consultancy, with the assisting agency meeting part of
the costs involved (Dawson, 1999; Wong, 1994).

5.3 Organization in Groups of Clustered Firms

In recent successful projects, the practical organization of the type of assistance
outlined above has involved targeting groups of geographically clustered producers in
the same industry, rather than scattered individual small companies with varied activi-
ties (Humphrey & Schmitz, 1996). Collective support to clusters has advantages. First,
it is simply more cost-effective for agencies to concentrate on the problems faced by
groups of similar producers in a few speci® c localities. Keeping costs down is especially
important since aid projects are increasingly under pressure from donors to ensure
® nancial sustainability of projects within a short period (Dawson & Jeans, 1997). It also
allows agency personnel to develop in-depth expertise about the technical, marketing
and other problems of speci® c industries in particular regions.15 Moreover, an agency
based close to its clients can develop a close working relationship with them, over-
coming mistrust and establishing its credibility.
Joint support can also stimulate intra-® rm learning through common problem-
solving, information exchange and overcoming lumpiness problems (e.g. establishment
of a common facility centre by the assisting agency, or joint investment by producers in
expensive equipment which cannot be operated pro® tably by only one company).
Finally, organization in groups can overcome contractual problems with clients and
enforce complaint behaviour by the companies participating in a project. Large clients
prefer to deal with an association of small manufacturers rather than to share out orders
 Technology Support 67

among individual members, monitor their progress and deal with problems relating to
order ful® lment. Furthermore, the manufacturers’ association can pressure its members
to perform according to the contract, e.g. by making the group jointly responsible for
delivering the outputs and honouring warranty claims.

5.4 Incentives that Promote Sustainability

Flawed incentives, both for assistance providers and bene® ciaries, were one of the chief
causes of failure in early support projects. In some of the recent projects, much
attention has been devoted to the careful design of a more appropriate incentive
structure (see especially Tendler & Amorim, 1996). This entails a combination of
ª carrotsº , potential rewards that motivate the participants to act, and ª sticksº , sanc-
tions that come into operation when they fail to do their best. Effective incentive
systems appear to be those that rely as much as possible on market discipline, especially
a payment system that rewards effective services offered by support staff and that
penalizes failure. Getting bene® ciaries to pay for (at least part of) the services offered
is also a way to ensure that they value the support and that they will utilize it.
That is not to deny that some project activities do require initial subsidies to
overcome market failures commonly associated with technological learning (see Section
2). However, this is not an argument for mounting interventions that distort, or even
completely replace the market by doling out permanently subsidized services of dubious
value. Linking project activities with market forces ensures that projects ª remain on
trackº by getting the right signals about their activities from their client base and that
the project staff remain motivated. Only then will projects be able to bring about the
cumulative effects that Humphrey and Schmitz refer to. Some projects now do not
merely aim for cumulativeness in terms of sustained bene® ts for small producers and
their customers (i.e improvement processes driven by ongoing learning); they also apply
the notion to delivery of the project services themselves. Dawson & Jeans (1997) argue
that projects must evolve institutional forms of self-help that will, over time, start to
function independently of external aid agencies. Organizing producers and helping
them to build strong local collective institutions (see above) can contribute to the
achievement of this goal.

6. Some Illustrations from Speci® c Projects

We illustrate the above points with some examples ranging from projects in advanced
developing economies in East Asia to schemes in low-income countries in Africa.
Among the higher-income industrializing countries in East and Southeast Asia and
Latin America, Singapore, Korea and Taiwan have the most elaborate technical
support systems for small manufacturers. These appear to be used quite intensively,
although most of the support is aimed at more advanced ª small/mediumº ® rms rather
than the small workshops that are the central focus of this article. The usefulness of
institutional schemes in other higher-income developing countries appears to be more
limited. For example, interactions with private parties, especially buyers and suppliers,
were much more important for technological upgrading among small exporters in
Indonesia and Colombia (Levy et al., 1999).
Here we focus mainly on an East Asian programme with a more ª workshop-likeº
focus, drawing on other NIC schemes where relevant. Singapore’ s Local Industry
Upgrading Programme (LIUP) was designed in 1989 by the Economic Development
Board (EDB) as part of its Small and Medium Enterprise Master Plan, to strengthen
68 H. Romijn

the technological capacity of suppliers to transnational corporations (TNCs) in pre-

cision engineering, which is crucial for the country’ s competitiveness in high-tech
electronics. 16 It has been reported that various signi® cant forms of technological
learning have taken place, including: learning through direct know-how transfers;
learning through feedback provided by stringent quality/performance control by the
TNCs; learning through exposure to information resources provided by the TNCs; and
learning through, and as a result of, investments in capital equipment and other forms
of new technology by the small ® rms that they would not have made in the absence of
their relationship with the supporting corporation.
An example of a project in a less prosperous setting is a public procurement scheme
for school furniture in the backward Brazilian State of CearaÂ.17 It was organized by the
Industry and Commerce Department (SIC) of the state government, together with
SEBRAE, the Brazilian SMIDO, after the central government decided to seek out
alternative, small-scale scale suppliers of wood products in regions that had been badly
hit by drought, as a sort of alternative to a public works programme. The success of this
project is evident from the upgraded skills, knowledge and management capabilities of
the participating enterprises, as well as substantial investments made in power tools,
expanded capacity, creation of many backward and forward linkages, and vastly
increased employment in local industry. The most powerful evidence of project sustain-
ability is the fact that the producers were able to use their sales to SIC as a starting point
to enter new markets and diversify their customer base. Five years after the start of the
project, 70% of their output was already going to the private sector. The association
had become an important institution locally, initiating many developmental activities
without help from the original assistance agencies.
We also draw on the experience of the Farm Implements and Tools (FIT) pro-
gramme implemented in Kenya by the International Labour Organization and a Dutch
NGO called TOOL.18 It targets (among others) small manufacturers of farm equip-
ment. Conditions are hostile: Kenya is currently a stagnant economy at a low stage of
economic, technological and infrastructure development. Indications of success in this
project have been that the participating small ® rms have continued within the pro-
gramme over a period of time and that they are paying at least part of the cost of the
services. They are also suggesting the introduction of new project activities, such as
help with establishing facilities for equipment testing and with resolving teething
problems encountered when new or improved farm implements are used in practice.

6.1 Demand Creation

Of the three projects, the SEBRAE scheme has been the most explicitly demand-
driven. The project was itself a response to the opening up of public procurement of
basic manufactured products by SIC to small-scale manufacturers, which created a
potential new market outlet that could be readily exploited. After the success of this
project, public procurement has begun to be more widely used as a vehicle for small
® rm support, but the scope for such public sector initiatives is inevitably somewhat
limited. Therefore, the experience of the other two projects is more instructive. They
show that demand-driven projects can also come about as a result of special efforts to
build demand from private agents into the project design. The LIUP scheme has
achieved this by forging close links between TNCs and their suppliers, focusing the
support around practical problems that crop up as the suppliers try to meet the demand
from their customers. The EDB assumes the role of ª network brokerº , an idea that has
also worked well in assistance projects in other countries.19 TNCs are approached by
 Technology Support 69

the EDB to participate in the project. When agreement is reached with a TNC, an
experienced engineer from the company is identi® ed and seconded to the EDB to act
as LIUP manager for 2± 3 years, identifying areas of focused assistance for the TNC’ s
suppliers. A participating TNC takes several small ® rms under its wing, and is expected
to provide training in areas such as management, quality control, process engineering
and industrial engineering through visits, workshops, consulting activities and so on, a
clear example of the practice of ª buyer-mentoringº discussed in Section 5. The EDB,
meanwhile, arranges access for the participating small companies to a variety of
® nancial support schemes operated by and through it.
FIT, too, is strongly demand-driven. According to the project organizers, ª ¼ the
demand by micro- and small-scale enterprises for services is ultimately ® nanced by sales
to their customers, and it is the demand and perceptions of these customers which are
therefore the origin of all sustainable activitiesº (Tanburn, 1996, p. 47). For that
reason, project activities must translate into substantial and reasonably quick improve-
ments in products that bene® t the customers. Unlike in the cases of SEBRAE and
LIUP, ensuring customer co-operation for technological upgrading of suppliers has
been dif® cult in a situation where the farmers whose demand for innovations is to be
stimulated are just as resource-poor as the entrepreneurs who make their ploughs and
hoes. Yet the project has had some success with the organization of regional user± pro-
ducer meetings where improved farm equipment developed by project participants is
displayed. Through these ª innovation fairsº , farmers developed a keen interest in
interacting with the producers, providing them with feedback on the performance of
their improved products and suggesting further improvements.

6.2 Capability Building

Upgrading of small producers’ technological capacity has been essential to satisfying
customer demand for new or improved products in all three programmes. LIUP
strengthens suppliers’ capabilities through a focus on increased operational ef® ciency
(ability to produce according to the exacting time schedules and quality standards
required by clients) and increased ability to perform ongoing incremental improve-
ments in products and processes. It aims to raise technical standards to a level where
the producers can compete successfully with leading foreign suppliers, forming
attractive partners for the local TNC community.
In the SEBRAE project, too, small woodworkers had to advance technologically in
order to reach the production standards called for by the public sector tenders. To this
end, the local woodworkers were provided with technical assistance and consulting
services over a period of time. The customer-driven approach led to a well-focused and
ef® cient form of assistance in which producers could learn incrementally: participants
in the project who encountered technical bottlenecks in the course of trying to ful® l
requirements could call on SEBRAE engineers to help them identify the speci® c
problems which needed to be solved to deliver the required quality at the agreed price
at the right time.
Likewise, FIT’ s objective has been to strengthen producers’ internal capacity to
undertake technological upgrading of their products, ultimately contributing to higher
productivity and incomes of farmers as well. The programme essentially functions as a
facilitator for technological learning in the participating enterprises. Several services
helping to create an information-rich environment (see Section 5) have been developed.
For example, by broadening access to market and product information, producers are
in a better position to analyse their own strengths and weaknesses in relation to others,
70 H. Romijn

learn more about the needs of their users and understand why the performance of their
products can fall short of user expectations. It also gives them new ideas for improve-
ments and for the introduction of new products. One activity is the organization of
group visits by small producers to bigger enterprises elsewhere in the country or abroad,
which has strong demonstration effects. Another is the facilitation of direct communi-
cation with the users of farm implements. ª Brokering workshopsº for groups of
entrepreneurs, initially designed to evaluate the impact of these two initiatives, became
an independent activity in itself because the informal information exchange obviously
met participants’ needs. Finally, help has also been offered to producers to identify new
markets through teaching a simple ª rapid market appraisalº .
Inter alia, provision of an information-rich environment has been a successful
assistance instrument for more advanced small and medium enterprises in higher-
income developing countries, too. Research about assistance to export-oriented small
and medium-sized manufacturers in Colombia, Indonesia and Korea (as well as Japan)
found that such ª broad-based supportº , which leaves ® rms to decide what information
sources might be most useful, was particularly valued by the ® rms. A key precondition
for its success, however, is that the macroeconomic environment should not be grossly
distorted through overvalued exchange rates, rationed credit and foreign exchange, and
so on (Section 3). Such distortions have created unfair advantages for large ® rms in the
past, which no amount of information can offset (Levy et al., 1999).
An advantage of information provision through mechanisms such as trade fairs,
specialized consultants, data banks, catalogues and so on, is that the institutional
requirements for this kind of assistance are relatively modest compared with ª high-
intensityº specialized technical support, which places high demands on the technologi-
cal and organizational capabilities of the supporting institutions. Few developing
countries can deliver such assistance effectively. An exception appears to be Korea,
where the central support providers have good procedures for evaluating the impact of
their programmes, distributing resources between and within agencies, matching supply
and demand of services, and hiring competent staff. But this is hard to replicate
in poorer countries with weaker governance structures. In any case, the Korean
case shows that demand for such specialized services becomes signi® cant only when
industries reach relatively high levels of technological complexity (Levy et al., 1999).

6.3 Organization in Groups

Advantages of group-based support are illustrated by the SEBRAE and FIT projects.
The SEBRAE project has been narrowly targeted at a small collective of small
producers in one particular locality operating in one particular activity (woodworking).
This had a number of advantages. First, in view of the large orders from SIC, which
exceeded the capacity of any one of the individual producers, it was practical to
contract with a group. A producer association was formed with encouragement from
SEBRAE, which assumed responsibility for ensuring product quality and honouring
product warranties, and co-ordinated the activities of the individual members. In case
of default by one of them, the association was responsible for honouring orders and
warranty claims. By making the association the one focal point in the transactions, it
was in the interest of members to monitor each other’ s performance. Meanwhile, the
existence of the association lowered the transaction costs involved in dealing with SIC
and SEBRAE. Moreover, group formation enabled the producers to engage in collec-
tive learning because they had to communicate and collaborate to solve common
problems, and co-ordinate their activities in order to meet large orders. SEBRAE
 Technology Support 71

engineers would typically come in only after the producers had collectively discussed
their problems and identi® ed their assistance priorities.
The FIT project also targets groups of metalworking producers operating in the same
area, although it does not seem to involve tightly-knit geographically con® ned clusters
as in the case of SEBRAE. This concept is in any case not so relevant in many African
rural areas where the volume of production is simply too low for such clusters to form.
Advantages of group-based assistance include a tight focus on common areas of interest
and concern, including specialized technical issues, which contributes to informal
information exchange and has sparked interactive learning among the producers. It has
also been a means to develop collective self-help institutions which can in due course
take over the project activities.

6.4 Appropriate Incentives

Appropriate incentive structures have been of paramount importance for the effective-
ness of the three projects. In LIUP, the main incentive to the participating TNCs is that
they bene® t from their efforts to upgrade the operational ef® ciency of client enterprises,
while they also receive considerable subsidies for making inputs into the project, up to
90% of the costs involved. These are partly borne by the EDB, but also partly by the
participating small producers themselves, which in turn helps to ensure their continued
support and commitment to the project. The programme is apparently effective because
ª ¼ the bene® ts are mutual, and market forces rather than administrative exhortation or
compulsion motivate the transfer of technology ¼ º (Wong, 1994, p. 82).
In the SEBRAE project too, incentives mimicked market forces and created pressures
on the assistance deliverers and bene® ciaries to perform. First, while a new potential
market opportunity was created, the customer was not obliged to proceed with the
procurement from small producers if quality remained below that of the regular
(large-scale) suppliers. Hence, rather than creating a protected market, the scheme
fostered competition between large and small producers. Second, by linking the
commission received by SEBRAE to the successful securing of orders by producers, the
agency’ s ® nancial position became partly dependent upon the effectiveness of its
assistance, thus creating a clear incentive to perform well.
In FIT, too, the design of a market-based incentive structure has played a crucial role
in achieving sustainability of project bene® ts. It is made known in advance that project
® nancing of the activities is always temporary, and even at the start of a new initiative
the participating producers, as well as their customers, must show willingness to pay at
least part of the costs. Moreover, producers must show willingness to take over the
organization of the activity after the external input has come to an end. When it is clear
that an activity does not meet these requirements, the project discontinues it quite
quickly. One could argue that improvements may take some time to materialize; and that
some degree of market failure is inevitable, justifying a long-term public subsidy.
However, in resource-poor countries like Kenya, the likelihood of such activities
attracting public money on a sustained basis is remote. In such conditions, projects like
FIT, which do not expect such subsidies and adhere to stringent ® nancial viabilit y
conditions, obviously stand the best chance of success.

7. Concluding Remarks
The practice of technology support for small producers has advanced since the early
1970s. First, and most important, the notion of ª successº in projects and programmes
72 H. Romijn

is evolving in a more realistic direction. It is no longer based on the idea that small
producers should essentially play the role of passive recipients and bene® ciaries of
improved technologies that have been developed elsewhere. We are moving towards the
understanding that durable competitiveness of small manufacturers must entail an
internal capability on their part to make improvements in products, processes and
organization on an ongoing basis. Only then can we expect such enterprises to make a
contribution to industrialization that goes beyond mere employment generation of last
resort. Truly progressive developing economies are those where ® rms of all sizes,
including the smallest ones, are actively involved in, and contribute to, the national
knowledge accumulation process (Bruton, 1985, p. 81).
Second, there is a growing understanding among LDC government agencies and
donors about the broad features associated with project success so de® ned. Although
regional and sectoral speci® cities exist, these principles appear to be common to
projects and programmes across regions that are technologically and economically quite
diverse. In particular, the development of producers’ capability will generally entail a
process of incremental and demand-focused technological learning and organizational
and institutional capacity-building by the people involved. The literature reviewed for
this paper also suggests that such learning is likely to be most effective when producers
can interact with each other and with other groups, especially customers.
Effective projects are those that stimulate these processes by establishing conditions
in which learning can occur. Linking producers to new markets is an important
precondition. Facilitating access to information, especially through interactions with
other parties, is another aspect. Organization in local groups with common interests
and building collective self-help institutions is yet another element. Finally, incentive
structures should spur producers’ and assistance agencies’ efforts (especially ones that
make rewards conditional upon performance).
Of course, this does not constitute a ready blueprint for technical assistance. Even
if there were many more detailed analytical case studies about successful projects than
can be drawn upon at presentÐ and more are certainly neededÐ an ideal model is
unlikely ever to emerge. The design and implementation of every new project must
always entail an act of creativity to adapt and operationalize general principles in such
a way as to ® t well within the local context within which it is to be implemented.

Notes
1. The main focus is on programmes and projects with a direct technological content, such as
development and commercialization of appropriate technologies, provision of technical
extension services and technical training. It does not deal with ® nancial support schemes,
even though these are very common in developing countries. Financing can be given for a
number of purposes, among which technological improvement is only oneÐ and often not
the most important one. Venture capital and R&D ® nancing for high-tech small ® rms are
exceptions, but these are not covered here because the scope of such schemes is still quite
limited in most developing countries.
2. This contrasts with the situation in developed economies where many small companies make
signi® cant contributions to the generation of innovations (e.g. Cosh et al., 1996; Rothwell &
Zegveld, 1985), and where small ® rms staffed by educated professionals are well represented
in new, knowledge-based industries such as information technology and biotechnology.
3. For an elaboration of the distinction between poverty alleviation and business growth as
objectives of small enterprise support projects, see Dawson & Jeans (1997).
4. Some notable exceptions did exist, especially Watanabe’ s (1983) study about subcontracting
linkages.
5. This argument is based on the ª rigid factor proportionsº problem, ® rst propounded by
Eckaus (1955).
 Technology Support 73
6. See Bongenaar & Sizrmai (1998) for a detailed case study of the top-down approach in
Tanzania.
7. Good examples are documented in Basant (1990), Ishikawa (1975) and Francks (1979).
8. See Basant (1990), Smillie (1991) and Powell (1995).
9. Their activities and performance have been evaluated in many studies, including several
commissioned by aid donors supporting the organizations. See, e.g. UNDP et al. (1988).
10. Few understood that the Japanese success in establishing elaborate and dynamic subcontrac-
tor networks was driven by extreme capital scarcity in the economy, which affected even the
largest keiretsu (Watanabe, 1983).
11. Some writers claim that substantial de-industrialization occurred in the region, although the
evidence on this is not conclusive. See World Bank (1994), pp. 149± 152, for a discussion
about the de-industrialization debate.
12. See Lall (1992) and UNCTAD (1996) for good reviews of the capability literature.
13. Their main sources of inspiration are economic organization literature (transaction costs
theory) and sociology. They are sensitive to the socio-political and institutional context
within which small industry clusters function. These aspects have received more attention
than ª hardº economic and technological ones.
14. This view is not uncontested. Some agencies, such as ApproTEC Kenya, operate projects
that adhere to the so-called ª corporate approachº (Havers, 1998). This approach continues
to give central importance to the introduction of technological hardware. A foreign NGO
designs useful technologies for the bene® t of primarily poor consumers, and disseminates
these as widely as possible. A few reasonably well-run small and medium-sized manufactur-
ers are contracted to make the products according to the speci® cations supplied by the NGO,
but any bene® t they receive is primarily a means of reaching a large number of poor
end-users. There is no explicit attempt to transfer any design skills. Although the corporate
approach is achieving good results (in terms of its own objectives), it does not seem to make
sense to view it as a full-¯ edged alternative to the indigenous technology development
approach (as done by the Donor Committee on Small Enterprise Development, 1997) since
it does not see small producers as the target bene® ciaries and since it does not explicitly aim
to raise their technological standards.
15. This is in line with the ª sub-sector approachº to small enterprise promotion, which advo-
cates that research and assistance should concentrate on commodity-speci® c sub-sectors. By
giving considerable weight to the study of interactions between ® rms of different sizes and at
different stages in the supply chain, this approach can provide a more thorough insight into
the competitive context in which the target enterprises operate (Boomgard et al., 1992).
16. The discussion about the Singaporean project is based on Wong (1994).
17. The discussion about the Brazilian project is based on Tendler & Amorim (1996).
18. The discussion about the Kenyan project is based on Tanburn (1996).
19. For example, in the case of Malaysia’ s car manufacturer Proton, small manufacturers gained
access to manufacturing know-how about complex car components in a similar way. A useful
review of successful and less successful policy experiences with subcontracting promotion in
different developing countries is contained in Altenburg (1997).

References
Albu, M. (1997) Technological learning and innovation in industrial clusters in the South, MSc
Thesis, Science Policy Review Unit, University of Sussex, Brighton.
Altenburg, T. (1997) Promoting ancillary industries in developing countries, Small Enterprise
Development, 8, pp. 24± 33.
Appropriate Technology (1997) 24(1), June, various articles.
Basant, R. (1990) Farmers, fabricators and formal R&DÐ the pipe frame multipurpose tool bar
in Gujarat, India, in: M.S. Gamser, H. Appleton & N. Carter (Eds) Tinker, Tiller, Technical
Change (London, Intermediate Technology Publications).
Best, M.H. (1990) The New Competition. Institutions of Industrial Restructuring (Cambridge, Polity
Press).
Bhalla, A.S. (Ed.) (1985) Technology and Employment in Industry (Geneva, International Labour
Of® ce).
Bhalla, A.S., James, D. & Stevens, Y. (1984) Blending of New and Traditional Technologies (Dublin,
Tycooly).
74 H. Romijn
Bongenaar, B. & Szirmai, A. (1998) The role of a research and development institute in the
development and diffusion of technology: the case of TIRDO, Paper presented in the EADI
Workshop on Innovation in Small Enterprises in the Third World, 18± 19 September, Institute
of Social Studies, The Hague.
Boomgard, J.J., Davies, S.P., Haggblade, S.J. & Mead, D.C. (1992) A subsector approach to
small enterprise promotion and research, World Development, 20, pp. 199± 212.
Bruton, H.J. (1985) On the production of a national technology, in: J. James & S. Watanabe
(Eds) Technology, Institutions and Government Policies (London, Macmillan), pp. 81± 115.
Ceglie, G. & Dini, M. (1999) SME cluster and network development in developing countries: the
experience of UNIDO, Paper for the International Conference on Building a Modern and
Effective Development Service Industry for Small Enterprises, 2± 5 March, Committee of
Donor Agencies for Small Enterprise Development, Rio de Janeiro.
Chon, S. (1996) Small and medium-sized enterprises in the Republic of Korea: implications for
the development of technology-intensive industries, Small Business Economics, 8, pp. 107± 120.
Chung, S. & Park, D.-H. (1998) Technological capabilities of Korean SMEs: a preliminary
analysis, Mimeo, Seoul, Science and Technology Policy Institute.
Cosh, A., Hughes, A. & Wood, E. (1996) Innovation in UK SMEs: causes and the consequences
for ® rm failure and acquisition, ESRC Centre for Business Research Working Paper 48,
University of Cambridge, Cambridge.
Dasgupta, P. (1987) The economic theory of technology policy: an introduction, in: P. Dasgupta
& P. Stoneman (Eds) Economic Policy and Technological Performance (Cambridge, Cambridge
University Press), pp. 7± 23.
Dawson, J. (1999) Donor sourcing from pro-poor business, Mimeo, Report for ActionAid.
Dawson, J. & Jeans, A. (1997) Looking Beyond Credit. Business Development Services and
the Promotion of Innovation among Small Producers (London, Intermediate Technology
Publications).
Donor Committee on Small Enterprise Development (1997) Business development services for
SMEs: preliminary guidelines for donor-funded interventions ([Link]
english/65entrep/bds/[Link]).
Eckaus, R.S. (1955) The factor proportions problem in underdeveloped areas, American Economic
Review, 45, September, pp. 539± 565.
Francks, P. (1979) The development of new techniques in agriculture: the case of the mechaniza-
tion of irrigation in the Saga Plain area of Japan, World Development, 7, pp. 531± 539.
Goldar, B. (1988) Relative ef® ciency of modern small scale industries in India, in: K.B. Suri (Ed.)
Small Scale Enterprises in Industrial Development. The Indian Experience (New Delhi, Sage
Publications), Chapter 5, pp. 95± 117.
Haggblade, S., Liedholm, C. & Mead, D.C. (1990) The effect of policy and policy reforms on
non-agricultural enterprises and employment in developing countries: a review of past experi-
ences, in: F. Stewart, H. Thomas & T. de Wilde (Eds) The Other Policy (London, Intermediate
Technology Publications).
Harper, M. (1984) Small Business in the Third World (London, Intermediate Technology
Publications).
Havers, M. (1998) ApproTEC, Kenya: developing technology-based business opportunities. A
case study on business development services for SMEs, Mimeo, The Spring® eld Centre for
Business in Development, Durham.
Humphrey, J. & Schmitz, H. (1996) The Triple C approach to local industrial policy, World
Development, 24, pp. 1259± 1277.
Ishikawa, S. (1975) The Chinese method of technological developmentÐ the case of the
agricultural machinery and implements industry, Developing Economies, 13, pp. 430± 458.
James, J. (1989) Improving Traditional Rural Technologies (London, Macmillan).
Jeans, A. (1999) Technology, NGOs and small enterprise: securing livelihoods through technical
change, in: K. King & S. McGrath (Eds) Enterprise in Africa. Between Poverty and Growth
(London, IT Publications), Chapter 13, pp. 169± 178.
King, K. (1996) Jua Kali Kenya (London, James Currey).
King, K. & McGrath, S. (Eds) (1999) Enterprise in Africa. Between Poverty and Growth (London,
Intermediate Technology Publications).
Lall, S. (1992) Technological capabilities and industrialisation, World Development, 20, pp.
165± 186.
Lall, S. (Ed.) (1999) The Technological Response to Import Liberalization in Subsaharan Africa
(London, Macmillan, on behalf of UNU-INTECH, Maastricht).
 Technology Support 75
Lall, S. & Teubal, M. (1998) ª Market-stimulatingº technology policies in developing countries:
a framework with examples from East Asia, World Development, 26, pp. 1369± 1385.
Lee, K.-U. (1992) Industrial organization: issues and recent developments, in: V. Corbo & S.-M.
Suh (Eds) Structural Adjustment in a Newly Industrialized Economy: The Korean Experience
(Washington, DC, Johns Hopkins University Press).
Leipziger, D. & Petri, P. (1993) Korean Industrial Policy: Legacies of the Past and Directions for the
Future (Washington, DC, The World Bank).
Levy, B., Berry, A. & Nugent, J.B. (1999) Ful® lling the Export Potential of Small and Medium Firms
(Boston, Kluwer Academic).
Little, I.M.D., Mazumdar, D. & Page, J.M. (1987) Small Manufacturing Enterprises. A Compara-
tive Analysis of India and Other Economies (New York, Oxford University Press).
Maldonado, C. & Sethuraman, S.V. (1992) Technological Capability in the Informal Sector. Metal
Manufacturing in Developing Countries (Geneva, International Labour Of® ce).
Massaquoi, J.G.M. (1995) The effect of some sectoral development policies on technologyÐ the
case of the informal sector, in: R. Heeks et al. (Eds) Technology and Developing Countries:
Practical Applications, Theoretical Issues (London, Frank Cass).
Meyanathan, S.D. (Ed.) (1994) Industrial Structures and the Development of Small and Medium
Enterprise Linkages. Examples from East Asia, EDI Seminar Series (Washington, DC, The World
Bank).
Nadvi, K.M. (1996) Small ® rm industrial districts in Pakistan, DPhil Thesis, Institute of
Development Studies, University of Sussex, Brighton.
Oladeji, S.I. (1998) Technology policy and the development of small and medium scale enter-
prises in contemporary Nigeria, Technovation, 18, pp. 125± 132.
Oyelaran-Oyeyinka, B. (1997) Technological learning in African industry: a study of engineering
® rms in Nigeria, Science and Public Policy, 24, pp. 309± 318.
Piore, M. & Sabel, C. (1984) The Second Industrial Divide: Possibilities for Prosperity (New York,
Macmillan).
Powell, J. (1995) The Survival of the Fitter. Lives of Some African Engineers (London, Intermediate
Technology Publications).
Pyke, F. & Sengenberger, W. (1992) Industrial Districts and Local Regeneration (Geneva, Inter-
national Institute for Labour Studies).
Rabellotti, R. (1995) External economies and cooperation in industrial districts: a comparison of
Italy and Mexico, DPhil Thesis, Institute of Development Studies, University of Sussex,
Brighton.
Romijn, H. (1999) Acquisition of Technological Capabilities in Small Firms in Developing Countries
(London, Macmillan).
Romijn, H. (1997) Acquisition of technological capabilities in development: a quantitative case
study of Pakistan’ s capital goods sector, World Development, 25, pp. 359± 377.
Rothwell, R. & Zegveld, W. (1985) Innovation and the Small and Medium Sized Firm (Boston, MA,
Kluwer).
Schmitz, H. (1995) Small shoemakers and fordist giants: tale of a supercluster, World Develop-
ment, 23, pp. 9± 28.
Sethuraman, S.V. (1977) Technology and small enterprise development, in: P.A. Neck & R.E.
Nelson (Eds) Small Enterprise Development: Policies and Programmes, Management Development
Series No. 14 (Geneva, International Labour Of® ce), Chapter 12, pp. 187± 202.
Sethuraman, S.V. (Ed.) (1981) The Urban Informal Sector in Developing Countries: Employment,
Poverty and Environment (Geneva, International Labour Of® ce).
Smillie, I. (1991) Light engineering and the very late starters, in: I. Smillie (Ed.) Mastering the
Machine, Poverty, Aid and Technology (London, Intermediate Technology Publications).
Stewart, F. (Ed.) (1987) Macro-policies for Appropriate Technology in Developing Countries (Boulder,
Westview Press).
Stewart, F. & Ranis, G. (1990) Macro-policies for appropriate technology: a synthesis of ® ndings,
in: F. Stewart, H. Thomas & T. de Wilde (Eds) The Other Policy (London, Intermediate
Technology Publications).
Stoneman, P. (Ed.) (1995) Handbook of the Economics of Innovation and Technological Change
(Oxford, Blackwell).
Tanburn, J. (1996) Towards success: impact and sustainability in the FIT Programme, Small
Enterprise Development, 7, pp. 42± 51.
Tendler, J. & Amorim, M. (1996) Small ® rms and their helpers: lessons in demand, World
Development, 24, pp. 407± 426.
76 H. Romijn
Tewari, M. (1996) When the marginal becomes mainstream: lessons from half-century
of dynamic small-® rm growth in Ludhiana, India, PhD Thesis, Massachusetts Institute of
Technology, Boston, Massachusetts.
UNCTAD (1996) Fostering Technological Dynamism: Evolution of Thought on Technological
Development Processes and Competitiveness. A Review of the Literature (Geneva and New York,
United Nations).
UNDP, UNIDO and Government of the Netherlands (1988) Development of Rural Small Indus-
trial Enterprises (Vienna, UNIDO).
UNIDO (1996) Industrial Development Global Report 1996 (Oxford, Oxford University Press).
Wangwe, S. (1993) Small and microenterprise promotion and technological policy implications,
in: A.H.J. Helmsing & Th. Kolstee (Eds) Small Enterprises and Changing Policies; Structural
Adjustment, Financial Policy and Assistance Programmes in Africa (London, Intermediate
Technology Publications).
Watanabe, S. (Ed.) (1983) Technology, Marketing and Industrialisation. Linkages Between Large and
Small Enterprises (Delhi, Macmillan).
Wong, S.T. (1994) Singapore, in: S.D. Meyanathan (Ed.) Industrial Structures and the Development
of Small and Medium Enterprise Linkages. Examples from East Asia, EDI Seminar Series (Wash-
ington, DC, The World Bank).
World Bank (1994) Adjustment in Africa. Reforms, Results and the Road Ahead (Oxford, Oxford
University Press).
World Development (1999) September issue, various articles.