1
SMALL BUSINESSES, JOB CREATION AND GROWTH:
FACTS, OBSTACLES AND BEST PRACTICES
2
Table of contents
Executive Summary…………………………………………………………………………………………………………………. 3
Section 1. Job Creation, Output and Productivity Growth………………………………………………………………. 7
Section 2. Financing Small and Medium Enterprises…………………………………………………………………… 17
Section 3. SMEs and Regulatory Reform………………………………………………………………………………….. 21
Section 4. Public Support to SMEs………………………………………………………………………………………….. 25
Section 5. Innovative SMEs……………………………………………………………………………………………………. 28
Section 6. High Growth SMEs………………………………………………………………………………………………… 34
Section 7. Women-owned SMEs……………………………………………………………………………………………… 38
Section 8. The Regional Dimension to Entrepreneurship……………………………………………………………… 41
Section 9. Best Practice Policies for SMEs………………………………………………………………………………… 44
3
EXECUTIVE SUMMARY
The importance of SMEs
SMEs (small and medium-sized enterprises) account for 60 to 70 per cent of jobs in most OECD
countries, with a particularly large share in Italy and Japan, and a relatively smaller share in the United
States. Throughout they also account for a disproportionately large share of new jobs, especially in those
countries which have displayed a strong employment record, including the United States and the
Netherlands. Some evidence points also to the importance of age, rather than size, in job creation: young
firms generate more than their share of employment. However, less than one-half of start-ups survive for
more than five years and only a fraction develop into the high-growth firms which make important
contributions to job creation. High job turnover poses problems for employment security; and small
establishments are often exempt from giving notice to their employees. Small firms also tend to invest
less in training and rely relatively more on external recruitment for raising competence.
The demand for reliable, relevant and internationally comparable data on SMEs is on the rise, and
statistical offices have started to expand their collection and publication of data. International
comparability is still weak, however, due to divergent size-class definitions and sector classifications. To
enable useful policy analysis, OECD governments need to improve their build-up of data, without creating
additional obstacles for firms through the burden of excessive paper work.
Problems confronted by SMEs
The greater variance in profitability, survival and growth of SMEs compared to larger firms
accounts for special problems in financing. SMEs generally tend to be confronted with higher interest
rates, as well as credit rationing due to shortage of collateral. The issues that arise in financing differ
considerably between existing and new firms, as well as between those which grow slowly and those that
grow rapidly. The expansion of private equity markets, including informal markets, has greatly improved
the access to venture capital for start-ups and SMEs, but considerable differences remain among countries.
Regulatory burdens remain a major obstacle for SMEs as these firms tend to be poorly equipped
to deal with the problems arising from regulations. Access to information about regulations should be
made available to SMEs at minimum cost. Policy makers must ensure that the compliance procedures
associated with, e.g. R&D and new technologies, are not unnecessarily costly, complex or lengthy.
Transparency is of particular importance to SMEs, and information technology has great potential to
narrow the information gap. It would be of great help to set up a “one-stop-shop system”, where all the
necessary information which affects firm strategies and decisions is made available in one place, as exists
already in some countries.
4
Support for SMEs
Most OECD countries have programmes which support SMEs. One-quarter of all public support
programmes reported to the OECD primarily target SMEs. Germany, Iceland, Japan and New Zealand
dedicated more than 50 per cent of their entire public support programmes to SMEs. In 1993, a total of
US$3.75 billion of public money was paid to help start-ups, the acquisition of equipment, R&D, training
and consultancy services, in the form of direct grants, tax concessions, low interest rate loans or loan
guarantees. More than 50 per cent of SME programmes are administered locally, making co-ordination
between authorities critical. There is potential for integrating programmes into fewer schemes but with a
wider scope, making it easier for SMEs to understand them, and lowering administrative costs. Almost 70
per cent of SME programmes last for more than five years. Stable and predictable programme
management is in the interest of users; however, a constant review process is vital to ensure quality and
flexibility. Governments need to intensify their efforts to disseminate information, eliminate unnecessary
red tape, and make programmes more responsive to the changing needs of SMEs.
Between 30 and 60 per cent of SMEs can be characterised as innovative, of which some 10 per
cent are technology-based. Innovative SMEs tend to be market-driven rather than research-driven, and
quicker in responding to new opportunities than large firms. They play a key role in pioneering and
developing new markets. Programmes for improving the diffusion of technology have shifted from a
supply focus to raising the capacity of SMEs to absorb technology. However, governments need to: reduce
uncertainties in the tax, regulatory and macroeconomic environment; make sure that business framework
conditions do not impact unfavourably on the risk/reward ratio; and encourage the mobility of human
resources and the markets for specialised services. Although these are important for the entire economy,
such actions will produce benefits of particular value to SMEs.
The successes
A limited number of so-called high-growth SMEs make important contributions to job creation and
productivity growth in the OECD area. At the earlier stages, management capabilities are crucial to
survival. As the firm matures, human resource and innovation strategies increase in importance. By the
time the firm has become established, innovation is crucial for growth. The fastest growing entrants are
those that translate strategy into action in the form of R&D, innovation and training, put great emphasis on
hiring skilled employees and motivating employees, and balance the enhancement of their capabilities in
different areas — the last being particularly important in high-knowledge sectors. The main barriers to the
development of high-growth SMEs are market failures in capital markets, government regulations,
indirect labour costs, access to foreign markets, and difficulties in recruiting qualified staff and skilled
workers.
Women-owned SMEs are growing at a faster rate than the economy as a whole in several OECD countries,
allowing capitalisation of the skills of educated and trained women who might otherwise be blocked in
corporate advancement because of the “glass ceiling”. The increased flexibility inherent in owning one’s
business allows women to contribute to the income of their families while balancing work and family
responsibilities. However, the economic potential of women entrepreneurs remains partly untapped;
measures are required to improve information and statistics in this field, as well as to strengthen the
preconditions for financing, networks and technology.
Entrepreneurship tends to vary markedly across regions. An increasing number of regions are known for
generating clusters of dynamic firms which benefit from “information spill-over” and other intangible
factors. Regional development policies have been introduced to assist regions suffering from declining
5
industries. The primary policy tools for attracting firms to disadvantaged regions are investment in
infrastructure, social assistance, training and other forms of public assistance. The regional dimension of
entrepreneurship is not limited to clusters of enterprises but also includes micro enterprises. Programmes
to assist the creation and development of micro enterprises in inner cities and remote rural areas have
become widespread policy tools. Governments wishing to adopt policies used successfully in other
regions or countries should take the regional context into account.
Best practice policies
The report ends with lessons from policies undertaken in five areas:
— Financing
The primary role of the public sector in supporting venture capital is to reduce the risk and cost
of private equity finance, complementing and encouraging the development of the private capital
industry. There is major variation across OECD countries in the use of funding methods for
SMEs, but the provision of equity financing to start-up companies is more advanced in the
United States and Canada than elsewhere. Taxation should not impose a disproportionately
heavy burden on SMEs.
— The business environment
This can be improved by systematic and careful scrutiny of new regulations and by
implementation of a business impact system to ensure the audit and monitoring of new
legislation. Canada, the United Kingdom and the Netherlands have successfully introduced
procedures to that end. The use of information technologies provides opportunities for reducing
bureaucratic burdens on all companies, including SMEs.
— Technology
Technology diffusion programmes should: ensure quality control; promote customer-
orientation; upgrade the innovative capacity of firms — including the promotion of general
awareness of the value of innovation among management — and stimulate demand for technical
and organisational change; build on existing inter-relationships in national innovation systems
and provide greater coherence between programme design (e.g. targets, objectives, modes of
support) and service delivery; build on evaluation and assessment. Technology diffusion
programmes should in particular have mechanisms for assessment which can guide and improve
their operation and management on a continuing basis. The United States has programmes
effectively stimulating quality in diffusion processes, while Germany has sophisticated
institutional set-up catalysing interactions between existing actors in the national innovation
system.
— Management capabilities
Several G7 governments have sought to enhance the “quality” of owner/managers of SMEs
either by encouraging training and/or by providing access to advisory and consultancy services.
The most extensive assistance is provided by Japan which has both a highly developed system of
advisory services and SME colleges. The United Kingdom and Italy have also implemented
interesting schemes. Subsidy-schemes aimed at enhancing the skill base of SMEs should take
6
the following into consideration: specification of objectives; situation after the removal of the
subsidy; collecting information from SMEs themselves. Measures to encourage information
networks must seek to customise databases and avoid information overload. Four approaches
have been developed to address these issues: know your customer; access; explicitly avoid
interference with market mechanisms; and subsidisation of information.
— Access to markets
Measures to ease access to markets have focused on international markets, on the one hand, and
public procurement, on the other. Japan has the most developed policy and institutional set-up
for the former, based upon the use of non-discriminatory measures which seek to support efforts
made by SMEs themselves. Policy in this area seeks to tackle the disadvantages experienced by
SMEs due to their lack of access to human resources, to external markets and to technology.
Regarding public procurement, the United States, and other OECD countries such as Australia,
have made comprehensive efforts to increase the “share” which small firms obtain of
government contracts.
7
SECTION 1
JOB CREATION, OUTPUT AND PRODUCTIVITY GROWTH
The term “SME” — small and medium-sized enterprises — covers a variety of definitions and
measures. In OECD Member countries, employment is the most widely used criterion for determining
firm size. SMEs are usually defined as firms with fewer than 500 employees, although a number of
countries — including those in the European Union — use a lower cut-off point of 250.
Employment and job creation
It is apparent that SMEs play an important role in all OECD economies: they make up over
95 per cent of enterprises and account for 60 to 70 per cent of jobs in most OECD countries. The share
tends to be somewhat lower in manufacturing, although it varies between 40 to 80 per cent of employment
in manufacturing (Table 1.1). The overall share of small firms in employment and output may be even
higher given that establishments or firms in the service sector are normally of smaller average size than in
manufacturing. Table 1.2 illustrates the variability across sectors: for example, wholesale and retail trade
and hotels and restaurants are dominated by SMEs. In construction SMEs account for 80 to 90 per cent of
all employment. The fact that these industries loom large in overall employment underscores the
importance of SMEs as sources of employment.
Furthermore, the share of large firms in employment and output has tended to show a certain
decline. The average establishment size in manufacturing has fallen since the early 1980s in Canada, the
United Kingdom and the United States, has remained constant in Germany, and has risen in Japan (Table
1.3a and 1.3b). The importance of smaller establishments in job dynamics is often assessed on the basis
of net employment changes. From the mid-1980s to the early 1990s, in all countries, small establishments
(fewer than 100 employees) displayed more rapid net employment growth than larger ones (OECD
Employment Outlook 1994, Chapter 3). However, there are at least four ways in which this should be put
into perspective:
First, it is not surprising that small enterprises/establishments play an important role in the job
creation process since they account for between 40 and 80 per cent of total manufacturing employment.
To see whether their role is disproportionately high, net job creation has to be expressed in relation to the
initial employment in small and large establishments. As shown in Table 1.4, net job creation rates are in
fact often higher for smaller size classes. However, for a number of countries it was found that the highest
net job creation rates were among very small firms whereas small to medium-sized firms (between 20 and
50 employees) did not perform better than large firms.
Second, methodology matters. An important technical issue in studies on net job creation rates
is how firms are allocated to size classes: for example, a firm can be considered “small” if it corresponds
to the criterion “small” in some base year. Any subsequent job creation is then attributed to the size class
“small”, irrespective of whether the firm has moved to a different size class by the end of the observation
period. Alternatively, a firm can be considered “small” if it corresponds to the criterion “small” on
average, over the entire period. It has been shown that net job creation rates of small and large firms are
highly sensitive to such changes in the size class allocation of firms.
8
Third, the customary estimation of net employment growth conceals the separate processes of
job creation and destruction. Plants of all sizes incur both job gains and job losses. Some idea of these
dynamics is given in Table 1.5. The data paint a picture of the concentration of gross job gains and losses
in very small and small establishments. Establishments employing fewer than 20 workers seem to
account for between 45 and 65 per cent of new job gains and 36 and 56 per cent of annual job losses (the
data for the United States are much lower, partly because firms employing fewer than five workers are
excluded and partly because they refer only to manufacturing and exclude services).
Fourth, recent research in the United States has pointed to age-related patterns in job creation:
one is that net job creation rates decline with plant age; the other is that employment volatility declines
with plant age. These patterns should not come unexpected, given that young firms are nearly always
small. However, small firms are not necessarily young. The distinction is important because, if age rather
than size was the criterion, policy should focus less on small firms and more on young firms. Put
differently, a policy in favour of SMEs would be replaced by a policy to promote entrepreneurship, for
example, through the removal of regulatory barriers to firm creation. Yet, more empirical evidence on the
significance of age as opposed to size is needed before clear policy conclusions can be put forward.
Information on job creation and destruction reveals a considerable amount of churning in all
labour markets. Annual job turnover rates — the sum of newly created jobs and jobs that have disappeared
— are of the order of 20 per cent per year in countries as diverse as France, Sweden and the United States.
Job turnover is a critical part of the competitive process, contributing to economic growth, productivity
and structural change. Excessive turnover, on the other hand, can deter businesses and workers from
investing optimally in training. A relative lack of skills can affect the ability of firms to adapt to change
via internal flexibility as opposed to external adjustment. Permanent job losses can lead to substantial
cuts in income for those affected as their accumulated firm-specific skills lose their value. Finally, when
turnover is associated with large-scale lay-offs or plant closures, substantial costs may be borne by regions
and communities.
The costs of churning have to be weighed against the positive effects of turbulence, such as
entrepreneurship and the search for new processes and products. As reported in OECD (1996),
Technology, Productivity and Job Creation, less than one-half of SME start-ups survive for five years;
only a small percentage of surviving SMEs turn into high-growth firms; and these high-growth firms
make important contributions to job creation and productivity growth. At the firm level, turnover could
be the result of a process of trial and error, with some enterprises failing almost from the start, some being
limited to the life span of a single innovation, and others enjoying sustained success.
Export, production and productivity
Overall, SMEs account for between 30 and 70 per cent of value added (Table 1.6) with
variations between countries and industries. Also, as would be expected, the likelihood that output is
exported is smaller for SMEs than for large enterprises: in very general terms and depending on the
country, SMEs contribute between 15 and 50 per cent of exports, while between 20 and 80 per cent of
SMEs are active exporters. Overall, it is estimated that SMEs contribute between 25 and 35 per cent of
world manufactured direct exports (OECD, 1997, Globalisation and Small and Medium Enterprises).
Where information exists, it points to these exports being concentrated around relatively few larger SMEs.
However, most of the growth of exports seems to be taking place in smaller SMEs.
The employment share of SMEs exceeds their share in value added, implying that value added
per employed person (a measure of labour productivity) is lower in smaller firms than in larger ones. Yet
it would be misleading to conclude from this observation that small firms necessarily contribute less than
9
large firms to economy-wide productivity growth. A more dynamic view should be taken. The static
observation of higher productivity levels linked to larger size ignores the process of productivity growth:
overall productivity changes occur because individual firms raise their productivity levels and because
they expand and displace low-productivity firms. Similarly, new entrants replace exits because they have
a higher level of productivity. As SMEs account for most of the entrants, exits, growth and decline, they
form an integral part of a competitive process that contributes significantly to aggregate productivity
growth — even if at any particular time, their level of productivity is below that of larger firms. This view
has been substantiated by several studies, which showed sizeable effects on aggregate productivity growth
from shifting market and employment shares and the entry and exit process. This is yet another positive,
dynamic effect associated with the turbulence of small firms.
Policy issues
Concerns about high employment turnover rates lead to two policy issues: i) the role of
governments and/or collective bargaining arrangements in providing some degree of employment security
(e.g. advance notice or consultation) as a way of dealing with some of the costs identified above and in
providing incentives for internal flexibility/training; and ii) the role of active labour market policies in
assisting the matching of workers and businesses. The first of these can be particularly controversial vis-
à-vis smaller establishments which are often exempt from notice requirements. Smaller firms also tend to
carry out less internal training and rely more on the external market, which can be problematic if a supply
of requisite skills is not available. There are numerous country experiences with programmes/policies to
“assist” smaller firms, sometimes allowing a pooling of resources (e.g. a French training levy can be
pooled, although evidence suggests that larger firms tend to make more use of this option).
There are several schools of thought on policies to help new and small firms mature. Some
argue that the role of government should only be to make it easier for firms to start up, while others focus
on helping those firms most likely to grow. The track record of special policies to encourage new firms
has not been particularly encouraging, although reforms to regulations and practices that hinder start-ups
and frustrate the operations of new small firms are desirable. Overall empirical knowledge of the issues
remains limited and governments may do well to invest more in evaluating what does, and does not, work
in terms of fostering an environment where new firms can grow successfully and create jobs.
Improving SME statistics
Demand for reliable, relevant and internationally comparable data on SMEs has been rising.
Statistical offices have started to collect and publish relevant data but serious shortcomings persist.
International comparability has remained weak, due to divergence of definitions for size-class and
treatment of underlying units (firms, establishments) and variation in industry classification and time
periods used for data collection. Many interesting issues in conjunction with SMEs can only be addressed
with sets of micro-level data that allow the tracing of individual firms or establishments over time. In
addition to problems of comparability, the sheer volume of the data sets involved, as well as questions of
confidentiality, have prevented rapid progress of international studies.
More generally, there is a conflict of interests between obtaining comprehensive and timely
information about SMEs and limiting the administrative burden associated with responding to statistical
questionnaires. As a result statisticians must increasingly endeavour to exploit creatively the existing
statistical or administrative sources to obtain better information on SMEs and to minimise future response
burdens.
10
Table 1.1 Size distributions in manufacturing industry
Number of enterprises/establishments Employment
of which in employment size class of which in employment size class
Year 1-19 20-99 100-499 500+ 1-19 20-99 100-499 500+
Percentages Percentages
United States 1993 73.7 19.8 5.1 1.4 7.4 14.6 16.5 61.5
Canada 1994 50.6 37.8 10.2 1.4 7.6 27.8 39.4 25.2
Mexico 1994 80.3 15.1 2.7 2.0 12.2 21.2 15.6 51
Japan 1994 74.3 21.6 3.6 0.5 22.4 30.9 25.0 21.6
Korea 1994 69.5 26.1 3.0 1.3 20.5 32.0 14.2 33.3
Australia 1994 82.0 14.1 3.4 0.4 22.3 27.5 32.7 17.5
New Zealand 1994 90.6 7.7 1.5 0.3 27.3 24.7 24.0 24.0
Austria 1993 43.2 41.5 10.0 5.2 4.3 26.9 23.4 45.5
Belgium 1993 80.4 15.3 3.7 0.6 .. .. .. ..
Denmark 1993 82.0 14.6 3.1 0.3 .. .. .. ..
Finland 1992 50.8 36.1 11.6 1.5 .. .. .. ..
Germany 1993 71.5 19.4 4.1 5.0 19.9 22.1 10.8 47.2
Greece 1992 59.0 34.3 6.0 0.7 20.4 35.0 27.5 17.2
Hungary 1994 76.8 18.3 3.9 1.1 .. .. .. ..
Iceland 1992 90.8 6.7 2.5 35.1 26.6 38.2
Italy 1992 89.7 9.0 1.2 0.2 38.7 25.0 17.3 19.0
Luxembourg 1992 79.4 15.0 4.7 0.9 13.0 22.1 35.0 29.9
Netherlands 1993 78.0 17.2 4.3 0.6 15.7 24.8 27.8 31.7
Norway 1994 40.2 47.4 7.5 4.9 9.3 34.9 18.2 37.6
Portugal 1994 85.8 11.8 2.2 0.2 23.5 32.3 27.8 16.5
Sweden 1993 44.4 40.8 12.4 2.4 6.9 23.1 35.3 34.7
Switzerland 1991 84.2 12.3 3.1 0.4 20.2 26.9 31.3 21.5
Czech Republic 1995 94.9 2.9 1.6 0.5 18.0 10.3 24.6 47.1
Turkey 1992 36.6 47.1 13.3 3.0 5.5 22.2 32.2 40.1
United Kingdom 1994 82.7 12.9 3.7 0.8 13.2 21.6 28.9 36.3
Note
: Statistical unit: establishment except for the United States, New Zealand, Czech Republic, Hungary, Italy,
Luxembourg, Portugal (enterprises). Size classes differ: Canada, New Zealand: 0-19; Mexico: 1-15; 16-100; 101-
250; 251+; Japan, Korea: 4-19; Finland: 10-19; Hungary: 0-9; 10-99; Iceland: 0-19; 20-60; 60+; Norway: 1-19
; 20-99 ; 100-199 ; 200+; Czech Republic: 0-24; 25-99.
Source
: OECD, Database on SME statistics; Eurostat (1996),
Enterprises in Europe
.
11
Table 1.2 Distribution of employment by firm size and sector
Year Employment size class
1-9 10-19 20-99 100-499 500+ Total
Percentages
United States
Manufacturing 1993 3.5 3.9 14.6 16.5 61.5 100.0
1988 3.1 3.7 14.5 16.1 62.6 100.0
Construction 1993 28.6 17.0 30.7 12.5 11.1 100.0
1988 25.5 16.5 31.7 14.8 11.5 100.0
Wholesale and retail trade, 1993 13.1 9.9 22.2 11.8 43.0 100.0
hotels and restaurants 1988 13.6 10.4 23.2 11.9 40.8 100.0
Finance, insurance and real estate 1993 12.5 5.5 13.8 12.2 56.1 100.0
1988 12.3 5.7 14.3 12.4 55.3 100.0
Total non-farm business sector 1993 10.7 7.7 18.8 13.1 49.6 100.0
1988 10.5 7.8 19.3 13.3 49.0 100.0
Canada
Manufacturing 1992 .. 11.4 20.3 22.0 46.3 100.0
1989 .. 9.9 21.3 22.8 46.0 100.0
Construction 1992 .. 56.0 26.7 12.3 5.0 100.0
1989 .. 51.0 29.4 14.5 5.1 100.0
Wholesale and retail trade, 1992 .. 33.6 25.7 13.2 27.5 100.0
hotels and restaurants 1989 .. 28.9 24.5 12.9 33.7 100.0
Finance, insurance and real estate 1992 .. 17.8 12.1 9.7 60.4 100.0
1989 .. 17.2 14.4 10.8 57.6 100.0
Total non-farm business sector 1992 .. 25.5 20.8 15.1 38.6 100.0
1989 .. 23.5 22.0 16.3 38.2 100.0
Japan
Manufacturing 1993 12.5 10.4 30.8 24.6 21.7 100.0
1986 13.7 11.0 30.6 23.5 21.3 100.0
France
Manufacturing 1992 8.1 5.0 22.4 23.6 40.9 100.0
1990 10.1 5.1 22.0 23.1 39.7 100.0
Construction 1992 29.0 11.5 27.5 14.0 18.0 100.0
1990 33.9 10.8 26.7 12.8 15.8 100.0
Wholesale and retail trade, 1992 32.9 10.3 25.8 12.4 18.5 100.0
hotels and restaurants 1990 38.6 9.8 23.7 11.2 16.8 100.0
Finance, insurance and real estate 1992 15.9 7.2 19.6 17.9 39.4 100.0
1990 19.7 7.4 18.9 17.4 36.6 100.0
Total non-farm business sector 1992 18.2 7.1 21.7 17.1 35.9 100.0
1990 22.0 7.0 21.0 16.2 33.7 100.0
Table 1.2 Distribution of employment by firm size and sector
(cont’d)
Year Employment size class
1-9 10-19 20-99 100-499 500+ Total
Percentages
Germany
Manufacturing 1992 7.8 6.2 16.3 21.6 48.2 100.0
1988 6.1 6.1 16.1 21.5 50.1 100.0
Construction 1992 27.1 21.6 28.0 13.7 9.6 100.0
1988 22.8 20.4 31.9 15.3 9.6 100.0
Wholesale and retail trade, 1992 37.8 12.6 19.2 11.9 18.5 100.0
hotels and restaurants 1988 29.0 14.6 22.2 13.5 20.7 100.0
Finance, insurance and real estate 1992 28.3 10.8 16.0 16.7 28.2 100.0
1988 21.3 10.3 16.9 17.9 33.6 100.0
Total non-farm business sector 1992 21.1 10.2 18.2 17.0 33.5 100.0
1988 15.3 10.2 19.2 18.0 37.2 100.0
Italy
Manufacturing 1991 24.2 14.8 24.4 16.9 19.7 100.0
Construction 1991 52.3 16.0 19.5 7.9 4.3 100.0
Wholesale and retail trade, 1991 74.6 9.7 8.8 3.1 3.8 100.0
