1. Objectives and challenges of entrepreneurship policy

Business creation is an important driver of economic growth and job creation, generating innovations and contributing to economic efficiency through competition (OECD, 2018a; OECD, 2017). It can also create wider social benefits by contributing to local economic development, supporting industrial transitions (OECD, 2019), and offering an alternative pathway into work for those at a disadvantage in the labour market or in search of more flexibility (OECD/European Union, 2016).

However, market, institutional and behavioural failures create barriers to entrepreneurship. The obstacles include barriers to entry in markets with large incumbents, administrative costs associated with registering a business, and information imperfections in finance markets, which create liabilities of newness and smallness (Stinchcombe, 1965). The basic rationale for entrepreneurship policy stems from the need to address these types of barriers in order to secure the economic and social benefits of entrepreneurship.

The aim of entrepreneurship policy should not solely be to increase the start-up rate, but also to improve the quality of the businesses created. Overall business start-up numbers are dominated by one-person/non-employer businesses. On average solo entrepreneurs offer a modest contribution to growth and employment. Some, such as skilled independent workers and networked entrepreneurs, can be very successful and innovative. The increasing prominence of platform-based work arrangements enables individuals to pursue independent work that may in future become the basis of larger entrepreneurial endeavours. Policies seeking a greater economic and social benefit should focus resources on start-ups with potential for sustainability and growth.

On the other hand, policy should not focus solely on the most dynamic start-ups – the “gazelles” (i.e. enterprises up to 5 years old with average annualised growth greater than at least 10% per annum over a three year period) and “unicorns” (i.e. privately-held start-ups with a valuation of over USD 1 billion). These types of start-ups have disproportionate impacts on job creation and innovation diffusion but are few in number. Furthermore, it is difficult to predict which firms will grow in advance, since start-ups with growth potential come from many different sectors and operate many different kinds of business models (Mason and Brown, 2013; Brown et al, 2017).

The main focus of this report is therefore on the intermediate target of promoting productive entrepreneurship (Box 1.1). Policy for productive entrepreneurship emphasises the creation of businesses with job creation and innovation potential. It focuses on businesses with the potential to employ more than the founder but also on more ordinary businesses that are unlikely to achieve the dramatic rates of growth of gazelles and unicorns.

There are significant and longstanding geographical variations in entrepreneurial activity within countries (OECD, 2018b; Fritsch and Wyrwich, 2014). These regional variations are greater in the case of high growth firms than for start-ups in general (Stam, 2015; Acs and Mueller, 2008). As a result, national programmes designed without account of regional differences are likely to have geographically variable impacts.

Entrepreneurship policies will therefore often benefit from taking into account regional variations in their design. This includes both systemic interventions (e.g. economic policies, fiscal policies), which may affect different segments of the business community differently, and business support programmes (e.g. start-up grants, training programmes, export support), which often show the highest take-up rates in the most entrepreneurial regions. As an example, the United Kingdom’s Loan Guarantee Scheme, which was set up in 1981 to help small firms borrow from banks, had the same requirements for all firms regardless of their location. However, there were large regional differences in the number of loans issued (standardised by the size of the region) and their value, with an over-representation of major financial centres (Cowling, 1998; Harrison and Mason, 1986). Attention to the conditions and targets of these types of programmes may help to support entrepreneurship more evenly within countries.

Incorporating a regional dimension in entrepreneurship policies can also be important to accommodate for structural differences across regions and address regional variations in the nature and intensity of barriers to entrepreneurship. Research finds that geographical variations in entrepreneurial activity can be related to various place-based structural factors. For example, in 2015, around half of firm creations in the OECD area occurred in predominantly urban regions, 36 % in intermediate regions and 13% in rural regions (OECD, 2018b). Capital cities, for example, tend to be entrepreneurship hubs: in 2015 in the OECD area, 29.5% of new firms were created in capital cities, while these cities hosted only 27.5% of existing firms and 20% of the population (OECD, 2018b). Large urban centres benefit from agglomeration effects that are conducive to business start-up. This includes large local markets, easier access to public research and education facilities, high quality of human capital and infrastructure and networking opportunities. Compensating support may be needed to promote entrepreneurship in rural and less dense regions.

The industrial and occupational structure of a city or region also affect entrepreneurial activity: employees in small firms are more likely to start their own business than those working in large businesses. Prior management experience also increases the likelihood that an individual will start their own business, as does family history of business ownership. Wider socio-economic conditions also affect entrepreneurial activities indirectly (e.g. home ownership facilitates access to bank loans). Stronger support may therefore be required in regions with low existing entrepreneurship and small business rates to overcome these disadvantages.

Rates of productive entrepreneurship in a region are affected by a wide range of regional conditions – culture, access to finance, skills, networks and so on. Therefore one of the jobs of entrepreneurship policy is to identify the different strengths and weaknesses of regions in these conditions and developed adapted and tailored policies to overcome the key constraints manifested in each region. Key policy success factors.

Entrepreneurship policy covers a wide range of intervention, with differentiated objectives. Several key success factors contribute to an effective policy mix in support of productive entrepreneurship:

  • Policy interventions should seek support a wide range of entrepreneurs, but focus resources on ventures with potential for growth rather than focusing exclusively on specific sectors or places. Productive entrepreneurship is not restricted to high-tech sectors and entrepreneurial hubs.

  • Institutional conditions can contribute as much to supporting entrepreneurship as directly targeted programmes. These include culture, taxation, competitive conditions and the regulatory framework.

  • Barriers to entrepreneurship are multifaceted and require comprehensive packages. Interventions combining several types of support are typically more effective. Linkages should be fostered between different programmes and across support organisations in the entrepreneurial ecosystems at national, regional and local level, as the support needs of firms change as they progress from idea, start-up, early-growth and scale-up

  • Policy has to be adapted to context to be efficient. Transplanting policies from one country or region to another is unlikely to be successful without adaptation. This includes taking into account structural conditions and sensitivity to new economic developments transforming entrepreneurship and policy delivery, such as digitalisation.

  • Entrepreneurship policies should be designed to minimise deadweight, displacement and distortion. Poorly designed support may create a displacement effect whereby publicly-supported new firms drive out existing businesses. Deadweight is a further threat – whereby support may be provided to enterprises that do not need it or do not change their behaviour as a result of it. Policy support can also lead to market distortion away from supply and prices that match with consumer preferences. Entrepreneurship policies should seek to reduce barriers to business creation without encouraging unsuited individuals to start unsustainable businesses or subsidising start-ups which do not need it.

  • Many of the impacts of entrepreneurship policies tend to occur over long timelines. The time needed to influence the entrepreneurial culture and the overall business birth rate of a place is likely to be much longer than the time to influence specific achievements among specific entrepreneurs, such as developing a new product of market. Therefore the judgements on whether policy is effective or not need to be made after allowing sufficient time for the policy to have an effect, and appropriately timed evaluations are needed.

  • Monitoring and evaluation should be built into policy and programmes from the start with proportionate but adequate resources allocated. Mechanisms to incorporate results into programme revision and future policy developments should be included.

Implementation is as important as design for policy effectiveness. Factors for success include choosing the appropriate delivery agency or partner and providing them with adequate resources, setting relevant targets and including process for feedback from frontline programme workers and users.

International experiences in entrepreneurship policy highlight effective good practices in design and implementation. Based on the selected 16 case studies presented in Part II of the present report, key lessons from these initiatives include:

  • Involving stakeholders in identifying issues and designing programmes is a success factor reported for initiatives in all domains (from regulatory improvements to direct support and ecosystem initiatives). This is essential in ensuring issues are identified accurately and solutions are delivered in a format that matches recipient capacities and takes full advantage of existing support actors. This is especially important when developing local level initiatives and initiatives seeking to develop ecosystems. This also includes building feedback loops so that that inputs from users and other stakeholders can be collected and analysed.

  • Setting clear goals and corresponding Key Performance Indicators is important for successful implementation. A lack of clear objectives is a common challenge to programme implementation. It also hinders effective monitoring and evaluation, which are essential to achieving results.

  • Programmes should consider investing resources in awareness raising among entrepreneurs. Well-designed programmes may face issues in up-take or in reaching to the appropriate entrepreneurs and enterprises if no resource is set aside for outreach. This is an especially common pitfall in schemes seeking to provide financial incentive to investors, as these usually include very diverse actors who traditionally do not seek information and advice from business support providers.

  • Programmes should seek to minimise “red tape” and allocate resources to administrative tasks. Delivery of programmes, especially those involving screening and/or funding can create significant burden on applicant firms and programme staff. This burden should not be underestimated to ensure that programmes have sufficient capacity to run smoothly without undue barriers to uptake.

  • Sufficient resources should be allocated to monitoring and evaluation, and programmes should consider measures to facilitate reporting and build capacity. Monitoring and evaluation are essential to achieving results, but require time and resources for the programme manager. This may be especially challenging for local providers with limited capacity. Reporting systems should be designed to minimise time involved and facilitate reporting by untrained staff. This may include the use of support tools and explicitly budgeting time and resources.

  • Programmes should be designed to fit within the existing policy portfolio and consider possible interactions with existing programmes. This is especially important for programmes seeking to create financial incentives (as their magnitude needs to match other incentive programmes to have an effect). It is also important to provide appropriate cross-referrals to enable entrepreneurs to benefit from other programmes that may help them overcome other challenges, particularly as start-ups grow and face new challenges. The fact that start-ups may benefit from packages of support should also be taken into account in evaluation.

  • Policymakers should consider embedding key programmes in national and local strategies, to foster cohesive action and signal political support. The latter is particularly important for entrepreneurial ecosystem initiatives and one-stop-shops, where strong political support may help reinforce the legitimacy of a body as a central facilitator in the system, when it may otherwise be perceived as a temporary initiative and/or another competitor in the business development support market.

  • Programmes benefit from leveraging existing capacity. This includes relying on existing ecosystem actors (including non-governmental providers), possibly developing a certification model to better link public and private sector business development support and help entrepreneurs navigate the offer. It may also involve leveraging actors for outreach and building around existing structures rather than starting new initiatives from scratch to facilitate uptake.

  • Policies should consider supporting ecosystem actors of different sizes. One-stop-shops and large-scale incubators and support providers have capacity to serve large numbers of firms and hold a range of capacities in-house. However, experiences from local programmes show that small scale development support providers can sometimes have advantages, especially when local knowledge is critical and when in-person interactions play a role (e.g. in incubator settings where inter-personal exchanges between entrepreneurs play as important a role as formal training)

  • Providing training and capacity building to delivery staff may be needed to ensure programmes are rolled out effectively. This is especially true when programmes involve the use of new tools or the launch of platforms.

  • Creating networks between local initiatives to facilitate peer-learning contributes to improved practices over time. This is especially important for “one-stop-shop” initiatives embedded at the local level, as a centralised hub and spokes model may miss opportunities for learning from experience.

  • Programmes seeking to reinforce entrepreneurial ecosystems should invest in building linkages between existing programmes and actors. Lack of linkages between ecosystem actors may affect the ability of entrepreneurs to find partners and suppliers, access resources and business support. Similarly, lack of co-ordination between support providers may undermine the efficiency of otherwise well-designed support programmes. Resources and actors in ecosystems are as important as the linkages between them.

Table 3.3 (in Chapter 3) highlights key success factors illustrated by the different case studies presented in the report.

Monitoring and evaluation are critical in providing information to guide the effective design and implementation of policy. The monitoring of a policy or programme describes the inputs (e.g. budget, resources), record the outputs (e.g. number of participants, take-up rates), and may collect the opinions of the managers of the programme, stakeholders and participants. Evaluation, in contrast, involves the application of sophisticated methodologies to quantitative and sometimes qualitative data, to formally measure what difference the policy or programme has made to the businesses that were assisted and to the wider economy.

Governments should monitor and evaluate policies to establish whether interventions have contributed to correcting the problems that they have set out to solve as well as to assess the economic efficiency of interventions. This is particularly important as entrepreneurship policies target complex mechanisms and are expected to create effects over long timeframes. Monitoring and evaluation are also required to improve the design and administration of interventions to increase performance and value-for-money.

Evaluation is also fundamental to public accountability. Entrepreneurship programmes involve substantial public expenditure, both direct and indirect (e.g. forgone tax revenues), creating a need for governments to demonstrate impact. As government budgets are constrained, monitoring and evaluation can help identify those interventions that lead to the most desirable outcomes and which could be expanded and those that could be contracted. As entrepreneurship policy portfolios are complex and policies interact with one another, systemic evaluation may help governments make appropriate adjustments to the entrepreneurship policy mix.

However, the cost – and therefore the sophistication – of the monitoring and evaluation methods selected should be proportionate to the cost of the assessed intervention. Interventions with smaller budgets, as is often the case for regional and local programmes, may therefore need to opt for less expensive evaluation methods.

The OECD Framework for the Evaluation of SME and Entrepreneurship Policies and Programmes (OECD, 2008) has identified a six-step approach to monitoring and evaluation, ranging from the simplest methodology (Step I) to the most sophisticated and reliable one (Step VI) (Table 1.1).

Monitoring (steps I-III) includes measuring take-up rates and recording applicants’ characteristics as well as asking participants for their views on the value of the programme and the impact they believe it had on their own behaviour or on their business. The evidence collected in steps II and III is typically qualitative, self-report data derived from surveys. It can be useful for tracking programme usage, but is not reliable for establishing programme impact.

Evaluation (steps IV-VI) involves a comparison between firms that received assistance (the treatment group) and those which did not (the control group) in order to establish the counterfactual situation. This comparison is based on outcome measures of policy impact, such as sales, employment or survival. Comparing the actual changes in the assisted firms with the counterfactual is interpreted as the impact of policy – the additionality. It can look at average firms (step IV), or, for more accuracy, compare treated firms with enterprises with similar characteristics that did not benefit from the programme. Indeed, entrepreneurship support participants may often differ from the average firm, which can lead to over- or under-estimating the impact of a programme. Step VI, which is the most complex methodology, includes methods to address selection bias. These methods are typically based on quantitative evidence (either from official data sources or collected directly from the firms themselves). Recent entrepreneurship policy evaluation has started to use Randomised Control Trials as an alternative approach to establishing impact. The higher levels of steps are likely to provide more reliable estimates of programme impacts.

Other factors that are helpful to consider while designing an evaluation framework include capturing deadweight (activity related to the intervention that would have gone ahead without even in the absence of support), displacement (when an activity subsidised or supported by the government displaces another activity) and spillover effects (when interventions to support specific firms have a positive impact on firms that are not the direct target population of the support).

Monitoring and evaluation should be designed to assess policy against carefully defined policy objectives. To measure progress towards these goals, appropriate key performance indicators (KPIs) – quantifiable measures that can be tracked over time – should be chosen.

KPIs should be set to monitor the activity of a programme. This includes metrics related to outreach and uptake, such as the number of programme participants, metrics describing the delivery process, such as indicators of costs, and measures of satisfaction of users and stakeholders.

Programmes should also select KPIs to quantify direct outputs and impacts. These can include measures of the number of firms created after a pre-start-up programme, number of jobs created by supported firms, amount of funding secured by firms supported in accessing finance, and survival rates and economic performance of supported entrepreneurs.

Finally, KPIs aiming to capture broader impact of a policy intervention should also be defined. These can include quantitative measures (such as growth in value added) and incorporate econometric techniques to attribute effects to an interventions. Measurable indicators seeking to capture more qualitative developments can also be set (e.g. estimates of the linkages between certain types of actor through formal collaboration as a proxy to ecosystem development).

The first step of designing KPIs is identifying the objectives of an intervention. A concrete and measurable output reflecting each objective should then be identified, and a target set for it. It is important to also develop a clear understanding of how the intervention is expected to lead to the desired outcome, i.e. through which channels will the intervention affect the variable observed. This step will be important in setting efficient KPIs to monitor the implementation of a policy or programme over time and adjust course in early stages, before outputs or impact can be accurately measured.

Once KPIs have been defined, a strategy for measurement should be clearly outlined. This may involve developing measurements and gathering data in-house (e.g. counting beneficiaries, developing satisfaction surveys) or using existing data sources. The methodology should involve a timeline and method for data collection and treatment. Finally, responsibility should be assigned for achieving different KPIs and resources should be set aside for tracking progress.

KPIs are not universal. The choice of KPIs depend on a number of factors, including the type of intervention (which affects the type of outcome expected and the timeline of results), its scale, and the capacity of the programme (i.e. resources to conduct measurements or reliance on existing sources, type of evaluation foreseen).

Indeed, while entrepreneurship policies seek out similar ultimate goals, their outputs can differ widely. For example, an intervention seeking to integrate entrepreneurship education into formal education curricula is expected to affect students’ mind-sets. However, it will not translate into significant changes in entrepreneurial activity for years or decades, and such changes would be extremely difficult to attribute to such an intervention. As a result, a programme of this type should set monitoring KPIs around the number of students exposed and the quality of their exposure, and set evaluation KPIs around measurements of the knowledge of students around entrepreneurship, their opinion on it, and measurable progress in concrete entrepreneurship skills rather than focus on “traditional” measures of entrepreneurship such as start-up rates.

By contrast, a programme seeking to support entrepreneurs in exporting may expect quantifiable short-term changes in participants’ activities. Such a programme may choose to use KPIs measuring the number of firms that export, their export intensity, and the number of external markets they export to.

Table 3.2 (Chapter 3) presents an overview of the main types of KPIs used to assess the entrepreneurship policies presented in this report.

An important challenge to the evaluation of entrepreneurship policy is the presence of unobservable selection bias: firms with growth ambitions can be expected to be more likely to apply for support. Even without support such firms would have likely outperformed the average, resulting in an over-estimation of the effects of the programme. A similar effect arises with schemes that involve a selection process, which is likely to eliminate firms that are likely to perform poorly (OECD, 2008). Conversely, schemes targeting laggard firms will have the opposite bias. Moreover, measurement of interventions that have an economy-wide impact (e.g. tax, culture) will be difficult in the absence of a natural control group.

It can also be difficult to isolate the impact of any single programme since policies in other areas also affect entrepreneurial activity. The identification of a policy effect also does not indicate the causal link between the input and the outcome – the mechanisms by which the policy inputs leads to the observed outcome. Feedback effects may also be at play.

Another challenge is selecting an appropriate timing for an evaluation, as impact may occur in a timeline ranging from months to years depending on the type of intervention and context. Moreover, some schemes will have a one-off impact whereas other programmes may have both immediate and longer-term effects.

Finally, the choice of the entity carrying out an evaluation can also be challenging, as concerns for in-depth understanding of a programme’s design, which might imply involvement of the programme managers in evaluation, may not align with the goal to ensure objectivity of the assessment, suggesting an external expert. For accurate impact assessment it is preferable to use an independent evaluator.

It is also important to adequately reflect on evaluation outcomes, assessing not only the effect of the policy being evaluated but also reviewing the accuracy of the initial diagnostic, the approach taken to address the issue and the way the intervention was delivered.

Several principles can be identified for good practice evaluation of entrepreneurship policies:

  • Policy interventions should be evaluated against clearly specified objectives. Setting clear objectives and identifying the mechanisms through which the intervention is expected to reach them is key to identifying targets or key performance indicators against which outcomes can be measured.

  • Monitoring and evaluation need to be based on appropriate data. SME policy evaluation can take advantage of existing relevant data held in different parts of the administration (e.g. tax records, unemployment registry data). Recent efforts to broaden access to public data for research conducted in many OECD countries could facilitate this process (OECD, 2018a). The development of “big data” also holds promise for improving evaluation. Data should be available at appropriate time intervals and levels of disaggregation and refer to an outcome indicator that is relevant for the foreseeable future. Some programmes will require dedicated data collection exercises (OECD, 2018c).

  • When possible, evaluation needs to take full advantage of the application of rigorous evaluation techniques that use appropriate counterfactuals and can correct for selection bias. One example is Randomised Control Trials (RCT), whereby the performance of a group of eligible recipients is compared over time with those eligible recipients who were randomly excluded in order to establish a counterfactual. A number of recent RCT evaluations have been undertaken in the area of SME and entrepreneurship policy (OECD, 2018c).

  • Evaluation techniques need to accommodate the diversity of entrepreneurship programmes – each with their own objectives, targets, instruments and impacts. This will involve different data sources and timescales, with some programmes intended to have short-term impacts (e.g. export support) whereas others will only have an impact in the longer-term (e.g. innovation support).

  • Monitoring and evaluation should take into account possible interactions between the outcomes of various entrepreneurship policies and programmes to avoid attributing outcomes to specific programmes that arise from other interventions. The impact of other policies affecting entrepreneurship (e.g. in the areas of taxation, social security, business regulation, immigration and competition) should also be taken into account in the evaluation of entrepreneurship policies.

  • Effective monitoring and evaluation are included from the early-stage of the policy-making process. Evaluations should be embedded in the policy cycle rather than be designed after the launch of a programme. Moreover, it is important that monitoring and evaluation evidence does not discourage or deter policy experimentation and acceptance of failure (OECD, 2018c).

  • Accurate evaluations should favour a broad focus, looking beyond outcomes to consider other factors that may impact the effectiveness of an intervention. This may include the definition of the eligibility criteria, awareness of the programme among the target groups, and local context.


Acs, Z. and P. Mueller (2008), Employment effects of business dynamics: Mice, Gazelles and Elephants, Small Business Economics, 2008, Vol. 30/1, pp. 85-100.

Baumol, W. J. (1990), Entrepreneurship: Productive, Unproductive, and Destructive, Journal of Political Economy, Vol. 98/5, pp. 893-921.

Brown, R. and C. Mason (2017), Looking Inside the Spiky Bits: A Critical Review and Conceptualisation of Entrepreneurial Ecosystems, Small Business Economics, Vol. 49/1, pp. 11-30.

Cowling, M. (1998), Regional Determinants of Small Firm Loans Under the U.K. Loan Guarantee Scheme, Small Business Economics, Vol. 11/2, pp. 155-167.

Fritsch, M. and M. Wyrwich (2014), The Long Persistence of Regional Levels of Entrepreneurship: Germany, 1925–2005, Regional Studies, Vol. 48/6, pp. 955-973.

Harrison, R. T. and C. M. Mason, (1986), The regional impact of the Small Firms Loan Guarantee Scheme in the United Kingdom, Regional Studies, Vol. 20, pp. 535-549.

Mason, C. and R. Brown (2013), Creating good public policy to support high growth firms, Small Business Economics, Vol. 40/2, pp. 211-225.

OECD (2019), Regions in Industrial Transition: Policies for People and Places, OECD Publishing, Paris, https://dx.doi.org/10.1787/c76ec2a1-en.

OECD (2018a), OECD Science, Technology and Innovation Outlook 2018: Adapting to Technological and Societal Disruption, OECD Publishing, Paris, https://dx.doi.org/10.1787/sti_in_outlook-2018-en.

OECD (2018b), OECD Regions and Cities at a Glance 2018, OECD Publishing, Paris, https://dx.doi.org/10.1787/reg_cit_glance-2018-en.

OECD (2018c), Monitoring and evaluation of SME and entrepreneurship programmes, Policy Note, SME Ministerial Conference, 22-23 February 2018, Mexico City, www.oecd.org/cfe/smes/ministerial/documents/2018-SME-Ministerial-Conference-Parallel-Session-6.pdf.

OECD (2017), Business Dynamics and Productivity, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264269231-en.

OECD (2008), OECD Framework for the Evaluation of SME and Entrepreneurship Policies and Programmes, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264040090-en.

OECD/European Union (2016), Inclusive Business Creation: Good Practice Compendium, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264251496-en.

Stam, E. (2015), Entrepreneurial Ecosystems and Regional Policy: A Sympathetic Critique, European Planning Studies, Vol.23/9, pp.1759-1769.

Stinchcombe, A.L. (1965), Social structure and organizations, in March, J. (Ed.) Handbook of Organizations, Rand McNally, Chicago, IL, pp. 142‐93.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2020

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.