Entrepreneurial Ecosystem Diagnostics

3. Cross-country diagnostic results

Abstract

This chapter presents the main findings from the analysis of entrepreneurial ecosystem diagnostics data. It offers a performance overview for each country on each of the ten ecosystem elements and explores areas of progress or decline relative to earlier periods. The chapter also examines entrepreneurship levels by country and the extent of regional and social variations.

Performance overview across elements

The core of the diagnostics is a set of composite indexes that measure the ten entrepreneurial ecosystem elements. While some countries tend to perform better than others on multiple elements, no single country attains the top score in every element. In only two cases does a country attain the highest score on two elements. This shows that countries have advanced unevenly in the construction of their ecosystems, and no economy today is yet ready to offer a perfect context for entrepreneurs to operate. Even the most “entrepreneurship friendly” economies have areas to improve and can learn from the experience and policies of other countries.

Element summary scores

Table 3.1 presents the entrepreneurial ecosystem element scores by country. These results show significant variation both in element scores across countries and across elements within countries. Both types of information are relevant for policy development, with low scores on either indicating potential for remediating policy actions.

The variation in scores across countries and elements indicates that there are no rigid patterns across scores. Countries may attain a relatively high (low) score in one element, even if they perform relatively poorly (strongly) in most of the other elements. However, looking horizontally at performance within countries across the different elements, it is possible to identify four general types of ecosystems: i) ecosystems that are mostly still underdeveloped, where a large number of elements have below-OECD-average scores (e.g. Colombia, Costa Rica); ii) ecosystems that have started to develop but attain a mostly average performance across all elements and do not excel in any aspect (e.g. Czechia, Slovenia); iii) ecosystems that have developed unevenly with about average performance in some elements and strong performance in few other elements (e.g. France, Germany); iv) and ecosystems that are well advanced on the majority of the elements (United States, United Kingdom, or Switzerland).

Table 3.1 also helps show the relative performance of a country on different elements compared to the international context, by looking vertically in the table. For example, Australia attains relatively high scores on Markets, Talent and Leadership, and Intermediate Services, while its lowest scores are on Institutions and Knowledge. These can be interpreted as the weakest links in the ecosystem, which should be addressed first. However, by contextualising these scores it is possible to see that the distance between Australia’s scores and top performer scores is larger on Knowledge than Institutions, which may indicate that addressing the Knowledge gap could be more impactful than addressing gaps in other elements.

Clearly, these data provide only an initial snapshot of the situation in a country. To move to action, policy makers should follow up the diagnosis produced by the tool with a deeper analysis on nature and causes of the performance.

Table 3.1. Summary scores by entrepreneurial ecosystem element, 2020-2023 period

Country	1. Institution	2. Culture	3. Networks	4. Infrastr.	5. Markets	6. Finance	7. Knowledge	8. Talent	9. Leadership	10. Interm. Services
Australia	40.3	.	49.9	55.0	68.0	.	43.2	64.0	73.6	65.9
Austria	49.6	43.0	49.3	59.8	51.9	50.8	57.0	55.4	47.9	54.8
Belgium	48.1	.	63.2	52.1	54.2	58.4	57.2	.	41.3	50.3
Canada	53.4	70.0	61.2	52.9	64.5	44.9	51.6	69.1	90.9	74.2
Chile	50.4	38.8	25.9	17.3	21.0	.	18.1	35.0	30.4	32.8
Colombia	6.6	14.7	.	5.3	25.1	.	17.8	3.0	37.5	17.7
Costa Rica	9.6	.	.	5.7	15.0	.	19.7	2.2	23.6	14.9
Czechia	51.7	.	44.9	32.3	36.0	39.7	42.2	.	27.7	36.2
Denmark	67.9	.	51.5	71.9	51.3	56.6	69.2	69.4	51.5	50.7
Estonia	65.3	43.7	44.8	51.4	24.1	45.6	48.9	61.3	33.8	48.6
Finland	68.1	.	65.4	71.5	51.0	57.6	72.1	52.9	60.2	61.1
France	51.0	27.1	40.4	63.5	70.3	44.1	48.9	42.0	70.0	51.5
Germany	46.9	53.1	50.8	59.9	69.2	49.5	60.2	52.1	84.0	49.3
Greece	31.8	37.3	6.7	46.4	26.3	38.2	29.4	27.7	32.7	34.4
Hungary	18.1	40.7	30.9	35.7	12.4	36.7	33.7	38.5	25.4	33.2
Iceland	56.3	.	50.8	57.2	4.4	.	65.2	62.6	7.0	46.9
Ireland	76.2	46.0	54.2	45.2	53.7	54.1	47.5	56.3	47.9	64.4
Israel	42.0	61.9	.	39.4	12.0	65.3	86.1	40.7	79.1	78.7
Italy	36.6	18.3	44.3	51.1	53.8	47.7	30.2	27.9	56.0	46.2
Japan	37.2	10.4	49.0	52.5	80.6	.	54.8	20.7	49.5	27.2
Korea	44.6	55.9	49.6	66.0	79.1	.	81.2	73.6	45.0	36.1
Latvia	54.8	28.0	27.5	39.5	20.7	36.6	30.9	57.6	29.7	27.2
Lithuania	62.5	43.3	42.9	44.4	33.9	37.3	35.8	50.0	17.2	30.1
Luxembourg	42.7	.	46.9	47.6	32.1	.	39.0	.	33.8	52.0
Mexico	4.6	28.3	.	6.9	35.3	.	14.1	3.7	50.1	29.3
Netherlands	60.1	.	57.8	55.3	70.9	57.3	64.6	51.5	55.2	69.1
New Zealand	48.6	.	.	44.4	45.9	.	44.7	49.0	37.1	50.7
Norway	66.0	82.0	.	59.3	59.0	52.6	58.5	57.1	53.4	43.9
Poland	58.6	28.7	17.4	29.4	63.3	41.3	34.6	53.3	49.5	32.6
Portugal	36.4	.	35.3	45.9	53.5	38.7	37.5	27.2	38.8	41.9
Slovak Republic	43.7	23.6	15.2	34.5	30.1	38.7	26.6	45.2	18.4	21.6
Slovenia	51.1	57.2	35.8	43.5	30.6	39.4	40.0	58.0	26.9	37.9
Spain	48.6	23.6	25.1	47.8	66.4	50.9	35.0	44.6	68.4	51.8
Sweden	70.1	67.1	54.6	69.7	60.9	46.7	79.7	57.8	62.9	67.9
Switzerland	63.0	33.7	.	77.0	55.0	52.2	87.3	62.5	68.4	76.6
Türkiye	3.0	39.8	23.1	25.2	31.4	.	21.7	14.8	45.3	33.4
United Kingdom	70.9	69.2	67.3	51.5	73.6	41.0	57.1	64.7	97.5	83.0
United States	52.6	67.0	95.3	54.0	86.8	59.3	65.5	70.6	100.0	79.8

Note: Scores are composited indices, expressed as a 0-100 value, computed as the geometric means of the underlying indicators. The methodology is described in Annex A. Continuous shading indicates the level of performance, with darker blue shadings indicating higher scores. A dot indicates that there were not enough data for the reference period to compute an element score for a country. Element scores are not computed for countries if more than 1 indicator composing an element is missing or imputed.

Source: OECD’s Entrepreneurial Ecosystem Diagnostics

Variation in country performance by ecosystem element

Figure 3.1 shows the share of OECD countries with scores in different brackets across the entrepreneurial ecosystem elements. Although the diagnostic tool reports relative performance across countries, there are differences in the variation of the scores around the averages. There are some elements with relatively high variation in scores across countries and others with relatively low variation. Elements with relatively high variation would tend to indicate greater scope for making improvements in entrepreneurial ecosystem conditions. One of the ways of exploring this further is to examine the share of countries with low scores on each element.

Institutions is the element where the smallest share of countries attained low scores. Only 24% of countries attained scores in the bottom 40% of values. On the other hand, larger shares of OECD countries attained low scores (in the bottom 20% of values) on Markets, Knowledge, Leadership, and Intermediate services. These are common weak links across OECD countries, which should be prioritised by the majority of OECD countries.

Figure 3.1. Country performance across entrepreneurial ecosystem elements

Note: These data show what percentage of the 38 OECD countries achieve an aggregate element score within each of the five bands (0-20 band, 20-40 band, 40-60 band, 60-80 band, 80 to 100 band) for the 2020-2023 period. Values below 5% are not visualised in the chart.

Source: OECD’s Entrepreneurial Ecosystem Diagnostics

The variability by element can also be analysed in terms of differences between top and bottom performers in each element. Comparing the average scores of the top three countries with the average scores of the bottom three countries, Leadership emerges as the area with the largest score difference between top and bottom performers. This indicates that it is one of the differentiating factors of ecosystems that can actively support entrepreneurs. The gap between top and bottom performers is the smallest for the Finance element overall score, suggesting less room for large numbers of countries to catch up with leaders, although within that element, venture capital investment gaps between top and bottom performers are important.

Figure 3.2. Performance difference between top and bottom scores by element

Note: Difference between the average score of the top-3 countries and the average score of the bottom-3 countries for the period 2020-2023.

Source: OECD’s Entrepreneurial Ecosystem Diagnostics

Performance on sub-elements and evolution over time

This section examines country performance by element. In each section, a first chart shows countries’ scores sorted on the 2020-2023 period’s element scores from weak performers (on the left of the charts) to top performers (to the right of the charts). The chart also features the evolution of country performance on each element over the three data periods (2016-2020, 2018-2022, and 2020-2023) of the tool. A second chart for each section then presents the values of the sub-indicators that contribute to the element summary scores. The data are presented in terms of normalised scores. They are sorted from weak performers (on the left of the charts) to top performers (to the right of charts). Countries for which data were not available for the 2020-2023 period are not featured in these charts. In some cases, data are available for the 2018-2022 period or the 2016-2020 period. These data are provided in the country profiles in chapter 4.

1. Institutions

The Institutions element measures the extent to which a country has in place an administrative system, regulatory structure and taxation levels that facilitate economic activity and allow entrepreneurs to operate. The institutional framework determines the conditions in which new businesses are created and incentivised to grow. It is measured through four indicators that capture a country’s control of corruption, rule of law (justice system, crime, property rights), product market regulation (degree to which policies are red tape promote or inhibit competition), and taxation rates.

Figure 3.3 shows aggregate scores on Institutions across the OECD countries, sorted from the lowest to the highest score in the 2020-2023 period. Ireland, the United Kingdom, and the Scandinavian countries (Denmark, Finland, Norway, Sweden) are among the countries offering the most conducive institutional contexts for productive entrepreneurship.

Most of these countries have maintained a relatively stable performance since the 2016-2020 period, with the United Kingdom’s score slightly improving over the past four years. Ireland has performed at the top throughout the period. Ireland scores slightly below Scandinavian countries in terms rule of law and control of corruption levels but it offers significantly more favourable corporate tax rates and product market regulation, making its ecosystem a particularly low-burden environment for start-ups and scale-ups.

Figure 3.3. Institutions element scores by country and year

Note: These scores are computed as the geometric mean for countries across the following indicators: i). Rule of law (crime, enforcement of contract, legal process/courts transparency, speed of judicial processes, risk of expropriation of foreign assets, intellectual property protection, private property rights, source: World Bank – Worldwide Governance Indicators; ii.) Control of corruption index, Source: Vi-dem Project; iii.) Effective tax rate, % taxable income, Source: OECD - Corporate Tax Statistics Database; iv. Product Market Regulation Index, Source: OECD. Before aggregation, data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023’s data and must be interpreted as relative performance the 2020-2023 period.

Among the countries where the Institution scores have varied the most over time, France, Belgium, and Luxembourg have improved, while Hungary, Portugal, Spain, and Australia have registered a relative decline in their performance.

The performance of countries varies significantly across the indicators composing the Institutions element (Figure 3.4). Two indicators (rule of law and control of corruption) capture important aspects of the functioning of the public administration, one indicator measures the extent to which policies promote or inhibit competition, and one indicator measures the tax burden for businesses.

Some countries perform consistently well over time across the two public administration indicators. Notably, Scandinavian countries are well positioned across both aspects. There are however countries that tend to perform better on one of these two dimensions than the other. For instance, in Slovenia, entrepreneurs can benefit from relatively sound rules of law, yet they need to cope with lower control of corruption levels than in other OECD countries. The opposite is true for Spain where control of corruption is above the OECD average, but the implementation of the rule of law is below that of many countries. Most countries tend to show relatively stable performances over time on these dimensions, underlining the structural feature of institutional mechanisms.

In terms of product market regulation (PMR), Sweden, Lithuania and Irelands are among the countries where entrepreneurs can benefit from accessible markets and absence of red tape. This is also a factor where countries do not change performance rapidly over time. The only exception is Greece, who managed to improve its PMR score significantly between the 2018-2022 period and the 2020-2023 period.

On taxation, corporate tax rates in Hungary, Ireland and Poland are among the lowest in the OECD contexts, contributing to incentivize business activities. Notably, in Hungary, an above average score on low tax rates, contrasts with below OECD-average performances on product market regulation control of corruption and rule of law. Over time, few countries, including France, and Costa Rica made important efforts in reducing their tax rates, resulting in scores upgrades. In several other countries the tax rates have increased including the United Kingdom, who, however, has remained one of the OECD countries applying moderate corporate tax rates.

Figure 3.4. Institution sub-element scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023’s period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

2. Culture

The culture element measures the extent to which people in the country tend to have a propensity to entrepreneurship and the extent to which a country’s social norms, values and customs reward entrepreneurial efforts. These cultural and value systems can play an important role in boosting or holding back entrepreneurship. The Culture element is measured through three indicators that capture the share of adults who consider entrepreneurship as a desirable career choice, the share of adults who believe that successful entrepreneurs attain a high social status, and the share of adults who believe that most people can be trusted.

The aggregate Culture element scores for all countries are presented in Figure 3.5. For many countries data on these dimensions have not been updated recently, thus only a sample of 27 OECD countries is available for the 2020-2023 period. Among these countries, the top three are Norway, Canada and the United Kingdom who all achieve score between 69 and 82/100. Other countries performing well, with scores above 60/100 are Israel, the United States, and Sweden.

Among these well-performing countries in this area, Norway has slightly declined since the 2016-2020 period, Israel has remained stable, and the other four countries have improved. The United Kingdom is the countries among the top performers who improved the most.

Beyond the top performers, countries who did register a significant shift, include Poland, France and to a lesser extent Switzerland and Colombia, have seen their culture scores declining, while Italy, Spain, Ireland, Korea and Slovenia’s scores are on an upward trend.

Figure 3.5. Culture element scores by country, and year

Note: These scores are computed as the geometric mean for countries across the following indicators: i). Percentage of 18-64 population who consider starting a business as a desirable career choice, Source: Global Entrepreneurship Monitor (GEM); ii.) Percentage of 18-64 population who agree that successful entrepreneurs receive high status, Source: Global Entrepreneurship Monitor (GEM); iii) Share of people who believe that most people can be trusted, Source: World Value Survey (WVS). Before aggregation, data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

Analysing countries’ performances at sub-element indicator level provides more detailed and diverse picture than the aggregate element scores. Among the three top element performers (Norway, Canada and the United Kingdom), Norway is among the top 5 performers on two indicators (trust in others and status of successful entrepreneurs). Canada and the United Kingdom also performs well on two indicators (Entrepreneurship as a good career choice and status of successful entrepreneurs) and less well on trust. None of the top countries in each indicator is among the elements’ best. Korea is first in terms of High status of successful entrepreneurs but performs below the OECD average on the other two indicators. Chile excels on perceptions of entrepreneurship as a good career choice but attains low levels of trust, while Denmark is the country where trust levels are the highest, but due to an incomplete dataset it cannot be reliably assessed on the other two aspects.

Notably, the scores on the indicators of entrepreneurship sentiment fluctuate over time significantly more than indicators measuring societal values such as trust in others, which are arguably structurally more difficult to change over a short time period. The score of the indicator Entrepreneurship as a good career choice and the scores of the indicator High status of successful entrepreneurs have increased in most OECD countries over the past four years. Notably, these shares were low around the period of the pandemic as conditions for entrepreneurship were seen less positively temporarily in many OECD countries. The upward trend in most OECD countries indicate a return of greater interest towards entrepreneurship post-pandemic. Among the countries that have registered the higher increase in perceptions in terms of entrepreneurship as a good career choice are the United Kingdom, the United States, Slovenia, Ireland, and Italy. While in terms of high status of successful entrepreneurs, sentiment increased particularly in Korea, Slovenia and Japan. There are however few countries where sentiment has been on a declining trend in this period including Poland, Colombia and Türkiye on entrepreneurship as a good career choice, and France and Colombia on High status of successful entrepreneurs. Going forward, it will be important to monitor the evolution of these perceptions over a longer time period to assess if there have been permanent effects on entrepreneurial intentions years after the pandemic.

Figure 3.6. Culture sub-element scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

3. Networks

The network element measures the extent to which start-ups and scale-ups can leverage collaborations with other firms and stakeholders so that they can access information, technology, support, finance, form partnerships, and identify customers and/or suppliers. It is measured through two indicators, the extent to which SMEs collaborate with each other on innovation, and the extent to which firms collaborate with universities on R&D.

The aggregate scores (provided in Figure 3.7), show that the United States, with a score of about 95/100 benefits from the most interconnected network, while Canada, Belgium, Finland, and the United Kingdom also perform among the best countries, all exceeding a score of 60/100.

Among the top performers, Canada has slightly improved its performance since the 2016-2020 period, while the other top countries have remained substantially stable. Among the other countries, Czechia, Italy and Hungary have improved their network scores over the past four years, while Japan has registered the most noticeable declines in score.

Figure 3.7. Networks element scores by country and year

Note: These scores are computed as the geometric mean for countries across the following indicators: i.) Share of SMEs with innovation cooperation activities, Source: European Commission - European innovation scoreboard; ii.) Extent to which businesses collaborate with universities on R&D, Source: World Economic Forum, Executive Opinion Survey. Before aggregation, data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023 data and must be interpreted as relative performance the 2020-2023 period.

Looking at the two indicators that compose the element, the United States is by far the countries where SMEs cooperate on innovation. About 70% of US SMEs engage in innovation cooperation activities, and they have done so on a consistent basis over the past four years at least. This share is much lower in other countries, which explains the difference in score between the United States and the United Kingdom, the second on this aspect.

These scores have remained substantially stable over time for most OECD countries. The only notable exception is Japan, where SMEs collaboration has been on decline over time past few years.

The other aspects on which countries are benchmarked against is the extent to which businesses (not just SMEs) interact with universities for R&D projects. Israel, Switzerland, and the United States stand out on this dimension, attaining scores above 90/100. They are followed by Netherlands, Canada, Finland, Belgium, and Sweden who also perform well above average, with scores between 70 and 80/100. Among them Canada has gradually improved its performance over time, while Finland has slightly declined. Other top countries have remained stable. There also some countries that have improved their performance recently but remained in the lower part of the distribution. Among them the two most noticeable progresses are those of Italy and Czechia.

Figure 3.8. Network sub-element scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

4. Infrastructure

The infrastructure element measures the extent to which transport and telecommunications provide effective ways for people to connect, this in turn can facilitate the absorption of new ideas and technologies, and the delivery of goods and services. It is measured through three indicators: use of fixed telecommunications (in terms of subscriptions), the amount of mobile data downloaded (in terms of Kilobites per subscription, per month), and the overall quality of transport infrastructure.

Switzerland attains the highest aggregate score on the infrastructure element. It is followed by Denmark and Finland. These three countries are the only ones attaining a score above 70/100. Other countries that perform relatively well include Sweden, Korea, and France, who attain scores between 60 and 70/100. On the opposite front, Mexico, Colombia, Costa Rica, and Chile still have yet to close significant infrastructure gaps vis-à-vis other OECD countries.

Figure 3.9. Infrastructure element scores by country and year

Note: These scores are computed as the geometric mean for countries across the following indicators: i.) Fix broadband, subscriptions per 100 population, Source: OECD - Telecommunications database, iii.) Mobile data use, Gb/per subscription/month, Source: OECD – Broadband and telecommunications database; iv.) Transport infrastructure quality index, Source: World Bank - Logistic Performance Index (LPI). Before aggregation, data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

On average, most countries are improving their performance over time, especially thanks to greater use of IT services and infrastructure. Improvements, however, vary considerably across countries. Some countries upgrade their scores very little, other have made important strides. For example, in the case of Chile and Türkiye, improvements have started to close the gap between these countries and the average OECD level, compared to the 2016-2020 period.

Looking at the indicators composing the element score, the use of fixed broadbands and mobile data do not always go hand in hand. Only Denmark and to a lesser extent Switzerland attain high scores on both aspects. France, and Norway instead attains the highest scores among OECD countries on the use of fixed broadband but are less advanced on the use of mobile data. The countries leading on the use of mobile data include Austria, Estonia, Finland, Iceland, Latvia, Lithuania. All of them feature significantly lower scores on the use of fixed broadband internet, possibly indicating that they rely more on mobile technology than on fixed broadband.

Over time, fixed broadband scores have slightly improved over time for most countries, with Portugal, Czechia, and Slovak Republic recording the most visible increases in scores. More variability is found in terms of mobile data use. This is a dimension where technologies have changed significantly recently, and most countries have increased their scores. Some countries were able to leapfrog while have only marginally improved. Among the countries that improved the most, Lithuania, Iceland, Austria, and Italy have all been able to expand the use of mobile data significantly, while Mexico and New Zealand have improve less than others, the facto losing ground on this technology.

In terms of transport infrastructure quality, Germany, Switzerland, Japan, Sweden, and the Netherlands attain the higher scores among OECD countries, passing the 80/100 mark. On the opposite front Colombia, Mexico, and Costa Rica have yet to bridge important gaps on transport infrastructure with respect most other OECD countries.

Transport quality scores have also shifted in most countries over time. Since the 2016-2020 period about two-thirds of OECD countries have experienced relative improvement in their scores, and roughly one-third of OECD countries have reduced their relative score. Among the countries that improved the most, Iceland, Greece, and Portugal stand out, while among the countries that experienced relatively large declines including Luxembourg, Czechia, and the United Kingdom.

Figure 3.10. Infrastructure sub-element scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

5. Markets

Start-ups and scale-ups often leverage domestic and international demand to take off. Access to a large (domestic and international) customer base, increases the likelihood that these ventures can expand their business. The empirical literature shows that firm entry rates tend to increase in expanding economies. Two indicators are used to measure the Markets element. The total GDP (expressed in Purchasing power parity terms) captures the size of the domestic market both in terms of consumer demand and business-to-business demand, while the trade facilitation index measures the access to foreign markets. To avoid that extremely large differences in total GDP drive the results, GDP values are first transformed to logarithm and then normalised. This allows to reduce the variance across countries while making sure that the larger economies are rewarded for offering significantly more opportunities to businesses and entrepreneurs.

Taking these two factors together, the United States, Japan, Korea are the largest markets among the OECD countries, all attaining a score between 80 and 87/100 (Figure 3.11). On the opposite front, Hungary, Iceland, and Israel have yet to develop their market potential.

Over time, the Markets element scores have improved over time in all countries, driven by GDP growth, better trade facilitation or both. Among them, Czechia, Mexico, Türkiye, and Greece are the countries who have improved the most since the 2016-2020 period, while the United States, Netherlands, and Iceland have improved very little in this time frame.

Figure 3.11. Markets element scores by country and year

Note: These scores are computed as the geometric mean for countries across the following indicators: i.) GDP, PPP, Source: OECD, ii.) Trade facilitation index, Source: OECD - Trade Facilitation Indicators. GDP data are transformed to logarithms before normalisation. Before aggregation, data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023 data and must be interpreted as relative performance the 2020-2023 period.

Among the indicators driving these summary scores (Figure 3.12), GDP rewards large countries such as the United States, Japan, and Germany, while small countries such as Iceland, Estonia and Latvia need to overcome their small domestic market size to offer comparable opportunities to their start-ups. GDP growth has been slow but positive in all countries after the pandemic which explains why all countries scores have improved, albeit to different extents. Norway, Ireland, and, to a lesser extent Türkiye are the countries that have managed to improve their GDP scores relatively more than others in this period, while the scores of United States, Japan and Iceland have barely moved, due to their already high/low position in the ranking.

In terms of international markets, Korea and the Netherlands stand out as countries with the most favourable to border procedures. They attain scores above 80/100, and their followed by Norway, Japan, Sweden, the United States, and Australia who achieve scores between 70 and 77/100. On the opposite front Israel, Hungary and Chile are the countries that have in place more challenging border controls, which constrain markets given the small size of the domestic economies.

Most OECD countries have been improving border access over the past few years. Notably Switzerland, Czechia and Portugal have improved border access scores significantly more than other countries, while Hungary, the Netherlands and Israel have improved little, virtually remaining at the same level as four years ago.

Figure 3.12. Market sub-element scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

6. Finance

Access to finance is frequently mentioned as a critical challenge for entrepreneurs. Newly-created firms often either lack collateral or have limited credibility to access bank credit or lack a sufficient track record to access equity finance. This matters especially at the pre-seed stage. However, insufficient access to finance can also hit at a later stage, when moving from a small scale to a larger scale can be limited by insufficient funds for investments. Four indicators are used to measure this element, seed-stage venture capital investments per capita, later-stage venture capital investments per capita, SME loans stock per capita, and factoring contract value per capita.

The United States and Israel are the best countries in granting access to capital to SMEs, start-up and scale-up, attaining scores between 60 and 65/100. They are followed by Belgium, Finland, and Netherlands, but the composition of their elements scores are different. The United States and Israel excel on venture capital investments but perform significantly less well on more traditional bank credit and factoring services. On the contrary, Belgium, Finland and the Netherland perform significantly less well on venture capital investments but offers SMEs financial access through factoring and loans.

Performances have remained stable over time for most countries, with only Denmark and the United States improving their performance since 2016-2020 period.

Figure 3.13. Finance element scores by country and year

Note: These scores are computed as the geometric mean for countries across the following indicators: i. Seed VC investment per thousand pop. Source: OECD Entrepreneurship Financing Database; ii. Growth start-up VC investments per thousand pop., Source: OECD Entrepreneurship Financing Database; Factoring, thousands USD per capita; iii. Outstanding loans to SMEs, thousands USD per capita, Source: OECD Financing SMEs and Entrepreneurs: An OECD Scoreboard, iv. Factoring value, USD per capita, Source: OECD Financing SMEs and Entrepreneurs: An OECD Scoreboard. Before aggregation, data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023 data and must be interpreted as relative performance the 2020-2023 period.

In terms of early-stage venture capital, Israel and the United States are the countries where investments per capita are at the highest by far, attaining a score over 98/100. The next two countries, Luxembourg and Canada, attain scores between 60 and 65, while Estonia, the last country in the top 5, achieves a score of 52/100. Notably these five countries are also those where venture capital (VC) investments per capita scores increased the most over the past four years. In most other countries, scores increased but by a significantly smaller margin.

Investments in later-stage venture capital follow a similar pattern, but with even larger gaps between the top two countries and other economies. On this particular segment both Israel and the United States achieve a score of 100/100, while Canada, Sweden and Denmark, the other three countries in the top 5 attain scores between 49 and 56/100. Differently to early-stage investments, however, the countries that have improved their scores the most on later-stage venture capital investments include both top countries (the United States, Canada and Sweden), but also countries that have attained lower scores such as Austria and Estonia. More generally, most countries have improved on this dimension, except Norway and Slovenia.

In addition to venture capital, the element also measures financial aspects that are less relevant to start-ups who growth through venture capital runaways, but are important to start-ups who need access to more traditional types of finance such as loans and factoring to expand.

In terms of SME loans (as a share of the population), Switzerland is the country who provides a higher volume of loans to SMEs, proportional to its population, resulting in the top score (100/100). It is followed by Japan with a score of 89/100, as well as Korea, Sweden, and Belgium with scores between 60 and 62/100. Notably, in less developed countries such as Chile, Mexico, Colombia and Türkiye access to bank credit remains significantly more difficult than in other OECD countries. These conditions have not changed significantly over time, however, about half of OECD countries have seen an increase of their scores or remained stable on this indicator, while the other half has seen a reduction in scores. These changes are small and barely visible in absolute term, but they can be important for local entrepreneurs. Among the countries who have changed their SME loans scores the most, Korea, Japan and France scores improved while Denmark, Sweden, and Greece scores have contracted.

Looking at factoring (as a share of the population), Belgium, Ireland, and the Netherlands who use this instrument the most, registering a score of 100/100. They are followed by Finland, Spain, and Italy with scores between 80 and 85/100. On the opposite front, Türkiye, Hungary, and Chile are the countries where factoring is less used. Notably, in Israel and the United States, two of the top countries on the finance element, factoring scores are below the OECD average.

Over the time frame considered, factoring scores have increased or remained stable in about two-thirds of OECD countries, and declined in one third of the countries. Countries that have registered the most important scores increase include Spain, Denmark, Germany, and Austria; while in Switzerland and Norway are the countries where factoring scores have decreased the most in the period considered.

Figure 3.14. Finance sub-element scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

7. Knowledge

The availability of a substantial knowledge base is a necessary condition for entrepreneurs to turn ideas and research outcomes into new and innovative commercial products and services. To measure the knowledge base, this element includes R&D expenditure (as a share of GDP), patents (per capita), and GitHub software uploads (per thousand people).

With scores between 80 and 87, Korea, Israel and Switzerland, are the OECD countries with the largest accumulation of knowledge and intangible assets, relative to the size of the countries (Figure 3.15). Sweden, Finland, and Denmark follow suit with scores between 69 and 80/100, just above the United States who attain a score of 65/100.

Among the top countries, Korea (since 2016-2020 period) increased its performance remarkably, while Israel and Switzerland improved as well but much less. Beyond the top performers, Iceland, Estonia, and Portugal have also managed to marginally improve their relative scores over time, while the scores of most other countries have remained substantially stable.

Figure 3.15. Knowledge element scores by country and year

Note: These scores are computed as the geometric mean for countries across the following indicators: i.) Patents, per million population, Source: OECD - Main Science and Technology Indicators, ii.) R&D expenditure, % GDP, Source: OECD - Main Science and Technology Indicators; iii.) GitHub software uploads, per thousand people, Source: GitHub. Before aggregation, data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023 data and must be interpreted as relative performance the 2020-2023 period.

Among the sub-element indicators that compose the element’s score two have remained virtually stable over time (Figure 3.16), pointing to structural differences across countries in their capacity to produce and accumulate knowledge over time. One indicator, measuring software production has instead increased over time for most OECD countries.

In terms of patent production, Switzerland and Japan, are in per capita terms the best positioned among OECD countries, attaining scores 15-25 points above those of Israel, Sweden and Korea. In per capita terms, these countries are also 40 points higher than the United States and Germany, which are regarded as innovation hubs, with a high share of total patents worldwide. In terms of R&D expenditure, another two relatively small economies, Israel and Korea, attain the highest scores among the OECD member countries. They are followed by the United States, with a score about 20 points below those of the top two countries, and Sweden, Belgium, Japan and Switzerland, with scores approximately 25 points lower. In terms of software production, Norway and Switzerland attain the two highest scores (both above 90/100), and they are followed closely by other Nordic countries (Iceland, Estonia, Denmark, Finland and Sweden), as well as Canada and Israel, all attaining scores between 75 and 90/100. These results show how some of the most entrepreneurial and innovative countries benefit from having a strong knowledge backbone.

Figure 3.16. Knowledge sub-element scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

8. Talent

Talent is another critical element in entrepreneurial ecosystems, both in terms of available entrepreneurial skills as well as broader technical skills available in the domestic labour market that start-up and scale-up businesses can tap into for their development. The element is measured through four indicators, a measure of perceived entrepreneurial capabilities (the percentage of adults who believe they have the skills necessary to start a business), two measures of average person education level and quality (mean years of schooling, and OECD PISA scores) and one proxy measure of digital skills (internet users as a share of total population).

Korea, the United States, Denmark and Canada are the best positioned countries on the Talent element, with scores above 69/100. Other strong performers are Estonia, Switzerland, Iceland, Australia, and the United Kingdom, which all achieve scores between 60 and 65/100.

Figure 3.17. Talent element scores by country and year

Note: These scores are computed as the geometric mean for countries across the following indicators: i.) Mean years of schooling, Source: UNESCO, ii.) Average of Math, Reading, and Science Pisa scores, Source: OECD; iii.) Percentage of 18-64 population who believe they have the required skills and knowledge to start a business, Source: Global Entrepreneurship Monitor (GEM); iv) Internet users, % of the population, Source: World Bank, World Development Indicators. Before aggregation, data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

Among the top-performing countries, Korea and the United Staes have increased their relative scores over the past four years, while other countries, have remained stable or slightly declined, leading Korea to attain the highest score in the latest iteration of the tool. Other countries who have increased their scores substantially over the past four years are Lithuania, Poland, Portugal and Slovenia. On the opposite front, Iceland is the country who has moved backwards the most, while Japan and the Netherlands’ score declined only marginally.

Looking at the indicators composing the element, Mean years of schooling scores features\ Germany as the country attaining the top score (83/100). It is followed by Canada, Switzerland, Iceland, Korea, the United States, Estonia, and Lithuania all achieving scores between 70 and 77/100. On the opposite front, Mexico, Colombia, Costa Rica, and Türkiye, still attain much lower education participation than most other OECD countries, explaining their lower scores. Countries have maintained stable performances on this indicator over the past four years, yet Spain, Portugal, and Australia managed to improve relatively more than other countries.

While means years of school measures the quantity of education, Pisa scores aim at measuring quality of educational attainment. On this aspect, the best positioned country is Japan, attaining a score above 90/100, followed by Korea and Estonia who attain scores between 80 and 86/100. On the opposite side, Costa Rica, Colombia, and Mexico have yet to close important educational gaps relative to OECD performances on this aspect. Notably, most OECD countries have reduced their scores over time, which indicates a worsening of student skills learnt in school. Countries that have seen a larger reduction of scores relative to other OECD members are Germany, Slovenia, Finland, the Netherlands, Norway, and Iceland. While most countries scores decline three countries (Türkiye, Japan, and Korea) increased their scores, albeit starting from different bases.

Moving to digital skills, as measures in terms of internet users (as a share of the total population), Iceland, Luxembourg, and Denmark achieve the highest scores ranging between 80 and 87/100. Norway, Korea, the United States, Switzerland, and New Zealand also perform well with score above 70/100. With the share of people using the internet close set above 90% in several OECD countries, having relatively lower shares of the population using the internet results in low scores. This is the case of Italy, Colombia, and Mexico, where internet users are below 80% of the population, explaining their performance. Another factor penalising slow adopters of digitalisation is the rapid increase in the use of digital technologies in virtually all OECD countries. Among the, Hungary, Slovenia, and Poland are some of the countries that increased their score the most since the 2016-2020 period edition, while Japan, in relative terms, moved backwards.

The last aspect composing the Talent element is the perceived entrepreneurial capabilities among the adult population. Chile and Mexico, with scores above 90/100 are the best performers on this dimension. Another two countries that have registered a strong performance, with scores between 74 and 85/100, are the United States, and Slovenia. Japan instead is the country where entrepreneurial capabilities perceptions are the lowest. These data are subject to some degree of variability, which, in most cases pointed towards an improvement over time. Notably, Norway, Italy and France scores have improved throughout the periods considered. Among the few countries where perception decreased over the period considered, Colombia, Slovak Republic and Israel registered the largest score reductions.

Figure 3.18. Talent sub-element scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

9. Leadership

Founders of start-ups and scale-ups can benefit significantly from exchanges with former entrepreneurs who can guide them on how to improve their business ideas, and sometimes connect them to other relevant stakeholders, including investors. These experts can often take on the role of mentors, and their leadership is highly regarded by entrepreneurs. They can also be a source of inspiration for aspiring entrepreneurs and the presence of these leaders can sometimes make the difference between successful and unsuccessful ventures. The elements are measured in terms of number of serial entrepreneurs active in a country. There is only one indicator in this element. This indicator is expressed in total count rather than as a ratio to the total population. This is because the number of leaders do not need to increase linearly with the total population for a country to have a sufficient leadership base. Also, it can be argued that countries where a critical mass of leaders is present can be better positioned to sort inspirational and vision-setting on other aspiring entrepreneurs. However, to avoid over-rewarding larger countries, a logarithm of the moving average value is applied before computing the normalised score.

The United States and the United Kingdom are the two countries with the highest numbers of these leaders available to entrepreneurs. They are followed by Canada and Germany. These four countries all attain scores between 84 and 100.

These scores have changed little over time among the top countries. However, among all other countries there have been a general increase in the availability of serial entrepreneurs. Among the countries who increased their scores the most Chile and Türkiye stand out, while also Costa Rica and Czechia have registered significant improvements. Nonetheless, the gaps between countries remain large, indicating that developing new leaders, and making them available to entrepreneurs and start-up founders takes time, and may require significant efforts.

Figure 3.19. Leadership element scores by country and year

Note: These scores correspond to the normalised values of the number of serial entrepreneurs total count. Source: Crunchbase and OECD. Values are transformed to logarithms before normalisation. Data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

10. Intermediate services

The availability of intermediate business services can play a role in reducing barriers to entry for entrepreneurs, helping them to channel their efforts and supporting them with improved marketing, pitching, network building, and connections with investors and customers. This element is measured with three indicators, the number of active incubators and accelerators (per million population), the number of available coaches or mentors, and the share of technical employees to total workers, a proxy measure for availability of experts in technical domains. Among these measures the number of coaches and mentors is expressed in absolute terms. Similarly to the logic applied to leaders, a countries’ number of coaches and mentors does not need to increase linearly with the population, and the availability of a critical mass of these experts plays an important role in offering help to aspiring and emerging entrepreneurs. However, to reduce cross-country variance, a logarithmic transformation is applied before computing the normalised score of this indicator.

The United Kingdom, the United States with scores between 79 and 83 are the countries with the most intermediate services available to entrepreneurs. Israel, Switzerland and Canada follow suit, with scores above 70/100.

These scores have remained essentially the same throughout the periods considered period, with no countries registering even marginal declines, while Lithuania, Netherlands, and Hungary managed to improve their scores relatively more than other countries.

Figure 3.20. Intermediate services element scores by country and year

Note: These scores are computed as the geometric mean for countries across the following indicators: i.) Number of incubators, accelerators and start-up support programmes per capita, Source: Crunchbase and OECD, ii.) Technical employment, % total employment, Source: OECD; iii. Coaches and mentors total count. Source: Crunchbase and OECD. Coaches and mentors values are transformed to logarithms before normalisation. Before aggregation, data are normalised using a min-max transformations where the max/min are equal to the sample mean +/- 2*sample standard deviations, relative to the average of data from the 2020-2023 period. 2016-2020 and 2018-2022 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

One of the critical services to start-ups are those provided by incubators and accelerators. On this aspect, Luxembourg and Iceland are the countries with the highest number of incubators relative to the population, resulting in top scores. They are followed by Switzerland, Israel, and Estonia with scores above 80/100. Most OECD countries have gradually improved their supply of incubations and acceleration services over time, resulting in slightly higher scores for all but two OECD countries, while Lithuania, Switzerland, and Luxembourg are those who progressed the most in the period considered.

Complementary to incubation services, the availability of coaches and mentors plays an important role in supporting start-ups development. On this aspect, the highest availability of coaches and mentors is found in United States, United Kingdom, and to a lesser extent Canada. All these countries attain scores above 90/100. In Lithuania and Costa Rica, instead, finding good mentors and coaches is significantly more difficult. However, the availability of this professionals is on the rise in all OECD countries, with particularly notable increases in Colombia, Italy, and Norway.

A further type of intermediate services to start-ups and entrepreneurs are those provided by technical experts (e.g. lawyers, IT technicians, accountants, etc.). Sweden and the United Kingdom are the countries with a relatively higher incidence of these experts, making it easier to access them. As a result, these countries attain the highest scores on this aspect, with values above 90/100. Other countries who perform well are Netherlands, Switzerland, and Israel with scores above 80/100.

In terms of technical employment, Belgium and Japan achieve the same very high score as Luxembourg, followed by Switzerland, the United Kingdom, and the Netherlands which register scores of between 75 and 78/100. Although technical employment incidence is a structural feature of countries’ economies, there has been some variability over the past years. Notably Slovenia, Netherlands, Hungary, and the United Kingdom have increased their relative score on this indicator, while Germany, Switzerland, and Iceland registered lower scores in the 2020-2023 period compared to previous instances.

Figure 3.21. Intermediate services sub-element scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

Performance on entrepreneurship outputs

In this section, countries are benchmarked on entrepreneurial ecosystem outputs in terms of enterprise creation, growth and survival. The presented indicators report statistics as normalised scores of moving averaged data. Normalised scores are useful for benchmarking the performance of countries vis-a-vis others, and since these scores are anchored to the distribution of values in the latest iteration of the tool covering the period 2020-2023, they also enable a comparison of progress over time. The key entrepreneurship data underlying these statistics are reported in Tables B.1 – B.3 of the Annex data for the most recent datapoint available for each indicator and country.

Early-stage entrepreneurship

This section provides the diagnostic of four early-stage entrepreneurship output measures: Birth rate of employer enterprises as a share of active businesses; Medium and high growth enterprises, which is the incidence of firms that grow their employment base rapidly, as a share of employer enterprises; Equity-based young firms per million population, measured as the number of companies newly registered in Crunchbase, with an emphasis on more innovative and venture capital ready enterprises; and the number of Unicorns per million population, i.e. new companies with over one million dollars market capitalisation.

The first two indicators (birth rate of employer enterprises and incidence of high growth firms) represent broader measures of entrepreneurship, whereas the latter two (emerging start-ups, unicorns) tend to pick up more innovative, venture-backed start-up enterprises. The performance of the countries varies by these two categories of indicators.

Figure 3.22 shows the normalised country scores on these indicators from 0-100, country rankings (from left to right of the charts) and changes over time.

Countries such as Colombia, Costa Rica and Türkiye have high performance on the more general measures (Panels A and B). These countries are doing well at getting large numbers of business starts and, in some cases, also to expand them in terms of headcounts. However, when more specific and sophisticated types of start-ups are investigated, in terms of the more innovation, equity seeking and growth-oriented companies captured in the Crunchbase database (Panel C), the countries attaining the highest scores (in per capita terms) are Netherlands, Estonia, and Switzerland. They are closely followed by the United Kingdom, Australia, Luxembourg, the United States and Israel. In the most extreme case, when the focus is only on a very specific subset of the highest growth entrepreneurial ventures of unicorns per capita population (Panel D), the United States, Israel and Ireland, and to a lesser extent Estonia, Norway and Finland stand out.

There are some significant changes over time in particular countries in many of these rates. Notably, in three out of four of these entrepreneurship outcomes, most OECD countries scores register a declining trend when compared to the 2020-2023 period, indicating a possible decline in entrepreneurial dynamism in this period. One possible explanation is the 2020-2023 data contain post-pandemic statistics, which registered a decline of business creation and growth, while the 2016-2020 and the 2018-2022 periods contain a mix of pre- and post-pandemic statistics. The registered decline in available statistics can thus be temporary, but it will be important to continue monitoring the evolution of these indicators to evaluate potential permanent effects in these trends.

The evolution of Unicorns scores (Figure 3.22, Panel D), however, in contrast with the indicators in panel A, B, and C presents an increase in several OECD countries. The upward trend appears more pronounced among top countries and subdued in countries where few unicorns are created. This dynamic highlights increasing concentration of unicorns creation in specific countries who act as hubs for the development of these types of venture. As a results gaps between top and less advanced ecosystems are widening with respect to large-capitalised start-ups.

Figure 3.22. Early stage entrepreneurship outputs – scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

Survival and dynamism

Figure 3.23 gives a second set of entrepreneurship output metrics more focused on enterprise survival and evolution. These measures are: the churn rate of employer enterprises (Panel A), the entrepreneurs expectations to create jobs (Panel B), the survival rate after 3 years from incorporation of employer enterprises (Panel C), and the share of employers enterprises that are at least 2-years old (Panel D).

All these metrics focus on more general types of entrepreneurial ventures, not only the most innovative ones, and aim at capturing the capacity of these firms to remain on the market, expand and employ workers.

Figure 3.23. Survival and dynamism outputs – scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

Churn rates (Panel A), combining entries and exits, show one specific aspect of an ecosystem vitality. Here, the most business dynamic country is Finland, followed by Hungary and Estonia, and one step forward down, also by France and Sweden. Over time churn rates have tended to decline in countries where time series were available. While, this can still partially reflect the effect of the pandemic, it is another indication of a reduced business dynamism in the post-pandemic period.

In terms of expectations to create jobs (Panel B) the top countries include Colombia, Türkiye, Chile, Luxembourg, and Ireland. They are followed by Mexico, the United States, Korea, and Latvia. These results suggest that employment growth can be driven by different types of entrepreneurial ventures, but in all cases job creation is an important sign of entrepreneurial dynamism. On this aspect, variations over time are somewhat more volatile and do not affect all countries in the same way. Among the countries where expectation have improved the most in the period considered, Mexico, Korea, Luxembourg stand out, while Germany, Switzerland, and Hungary are the countries where scores have declined the most.

In terms of start-up survival rates (Panel C), the highest share by far is found in Sweden, closely followed by Ireland and Belgium. All these countries’ scores are between 76 and 89/100. Latvia, Portugal, and Germany are next with scores between 64 and 67/100. Most OECD countries’ performances have remained stable over time, but there are some notable exceptions. Hungary, Lithuania and Denmark have registered the most noticeable declines, while Greece, Belgium, and Latvia are those who have increased their scores the most in the period considered.

2-years old employer firms (Panel D) are also partially related to survival but focus the attention of the incidence of young firm on the total population of firms. On this element Finland reaches the highest score among OECD countries. It is followed by Türkiye and Colombia. The shares are significantly lower after the top 3, indicating a particularly high incidence of young firms on these economies, which however, does not distinguish between highly innovative and more traditional young firms. These scores have remained mostly stable in OECD countries, with Finland being the only country registering an outstanding upward swing, while Denmark, Hungary, Netherlands, Iceland, Belgium, and Greece are the countries where scores reduced the most in the period considered.

Ecosystem variation

In addition to the national average rates of entrepreneurship presented above, it is important to keep track of the variation within national entrepreneurial ecosystems in terms of geographical concentration and social inclusion in entrepreneurship activity. This section provides this information. The data are useful as an indication of the extent to which the national average entrepreneurship ecosystem conditions reported above are likely to be representative of conditions in the ecosystem sub-systems affecting specific regions and social groups in a country. Strong regional and social heterogeneity is suggestive of a greater need for disaggregated analysis and more differentiated entrepreneurship support policies.

The social variation of the ecosystems is measured in terms of two indicators: the share of women entrepreneurs, founders and CEOs, among the total number of these categories, and the share of “missing entrepreneurs” in each country following the methodology of the OECD/European Union Missing Entrepreneurs work. The variation in terms of regional distribution is measured in terms of the Herfindal index of start-ups by cities as reported by Crunchbase. As per the entrepreneurship output data, the indicators are presented as normalised scores of moving averaged data.

Social variation indicators are picked up in the upper part of Figure 3.24. Panel A reports the ‘Missing entrepreneurs’ rate. Panel B reports the share of women founders.

The highest level of women founders is in the United States, where almost 27% of start-up founders are women. Other countries with relatively high shares of women founders include France and the United Kingdom, where women represent about 25% of start-up founders. The median level among the few OECD countries for which recent data are available is 18.8%, which shows that women’s participation in productive entrepreneurship is still well below that of men. This is also partially reflected in the statistics on the missing entrepreneurs, which capture the extent to which women, youth, seniors, and immigrants are less represented in entrepreneurship activities relative to men of 30-49 years old (the group with the highest entrepreneurship participation rate). To a significant extent differences in women’s entrepreneurship rates are also reflected in the missing entrepreneurs rates in countries, given that women generally make up a high share of the missing entrepreneurs. This is the case of Italy, where gender imbalances drive a high rate of missing entrepreneurs. In contrast, countries with relatively homogenous entrepreneurship rates across different social groups are Colombia, Greece, Ireland, Mexico, and Sweden.

There have also been some strong changes in country performance on these variables over time. Spain is the country where women entrepreneurship scores have increased the most, while France and Israel have recoded small but important declines. In terms of the broader set of social groups composing the missing entrepreneurs, Finland and Switzerland have improved their scores the most, while Poland has moved backwards.

Figure 3.24. Entrepreneurship regional and social variation – scores by country and year

Note: These scores are computed in two steps. In the first step a moving average of the data within each period is calculated. 2020-2023 values are moving averages of 2020-2023 data, 2018-2022 values are moving average of 2018-2022 data, and 2016-2020 values are moving averages of 2016-2020 data. All averages are computed using the datapoints available for a specific country and indicator in, each period. In the second step scores are computed using a min-max transformations where the max/min are equal to the 2020-2023 period sample mean +/- 2*sample standard deviations. 2018-2022 and 2016-2020 scores are anchored to the 2020-2023 data and must be interpreted as relative performance to the 2020-2023 period.

The regional distribution of entrepreneurship is picked up in the bottom part of Figure 3.24 (Panel C), which reports statistics on concentrations of start-ups across cities within countries.1

The United States has relatively homogenous levels of entrepreneurship activity across multiple locations. The Netherlands, Belgium, Italy and Switzerland also have fairly well distributed start-up activities across regions. In contrast, in Iceland, Estonia, and Lithuania particularly, start-ups tend to cluster around the main country hub, often corresponding to the capital city.

The rates of regional dispersion or homogeneity have been fairly persistent over time. In the period considered, only Hungary and Japan became significantly more even in terms of regional entrepreneurship rates, while regional entrepreneurship activity in the United Kingdom and Slovenia became less homogenous.

Note

← 1.  The indicator “Regional dispersion of employer enterprises birth, standard deviations” has been collected and included in the framework, however, scores are not computed for any country as the data have not been updated after 2016.