Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A Global Entrepreneurship Efficiency Benchmarking and Comparison Study based on National Systems of Entrepreneurship and Early-Stage Business: A Data Envelopment Analysis Approach

A Global Entrepreneurship Efficiency Benchmarking and Comparison Study based on National Systems... National Systems of Entrepreneurship is defined as a nation’s resource allocation structure leading to entrepreneurial behaviors. However, the existing indicators of national framework conditions may have limitations in comparing the entrepreneurial efficiency of countries. Based on institutional theory, this paper presents a model to examine the efficiency of entrepreneurial activities stemming from the given conditions of a country and find benchmarks based on data envelopment analysis by scrutinizing inputs and outputs with static efficiency, dynamic efficiency, and strategic quadrant analysis. For this purpose, the study utilizes the Global Entrepreneurship Monitor dataset from 2015 to 2020 for 24 countries and presents the research questions regarding the differences in global entrepreneurial efficiencies, the countries for benchmarking, and the implications for entrepreneurial activities. The research implications suggest that diversifying the views on entrepreneurial efficiency may be valuable, and policymakers may focus on institutional conditions and entrepreneurial efficiency regarding the activity of early-stage businesses. Keywords data envelopment analysis, entrepreneurial efficiency, global entrepreneurship monitor, national systems of entrepreneurship, total early-stage entrepreneurial activity (Ács et al., 2014). Entrepreneurship is a complex phenome- Introduction non, and many determinants affect its occurrence. A coun- Entrepreneurs play a substantial role in a country’s economic try’s resource allocation systems are affected by the development. They are the agents of creative destruction individual’s opportunity pursuit and new business creations, (Schumpeter, 1934), the epicenter of innovation (Ács, and these activities and outcomes are influenced by the coun- Audretsch, 1988), and actors of job creation and knowledge try’s particular institutional specificities (Ács et al., 2014). spillover (e.g., Blanchflower, 2000; Parker, 2009; van Praag Then, it could be assumed that individual entrepreneurial & Versloot, 2007). Despite its positive contributions to the activities and the birth of early-stage businesses could be economy and academic recognition, still defining entrepre- viewed as a result of the national systems at a country level neurship is challenging. Various scholars have strived to (Ács et al., 2014; Inacio Junior et al., 2021; Tasnim & Afzal, make a consensus presenting different views on defining entrepreneurship (Anderson & Starnawska, 2008; Gedeon, Seoul School of Integrated Sciences and Technologies, Seodaemun-gu, 2010; Sendra-Pons et al., 2022) over time. It is an economic Seoul, Republic of Korea novelty-introducing (Schumpeter, 1934), firm-level behav- The Institute for Industrial Policy Studies (IPS), Seodaemun-gu, Seoul, ioral disposition or individual cognitive attribute for business Republic of Korea opportunity (e.g., Lumpkin & Dess, 1996; Shane & 3 Business School Lausanne, Lausanne, Chavannes, Switzerland Venkataraman, 2000), or activities such as self-employment Corresponding Author: or new business creation (i.e., Reynolds et al., 2005). Sehoon Kim, Business School Lausanne, Route de la Maladière 21, Various academic debates about entrepreneurial theories Lausanne, Chavannes 1022, Switzerland. and their academic background need further elaboration Email: sehoon.kim@bsl-lausanne.ch Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (https://creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 SAGE Open 2018). Also, the concept of National Systems of businesses), enabling efficiency comparison at the global Entrepreneurship (NSE) is central to assessing a country’s level and help to find comparable benchmarks. Benchmarking contextual conditions upon the entrepreneurial perfor- refers to exploring and utilizing other organizations’ best mance (Feldman, 2001; Sternberg, 2009), which influ- practices, services, and products to improve organizational enced the policymakers to form a framework to understand performance (Main, 1992). However, there is a risk of taking entrepreneurial activities in the overall context. NSE, a wrong benchmark that ignores NSE differences if policy- affected by the National Systems of Innovation (NSI) con- makers only focus on entrepreneurial performance, such as cept, is defined as a nation’s resource allocation system the high birth rate of start-up businesses. In other words, it and structure where individual entrepreneurial behaviors would be a reasonable approach to choose a benchmark and initiatives occur (Ács et al., 2014; Inacio Junior et al., country that produces optimum performance in a similar 2021; Tasnim & Afzal, 2018). NSE focuses on systemic social and economic environment (Lins et al., 2003). factors driving entrepreneurial activities (Ács et al., 2014), Efficiency theory can be applied when selecting a benchmark and it views the creation of businesses as the product of a with excellent output in a similar input structure. process affected by combined factors that arise from the Efficiency is defined as the relationship between input and system (Inacio Junior et al., 2021). However, referring output or evaluating the effectiveness of output factors for only to the rankings of global surveys that measure national input (Ramanathan, 2003; Thornton et al., 1982). Efficiency institutional conditions where developed economies hold is a significant concept in economics. Low efficiency could the entrepreneurial leadership position (the highest national mean poorly performing system components that hinder potential for entrepreneurship) may not help compare the overall performance (Tasnim & Afzal, 2018). The application entrepreneurial activities of countries since those rankings of the efficiency theory could help find high-performance are not designed to evaluate the simple efficiency such as units and units with similar input conditions as benchmarks. the ratio between input and output (Inacio Junior et al., Parametric and non-parametric methods can be considered 2021). Top rankers in those global surveys may not always for efficiency evaluation (Smith & Street, 2005; Stone, 2002). be efficient in generating new businesses, and they may also Data envelopment analysis (DEA) is a non-parametric statis- strengthen the belief that a national system with the highest tical technique that measures efficiency by comparing rela- ranking should be chosen as a benchmark (Bogetoft, 2012). tive inputs and outputs (Casu & Molyneux, 2003) and As such, quantifying and measuring conditions may not estimates the optimal level of output based on the mixed input be sufficient in understanding the actual performance of a elements (Smith & Street, 2005). In addition, the DEA model national system (Inzelt, 2004; Kuhlmann, 2003), and indices enables relative evaluation between DMUs (decision-making not considering the input-output efficiency may not fulfill units) by comparing the empirical productivity frontiers in a the need to gauge the country’s entrepreneurial performance similar environment and provides information on benchmarks (Edquist & Zabala-Iturriagagoitia, 2015). Due to its substan- with optimal performance. Efficiency or productivity focuses tial role in decision-making policies and nurturing socioeco- on selecting a target that achieves superior performance (e.g., nomic development, academic interests and attention have services, outcomes, or activities) under a similar environment examined the causal relationships between innovative entre- (e.g., socioeconomic factors, budget) (Shephard, 1970; Stone, preneurial activities and a country’s international competi- 2002). The environment involves national contexts, levels of tiveness (Ferreira et al., 2017). While valuable, the existing NSE, and stages of economic growth, such as GDP (Gross global entrepreneurship surveys focus on how supportive the Domestic Product) per capita (i.e., Solow, 1957), where a entrepreneurial environment is and may have limitations in country can foster entrepreneurship. considering national characteristics and specificities to eval- To date, global entrepreneurship research investigating uate entrepreneurial activities’ efficiency as the output of entrepreneurial efficiency frontiers based on the institutional national systems (Edquist & Zabala-Iturriagagoitia, 2015). factors and early-stage entrepreneurs for the multi-faceted The entrepreneurship literature has neglected to highlight efficiency and longitudinal analyses as in this paper are efficiency by focusing on advanced economies with the best somewhat limited. Some previous studies suggesting NSE framework conditions, such as the US (Inacio Junior et al., constructs (regional framework conditions) such as output 2021). Hence, examining how a nation’s entrepreneurial measures, attitude measures, and framework measures but activities could be evaluated and compared apple-to-apple without efficiency comparison (i.e., Ács et al., 2014) failed based on the “output” of the given systems would add value to examine longitudinal dynamics (e.g., Das & Kundu, 2019; to the literature. Inacio Junior et al., 2021; Tasnim & Afzal, 2018) or set mis- Consequently, this paper seeks to analyze entrepreneur- leading causal relationships between inputs and outputs for ship efficiency that is measured and calculated by consider- efficiency analysis (i.e., Faghih et al., 2021). Overall, it is ing inputs (the institutional factors that make up the national understood that global entrepreneurship literature further framework condition, such as evaluation results of mecha- requires; (a) an efficiency study based on NSE, (b) empirical nisms and processes) and the output (the entrepreneurial evidence for benchmarking, and (c) integrated efficiency activities of a country represented by the creation of new measures that would advance future studies. Kim 3 In order to reduce the research gap, this paper aims to NSE concept evolved into the “entrepreneurial ecosystem” contribute to the knowledge of country-level efficiency concept (i.e., Szerb et al., 2019), which views the regional based on the NSE concept and provide a valuable integrated ecosystem may affect different types of entrepreneurial efficiency model for benchmarking based on empirical evi- activities (Kirznerian or Schumpeterian). A healthier ecosys- dence gathered throughout the research. This paper mainly tem grants greater capacity for higher business formation, deals with the global entrepreneurship phenomenon from a while a weak ecosystem leads to innovation of entrepreneurs “systemic efficiency” viewpoint, not “the superiority of due to a lack of support policies or economic support (Szerb entrepreneurial environment.” To this end, this paper focuses et al., 2019). NSE also represents the dynamism and interac- on cross-sectional (static) and longitudinal (dynamic) effi- tion among individual attitudes, aspirations, and abilities that ciency analyses. Also, the paper aims to assess global entre- stimulate resource allocation via new business creation and preneurship competitiveness based on strategic quadrant operation (Ács et al., 2014). In a country, individuals pursue analysis to compare each country’s position with peers. For entrepreneurial opportunities and evaluate the desirability this purpose, this study utilizes the Global Entrepreneurship and feasibility of the pursuit, which is also determined by a Monitor (GEM) dataset for 6 years, from 2015 to 2020, for particular country’s institutional (contextual) factors, such 24 countries. Consequently, this paper presents the fol- as distribution of available resources, social norms, and lowing research questions; (a) What are the differences in attitudes (Ács et al., 2014). Some NSE output indicators entrepreneurial efficiencies of countries? (b) Who are the evaluate the degree of emergence for the newly born entre- benchmarking countries and regions from a static and preneurs within a given population, and the most widely dynamic efficiency perspective? (c) What are the strategic used example is GEM’s total early-stage entrepreneurial implications of the efficiency-stability analysis result? activity rate, TEA (Autio, 2007). Global entrepreneurship This study is structured as follows. Section 2 deals with indicators such as GEDI (Global Entrepreneurship and the literature review. Section 3 provides the research Development Index) and GEI (Global Entrepreneurship design and methodology. Section 4 deals with the specific Index) deliver multi-faceted information on the NSE per- result of the study. Section 5 discusses and summarizes formance at a system level by profiling each NSE. the conclusion, and Section 6 provides theoretical and However, these indicator-based approaches might be less policymaking implications. Lastly, Section 7 presents valuable in contemplating the relationships between input suggestions for future research. factors and the productivity or efficiency of analyzed units (Inacio Junior et al., 2021). Also, considering the countries with the highest values in NSE indicators as benchmarks Literature Review could lead to severe bias or failure in decision-making pro- cesses (Bogetoft, 2012). National Systems of Entrepreneurship Understanding the system performance requires a holistic National Systems of Innovation (NSI) literature comprised approach (Inzelt, 2004; Kuhlmann, 2003), and quantifying the holistic relationships toward interactive, iterative, and the national systems may not be sufficient if not followed by cumulative innovation processes at a country level (e.g., the productivity analysis considering inputs and outputs Freeman, 1987; Lundvall et al., 2002). NSI’s institutional (Edquist & Zabala-Iturriagagoitia, 2015; Inacio Junior et al., and structural focus inspired policymakers with a frame- 2021). It is reasonable to identify the national system’s per- work-based performance evaluation (Nelson, 1993) and formance based on the efficiency concept, where a unit’s made stakeholders consider the influence of regulatory inter- productivity is based on the rate of a total of outputs by a vention in national entrepreneurship (Inacio Junior et al., total of inputs (Ramanathan, 2003). Also, it contributes to the 2021). As a result, entrepreneurship literature has strived to existing literature to evaluate and review if a particular coun- study systemic context and system-level conditions of entre- try’s institutional condition leads to the high efficiency of preneurial action (Ács et al., 2014). entrepreneurial activities. This paper seeks to provide an National Systems of Entrepreneurship (NSE) concept is empirical analysis of entrepreneurial efficiency based on the about institutional resource allocation closely related to the relational assumptions among the national framework condi- entrepreneurial attitudes, abilities, and aspirations of indi- tions and individual entrepreneurial outcomes as hypothe- viduals leading to the creation and operation of new busi- sized in the NSE concept. nesses (Ács et al., 2014; Tasnim & Afzal, 2018). In summary, the NSE depicts national capacity leading to entrepreneurial Global Entrepreneurship Efficiency Literature activities and presents the dynamic, inter-correlated relation- ships among the multi-dimensional economic, social, and Different studies suggested the methodological approach for institutional frameworks (Ács et al., 2014), and also helps efficiency comparison and analysis for global entrepreneur- explain the national economic capability for potential entre- ship. Bygrave et al. (2003) examined the correlation of entre- preneurial activities (Ács, Szerb, et al. 2014; Inacio Junior preneurship from a push-pull perspective in 29 countries et al., 2021; Tasnim & Afzal, 2018). Further recently, the participating in the GEM program, finding the relationship 4 SAGE Open between opportunity-pull entrepreneurship and variables with less-supportive institutional conditions might be lim- such as informal investment, entrepreneurial capacity, and ited. If a country’s NSE efficiency is defined by the total perception of start-up opportunities. Sternberg and input-output rate (Ramanathan, 2003), the variable selection Wennekers (2005) provided empirical evidence that the role might be reconsidered (Inacio Junior et al., 2021). of entrepreneurial activity varies according to economic Consequently, providing objective logic for national entre- development stages. Also, they suggested that regional preneurial efficiency and criteria for global comparison framework conditions, such as infrastructure and policies, benchmarks is expected to contribute valuable insights into need to be considered to investigate the regional entrepre- fostering early-stage businesses. neurial context. Lafuente et al. (2016), based on the data of 63 countries from 2012, tested the knowledge spillover the- Data Envelopment Analysis (DEA) ory of entrepreneurship using DEA, highlighting the rela- tionship between NSE and knowledge spillovers leading to Data envelopment analysis (DEA) was proposed by Charnes better efficiency. Tasnim and Afzal (2018) compared coun- et al. (1978). DEA is a part of the linear planning methodol- try-level efficiency based on the NSE perspective, hypothe- ogy and can evaluate the relative efficiency of each DMU sizing that individual-level actions and country-level (decision-making unit) by applying multiple input and out- framework conditions interact. Das and Kundu (2019) com- put factors (Casu & Molyneux, 2003). DEA is a non-para- pared countrywide efficiency based on GEDI and GEM data- metric statistical technique used in various fields because it sets from 2012 to 2016 to understand small-sized enterprise can be applied even when the measurement scale of each performance. They utilized attitudinal parameters as inde- variable is different (Barros & Alves, 2004; Nguyen et al., pendent variables and adopted dependent variables, such as 2016; Wu et al., 2009). DEA evaluates units that produce the nascent entrepreneurship rate. Also, Lafuente et al. maximum (weighted sum) outputs at minimum (weighted (2021) conducted an efficiency analysis using the non-para- sum) input levels as efficient DMUs (Smith & Street, 2005; metric technique to figure out how country-level entrepre- Stone, 2002). It informs the level of improvement and neurship relates to total factor productivity using global decrease of the input variable, has the advantage of express- entrepreneurship data from 2002 to 2013. Faghih et al. ing the growth potential of the output variable numerically, (2021) investigated entrepreneurship “overall efficiency,” and is used in various research fields (e.g., Nguyen et al., assessing national entrepreneurship attitude toward the 2016; Wu et al., 2009). entrepreneurship system, based on GEM 2018 data. Inacio DEA directly compares efficient and inefficient DMUs Junior et al. (2021) conducted an efficiency analysis of the and provides information on which DMUs can be bench- GEI data to demonstrate that the global entrepreneurial rank- marked to improve efficiency. It has two models, input-ori- ings may misinterpret NSE dynamics with the impression ented and output-oriented. After fixing the output value, the that countries showing higher national framework indicators former calculates the efficiency index based on the input fac- are more productive in generating actual entrepreneurial tor values, focusing on the input reduction. The latter fixes activity. The authors claimed that there might be another per- the input factors and pays attention to how much the output spective that prioritizes the efficiency of actual entrepreneur- can be improved. In addition, DEA has CCR (Charnes- ial activities. They pointed out that lower-positioned Cooper-Rhodes) model (Charnes et al., 1978) that assumes countries in framework conditions often present higher effi- the Constant Return to Scale (CRS) and BCC (Banker- ciency, producing more entrepreneurial business opportuni- Charnes-Cooper) model (Banker et al., 1984) based on ties at the individual level. Variable Return to Scale (VRS) assumption. The previous studies can be divided into three streams; (a) Research that dealt with the level of NSEs without efficiency Technical Efficiency (TE). The first concept proposed in analysis among countries (i.e., Ács et al., 2014), (b) Research efficiency research was Technical Efficiency (TE). TE calcu- that compared efficiency among countries without longitudi- lates the efficiency index through relative comparison using nal or time-series consideration (e.g., Inacio Junior et al., input-output data and assumes the constant return to scale. 2021; Tasnim & Afzal, 2018), and (c) Research that needs Technical efficiency theory was first introduced by Far- further probe in setting input-output relationship (i.e., Faghih rell (1957) and later rediscovered by Charnes et al. (1978) et al., 2021). In Faghih et al.’s efficiency study (2021), GDP and subsequently re-labeled as CCR-efficiency under DEA is defined as the output of the overall entrepreneurial activity. (Cooper et al., 2000). The CCR model measures TE, the Under the research setting, the author consequently priori- maximum output capacity for input, assuming that the scale tizes a country with less entrepreneurial activity (less input) value is invariant. However, this model works with DMUs and higher GDP (more output) as an efficient benchmark. If operating at the optimal scale and has the disadvantage of the researchers adopt this theoretical assumption, entrepre- not distinguishing between Scale Efficiency (SE) and Pure neurship-nurturing implications for developing economies Technical Efficiency (PTE). Kim 5 Pure Technical Efficiency (PTE). The BCC model calculates Table 1. DEA/Window Analysis Formula. the efficiency assuming that production constraints occur in Item Calculation the input-output relationship. The BCC model was proposed to overcome the shortcomings of the CCR model (Banker Number of windows k–p + 1 Number of DMUs in each window p × n et al., 1984), reflecting the variability of returns to scale. It Window Width k (odd number) (k + 1)/2 is a concept that examines the change in the outputs when k (even number) {(k + 1)/2} ± 1/2 the scale extends while keeping the ratio of input factors constant. BCC considers Increasing Return to Scale and Note. k: time, p: window width, w: # of windows. Decreasing Return to Scale, yielding Pure Technical Effi- ciency of DMUs. pure technical efficiency (PTE), and scale efficiency (SE) of DMUs. Fourth, dynamic efficiency analysis covered the lon- Scale Efficiency (SE). Scale Efficiency (SE) is calculated gitudinal examination using the DEA/Window method. from TE value (measured by CCR model) divided by PTE Fifth, the efficiency-stability matrix based on quadrant anal- value (measured by BCC model). SE value of “1” means the ysis was drawn, grouping the similarly-characterized coun- optimal scale state, and if it is less than “1,” the current input tries. The overall research framework is presented below and output are not achieving scale efficiency. SE value could (see Figure 1). provide the direction of efficiency improvement by identi- fying whether the cause of the inefficiency of the DMU is purely technical or in terms of scale through the relationship Data Collection between PTE and SE. This study utilized the Global Entrepreneurship Monitor (GEM) dataset from 2015 to 2020. GEM is a comparative DEA/Window Analysis. Although DEA has been used in study tracing entrepreneurial thoughts, behaviors, and activi- various analytical studies due to its cross-sectional charac- ties having over 150,000 participants from over 50 countries teristics, it has not frequently been applied to longitudinal (GEM, 2020). The research consists of Adult Population research. This study utilized the DEA/Window analysis, Survey (APS) and National Expert Survey (NES) and pro- which complements the cross-sectional characteristics of vides primary data on the competitiveness of each country’s DEA. Through this method, it is possible to evaluate how the NSE by surveying the adult population and experts (Das & performance of the same DMU changes according to time, Kundu, 2019). The countrywide dataset from 2015 to 2020 and DMUs are analyzed and divided into several windows consisted of 335 NES and 320 APS instances. The 24 coun- of multiple periods (Charnes & Cooper, 1984). In this analy- tries consecutively participated throughout the whole period sis, individual DMUs can be evaluated by referencing other (6 years, the time frame of this research) were chosen as DMUs belonging to different periods if the primary input- DMUs, and a total of 144 data instances were finally col- output relationship does not change over time. lected for analysis. As for the DMUs, four countries were For analysis, a researcher determines the window width from Latin America and Caribbean (17%), fourteen were (p) after collecting the longitudinal data for a certain period. from Europe and North America (58%), two were from The number of windows (w) is “k-p+1.” After analyzing the Africa (8%), and four were from Asia and Oceania (17%) first (p) period, the first analytical unit (e.g., year, month) is (see Table 3). replaced by the new unit to analyze the next window. This process is repeated until the final period k (see Table 1). The efficiency trend and stability of each DMU can be analyzed Data Processing by assessing the efficiency of each window. The DEA model This study adopted a computational analytical tool to elimi- used in this study is as follows (see Table 2). nate human errors. Efficiency Measurement System (Scheel, 2000) was utilized as a reliable DEA statistical tool (Afzal & Method Lawrey, 2012). This package provides the CCR, BCC, DEA/ Windows, and Malmquist analysis. Research Design This study evaluated global entrepreneurship’s static and Theoretical Foundation and Variable Selection dynamic efficiency and explored the benchmarking coun- tries for entrepreneurship efficiency. The analytical proce- A country’s institutional factors (framework conditions) are dure is as follows. First, the data collection process chose the known to be fundamentally correlated to formal structure/the primary data for the research. Second, the pre-processing norms of regulatory bodies and cultural/social practices while stage identified the input and output variables. Third, static playing a decisive role in promoting individual entrepreneur- efficiency analysis calculated the technical efficiency (TE), ial behaviors (Bianchi et al., 2015; Boudreaux et al., 2019; 6 SAGE Open Table 2. DEA Models Used in This Study. Model Description Author CCR Constant Returns to Scale (unchanged scale) Charnes et al. (1978) Technical Efficiency Suitable for DMUs operating at optimal scale BCC Variable Returns to Scale (variable scale) Banker et al. (1984) Pure Technical Efficiency Increasing Returns to Scale (IRS) and Decreasing Returns to Scale (DRS) DEA/Window Used to analyze dynamic changes of relative efficiency of DMUs over time Charnes and Cooper (1984) Figure 1. Research design. Table 3. Regional Composition. ID Region Country # % 1 Latin America and Caribbean Brazil, Chile, Colombia, Guatemala 4 17 2 Europe and North America Croatia, Germany, Greece, Italy, Luxembourg, Netherlands, Poland, Slovak 14 58 Republic, Slovenia, Spain, Sweden, Switzerland, United Kingdom, United States 3 Africa Egypt, Morocco 2 8 4 Asia and Oceania India, Israel, South Korea, Taiwan 4 17 Total 24 100 Note. Countries consecutively participated from 2015 to 2020. n = 24. Bylund & McCaffrey, 2017; Sendra-Pons et al., 2022). in various research topics, recently widening its presence in Therefore, it is of theoretical value and interest to view the studying determinant factors leading to the success of new entrepreneurship phenomenon from an institutional theory entrepreneurial activities (e.g., Bruton et al., 2010; Peng, standpoint regarding the remarkable contextual influence on 2001). The institutional environment, such as accountability, promoting entrepreneurial activities stemming from the effectiveness, and quality of related authorities, policies, and institutional environment (e.g., Bruton et al., 2010; Sendra- support organizations, may enhance or hinder the develop- Pons et al., 2022; Sinha et al., 2019). The institutional theory ment of businesses (Sendra-Pons et al., 2022). In this regard, assumes that the regulatory, social, and cultural elements the paper hypothesizes and selects variables based on the affect organizations to promote legitimacy and be poised for assumptions from the institutional theory point of view to survival (e.g., Ahlstrom & Bruton, 2003; Fang, 2010). The build an efficiency model. Determining valid inputs and out- theory has been broadly accepted as a theoretical foundation puts is substantial in analyzing efficiency (Chan & Karim, Kim 7 Table 4. Variables. Variable Description Source Input 1 GDP per capita in US dollars World bank (2021), IMF (2021) 2 Established Business Business owners % in 18–64 population Ownership (EBO) Financial value creation for more than 42 months 3 National Expert Composite Mean value of below 11 items GEM (2020) Index (NECI) *Entrepreneurial finance One recent survey item excluded *Government policies: support and relevance (Entrepreneurial post-school education) *Government policies: taxes and bureaucracy *Government entrepreneurship programs *Entrepreneurial education at school stage *R&D transfer *Commercial and legal infrastructure *Internal market dynamics *Internal market burdens or entry regulation *Physical infrastructure *Cultural and social norms Output 4 Total early-stage New business owner % in 18–64 population GEM (2020) Entrepreneurial Activity (TEA) Adult population starting or running a new business Note. NECI: based on National Expert Survey (NES). All variables: 2015 to 2020. GDP per capita: World bank, only Taiwan data from IMF. 2012). The following variables were selected to create the institutional theory. Thus, comparing the TEA level with efficiency model. The variables were determined by the con- national framework conditions might be a legitimate way to sensus of three researchers and 2 Ph.D. holders after a peer compare national entrepreneurship efficiency. review for validity. Finally, three input variables and one out- Consequently, this study set TEA as output and analyzed put variable were selected (see Table 4). the efficiency based on the weighted averaged inputs. Output Variable Input Variable (a) Total Early-stage Entrepreneurial Activity (TEA): (a) GDP per capita: A country’s population and GDP are TEA, a part of the GEM dataset, was selected as the output absolute indicators that could help understand its socioeco- variable. GEM describes entrepreneurship activity in a nomic resources. Each country has a different institutional country through a variety of indicators. TEA is defined as framework resulting from various aspects, and one of the the proportion of those who participate in early-stage busi- significant elements is the economic development stage of ness activities among the adult population aged 18 to 64. the country (Eijdenberg et al., 2019). Some researchers GEM identifies entrepreneurs into four levels; (a) potential viewed that entrepreneurship affected economic growth entrepreneur, (b) nascent entrepreneur (less than 3 months), (e.g., Doran et al., 2018; Stoica et al., 2020), while other (c) new business (less than 42 months), and (d) established researchers presented a different way of understanding the business (more than 42 months). TEA represents the pro- causal direction between the two variables, hypothesizing portion of nascent entrepreneurs and new business owners that GDP is among the other factors that may significantly among all entrepreneurs. TEA is a valuable benchmark for affect the emergence of the nascent entrepreneurs (e.g., entrepreneurial activity and an indicator adopted by many Micozzi, 2013; Rusu & Roman, 2017). Being a country’s economies to broadly gauge the degree of newly born economic development stage, GDP could be considered a entrepreneurs across the country (Wright, 2019). The emer- legitimate input factor in an efficiency analysis model. gence of new businesses could be viewed as an output mea- Additional empirical research shows that GDP per capita sig- sure under NSE (Ács et al., 2014), and there have been nificantly impacts entrepreneurial activity (Shane, 2010). academic approaches that adopted TEA as a dependent Since the research model adopts TEA, a ratio index, as an variable to investigate its affecting factors or to compare output variable, GDP per capita, which enables countrywide entrepreneurial performance among countries (e.g., comparison, was chosen as an input variable. Except for Micozzi, 2013; Rusu & Roman, 2017) based on the Taiwan, the 2015-2020 GDPPC data were secured from the 8 SAGE Open Table 5. Variable Descriptive Data. GDP per capita (USD) NECI EBO (%) TEA (%) Mean 30,489 2.8 7.9 11.3 Median 25,120 2.7 6.8 9.2 Standard deviation 26,614 0.4 3.6 6.6 Minimum 1,606 2.1 1.5 1.9 Maximum 116,654 3.6 20.3 36.7 25th percentile 11,319 2.5 5.2 7.1 50th percentile 25,120 2.7 6.8 9.2 75th percentile 43,595 3.0 10.4 13.0 Note. n = 144 (24 countries, 2015–2020). Worldbank (2021). Separate GDPPC data for Taiwan was results. The number of input and output factors is closely obtained from IMF (2021). related to the number of DMUs, and for validity, the num- (b) Established Business Ownership (EBO): EBO refers ber of DMUs should be over twice the product of the num- to the proportion of businesses that have been in operation for ber of input and output factors (Banker et al., 1984). This over 42 months and is calculated by the proportion of estab- study chose three inputs, one output factor, and 24 DMUs, lished business owners among the adult population aged 18 to satisfying the condition. 64. GEM recognizes the EBO rate as a part of the supportive The descriptive statistics of the variables are as follows environment leading to new businesses and a health indicator (see Table 5). From the data for 24 countries from 2015 to for a country’s entrepreneurial ecosystem (GEM, 2021). After 2020, the mean of GDPPC was 30,489 USD (min: 1,606, establishing a new business, it is essential to continue and max: 116,654). For NECI, the mean value was 2.8 (min: scale it. Comparing the EBO (input, the existing and estab- 2.1, max: 3.6), and the EBO rate was 7.9% (min: 1.5, max: lished business fundamental) and TEA (output, new busi- 20.3). TEA, the output variable, was 11.3% on average nesses) can signal how the outcome of the institutional (min: 1.9, max: 36.7). environment is resilient or productive regardless of the cur- rent entrepreneurial context within the country (Wright, Efficiency Analysis 2020), highlighting the country’s capacity to create new busi- ness opportunities. It is also considered one of the factors Static efficiency analysis. An output-oriented model was affecting new entrepreneurial activities (Almodóvar- applied to all analyses in this study. As discussed previ- González et al., 2020). By comparing EBO and TEA rates, the ously, national framework conditions are inputs, and ini- viability and vitality of a new business could be measured and tial entrepreneurship activities are output. Under these translated into efficiency. assumptions, reducing the inputs for efficiency (input-ori- (c) National Expert Composite Index (NECI): GEM’s ented model) is not reasonable, and improving the output sub-item, National Expert Survey (NES), supports the NSE under given conditions would be adequate. In addition, concept by evaluating expert opinions on national entrepre- since the efficiency level cannot be ranked among the effi- neurial conditions through a scale measure. Over 36 experts cient DMUs whose index is 1 (100%), the super-efficiency in 9 fields from each country participate in evaluating 12 model (Andersen & Petersen, 1993) is applied to rank effi- items such as the economic environment, market access, cient DMUs. The output-oriented values are converted into government policies, and entrepreneurship education. This inverse numbers to make comparison easier. Static analy- study calculated NECI by averaging the results of 11 items, sis proceeds in three steps; (a) TE analysis by CCR model, excluding the recently included item, entrepreneurial post- (b) PTE analysis by BCC model, and (c) SE analysis cal- school education. GEM considers NECI as a measure to culated by TE/PTE. gauge the easiness of beginning and developing a new busi- ness (GEM, 2020). Overall, NECI summarizes the evalua- Dynamic efficiency analysis. DEA/Window analysis is a tion results of a country’s entrepreneurial framework valuable method to evaluate how the performance of the conditions into a composite score (GEM, 2020) and may same DMU changes over time. Enabling comparative evalu- help efficiently compare the processes or the mechanisms ation over time can allow researchers to focus on the lon- leading to entrepreneurial behaviors, such as TEA. As a gitudinal side of the data. In this study, dynamic efficiency result, NECI was selected as a legitimate input variable for analysis is performed to measure changes in the effective- the efficiency analysis model. ness of entrepreneurial activities. The period is 6 years The total number of inputs and outputs in DEA should (2015–2020), and the window width (p) is three based on be limited to a minimum to avoid unrealistic evaluation the formula of ((k + 1)/2) ± 1/2, and through four windows, Kim 9 changes in the efficiency were measured. In addition, by Global entrepreneurship benchmarking: SE. SE was calcu- calculating the mean, standard deviation, LDY (largest dif- lated as the second step of the static efficiency analysis. SE ference between scores in the same year), and LDP (largest value smaller than one means that scale inefficiency exists. difference between scores across the entire period) within the From the relationship between PTE and SE, a researcher can analysis period, the entrepreneurship performance based on suggest directions for efficiency improvement by identifying the efficiency-stability of each country was calculated. whether the cause of the inefficiency is a purely technical issue or a scale issue. For example, in the SE results of 2020, Strategic quadrant analysis. This study also proposes a six DMUs have PTE values greater than SE values out of 24 strategic quadrant analysis based on the stability (LDP countries (Chile, Croatia, Egypt, India, Italy, and Morocco). value) and the mean efficiency of each DMU based on the It is understood that these countries can improve efficiency dynamic efficiency results. Unlike the existing literature, through economies of scale (SE) rather than resource conver- this study adds value by including another performance axis, sion (PTE) for entrepreneurial activities (see Table 7). entrepreneurial stability, suggesting a matrix framework to The five DMUs with the highest mean SE values over the compare the relative positions with peers. whole period were Guatemala, Colombia, Chile, Sweden, and United Kingdom. On the other hand, the bottom five countries were Morocco, Croatia, Italy, Egypt, and the Results Slovak Republic. The year with the highest number of effi- cient DMUs was 2015 (6 countries, % of efficient Static Efficiency Comparison DMU = 25.0%). On the other hand, the years with the lowest Global entrepreneurship benchmarking: TE. TE analysis number of efficient DMUs were 2018 and 2020 (2 countries, compares entrepreneurship efficiency and ranks the effi- % of efficient DMU = 8.3%). ciency levels, suggesting benchmarks among reference The regional analysis results are as follows (see Table 8). DMUs. The results are as follows (see Table 6). In 2015, Latin America and Caribbean area was the highest region in five DMUs were found effective in entrepreneurship activi- TE and SE except for 2016. Africa has been consistently ties (Colombia, India, Guatemala, Chile, and Brazil), and effective in terms of PTE for 6 years. Colombia was the most efficient and referenced country Overall efficiency averaging of all countries is as follows (inversed super-efficiency θ = 1.578, referenced = 16). As of (see Figure 2). It can be seen that TE, PTE, and SE have been 2016, three countries were found efficient, with Colombia on a downward trend since 2015. being the most efficient DMU (θ = 1.311, referenced = 20). India was second (θ = 1.270, referenced = 2), and Guatemala Dynamic Efficiency Comparison was third (θ = 1.004, referenced = 2). In 2017, four DMUs (Guatemala, Israel, Egypt, and Chile) proved themselves as DEA/Window analysis. A dynamic efficiency analysis was efficient, and the most efficient was Guatemala (θ = 1.676, performed to measure the efficiency change over time. DEA/ referenced = 8). The most referenced was Chile (θ = 1.080, Window analysis, which complements the cross-sectional referenced = 15), ranked fourth in efficiency, followed by characteristics of DEA, enables longitudinal comparisons Israel (θ = 1.535, referenced = 12). among DMUs by evaluating the performance based on Regarding 2018 results, two countries were found to be smoothed average values. The analysis period is from 2015 efficient; Guatemala (θ = 1.805, referenced = 13) and to 2020, and the window width (p) is set to 3, with four Colombia (θ = 1.221, referenced = 20). In 2019, four coun- windows (2015–2017, 2016–2018, 2017–2019, and 2018– tries were efficient (Guatemala, Colombia, Chile, and India). 2020). This analysis included the mean value of 24 countries Same as the previous year, Guatemala (θ = 1.442, refer- that presents the average value of entrepreneurial efficiency enced = 2) was the most efficient. Colombia (θ = 1.433, refer- for all countries included in the same time frame, which is a enced = 15) and Chile (θ = 1.364, referenced = 15) were the window of 4 years (Mean A), and the mean value of 4 win- most referenced. Concerning the TE results in 2020, two dows that can show a smoothed efficiency of a country over countries were picked as efficient DMUs; Colombia time (Mean B). The value of 0.471 shown where the x-axis (θ = 1.913, referenced = 20) and Guatemala (θ = 1.054, (Mean A) and y-axis (Mean B) meet is the average efficiency referenced = 17). of all DMUs within the entire window period (see Table 9). The overall TE analysis results demonstrate that Guatemala Mean A suggested the average value of all DMUs’ effi- has been an efficient unit throughout the whole period (six ciency scores in the same window. The overall average of times). Four countries (Guatemala 6, Colombia 5, Chile 3, and Mean A was 0.471, and Window 2 (2016–2018) showed the India 3 times) were selected as efficient DMUs over three highest efficiency (0.523). The window efficiency slightly times (17% of the entire countries). Three countries other than improved in the Window 2 period, then showed a downtrend India belonged to Latin America and Caribbean, showing high passing Windows 3 and 4. It could be interpreted that the competitiveness in entrepreneurial activity efficiency. overall global entrepreneurship efficiency trend has declined 10 Table 6. Entrepreneurship Technical Efficiency (2015–2020). 2015 2016 2017 2018 2019 2020 CCR CCR CCR CCR CCR CCR No. DMU SE Rank # Ref. SE Rank # Ref. SE Rank # Ref. SE Rank # Ref. SE Rank # Ref. SE Rank # Ref. 1 Brazil 1.007 5 0.811 8 0.798 7 0.635 10 0.762 6 0.783 4 2 Chile 1.029 4 9 0.983 4 1.080 4 15 0.992 3 1.364 3 15 0.856 3 3 Colombia 1.578 1 16 1.311 1 20 0.951 5 1.221 2 20 1.433 2 15 1.913 1 20 4 Croatia 0.630 8 0.650 9 0.756 8 0.701 9 0.563 8 0.535 6 5 Egypt 0.585 9 0.829 7 1.144 3 4 0.818 7 0.855 5 0.538 5 6 Germany 0.224 22 0.213 24 0.297 22 0.205 22 0.287 16 0.143 23 7 Greece 0.315 17 0.225 23 0.189 24 0.234 21 0.240 19 0.278 15 8 Guatemala 1.123 3 1 1.004 3 2 1.676 1 8 1.805 1 13 1.442 1 2 1.054 2 17 9 India 1.518 2 1 1.270 2 2 0.843 6 0.930 5 1.319 4 0.453 8 10 Israel 0.693 7 0.918 6 1.535 2 12 0.927 6 0.479 10 0.358 12 11 Italy 0.249 20 0.275 20 0.256 23 0.203 23 0.118 24 0.153 22 12 Luxembourg 0.708 6 0.934 5 0.711 9 0.965 4 0.420 12 0.393 10 13 Morocco 0.325 14 0.379 16 0.521 11 0.546 12 0.730 7 0.389 11 14 Netherlands 0.220 23 0.350 17 0.409 16 0.361 16 0.255 17 0.291 14 15 Poland 0.362 12 0.490 12 0.379 17 0.158 24 0.152 23 0.100 24 16 Slovak Republic 0.400 11 0.506 11 0.510 12 0.807 8 0.540 9 0.479 7 17 Slovenia 0.322 15 0.388 15 0.379 18 0.289 18 0.242 18 0.188 20 18 South Korea 0.351 13 0.330 19 0.490 14 0.449 13 0.368 13 0.350 13 19 Spain 0.233 21 0.272 21 0.334 19 0.322 17 0.210 21 0.161 21 20 Sweden 0.317 16 0.549 10 0.504 13 0.393 14 0.326 14 0.231 17 21 Switzerland 0.219 24 0.240 22 0.323 20 0.236 20 0.233 20 0.255 16 22 Taiwan 0.270 19 0.346 18 0.305 21 0.259 19 0.191 22 0.211 19 23 United Kingdom 0.298 18 0.469 13 0.448 15 0.393 15 0.288 15 0.228 18 24 United States 0.420 10 0.445 14 0.658 10 0.605 11 0.453 11 0.443 9 Note. CCR & SE (super-efficiency): inverse number / Bold: SE (CCR = 1.000). Kim 11 Table 7. Entrepreneurship scale efficiency (2015–2020). 2015 2016 2017 2018 2019 2020 DMU TE PTE SE TE PTE SE TE PTE SE TE PTE SE TE PTE SE TE PTE SE Brazil 1.000 1.000 1.000 0.811 1.000 0.811 0.798 0.819 0.974 0.635 0.651 0.975 0.762 0.777 0.981 0.783 0.783 1.000 Chile 1.000 1.000 1.000 0.983 0.989 0.993 1.000 1.000 1.000 0.992 1.000 0.992 1.000 1.000 1.000 0.856 0.926 0.924 Colombia 1.000 1.000 1.000 1.000 1.000 1.000 0.951 0.965 0.985 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Croatia 0.630 1.000 0.630 0.650 1.000 0.650 0.756 1.000 0.756 0.701 1.000 0.701 0.563 1.000 0.563 0.535 1.000 0.535 Egypt 0.585 1.000 0.585 0.829 1.000 0.829 1.000 1.000 1.000 0.818 1.000 0.818 0.855 1.000 0.855 0.538 1.000 0.538 Germany 0.224 0.233 0.964 0.213 0.216 0.990 0.297 0.303 0.979 0.205 0.216 0.950 0.287 0.311 0.921 0.143 0.154 0.924 Greece 0.315 0.330 0.956 0.225 0.256 0.880 0.189 0.194 0.974 0.234 0.236 0.993 0.240 0.256 0.940 0.278 0.281 0.987 Guatemala 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 India 1.000 1.000 1.000 1.000 1.000 1.000 0.843 1.000 0.843 0.930 1.000 0.930 1.000 1.000 1.000 0.453 1.000 0.453 Israel 0.693 0.810 0.856 0.918 1.000 0.918 1.000 1.000 1.000 0.927 1.000 0.927 0.479 0.508 0.941 0.358 0.434 0.825 Italy 0.249 0.332 0.752 0.275 0.341 0.805 0.256 0.277 0.923 0.203 0.214 0.951 0.118 0.120 0.982 0.153 1.000 0.153 Luxembourg 0.708 0.942 0.751 0.934 1.000 0.934 0.711 0.711 1.000 0.965 1.000 0.965 0.420 0.440 0.954 0.393 0.560 0.702 Morocco 0.325 1.000 0.325 0.379 1.000 0.379 0.521 1.000 0.521 0.546 1.000 0.546 0.730 1.000 0.730 0.389 1.000 0.389 Netherlands 0.220 0.278 0.793 0.350 0.401 0.873 0.409 0.458 0.894 0.361 0.447 0.807 0.255 0.283 0.899 0.291 0.370 0.786 Poland 0.362 0.392 0.922 0.490 0.508 0.963 0.379 0.382 0.992 0.158 0.189 0.837 0.152 0.157 0.971 0.100 0.101 0.990 Slovak Republic 0.400 0.413 0.969 0.506 0.555 0.912 0.510 0.524 0.975 0.807 0.984 0.819 0.540 0.915 0.590 0.479 0.620 0.772 Slovenia 0.322 0.359 0.897 0.388 0.406 0.955 0.379 0.379 0.999 0.289 0.294 0.982 0.242 0.245 0.986 0.188 0.193 0.974 South Korea 0.351 0.378 0.929 0.330 0.336 0.980 0.490 0.532 0.921 0.449 0.535 0.841 0.368 0.406 0.906 0.350 0.418 0.838 Spain 0.233 0.236 0.987 0.272 0.282 0.965 0.334 0.361 0.924 0.322 0.322 0.997 0.210 0.230 0.914 0.161 0.167 0.962 Sweden 0.317 0.317 1.000 0.549 0.580 0.945 0.504 0.510 0.987 0.393 0.397 0.991 0.326 0.350 0.929 0.231 0.235 0.986 Switzerland 0.219 0.282 0.777 0.240 0.299 0.802 0.323 0.353 0.915 0.236 0.269 0.876 0.233 0.267 0.871 0.255 0.296 0.863 Taiwan 0.270 0.282 0.958 0.346 0.348 0.993 0.305 0.348 0.876 0.259 0.345 0.750 0.191 0.229 0.834 0.211 0.270 0.783 United Kingdom 0.298 0.303 0.986 0.469 0.485 0.966 0.448 0.455 0.986 0.393 0.393 0.999 0.288 0.298 0.964 0.228 0.251 0.910 United States 0.420 0.477 0.880 0.445 0.460 0.967 0.658 0.670 0.982 0.605 0.652 0.929 0.453 0.475 0.954 0.443 0.495 0.895 Mean 0.506 0.598 0.872 0.567 0.644 0.896 0.586 0.635 0.934 0.559 0.631 0.899 0.488 0.553 0.904 0.409 0.565 0.799 No. efficient DMU 5 8 6 3 9 3 4 7 5 2 9 2 4 7 4 2 7 2 % of efficient DMU 20.8 33.3 25.0 12.5 37.5 12.5 16.7 29.2 20.8 8.3 37.5 8.3 16.7 29.2 16.7 8.3 29.2 8.3 Table 8. Entrepreneurship scale efficiency by region (2015–2020). 2015 2016 2017 2018 2019 2020 Region TE PTE SE TE PTE SE TE PTE SE TE PTE SE TE PTE SE TE PTE SE Latin America and Caribbean 1.000 1.000 1.000 0.948 0.997 0.951 0.937 0.946 0.990 0.907 0.913 0.992 0.940 0.944 0.995 0.910 0.927 0.981 Europe and North America 0.351 0.421 0.876 0.429 0.485 0.900 0.439 0.470 0.949 0.419 0.472 0.914 0.309 0.382 0.888 0.277 0.409 0.817 Africa 0.455 1.000 0.455 0.604 1.000 0.604 0.760 1.000 0.760 0.682 1.000 0.682 0.793 1.000 0.793 0.464 1.000 0.464 Asia and Oceania 0.579 0.617 0.936 0.648 0.671 0.973 0.660 0.720 0.910 0.641 0.720 0.862 0.509 0.536 0.920 0.343 0.530 0.725 Note. Efficiency: mean value of 24 DMUs. since the Window 2 period (2016–2018), and the dynamics, drivers, and relevant contextual factors that soared the global entrepreneurial efficiency during this period may require fur- ther research. Mean B can be interpreted as representing the continued overall efficiency of each DMU over 6 years. According to the result, Latin America and Caribbean DMUs, such as Chile, Colombia, and Guatemala, were among the top three countries. Four additional indicators were calculated based on the above results (see Table 10). First, the four windows’ average value and standard deviation were calculated. Second, LDY (largest difference between scores in the same year) means the maximum value among the differences in efficiency values for Figure 2. Entrepreneurship efficiency: TE, PTE, SE (2015–2020). each country in the same year. Third, LDP (largest difference 12 SAGE Open Table 9. DEA/Window Analysis. Window DMU (1) 2015–2017 (2) 2016–2018 (3) 2017–2019 (4) 2018–2020 Mean (B) Brazil 0.806 0.693 0.673 0.690 0.716 Chile 0.892 0.909 0.826 0.858 0.871 Colombia 0.893 0.898 0.843 0.879 0.878 Croatia 0.517 0.622 0.492 0.487 0.530 Egypt 0.799 0.834 0.846 0.655 0.784 Germany 0.198 0.221 0.214 0.182 0.204 Greece 0.223 0.199 0.201 0.229 0.213 Guatemala 0.934 0.924 0.939 0.964 0.940 India 0.914 0.897 0.875 0.770 0.864 Israel 0.743 0.868 0.600 0.435 0.661 Italy 0.210 0.224 0.149 0.134 0.179 Luxembourg 0.666 0.755 0.517 0.446 0.596 Morocco 0.383 0.432 0.504 0.434 0.438 Netherlands 0.269 0.340 0.270 0.268 0.287 Poland 0.358 0.308 0.183 0.126 0.244 Slovak Republic 0.414 0.556 0.485 0.465 0.480 Slovenia 0.303 0.322 0.238 0.201 0.266 South Korea 0.337 0.382 0.364 0.349 0.358 Spain 0.226 0.280 0.222 0.175 0.226 Sweden 0.367 0.438 0.307 0.252 0.341 Switzerland 0.231 0.240 0.213 0.221 0.226 Taiwan 0.264 0.277 0.207 0.199 0.237 United Kingdom 0.323 0.401 0.294 0.249 0.317 United States 0.438 0.528 0.456 0.418 0.460 Mean (A) 0.488 0.523 0.455 0.420 0.471 Note. Countrywide mean of each window’s CCR-output oriented model value, inversed. between scores across the entire period) is the difference need to balance their framework conditions to pursue effi- between maximum and minimum scores of a country within ciency and stability. Seven countries are in Quadrant 2, and the entire period. Through these indicators, the efficiency-sta- twelve are in Quadrant 3. Three countries (Guatemala, bility of entrepreneurial performance can be identified. Brazil, and Croatia) located in Quadrant 1 with higher aver- Guatemala showed the highest efficiency (0.940), and age entrepreneurial efficiency and stability are considered Italy had the lowest value (0.179). The DMU with the lowest benchmarks for DMUs in other quadrants. DMUs in inter-window standard deviation was Switzerland Quadrant 1 are considered to have shown higher-than-aver- (SD = 0.010), and Israel showed the largest value (SD = 0.162). age entrepreneurial efficiency continuously regarding their Greece was the DMU with the lowest LDY (0.028), which framework conditions. The two countries included in showed stable entrepreneurship performance. The Slovak Quadrant 4 display low efficiency and low stability. These Republic indicated the largest LDY value (0.300). DMUs countries may need attention and efforts at the national level with low LDP values were Switzerland (0.076) and Greece to promote and foster entrepreneurship activities. (0.113), which showed the slightest change in efficiency dur- ing the entire period. On the other hand, Israel showed the Conclusion most significant change (0.642). This paper aimed to present a diverse way to understand entrepreneurial activities as the outcome of institutional con- Efficiency-Stability Model ditions and benchmark for country-level efficiency, based on A strategic quadrant matrix analysis presented efficiency- the NSE concept from a “systemic efficiency” viewpoint. As stability axes based on dynamic efficiency results (see Figure a result, this paper presented research questions regarding; 3). A model with four quadrants consisting of the X-axis (a) differences in entrepreneurial efficiencies at the global (Mean, average = 0.471) and Y-axis (LDP, average = 0.266) level, (b) the countries for benchmarking, and c) the implica- was created. DMUs belonging to Quadrant 2 (High effi- tions for entrepreneurial activities. The study conducted ciency–Low stability) and 3 (Low efficiency–High stability) static/dynamic efficiency analysis and strategic quadrant Kim 13 Table 10. Dynamic Efficiency Analysis. Measure Rank DMU Mean SD LDY LDP Mean SD LDY LDP Brazil 0.716 0.053 0.104 0.246 6 16 12 13 Chile 0.871 0.032 0.205 0.306 3 8 19 17 Colombia 0.878 0.021 0.288 0.323 2 6 22 18 Croatia 0.530 0.054 0.266 0.266 9 17 21 15 Egypt 0.784 0.076 0.125 0.453 5 21 14 22 Germany 0.204 0.015 0.064 0.145 23 4 7 5 Greece 0.213 0.013 0.028 0.113 22 2 1 2 Guatemala 0.940 0.015 0.080 0.149 1 3 10 7 India 0.864 0.056 0.055 0.565 4 19 4 23 Israel 0.661 0.162 0.296 0.642 7 24 23 24 Italy 0.179 0.038 0.058 0.148 24 11 6 6 Luxembourg 0.596 0.121 0.255 0.423 8 23 20 21 Morocco 0.438 0.043 0.158 0.279 12 13 17 16 Netherlands 0.287 0.031 0.078 0.139 16 7 9 4 Poland 0.244 0.093 0.058 0.378 18 22 5 19 Slovak Republic 0.480 0.051 0.300 0.398 10 15 24 20 Slovenia 0.266 0.049 0.090 0.187 17 14 11 9 South Korea 0.358 0.017 0.076 0.191 13 5 8 10 Spain 0.226 0.037 0.116 0.154 21 10 13 8 Sweden 0.341 0.069 0.144 0.250 14 20 16 14 Switzerland 0.226 0.010 0.040 0.076 20 1 2 1 Taiwan 0.237 0.034 0.048 0.137 19 9 3 3 United Kingdom 0.317 0.055 0.128 0.210 15 18 15 12 United States 0.460 0.041 0.186 0.201 11 12 18 11 Note. SD: Standard Deviation (population); LDY: Largest Difference between scores in the same Year; LDP: Largest Difference between scores across the entire Period. Efficiency (CCR) mean: 0.471, Stability (LDP) mean: 0.266. analysis for 24 countries throughout the 2015 to 2020 period. national framework conditions (NECI), and (c) the vitality of The results are summarized as follows. existing businesses (EBO). TEA was selected as the output From TE results of static efficiency analysis, Guatemala variable. Different from the conclusions of previous litera- was an efficient DMU with outstanding input-output com- ture that centered on the superiority of regional framework petitiveness regarding regional framework conditions among conditions (i.e., Ács et al., 2014) or that included the degree peer countries. Additionally, from the SE point of view, it of economic development stage (GDP) as an output variable was confirmed that the DMUs - such as Guatemala, (i.e., Das & Kundu, 2019), this study highlighted the output Colombia, and Chile - in Latin America and Caribbean coun- of actual entrepreneurial activity stemming from the given tries showed higher efficiency, leading to the birth of nascent conditions. The US, having been considered to have the and new entrepreneurs. The dynamic efficiency analysis highest level of NSE already, also proved itself as a produc- results also revealed that the efficiency of countries in this tive DMU in the strategic quadrant analysis in this study. region, such as Guatemala, was outstanding. The setting of However, in terms of efficiency, which is output versus input, national direction or priority in policymaking might differ by it could be understood that developing economies have country when weighing the two axes provided by this study higher efficiency than advanced economies with better (efficiency or stability). However, when viewed from a framework conditions. global entrepreneurship perspective, the benchmarking focus The higher efficiency in Latin America and Caribbean could be on efficiency first, prioritizing the activity of newly countries might be attributed to relative resource constraints, born businesses. Then the strategic priority could be thought limited access to business, financial infrastructure, or the of in the order of Quadrants 1, 2, 3, and 4. This research was inefficiency of the formal labor market (GEM, 2019). based on the theoretical assumptions of the institutional the- However, at the same time, considering how to foster initial ory and determined input variables such as (a) the economic entrepreneurial activities as a significant axis of economic development stage of a country (GDP per capita), (b) overall development despite the limitations of given national 14 SAGE Open Figure 3. Entrepreneurship efficiency-stability model. systems conditions would provide academic and policymak- enhanced the theoretical value of NSE by synthesizing the ing insight on finding proper benchmarks. efficiency theory to examine what kind of entrepreneurial result the environmental and contextual inputs have created. The study contributes to the literature by expanding and Implications diversifying the views on entrepreneurial efficiency by This study suggests several theoretical implications. First, highlighting TEA, the less studied output under NSE the paper synthesized the National Systems of assumptions, as a dependent variable that may help establish Entrepreneurship concept and efficiency theory to provide a a feasible efficiency comparison model based on DEA. novel way of assessing the national entrepreneurial activi- Second, a multi-faceted evaluation model that could com- ties based on the institutional theory. Having excellent infra- pare global entrepreneurship efficiency was presented. This structure for entrepreneurship and making as many research suggested a novel, integrated methodology using early-stage businesses as possible are two different issues. realistic input-output variables. This approach applies to This study does not merely compare national NSEs but countries at various stages of economic development looking advances a perspective that might demonstrate the causal for a similar-level benchmark. This study revealed three relationship of a national entrepreneurial system. For exam- evaluation methodologies to consider for an overall picture; ple, Ács, Autio, et al (2014) presented GEDI as a valuable (a) static analysis, (b) dynamic analysis, and (c) strategic framework for NSE. The indicators under the concept depict quadrant analysis. Few studies on global entrepreneurship how a nation’s framework conditions are prepared and brought intensive evaluation perspectives to date to identify might be adequately viewed as a process rather than an out- the comparable positions of DMUs in a global landscape put. It would be challenging to consider the evaluation based on an efficiency evaluation. In particular, the strategic results of the institutional contexts as a country’s “systemic quadrant analysis framework presented evaluation axes, effi- efficiency.” However, this paper, standing on a simple effi- ciency-stability, presenting two essential perspectives that ciency relationship between major factors, has further should be considered for entrepreneurship policymaking. Kim 15 The findings of this study suggested implications for resource based on the input-output efficiency) may still exist. Future management and the significance of measuring efficiency scholars may want to consider a more sophisticated research from multiple perspectives. design using other models. The policymaking implications are as follows. This study provides a valuable reference for entrepreneurship policy Declaration of Conflicting Interests planning and implementation. Setting a reasonable bench- The author declared no potential conflicts of interest with respect to mark is essential, and only evaluating a benchmark’s frame- the research, authorship, and/or publication of this article. work conditions would hinder the policy’s effectiveness. Benchmarking refers to exploring and utilizing other organi- Funding zations’ excellent practices, services, and products (Main, The author received no financial support for the research, author- 1992). Selecting a country with excellent entrepreneurial ship, and/or publication of this article. efficiency in a similar socioeconomic environment is reason- able considering national differences in institutional contexts ORCID iD such as framework conditions (Lins et al., 2003). Unlike previous literature, this study suggested a dynamic Sehoon Kim https://orcid.org/0000-0002-6345-7433 efficiency analysis (longitudinal) methodology and a cross- sectional (static) perspective to find a benchmark that shows References outstanding output in a similar input structure. In addition, Ács, Z. J., & Audretsch, D. B. (1988). Innovation in large and small this study presented the global entrepreneurial frontier by firms: An empirical analysis. The American Economic Review, including multiple aspects of entrepreneurial efficiency with 78(4), 678–690. regional analyses. Entrepreneurship-supporting entities and Ács, Z. J., Autio, E., & Szerb, L. (2014). National systems of entre- policymakers in each country will be able to get insight into preneurship: Measurement issues and policy implications. Research Policy, 43(3), 476–494. improving their national infrastructure, which would eventu- Ács, Z. J., Szerb, L., & Autio, E. (2014). Global entrepreneur- ally lead to the birth of initial stage entrepreneurs based on ship and development index 2014. CreateSpace Independent the findings presented in this study. Publishing Platform. Afzal, M. N. I., & Lawrey, R. (2012). Knowledge-based econ- Limitations and Future Research omy (KBE) frameworks and empirical investigation of KBE input-output indicators for ASEAN. International Journal of This study has several limitations. First, the judgment on Economics and Finance, 4(9), 13–22. entrepreneurial efficiency may differ depending on the Ahlstrom, D., & Bruton, G. D. (2003). An institutional perspective viewpoint. However, critical views see entrepreneurship as on the role of culture in shaping strategic actions by technol- a national phenomenon beyond mere new business creation ogy-focused entrepreneurial firms in China. Entrepreneurship rate (Ács et al., 2014; Lafuente et al., 2021). Also, this paper Theory and Practice, 26(4), 53–68. Almodóvar-González, M., Fernández-Portillo, A., & Díaz-Casero, did not clarify the motivation for early-stage entrepreneurs J. C. (2020). Entrepreneurial activity and economic growth. A in the research design. Studying decisive factors for neces- multi-country analysis. European Research on Management sity-driven and improvement-driven entrepreneurship and Business Economics, 26(1), 9–17. (GEM, 2021) may interest researchers. Second, there may Andersen, P., & Petersen, N. C. (1993). A procedure for ranking exist more substantial factors that influence early-stage efficient units in data envelopment analysis. Management entrepreneurs. The selection of input factors is crucial in Science, 39(10), 1261–1264. creating a DEA model (Soares de Mello et al., 2008), which Anderson, A. R., & Starnawska, M. (2008). Research Practices affects the overall assessment. Future researchers could in Entrepreneurship: Problems of definition, description and apply hypotheses or assumptions different from those meaning. International Journal of E-Entrepreneurship and applied to this study and suggest a new model. Third, differ- Innovation, 9(4), 221–230. ences in the socio-cultural environment could be consid- Autio, E. (2007). GEM 2007 Report on High-Growth Entrepreneurship. ered. Cultural factors may affect businesses and Banker, R. D., Charnes, A., & Cooper, W. W. (1984). Some models entrepreneurial activities (i.e., Hofstede, 1984). Studying for estimating technical and scale inefficiencies in data envel- these relationships may also contribute to advancing entre- opment Analysis. Management Science, 30(9), 1078–1092. preneurship literature. Fourth, this study did not fully con- Barros, C. P., & Alves, F. P. (2004). Productivity in the tourism sider the economic differences among countries. Even industry. International Advances in Economic Research, 10(3), between two countries having similar GDP levels, the compo- 215–225. sition of the working-age population or the economy’s job- Bianchi, C. G., Borini, F. M., & Ogasavara, M. H. (2015). Creative creating abilities may differ. Alternatively, in an economy industry internationalisation: The impact of regulative, norma- where decent jobs are continuously provided, the desire to tive and cognitive institutions in developed and developing start a business may be limited. Lastly, the shortcomings of economies. Journal for Global Business Advancement, 8(4), DEA modeling (oversimplifies the complicated phenomenon 451–468. 16 SAGE Open Blanchflower, D. G. (2000). Self-employment in OECD countries. Feldman, M. P. (2001). The Entrepreneurial Event Revisited: Firm Labour Economics, 7(5), 471–505. formation in a Regional Context. Industrial and Corporate Bogetoft, P. (2012). Performance benchmarking: Measuring and Change, 10(4), 861–891. managing performance. Springer. Ferreira, J. J., Fernandes, C. I., & Ratten, V. (2017). Entrepreneurship, Boudreaux, C. J., Nikolaev, B. N., & Klein, P. (2019). Socio- innovation and competitiveness: What is the connection? cognitive traits and entrepreneurship: The moderating role of International Journal of Business and Globalisation, 18(1), economic institutions. Journal of Business Venturing, 34(1), 73–95. 178–196. Freeman, C. (1987). Technology Policy and Economic Performance: Bruton, G. D., Ahlstrom, D., & Li, H. (2010). Institutional theory Lessons from Japan. Pinter Publishers. and entrepreneurship: Where are We now and where do we Gedeon, S. (2010). What is entrepreneurship? Entrepreneurial need to move in the future? Entrepreneurship Theory and Practice Review, 1(3), 16–35. Practice, 34(3), 421–440. GEM. (2019). How policymakers in Guatemala can foster enter- Bygrave, W., Hay, M., Ng, E., & Reynolds, P. (2003). Executive preneurship. https://www.gemconsortium.org/news/how-poli- forum: A study of informal investing in 29 nations composing cymakers-in-guatemala-can-foster-entrepreneurship the Global Entrepreneurship Monitor. Venture Capital, 5(2), GEM. (2020). GEM 2019 / 2020 Global Report. https://www.gem- 101–116. consortium.org/report Bylund, P. L., & McCaffrey, M. (2017). A theory of entrepre- GEM. (2021). GEM 2020 / 2021 Global Report. https://www.gem- neurship and institutional uncertainty. Journal of Business consortium.org/reports/latest-global-report Venturing, 32(5), 461–475. Hofstede, G. (1984). Cultural dimensions in management and plan- Casu, B., & Molyneux, P. (2003). A comparative study of efficiency ning. Asia Pacific Journal of Management, 1(2), 81–99. in European banking. Applied Economics, 35(17), 1865–1876. IMF. (2021). GDP per capita, current prices. https://www.imf.org/ Chan, S.-G., & Karim, M. (2012). Public spending efficiency and external/datamapper/NGDPDPC@WEO/AZE political and economic factors: Evidence from selected East Inacio Junior, E., Dionisio, E. A., Fischer, B. B., Li, Y., & Meissner, Asian countries. Economic Annals, 57(193), 7–23. D. (2021). The global entrepreneurship index as a benchmark- Charnes, A., & Cooper, W. W. (1984). Preface to topics in data ing tool? Criticisms from an efficiency perspective. Journal of envelopment analysis. Annals of Operations Research, 2(1), Intellectual Capital, 22(1), 190–212. 59–94. Inzelt, A. (2004). The evolution of university–industry–govern- Charnes, A., Cooper, W. W., & Rhodes, E. (1978). Measuring ment relationships during transition. Research Policy, 33(6-7), the efficiency of decision making units. European Journal of 975–995. Operational Research, 2(6), 429–444. Kuhlmann, S. (2003). Evaluation of research and innovation poli- Cooper, W. W., Seiford, L. M., & Tone, K. (2000). Data cies: A discussion of trends with examples from Germany. Envelopment Analysis: A comprehensive text with models, International Journal of Technology Management, 26(2/3/4), applications, and references. Kluwer. 131–149. Cs, Z. J., Szerb, L., & Autio, E. (2014). Global Entrepreneurship Lafuente, E., Acs, Z. J., Sanders, M., & Szerb, L. (2021). Correction and Development Index 2014. CreateSpace Independent to: The global technology frontier: Productivity growth and the Publishing Platform. relevance of Kirznerian and Schumpeterian entrepreneurship. Das, S., & Kundu, A. (2019). Benchmarking a country for effi- Small Business Economics, 56(4), 1783–2178. ciency improvement: A DEA-based approach. Journal of Lafuente, E., Szerb, L., & Acs, Z. J. (2016). Country level efficiency Global Entrepreneurship Research, 9(1), 49. and national systems of entrepreneurship: A data envelopment Doran, J., McCarthy, N., & O’Connor, M. (2018). The role of analysis approach. The Journal of Technology Transfer, 41(6), entrepreneurship in stimulating economic growth in developed 1260–1283. and developing countries. Cogent Economics & Finance, 6(1), Lins, M. P. E., Gomes, E. G., Soares de Mello, J. C. C. B., & Soares 1442093. de Mello, A. J. R. (2003). Olympic ranking based on a zero Edquist, C., & Zabala-Iturriagagoitia, J. M. (2015). The Innovation sum gains DEA model. European Journal of Operational Union scoreboard is flawed: The case of Sweden – not being Research, 148(2), 312–322. the innovation leader of the EU. Research and Competences in Lumpkin, G. T., & Dess, G. G. (1996). Clarifying the entrepre- the Learning Economy, 16, 1–15. neurial orientation construct and linking it to performance. Eijdenberg, E. L., Thompson, N. A., Verduijn, K., & Essers, C. Academy of Management Review, 21(1), 135–172. (2019). Entrepreneurial activities in a developing country: Lundvall, B.-Å., Johnson, B., Andersen, E. S., & Dalum, B. (2002). An institutional theory perspective. International Journal of National systems of production, innovation and competence Entrepreneurial Behaviour & Research, 25(3), 414–432. building. Research Policy, 31(2), 213–231. Faghih, N., Bonyadi, E., & Sarreshtehdari, L. (2021). Assessing the Main, J. (1992). How to steal the best ideas around. Fortune, entrepreneurship process: An application of the data envelop- 126(8), 102–106. ment analysis. Journal of Global Entrepreneurship Research. Micozzi, A. (2013). Factors affecting the entrepreneurial dynamics Fang, T. (2010). Asian management research needs more self- in Italy: A comparison across European countries. Economia confidence: Reflection on Hofstede (2007) and beyond. Asia Marche Journal of Applied Economics, 32(1), 79–95. Pacific Journal of Management, 27(1), 155–170. Nelson, R. R. (1993). National innovation systems: A comparative Farrell, M. J. (1957). The measurement of productive efficiency. analysis. Oxford University Press. Journal of the Royal Statistical Society Series A (General), Nguyen, H.-O., Nguyen, H.-V., Chang, Y.-T., Chin, A. T. H., 120(3), 253–281. & Tongzon, J. (2016). Measuring port efficiency using Kim 17 bootstrapped DEA: The case of Vietnamese ports. Maritime Solow, R. M. (1957). Technical change and the aggregate produc- Policy & Management, 43(5), 644–659. tion function. The Review of Economics and Statistics, 39, Parker, S. C. (2009). The economics of entrepreneurship. Cambridge 312–320. University Press. Sternberg, R. (2009). Regional dimensions of entrepreneurship. Peng, M. W. (2001). How entrepreneurs create wealth in transition Foundations and Trends in Entrepreneurship, 5(4), 211–340. economies. The Academy of Management Perspectives, 15(1), Sternberg, R., & Wennekers, S. (2005). Determinants and effects of 95–108. new business creation using Global Entrepreneurship Monitor Ramanathan, R. (2003). An introduction to data envelopment anal- Data. Small Business Economics, 24(3), 193–203. ysis: A tool for performance measurement (1st ed.). SAGE Stoica, O., Roman, A., & Rusu, V. D. (2020). The nexus between Publications. entrepreneurship and economic growth: A comparative analy- Reynolds, P., Bosma, N., Autio, E., Hunt, S., De Bono, N., Servais, sis on groups of countries. Sustainability, 12(3), 1186. I., Lopez-Garcia, P., & Chin, N. (2005). Global entrepreneur- Stone, M. (2002). How not to measure the efficiency of public ser- ship monitor: Data collection design and implementation 1998- vices (and how one might). Journal of the Royal Statistical 2003. Small Business Economics, 24(3), 205–231. Society Series A (General), 165(3), 405–434. Rusu, V., & Roman, A. (2017). Entrepreneurial activity in the EU: Szerb, L., Lafuente, E., Horváth, K., & Páger, B. (2019). The rel- An empirical evaluation of its determinants. Sustainability, evance of quantity and quality entrepreneurship for regional 9(10), 1679. performance: The moderating role of the entrepreneurial eco- Scheel, H. (2000). EMS: Efficiency measurement system. http:// system. Regional Studies, 53(9), 1308–1320. www.holger-scheel.de/ems/ Tasnim, N., & Afzal, M. N. I. (2018). An empirical investiga- Schumpeter, J. A. (1934). The theory of economic development: An tion of country level efficiency and national systems of inquiry into profits, capital, credit, interest, and the business entrepreneurship using data envelopment analysis (DEA) cycle. Harvard University Press. and the TOBIT model. Journal of Global Entrepreneurship Sendra-Pons, P., Comeig, I., & Mas-Tur, A. (2022). Institutional factors Research, 8(1), 37. affecting entrepreneurship: A QCA analysis. European Research Thornton, C. V. D., Greiner, J. M., Hatry, H. P., Koss, M. P., Millar, on Management and Business Economics, 28(3), 100187. A. P., & Woodward, J. P. (1982). Productivity and motivation: Shane, S., & Venkataraman, S. (2000). The promise of entre- A review of state and local government Initiatives. Journal of preneurship as a field of research. Academy of Management Policy Analysis and Management, 1, 570. Review, 25(1), 217–226. van Praag, C. M., & Versloot, P. H. (2007). What is the value of Shane, S. A. (2010). The illusion of entrepreneurship: The costly entrepreneurship? A review of recent research. Small Business myths that entrepreneurs, investors and policymakers live by. Economics, 29(4), 351–382. Yale University Press. Worldbank. (2021). World development indicators. https://data- Shephard, R. W. (1970). The theory of cost and production func- bank.worldbank.org/reports.aspx?source=2&series=NY.GDP. tions. Princeton University Press. PCAP.CD&country Sinha, S., Singh, S., Mukunda Das, V., & Sharma, A. (2019). A Wright, F. (2019). Spotlighting trends in early-stage entrepre- framework for linking entrepreneurial ecosystem with institu- neurial activity. GEM. https://www.gemconsortium.org/ tional factors: A modified total interpretive structural modelling news/Spotlighting%20Trends%20in%20Early-Stage%20 approach. Journal for Global Business Advancement, 12(3), Entrepreneurial%20Activity 382–404. Wright, F. (2020). How Colombia’s recent entrepreneurial success Smith, P. C., & Street, A. (2005). Measuring the efficiency of public can continue despite Coronavirus. GEM. https://www.gem- services: The limits of analysis. Journal of the Royal Statistical consortium.org/news/how-colombia%E2%80%99s-recent- Society Series A (General), 168(2), 401–417. entrepreneurial-success-can-continue-despite-coronavirus Soares de Mello, J. C. C. B., Meza, L. A., & Silva, B. B. D. (2008). Wu, J., Liang, L., & Yang, F. (2009). Achievement and bench- Some rankings for the Athens Olympic Games using DEA marking of countries at the Summer Olympics using cross effi- models with a constant input. Investigação Operacional, 28(1), ciency evaluation method. European Journal of Operational 77–89. Research, 197(2), 722–730. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png SAGE Open SAGE

A Global Entrepreneurship Efficiency Benchmarking and Comparison Study based on National Systems of Entrepreneurship and Early-Stage Business: A Data Envelopment Analysis Approach

Kim, Sehoon
SAGE Open , Volume OnlineFirst: 1 – Sep 19, 2022
Download PDF
17 pages

Loading next page...
 
/lp/sage/a-global-entrepreneurship-efficiency-benchmarking-and-comparison-study-r0Scvnzfku

References (95)

Publisher
SAGE
Copyright
© The Author(s) 2022
ISSN
2158-2440
eISSN
2158-2440
DOI
10.1177/21582440221123252
Publisher site
See Article on Publisher Site

Abstract

National Systems of Entrepreneurship is defined as a nation’s resource allocation structure leading to entrepreneurial behaviors. However, the existing indicators of national framework conditions may have limitations in comparing the entrepreneurial efficiency of countries. Based on institutional theory, this paper presents a model to examine the efficiency of entrepreneurial activities stemming from the given conditions of a country and find benchmarks based on data envelopment analysis by scrutinizing inputs and outputs with static efficiency, dynamic efficiency, and strategic quadrant analysis. For this purpose, the study utilizes the Global Entrepreneurship Monitor dataset from 2015 to 2020 for 24 countries and presents the research questions regarding the differences in global entrepreneurial efficiencies, the countries for benchmarking, and the implications for entrepreneurial activities. The research implications suggest that diversifying the views on entrepreneurial efficiency may be valuable, and policymakers may focus on institutional conditions and entrepreneurial efficiency regarding the activity of early-stage businesses. Keywords data envelopment analysis, entrepreneurial efficiency, global entrepreneurship monitor, national systems of entrepreneurship, total early-stage entrepreneurial activity (Ács et al., 2014). Entrepreneurship is a complex phenome- Introduction non, and many determinants affect its occurrence. A coun- Entrepreneurs play a substantial role in a country’s economic try’s resource allocation systems are affected by the development. They are the agents of creative destruction individual’s opportunity pursuit and new business creations, (Schumpeter, 1934), the epicenter of innovation (Ács, and these activities and outcomes are influenced by the coun- Audretsch, 1988), and actors of job creation and knowledge try’s particular institutional specificities (Ács et al., 2014). spillover (e.g., Blanchflower, 2000; Parker, 2009; van Praag Then, it could be assumed that individual entrepreneurial & Versloot, 2007). Despite its positive contributions to the activities and the birth of early-stage businesses could be economy and academic recognition, still defining entrepre- viewed as a result of the national systems at a country level neurship is challenging. Various scholars have strived to (Ács et al., 2014; Inacio Junior et al., 2021; Tasnim & Afzal, make a consensus presenting different views on defining entrepreneurship (Anderson & Starnawska, 2008; Gedeon, Seoul School of Integrated Sciences and Technologies, Seodaemun-gu, 2010; Sendra-Pons et al., 2022) over time. It is an economic Seoul, Republic of Korea novelty-introducing (Schumpeter, 1934), firm-level behav- The Institute for Industrial Policy Studies (IPS), Seodaemun-gu, Seoul, ioral disposition or individual cognitive attribute for business Republic of Korea opportunity (e.g., Lumpkin & Dess, 1996; Shane & 3 Business School Lausanne, Lausanne, Chavannes, Switzerland Venkataraman, 2000), or activities such as self-employment Corresponding Author: or new business creation (i.e., Reynolds et al., 2005). Sehoon Kim, Business School Lausanne, Route de la Maladière 21, Various academic debates about entrepreneurial theories Lausanne, Chavannes 1022, Switzerland. and their academic background need further elaboration Email: sehoon.kim@bsl-lausanne.ch Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (https://creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 SAGE Open 2018). Also, the concept of National Systems of businesses), enabling efficiency comparison at the global Entrepreneurship (NSE) is central to assessing a country’s level and help to find comparable benchmarks. Benchmarking contextual conditions upon the entrepreneurial perfor- refers to exploring and utilizing other organizations’ best mance (Feldman, 2001; Sternberg, 2009), which influ- practices, services, and products to improve organizational enced the policymakers to form a framework to understand performance (Main, 1992). However, there is a risk of taking entrepreneurial activities in the overall context. NSE, a wrong benchmark that ignores NSE differences if policy- affected by the National Systems of Innovation (NSI) con- makers only focus on entrepreneurial performance, such as cept, is defined as a nation’s resource allocation system the high birth rate of start-up businesses. In other words, it and structure where individual entrepreneurial behaviors would be a reasonable approach to choose a benchmark and initiatives occur (Ács et al., 2014; Inacio Junior et al., country that produces optimum performance in a similar 2021; Tasnim & Afzal, 2018). NSE focuses on systemic social and economic environment (Lins et al., 2003). factors driving entrepreneurial activities (Ács et al., 2014), Efficiency theory can be applied when selecting a benchmark and it views the creation of businesses as the product of a with excellent output in a similar input structure. process affected by combined factors that arise from the Efficiency is defined as the relationship between input and system (Inacio Junior et al., 2021). However, referring output or evaluating the effectiveness of output factors for only to the rankings of global surveys that measure national input (Ramanathan, 2003; Thornton et al., 1982). Efficiency institutional conditions where developed economies hold is a significant concept in economics. Low efficiency could the entrepreneurial leadership position (the highest national mean poorly performing system components that hinder potential for entrepreneurship) may not help compare the overall performance (Tasnim & Afzal, 2018). The application entrepreneurial activities of countries since those rankings of the efficiency theory could help find high-performance are not designed to evaluate the simple efficiency such as units and units with similar input conditions as benchmarks. the ratio between input and output (Inacio Junior et al., Parametric and non-parametric methods can be considered 2021). Top rankers in those global surveys may not always for efficiency evaluation (Smith & Street, 2005; Stone, 2002). be efficient in generating new businesses, and they may also Data envelopment analysis (DEA) is a non-parametric statis- strengthen the belief that a national system with the highest tical technique that measures efficiency by comparing rela- ranking should be chosen as a benchmark (Bogetoft, 2012). tive inputs and outputs (Casu & Molyneux, 2003) and As such, quantifying and measuring conditions may not estimates the optimal level of output based on the mixed input be sufficient in understanding the actual performance of a elements (Smith & Street, 2005). In addition, the DEA model national system (Inzelt, 2004; Kuhlmann, 2003), and indices enables relative evaluation between DMUs (decision-making not considering the input-output efficiency may not fulfill units) by comparing the empirical productivity frontiers in a the need to gauge the country’s entrepreneurial performance similar environment and provides information on benchmarks (Edquist & Zabala-Iturriagagoitia, 2015). Due to its substan- with optimal performance. Efficiency or productivity focuses tial role in decision-making policies and nurturing socioeco- on selecting a target that achieves superior performance (e.g., nomic development, academic interests and attention have services, outcomes, or activities) under a similar environment examined the causal relationships between innovative entre- (e.g., socioeconomic factors, budget) (Shephard, 1970; Stone, preneurial activities and a country’s international competi- 2002). The environment involves national contexts, levels of tiveness (Ferreira et al., 2017). While valuable, the existing NSE, and stages of economic growth, such as GDP (Gross global entrepreneurship surveys focus on how supportive the Domestic Product) per capita (i.e., Solow, 1957), where a entrepreneurial environment is and may have limitations in country can foster entrepreneurship. considering national characteristics and specificities to eval- To date, global entrepreneurship research investigating uate entrepreneurial activities’ efficiency as the output of entrepreneurial efficiency frontiers based on the institutional national systems (Edquist & Zabala-Iturriagagoitia, 2015). factors and early-stage entrepreneurs for the multi-faceted The entrepreneurship literature has neglected to highlight efficiency and longitudinal analyses as in this paper are efficiency by focusing on advanced economies with the best somewhat limited. Some previous studies suggesting NSE framework conditions, such as the US (Inacio Junior et al., constructs (regional framework conditions) such as output 2021). Hence, examining how a nation’s entrepreneurial measures, attitude measures, and framework measures but activities could be evaluated and compared apple-to-apple without efficiency comparison (i.e., Ács et al., 2014) failed based on the “output” of the given systems would add value to examine longitudinal dynamics (e.g., Das & Kundu, 2019; to the literature. Inacio Junior et al., 2021; Tasnim & Afzal, 2018) or set mis- Consequently, this paper seeks to analyze entrepreneur- leading causal relationships between inputs and outputs for ship efficiency that is measured and calculated by consider- efficiency analysis (i.e., Faghih et al., 2021). Overall, it is ing inputs (the institutional factors that make up the national understood that global entrepreneurship literature further framework condition, such as evaluation results of mecha- requires; (a) an efficiency study based on NSE, (b) empirical nisms and processes) and the output (the entrepreneurial evidence for benchmarking, and (c) integrated efficiency activities of a country represented by the creation of new measures that would advance future studies. Kim 3 In order to reduce the research gap, this paper aims to NSE concept evolved into the “entrepreneurial ecosystem” contribute to the knowledge of country-level efficiency concept (i.e., Szerb et al., 2019), which views the regional based on the NSE concept and provide a valuable integrated ecosystem may affect different types of entrepreneurial efficiency model for benchmarking based on empirical evi- activities (Kirznerian or Schumpeterian). A healthier ecosys- dence gathered throughout the research. This paper mainly tem grants greater capacity for higher business formation, deals with the global entrepreneurship phenomenon from a while a weak ecosystem leads to innovation of entrepreneurs “systemic efficiency” viewpoint, not “the superiority of due to a lack of support policies or economic support (Szerb entrepreneurial environment.” To this end, this paper focuses et al., 2019). NSE also represents the dynamism and interac- on cross-sectional (static) and longitudinal (dynamic) effi- tion among individual attitudes, aspirations, and abilities that ciency analyses. Also, the paper aims to assess global entre- stimulate resource allocation via new business creation and preneurship competitiveness based on strategic quadrant operation (Ács et al., 2014). In a country, individuals pursue analysis to compare each country’s position with peers. For entrepreneurial opportunities and evaluate the desirability this purpose, this study utilizes the Global Entrepreneurship and feasibility of the pursuit, which is also determined by a Monitor (GEM) dataset for 6 years, from 2015 to 2020, for particular country’s institutional (contextual) factors, such 24 countries. Consequently, this paper presents the fol- as distribution of available resources, social norms, and lowing research questions; (a) What are the differences in attitudes (Ács et al., 2014). Some NSE output indicators entrepreneurial efficiencies of countries? (b) Who are the evaluate the degree of emergence for the newly born entre- benchmarking countries and regions from a static and preneurs within a given population, and the most widely dynamic efficiency perspective? (c) What are the strategic used example is GEM’s total early-stage entrepreneurial implications of the efficiency-stability analysis result? activity rate, TEA (Autio, 2007). Global entrepreneurship This study is structured as follows. Section 2 deals with indicators such as GEDI (Global Entrepreneurship and the literature review. Section 3 provides the research Development Index) and GEI (Global Entrepreneurship design and methodology. Section 4 deals with the specific Index) deliver multi-faceted information on the NSE per- result of the study. Section 5 discusses and summarizes formance at a system level by profiling each NSE. the conclusion, and Section 6 provides theoretical and However, these indicator-based approaches might be less policymaking implications. Lastly, Section 7 presents valuable in contemplating the relationships between input suggestions for future research. factors and the productivity or efficiency of analyzed units (Inacio Junior et al., 2021). Also, considering the countries with the highest values in NSE indicators as benchmarks Literature Review could lead to severe bias or failure in decision-making pro- cesses (Bogetoft, 2012). National Systems of Entrepreneurship Understanding the system performance requires a holistic National Systems of Innovation (NSI) literature comprised approach (Inzelt, 2004; Kuhlmann, 2003), and quantifying the holistic relationships toward interactive, iterative, and the national systems may not be sufficient if not followed by cumulative innovation processes at a country level (e.g., the productivity analysis considering inputs and outputs Freeman, 1987; Lundvall et al., 2002). NSI’s institutional (Edquist & Zabala-Iturriagagoitia, 2015; Inacio Junior et al., and structural focus inspired policymakers with a frame- 2021). It is reasonable to identify the national system’s per- work-based performance evaluation (Nelson, 1993) and formance based on the efficiency concept, where a unit’s made stakeholders consider the influence of regulatory inter- productivity is based on the rate of a total of outputs by a vention in national entrepreneurship (Inacio Junior et al., total of inputs (Ramanathan, 2003). Also, it contributes to the 2021). As a result, entrepreneurship literature has strived to existing literature to evaluate and review if a particular coun- study systemic context and system-level conditions of entre- try’s institutional condition leads to the high efficiency of preneurial action (Ács et al., 2014). entrepreneurial activities. This paper seeks to provide an National Systems of Entrepreneurship (NSE) concept is empirical analysis of entrepreneurial efficiency based on the about institutional resource allocation closely related to the relational assumptions among the national framework condi- entrepreneurial attitudes, abilities, and aspirations of indi- tions and individual entrepreneurial outcomes as hypothe- viduals leading to the creation and operation of new busi- sized in the NSE concept. nesses (Ács et al., 2014; Tasnim & Afzal, 2018). In summary, the NSE depicts national capacity leading to entrepreneurial Global Entrepreneurship Efficiency Literature activities and presents the dynamic, inter-correlated relation- ships among the multi-dimensional economic, social, and Different studies suggested the methodological approach for institutional frameworks (Ács et al., 2014), and also helps efficiency comparison and analysis for global entrepreneur- explain the national economic capability for potential entre- ship. Bygrave et al. (2003) examined the correlation of entre- preneurial activities (Ács, Szerb, et al. 2014; Inacio Junior preneurship from a push-pull perspective in 29 countries et al., 2021; Tasnim & Afzal, 2018). Further recently, the participating in the GEM program, finding the relationship 4 SAGE Open between opportunity-pull entrepreneurship and variables with less-supportive institutional conditions might be lim- such as informal investment, entrepreneurial capacity, and ited. If a country’s NSE efficiency is defined by the total perception of start-up opportunities. Sternberg and input-output rate (Ramanathan, 2003), the variable selection Wennekers (2005) provided empirical evidence that the role might be reconsidered (Inacio Junior et al., 2021). of entrepreneurial activity varies according to economic Consequently, providing objective logic for national entre- development stages. Also, they suggested that regional preneurial efficiency and criteria for global comparison framework conditions, such as infrastructure and policies, benchmarks is expected to contribute valuable insights into need to be considered to investigate the regional entrepre- fostering early-stage businesses. neurial context. Lafuente et al. (2016), based on the data of 63 countries from 2012, tested the knowledge spillover the- Data Envelopment Analysis (DEA) ory of entrepreneurship using DEA, highlighting the rela- tionship between NSE and knowledge spillovers leading to Data envelopment analysis (DEA) was proposed by Charnes better efficiency. Tasnim and Afzal (2018) compared coun- et al. (1978). DEA is a part of the linear planning methodol- try-level efficiency based on the NSE perspective, hypothe- ogy and can evaluate the relative efficiency of each DMU sizing that individual-level actions and country-level (decision-making unit) by applying multiple input and out- framework conditions interact. Das and Kundu (2019) com- put factors (Casu & Molyneux, 2003). DEA is a non-para- pared countrywide efficiency based on GEDI and GEM data- metric statistical technique used in various fields because it sets from 2012 to 2016 to understand small-sized enterprise can be applied even when the measurement scale of each performance. They utilized attitudinal parameters as inde- variable is different (Barros & Alves, 2004; Nguyen et al., pendent variables and adopted dependent variables, such as 2016; Wu et al., 2009). DEA evaluates units that produce the nascent entrepreneurship rate. Also, Lafuente et al. maximum (weighted sum) outputs at minimum (weighted (2021) conducted an efficiency analysis using the non-para- sum) input levels as efficient DMUs (Smith & Street, 2005; metric technique to figure out how country-level entrepre- Stone, 2002). It informs the level of improvement and neurship relates to total factor productivity using global decrease of the input variable, has the advantage of express- entrepreneurship data from 2002 to 2013. Faghih et al. ing the growth potential of the output variable numerically, (2021) investigated entrepreneurship “overall efficiency,” and is used in various research fields (e.g., Nguyen et al., assessing national entrepreneurship attitude toward the 2016; Wu et al., 2009). entrepreneurship system, based on GEM 2018 data. Inacio DEA directly compares efficient and inefficient DMUs Junior et al. (2021) conducted an efficiency analysis of the and provides information on which DMUs can be bench- GEI data to demonstrate that the global entrepreneurial rank- marked to improve efficiency. It has two models, input-ori- ings may misinterpret NSE dynamics with the impression ented and output-oriented. After fixing the output value, the that countries showing higher national framework indicators former calculates the efficiency index based on the input fac- are more productive in generating actual entrepreneurial tor values, focusing on the input reduction. The latter fixes activity. The authors claimed that there might be another per- the input factors and pays attention to how much the output spective that prioritizes the efficiency of actual entrepreneur- can be improved. In addition, DEA has CCR (Charnes- ial activities. They pointed out that lower-positioned Cooper-Rhodes) model (Charnes et al., 1978) that assumes countries in framework conditions often present higher effi- the Constant Return to Scale (CRS) and BCC (Banker- ciency, producing more entrepreneurial business opportuni- Charnes-Cooper) model (Banker et al., 1984) based on ties at the individual level. Variable Return to Scale (VRS) assumption. The previous studies can be divided into three streams; (a) Research that dealt with the level of NSEs without efficiency Technical Efficiency (TE). The first concept proposed in analysis among countries (i.e., Ács et al., 2014), (b) Research efficiency research was Technical Efficiency (TE). TE calcu- that compared efficiency among countries without longitudi- lates the efficiency index through relative comparison using nal or time-series consideration (e.g., Inacio Junior et al., input-output data and assumes the constant return to scale. 2021; Tasnim & Afzal, 2018), and (c) Research that needs Technical efficiency theory was first introduced by Far- further probe in setting input-output relationship (i.e., Faghih rell (1957) and later rediscovered by Charnes et al. (1978) et al., 2021). In Faghih et al.’s efficiency study (2021), GDP and subsequently re-labeled as CCR-efficiency under DEA is defined as the output of the overall entrepreneurial activity. (Cooper et al., 2000). The CCR model measures TE, the Under the research setting, the author consequently priori- maximum output capacity for input, assuming that the scale tizes a country with less entrepreneurial activity (less input) value is invariant. However, this model works with DMUs and higher GDP (more output) as an efficient benchmark. If operating at the optimal scale and has the disadvantage of the researchers adopt this theoretical assumption, entrepre- not distinguishing between Scale Efficiency (SE) and Pure neurship-nurturing implications for developing economies Technical Efficiency (PTE). Kim 5 Pure Technical Efficiency (PTE). The BCC model calculates Table 1. DEA/Window Analysis Formula. the efficiency assuming that production constraints occur in Item Calculation the input-output relationship. The BCC model was proposed to overcome the shortcomings of the CCR model (Banker Number of windows k–p + 1 Number of DMUs in each window p × n et al., 1984), reflecting the variability of returns to scale. It Window Width k (odd number) (k + 1)/2 is a concept that examines the change in the outputs when k (even number) {(k + 1)/2} ± 1/2 the scale extends while keeping the ratio of input factors constant. BCC considers Increasing Return to Scale and Note. k: time, p: window width, w: # of windows. Decreasing Return to Scale, yielding Pure Technical Effi- ciency of DMUs. pure technical efficiency (PTE), and scale efficiency (SE) of DMUs. Fourth, dynamic efficiency analysis covered the lon- Scale Efficiency (SE). Scale Efficiency (SE) is calculated gitudinal examination using the DEA/Window method. from TE value (measured by CCR model) divided by PTE Fifth, the efficiency-stability matrix based on quadrant anal- value (measured by BCC model). SE value of “1” means the ysis was drawn, grouping the similarly-characterized coun- optimal scale state, and if it is less than “1,” the current input tries. The overall research framework is presented below and output are not achieving scale efficiency. SE value could (see Figure 1). provide the direction of efficiency improvement by identi- fying whether the cause of the inefficiency of the DMU is purely technical or in terms of scale through the relationship Data Collection between PTE and SE. This study utilized the Global Entrepreneurship Monitor (GEM) dataset from 2015 to 2020. GEM is a comparative DEA/Window Analysis. Although DEA has been used in study tracing entrepreneurial thoughts, behaviors, and activi- various analytical studies due to its cross-sectional charac- ties having over 150,000 participants from over 50 countries teristics, it has not frequently been applied to longitudinal (GEM, 2020). The research consists of Adult Population research. This study utilized the DEA/Window analysis, Survey (APS) and National Expert Survey (NES) and pro- which complements the cross-sectional characteristics of vides primary data on the competitiveness of each country’s DEA. Through this method, it is possible to evaluate how the NSE by surveying the adult population and experts (Das & performance of the same DMU changes according to time, Kundu, 2019). The countrywide dataset from 2015 to 2020 and DMUs are analyzed and divided into several windows consisted of 335 NES and 320 APS instances. The 24 coun- of multiple periods (Charnes & Cooper, 1984). In this analy- tries consecutively participated throughout the whole period sis, individual DMUs can be evaluated by referencing other (6 years, the time frame of this research) were chosen as DMUs belonging to different periods if the primary input- DMUs, and a total of 144 data instances were finally col- output relationship does not change over time. lected for analysis. As for the DMUs, four countries were For analysis, a researcher determines the window width from Latin America and Caribbean (17%), fourteen were (p) after collecting the longitudinal data for a certain period. from Europe and North America (58%), two were from The number of windows (w) is “k-p+1.” After analyzing the Africa (8%), and four were from Asia and Oceania (17%) first (p) period, the first analytical unit (e.g., year, month) is (see Table 3). replaced by the new unit to analyze the next window. This process is repeated until the final period k (see Table 1). The efficiency trend and stability of each DMU can be analyzed Data Processing by assessing the efficiency of each window. The DEA model This study adopted a computational analytical tool to elimi- used in this study is as follows (see Table 2). nate human errors. Efficiency Measurement System (Scheel, 2000) was utilized as a reliable DEA statistical tool (Afzal & Method Lawrey, 2012). This package provides the CCR, BCC, DEA/ Windows, and Malmquist analysis. Research Design This study evaluated global entrepreneurship’s static and Theoretical Foundation and Variable Selection dynamic efficiency and explored the benchmarking coun- tries for entrepreneurship efficiency. The analytical proce- A country’s institutional factors (framework conditions) are dure is as follows. First, the data collection process chose the known to be fundamentally correlated to formal structure/the primary data for the research. Second, the pre-processing norms of regulatory bodies and cultural/social practices while stage identified the input and output variables. Third, static playing a decisive role in promoting individual entrepreneur- efficiency analysis calculated the technical efficiency (TE), ial behaviors (Bianchi et al., 2015; Boudreaux et al., 2019; 6 SAGE Open Table 2. DEA Models Used in This Study. Model Description Author CCR Constant Returns to Scale (unchanged scale) Charnes et al. (1978) Technical Efficiency Suitable for DMUs operating at optimal scale BCC Variable Returns to Scale (variable scale) Banker et al. (1984) Pure Technical Efficiency Increasing Returns to Scale (IRS) and Decreasing Returns to Scale (DRS) DEA/Window Used to analyze dynamic changes of relative efficiency of DMUs over time Charnes and Cooper (1984) Figure 1. Research design. Table 3. Regional Composition. ID Region Country # % 1 Latin America and Caribbean Brazil, Chile, Colombia, Guatemala 4 17 2 Europe and North America Croatia, Germany, Greece, Italy, Luxembourg, Netherlands, Poland, Slovak 14 58 Republic, Slovenia, Spain, Sweden, Switzerland, United Kingdom, United States 3 Africa Egypt, Morocco 2 8 4 Asia and Oceania India, Israel, South Korea, Taiwan 4 17 Total 24 100 Note. Countries consecutively participated from 2015 to 2020. n = 24. Bylund & McCaffrey, 2017; Sendra-Pons et al., 2022). in various research topics, recently widening its presence in Therefore, it is of theoretical value and interest to view the studying determinant factors leading to the success of new entrepreneurship phenomenon from an institutional theory entrepreneurial activities (e.g., Bruton et al., 2010; Peng, standpoint regarding the remarkable contextual influence on 2001). The institutional environment, such as accountability, promoting entrepreneurial activities stemming from the effectiveness, and quality of related authorities, policies, and institutional environment (e.g., Bruton et al., 2010; Sendra- support organizations, may enhance or hinder the develop- Pons et al., 2022; Sinha et al., 2019). The institutional theory ment of businesses (Sendra-Pons et al., 2022). In this regard, assumes that the regulatory, social, and cultural elements the paper hypothesizes and selects variables based on the affect organizations to promote legitimacy and be poised for assumptions from the institutional theory point of view to survival (e.g., Ahlstrom & Bruton, 2003; Fang, 2010). The build an efficiency model. Determining valid inputs and out- theory has been broadly accepted as a theoretical foundation puts is substantial in analyzing efficiency (Chan & Karim, Kim 7 Table 4. Variables. Variable Description Source Input 1 GDP per capita in US dollars World bank (2021), IMF (2021) 2 Established Business Business owners % in 18–64 population Ownership (EBO) Financial value creation for more than 42 months 3 National Expert Composite Mean value of below 11 items GEM (2020) Index (NECI) *Entrepreneurial finance One recent survey item excluded *Government policies: support and relevance (Entrepreneurial post-school education) *Government policies: taxes and bureaucracy *Government entrepreneurship programs *Entrepreneurial education at school stage *R&D transfer *Commercial and legal infrastructure *Internal market dynamics *Internal market burdens or entry regulation *Physical infrastructure *Cultural and social norms Output 4 Total early-stage New business owner % in 18–64 population GEM (2020) Entrepreneurial Activity (TEA) Adult population starting or running a new business Note. NECI: based on National Expert Survey (NES). All variables: 2015 to 2020. GDP per capita: World bank, only Taiwan data from IMF. 2012). The following variables were selected to create the institutional theory. Thus, comparing the TEA level with efficiency model. The variables were determined by the con- national framework conditions might be a legitimate way to sensus of three researchers and 2 Ph.D. holders after a peer compare national entrepreneurship efficiency. review for validity. Finally, three input variables and one out- Consequently, this study set TEA as output and analyzed put variable were selected (see Table 4). the efficiency based on the weighted averaged inputs. Output Variable Input Variable (a) Total Early-stage Entrepreneurial Activity (TEA): (a) GDP per capita: A country’s population and GDP are TEA, a part of the GEM dataset, was selected as the output absolute indicators that could help understand its socioeco- variable. GEM describes entrepreneurship activity in a nomic resources. Each country has a different institutional country through a variety of indicators. TEA is defined as framework resulting from various aspects, and one of the the proportion of those who participate in early-stage busi- significant elements is the economic development stage of ness activities among the adult population aged 18 to 64. the country (Eijdenberg et al., 2019). Some researchers GEM identifies entrepreneurs into four levels; (a) potential viewed that entrepreneurship affected economic growth entrepreneur, (b) nascent entrepreneur (less than 3 months), (e.g., Doran et al., 2018; Stoica et al., 2020), while other (c) new business (less than 42 months), and (d) established researchers presented a different way of understanding the business (more than 42 months). TEA represents the pro- causal direction between the two variables, hypothesizing portion of nascent entrepreneurs and new business owners that GDP is among the other factors that may significantly among all entrepreneurs. TEA is a valuable benchmark for affect the emergence of the nascent entrepreneurs (e.g., entrepreneurial activity and an indicator adopted by many Micozzi, 2013; Rusu & Roman, 2017). Being a country’s economies to broadly gauge the degree of newly born economic development stage, GDP could be considered a entrepreneurs across the country (Wright, 2019). The emer- legitimate input factor in an efficiency analysis model. gence of new businesses could be viewed as an output mea- Additional empirical research shows that GDP per capita sig- sure under NSE (Ács et al., 2014), and there have been nificantly impacts entrepreneurial activity (Shane, 2010). academic approaches that adopted TEA as a dependent Since the research model adopts TEA, a ratio index, as an variable to investigate its affecting factors or to compare output variable, GDP per capita, which enables countrywide entrepreneurial performance among countries (e.g., comparison, was chosen as an input variable. Except for Micozzi, 2013; Rusu & Roman, 2017) based on the Taiwan, the 2015-2020 GDPPC data were secured from the 8 SAGE Open Table 5. Variable Descriptive Data. GDP per capita (USD) NECI EBO (%) TEA (%) Mean 30,489 2.8 7.9 11.3 Median 25,120 2.7 6.8 9.2 Standard deviation 26,614 0.4 3.6 6.6 Minimum 1,606 2.1 1.5 1.9 Maximum 116,654 3.6 20.3 36.7 25th percentile 11,319 2.5 5.2 7.1 50th percentile 25,120 2.7 6.8 9.2 75th percentile 43,595 3.0 10.4 13.0 Note. n = 144 (24 countries, 2015–2020). Worldbank (2021). Separate GDPPC data for Taiwan was results. The number of input and output factors is closely obtained from IMF (2021). related to the number of DMUs, and for validity, the num- (b) Established Business Ownership (EBO): EBO refers ber of DMUs should be over twice the product of the num- to the proportion of businesses that have been in operation for ber of input and output factors (Banker et al., 1984). This over 42 months and is calculated by the proportion of estab- study chose three inputs, one output factor, and 24 DMUs, lished business owners among the adult population aged 18 to satisfying the condition. 64. GEM recognizes the EBO rate as a part of the supportive The descriptive statistics of the variables are as follows environment leading to new businesses and a health indicator (see Table 5). From the data for 24 countries from 2015 to for a country’s entrepreneurial ecosystem (GEM, 2021). After 2020, the mean of GDPPC was 30,489 USD (min: 1,606, establishing a new business, it is essential to continue and max: 116,654). For NECI, the mean value was 2.8 (min: scale it. Comparing the EBO (input, the existing and estab- 2.1, max: 3.6), and the EBO rate was 7.9% (min: 1.5, max: lished business fundamental) and TEA (output, new busi- 20.3). TEA, the output variable, was 11.3% on average nesses) can signal how the outcome of the institutional (min: 1.9, max: 36.7). environment is resilient or productive regardless of the cur- rent entrepreneurial context within the country (Wright, Efficiency Analysis 2020), highlighting the country’s capacity to create new busi- ness opportunities. It is also considered one of the factors Static efficiency analysis. An output-oriented model was affecting new entrepreneurial activities (Almodóvar- applied to all analyses in this study. As discussed previ- González et al., 2020). By comparing EBO and TEA rates, the ously, national framework conditions are inputs, and ini- viability and vitality of a new business could be measured and tial entrepreneurship activities are output. Under these translated into efficiency. assumptions, reducing the inputs for efficiency (input-ori- (c) National Expert Composite Index (NECI): GEM’s ented model) is not reasonable, and improving the output sub-item, National Expert Survey (NES), supports the NSE under given conditions would be adequate. In addition, concept by evaluating expert opinions on national entrepre- since the efficiency level cannot be ranked among the effi- neurial conditions through a scale measure. Over 36 experts cient DMUs whose index is 1 (100%), the super-efficiency in 9 fields from each country participate in evaluating 12 model (Andersen & Petersen, 1993) is applied to rank effi- items such as the economic environment, market access, cient DMUs. The output-oriented values are converted into government policies, and entrepreneurship education. This inverse numbers to make comparison easier. Static analy- study calculated NECI by averaging the results of 11 items, sis proceeds in three steps; (a) TE analysis by CCR model, excluding the recently included item, entrepreneurial post- (b) PTE analysis by BCC model, and (c) SE analysis cal- school education. GEM considers NECI as a measure to culated by TE/PTE. gauge the easiness of beginning and developing a new busi- ness (GEM, 2020). Overall, NECI summarizes the evalua- Dynamic efficiency analysis. DEA/Window analysis is a tion results of a country’s entrepreneurial framework valuable method to evaluate how the performance of the conditions into a composite score (GEM, 2020) and may same DMU changes over time. Enabling comparative evalu- help efficiently compare the processes or the mechanisms ation over time can allow researchers to focus on the lon- leading to entrepreneurial behaviors, such as TEA. As a gitudinal side of the data. In this study, dynamic efficiency result, NECI was selected as a legitimate input variable for analysis is performed to measure changes in the effective- the efficiency analysis model. ness of entrepreneurial activities. The period is 6 years The total number of inputs and outputs in DEA should (2015–2020), and the window width (p) is three based on be limited to a minimum to avoid unrealistic evaluation the formula of ((k + 1)/2) ± 1/2, and through four windows, Kim 9 changes in the efficiency were measured. In addition, by Global entrepreneurship benchmarking: SE. SE was calcu- calculating the mean, standard deviation, LDY (largest dif- lated as the second step of the static efficiency analysis. SE ference between scores in the same year), and LDP (largest value smaller than one means that scale inefficiency exists. difference between scores across the entire period) within the From the relationship between PTE and SE, a researcher can analysis period, the entrepreneurship performance based on suggest directions for efficiency improvement by identifying the efficiency-stability of each country was calculated. whether the cause of the inefficiency is a purely technical issue or a scale issue. For example, in the SE results of 2020, Strategic quadrant analysis. This study also proposes a six DMUs have PTE values greater than SE values out of 24 strategic quadrant analysis based on the stability (LDP countries (Chile, Croatia, Egypt, India, Italy, and Morocco). value) and the mean efficiency of each DMU based on the It is understood that these countries can improve efficiency dynamic efficiency results. Unlike the existing literature, through economies of scale (SE) rather than resource conver- this study adds value by including another performance axis, sion (PTE) for entrepreneurial activities (see Table 7). entrepreneurial stability, suggesting a matrix framework to The five DMUs with the highest mean SE values over the compare the relative positions with peers. whole period were Guatemala, Colombia, Chile, Sweden, and United Kingdom. On the other hand, the bottom five countries were Morocco, Croatia, Italy, Egypt, and the Results Slovak Republic. The year with the highest number of effi- cient DMUs was 2015 (6 countries, % of efficient Static Efficiency Comparison DMU = 25.0%). On the other hand, the years with the lowest Global entrepreneurship benchmarking: TE. TE analysis number of efficient DMUs were 2018 and 2020 (2 countries, compares entrepreneurship efficiency and ranks the effi- % of efficient DMU = 8.3%). ciency levels, suggesting benchmarks among reference The regional analysis results are as follows (see Table 8). DMUs. The results are as follows (see Table 6). In 2015, Latin America and Caribbean area was the highest region in five DMUs were found effective in entrepreneurship activi- TE and SE except for 2016. Africa has been consistently ties (Colombia, India, Guatemala, Chile, and Brazil), and effective in terms of PTE for 6 years. Colombia was the most efficient and referenced country Overall efficiency averaging of all countries is as follows (inversed super-efficiency θ = 1.578, referenced = 16). As of (see Figure 2). It can be seen that TE, PTE, and SE have been 2016, three countries were found efficient, with Colombia on a downward trend since 2015. being the most efficient DMU (θ = 1.311, referenced = 20). India was second (θ = 1.270, referenced = 2), and Guatemala Dynamic Efficiency Comparison was third (θ = 1.004, referenced = 2). In 2017, four DMUs (Guatemala, Israel, Egypt, and Chile) proved themselves as DEA/Window analysis. A dynamic efficiency analysis was efficient, and the most efficient was Guatemala (θ = 1.676, performed to measure the efficiency change over time. DEA/ referenced = 8). The most referenced was Chile (θ = 1.080, Window analysis, which complements the cross-sectional referenced = 15), ranked fourth in efficiency, followed by characteristics of DEA, enables longitudinal comparisons Israel (θ = 1.535, referenced = 12). among DMUs by evaluating the performance based on Regarding 2018 results, two countries were found to be smoothed average values. The analysis period is from 2015 efficient; Guatemala (θ = 1.805, referenced = 13) and to 2020, and the window width (p) is set to 3, with four Colombia (θ = 1.221, referenced = 20). In 2019, four coun- windows (2015–2017, 2016–2018, 2017–2019, and 2018– tries were efficient (Guatemala, Colombia, Chile, and India). 2020). This analysis included the mean value of 24 countries Same as the previous year, Guatemala (θ = 1.442, refer- that presents the average value of entrepreneurial efficiency enced = 2) was the most efficient. Colombia (θ = 1.433, refer- for all countries included in the same time frame, which is a enced = 15) and Chile (θ = 1.364, referenced = 15) were the window of 4 years (Mean A), and the mean value of 4 win- most referenced. Concerning the TE results in 2020, two dows that can show a smoothed efficiency of a country over countries were picked as efficient DMUs; Colombia time (Mean B). The value of 0.471 shown where the x-axis (θ = 1.913, referenced = 20) and Guatemala (θ = 1.054, (Mean A) and y-axis (Mean B) meet is the average efficiency referenced = 17). of all DMUs within the entire window period (see Table 9). The overall TE analysis results demonstrate that Guatemala Mean A suggested the average value of all DMUs’ effi- has been an efficient unit throughout the whole period (six ciency scores in the same window. The overall average of times). Four countries (Guatemala 6, Colombia 5, Chile 3, and Mean A was 0.471, and Window 2 (2016–2018) showed the India 3 times) were selected as efficient DMUs over three highest efficiency (0.523). The window efficiency slightly times (17% of the entire countries). Three countries other than improved in the Window 2 period, then showed a downtrend India belonged to Latin America and Caribbean, showing high passing Windows 3 and 4. It could be interpreted that the competitiveness in entrepreneurial activity efficiency. overall global entrepreneurship efficiency trend has declined 10 Table 6. Entrepreneurship Technical Efficiency (2015–2020). 2015 2016 2017 2018 2019 2020 CCR CCR CCR CCR CCR CCR No. DMU SE Rank # Ref. SE Rank # Ref. SE Rank # Ref. SE Rank # Ref. SE Rank # Ref. SE Rank # Ref. 1 Brazil 1.007 5 0.811 8 0.798 7 0.635 10 0.762 6 0.783 4 2 Chile 1.029 4 9 0.983 4 1.080 4 15 0.992 3 1.364 3 15 0.856 3 3 Colombia 1.578 1 16 1.311 1 20 0.951 5 1.221 2 20 1.433 2 15 1.913 1 20 4 Croatia 0.630 8 0.650 9 0.756 8 0.701 9 0.563 8 0.535 6 5 Egypt 0.585 9 0.829 7 1.144 3 4 0.818 7 0.855 5 0.538 5 6 Germany 0.224 22 0.213 24 0.297 22 0.205 22 0.287 16 0.143 23 7 Greece 0.315 17 0.225 23 0.189 24 0.234 21 0.240 19 0.278 15 8 Guatemala 1.123 3 1 1.004 3 2 1.676 1 8 1.805 1 13 1.442 1 2 1.054 2 17 9 India 1.518 2 1 1.270 2 2 0.843 6 0.930 5 1.319 4 0.453 8 10 Israel 0.693 7 0.918 6 1.535 2 12 0.927 6 0.479 10 0.358 12 11 Italy 0.249 20 0.275 20 0.256 23 0.203 23 0.118 24 0.153 22 12 Luxembourg 0.708 6 0.934 5 0.711 9 0.965 4 0.420 12 0.393 10 13 Morocco 0.325 14 0.379 16 0.521 11 0.546 12 0.730 7 0.389 11 14 Netherlands 0.220 23 0.350 17 0.409 16 0.361 16 0.255 17 0.291 14 15 Poland 0.362 12 0.490 12 0.379 17 0.158 24 0.152 23 0.100 24 16 Slovak Republic 0.400 11 0.506 11 0.510 12 0.807 8 0.540 9 0.479 7 17 Slovenia 0.322 15 0.388 15 0.379 18 0.289 18 0.242 18 0.188 20 18 South Korea 0.351 13 0.330 19 0.490 14 0.449 13 0.368 13 0.350 13 19 Spain 0.233 21 0.272 21 0.334 19 0.322 17 0.210 21 0.161 21 20 Sweden 0.317 16 0.549 10 0.504 13 0.393 14 0.326 14 0.231 17 21 Switzerland 0.219 24 0.240 22 0.323 20 0.236 20 0.233 20 0.255 16 22 Taiwan 0.270 19 0.346 18 0.305 21 0.259 19 0.191 22 0.211 19 23 United Kingdom 0.298 18 0.469 13 0.448 15 0.393 15 0.288 15 0.228 18 24 United States 0.420 10 0.445 14 0.658 10 0.605 11 0.453 11 0.443 9 Note. CCR & SE (super-efficiency): inverse number / Bold: SE (CCR = 1.000). Kim 11 Table 7. Entrepreneurship scale efficiency (2015–2020). 2015 2016 2017 2018 2019 2020 DMU TE PTE SE TE PTE SE TE PTE SE TE PTE SE TE PTE SE TE PTE SE Brazil 1.000 1.000 1.000 0.811 1.000 0.811 0.798 0.819 0.974 0.635 0.651 0.975 0.762 0.777 0.981 0.783 0.783 1.000 Chile 1.000 1.000 1.000 0.983 0.989 0.993 1.000 1.000 1.000 0.992 1.000 0.992 1.000 1.000 1.000 0.856 0.926 0.924 Colombia 1.000 1.000 1.000 1.000 1.000 1.000 0.951 0.965 0.985 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Croatia 0.630 1.000 0.630 0.650 1.000 0.650 0.756 1.000 0.756 0.701 1.000 0.701 0.563 1.000 0.563 0.535 1.000 0.535 Egypt 0.585 1.000 0.585 0.829 1.000 0.829 1.000 1.000 1.000 0.818 1.000 0.818 0.855 1.000 0.855 0.538 1.000 0.538 Germany 0.224 0.233 0.964 0.213 0.216 0.990 0.297 0.303 0.979 0.205 0.216 0.950 0.287 0.311 0.921 0.143 0.154 0.924 Greece 0.315 0.330 0.956 0.225 0.256 0.880 0.189 0.194 0.974 0.234 0.236 0.993 0.240 0.256 0.940 0.278 0.281 0.987 Guatemala 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 India 1.000 1.000 1.000 1.000 1.000 1.000 0.843 1.000 0.843 0.930 1.000 0.930 1.000 1.000 1.000 0.453 1.000 0.453 Israel 0.693 0.810 0.856 0.918 1.000 0.918 1.000 1.000 1.000 0.927 1.000 0.927 0.479 0.508 0.941 0.358 0.434 0.825 Italy 0.249 0.332 0.752 0.275 0.341 0.805 0.256 0.277 0.923 0.203 0.214 0.951 0.118 0.120 0.982 0.153 1.000 0.153 Luxembourg 0.708 0.942 0.751 0.934 1.000 0.934 0.711 0.711 1.000 0.965 1.000 0.965 0.420 0.440 0.954 0.393 0.560 0.702 Morocco 0.325 1.000 0.325 0.379 1.000 0.379 0.521 1.000 0.521 0.546 1.000 0.546 0.730 1.000 0.730 0.389 1.000 0.389 Netherlands 0.220 0.278 0.793 0.350 0.401 0.873 0.409 0.458 0.894 0.361 0.447 0.807 0.255 0.283 0.899 0.291 0.370 0.786 Poland 0.362 0.392 0.922 0.490 0.508 0.963 0.379 0.382 0.992 0.158 0.189 0.837 0.152 0.157 0.971 0.100 0.101 0.990 Slovak Republic 0.400 0.413 0.969 0.506 0.555 0.912 0.510 0.524 0.975 0.807 0.984 0.819 0.540 0.915 0.590 0.479 0.620 0.772 Slovenia 0.322 0.359 0.897 0.388 0.406 0.955 0.379 0.379 0.999 0.289 0.294 0.982 0.242 0.245 0.986 0.188 0.193 0.974 South Korea 0.351 0.378 0.929 0.330 0.336 0.980 0.490 0.532 0.921 0.449 0.535 0.841 0.368 0.406 0.906 0.350 0.418 0.838 Spain 0.233 0.236 0.987 0.272 0.282 0.965 0.334 0.361 0.924 0.322 0.322 0.997 0.210 0.230 0.914 0.161 0.167 0.962 Sweden 0.317 0.317 1.000 0.549 0.580 0.945 0.504 0.510 0.987 0.393 0.397 0.991 0.326 0.350 0.929 0.231 0.235 0.986 Switzerland 0.219 0.282 0.777 0.240 0.299 0.802 0.323 0.353 0.915 0.236 0.269 0.876 0.233 0.267 0.871 0.255 0.296 0.863 Taiwan 0.270 0.282 0.958 0.346 0.348 0.993 0.305 0.348 0.876 0.259 0.345 0.750 0.191 0.229 0.834 0.211 0.270 0.783 United Kingdom 0.298 0.303 0.986 0.469 0.485 0.966 0.448 0.455 0.986 0.393 0.393 0.999 0.288 0.298 0.964 0.228 0.251 0.910 United States 0.420 0.477 0.880 0.445 0.460 0.967 0.658 0.670 0.982 0.605 0.652 0.929 0.453 0.475 0.954 0.443 0.495 0.895 Mean 0.506 0.598 0.872 0.567 0.644 0.896 0.586 0.635 0.934 0.559 0.631 0.899 0.488 0.553 0.904 0.409 0.565 0.799 No. efficient DMU 5 8 6 3 9 3 4 7 5 2 9 2 4 7 4 2 7 2 % of efficient DMU 20.8 33.3 25.0 12.5 37.5 12.5 16.7 29.2 20.8 8.3 37.5 8.3 16.7 29.2 16.7 8.3 29.2 8.3 Table 8. Entrepreneurship scale efficiency by region (2015–2020). 2015 2016 2017 2018 2019 2020 Region TE PTE SE TE PTE SE TE PTE SE TE PTE SE TE PTE SE TE PTE SE Latin America and Caribbean 1.000 1.000 1.000 0.948 0.997 0.951 0.937 0.946 0.990 0.907 0.913 0.992 0.940 0.944 0.995 0.910 0.927 0.981 Europe and North America 0.351 0.421 0.876 0.429 0.485 0.900 0.439 0.470 0.949 0.419 0.472 0.914 0.309 0.382 0.888 0.277 0.409 0.817 Africa 0.455 1.000 0.455 0.604 1.000 0.604 0.760 1.000 0.760 0.682 1.000 0.682 0.793 1.000 0.793 0.464 1.000 0.464 Asia and Oceania 0.579 0.617 0.936 0.648 0.671 0.973 0.660 0.720 0.910 0.641 0.720 0.862 0.509 0.536 0.920 0.343 0.530 0.725 Note. Efficiency: mean value of 24 DMUs. since the Window 2 period (2016–2018), and the dynamics, drivers, and relevant contextual factors that soared the global entrepreneurial efficiency during this period may require fur- ther research. Mean B can be interpreted as representing the continued overall efficiency of each DMU over 6 years. According to the result, Latin America and Caribbean DMUs, such as Chile, Colombia, and Guatemala, were among the top three countries. Four additional indicators were calculated based on the above results (see Table 10). First, the four windows’ average value and standard deviation were calculated. Second, LDY (largest difference between scores in the same year) means the maximum value among the differences in efficiency values for Figure 2. Entrepreneurship efficiency: TE, PTE, SE (2015–2020). each country in the same year. Third, LDP (largest difference 12 SAGE Open Table 9. DEA/Window Analysis. Window DMU (1) 2015–2017 (2) 2016–2018 (3) 2017–2019 (4) 2018–2020 Mean (B) Brazil 0.806 0.693 0.673 0.690 0.716 Chile 0.892 0.909 0.826 0.858 0.871 Colombia 0.893 0.898 0.843 0.879 0.878 Croatia 0.517 0.622 0.492 0.487 0.530 Egypt 0.799 0.834 0.846 0.655 0.784 Germany 0.198 0.221 0.214 0.182 0.204 Greece 0.223 0.199 0.201 0.229 0.213 Guatemala 0.934 0.924 0.939 0.964 0.940 India 0.914 0.897 0.875 0.770 0.864 Israel 0.743 0.868 0.600 0.435 0.661 Italy 0.210 0.224 0.149 0.134 0.179 Luxembourg 0.666 0.755 0.517 0.446 0.596 Morocco 0.383 0.432 0.504 0.434 0.438 Netherlands 0.269 0.340 0.270 0.268 0.287 Poland 0.358 0.308 0.183 0.126 0.244 Slovak Republic 0.414 0.556 0.485 0.465 0.480 Slovenia 0.303 0.322 0.238 0.201 0.266 South Korea 0.337 0.382 0.364 0.349 0.358 Spain 0.226 0.280 0.222 0.175 0.226 Sweden 0.367 0.438 0.307 0.252 0.341 Switzerland 0.231 0.240 0.213 0.221 0.226 Taiwan 0.264 0.277 0.207 0.199 0.237 United Kingdom 0.323 0.401 0.294 0.249 0.317 United States 0.438 0.528 0.456 0.418 0.460 Mean (A) 0.488 0.523 0.455 0.420 0.471 Note. Countrywide mean of each window’s CCR-output oriented model value, inversed. between scores across the entire period) is the difference need to balance their framework conditions to pursue effi- between maximum and minimum scores of a country within ciency and stability. Seven countries are in Quadrant 2, and the entire period. Through these indicators, the efficiency-sta- twelve are in Quadrant 3. Three countries (Guatemala, bility of entrepreneurial performance can be identified. Brazil, and Croatia) located in Quadrant 1 with higher aver- Guatemala showed the highest efficiency (0.940), and age entrepreneurial efficiency and stability are considered Italy had the lowest value (0.179). The DMU with the lowest benchmarks for DMUs in other quadrants. DMUs in inter-window standard deviation was Switzerland Quadrant 1 are considered to have shown higher-than-aver- (SD = 0.010), and Israel showed the largest value (SD = 0.162). age entrepreneurial efficiency continuously regarding their Greece was the DMU with the lowest LDY (0.028), which framework conditions. The two countries included in showed stable entrepreneurship performance. The Slovak Quadrant 4 display low efficiency and low stability. These Republic indicated the largest LDY value (0.300). DMUs countries may need attention and efforts at the national level with low LDP values were Switzerland (0.076) and Greece to promote and foster entrepreneurship activities. (0.113), which showed the slightest change in efficiency dur- ing the entire period. On the other hand, Israel showed the Conclusion most significant change (0.642). This paper aimed to present a diverse way to understand entrepreneurial activities as the outcome of institutional con- Efficiency-Stability Model ditions and benchmark for country-level efficiency, based on A strategic quadrant matrix analysis presented efficiency- the NSE concept from a “systemic efficiency” viewpoint. As stability axes based on dynamic efficiency results (see Figure a result, this paper presented research questions regarding; 3). A model with four quadrants consisting of the X-axis (a) differences in entrepreneurial efficiencies at the global (Mean, average = 0.471) and Y-axis (LDP, average = 0.266) level, (b) the countries for benchmarking, and c) the implica- was created. DMUs belonging to Quadrant 2 (High effi- tions for entrepreneurial activities. The study conducted ciency–Low stability) and 3 (Low efficiency–High stability) static/dynamic efficiency analysis and strategic quadrant Kim 13 Table 10. Dynamic Efficiency Analysis. Measure Rank DMU Mean SD LDY LDP Mean SD LDY LDP Brazil 0.716 0.053 0.104 0.246 6 16 12 13 Chile 0.871 0.032 0.205 0.306 3 8 19 17 Colombia 0.878 0.021 0.288 0.323 2 6 22 18 Croatia 0.530 0.054 0.266 0.266 9 17 21 15 Egypt 0.784 0.076 0.125 0.453 5 21 14 22 Germany 0.204 0.015 0.064 0.145 23 4 7 5 Greece 0.213 0.013 0.028 0.113 22 2 1 2 Guatemala 0.940 0.015 0.080 0.149 1 3 10 7 India 0.864 0.056 0.055 0.565 4 19 4 23 Israel 0.661 0.162 0.296 0.642 7 24 23 24 Italy 0.179 0.038 0.058 0.148 24 11 6 6 Luxembourg 0.596 0.121 0.255 0.423 8 23 20 21 Morocco 0.438 0.043 0.158 0.279 12 13 17 16 Netherlands 0.287 0.031 0.078 0.139 16 7 9 4 Poland 0.244 0.093 0.058 0.378 18 22 5 19 Slovak Republic 0.480 0.051 0.300 0.398 10 15 24 20 Slovenia 0.266 0.049 0.090 0.187 17 14 11 9 South Korea 0.358 0.017 0.076 0.191 13 5 8 10 Spain 0.226 0.037 0.116 0.154 21 10 13 8 Sweden 0.341 0.069 0.144 0.250 14 20 16 14 Switzerland 0.226 0.010 0.040 0.076 20 1 2 1 Taiwan 0.237 0.034 0.048 0.137 19 9 3 3 United Kingdom 0.317 0.055 0.128 0.210 15 18 15 12 United States 0.460 0.041 0.186 0.201 11 12 18 11 Note. SD: Standard Deviation (population); LDY: Largest Difference between scores in the same Year; LDP: Largest Difference between scores across the entire Period. Efficiency (CCR) mean: 0.471, Stability (LDP) mean: 0.266. analysis for 24 countries throughout the 2015 to 2020 period. national framework conditions (NECI), and (c) the vitality of The results are summarized as follows. existing businesses (EBO). TEA was selected as the output From TE results of static efficiency analysis, Guatemala variable. Different from the conclusions of previous litera- was an efficient DMU with outstanding input-output com- ture that centered on the superiority of regional framework petitiveness regarding regional framework conditions among conditions (i.e., Ács et al., 2014) or that included the degree peer countries. Additionally, from the SE point of view, it of economic development stage (GDP) as an output variable was confirmed that the DMUs - such as Guatemala, (i.e., Das & Kundu, 2019), this study highlighted the output Colombia, and Chile - in Latin America and Caribbean coun- of actual entrepreneurial activity stemming from the given tries showed higher efficiency, leading to the birth of nascent conditions. The US, having been considered to have the and new entrepreneurs. The dynamic efficiency analysis highest level of NSE already, also proved itself as a produc- results also revealed that the efficiency of countries in this tive DMU in the strategic quadrant analysis in this study. region, such as Guatemala, was outstanding. The setting of However, in terms of efficiency, which is output versus input, national direction or priority in policymaking might differ by it could be understood that developing economies have country when weighing the two axes provided by this study higher efficiency than advanced economies with better (efficiency or stability). However, when viewed from a framework conditions. global entrepreneurship perspective, the benchmarking focus The higher efficiency in Latin America and Caribbean could be on efficiency first, prioritizing the activity of newly countries might be attributed to relative resource constraints, born businesses. Then the strategic priority could be thought limited access to business, financial infrastructure, or the of in the order of Quadrants 1, 2, 3, and 4. This research was inefficiency of the formal labor market (GEM, 2019). based on the theoretical assumptions of the institutional the- However, at the same time, considering how to foster initial ory and determined input variables such as (a) the economic entrepreneurial activities as a significant axis of economic development stage of a country (GDP per capita), (b) overall development despite the limitations of given national 14 SAGE Open Figure 3. Entrepreneurship efficiency-stability model. systems conditions would provide academic and policymak- enhanced the theoretical value of NSE by synthesizing the ing insight on finding proper benchmarks. efficiency theory to examine what kind of entrepreneurial result the environmental and contextual inputs have created. The study contributes to the literature by expanding and Implications diversifying the views on entrepreneurial efficiency by This study suggests several theoretical implications. First, highlighting TEA, the less studied output under NSE the paper synthesized the National Systems of assumptions, as a dependent variable that may help establish Entrepreneurship concept and efficiency theory to provide a a feasible efficiency comparison model based on DEA. novel way of assessing the national entrepreneurial activi- Second, a multi-faceted evaluation model that could com- ties based on the institutional theory. Having excellent infra- pare global entrepreneurship efficiency was presented. This structure for entrepreneurship and making as many research suggested a novel, integrated methodology using early-stage businesses as possible are two different issues. realistic input-output variables. This approach applies to This study does not merely compare national NSEs but countries at various stages of economic development looking advances a perspective that might demonstrate the causal for a similar-level benchmark. This study revealed three relationship of a national entrepreneurial system. For exam- evaluation methodologies to consider for an overall picture; ple, Ács, Autio, et al (2014) presented GEDI as a valuable (a) static analysis, (b) dynamic analysis, and (c) strategic framework for NSE. The indicators under the concept depict quadrant analysis. Few studies on global entrepreneurship how a nation’s framework conditions are prepared and brought intensive evaluation perspectives to date to identify might be adequately viewed as a process rather than an out- the comparable positions of DMUs in a global landscape put. It would be challenging to consider the evaluation based on an efficiency evaluation. In particular, the strategic results of the institutional contexts as a country’s “systemic quadrant analysis framework presented evaluation axes, effi- efficiency.” However, this paper, standing on a simple effi- ciency-stability, presenting two essential perspectives that ciency relationship between major factors, has further should be considered for entrepreneurship policymaking. Kim 15 The findings of this study suggested implications for resource based on the input-output efficiency) may still exist. Future management and the significance of measuring efficiency scholars may want to consider a more sophisticated research from multiple perspectives. design using other models. The policymaking implications are as follows. This study provides a valuable reference for entrepreneurship policy Declaration of Conflicting Interests planning and implementation. Setting a reasonable bench- The author declared no potential conflicts of interest with respect to mark is essential, and only evaluating a benchmark’s frame- the research, authorship, and/or publication of this article. work conditions would hinder the policy’s effectiveness. Benchmarking refers to exploring and utilizing other organi- Funding zations’ excellent practices, services, and products (Main, The author received no financial support for the research, author- 1992). Selecting a country with excellent entrepreneurial ship, and/or publication of this article. efficiency in a similar socioeconomic environment is reason- able considering national differences in institutional contexts ORCID iD such as framework conditions (Lins et al., 2003). Unlike previous literature, this study suggested a dynamic Sehoon Kim https://orcid.org/0000-0002-6345-7433 efficiency analysis (longitudinal) methodology and a cross- sectional (static) perspective to find a benchmark that shows References outstanding output in a similar input structure. In addition, Ács, Z. J., & Audretsch, D. B. (1988). Innovation in large and small this study presented the global entrepreneurial frontier by firms: An empirical analysis. The American Economic Review, including multiple aspects of entrepreneurial efficiency with 78(4), 678–690. regional analyses. Entrepreneurship-supporting entities and Ács, Z. J., Autio, E., & Szerb, L. (2014). National systems of entre- policymakers in each country will be able to get insight into preneurship: Measurement issues and policy implications. Research Policy, 43(3), 476–494. improving their national infrastructure, which would eventu- Ács, Z. J., Szerb, L., & Autio, E. (2014). Global entrepreneur- ally lead to the birth of initial stage entrepreneurs based on ship and development index 2014. CreateSpace Independent the findings presented in this study. Publishing Platform. Afzal, M. N. I., & Lawrey, R. (2012). Knowledge-based econ- Limitations and Future Research omy (KBE) frameworks and empirical investigation of KBE input-output indicators for ASEAN. International Journal of This study has several limitations. First, the judgment on Economics and Finance, 4(9), 13–22. entrepreneurial efficiency may differ depending on the Ahlstrom, D., & Bruton, G. D. (2003). An institutional perspective viewpoint. However, critical views see entrepreneurship as on the role of culture in shaping strategic actions by technol- a national phenomenon beyond mere new business creation ogy-focused entrepreneurial firms in China. Entrepreneurship rate (Ács et al., 2014; Lafuente et al., 2021). Also, this paper Theory and Practice, 26(4), 53–68. Almodóvar-González, M., Fernández-Portillo, A., & Díaz-Casero, did not clarify the motivation for early-stage entrepreneurs J. C. (2020). Entrepreneurial activity and economic growth. A in the research design. Studying decisive factors for neces- multi-country analysis. European Research on Management sity-driven and improvement-driven entrepreneurship and Business Economics, 26(1), 9–17. (GEM, 2021) may interest researchers. Second, there may Andersen, P., & Petersen, N. C. (1993). A procedure for ranking exist more substantial factors that influence early-stage efficient units in data envelopment analysis. Management entrepreneurs. The selection of input factors is crucial in Science, 39(10), 1261–1264. creating a DEA model (Soares de Mello et al., 2008), which Anderson, A. R., & Starnawska, M. (2008). Research Practices affects the overall assessment. Future researchers could in Entrepreneurship: Problems of definition, description and apply hypotheses or assumptions different from those meaning. International Journal of E-Entrepreneurship and applied to this study and suggest a new model. Third, differ- Innovation, 9(4), 221–230. ences in the socio-cultural environment could be consid- Autio, E. (2007). GEM 2007 Report on High-Growth Entrepreneurship. ered. Cultural factors may affect businesses and Banker, R. D., Charnes, A., & Cooper, W. W. (1984). Some models entrepreneurial activities (i.e., Hofstede, 1984). Studying for estimating technical and scale inefficiencies in data envel- these relationships may also contribute to advancing entre- opment Analysis. Management Science, 30(9), 1078–1092. preneurship literature. Fourth, this study did not fully con- Barros, C. P., & Alves, F. P. (2004). Productivity in the tourism sider the economic differences among countries. Even industry. International Advances in Economic Research, 10(3), between two countries having similar GDP levels, the compo- 215–225. sition of the working-age population or the economy’s job- Bianchi, C. G., Borini, F. M., & Ogasavara, M. H. (2015). Creative creating abilities may differ. Alternatively, in an economy industry internationalisation: The impact of regulative, norma- where decent jobs are continuously provided, the desire to tive and cognitive institutions in developed and developing start a business may be limited. Lastly, the shortcomings of economies. Journal for Global Business Advancement, 8(4), DEA modeling (oversimplifies the complicated phenomenon 451–468. 16 SAGE Open Blanchflower, D. G. (2000). Self-employment in OECD countries. Feldman, M. P. (2001). The Entrepreneurial Event Revisited: Firm Labour Economics, 7(5), 471–505. formation in a Regional Context. Industrial and Corporate Bogetoft, P. (2012). Performance benchmarking: Measuring and Change, 10(4), 861–891. managing performance. Springer. Ferreira, J. J., Fernandes, C. I., & Ratten, V. (2017). Entrepreneurship, Boudreaux, C. J., Nikolaev, B. N., & Klein, P. (2019). Socio- innovation and competitiveness: What is the connection? cognitive traits and entrepreneurship: The moderating role of International Journal of Business and Globalisation, 18(1), economic institutions. Journal of Business Venturing, 34(1), 73–95. 178–196. Freeman, C. (1987). Technology Policy and Economic Performance: Bruton, G. D., Ahlstrom, D., & Li, H. (2010). Institutional theory Lessons from Japan. Pinter Publishers. and entrepreneurship: Where are We now and where do we Gedeon, S. (2010). What is entrepreneurship? Entrepreneurial need to move in the future? Entrepreneurship Theory and Practice Review, 1(3), 16–35. Practice, 34(3), 421–440. GEM. (2019). How policymakers in Guatemala can foster enter- Bygrave, W., Hay, M., Ng, E., & Reynolds, P. (2003). Executive preneurship. https://www.gemconsortium.org/news/how-poli- forum: A study of informal investing in 29 nations composing cymakers-in-guatemala-can-foster-entrepreneurship the Global Entrepreneurship Monitor. Venture Capital, 5(2), GEM. (2020). GEM 2019 / 2020 Global Report. https://www.gem- 101–116. consortium.org/report Bylund, P. L., & McCaffrey, M. (2017). A theory of entrepre- GEM. (2021). GEM 2020 / 2021 Global Report. https://www.gem- neurship and institutional uncertainty. Journal of Business consortium.org/reports/latest-global-report Venturing, 32(5), 461–475. Hofstede, G. (1984). Cultural dimensions in management and plan- Casu, B., & Molyneux, P. (2003). A comparative study of efficiency ning. Asia Pacific Journal of Management, 1(2), 81–99. in European banking. Applied Economics, 35(17), 1865–1876. IMF. (2021). GDP per capita, current prices. https://www.imf.org/ Chan, S.-G., & Karim, M. (2012). Public spending efficiency and external/datamapper/NGDPDPC@WEO/AZE political and economic factors: Evidence from selected East Inacio Junior, E., Dionisio, E. A., Fischer, B. B., Li, Y., & Meissner, Asian countries. Economic Annals, 57(193), 7–23. D. (2021). The global entrepreneurship index as a benchmark- Charnes, A., & Cooper, W. W. (1984). Preface to topics in data ing tool? Criticisms from an efficiency perspective. Journal of envelopment analysis. Annals of Operations Research, 2(1), Intellectual Capital, 22(1), 190–212. 59–94. Inzelt, A. (2004). The evolution of university–industry–govern- Charnes, A., Cooper, W. W., & Rhodes, E. (1978). Measuring ment relationships during transition. Research Policy, 33(6-7), the efficiency of decision making units. European Journal of 975–995. Operational Research, 2(6), 429–444. Kuhlmann, S. (2003). Evaluation of research and innovation poli- Cooper, W. W., Seiford, L. M., & Tone, K. (2000). Data cies: A discussion of trends with examples from Germany. Envelopment Analysis: A comprehensive text with models, International Journal of Technology Management, 26(2/3/4), applications, and references. Kluwer. 131–149. Cs, Z. J., Szerb, L., & Autio, E. (2014). Global Entrepreneurship Lafuente, E., Acs, Z. J., Sanders, M., & Szerb, L. (2021). Correction and Development Index 2014. CreateSpace Independent to: The global technology frontier: Productivity growth and the Publishing Platform. relevance of Kirznerian and Schumpeterian entrepreneurship. Das, S., & Kundu, A. (2019). Benchmarking a country for effi- Small Business Economics, 56(4), 1783–2178. ciency improvement: A DEA-based approach. Journal of Lafuente, E., Szerb, L., & Acs, Z. J. (2016). Country level efficiency Global Entrepreneurship Research, 9(1), 49. and national systems of entrepreneurship: A data envelopment Doran, J., McCarthy, N., & O’Connor, M. (2018). The role of analysis approach. The Journal of Technology Transfer, 41(6), entrepreneurship in stimulating economic growth in developed 1260–1283. and developing countries. Cogent Economics & Finance, 6(1), Lins, M. P. E., Gomes, E. G., Soares de Mello, J. C. C. B., & Soares 1442093. de Mello, A. J. R. (2003). Olympic ranking based on a zero Edquist, C., & Zabala-Iturriagagoitia, J. M. (2015). The Innovation sum gains DEA model. European Journal of Operational Union scoreboard is flawed: The case of Sweden – not being Research, 148(2), 312–322. the innovation leader of the EU. Research and Competences in Lumpkin, G. T., & Dess, G. G. (1996). Clarifying the entrepre- the Learning Economy, 16, 1–15. neurial orientation construct and linking it to performance. Eijdenberg, E. L., Thompson, N. A., Verduijn, K., & Essers, C. Academy of Management Review, 21(1), 135–172. (2019). Entrepreneurial activities in a developing country: Lundvall, B.-Å., Johnson, B., Andersen, E. S., & Dalum, B. (2002). An institutional theory perspective. International Journal of National systems of production, innovation and competence Entrepreneurial Behaviour & Research, 25(3), 414–432. building. Research Policy, 31(2), 213–231. Faghih, N., Bonyadi, E., & Sarreshtehdari, L. (2021). Assessing the Main, J. (1992). How to steal the best ideas around. Fortune, entrepreneurship process: An application of the data envelop- 126(8), 102–106. ment analysis. Journal of Global Entrepreneurship Research. Micozzi, A. (2013). Factors affecting the entrepreneurial dynamics Fang, T. (2010). Asian management research needs more self- in Italy: A comparison across European countries. Economia confidence: Reflection on Hofstede (2007) and beyond. Asia Marche Journal of Applied Economics, 32(1), 79–95. Pacific Journal of Management, 27(1), 155–170. Nelson, R. R. (1993). National innovation systems: A comparative Farrell, M. J. (1957). The measurement of productive efficiency. analysis. Oxford University Press. Journal of the Royal Statistical Society Series A (General), Nguyen, H.-O., Nguyen, H.-V., Chang, Y.-T., Chin, A. T. H., 120(3), 253–281. & Tongzon, J. (2016). Measuring port efficiency using Kim 17 bootstrapped DEA: The case of Vietnamese ports. Maritime Solow, R. M. (1957). Technical change and the aggregate produc- Policy & Management, 43(5), 644–659. tion function. The Review of Economics and Statistics, 39, Parker, S. C. (2009). The economics of entrepreneurship. Cambridge 312–320. University Press. Sternberg, R. (2009). Regional dimensions of entrepreneurship. Peng, M. W. (2001). How entrepreneurs create wealth in transition Foundations and Trends in Entrepreneurship, 5(4), 211–340. economies. The Academy of Management Perspectives, 15(1), Sternberg, R., & Wennekers, S. (2005). Determinants and effects of 95–108. new business creation using Global Entrepreneurship Monitor Ramanathan, R. (2003). An introduction to data envelopment anal- Data. Small Business Economics, 24(3), 193–203. ysis: A tool for performance measurement (1st ed.). SAGE Stoica, O., Roman, A., & Rusu, V. D. (2020). The nexus between Publications. entrepreneurship and economic growth: A comparative analy- Reynolds, P., Bosma, N., Autio, E., Hunt, S., De Bono, N., Servais, sis on groups of countries. Sustainability, 12(3), 1186. I., Lopez-Garcia, P., & Chin, N. (2005). Global entrepreneur- Stone, M. (2002). How not to measure the efficiency of public ser- ship monitor: Data collection design and implementation 1998- vices (and how one might). Journal of the Royal Statistical 2003. Small Business Economics, 24(3), 205–231. Society Series A (General), 165(3), 405–434. Rusu, V., & Roman, A. (2017). Entrepreneurial activity in the EU: Szerb, L., Lafuente, E., Horváth, K., & Páger, B. (2019). The rel- An empirical evaluation of its determinants. Sustainability, evance of quantity and quality entrepreneurship for regional 9(10), 1679. performance: The moderating role of the entrepreneurial eco- Scheel, H. (2000). EMS: Efficiency measurement system. http:// system. Regional Studies, 53(9), 1308–1320. www.holger-scheel.de/ems/ Tasnim, N., & Afzal, M. N. I. (2018). An empirical investiga- Schumpeter, J. A. (1934). The theory of economic development: An tion of country level efficiency and national systems of inquiry into profits, capital, credit, interest, and the business entrepreneurship using data envelopment analysis (DEA) cycle. Harvard University Press. and the TOBIT model. Journal of Global Entrepreneurship Sendra-Pons, P., Comeig, I., & Mas-Tur, A. (2022). Institutional factors Research, 8(1), 37. affecting entrepreneurship: A QCA analysis. European Research Thornton, C. V. D., Greiner, J. M., Hatry, H. P., Koss, M. P., Millar, on Management and Business Economics, 28(3), 100187. A. P., & Woodward, J. P. (1982). Productivity and motivation: Shane, S., & Venkataraman, S. (2000). The promise of entre- A review of state and local government Initiatives. Journal of preneurship as a field of research. Academy of Management Policy Analysis and Management, 1, 570. Review, 25(1), 217–226. van Praag, C. M., & Versloot, P. H. (2007). What is the value of Shane, S. A. (2010). The illusion of entrepreneurship: The costly entrepreneurship? A review of recent research. Small Business myths that entrepreneurs, investors and policymakers live by. Economics, 29(4), 351–382. Yale University Press. Worldbank. (2021). World development indicators. https://data- Shephard, R. W. (1970). The theory of cost and production func- bank.worldbank.org/reports.aspx?source=2&series=NY.GDP. tions. Princeton University Press. PCAP.CD&country Sinha, S., Singh, S., Mukunda Das, V., & Sharma, A. (2019). A Wright, F. (2019). Spotlighting trends in early-stage entrepre- framework for linking entrepreneurial ecosystem with institu- neurial activity. GEM. https://www.gemconsortium.org/ tional factors: A modified total interpretive structural modelling news/Spotlighting%20Trends%20in%20Early-Stage%20 approach. Journal for Global Business Advancement, 12(3), Entrepreneurial%20Activity 382–404. Wright, F. (2020). How Colombia’s recent entrepreneurial success Smith, P. C., & Street, A. (2005). Measuring the efficiency of public can continue despite Coronavirus. GEM. https://www.gem- services: The limits of analysis. Journal of the Royal Statistical consortium.org/news/how-colombia%E2%80%99s-recent- Society Series A (General), 168(2), 401–417. entrepreneurial-success-can-continue-despite-coronavirus Soares de Mello, J. C. C. B., Meza, L. A., & Silva, B. B. D. (2008). Wu, J., Liang, L., & Yang, F. (2009). Achievement and bench- Some rankings for the Athens Olympic Games using DEA marking of countries at the Summer Olympics using cross effi- models with a constant input. Investigação Operacional, 28(1), ciency evaluation method. European Journal of Operational 77–89. Research, 197(2), 722–730.

Journal

SAGE Open – SAGE

Published: Sep 19, 2022

Keywords: data envelopment analysis; entrepreneurial efficiency; global entrepreneurship monitor; national systems of entrepreneurship; total early-stage entrepreneurial activity

There are no references for this article.