Introduction - Competence Centre on Composite Indicators ...



Entrepreneurship and Growth: A Structural Equation Model and a Composite Index Cesare A.F. RiilloSTATEC-Research, Centre Administratif Pierre Werner, 13 rue Erasme - B.P.304, L-2013,Luxembourg. Cesare.riilo@statec.etat.lu15/04/2019Entrepreneurship is a multifaceted phenomenon that takes many forms such as business creation and innovative activities within a company. A synthetic measure of entrepreneurship that captures the different forms of entrepreneurship is needed for cross-country comparisons. This paper proposes a formative Structural Equation Model (SEM) to construct a synthetic indicator that summarizes the four entrepreneurship indicators of the Global Entrepreneurship Monitor (GEM) using 2017 data on 54 GEM countries. Compared with traditional GEM measures such as Total Entrepreneurship Activity, the proposed composite index offers a more comprehensive view of country entrepreneurship activity and it is able to captures cross-country variation in GDP growth more effectively than traditional measures (TEA). A sensitivity analysis illustrate uncertainty about the estimates. Main results are robust to lagged measures of entrepreneurship, non-linear relationship and different development stages of development. Future model developments are suggested.Keywords: Entrepreneurship, GEM, Growth, Composite index, structural equation modelsIntroductionThe purpose of this paper is to simultaneously develop a composite index of entrepreneurship activity and test whether it plays a role in explaining economic growth. Developing sound and comprehensive metrics of entrepreneurship is a difficult task for various reasons related to (i) the complexity and multidimensionality of the concept; (ii) the lack of consensus on available indicators; (iii) limited agreement on an operational definition; (iv) the limited availability of data across countries and over time; and finally, (v) disagreement regarding the role of entrepreneurship across stages of economic development. Given the critical role that measures play in evidence-based policy-making, the debate on how to adequately measure countries’ entrepreneurship is, unsurprisingly, high. This paper argues that Structural Equation Modelling (SEM) offers a suitable methodological framework for measuring entrepreneurship. In this framework, I treat entrepreneurship as a latent (multidimensional) phenomenon that is measured through a number of observable indicators describing different dimensions of such a latent construct. Additionally, the latent variable methodology is also able to simultaneously test the empirical association of the estimated entrepreneurship measure with economic outcomes. By applying the proposed methodology to countries participating in GEM 2017, the paper goes beyond the measurement of multidimensional entrepreneurship, by also modelling entrepreneurship and economic growth at the cross-country level. Briefly, the proposed SEM model consists of a system of simultaneous equations modelling relationship among GDP (observed) and Entrepreneurship Activity (latent). Entrepreneurship is a latent construct that is formed by distinct (and observed) dimensions of entrepreneurship. These dimensions are: Nascent independent entrepreneur/sNew independent entrepreneursEstablished independent entrepreneursEntrepreneurial employeesThese dimensions of entrepreneurship are directly derived from the GEM definition of entrepreneurship. While entrepreneurship is a multifaceted phenomenon with many meanings and definitions, GEM defines entrepreneurship as:?"Any attempt at new business or new venture creation, such as self-employment, a new business organization, or the expansion of an existing business, by an individual, a team of individuals, or an established business". [GEM wiki, emphasis added]. It is worth to note that, the current de facto index of GEM (TEA), covers only the first two dimension of entrepreneurship (nascent and new). TEA (Total Early-Stage Entrepreneurial Activity) is the most important indicator of entrepreneurship produced by GEM and measures the share of the active population that are nascent entrepreneurs or are leading new businesses. In this respect proposed index (GEM-COIN) encompasses TEA and it is more in line with GEM definition of entrepreneurship. The paper implements the SEM on GEM APS data 2017 to construct a meaningful composite indicator of the entrepreneurial activities (EA). This composite indicator named GEM-COIN allows the ranking of countries based on comprehensive measure of entrepreneurship. Economies ranking and other measures are reported in associated excel file. The paper proceeds as follows. Section 2 provides a brief overview of the theoretical framework and the empirical results of previous studies examining the link between entrepreneurship and economic activity. Section 3 discusses the SEM methodology, and its application. The estimation results are presented in Section 4 and Section 5 presents how the GEM-COIN is derived based on the SEM model. Section 6 closes the paper with conclusions, future developments and recommendations.Theoretical frameworkThis Section re-elaborates the GEM framework presents framework for the analysis of Entrepreneurial Activity -EA-. EA is a complex phenomenon: not only is it multidimensional, but it is also the outcome of a process of achievement, in which the dimensions of EA are also interdependent. New products, new firms are constantly replacing, expanding and creating new markets. Success of new venture is often associated with difficulties, shrinking, and failures of exiting firms. This process of creation and destruction is often described as creative destruction (Schumpeter, 1934). The net effect of creative destruction is the main driver of economic growth. Main studies investigating the link entrepreneurship and economic growth are presented in next section.Previous studies: Economic growth and entrepreneurshipThere is a long standing tradition of belief in the value of entrepreneurship as a in economic growth. Economic growth models have expanded to incorporate various measures of entrepreneurship. Starting with the basic model ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2307/1884513","ISSN":"00335533","author":[{"dropping-particle":"","family":"Solow","given":"Robert M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"The Quarterly Journal of Economics","id":"ITEM-1","issue":"1","issued":{"date-parts":[["1956","2"]]},"page":"65","title":"A Contribution to the Theory of Economic Growth","type":"article-journal","volume":"70"},"uris":[""]}],"mendeley":{"formattedCitation":"(Solow, 1956)","plainTextFormattedCitation":"(Solow, 1956)","previouslyFormattedCitation":"(Solow, 1956)"},"properties":{"noteIndex":0},"schema":""}(Solow, 1956), researchers have sought to expand the list of economic factors that may contribute to observed economic growth, one of them being the role of entrepreneurship. This has led to valuable insights regarding the dynamic process by which economic growth occurs.ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"author":[{"dropping-particle":"","family":"Holcombe","given":"Randall G","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issue":"2","issued":{"date-parts":[["1998"]]},"page":"45-62","title":"ENTREPRENEURSHIP AND ECONOMIC GROWTH","type":"article-journal","volume":"2"},"uris":[""]}],"mendeley":{"formattedCitation":"(Holcombe, 1998)","manualFormatting":" Holcombe (1998)","plainTextFormattedCitation":"(Holcombe, 1998)","previouslyFormattedCitation":"(Holcombe, 1998)"},"properties":{"noteIndex":0},"schema":""} Holcombe (1998), for example, argues that entrepreneurship, once included into the standard neo‐classical growth model fleshes out the process by which the factors of production, namely, capital, labor and technology, interact to create economic growth. Incorporating some measure of entrepreneurship into a model of economic growth makes it “apparent that the engine of economic growth is entrepreneurship, not technological advance or investment in human capital per se” ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"author":[{"dropping-particle":"","family":"Holcombe","given":"Randall G","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issue":"2","issued":{"date-parts":[["1998"]]},"page":"45-62","title":"ENTREPRENEURSHIP AND ECONOMIC GROWTH","type":"article-journal","volume":"2"},"uris":[""]}],"mendeley":{"formattedCitation":"(Holcombe, 1998)","manualFormatting":" Holcombe (1998, p 60)","plainTextFormattedCitation":"(Holcombe, 1998)","previouslyFormattedCitation":"(Holcombe, 1998)"},"properties":{"noteIndex":0},"schema":""} Holcombe (1998, p 60).A standard model of economic growth represents some measure of real output, capital, labour, and technology. Historically, in estimating this model, technology was often estimated as the constant term since few direct measures of “technology” are available. Where technology is included it is usually proxied by a simple time‐trend measure reflecting the assumption of advances over time.Developments in growth theory have focussed on explaining the process by which technology advances ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"author":[{"dropping-particle":"","family":"Romer","given":"Paul M","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issue":"1","issued":{"date-parts":[["1994"]]},"page":"3-22","title":"The Origins of Endogenous Growth","type":"article-journal","volume":"8"},"uris":[""]}],"mendeley":{"formattedCitation":"(Romer, 1994)","plainTextFormattedCitation":"(Romer, 1994)","previouslyFormattedCitation":"(Romer, 1994)"},"properties":{"noteIndex":0},"schema":""}(Romer, 1994) ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"author":[{"dropping-particle":"","family":"Lucas","given":"Robert E","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issue":"August 1987","issued":{"date-parts":[["1988"]]},"page":"3-42","title":"ON THE MECHANICS OF ECONOMIC DEVELOPMENT* Robert E. LUCAS, Jr.","type":"article-journal","volume":"22"},"uris":[""]}],"mendeley":{"formattedCitation":"(Lucas, 1988)","plainTextFormattedCitation":"(Lucas, 1988)","previouslyFormattedCitation":"(Lucas, 1988)"},"properties":{"noteIndex":0},"schema":""}(Lucas, 1988). Endogenous growth theory emerged because the original Solow model did not address “black box” aspects of dynamic economic growth. That is, could it be that as one person's human capital is advanced there are positive externalities that enhance the productivity of others? If the answer is yes, and Romer's theoretical work suggests that it is, the stark Solow model is ill‐equipped to address the complexities of dynamic growth. Although endogenous growth theory provides a mechanism by which human capital development can be explained and helps explain observed diffusion in economic growth patterns, it focusses on the inputs to the production process, rather than the process itself. Extensions of the Solow model have tried to address this problem. There are many surveys of the existing growth literature. Suffice it to say that previous work, mostly conducted at the national level, has tested for the relative importance of health, geography, educational attainment, social institutions (such as property rights, religion, and corruption) and measures of general intelligence as predictors of economic growth ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"ISBN":"2033670007","author":[{"dropping-particle":"","family":"Sala-i-martin","given":"By Xavier","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Doppelhofer","given":"Gernot","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Miller","given":"Ronald I","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"The American Economic Review","id":"ITEM-1","issue":"4","issued":{"date-parts":[["2004"]]},"page":"813-835","title":"Determinants of Long-Term Growth : A Bayesian Averaging of Classical Estimates ( BACE ) Approach","type":"article-journal","volume":"94"},"uris":[""]}],"mendeley":{"formattedCitation":"(Sala-i-martin, Doppelhofer, & Miller, 2004)","plainTextFormattedCitation":"(Sala-i-martin, Doppelhofer, & Miller, 2004)","previouslyFormattedCitation":"(Sala-i-martin, Doppelhofer, & Miller, 2004)"},"properties":{"noteIndex":0},"schema":""}(Sala-i-martin, Doppelhofer, & Miller, 2004). Although the actions of entrepreneurs and their effect on economic activity have long been recognized ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"ISBN":"0878556982","abstract":"\"Original material copyright ?1934 by the President and Fellows of Harvard College\". -- verso page. Translation of: Theorie der wirtschaftlichen Entwicklung. Reprint. Originally published: Cambridge, Mass. : Harvard University Press, 1934. (Harvard economic studies ; v. 46) With new introduction. \"Transaction edition\"--Introduction. Overview: Schumpeter proclaims in this classical analysis of capitalist society first published in 1911 that economics is a natural self-regulating mechanism when undisturbed by \"social and other meddlers.\" In his preface he argues that despite weaknesses, theories are based on logic and provide structure for understanding fact. Of those who argue against him, Schumpeter asks a fundamental question: \"Is it really artificial to keep separate the phenomena incidental to running a firm and the phenomena incidental to creating a new one?\" In his answers, Schumpeter offers guidance to Third World politicians no less than First World businesspeople. In his substantial new introduction, John E. Elliott discusses the salient ideas of The Theory of Economic Development against the historical background of three great periods of economic thought in the last two decades. Translator's note -- Introduction to the transaction edition -- Preface to the English edition -- Circular flow of economic life as conditioned by given circumstances -- Fundamental phenomenon of economic development -- Credit and capital -- Entrepreneurial profit -- Interest on capital -- Business cycle.","author":[{"dropping-particle":"","family":"Schumpeter","given":"Joseph A.","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issued":{"date-parts":[["1934"]]},"number-of-pages":"255","publisher":"Transaction Publishers","publisher-place":"Harvard University Press.","title":"The theory of economic development : an inquiry into profits, capital, credit, interest, and the business cycle","type":"book"},"uris":[""]}],"mendeley":{"formattedCitation":"(Schumpeter, 1934)","manualFormatting":"(e.g. Schumpeter, 1934)","plainTextFormattedCitation":"(Schumpeter, 1934)","previouslyFormattedCitation":"(Schumpeter, 1934)"},"properties":{"noteIndex":0},"schema":""}(e.g. Schumpeter, 1934) empirically assessing the explanatory power of entrepreneurship is relatively new.How can entrepreneurship affect economic growth? It may not, as already noted, affect the inputs per se but can influence the process by which those inputs are combined to produce goods and services. Some have introduced entrepreneurship (in various forms) into endogenous growth models, with varying outcomes ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1007/978-1-4419-1191-9","ISBN":"9781441911919","author":[{"dropping-particle":"","family":"Carree","given":"Martin A","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Thurik","given":"A Roy","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Handbook of Entrepreneurship Research","editor":[{"dropping-particle":"","family":"Z.J. Acs","given":"","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Audretsch","given":"D.B.","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issue":"2","issued":{"date-parts":[["2010"]]},"page":"557-594","publisher":"Springer","title":"The Impact of Entrepreneurship on Economic Growth","type":"chapter"},"uris":[""]}],"mendeley":{"formattedCitation":"(Martin A Carree & Thurik, 2010)","plainTextFormattedCitation":"(Martin A Carree & Thurik, 2010)","previouslyFormattedCitation":"(Martin A Carree & Thurik, 2010)"},"properties":{"noteIndex":0},"schema":""}(Martin A Carree & Thurik, 2010). growth is generated by innovative entrepreneurs that drive creative destruction ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1007/s11187-016-9812-z","ISSN":"0921-898X","abstract":"This lecture is the story of an intellectual journey, that of elaborating a new—Schumpeterian— theory of economic growth.Atheorywhere (i) growth is generated by innovative entrepreneurs; (ii) entrepre- neurial investments respond to incentives that are them- selves shaped by economic policies and institutions; (iii) new innovations replace old technologies: in other words, growth involves creative destruction and there- fore involves a permanent conflict between incumbents and newentrants. First,wemotivate and then lay out the Schumpeterian paradigm and point to a set of empirical predictions which distinguish this paradigm from other growth models. Second, we raise four debates on which the Schumpeterian approach sheds newlight: themiddle income trap, secular stagnation, the recent rise in top income inequality, and firm dynamics. Third and last, we show how the paradigm can be used to think (or rethink) about growth policy design.","author":[{"dropping-particle":"","family":"Aghion","given":"Philippe","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Small Business Economics","id":"ITEM-1","issued":{"date-parts":[["2016"]]},"page":"9-24","publisher":"Small Business Economics","title":"Entrepreneurship and growth: lessons from an intellectual journey","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(Aghion, 2016)","plainTextFormattedCitation":"(Aghion, 2016)","previouslyFormattedCitation":"(Aghion, 2016)"},"properties":{"noteIndex":0},"schema":""}(Aghion, 2016). Entrepreneurs may impact growth through innovation and introducing new production processes ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"author":[{"dropping-particle":"","family":"Griliches","given":"Zvi","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"The Bell Journal of Economics","id":"ITEM-1","issue":"1","issued":{"date-parts":[["1979"]]},"page":"92-116","title":"Issues in assessing the contribution and development of research to productivity growth","type":"article-journal","volume":"10"},"uris":[""]}],"mendeley":{"formattedCitation":"(Griliches, 1979)","plainTextFormattedCitation":"(Griliches, 1979)","previouslyFormattedCitation":"(Griliches, 1979)"},"properties":{"noteIndex":0},"schema":""}(Griliches, 1979). ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1093/oxrep/grm001","ISBN":"9780511071072","ISSN":"0266903X","abstract":"This paper shows how and why the Solow growth accounting framework is useful for linking entrepreneurship capital to economic growth. The knowledge filter impedes the spillover of knowledge for commercialization, thereby weakening the impact of knowledge investments on economic growth. By serving as a conduit for knowledge spillovers, entrepreneurship is the missing link betwn investments in new knowledge and economic growth. Entrepreneurship is an important mechanism permeating the knowledge filter to facilitate the spillover of knowledge and ultimately generate economic growth. The emergence of entrepreneurship policy to promote economic growth is interpreted as an attempt to promote entrepreneurship capital, or the capacity of an economy to generate the start-up and growth of new firms.","author":[{"dropping-particle":"","family":"Audretsch","given":"David B.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Oxford Review of Economic Policy","id":"ITEM-1","issue":"1","issued":{"date-parts":[["2007"]]},"page":"63-78","title":"Entrepreneurship capital and economic growth","type":"article-journal","volume":"23"},"uris":[""]}],"mendeley":{"formattedCitation":"(Audretsch, 2007)","plainTextFormattedCitation":"(Audretsch, 2007)","previouslyFormattedCitation":"(Audretsch, 2007)"},"properties":{"noteIndex":0},"schema":""}(Audretsch, 2007) argues that the entrepreneur is able to exploit those new knowledge opportunities more fully than an organization within which such ideas may arise. Entrepreneurs can enhance the dissemination of new information and production techniques.The evidence at the country level is complex and mixed ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1007/s11187-005-1974-z","ISBN":"0921898X","ISSN":"0921898X","PMID":"15974422","abstract":"This paper is an introduction to the present special issue dedicated to scientific research using data col- lected as part of the Global Entrepreneurship Monitor (GEM) and considering new venture creation as the hall- mark of entrepreneurship. After a short description of GEM’s theoretical and methodological background, this introduction highlights the main results of seven papers which were presented at the First GEM Research Confer- ence in Berlin from 1 to 3 April 2004. First, there is empiri- cal evidence that the role of entrepreneurial activity differs across the stages of economic development, in that there appears to be a U-shaped relationship between the level of development and the rate of entrepreneurship. Conse- quently, a positive effect of entrepreneurial activity on eco- nomic growth is found for highly developed countries but a negative effect for developing nations. Second, it is shown that different types of entrepreneurship may have a different impact on a nation’s innovativeness and economic growth rate. In particular, potentially high-growth business start- ups and so-called opportunity entrepreneurship enhance knowledge spillovers and economic growth. Third, entre- preneurship is again shown to be a regional event that can only be understood if regional framework conditions, including networks and regional policies, are taken into consideration.","author":[{"dropping-particle":"","family":"Sternberg","given":"Rolf","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Wennekers","given":"Sander","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Small Business Economics","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2005"]]},"page":"193-203","title":"Determinants and effects of new business creation using global entrepreneurship monitor data","type":"article-journal","volume":"24"},"uris":[""]}],"mendeley":{"formattedCitation":"(Sternberg & Wennekers, 2005)","manualFormatting":"(Sternberg & Wennekers, 2005)","plainTextFormattedCitation":"(Sternberg & Wennekers, 2005)","previouslyFormattedCitation":"(Sternberg & Wennekers, 2005)"},"properties":{"noteIndex":0},"schema":""}(Sternberg & Wennekers, 2005). ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1007/s11187-005-2000-1","ISBN":"0921898X","ISSN":"0921898X","PMID":"220910300","abstract":"Studies on the impact of technological innovation on growth have been largely mute on the role of new firm formation. Using cross-sectional data on the 37 countries participating in GEM 2002, this paper uses an augmented Cobb-Douglas production to explore firm formation and technological innovation as separate deter- minants of growth. One area of interest is the contrast between different types of entrepreneurial activities as measured using GEM Total Entrepreneurial Activity (TEA) rates - high growth potential TEA, necessity TEA, opportunity TEA and overall TEA. Of the four types of entrepreneurship, only high growth potential entrepre- neurship is found to have a significant impact on eco- nomic growth. This finding is consistent with extant findings in the literature that it is fast growing new firms, not new firms in general, that accounted for most of the new job creation by small and medium enterprises in advanced countries","author":[{"dropping-particle":"","family":"Wong","given":"Poh Kam","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Ho","given":"Yuen Ping","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Autio","given":"Erkko","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Small Business Economics","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2005"]]},"page":"335-350","title":"Entrepreneurship, innovation and economic growth: Evidence from GEM data","type":"article-journal","volume":"24"},"uris":[""]}],"mendeley":{"formattedCitation":"(Wong, Ho, & Autio, 2005)","manualFormatting":"Wong, Ho, & Autio, (2005)","plainTextFormattedCitation":"(Wong, Ho, & Autio, 2005)","previouslyFormattedCitation":"(Wong, Ho, & Autio, 2005)"},"properties":{"noteIndex":0},"schema":""}Wong, Ho, & Autio, (2005) use subcomponents of the GEM index (high growth potential TEA, necessity TEA, opportunity TEA and overall TEA) and find that only “high‐potential entrepreneurship” significantly and positively impacts economic growth. This, they argue, is consistent with the notion that fast‐growing firms, not simply new firms in general, account for job creation and, thus, economic growth. ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1007/s11187-005-1996-6","ISBN":"0921898X","ISSN":"0921898X","PMID":"16731569","abstract":"Entrepreneurial activity is generally assumed to be an important aspect of the organization of industries most conducive to innovative activity and unre- strained competition. This paper investigates whether total entrepreneurial activity (TEA) influences GDP growth for a sample of 36 countries. We test whether this influence depends on the level of economic development measured as GDP per capita. Adjustment is made for a range of alternative explanations for achieving economic growth by incorporating the Growth Competitiveness Index (GCI). We find that entrepreneurial activity by nascent entrepre- neurs and owner/managers of young businesses affects economic growth, but that this effect depends upon the level of per capita income. This suggests that entrepre- neurship plays a different role in countries in different stages of economic development.","author":[{"dropping-particle":"Van","family":"Stel","given":"André","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Carree","given":"Martin","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Thurik","given":"Roy","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Small Business Economics","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2005"]]},"page":"311-321","title":"The effect of entrepreneurial activity on national economic growth","type":"article-journal","volume":"24"},"uris":[""]}],"mendeley":{"formattedCitation":"(Stel, Carree, & Thurik, 2005)","manualFormatting":"Stel, Carree, & Thurik, (2005)","plainTextFormattedCitation":"(Stel, Carree, & Thurik, 2005)","previouslyFormattedCitation":"(Stel, Carree, & Thurik, 2005)"},"properties":{"noteIndex":0},"schema":""}Stel, Carree, & Thurik, (2005) report that, after controlling for initial income and measures of global competitiveness, entrepreneurship has a positive and significant impact on national growth. But their results are tempered by the fact that this outcome holds only for relatively prosperous countries: increased entrepreneurship is found to negatively affect economic growth in developing countries. Entrepreneurial ecosystem looks playing a major role in explaining cross-country differences in economic growth ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"ISBN":"1118701800139","author":[{"dropping-particle":"","family":"Acs","given":"Zoltan J","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Estrin","given":"Saul","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Mickiewicz","given":"Tomasz","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issued":{"date-parts":[["2018"]]},"page":"501-514","publisher":"Small Business Economics","title":"Entrepreneurship , institutional economics , and economic growth : an ecosystem perspective","type":"article-journal"},"uris":[""]}],"mendeley":{"formattedCitation":"(Acs, Estrin, & Mickiewicz, 2018)","plainTextFormattedCitation":"(Acs, Estrin, & Mickiewicz, 2018)","previouslyFormattedCitation":"(Acs, Estrin, & Mickiewicz, 2018)"},"properties":{"noteIndex":0},"schema":""}(Acs, Estrin, & Mickiewicz, 2018).This study focusses on the entrepreneurship‐economic growth link at the country level. This paper differs from much of the previous work because proposes a combination of different dimensions of entrepreneurship to investigate the link between entrepreneurship and economic growth.MethodologyStructural Equation Models differ on both statistical tools and on the assumptions they make regarding the nature of the associations between the variables. Nonetheless, different methods share the assumption that a latent construct can be estimated through a set of observable indicators, which represent linear and noisy representations of the phenomenon itself ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1080/10705519809540105","ISBN":"1070-5511, Print\\r1532-8007, Electronic","ISSN":"10705511","PMID":"25522096","abstract":"Interactions of variables occur in a variety ofstatistical analyses. Tbe best known procedures for models with interactions of latent variables are technically demanding. Not only does the potential user need to be familiar with structural equation modeling (SEM), but the researcher must be familiar with programming nonlinear and linear constraints and must be comfortable with fairly large and complicated models. Tbis article provides a largely nontechnical description of an alternative two-stage least squares (2SLS) technique to include interactions of latent variables in SEM. Tbe method requires the selection of instrumental variables and we give mies for their selection in the most common cases. We compare the 2SLS method to the alternatives. Some of the important advantages of the 2SLS are that it can handle nonnormal observed variables, is readily available in major statistical software packages, and has a known asymptotic distribution. In providing the comparisons, we reanalyze aB the interaction examples from Kenny and Judd's (1984) arüde with the 2SLS method. We also give a new empirical example, and list SAS programs for aB examples.","author":[{"dropping-particle":"","family":"Bollen","given":"Kenneth A.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Paxton","given":"Pamela","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Structural Equation Modeling","id":"ITEM-1","issue":"3","issued":{"date-parts":[["1998"]]},"page":"267-293","title":"Interactions of latent variables in structural equation models","type":"article-journal","volume":"5"},"uris":[""]}],"mendeley":{"formattedCitation":"(Bollen & Paxton, 1998)","plainTextFormattedCitation":"(Bollen & Paxton, 1998)","previouslyFormattedCitation":"(Bollen & Paxton, 1998)"},"properties":{"noteIndex":0},"schema":""}(Bollen & Paxton, 1998). SEM is appealing for measuring multidimensional phenomena as it addresses the lack of agreement on the weights to be used to aggregate multiple indicators into a single metrics. SEM is particularly suitable to test research framework. The researcher has to put forward hypotheses on the number of latent factors and how they are associated with the observed indicators, and later to check the consistency of the theory with sample data. Additionally, SEM models aim at testing the association between the latent constructs and some exogenous variables that are hypothesised as influencing the latent factors. This paper adopts a Multiple Indicators Multiple Causes model -MIMIC-, ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2307/2285946","ISBN":"01621459","ISSN":"01621459","PMID":"402","abstract":"We consider a model in which one observes multiple indicators and multiple causes of a single latent variable. In terms of the multivariate regression of the indicators on the causes, the model implies restrictions of two types: (i) the regression coefficient matrix has rank one, (ii) the residual variance-covariance matrix satisfies a factor analysis model with one common factor. The first type of restriction is familiar to econometricians and the second to psychometricians. We derive the maximum-likelihood estimators and their asymptotic variance-covariance matrix. Two alternative \"limited information\" estimators are also considered and compared with the maximum-likelihood estimators in terms of efficiency.","author":[{"dropping-particle":"","family":"Joreskog","given":"Karl G.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Goldberger","given":"Arthur S.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Journal of the American Statistical Association","id":"ITEM-1","issue":"351","issued":{"date-parts":[["1975"]]},"page":"631","title":"Estimation of a Model with Multiple Indicators and Multiple Causes of a Single Latent Variable","type":"article-journal","volume":"70"},"uris":[""]}],"mendeley":{"formattedCitation":"(Joreskog & Goldberger, 1975)","manualFormatting":"(Joreskog & Goldberger (1975)","plainTextFormattedCitation":"(Joreskog & Goldberger, 1975)","previouslyFormattedCitation":"(Joreskog & Goldberger, 1975)"},"properties":{"noteIndex":0},"schema":""}(Joreskog & Goldberger (1975) to measure EA. MIMIC models are characterized by two types of equations: a “measurement equation”, which models the relationship between the latent phenomenon and its observed indicators, and a “structural equation”, which links the latent variable to a set of exogenous indicators. This general theoretical model can be characterised in the following way:EA or countries’ entrepreneurship activity is considered as a latent and endogenous factor in the structural model;Observable indicators of EA are modelled as constituents of the latent construct in the set of measurement equations. The latent “entrepreneurship activity”, is influenced by institutional, and economic elements. These are linked to the endogenous construct through a structural equation. To introduce some basic notation: Lety*a scalar of latent country ’s EA, or ‘entrepreneurship activity;y a (p x 1) vector of observed indicators representing the manifested associated association with the latent construct;λ a (p x 1) vector of factor loadings. These estimate the direct effects of the indicators on the latent construct and are interpreted as regression coefficients. In the case of standardised factor loadings, these represent the estimated correlation between the indicators and the underlying factor;x a (k x 1) vector of exogenous causes of y*;β’ a (1 x p) vector of path coefficients. These can be interpreted as regression coefficients.On this basis of the conceptual framework sketched above, we can introduce the following MIMIC model ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2307/2285946","ISBN":"01621459","ISSN":"01621459","PMID":"402","abstract":"We consider a model in which one observes multiple indicators and multiple causes of a single latent variable. In terms of the multivariate regression of the indicators on the causes, the model implies restrictions of two types: (i) the regression coefficient matrix has rank one, (ii) the residual variance-covariance matrix satisfies a factor analysis model with one common factor. The first type of restriction is familiar to econometricians and the second to psychometricians. We derive the maximum-likelihood estimators and their asymptotic variance-covariance matrix. Two alternative \"limited information\" estimators are also considered and compared with the maximum-likelihood estimators in terms of efficiency.","author":[{"dropping-particle":"","family":"Joreskog","given":"Karl G.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Goldberger","given":"Arthur S.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Journal of the American Statistical Association","id":"ITEM-1","issue":"351","issued":{"date-parts":[["1975"]]},"page":"631","title":"Estimation of a Model with Multiple Indicators and Multiple Causes of a Single Latent Variable","type":"article-journal","volume":"70"},"uris":[""]}],"mendeley":{"formattedCitation":"(Joreskog & Goldberger, 1975)","manualFormatting":"(Joreskog & Goldberger (1975)","plainTextFormattedCitation":"(Joreskog & Goldberger, 1975)","previouslyFormattedCitation":"(Joreskog & Goldberger, 1975)"},"properties":{"noteIndex":0},"schema":""}(Joreskog & Goldberger (1975):y*=λy+ε iy* = β’x + ν(ii)The first set of equations represents the measurement model, which specifies how the observed indicators are manifestations of the latent construct, the ‘activity entrepreneurship, plus an error term. The second equation specifies the structural model, which explains the latent construct as a function of a set of observed exogenous variables. Vectors ε and ν are the respective error terms in the measurement and structural equations, with zero expectations and uncorrelated between the two parts. In particular, ε captures uncertainty in the relationship between true EA and the observed indicators. Note that in MIMIC the exogenous covariates are modelled as error-free ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1037/a0024448.Three","abstract":"In the last two decades attention to causal (and formative) indicators has grown. Accompanying this growth has been the belief that we can classify indicators into two categories, effect (reflective) indicators and causal (formative) indicators. This paper argues that the dichotomous view is too simple. Instead, there are effect indicators and three types of variables on which a latent variable depends: causal indicators, composite (formative) indicators, and covariates (the “three Cs”). Causal indicators have conceptual unity and their effects on latent variables are structural. Covariates are not concept measures, but are variables to control to avoid bias in estimating the relations between measures and latent variable(s). Composite (formative) indicators form exact linear combinations of variables that need not share a concept. Their coefficients are weights rather than structural effects and composites are a matter of convenience. The failure to distinguish the “three Cs” has led to confusion and questions such as: are causal and formative indicators different names for the same indicator type? Should an equation with causal or formative indicators have an error term? Are the coefficients of causal indicators less stable than effect indicators? Distinguishing between causal and composite indicators and covariates goes a long way toward eliminating this confusion. We emphasize the key role that subject matter expertise plays in making these distinctions. We provide new guidelines for working with these variable types, including identification of models, scaling latent variables, parameter estimation, and validity assessment. A running empirical example on self-perceived health illustrates our major points. Keywords","author":[{"dropping-particle":"","family":"Bauldry","given":"Shawn","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Bollen","given":"Kenneth A.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Psychological methods","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2011"]]},"page":"265-284","title":"Three Cs in measurement models: causal indicators, composite indicators, and covariates.","type":"article-journal","volume":"16"},"uris":[""]}],"mendeley":{"formattedCitation":"(Bauldry & Bollen, 2011)","plainTextFormattedCitation":"(Bauldry & Bollen, 2011)","previouslyFormattedCitation":"(Bauldry & Bollen, 2011)"},"properties":{"noteIndex":0},"schema":""}(Bauldry & Bollen, 2011). The above relations are specified in Figure 2 below. ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.2307/2285946","ISBN":"01621459","ISSN":"01621459","PMID":"402","abstract":"We consider a model in which one observes multiple indicators and multiple causes of a single latent variable. In terms of the multivariate regression of the indicators on the causes, the model implies restrictions of two types: (i) the regression coefficient matrix has rank one, (ii) the residual variance-covariance matrix satisfies a factor analysis model with one common factor. The first type of restriction is familiar to econometricians and the second to psychometricians. We derive the maximum-likelihood estimators and their asymptotic variance-covariance matrix. Two alternative \"limited information\" estimators are also considered and compared with the maximum-likelihood estimators in terms of efficiency.","author":[{"dropping-particle":"","family":"Joreskog","given":"Karl G.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Goldberger","given":"Arthur S.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Journal of the American Statistical Association","id":"ITEM-1","issue":"351","issued":{"date-parts":[["1975"]]},"page":"631","title":"Estimation of a Model with Multiple Indicators and Multiple Causes of a Single Latent Variable","type":"article-journal","volume":"70"},"uris":[""]}],"mendeley":{"formattedCitation":"(Joreskog & Goldberger, 1975)","manualFormatting":"(Joreskog & Goldberger (1975)","plainTextFormattedCitation":"(Joreskog & Goldberger, 1975)","previouslyFormattedCitation":"(Joreskog & Goldberger, 1975)"},"properties":{"noteIndex":0},"schema":""}(Joreskog & Goldberger (1975) showed that the latent factor scores can be estimated by: (iii)With V(ε) = Ψ, V(ν) = σ2I, and Ω = λλ’ + Ψ. In general Ψ is assumed to be diagonal in the literature on latent variable models. SEM models estimate parameters to best reproduce observe covariance matrix.Empirical ApplicationSampleThe sample relates to a cross-section of 54 countries for the year 2017 with data from the International Investment Fund. GDP is measured as GDP in purchase power parity per capita. Although the sample is a quite limited for SEM, it is akin to the ones of analogous literature that uses the same data for cross-country comparisons. For instance, ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1007/s11187-005-1996-6","ISBN":"0921898X","ISSN":"0921898X","PMID":"16731569","abstract":"Entrepreneurial activity is generally assumed to be an important aspect of the organization of industries most conducive to innovative activity and unre- strained competition. This paper investigates whether total entrepreneurial activity (TEA) influences GDP growth for a sample of 36 countries. We test whether this influence depends on the level of economic development measured as GDP per capita. Adjustment is made for a range of alternative explanations for achieving economic growth by incorporating the Growth Competitiveness Index (GCI). We find that entrepreneurial activity by nascent entrepre- neurs and owner/managers of young businesses affects economic growth, but that this effect depends upon the level of per capita income. This suggests that entrepre- neurship plays a different role in countries in different stages of economic development.","author":[{"dropping-particle":"Van","family":"Stel","given":"André","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Carree","given":"Martin","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Thurik","given":"Roy","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Small Business Economics","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2005"]]},"page":"311-321","title":"The effect of entrepreneurial activity on national economic growth","type":"article-journal","volume":"24"},"uris":[""]}],"mendeley":{"formattedCitation":"(Stel et al., 2005)","manualFormatting":"Stel et al., (2005)","plainTextFormattedCitation":"(Stel et al., 2005)","previouslyFormattedCitation":"(Stel et al., 2005)"},"properties":{"noteIndex":0},"schema":""}Stel et al., (2005) have a sample of 37 countries. IndicatorsThe indicators in the measurement part, which aim to capture the latent ‘entrepreneurship activity’, are outcome indicators of distinct dimensions of EA. They are formative indicators and that are current collected using the core APS questions. These indicators are: Nascent independent entrepreneur/sNew independent entrepreneursEstablished independent entrepreneursEntrepreneurial employeesControlsDevelopment stages: Factor driven; Factor/Efficient; Efficient Driven; Efficient/Innovation ; Innovation Driven as defined by the World Economic Forum GDP level in 2016 in purchase power parity per capita to account for possible catch up ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1007/s11187-005-1996-6","ISBN":"0921898X","ISSN":"0921898X","PMID":"16731569","abstract":"Entrepreneurial activity is generally assumed to be an important aspect of the organization of industries most conducive to innovative activity and unre- strained competition. This paper investigates whether total entrepreneurial activity (TEA) influences GDP growth for a sample of 36 countries. We test whether this influence depends on the level of economic development measured as GDP per capita. Adjustment is made for a range of alternative explanations for achieving economic growth by incorporating the Growth Competitiveness Index (GCI). We find that entrepreneurial activity by nascent entrepre- neurs and owner/managers of young businesses affects economic growth, but that this effect depends upon the level of per capita income. This suggests that entrepre- neurship plays a different role in countries in different stages of economic development.","author":[{"dropping-particle":"Van","family":"Stel","given":"André","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Carree","given":"Martin","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Thurik","given":"Roy","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Small Business Economics","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2005"]]},"page":"311-321","title":"The effect of entrepreneurial activity on national economic growth","type":"article-journal","volume":"24"},"uris":[""]}],"mendeley":{"formattedCitation":"(Stel et al., 2005)","plainTextFormattedCitation":"(Stel et al., 2005)","previouslyFormattedCitation":"(Stel et al., 2005)"},"properties":{"noteIndex":0},"schema":""}(Stel et al., 2005)GDP growth in 2016 in purchase power parity per capita (to capture short term business cycle Estimation resultsFollowing table shows standardized results of a SEM model formulated as a Multiple Indicators multiple causes (MIMIC) using individual data aggregate at country level ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"ISBN":"9781597181396","abstract":"1st ed. Discovering Structural Equation Modeling Using Stata is devoted to Stata's sem command and all it can do. Learn about its capabilities in the context of confirmatory factor analysis, path analysis, structural equation modeling, longitudinal models, and multiple-group analysis. Each model covered is presented along with the necessary Stata code, which is parsimonious, powerful, and can be modified to fit a wide variety of models. The datasets used are downloadable, and you are encouraged to run the programs in a hands-on approach to learning. A particularly exciting feature of Stata is the SEM builder. This graphic interface for structural equation modeling allows you to draw publication-quality path diagrams and to fit the models without writing any programming code. When you fit a model with the SIM builder, Stata automatically generates the complete code that you can save for future use. Use of this unique tool is extensively covered in an appendix, and brief examples appear throughout the text. A miminal background in multiple regression is sufficient to benefit from this text. While it would be helpful to have some experience with Stata, it is not essential. Though the primary audience is those who are new to structural equation modeling, those who are already familiar with it will find this text useful for the Stata code it covers. Overall, the text is intended to be practical and will serve as a useful reference. Introduction to confirmatory factor analysis -- Using structural equation modeling for path models -- Structural equation modeling -- Latent growth curves -- Group comparisons -- Epilogue : what now? -- Graphical user interface -- Entering data from summary statistics.","author":[{"dropping-particle":"","family":"Acock","given":"Alan C.","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issued":{"date-parts":[["2013"]]},"number-of-pages":"304","publisher":"Stata Press","title":"Discovering structural equation modeling using Stata","type":"book"},"uris":[""]}],"mendeley":{"formattedCitation":"(Acock, 2013)","manualFormatting":"(see Acock, 2013)","plainTextFormattedCitation":"(Acock, 2013)","previouslyFormattedCitation":"(Acock, 2013)"},"properties":{"noteIndex":0},"schema":""}(see Acock, 2013). The estimation, implemented using Stata Version 15, confirmed identification.Table 1 the estimation results for the measurement part of the MIMIC model, reporting both normal and standardised coefficients. The model computes the coefficients to maximize the fit of data to the model. The standardized coefficients can also be interpreted as z-scores (Brown 2009). Basically, standardised coefficients report the weights to combine constituent variables (proportion of nascent, new, established independent entrepreneurs and entrepreneurial employees with respect to population 18-64 years old). REF _Ref535336301 \h Figure 1 offers a graphical representation of the model, and the standardized results. The upper part of REF _Ref535336301 \h Figure 1 shows how the latent variable GEM-COIN (i.e. the Entrepreneurship Composite Index) is formed and associated to GDP growth. The lower part shows how all variables that influence the GDP growth. Full results are reported in REF _Ref535336757 \h Table 1 and REF _Ref535336764 \h Table 2.Figure SEQ Figure \* ARABIC 1 SEM model estimatesNotes: MEASUREMENT: 4 observable indicators are forming the latent construct of “Entrepreneurship activity” (i.e. GEM-COIN) Nascent independent entrepreneurs (0-3 months)New independent entrepreneurs (3-42 months)Established independent entrepreneurs (>42 months)EEA Entrepreneurial employeesSTRUCTURAL: “Entrepreneurship activity”, is linked to economic growth (PPP GDP per capita) controlling for institutional, and economic environment.Dev. Stages: World Economic Forum stages of developments (-resource, efficiency and innovation driven- )GDP growth t-1 (business circle)GDP level t-1. (catch up) REF _Ref535336757 \h Table 1 shows that an increase of one unit in the item “EEA”, will result in an increase in 0.656 standardised units of GEM-COIN Similarly, an increase of one unit in the item “Nascent” will result in a decrease of 0.575 standardised units of GEM-COIN Although the ranking l has been found to be unaffected by the choice of the variable for scaling the latent factor, EEA was chosen as the reference variable because it is the indicator with the highest loading. It is important to note that the role of intrapreneurs is prominent in shaping the GEM-COIN in line with the studies that emphasize the importance of the innovation and diversification of businesses.One can note that nascent, new, established have negative weights while EEA is positive. This result may seem surprising to someone, but it is possibly due to difficulties of starting a new business alone outside of the resources available to larger organizations. This suggests that more mature/successful/impactful entrepreneurs (note that the negative coefficients decreases from new to established firms) or entrepreneurial employees (possibly entrepreneurs of more established firms and with more resources of independent entrepreneurs) are stronger drivers of economic growth. This result is consistent with Schumpeterian creative destruction and with the argument that innovation (more than the mere incorporation of low value new ventures) matters for economic growth. This result is also consistent with the empirical evidence of U- shaped relationship with TEA and GDP. Countries with higher TEA – nascent and new business- experience low GDP while countries with low TEA – and high EEA- experience.It is important to note also that established firms were nascent firms in previous years. This raises a more general issue of interplay between EA and Economic growth. New entrepreneurial activities may need some time to prove they are profitable and then contribute to GDP. In this sense EA anticipates economic growth, and we observe first a rise in EA and then higher economic growth in the data. In this case, we observe. At the same time, a potential entrepreneur may observe high economic activity and start a firm when he/she expect higher economic. expecting that the higher economic activity continues in the future. In this case, we observe first a higher economic activity and then a higher entrepreneurial activity. For sake of simplicity this study assumes that EA in one year has an impact in the same year. This assumption is also justified by the relative stability of entrepreneurship activities over time. For robustness the same model with different lagged values of nascent, new and established are estimated and results are reported Table SEQ Table \* ARABIC 1 Measurement part linking GEM-COIN and Entrepreneurial Activities: INCLUDETEXT "D:/gem/composite_indicator/tab/sem_coin_measu.rtf" \* MERGEFORMAT Measurement model---------------------------------------------------- (1) (2) Coeff. Stand. coeff. ----------------------------------------------------GEM-COIN Nascent -0.553 -0.575** (0.387) (0.279) New -0.375 -0.364 (0.573) (0.505) Established 0.0988 0.150 (0.328) (0.493) EEA 1 0.656*** (.) (0.226) ----------------------------------------------------obs. 54 ----------------------------------------------------Source: APS GEM 2017 and IMF* p<0.1, ** p<0.05, *** p<0.01The estimated coefficients in the structural part of the model can be interpreted as in multivariate regression analysis ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"ISBN":"1606238760","abstract":"3rd ed. \"This bestselling text provides a balance between the technical and practical aspects of structural equation modeling (SEM). Using clear and accessible language, Rex B. Kline covers core techniques, potential pitfalls, and applications across the behavioral and social sciences. Some more advanced topics are also covered, including estimation of interactive effects of latent variables and multilevel SEM. The companion Web page offers downloadable syntax, data, and output files for each detailed example for EQS, LISREL, and Mplus, allowing readers to view the results of the same analysis generated by three different computer tools.\"--Pub. desc. Part I: Concepts and tools. Introduction ; Fundamental concepts ; Data preparation ; Computer tools -- Part II: Core techniques. Specification ; Identification ; Estimation ; Hypothesis testing ; Measurement models and confirmatory factor analysis ; Structural regression models -- Part III: Advanced techniques, avoiding mistakes. Mean structures and latent growth models ; Interaction effects and multilevel SEM ; How to fool yourself with SEM.","author":[{"dropping-particle":"","family":"Kline","given":"Rex B.","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issued":{"date-parts":[["2011"]]},"number-of-pages":"427","publisher":"Guilford Press","title":"Principles and practice of structural equation modeling","type":"book"},"uris":[""]}],"mendeley":{"formattedCitation":"(Kline, 2011)","plainTextFormattedCitation":"(Kline, 2011)","previouslyFormattedCitation":"(Kline, 2011)"},"properties":{"noteIndex":0},"schema":""}(Kline, 2011). In this case, the structural part of the MIMIC provides estimates of the associations between the GEM-COIN and economic growth. In REF _Ref535336764 \h Table 2 the dependent variable is GDP growth 2016/2017 that is positively correlated with GEM-COIN. The rest of coefficients show associations between GDP growth 2016/17 and GDP 2016, GDP growth 2015/16 and the development stages. REF _Ref535336764 \h Table 2 (Structural) shows that GEM-COIN is positively correlated with GDP 2017 even when controlling for economic environment in terms of development stage of countries, previous year level and growth of GDP. Note that previous year GDP captures other important drivers of economic growth (among many others, infrastructures, institution, education, inflation, exchange rate, monetary and fiscal policy). GDP growth rate captures business cycles effects.Based on REF _Ref535336764 \h Table 2 conclude that when GEM-COIN increases in 1 unit, the GDP growth 2016/2017 increases on average of 0.3 unit, which can be interpreted as a positive impact of GEM-COIN on economic growth.Table SEQ Table \* ARABIC 2 Structural part linking GDP growth and GEM-COIN INCLUDETEXT "D:/gem/composite_indicator/tab/sem_coin_struct.rtf" \* MERGEFORMAT Structural model---------------------------------------------------- (1) (2) Coeff. Stand. coe~. ----------------------------------------------------GPD growth 2016/17 GEM-COIN 14.06* 0.308** (8.329) (0.127) GPD 2016 -0.0000265** -0.330** (0.000) (0.132) GPD growth 2015/16 0.422*** 0.579*** (0.067) (0.076) Dev. stage 1 ref. ref. Dev. stage 2 0.730 0.0739 (1.289) (0.130) Dev. stage 3 -0.526 -0.126 (0.942) (0.226) Dev. stage 4 0.191 0.0413 (0.964) (0.208) Dev. stage 5 -0.604 -0.161 (1.066) (0.284) Constant 4.001*** 2.145*** (1.118) (0.618) ----------------------------------------------------obs 54 ll -388.2 Coef. determination 0.613 ----------------------------------------------------Source: APS GEM 2017 and IMF* p<0.1, ** p<0.05, *** p<0.01 REF _Ref535330448 \h Figure 2 graphically shows the association between GEM-COIN and GDP growth and precise values are reported in online Appendix. Given the difficulties to explain GDP growth, a value of R2 of 0.61 is fairly acceptable. Nevertheless, the model does not success to account country specific idiosyncrasies such us, exceptional development of China, India and Viet Nam and the low performance of Gulf countries (probably due to large fluctuation of oil price) and Puerto Rico (that recently suffered severe recession). Figure SEQ Figure \* ARABIC 2 Association GEM-COIN and GDP growthRobustness checks Lags As discussed above, economic growth may have complex interplay with entrepreneurship activity (EA). New entrepreneurial activities may need some time to prove they are profitable and then contribute to GDP. In this sense EA anticipates economic growth. At the same time, a potential entrepreneur may start a venture when he/she expects higher economic activity in the future. In this case, we observe first a higher economic activity and then a higher entrepreneurial activity. Previous studies show that changes in the number of business owners have different impacts on GDP growth over time suggesting a U-shaped relationship ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1007/s11187-006-9007-0","ISBN":"0921-898X","ISSN":"0921898X","abstract":"This paper investigates the impact of changes in the number of business owners on three measures of economic performance, viz. employment growth, GDP growth and labor productivity growth. Particular attention is devoted to the lag structure. The analysis is performed at the country level for 21 OECD countries. Our results confirm earlier evidence on three stages in the impact of entry on economic performance: an initial direct positive effect, followed by a negative effect due to exiting capacities and finally a stage of positive supply-side effects. The net effect is positive for employment and GDP growth. Changes in the number of business owners have no effect on labor productivity.","author":[{"dropping-particle":"","family":"Carree","given":"M. A.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Thurik","given":"A. R.","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Small Business Economics","id":"ITEM-1","issue":"1","issued":{"date-parts":[["2008"]]},"page":"101-110","title":"The lag structure of the impact of business ownership on economic performance in OECD countries","type":"article-journal","volume":"30"},"uris":[""]}],"mendeley":{"formattedCitation":"(M. A. Carree & Thurik, 2008)","manualFormatting":"(Carree & Thurik, 2008)","plainTextFormattedCitation":"(M. A. Carree & Thurik, 2008)","previouslyFormattedCitation":"(M. A. Carree & Thurik, 2008)"},"properties":{"noteIndex":0},"schema":""}(Carree & Thurik, 2008). At the same time literature suggests that entrepreneurship is a structural characteristic of an economy relative stable over time ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"author":[{"dropping-particle":"","family":"Reynolds","given":"P. D.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Bygrave","given":"W.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Autio","given":"E.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Cox","given":"L.W.","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Hay","given":"M.","non-dropping-particle":"","parse-names":false,"suffix":""}],"id":"ITEM-1","issued":{"date-parts":[["2002"]]},"publisher":"Babson College","publisher-place":"Wellesley, MA","title":"Global Entrepreneurship Monitor 2002 Executive Report","type":"book"},"uris":[""]}],"mendeley":{"formattedCitation":"(Reynolds, Bygrave, Autio, Cox, & Hay, 2002)","plainTextFormattedCitation":"(Reynolds, Bygrave, Autio, Cox, & Hay, 2002)","previouslyFormattedCitation":"(Reynolds, Bygrave, Autio, Cox, & Hay, 2002)"},"properties":{"noteIndex":0},"schema":""}(Reynolds, Bygrave, Autio, Cox, & Hay, 2002). For this reason we run a robustness analysis using lagged values of nascent, new and established firms. Estimated are reported in REF _Ref535331514 \h Table 3 and REF _Ref535331519 \h Table 4. GEM data are not available for all economies all the year, therefore for comparability; the analysis is restricted to countries that are constantly participating in GEM in the period 2014- 2017. Specification (1) in column 1 of REF _Ref535331514 \h Table 3 and REF _Ref535331519 \h Table 4 report results of the SEM model that is our baseline. Specification (2) adds EA for 2015 to the base line to verify if EA in t-2 has an influence on GDP growth in t. Specification (3) adds EA for 2014 to the base line and specification (4) includes both EA 2015 and EA 2014. As specification (4) may run short of degree of freedom making estimates instable, specification (5) is more parsimonious and includes only EEA and Established in 2016, new 2015, nascent 2014. Finally, specification (6) includes nascent2015 and 2014 to the baseline (1).Overall, REF _Ref535331514 \h Table 3 confirms that the positive association of GDP growth and GEM-COIN for all 6 specifications. REF _Ref535331519 \h Table 4 shows that nascent and new are never positive and statistically significant confirming the main features of the SEM model presented in the previous section. Table SEQ Table \* ARABIC 3 Structural part linking GDP growth and GEM-COIN: 6 specifications with different lags (1)(2)(3)(4)(5)(6)Dep variable:GDP growth 2016/2017GEM-COIN0.41**0.46***0.47***0.56***0.43***0.46***(0.16)(0.15)(0.16)(0.10)(0.15)(0.16)Innovation driven-0.15-0.14-0.19-0.18-0.16-0.19(0.18)(0.18)(0.18)(0.16)(0.18)(0.18)GDP 2016-0.25-0.27-0.24-0.20-0.25-0.25(0.19)(0.18)(0.20)(0.18)(0.18)(0.18)GDP growth 2015/20160.60***0.59***0.61***0.67***0.60***0.59***(0.08)(0.09)(0.08)(0.08)(0.09)(0.09)_cons2.91***2.88***3.04***2.67***3.06***3.25***(0.92)(0.97)(0.94)(0.85)(0.91)(0.96)Table SEQ Table \* ARABIC 4 Measurement part linking GEM-COIN and Entrepreneurial Activities: different lags(1)(2)(3)(4)(5)(6)Dep variable:GEM-COIN2016 nascent-0.80**-0.71-0.21-0.140.11(0.33)(0.82)(0.62)(0.67)(0.89)2016 new-0.060.020.150.15-0.08(0.53)(0.70)(0.60)(0.58)(0.48)2016 established0.170.83-0.290.98*0.060.11(0.41)(0.79)(0.55)(0.58)(0.35)(0.37)2016 eea0.50*0.000.830.480.380.40(0.28)(0.00)(0.53)(0.52)(0.29)(0.28)2015 nascent-0.12-0.32-0.37(0.84)(0.59)(0.81)2015 new0.091.00*-0.05(0.61)(0.52)(0.42)2015 established-0.81-1.91***(0.80)(0.65)2015 eea0.54**0.73(0.24)(0.51)2014 nascent-0.39-0.02-0.81***-0.56(0.59)(0.47)(0.26)(0.52)2014 new-0.55-1.55***(0.60)(0.49)2014 established0.691.42***(0.50)(0.42)2014 eea-0.52-1.02**(0.60)(0.47)/var(e.grow2017)0.45***0.41***0.41***0.29***0.44***0.43***(0.09)(0.08)(0.08)(0.06)(0.09)(0.09)var(e.GEM-COIN)0.000.000.000.000.000.00_ll-1059.87-1329.82-1374.90-1618.80-1068.33-1228.04N424242424242bic215327002798330421702497.201Non linearities Previous studies suggest a U-shaped relationship between start-up rates of enterprise and levels of economic development ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1561/0300000023","ISSN":"1551-3114","author":[{"dropping-particle":"","family":"Wennekers","given":"Sander","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"van","family":"Stel","given":"Andre","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Carree","given":"Martin","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Thurik","given":"Roy","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Foundations and Trends? in Entrepreneurship","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2010"]]},"page":"167-237","title":"Towards a Psychology of Entrepreneurship: An Action Theory Perspective","type":"article-journal","volume":"6"},"uris":[""]}],"mendeley":{"formattedCitation":"(Wennekers, Stel, Carree, & Thurik, 2010)","plainTextFormattedCitation":"(Wennekers, Stel, Carree, & Thurik, 2010)","previouslyFormattedCitation":"(Wennekers, Stel, Carree, & Roy, 2010)"},"properties":{"noteIndex":0},"schema":""}(Wennekers, Stel, Carree, & Thurik, 2010). To test for possible non linearities, a common method is to compute GEM-COIN with the measurement part and plug the square in the structural part ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1016/j.dcn.2011.01.002.The","ISBN":"6176321972","ISSN":"15378276","author":[{"dropping-particle":"","family":"Nelson","given":"Eric E","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Guyer","given":"Amanda E","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Psychol Methods","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2012"]]},"page":"233-245","title":"A Comparison of Methods for Estimating Quadratic Effects in Nonlinear Structural Equation Models","type":"article-journal","volume":"1"},"uris":[""]}],"mendeley":{"formattedCitation":"(Nelson & Guyer, 2012)","plainTextFormattedCitation":"(Nelson & Guyer, 2012)","previouslyFormattedCitation":"(Nelson & Guyer, 2012)"},"properties":{"noteIndex":0},"schema":""}(Nelson & Guyer, 2012). REF _Ref536815737 \h Figure 3 and the LR chi square shows that the squared term is not fitting the data better and the linear model is preferred Figure SEQ Figure \* ARABIC 3 Non Linearities between GEM COIN and GDP growthDifferent development stages Previous literature suggests that entrepreneurship can play different role at different stages of development ADDIN CSL_CITATION {"citationItems":[{"id":"ITEM-1","itemData":{"DOI":"10.1007/s11187-005-1996-6","ISBN":"0921898X","ISSN":"0921898X","PMID":"16731569","abstract":"Entrepreneurial activity is generally assumed to be an important aspect of the organization of industries most conducive to innovative activity and unre- strained competition. This paper investigates whether total entrepreneurial activity (TEA) influences GDP growth for a sample of 36 countries. We test whether this influence depends on the level of economic development measured as GDP per capita. Adjustment is made for a range of alternative explanations for achieving economic growth by incorporating the Growth Competitiveness Index (GCI). We find that entrepreneurial activity by nascent entrepre- neurs and owner/managers of young businesses affects economic growth, but that this effect depends upon the level of per capita income. This suggests that entrepre- neurship plays a different role in countries in different stages of economic development.","author":[{"dropping-particle":"Van","family":"Stel","given":"André","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Carree","given":"Martin","non-dropping-particle":"","parse-names":false,"suffix":""},{"dropping-particle":"","family":"Thurik","given":"Roy","non-dropping-particle":"","parse-names":false,"suffix":""}],"container-title":"Small Business Economics","id":"ITEM-1","issue":"3","issued":{"date-parts":[["2005"]]},"page":"311-321","title":"The effect of entrepreneurial activity on national economic growth","type":"article-journal","volume":"24"},"uris":[""]}],"mendeley":{"formattedCitation":"(Stel et al., 2005)","plainTextFormattedCitation":"(Stel et al., 2005)","previouslyFormattedCitation":"(Stel et al., 2005)"},"properties":{"noteIndex":0},"schema":""}(Stel et al., 2005). For this reason the SEM model is estimated separately for innovation driven and not innovation driven countries and tested for invariance of GEM-COIN parameters across innovation and non-innovation driven countries. Due to the low tuberosity of factor driven countries in the sample, Factor driven and efficiency driven countries are pooled together and are defined as “not innovation driven countries”. REF _Ref536817120 \h Figure 4 shows that the GEM-COIN and GDP growth slope is not statistically significant over innovation and non-innovation for most values of GEM-COIN. At higher values of GEM-COIN, GEM-COIN has stronger impact on non-innovation countries then innovation. However, formal tests of for invariance of parameters between innovation and non- innovation suggest that, overall there is no evidence to reject the hypothesis that the slope of GEM-COIN on GDP. Results of separed SEM models for innovation and not innovation are reported in REF _Ref536817839 \h Table 5 REF _Ref536817841 \h Table 6 in appendix, the formal tests are in REF _Ref536817506 \h Table 7.Figure SEQ Figure \* ARABIC 4 GEM-COIN and GDP growth in innovation and non-innovation driven countries A multidimensional index of countries Entrepreneurship: GEM-COINBased on the results of the MIMIC model presented in previous sections, the latent ‘entrepreneurship activity’ was estimated. The GEM-COIN is obtained by normalizing the scores on a scale from 1 to 100, where 1 indicates a situation of minimum EA and 100 the maximum entrepreneurship activity. This paper argues that GEM-COIN has three features that render it appealing for measuring EA: first, by including outcome indicators on different dimensions of the concept, GEM-COIN is more able to capture the complexity of EA than single indicators alone as TEA. Secondly, through the combination of different indicators, GEM-COIN reduces the impact of random measurement error in single indicators. Finally, as shown by Table 2, the strength of association is higher for the composite index than for its components. This feature shows the ability of the GEM-COIN to better capture entrepreneurship than its single components, and hence to provide a comprehensive, yet summary, view of overall entrepreneurship activity. Uncertainty analysisConstructing composite indicator may require taking choices that may impact influence empirical results. Is GPD growth in 2017 associated with GDP grow of 2016 or 2015? Are entrepreneurs that are active and leading important or all intrapreneurs, regard less they are leading the new process and products within the enterprise (Note that TEA does not include all owner and managers of new ventures). Do entrepreneurship matters for GDP growth only or also for levels of GDP?. As answers to these questions are not unique, as suggested in the OECD handbook to construct composite indicator, am extensive uncertainty analysis is conducted.The SEM model is re-estimated substituting EEA defined as active and leading entrepreneurs in the last three years-eea1- with EEA defined as active entrepreneurs in the last three years, regardless of the leading position-EEA-. REF _Ref963926 \h Figure 5 shows that EEA has a larger variability than eea1 (see axis x) but the two GEM-COIN provide similar results and they are highly correlated (0.97). This increases confidence in our estimates. Results of the SEM model using GDP in level are reported in appendix and confirm the appropriateness of GEM-COIN in predicting GDP and importance of EEA as indicator of national entrepreneurship. Figure SEQ Figure \* ARABIC 5 Difference between alternative measures of EEASeveral SEM specifications have being estimated to conduct the uncertainty analysis. Each specification includes development stages, GDP growth 2016 and GDP level 2016, but they differ in terms of: Dependent variable (GDP2017in level of growth)EEA definition (considering leadership or not)Lag of GDP growth (from 2013 to 2015).Each specification is estimated, GEM-COIN computed and countries ranked on the basis of the estimated GEM-COIN Overall, 16 different specifications are estimates. Figure 3 report the Inter Quantile Range -IQR- of the ranking of these 16 specifications. The figure shows that the ranking remains relatively stable – countries in the top distribution remain in the top while countries in the bottom fluctuate in the bottom of the ranking, even if some countries experience larger variability in position than others, probably do to idiosyncratic pattern of entrepreneurship component and economic development.Overall, Figure 3 calls for caution when interpreting any composite index, because it comes with a lot of uncertainty. Figure SEQ Figure \* ARABIC 6 Uncertainty analysis of GEM-COINConclusions and Future developments This paper has shown how SEM model rooted in GEM framework allows constructing meaningful entrepreneurship composite index that is positively associated with economic growth. A new entrepreneurship composite index is presented. As each model, the SEM model (and GEM-COIN) hinges on some assumptions. Relaxing them will enhance the GEM-COIN. Interpretation of causality using observational data is no less problematic than the one of standard regression models. As such, it is important to stress that the estimated coefficients in the empirical applications are measures of statistical associations and not necessarily of causality in the sense of Rubin or Granger. In this paper, I controlled for some lag structure of GEM-COIN but a proper panel analysis can better investigate how much time is needed to nascent and new to display their supposed positive effect on economic growth. The negative contribution of nascent entrepreneurship on the GEM-COIN it may look surprising but this result is robust event to lagged values. This is consistent with Schumpeterian creative destruction with the argument that innovation (more than the mere incorporation of low value new ventures) matters for economic growth. This result is also consistent with the empirical evidence of U- shaped relationship with TEA and GDP. Countries with higher TEA – nascent and new business- experience low GDP while countries with low TEA – and high EEA- experience. Possible developments of GEM-COIN are: Panel analysis to investigate lag of GEM-COIN and GDP (GEM-COIN predict GDP or GDP predicts entrepreneurship? Is forecasted GDP in year x+1 influencing entrepreneurship in year x?)Test the whole GEM framework including other components of GEM framework as additional variable Investigate impactful entrepreneurship (GEM-COIN as TEA consider all entrepreneurs the same impactful -next google is different from next pizzeria!) – Increasing sample size. It requires missing data imputation and there is not much consensus on the appropriateness of this technique.Include Growth Competitiveness Index Global Competitively Index of the World Economic Forum as control. However it is important based on the result of the Executive Opinion Survey whose relatively small sample size for each country may raise some issue. In 2017 there are only 43 interviews in Luxembourg and detailed information on the quality of the survey by country was not published (see ). Distinguish between Employee Entrepreneurial Activity -EEA-and Total Early Activity -TEA-(avoiding possible double counting issues).ReferencesADDIN Mendeley Bibliography CSL_BIBLIOGRAPHY Acock, A. C. (2013). Discovering structural equation modeling using Stata. Stata Press.Acs, Z. J., Estrin, S., & Mickiewicz, T. (2018). Entrepreneurship , institutional economics , and economic growth?: an ecosystem perspective, 501–514.Aghion, P. (2016). Entrepreneurship and growth: lessons from an intellectual journey. Small Business Economics, 9–24. , D. B. (2007). Entrepreneurship capital and economic growth. Oxford Review of Economic Policy, 23(1), 63–78. , S., & Bollen, K. A. (2011). Three Cs in measurement models: causal indicators, composite indicators, and covariates. Psychological Methods, 16(3), 265–284. , K. A., & Paxton, P. (1998). Interactions of latent variables in structural equation models. Structural Equation Modeling, 5(3), 267–293. , M. A., & Thurik, A. R. (2008). The lag structure of the impact of business ownership on economic performance in OECD countries. Small Business Economics, 30(1), 101–110. , M. A., & Thurik, A. R. (2010). The Impact of Entrepreneurship on Economic Growth. In Z.J. Acs & D. B. Audretsch (Eds.), Handbook of Entrepreneurship Research (pp. 557–594). Springer. , Z. (1979). Issues in assessing the contribution and development of research to productivity growth. The Bell Journal of Economics, 10(1), 92–116.Hafer, R. W. (2013). Entrepreneurship and state economic growth. Journal of Entrepreneurship and Public Policy, 2(1), 67–79. , R. G. (1998). ENTREPRENEURSHIP AND ECONOMIC GROWTH, 2(2), 45–62.Joreskog, K. G., & Goldberger, A. S. (1975). Estimation of a Model with Multiple Indicators and Multiple Causes of a Single Latent Variable. Journal of the American Statistical Association, 70(351), 631. , R. B. (2011). Principles and practice of structural equation modeling. Guilford Press.Lucas, R. E. (1988). ON THE MECHANICS OF ECONOMIC DEVELOPMENT* Robert E. LUCAS, Jr., 22(August 1987), 3–42.Nelson, E. E., & Guyer, A. E. (2012). A Comparison of Methods for Estimating Quadratic Effects in Nonlinear Structural Equation Models. Psychol Methods, 1(3), 233–245. , P. D., Bygrave, W., Autio, E., Cox, L. W., & Hay, M. (2002). Global Entrepreneurship Monitor 2002 Executive Report. Wellesley, MA: Babson College.Romer, P. M. (1994). The Origins of Endogenous Growth, 8(1), 3–22.Sala-i-martin, B. X., Doppelhofer, G., & Miller, R. I. (2004). Determinants of Long-Term Growth?: A Bayesian Averaging of Classical Estimates ( BACE ) Approach. The American Economic Review, 94(4), 813–835.Schumpeter, J. A. (1934). The theory of economic development?: an inquiry into profits, capital, credit, interest, and the business cycle. Harvard University Press.: Transaction Publishers. Retrieved from , R. M. (1956). A Contribution to the Theory of Economic Growth. The Quarterly Journal of Economics, 70(1), 65. , A. Van, Carree, M., & Thurik, R. (2005). The effect of entrepreneurial activity on national economic growth. Small Business Economics, 24(3), 311–321. , R., & Wennekers, S. (2005). Determinants and effects of new business creation using global entrepreneurship monitor data. Small Business Economics, 24(3), 193–203. , S., Stel, A. van, Carree, M., & Thurik, R. (2010). Towards a Psychology of Entrepreneurship: An Action Theory Perspective. Foundations and Trends? in Entrepreneurship, 6(3), 167–237. , P. K., Ho, Y. P., & Autio, E. (2005). Entrepreneurship, innovation and economic growth: Evidence from GEM data. Small Business Economics, 24(3), 335–350. SEQ Figure \* ARABIC 7 GEM -COIN and GDP2017 in levelFirst panel (Structural) of the table below shows that Entrepreneurship (Entre) is positively correlated with (ln) of GDP even when controlling for previous year level of GDP and development stage of countries. The magnitude of the effect of Entre (0.0156) is small but reasonable when compared to the effect of GDP of previous year on current year (0.989). Note that previous year GDP captures other important drivers of economic growth (among many others investments infrastructures, capital, education etc etc).Second panel (Entre) shows how the latent variable ENTRE (i.e. the Entrepreneurship composite Index or GEM-COIN or Entre COIN) is formed. Basically, it reports the weights to combine constituent variables (proportion of nascent, new, established independent entrepreneurs and entrepreneurial employees with respect to population 18-64 years old). The weights are computed to maximize the fit of data to the model. One can note that nascent, new, established have negative weights while EEA is positive. This suggests that more mature/successful/impactful entrepreneurs (note that the negative coefficients decrease from new to established) or entrepreneurial employees (possibly entrepreneurs of more established firms and with more resources of independent entrepreneurs) are stronger drivers of economic growth. Legend and notes:lnam= natural logarithm of GDP per capita in international $ (purchase power parity) in 2017lnam= natural logarithm of GDP per capita in international $ (purchase power parity) in 2016dev1_1 = dummy variable for country in stage 1 of development (factor driven) according to wef classification (not reported because reference category)dev1_2 = dummy variable for country in transition from stage 1 to stage 2dev1_3 = dummy variable transition in stage 2 (efficiency driven)dev1_4 = dummy variable transition from stage 2 to stage 3dev1_5 = dummy variable transition in stage 3 (innovation driven)Var (e.Entre) constrained to 0 means that Entre is measured without errorsUnit of analyse 54 countries participated in APS GEM 2017Source: APS GEM 2017 and IMFTable SEQ Table \* ARABIC 5 SEM model for non-innovation driven countries Table SEQ Table \* ARABIC 6 SEM model for Innovation driven countriesTable SEQ Table \* ARABIC 7 Test for invariance of parameters across innovation and non-innovationWald tests are reported for parameters that were not constrained. The null hypothesis is that aconstraint would be valid. Results show that the hypothesis that GEM-COIN has the same slope in innovation and not innovation driven countries cannot be reject. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download