Published online: 03 Feb 2014.

Transcription

1 This artice was downoaded by: [Universidad de Chie] On: 09 October 2014, At: 11:35 Pubisher: Routedge Informa Ltd Registered in Engand and Waes Registered Number: Registered office: Mortimer House, Mortimer Street, London W1T 3JH, UK Appied Economics Letters Pubication detais, incuding instructions for authors and subscription information: The factors affecting the risk attitude in entrepreneurship: evidence from Latin America Jean P. Sepúveda a & Caudio A. Bonia b a Schoo of Business and Economics, Universidad de Desarroo, Concepción, Chie b Schoo of Business and Economics, Universidad de Chie, Santiago, Chie Pubished onine: 03 Feb To cite this artice: Jean P. Sepúveda & Caudio A. Bonia (2014) The factors affecting the risk attitude in entrepreneurship: evidence from Latin America, Appied Economics Letters, 21:8, , DOI: / To ink to this artice: PLEASE SCROLL DOWN FOR ARTICLE Tayor & Francis makes every effort to ensure the accuracy of a the information (the Content ) contained in the pubications on our patform. However, Tayor & Francis, our agents, and our icensors make no representations or warranties whatsoever as to the accuracy, competeness, or suitabiity for any purpose of the Content. Any opinions and views expressed in this pubication are the opinions and views of the authors, and are not the views of or endorsed by Tayor & Francis. The accuracy of the Content shoud not be reied upon and shoud be independenty verified with primary sources of information. Tayor and Francis sha not be iabe for any osses, actions, caims, proceedings, demands, costs, expenses, damages, and other iabiities whatsoever or howsoever caused arising directy or indirecty in connection with, in reation to or arising out of the use of the Content. This artice may be used for research, teaching, and private study purposes. Any substantia or systematic reproduction, redistribution, reseing, oan, sub-icensing, systematic suppy, or distribution in any form to anyone is expressy forbidden. Terms & Conditions of access and use can be found at

2 Appied Economics Letters, 2014 Vo. 21, No. 8, , The factors affecting the risk attitude in entrepreneurship: evidence from Latin America Jean P. Sepúveda a, * and Caudio A. Bonia b a Schoo of Business and Economics, Universidad de Desarroo, Concepción, Chie b Schoo of Business and Economics, Universidad de Chie, Santiago, Chie Downoaded by [Universidad de Chie] at 11:35 09 October 2014 This artice departs from the traditiona anaysis of the effects of risk aversion in entrepreneurship to study the determinants of entrepreneuria risk aversion in deveoping a new venture and becoming an entrepreneur. We took fear of faiing as a proxy for risk aversion and appied our anaysis to the most important Latin American economies. We observed that being mae, having more years of forma education and beieving to have the necessary skis to deveop a new venture decreased the probabiity of feeing a fear of faiing and, thus, eventuay increased the probabiity of deveoping a new venture. Age affects risk quadraticay (first positivey, but after some point, negativey), and if there is a prior experience of having shut down a business, risk aversion increases, that is, the probabiity of feeing a fear of faiing, which reduces the probabiity of becoming an entrepreneur. Keywords: risk aversion; entrepreneurship; fear of faiing JEL Cassification: L26; M13 I. Introduction Traditionay, entrepreneurship has been a subject of study in management using the case study methodoogy. Ony recenty have a arger number of economists reaized that entrepreneurship is at the core of the fied of economics and, therefore, the subject deserves to be treated from the perspective of the rationa choice paradigm using a the empirica toos and theoretica methods avaiabe to modern microeconomic anaysis. From the standpoint of economic theory, one of the focuses of entrepreneurship iterature has been on the reationship between risk aversion and entrepreneurship. Literature has shown that more risk-averse individuas are more ikey to become empoyees, whie ess risk-averse individuas are more ikey to become entrepreneurs. Important eary contributions here are the works of Kihstrom and Laffont (1979), who deveoped a mode that, in equiibrium, produces ess risk-averse entrepreneurs, and the work of Kanbur (1979), who deveoped a mode that disputes Friedman s(1953) caim that different attitudes towards risk resut in income inequaity. We shoud aso mention here the cassic work of Knight (1971), who emphasizes the wiingness (and) power to give satisfactory guarantees, that is, the wiingness to assume risk, as a fundamenta factor determining the abundance of entrepreneurs. Since then, a considerabe body of research that studies the risk aversion-entrepreneurship reationship has emerged (Parker, 1997; Cressy, 2000; Van Praag and Cramer, 2001; Newman, 2007; Parker, 2007). We wi briefy discuss some of these artices in the next section. *Corresponding author. E-mai: jeansepuveda@udd.c 2014 Tayor & Francis 573

3 574 J. P. Sepúveda and C. A. Bonia Downoaded by [Universidad de Chie] at 11:35 09 October 2014 On the empirica side, much of the research on entrepreneurship has focused on correating entrepreneurship to economic growth (see, for exampe, Acs, 1996; Acs and Varga, 2005; Van Ste et a., 2005; Wennekers et a., 2005; Wong et a., 2005; Acs, 2006; Acs and Szerb, 2007) and on expaining the determinants of entrepreneurship (see, for exampe, Demar and Davidsson, 2000; Davidsson and Honig, 2003; Parker, 2004; Davidsson, 2006; Grio and Thurik, 2008). In these cases, different proxies of risk aversion are usuay considered to be determining factors affecting entrepreneurship. In this artice we do something innovative. Instead of studying the effect of risk attitudes on entrepreneurship, we anayse the factors that can expain the individua s attitude towards risk at the microeconomic eve. In other words, whether how wiing an individua is to bear the risk inherent in deveoping a new venture is one of the precursors to entrepreneurship. From an empirica point of view, the attitude towards risk has been taken into account as a determinant of entrepreneurship. For instance, Ardagna and Lusardi (2008) use a proxy for risk aversion as an exogenous variabe when expaining internationa differences in entrepreneurship eves and Cramer et a. (2002) use Dutch data to show that risk aversion negativey affects entrepreneurship. To our knowedge, the ony study trying to estabish what are the determinants of risk aversion (fear of faiing) is the one by Sepúveda and Bonia (2011), which research the issue for Chie. In this research, we expand the previous one beyond Chie to cover five Latin American countries where Goba Entrepreneurship Monitor (GEM) data is avaiabe. We want to study what the factors are that affect, at the individua eve, the attitude towards the risk of starting a new business. We want to know whether age, gender, skis, previous faied experiences in a business and educationa eves, among others, can hep to differentiate among individuas more or ess wiing to bear the risk of becoming entrepreneurs. We are aso interested in studying whether there are some country effects that may hep to expain specific characteristics of subgroups of the data. We focus on Latin America because a these countries are in the group of deveoping economies and, therefore, they face simiar economic chaenges for the future. For them, choosing between a good or bad production poicy can make the difference between achieving deveopment and maintaining poor economic conditions for a ong time. Aso, Latin American countries usuay share some degree of poitica instabiity (Banco and Grier, 2009) and we observe in these countries, from time to time, the emergence of popuist eaders that damage the poitica and economic institutions, property rights and the fisca baance, which directy affects the risk of any entrepreneuria activity. We want to see how this context impacts the resuts of our anaysis. To this end, we take advantage of a question in the Adut Popuation Survey (APS) of the GEM project that asks individuas whether fear of faiure woud prevent them from starting a new business. This is a yes or no question, which we use to mode fear of faiure as a proxy for the attitude towards risk. We assume that a person who decares that fear of faiure woud prevent him or her from starting a venture has a more risk-averse attitude than one who is not afraid to take on the chaenge. We estimate a Probit mode that can correcty predict, about 67% of the time, whether an individua has a greater or esser fear of faiing. We found that women have more fear of faiing than men do. Each extra year of education above the mean of the sampe can decrease the probabiity of feeing a fear of faiing by 0.7%. Individuas who beieve they have the skis necessary to be entrepreneurs have a 14% ower probabiity of feeing a fear of faiing. Individuas with a previous faiure or individuas who know a nascent entrepreneur are ess ikey to experience the fear of faiure. The artice continues as foows: Section II briefy describes the iterature that studies the effects of entrepreneurship and economics; Section III describes the methodoogy used in this study; Section IV presents the main resuts of our anaysis; and Section V expains the concusions. II. Entrepreneurship and Economics Entrepreneurship is a key eement in the growth of nations. The entrepreneur's roe in the process of deveopment has been emphasized in the iterature for a ong time. Schumpeter (1934) argues that the existence of innovative entrepreneurs drives the process of creative destruction by which new firms repace the od ones through efficiency improvements, which resuts in higher eves of productivity. Kirzner (1973, 1979) highights the roe of the entrepreneur as the one who is abe to recognize opportunities where the ordinary person cannot see any. Knight (1971) recognizes that the entrepreneur is a eader who organizes production inputs and is wiing to confront the uncertainty that characterizes a new endeavour. More recenty, Baumo (2002) argues that the entrepreneuria spirit is the main factor expaining the growth mirace of capitaism. Through innovation, the entrepreneur is then the engine of growth. Acs (2006) expains how entrepreneurship is good for economic growth by separating necessity entrepreneurs those who become entrepreneurs because no other option is avaiabe from opportunity entrepreneurs, those who become entrepreneurs ooking for an undiscovered opportunity. Acs argues that the first type of entrepreneurship has no effect on economic deveopment, whie the second has a

4 Risk attitude and entrepreneurship 575 Downoaded by [Universidad de Chie] at 11:35 09 October 2014 positive and significant effect on economic deveopment and growth. Given the importance of entrepreneurship on economic growth, most of the recent work has focused on the determinants of entrepreneuria activity, ike the roe of individua characteristics on entrepreneurship. For instance, Evans and Jovanovic (1989) investigate the ink between initia weath, the reguatory and institutiona environment in which entrepreneurs have to operate and the choice of becoming an entrepreneur. Hotz- Eakin et a. (1994) inspect the reationship between access to capita and entrepreneuria surviva. Johnson et a. (2002) anayse the nexus between property rights and entrepreneurship. Lerner and Mamendier (2007) study the effect of entrepreneuria peers on entrepreneurship rates. Giannetti and Simonov (2009) examinethe effect of socia interaction on the eves of entrepreneuria activity, and Ardagna and Lusardi (2008) expain internationa differences in entrepreneurship by highighting the roe of individua characteristics and reguatory constraints. The theoretica iterature about the choice of becoming an entrepreneur puts the emphasis on risk aversion. Indeed, since the work of Knight (1971), who emphasizes the wiingness to bear risk as a fundamenta factor determining the abundance of entrepreneurs, the iterature describes entrepreneurship as the resut of choosing between working for someone ese (ow risk) and sef-empoyment (high risk). Within this notion, more recenty, Cressy (2000) proposes that the positive effect of weath on business start-ups, as advanced by Evans and Jovanovic (1989) and corroborated by most of the empirica iterature, may be the resut of decreasing absoute risk aversion (DARA). Kan and Tsai (2006) empiricay prove Cressy s argument and find that Evans and Jovanovic s caim is robust to the incusion of risk aversion. Newman (2007), on the other hand, deveops a principa-agent mode to chaenge the risk aversionbased expanation for the existence of empoyees and entrepreneurs by endogenizing entrepreneuria risk and aowing for optima insurance contracts as we as occupationa sef-seection. A the previous discussion is about the effect of risk aversion or about genera attitudes towards risk in the decision to become an entrepreneur. Indeed, an individua s attitude towards risk is a key precursor to entrepreneuria activity. For this reason, it is important to study and understand what factors can infuence attitudes towards risk. To our knowedge, no empirica work has been done on the factors that affect an individua's attitudes towards risk and that individua s choice to become an entrepreneur. In this artice we present an empirica investigation that aims to shed ight on the factors that affect an individua s attitude towards the risk invoved in becoming an entrepreneur. III. Empirica Methodoogy The GEM is an internationa research program started in 1999 that coects, annuay, cross-country harmonized data on entrepreneurship. The project consists of two surveys: (i) the APS, and the (ii) Expert Questionnaire Data. The APS surveys about 2000 individuas, seected at random, from each of the participating countries. The number of surveyed countries has increased from 10 in 1999 to 85 in The GEM project contains, among severa questions, a query that asks individuas: woud fear of faiure prevent you from starting a business? We use the answers to this question as a proxy for the attitude towards risk of the individua answering the survey. We consider that the person who answers yes to this query is ess wiing to bear the risk of becoming an entrepreneur than the person who answers no. We use data from the APS appied in the foowing emerging markets in Latin America: Argentina, Brazi, Chie, Mexico and Venezuea, a in Estimation The answers to the fear of faiing question adopts the vaue of 1 if answer is yes and 0 if answer is no. For this reason, we estimated the foowing Probit mode: FF i ¼ γ 0 þ γ 1 AGE i þ γ 2 AGE 2 i þ γ 3 WOMEN i þ γ 4 SUSKILL i þ γ 5 KNOWENT i þ γ 6 CLOSE i þ γ 7 OPPORT i þ γ 8 EDUCSD i þ γ 9 EDUCPS i þ γ 10 EDUCGE i þ γ 11 TEA05OPP i þ COUNTRY DUMMIES þ INTERACTION EFFECTS þ ε i where: (1) i denotes each individua in the survey, FF is a dummy variabe that represents the answer to the question: Woud fear of faiure prevent you from starting a business?, AGE represents the age, in years, of the individua answering the survey, WOMEN is a dummy variabe equa to 1 if the person is a femae and 0 otherwise, SUSKILL is a dummy variabe that equas 1 if the person answers yes to the question: Do you have the knowedge, ski and experience required to start a new business? and 0 otherwise, KNOWENT equas 1 if the individua personay knows someone that started a business in the past 2 years,

5 576 J. P. Sepúveda and C. A. Bonia Downoaded by [Universidad de Chie] at 11:35 09 October 2014 CLOSE equas 1 if the person has, in the past 12 months, shut down, discontinued or quit a business he or she owned and managed and 0 otherwise, OPPORT is a dummy variabe representing the yes or no answer to the question: In the next six months wi there be good opportunities for starting a business in the area where you ive?, EDUCSD, EDUCPS and EDUCGE are dummy variabes that equa 1 if the person answering the survey had some secondary degree of education (but did not finish it), had some post-secondary degree of education or graduate education, respectivey, and 0 otherwise, TEA05OPP is Tota Entrepreneuria Activity Index 2005 that reports opportunity as the major motive. We aso incude country dummies and interaction dummies to check for sope changes among countries. ε is a normay distributed disturbance with mean 0 and unit variance. Tabes 1 6 present a summary of the aggregated data and by country. Tabe 1. Data summary: a countries Variabe Obs Mean SD Min Max Fear of faiure Age Women Skis Some secondary Secondary Post secondary Graduate Cose before Know someone Opportunity Tabe 2. Data summary: Argentina Variabe Obs Mean SD Min Max Fear of faiure Age Women Skis Some secondary Secondary Post secondary Graduate Cose before Know someone Opportunity Tabe 3. Data summary: Brazi Variabe Obs Mean SD Min Max Fear of faiure Age Women Skis Some secondary Secondary Post secondary Graduate Cose before Know someone Opportunity Tabe 4. Data summary: Chie Variabe Obs Mean SD Min Max Fear of faiure Age Women Skis Some secondary Secondary Post secondary Graduate Cose before Know someone Opportunity Tabe 5. Data summary: Mexico Variabe Obs Mean SD Min Max Fear of faiure Age Women Skis Some secondary Secondary Post secondary 1884 Graduate Cose before Know someone Opportunity IV. Resuts To preview the estimation of equation (1), we obtained the pair-wise correation among the variabes as shown in Tabe 7. We observed that in most cases, the correation is significanty different from zero at conventiona eves and none is sufficienty high to make coinearity a probem.

6 Risk attitude and entrepreneurship 577 Downoaded by [Universidad de Chie] at 11:35 09 October 2014 Tabe 6. Data summary: Venezuea Variabe Obs Mean SD Min Max Fear of Faiure Age Women Skis Some secondary Secondary Post secondary Graduate Cose before Know someone Opportunity Tabe 8 dispays the resuts of the coefficient estimates and the margina effects for four different specifications. As can be seen, in its four specifications, the mode is capabe of correcty cassifying about 67% of the outcomes. The second and third coumns show the resut for the baseine mode. We can observe that age affects fear of faiing quadraticay. When young, a person experiences an increase in the probabiity of feeing a fear of faiing for each year that his age increases. However, after about 40 years of age, the probabiity of feeing a fear of faiing goes down as age increases. The same effect Tabe 7. Correation matrix appears in each one of the countries anaysed. Indeed, whie in Argentina the probabiity of feeing a fear of faiing goes down after 30 years of age, in Chie it decreases after 53 years of age and in Brazi after 51 years of age. In Mexico and Venezuea, the probabiity of feeing a fear of faiing is not affected by age as the coefficients turn out to be nonsignificant. 1 A possibe expanation for this resut is that age may actuay be capturing a weath effect. Indeed, we do not have a proxy for weath in the APS, and if weath is positivey correated with age, then higher eves of weath may reduce the person s dependency on the future fows of the project. In this sense, the individua may experience ower eves of fear of faiing because a faiure woud not mean financia disaster or, aternativey, as Cressy (2000) suggests, because individuas present DARA preferences. The dummy WOMEN is statisticay significant and has a positive sign, which indicates that women fee more fear of faiing than men. In this sense, women woud be more risk averse than men with respect to starting a new venture. Indeed, the probabiity of feeing a fear of faiing increases by about 6.5% (margina effects) if the person is a woman. This shoud be one of the factors expaining why women tend to embrace entrepreneuria activity at ower rates than men. For instance, the GEM report of 2008 shows that for a the countries surveyed that year, the tota entrepreneurship activity is ower for women. Fear F. Age Women Skis Educss Educsd Educps Educge Cose Know Opport Fear F. 1 Age (0.390) Women (0.00)*** (0.006) Skis (0.00)*** (0.084) (0.00)*** Educss (0.00)*** (0.00)*** (0.002) (0.00)*** Educsd (0.982) (0.00)*** (0.251) (0.028) (0.00)*** Educps (0.030) (0.00)*** (0.250) (0.00)*** (0.00)*** (0.00)*** Educge (0.002) (0.00)*** (0.001) (0.00)*** (0.00)*** (0.00)*** (0.00)*** Cose (0.004) (0.014) (0.00)*** (0.00)*** (0.00)*** (0.783) (0.074) (0.00)*** Know (0.112) (0.00)*** (0.00)*** (0.00)*** (0.00)*** (0.00)*** (0.00)*** (0.00)*** (0.00)*** Opport (0.017) (0.00)*** (0.00)*** (0.00)*** (0.013) (0.405) (0.00)*** (0.610) (0.00)*** (0.00)*** Note: Numbers in parentheses are p-vaues for the nu hypothesis that the correation is zero. 1 Resuts for each country are avaiabe upon request.

7 578 J. P. Sepúveda and C. A. Bonia Tabe 8. Coefficient and margina effects estimates Variabe Mode 1 Margina effect Mode 2 Margina effect Mode 3 Margina effect Mode 4 Margina effect Downoaded by [Universidad de Chie] at 11:35 09 October 2014 Age *** *** *** *** *** *** *** *** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Age *** *** *** *** *** *** *** *** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Women *** *** *** *** *** *** *** *** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Cose *** *** *** *** *** *** *** *** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Suski *** *** *** *** *** *** *** *** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Knowent * * * * (0.168) (0.168) (0.218) (0.218) (0.088) (0.088) (0.097) (0.097) Opport (0.824) (0.824) (0.390) (0.390) (0.756) (0.756) (0.555) (0.555) Educsd (0.580) (0.580) (0.102) (0.102) (0.191) (0.188) (0.146) (0.143) Educps * ** *** *** *** *** *** *** (0.053) (0.053) (0.000) (0.000) (0.002) (0.001) (0.002) (0.001) Educge *** *** *** *** *** *** *** *** (0.005) (0.005) (0.000) (0.000) (0.002) (0.002) (0.002) (0.001) Argentina *** *** *** *** *** *** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Brazi *** *** *** *** *** *** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Chie *** *** *** *** *** *** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Venezuea *** *** *** *** *** *** (0.001) (0.001) (0.000) (0.000) (0.000) (0.000) Tea05opp *** *** *** *** (0.000) (0.000) (0.000) (0.000) Cose Argentina *** *** (0.000) (0.000) Cose Brazi *** *** (0.004) (0.000) Cose Chie *** *** Cose Venezuea (0.008) (0.002) *** *** (0.002) (0.000) Constant *** *** *** *** (0.000) (0.000) (0.000) (0.000) Observations Pseudo R C.C. proportion 67.23% 67.23% 67.18% 67.53% LR Chi Log Likeihood Notes: p-vaues in parentheses, ***p < 0.01, **p < 0.05, *p < 0.1. C.C.: correcty cassified. There is aso evidence that women tend to be more risk averse than men when confronting financia decisions (See Jianakopos and Bernasek, 1998) and consumer decisions ike smoking and wearing seat bets (Hersch, 1996). The variabe CLOSE is statisticay significant at conventiona eves and shows that peope who have cosed down a business in the ast tweve months have a higher eve of a fear of faiing. Coumn three indicates that having the experience of shutting down a business in the recent past increases the probabiity of feeing a fear of faiure by 7.7%. Having a previous faiure has both costs and benefits. On the one hand, a person who recenty shut down a business coud be more prone to feeing fear of a new faiure because of the materia osses he has just endured. On the other hand, the experience of cosing down a business may increase his stock of experience and thus decrease fear of faiure in the future. Shepherd (2003) shows how business faiure can become a positive

8 Risk attitude and entrepreneurship 579 Downoaded by [Universidad de Chie] at 11:35 09 October 2014 experience for future endeavours. Sitkin (1992) indicates that faiure is more important than success as a source of earning. The positive sign in this case woud therefore indicate that in these Latin American countries, the costs of shutting down a business outweigh the potentia benefits from that experience. A possibe expanation for this resut may be a stigma effect on faied entrepreneurs in these economies. The stigma originates in the fact that in Latin America, there is itte trust in institutions ike the judicia system or the government 2 (see, for exampe, The Word Justice Project) and, therefore, there is a socia presumption of fraud by the faied entrepreneur that makes it difficut for the entrepreneur to make a comeback, principay since in these countries, there are no we-deveoped chapters on business reorganization in the bankruptcy aws and the rue of aw is weak. There is aso a perception of a ow eve of enforcement of the aw. Therefore, the margina expected cost of frauduent behaviour may be ower than the margina expected benefit. The variabe SUSKILL is aso statisticay significant at conventiona eves and ceary shows that peope who beieve that they have the skis necessary to start a new business are ess ikey to experience a fear of faiing. In quantitative terms, this is the most important factor affecting the probabiity of experiencing the fear of faiing. The margina effect indicates that if an individua beieves that he has the knowedge and ski needed to start a new business, then the probabiity of feeing a fear of faiing decreases by about 12%. The coefficients in the variabes KNOWENT and OPPORT show the expected sign, but they are not statisticay significant. Thus, knowing an entrepreneur and beieving that there is an economic opportunity on the market does not seem to affect the probabiity of feeing a fear of faiing. The effect of education on the fear of faiing is studied through the variabes EDUCSD, EDUCPS and EDUCGE, as described above. The estimated resuts indicate that having post-secondary and graduate education decreases the fear of faiing. Indeed, a person who attains these eves of education decreases his probabiity of feeing the fear of faiing by about 4% in both cases. So, the cumuative effect of education woud decrease the probabiity of feeing the fear of faiing by about 8%. In the four and fifth coumns we show the resuts of estimating mode 2, which incorporates dummies for each country in addition to the aforementioned variabes. We use Mexico as the base. The estimated coefficients show that peope in Argentina, Brazi, Chie and Venezuea begin at higher eves of a fear of faiing than peope in Mexico. For instance, whie peope in Venezuea have an 8% higher probabiity of feeing the fear of faiing than peope in Mexico, peope in Chie have a 15% higher probabiity of feeing the fear of faiing than peope in Mexico. Mode 3 (sixth and seventh coumns) adds the variabe TEA05OPP, which is incuded to contro for those peope who report opportunity as their major motive for embarking on an entrepreneuria endeavour. Indeed, the probabiity of feeing the fear of faiing goes down by 15% when the person undertakes an entrepreneurship in the search for an opportunity, that is, the person interviewed is not doing it because of necessity. Mode 4 (ast two coumns) expand the mode by incuding the interaction between the variabe CLOSE and each one of the countries, except Mexico. In a cases, the coefficient and the margina effects are negative. This aows us to observe the net effect, by country, that having shut down a business in the recent past (12 months) has on the probabiity of having a fear of faiure today. For instance, whie in Argentina having cosed a business in the past 12 months increases the probabiity of feeing the fear of faiure by 3.26% ( ), in Chie, the same experience increases the probabiity by 11.7% ( ). We aso estimated the mode with more interactions, but none of them turned out to be significant. Hence, the predictive capacity of the mode was not improved. Figures 1 3 show the margina effect of the CLOSE, WOMEN and SUSKILL variabes for different ages whie evauating the other independent variabes at their mean. As can be seen, in a cases the effect is quadratic, driven by age. V. Concusions We have studied the factors that, at the individua eve, affect the attitude towards the risk of starting a new business. Our focus has been the main Latin American dy dx Age Fig. 1. Margina effect of CLOSE on the probabiity of feeing fear of faiing for different ages 2 The exception to this is Chie.

9 580 J. P. Sepúveda and C. A. Bonia dy dx Age Peope beieving that they have the skis to deveop a new business is the most important factor in reducing the fear of faiing and this provides an important insight for poicy makers about entrepreneuria and business education at a eves. Knowing an entrepreneur and beieving that a market opportunity exists do not affect the fear of faiing and this resut supports the Larrouet and Couyooumdjian (2009) expanation of the Latin American paradox that contends that much of the Latin American entrepreneurship is a necessity entrepreneurship instead of an opportunity entrepreneurship. Finay, more years of education decrease the probabiity of feeing the fear of faiing. Downoaded by [Universidad de Chie] at 11:35 09 October 2014 Fig. 2. Margina effect of WOMAN on the probabiity of feeing fear of faiing for different ages.09.1 dy dx Age Fig. 3. Margina effect of SUSKILL on the probabiity of feeing fear of faiing for different ages countries because they face simiar economic and poitica conditions. In addition, for these countries the difference between choosing a good or bad entrepreneuria and production poicy may account, in the future, for the difference between achieving deveopment and maintaining poor economic conditions. Our findings suggest that age affects the fear of faiing (our proxy for risk aversion) quadraticay. Before the age of 40, age increases the probabiity of feeing a fear of faiing, but after 40, every additiona year decreases that probabiity. Being a woman aso increases the probabiity of feeing a fear of faiing, something that is consistent with previous iterature about gender and risk aversion. Having shut down a business recenty increases the fear of faiing. This resut contradicts the entrepreneurship iterature appied to deveoped countries, and we hypothesized that the ow eve of trust and the weaknesses in institutions in Latin America generate a stigma effect that heps to expain this resut. 80 References Acs, J. (2006) How is entrepreneurship good for economic growth, Innovations: Technoogy, Governance, Gobaization, 1, Acs, Z. and Szerb, L. (2007) Entrepreneurship, economic growth and pubic poicy, Sma Business Economics, 28, Acs, Z. J. (1996) Sma firms and economic growth, in Sma Business in the Modern Economy, De Vries Lectures in Economics, Admiraa, P. H. (Ed.), Backwe Pubishers, Oxford, pp Acs, Z. J. and Varga, A. (2005) Entrepreneurship, aggomeration and technoogica change, Sma Business Economics, 24, Ardagna, S. and Lusardi, A. (2008) Expaining internationa differences in entrepreneurship: the roe of individua characteristics and reguatory constraints, NBER Working Paper No , May, Nationa Bureau of Economic Research, Boston, MA. Baumo, W. (2002) The Free Market Innovation Machine: Anayzing the Growth Mirace of Capitaism, Princeton University Press, Princeton NJ. Banco, L. and Grier, R. (2009) Long ive democracy: the determinants of poitica instabiity in Latin America, Journa of Deveopment Studies, 45, Cramer, J., Hartog, J., Jonker, N. et a. (2002) Low risk aversion encourages the choice for entrepreneurship: an empirica test of a truism, Journa of Economic Behavior & Organization, 48, Cressy, R. (2000) Credit rationing or entrepreneuria aversion? An aternative expanation for the Evans and Jovanovic finding, Economic Letters, 66, Davidsson, P. (2006) Nascent entrepreneurship: empirica studies and deveopments., Foundations and Trends in Entrepreneurship Research, 2, Davidsson, P. and Honig, B. (2003) The roe of socia and human capita among nascent entrepreneurs, Journa of Business Venturing, 18, Demar, F. and Davidsson, P. (2000) Where do they come from? Prevaence and characteristics of nascent entrepreneurs, Entrepreneurship and Regiona Deveopment, 12, Evans, D. S. and Jovanovic, B. (1989) An estimated mode of entrepreneuria choice under iquidity constraints, Journa of Poitica Economy, 97, Friedman, M. (1953) Choice, chance and the persona distribution of income, Journa of Poitica Economy, 61,

10 Risk attitude and entrepreneurship 581 Downoaded by [Universidad de Chie] at 11:35 09 October 2014 Giannetti, M. and Simonov, A. (2009) Socia interactions and entrepreneuria activity, Journa of Economics & Management Strategy, 18, Grio, I. and Thurik, R. (2008) Determinants of entrepreneuria engagement eves in Europe and the US, Industria and Corporate Change, 17, Hersch, J. (1996) Smoking, seat bets, and other risky consumer decisions: differences by gender and race, Manageria and Decision Economics, 17, Hotz-Eakin, D., Joufaian, D. and Rosen, H. S. (1994) Sticking it out: entrepreneuria surviva and iquidity constraints, Journa of Poitica Economy, 102, Jianakopos, N. and Bernasek, A. (1998) Are women more risk averse?, Economic Inquiry, 36, Johnson, S., McMian, J. and Woodruff, C. (2002) Property rights and finance, American Economic Review, 92, Kan, K. and Tsai, W. (2006) Entrepreneurship and risk aversion, Sma Business Economics, 26, Kanbur, S. M. (1979) Of the risk taking and the persona distribution of income, Journa of Poitica Economy, 87, Kihstrom, R. E. and Laffont, J. J. (1979) A genera equiibrium entrepreneuria theory of firm formation based on risk aversion, Journa of Poitica Economy, 87, Kirzner, I. M. (1973) Competition and Entrepreneurship, The University of Chicago Press, Chicago, IL. Kirzner, I. M. (1979) Perception, Opportunity, and the Profit: Studies in the Theory of Entrepreneurship, University of Chicago Press, Chicago IL. Knight, F. (1971) Risk Uncertainty and Profit, University of Chicago Press, Chicago IL. First pubished in Larrouet, C. and Couyooumdjian, J. (2009) Entrepreneurship and growth: a Latin American paradox, The Independent Review, 14, Lerner, J. and Mamendier, U. (2007) With a itte hep from my (random) friends: success and faiure in post-business schoo entrepreneurship, Unpubished Working Paper, Harvard University, Boston, MA. Newman, A. (2007) Risk-bearing and entrepreneurship, Journa of Economic Theory, 137, Parker, S. (1997) The effects of risk on sef-empoyment, Sma Business Economics, 9, Parker, S. (2007) Which firms do the entrepreneurs come from?, Economics Buetin, 10, 1 9. Parker, S. C. (2004) The Economics of Sef-Empoyment and Entrepreneurship, Cambridge University Press, Cambridge. Schumpeter, J. (1934) The Theory of Economic Deveopment: An Inquiry Into Profits, Credit, Interest and the Business Cyce, 2nd edn, Harvard University Press, Cambridge MA. First pubished in Sepúveda, J. and Bonia, C. (2011) The attitude toward the risk of entrepreneuria activity: evidence from Chie, Academia, 46, Shepherd, A. (2003) Learning from business faiure: propositions of grief recovery for the sef-empoyed, The Academy of Management Review, 28, Sitkin, S. (1992) Learning through faiure: the strategy of sma osses, in Research in Organizationa Behavior, vo. 14, Staw, B. M. and Cummings, L. L. (Eds), JAI Press, Greenwich, CT, pp Van Praag, C. and Cramer, J. (2001) The roots of entrepreneurship and abour demand: individua abiity and ow risk aversion, Economica, 68, Van Ste, A., Carree, M. and Thurik, R. (2005) The effect of entrepreneuria activity on nationa economic growth, Sma Business Economics, 24, Wennekers, A. R. M., Van Ste, A. J., Thurik, A. R. et a. (2005) Nascent entrepreneurship and the eve of economic deveopment, Sma Business Economics, 24, Wong, P. K., Ho, Y. P. and Autio, E. (2005) Entrepreneurship, innovation and economic growth: evidence from GEM data, Sma Business Economics, 24,