The Attractiveness of Entrepreneurship for Females and Males in a Developing Arab Muslim Country ; Entrepreneurial Intentions in Tunisia

This article examines the entrepreneurial intentions of young well educated people in Tunisia. Tunisia is a Muslim country which is “catching up”. Hence, understanding the attitudes and entrepreneurial intentions of females, and contrasting these with males, will provide an account of cultural attitudes towards female enterprise. Our survey of 332 students showed that male and female intentions were very positive, but varied by gender. We found that cultural gender expectations continued to play a major role, but that this was most obvious in how female enterprise could be operated, rather than in the attractiveness of entrepreneurship itself.


Introduction
In some Arabic Muslim countries, culture and attitudes may prevent women from playing a full economic role.Yet, it is very well recognized that entrepreneurship plays a crucial role in enhancing economic development (Carree and Thurik 2003), so that the part women play in entrepreneurship may be critical in achieving development.Tunisia, an emerging Muslim nation located in North Africa, is a modern Arab country striving to catch up with more developed nations.We argue that establishing the entrepreneurial intentions of young well educated Tunisian women may be useful in helping us to understand future developments.The entrepreneurial intentions of well educated young people of both genders are important because they are the most likely group to influence future events and the most likely to be future leaders.Nonetheless, male entrepreneurship is likely to represent an economic "base" line, whilst younger female entrepreneurial intentions are more likely to represent social change.Accordingly, capturing the intentions of young well educated females will be useful, not only in terms of what themselves may do, but also in seeing how the culture is changing and, indeed their role in influencing future cultural attitudes.As Idris (2008) noticed, studying women's participation will have strategic implications on a firm operating in a gender-sensitive society.
Across the world, awareness of female entrepreneurship has increased and is characterised by the increasing numbers of women setting up in business (Buttner 1993).Female entrepreneurs are thus recognised as making an important contribution to the economy, both as a source of employment and of potential growth (Global Entrepreneurship Monitor: GEM, 2007).Nonetheless, entrepreneurship is frequently described as a form of masculinity (Bird and Brush 2002;Bruni and al 2004) and women remain under-represented in the entrepreneurial population.So that, a significant gender gap exists in rates of entrepreneurial activity across genders (GEM, 2006).Moreover, there are considerable differences between male and female entrepreneurship; motivations, financing strategies, risk perception and even success criteria are not the same (Galloway and al 2002;Buttner and Moore 1997).There are also considerable variations between countries; the GEM study (2007) notes that, with the exception of Brazil, Peru, Japan and Thailand, men are more likely to engage in entrepreneurial activity.For example, the activity rate for US females is 25 percent, but in Thailand 95 percent.Yet, Turkey, a largely Muslim country, has a tiny participation rate of 2.41 percent.It can be argued that the distinctive masculine dominated culture of such countries is historically and religiously rooted and thus strongly influences female participation.Indeed, Roomi and Parrot (2008) argue that many of the problems and challenges faced by Pakistani women entrepreneurs originate from the structurally enforced status of women within an Islamic society.When studying Omani entrepreneurial women's barriers to success, McElwee and Al-Riyami (2003) argue that Omani female entrepreneurship has to be seen in the light of their role in Arab Muslim societies where gender differences are enshrined in Shari'a (the Islamic Law).
Nevertheless, there are considerable changes in the role of women in Muslim countries; this is particularly true in business and in entrepreneurship.The number of females working in business has increased but the evidence to date suggests a large gender gap in venture creation and in business participation more generally in Muslim countries.Female workforce participation in Tunisia is some 27 percent but is characterised by a very limited presence in senior roles.As an example, women have only gained 12 percent of parliamentary seats in the last two elections.There is unfortunately no specific data available about entrepreneurial participation rates.However, even these indicative low female participation rates mask the differences in employment levels by income groups.Although female employment participation levels average 27 percent, the level for low and middle income groups is 42 percent.This suggests that the better off female is much less likely to be in employment and probably much less likely to have their own business.This correspondence between income and labour force participation is likely to be mirrored in entrepreneurship.The GEM study (2006) notes that, internationally but particularly in less well developed countries; a high proportion of female entrepreneurship is necessity enterprise.In such cases women may have been pushed into enterprise to escape unemployment rather than being drawn to entrepreneurship as opportunity (Orhan and Scott 2001).For example, Turkey has the lowest rate of opportunity to necessity early stage entrepreneurship (0.32 percent).This is particularly significant for our study because opportunity entrepreneurship is much more likely to lead to economic growth.Additionally, opportunity enterprises often reflect the desire for independence and individual freedom (Cromie 1987;;Muir 1999).Women are particularly constrained in these terms by the dual role, or double burden of cultural expectations about homemaking.But the flexibility offered by being your own boss is seen by some as a solution to balance work and family responsibilities (Goffee and Scase 1983;Scott 1986;Moult and Anderson 2004;Buttner 1993;Buttner and Moore 1997;Maysami and Goby 1999).Yet, even with these constraints, starting a business represents an expression of self accomplishment and the desirability of controlling one's own destiny.So an appreciation of well educated young womens' attitudes about becoming their own boss should present a measure of cultural change and impact.
One aspect of cultural shift is education.In comparison with other MENA (Middle East North Africa) countries, Tunisian women are now better educated.The proportion of women who have received a secondary or higher level of education rose from 22 percent in 1994 to 35 percent in 2004.We argue that these well educated women are most likely to be the feed stock of new opportunity driven enterprise.In this study we are interested in female entrepreneurial intentions.Intentions are considered to be a more advanced stage in the behavioural process.Intention as presented in the theory of planned behaviour (Ajzen 1991) and is explained by the perception of desirability (which includes attitude towards entrepreneurship and perception of social norms) and perception of feasibility towards entrepreneurship.Consequently, we will shed some light on differences between male and female in terms of the feasibility, desirability and social norms regarding intention to start a business.We also add supplementary hypothesis about the perception of informational and financial resources availability, thus gauging how entrepreneurial possibilities are perceived.Thus, we decided to explore female intentions in Tunisia.Analytically, we use chi square tests to capture differences/similarities between the two genders regarding to the intention constituents and then conduct a discriminant analysis, which allows us to pursue the same object but simultaneously considering all relevant variables.The paper is structured as follows: section two briefly explains our hypotheses; section three presents our methodology and section four presents the empirical result, whilst section five concludes.

Hypotheses
According to the Ajzen's theory of planned behaviour (1991), attitude toward behaviour refers to the degree to which the individual has a favourable or unfavourable evaluation or appraisal of the behaviour in question.Accordingly, the following hypothesis is considered: Hypothesis 1: There is a significant difference between genders in the attitude towards starting a business.
According to Shapero (1982), perceived feasibility is the degree to which one feels personally capable of starting a business.Krueger and al (2000) note that feasibility perceptions drive career-related choices, including self-employment as an entrepreneur.Thus: Hypothesis 2: There is a significant difference between genders in the perception of feasibility towards starting a business.
Subjective norm refers to the perceived social pressure to perform or not to perform the considered behaviour.Thus: Hypothesis 3: There is a significant difference between genders according to perception of social norm.
Hypothesis 3a: There is a significant difference between genders according to the perception of the opinion of closest family and the opinion of people who are important in the respondent's lives toward starting their own business.
Family situation is assumed to impact career decision.So, individuals with families are hypothesised to be more tied to their current community are less flexible and they evaluate job alternatives more critically (Hooft 2005).Thus: Hypothesis 3b: There is a significant difference between genders in the perceptions of the impact of future family commitments on the decision to start a business.
Moreover in this paper we expect that attitude toward female entrepreneurship differs relatively to gender.Thus: Hypothesis 3c: There is a significant difference between genders according to attitude towards female entrepreneurship.
The extant literature suggests that the interpretations of the entrepreneur are of importance when resource availability is considered (Bruno and Tyebjee 1982;Krueger and Brazeal 1994) and that the perceived resource availability is more important than actual resource availability.Moreover, perceived resource availability may influences entrepreneurial intentions (Grundstèn 2004).Thus: Hypothesis 4: There is a significant difference between genders relatively to their perception of resources availability.
Hypothesis 4a: There is a significant difference between genders in the perception of availability of financial resources.
Hypothesis 4b: There is a significant difference between genders in the perception of availability of informational resources.

Sample, data collection and procedure
This study was conducted within 9 faculties in Sousse University (Note1).Our sample consisted of 332 final year undergraduate students from different disciplines.It is representative of the distribution of gender and speciality at Sousse University.To ensure the clarity of wording, we first piloted the questionnaire with 14 final year undergraduate students and made some subsequent improvements to the questionnaire.We launched the survey in March 2007 and completed collecting data in April 2007.The questionnaire was distributed to 350 final year undergraduate students in the classroom and was completed in our presence.From the 350 questionnaires received, 18 containing incomplete responses were removed, reducing the sample to 332 students.Accordingly, our approach to data collection ensured a high response rate.To verify if there are significant differences between men and women concerning some entrepreneurial aspects we have performed the chi square tests using SPSS 11.0.After that, a discriminant analysis was performed using SPAD 5.0.

3.1Measures
The measure of attractiveness toward starting a business was adopted from Krueger, Reilly, and Carsrud (2000).Hence, we have asked students the following: Is starting your own business an attractive idea to you? (1= not at all; 2=neutral; 3= very attractive).
The measure of perceived feasibility towards starting a business was adopted from Krueger, Reilly, and Carsrud (2000).So, the following question was considered: How practical is it for you to start your own business?(1= not very practical; 2=neutral; 3= very practical).
To measure subjective norm we introduced some changes to Kolvereid's (1996a) question about this variable; we have broken up the item proposed by Kolvereid (1996a) "I believe that my closest family thinks that I should not (I should) pursue a career as self employed" into different parts.This is because usually there are differences in points of views amongst Tunisian family members.Accordingly we asked respondents to answer the five following questions: (1).I believe that my father thinks that I should (I should not) pursue a career as self employed.
(2).I believe that my mother thinks that I should (I should not) pursue a career as self employed.
(3).I believe that my brother (and/or sister) thinks that I should (I should not) pursue a career as self employed.
(4).I believe that my spouse thinks that I should (I should not) pursue a career as self employed.
(5).I believe that persons who are important to me think that I should (I should not) pursue a career as self employed.
All responses were given on a three point Likert scale (1= I should not; 2= neutral; 3= I should).
Moreover, in order to asses motivation to comply, respondents were asked along three point Likert scale (1= not important; 2= neutral; 3= very important), about the extent to which they cared about the opinions of the different persons cited above when they decide to start their own business.
Unlike Kolvereid (1994 a) who summed all items described above to measure social norm, we decided to consider each item separately in order to pick out where differences exist between females and males relatively to the each element of the perceived social norm.We also added two other variables to the measure of social norm.So we asked students questions about their perception of the impact of future family commitments on their decision to start a business and their attitude toward female entrepreneurship.Therefore the following questions were considered: (1).Future family commitments may make it difficult for me to start my own business (1= unlikely; 2= neural; 3= very likely).
( 2).Indicate to what extent you are approve of the creation of a business by woman (1= not at all; 2= neutral; 3= agree).Furthermore, using a three point Likert scale, we asked students about their perception of the availability of financial and informational resources, and their assessment of the incentives available in the institutional environment for entrepreneurship in Tunisia.Hence, institutions that economic agents (including entrepreneurs) operate in -political, legal and cultural-directly influence their activity and hence economic development (Baumol 1990;Olson 1996).Consequently, incentives matter and the institutional environment in which the economic agent acts, serves as an incentive structure which guides and influences action, (Boettke and Coyne 2003).

Descriptive statistics
Respondents consisted of 38.6 percent male and 61.4 percent female undergraduate students.There are more female than male students in higher education in Tunisia.For example in 2005/2006, woman represented 58.1 percent of the total in higher education.The medium age of the sample is 23.36 years old and 94.6 percent are single.The mainstream of the students came from urban areas (78.6 percent).In terms of work experience, 39.5 percent had work experience.Interestingly, more than the half (61.4 percent) have attended an entrepreneurial course.
The great majority of Tunisian students (more than 70 percent) perceive a favourable attitude of all members of their family and of other important persons to them toward starting their own business.The opinion of family and other important people seem playing an important role.Indeed, the overwhelming majority of our respondents (more than 70 percent) consider the opinion of their father, mother, spouse, brother and/or sister as important.
Additionally, more than 70 percent of respondents believe that the incentives and business environment in Tunisia (support structure for start-up, formalities involved in setting up a business, tax and financial incentives, etc) encourage entrepreneurship.This result indicates a perceived success for the government initiatives to promote new business formation.Moreover more than 60 percent of respondents believed that the informational resources at their disposal are favourable for starting a business.Unsurprisingly, a considerable proportion of our respondents appear to be financially constrained.Only 39.2 percent of our respondents reported favourably on the financial resources at their disposal for the creation of their own business.

Chi square tests results
The key results obtained from the chi square tests are summarized in Table 1 presented below.The chi square test of independence indicates that there is, at a significance level of 5 percent, a relationship between genders and perception of desirability towards starting a business.Hence, hypothesis 1 is supported.Similarly, hypothesis 2 is supported.Therefore, we have found a significant relationship between genders and the perception of feasibility towards starting a business.At the same level of significance (5 percent), the hypothesis 3a was rejected.Thus, and rather surprisingly, we found no significant relationship between gender and the perception of the opinion of the closest family and other important people who are important to the respondents towards starting a business.However, hypothesis 3b was supported.Accordingly, there is a significant relationship between gender and the attitude towards the impact of future family commitments on the decision to launch a business.Although very interested in starting a business, females are still significantly more worried than males about the impact of future family commitments on the realization of their intention to start a business.Furthermore, there is also a significant relationship, at a 5 percent significance level, between gender and attitude towards female entrepreneurship.Consequently, hypothesis 3c was supported.Men are significantly less taken with the idea of female entrepreneurship.This result suggests that men and women approach entrepreneurship differently and that entrepreneurship is "welcomed" differently according to gender.As a result, entrepreneurship, considered as a mean of economic and social promotion, could be transformed into a vehicle of gender discrimination and so social regression.Hypotheses H4 (a) and H4 (b) were rejected when considering, genders and perception of resources availability.Therefore, there is no statistically significant difference neither between gender and perception of the availability of financial resources nor between gender and perception of the availability of informational resources.

Discriminant analysis results
After the Chi squared analysis, we use discriminant analysis, a multivariate technique which allows the study of the differences between two (or more) groups with respect to several variables simultaneously.As it is normal practice, we only consider the variables from the Chi squared results that imply significant differences between genders.The goal of discriminant analysis is to optimally separate groups through constructed discriminating axes (calculated as a linear combination of the initial variables), in such a way that optimal separation of the predefined groups is attained.The linear discriminant equation is constructed such that the two groups are separated as much as possible.Thus, discriminant analysis permits us to examine group differences on considered covariates and to predict group membership from the covariates.
Regarding to the object of our research, we exploit the geometrical principals of this method, by searching axes that best separate the variables to find the significant predicators that best explain the distance between genders in the multivariate space.In its classical form, the method uses continuous variable measurements observed in the two gender groups to highlight aspects that distinguish the groups and to use these measurements to classify new individuals.Thus the explanatory variables have to be independent and normally distributed.Nevertheless, in our study, all the considered variables are qualitative ones.The "Disqual" approaches in Spad software can resolve this problem, in that it implies the construction of factorial axes with the MCA analysis.These axes are by construction quantitative variables, unimodales, and symmetric so they satisfy the conditions of discriminant analysis.Then the discriminant analysis is performed on these axes.
Mahalanobis distance and the Hotelling test are used to compare the means of the two samples.More specifically, they are utilised to test the null hypothesis where the two centroids (means in a multivariate space) of the two gender groups are identical.Thus they are the privileged tools for the appreciation of the discrimination quality.Our results show that the probabilities of the Fisher tests tend to 0, so the Mahalanobis distance is significantly different from zero.The Hotelling test also allows rejecting at the significant level of 5 percent the equality of the two centroids.Consequently there is to a significant difference between the two centroides of the two gender groups.Hence the two groups are significantly separated.(See table 2 for the main results of the discriminant analysis).

[Insert table 2]
The quality of the representation is depicted through the percentage of well classified individuals.190 female in our case are correctly classified according to the obtained discriminant function and only 14 were incorrectly classified.In the total 247 individual were correctly classified with a rate of 74.4 percent of success.The R2 measures the inter-class variance related to the total variance, the R2 in our case is equal to 23.28 percent.The coefficients of the discriminate function reflect the impact of covariates on the gender.
We turn now to consider the most discriminant variables.According to the results, the variable "attitude towards female entrepreneurship" is the most discriminant variable between genders.Interestingly, the desirability, feasibility and the attitude towards family commitment plays only a minor role in the explanation of intention differences between genders.

Discussion and Concluding Remarks
Although our results show a very promising attitude to entrepreneurship by the young well educated females, it is also very clear from that male and female attitudes to entrepreneurship differ significantly.Thus it seems that in a general sense, entrepreneurship is an attractive option for all young well educated people, there remains considerable cultural constraints for females.Interestingly, we note that feasibility, capacity nor even the point of view of the closest relatives are not responsible for any gender discrimination.It seems that the attitudes of educated young people themselves clearly do not approve of womens' venture creation.Thus, in spite of the fact that they are well educated, the perceived social norms seem to confine female to a more traditional path.Therefore, any policy targeting the encouragement of female entrepreneurship would need to work to change the image female entrepreneurship.In other respects policy seems to be working.Culture, which is notoriously slow to change, seems to be at least beginning to adapt to make female entrepreneurship a possibility.We had expected the role of close relatives, especially fathers to act as a constraint.We anticipated that they would represent more old fashioned attitudes towards discouraging full female participation but that does not seem to be the case.Indeed the contemporaries of our females, their male counterparts, seemed to be more negative towards the idea of female enterprises.